Add files using upload-large-folder tool

processed/00002_W2122361802/document.md

@@ -0,0 +1,559 @@
+# Navigating the Patent Thicket: Cross Licenses, Patent Pools, and Standard­Setting
+
+Carl Shapiro University of California at Berkeley
+
+March 2001
+
+## Abstract
+
+In several key industries, including semiconductors, biotechnology, computer software, and the Internet, our patent system is creating a patent thicket: an overlapping set of patent rights requiring that those seeking to commercialize new technology obtain licenses from multiple patentees. The patent thicket is especially thorny when combined with the risk of hold-up, namely the danger that new products will inadvertently infringe on patents issued after these products were designed. The need to navigate the patent thicket and hold-up is especially pronounced in industries such as telecommunications and computing in which formal standard-setting is a core part of bringing new technologies to market. Cross-licenses and patent pools are two natural and effective methods used by market participants to cut through the patent thicket, but each involves some transaction costs. Antitrust law and enforcement, with its historical hostility to cooperation among horizontal rivals, can easily add to these transaction costs. Yet a few relatively simple principles, such as the desirability package licensing for complementary patents but not for substitute patents, can go a long way towards insuring that antitrust will help solve the problems caused by the patent thicket and by hold-up rather than exacerbating them.
+
+Forthcoming, Innovation Policy and the Economy, Volume I , Adam Jaffe, Joshua Lerner, and Scott Stern, eds., MIT Press, 2001. This paper is available at http://haas.berkeley.edu/~shapiro/thicket.pdf . Comments are welcomed; please direct comments to shapiro@haas.berkeley.edu .
+
+---
+
+## CONTENTS
+
+I. THE PATENT THICKET......1 II. MARKET RESPONSES TO OVERLAPPING PATENTS......4 A. The Economic Theory of Complements........................................................4 B. The Hold-Up Problem........................................................................6 C. Overlapping Patents and Business Strategy in Practice......8 D. Antitrust Limits........................................................................11 III. CROSS LICENSING......12 A. Cross Licenses and Design Freedom........................................12 B. Intel’s Policy of “IP For IP”......13 IV. PATENT POOLS......17 A. Essential Patents vs. Rival Patents......17 B. A Patent Pool Created to Resolve Claims of Blocking Patents......18 V. COOPERATIVE STANDARD SETTING......19 A. The Costs and Benefits of Compatibility and Standards......21 B. Legal Treatment of Cooperative Standard Setting......22 C. Hidden Patents and Hold-Up in Standard Setting......24 VI. SETTLEMENTS OF PATENT DISPUTES......26 VII. CONCLUSIONS......28 TECHNICAL APPENDIX......31
+
+---
+
+## I. The Patent Thicket
+
+Is our patent system slowing down the commercialization of new technologies?
+
+The essence of science is cumulative investigation combined with hypothesis testing. The notion of “cumulative innovation,” each discovery building on many previous findings, is central to the scientific method. Indeed, no respectable scientist would fail to recognize and acknowledge the crucial role played by his or her predecessors in establishing a foundation from which progress could be made. As Sir Isaac Newton put it, each scientist “stands on the shoulders of giants” to reach new heights.
+
+Today, most basic and applied researchers are effectively standing on top of a huge pyramid, not just on one set of shoulders. Of course, a pyramid can rise to far greater heights than could any one person, especially if the foundation is strong and broad. But what happens if, in order to scale the pyramid and place a new block on the top, a researcher must gain the permission of each person who previously placed a block in the pyramid, perhaps paying a royalty or tax to gain such permission? Would this system of intellectual property rights slow down the construction of the pyramid or limit its height?
+
+Clearly, pyramid building, namely research and development (R & D), is taking place at an impressive pace today, so there is no great cause for alarm, especially in the area of basic research where the “ royalty ” is often (but not always) nothing more than a citation. As we move from pure “ R ” to applied “ R ” and ultimately to “ D, ” however, one can fairly ask whether our legal and commercial institutions are in fact properly designed to promote rather than discourage the creation of products and services that draw on many strands of innovation and thus potentially require licenses from multiple patent holders. To complete the analogy, blocking patents play the role of the pyramid's building blocks.
+
+Mixing metaphors, thoughtful observers are increasingly expressing concerns that our patent (and copyright) system is in fact creating a patent thicket, a dense web of overlapping intellectual property rights that a company must hack its way through in order to actually
+
+Shapiro on Patent Thicket, Innovation Policy and the Economy, Page 1 of 32
+
+---
+
+commercialize new technology. With cumulative innovation and multiple blocking patents, stronger patent rights can have the perverse effect of stifling, not encouraging, innovation. $^{1}$
+
+In fact, even while a consensus has emerged that innovation is the main driver of economic growth, we are witnessing somewhat of a backlash against the patent system as it is currently operating. Especially unpopular are patents on “business methods,” such as Priceline.com's patent on “buyer-driven conditional purchase offers” (asserted against Microsoft) or Amazon's patent on a one-click on-line shopping system (asserted against Barnes & Noble). The Patent and Trademark Office does indeed seem to have allowed a number of patents on ideas that would not appear offhand to meet the usual standards for novelty and non-obviousness, such as the patent held by Sightsound.com which reputedly covers “the sale of audio or video recordings in download fashion over the Internet.” Emboldened by a key appeals court decision in 1998 supporting a patent for a business method enabled by computer software, patent applications for computer-related business methods have jumped from about 1000 in 1997 to over 2500 in 1999. In an attempt to call a truce in what could otherwise prove to be a mutually destructive patent battle, Jeff Bezos, the Chairman of Amazon.com, recently suggested that patents on software and Internet business methods be limited to three or five years, rather than the usual twenty years from the date of application. $^2$
+
+But concerns about a patent thicket, and excessively loose standards at the Patent and Trademark Office (PTO), are hardly confined to e-commerce and business-method patents. For example, in the semiconductor industry, companies like IBM, Intel, or Motorola find it all too easy to unintentionally infringe on a patent in designing a microprocessor, potentially exposing themselves to billions of dollars of liability and/or an injunction forcing them to cease production
+
+For example, in 1995 Joseph Stiglitz, then Chairman of the Council of Economic Advisors, stated at the opening of the Federal Trade Commission’s hearings on Competition Policy in the New High-Tech, Global Marketplace, that “some people jump .. to the conclusion that the broader the patent rights are, the better it is for innovation, and that isn’t always correct, because we have an innovation system in which one innovation builds on another. If you get monopoly rights down at the bottom, you may stifle competition that uses those patents later on and so ... the breadth and utilization of patent rights can be used not only to stifle competition, but also have adverse effects in the long run on innovation.” See FTC Staff Report, p. 6.
+
+2 See http://www.amazon.com/exec/obidos/subst/misc/patents.html/103-4266077-5496631.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 2 of 32
+
+---
+
+of key products. $^3$ So-called “ submarine patents, ” which take years if not decades to work their way through the Patent and Trademark Office, are another great source of anxiety, especially for large manufacturing firms. Plus, more and more companies are following the lead of Texas Instruments and engaging in “ patent mining, ” trying to get the most out of their patents by asserting them more aggressively than ever against possible infringing firms, even those who are not rivals. And considerable research shows that companies are increasingly inclined to seek patents, causing an increase in the “ propensity to patent, ” as well as an increase in the practice of “ defensive patenting. ” $^4$
+
+In short, our patent system, while surely a spur to innovation overall, is in danger of imposing an unnecessary drag on innovation by enabling multiple rights owners to “ tax ” new products, processes and even business methods. The vast number of patents currently being issued creates a very real danger that a single product or service will infringe on many patents. Worse yet, many patents cover products or processes already being widely used when the patent issued, making it harder for the companies actually building businesses and manufacturing products to invent around these patents. Add in the fact that a patent holder can seek injunctive relief, i.e., can threaten to shut down the operations of the infringing company, and the possibility for “ hold up ” becomes all too real.
+
+This paper takes as given the flood of patents currently being issued by the PTO, and assumes that these patents are indeed creating a “patent thicket” in the sense that many new products would likely infringe on multiple patents. Remaining agnostic (but suspicious) about whether the PTO is too lax in granting patents (especially software patents), or whether the courts are too generous in upholding patents that are granted, I look at the business arrangements that are being used to cut through the patent thicket.
+
+More specifically, I consider the evolving and growing role of cross licenses and patent pools to solve the “complements problem” that arises when multiple patent holders can potentially block a given product. I discuss specifically the standard-setting process, which increasingly
+
+3 Nearly 5000 patents were granted in the U.S. in a recent single year, 1998, relating to "microprocessors" alone, not to mention semiconductors more broadly.
+
+4 See, for example, Kortum and Lerner (1997), Cohen et. al. (1997), and Hall and Ham (1998).
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 3 of 32
+
+---
+
+involves complex negotiations over patent rights and licensing terms. I also consider other ways in which companies resolve disputes over intellectual property, including acquisitions.
+
+For each business practice, in addition to describing the economics underlying that practice and examples of its use, I consider whether antitrust limits are contributing to the problems caused by the patent system. Unfortunately, antitrust enforcement and antitrust law have a deep-rooted suspicion of cooperative activities involving direct competitors. But such cooperation, in one form or another, may be precisely what is required to navigate the patent thicket. As a result, unless antitrust law and enforcement are quite sensitive to the problems posed by the patent thicket, they can have the perverse effect of slowing down the commercialization of new discoveries and ultimately retarding innovation, precisely the opposite of the intent of both the patent laws and the antitrust laws.
+
+## II. Market Responses to Overlapping Patents
+
+### A. The Economic Theory of Complements
+
+The generic problem inherent in the “ patent thicket ” is well understood as a matter of economic theory, at least in its static version. Consider, for example, a company seeking to manufacture a new graphics chip for use in personal computers or video game consoles. (Substitute a biotech firm using patented tools for genetic engineering, or an e-commerce firm using patented business methods, if you would prefer.) Suppose that the company's preferred design for this chip is likely to infringe on a number of patents; the process manufacturing methods used to actually produce the chip infringe on a number of additional patents. In order the produce the chip as designed, the company needs to obtain licenses from a number, call it $N$ , of separate rights holders.
+
+This situation is precisely the classic “ complements problem ” originally studied by Cournot in 1838. Cournot considered the problem faced by a manufacturer of brass who had to purchase two key inputs, copper and zinc, each controlled by a monopolist. 5 As Cournot demonstrated, the resulting price of brass was higher than would arise if a single firm controlled
+
+5 For a brief description of Cournot's original work on complements, and modern extensions, see Shapiro (1989), p. 339.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 4 of 32
+
+---
+
+trade in both copper and zinc, and sold these inputs to a competitive brass industry (or made the brass itself). Worse yet, the combined profits of the producers were lower as well in the presence of complementary monopolies . So, the sad result of the balkanized rights to copper and zinc was to harm both consumers and producers. $^6$ The same applies today when multiple companies control blocking patents for a particular product, process, or business method.
+
+How can the inefficiency associated with multiple blocking patents be eliminated? One natural and attractive solution is for the copper and zinc suppliers to join forces and offer their inputs for a single, package price to the brass industry. The two monopolist suppliers will find it in their joint interests to offer a package price that is less than these two components sold for when priced separately. The blocking patent version of this principle is the rights holders will find it attractive to create a package license or patent pool, or in some situation to simply engage in cross licensing so they can each produce final products themselves.
+
+The Appendix offers a short, modern, and more general version of Cournot's theory of complements, cast in terms of blocking patents. This basic theory of complements (used in fixed proportions) gives strong support for businesses to adopt, and for competition authorities to welcome, either cross-licensees, package licenses, or patent pools to clear such blocking positions. If two patent holders are the only companies realistically capable of manufacturing products that utilize their intellectual property rights, a royalty-free cross license is ideal from the point of view of competition. But any cross license is superior to a world in which the patents holders fail to cooperate, since neither could proceed with actual production and sale in that world without infringing on the other's patents. Alternatively, if the two patent holders see benefits from enabling many others to make products that utilize their intellectual property rights, a patent pool, under which all the blocking patents are licensed in a coordinated fashion as a package, can be an ideal outcome. The simple theory, which is sketched out in the Appendix,
+
+6 Cournot assumed that the two inputs, copper and zinc, were required in certain fixed proportions for the production of brass. If one input can be substituted for the other, they have properties of substitutes as well as complements , in which case competition between the two input owners can go far to solving the problem posed here. Throughout this paper, I am assuming that the company in question requires rights to practice each of several patents, and that one patent license cannot substitute for another. Clearly, to the extent that a manufacturer, for example, can rely on multiple designs or production processes covered by separate patents with separate owners, the “patent thicket” is far less of a problem. But even in this relatively friendly setting, extra difficulties can still be raised by the “hold-up problem,” discussed below.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 5 of 32
+
+---
+
+suggests that coordinating such licensing can lead to lower royalty rates than would independent pricing (licensing) of the two companies' patents.
+
+In other words, without cross-licenses or patent pools, there is a tendency for products to bear “ multiple patent burdens. ” The buildup of licensing fees can have several unattractive consequences. First, the well-known costs of static monopoly power are magnified: prices are well above marginal costs, causing inefficiently low use of these products. As shown in the Appendix, with $N$ rights holders, equilibrium markups are $N$ times the monopoly level. Of course, this is merely a magnified version of the monopoly “ burden ” resulting from the patent system itself, but it is well to remember Cournot's lesson that the multiple burdens reduce both consumer welfare and the profits of patentees in comparison with a coordinated licensing approach. Second, these burdens may cause certain products not to be produced at all, if that production is subject to economies of scale. Third (this is a dynamic version of the previous point), the prospect of paying such royalties necessarily reduces the return to new product design and development, and thus can easily be a drag on innovation and commercialization of new technologies.
+
+Heller and Eisenberg (1998) discuss the “ complements problem ” in the context of biotechnology patents, making a nice comparison to the classic “ tragedy of the commons. ” The well-known tragedy of the commons refers to the fact that a resource can be overused if it is not protected by property rights; fishing grounds and clean water are standard examples. Heller and Eisenberg point out that quite a different problem arises when there are multiple blocking patents; they label this problem the “ tragedy of the anti-commons. ” The tragedy of the anti-commons arises when there are multiple gatekeepers, each of whom must grant permission before a resource can be used. With such “ excessive ” property rights, the resource is likely to be underused. In the case of patents, innovation is stifled.
+
+## B. The Hold-Up Problem
+
+As noted above, the “ complements problem ” is at its worst when the downstream firms using the various inputs truly require each input to make their products. In the patent context, if a manufacturer finds it relatively easy to design around a given patent, the royalties that the patentee can assert are necessarily limited. So, unless the patent in question is quite broad, one
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 6 of 32
+
+---
+
+might think that any burden on the manufacturer would be modest, and arguably the very return we wish to provide to the patentee as a reward for innovation.
+
+Unfortunately, this rather romantic view of patents is less and less applicable in our economy, for three reasons. First, even a modest “tax” is counterproductive if the patent was improperly granted, i.e., if the patentee did not truly made a new and useful discovery, or if the patent as granted was too broad, covering some prior art as well as something truly new. Second, the cumulative effect of many small “taxes” can become quite large; there are sound reasons to believe that the (static) deadweight loss associated with these royalties is increasing and convex in the “tax” rate, at least over some range of royalties. The danger of paying royalties to multiple patent owners is hardly a theoretical curiosity in industries such as semiconductors in which many thousands of patents are issued each year and manufacturers can potentially infringe on hundreds of patents with a single product.
+
+Third, and most important, is timing. Suppose that our representative manufacturer could, with ease, invent around a given patent, if that manufacturer were aware of the patent and afforded sufficient lead time. Clearly, in this case the patented technology contributes little if anything to the final product, and any “reasonable” royalty would be modest at best. But, oh, how the situation changes if the manufacturer has already designed its product and placed it into large-scale production before the patent issues. In this case, even though the timing is strongly suggestive that the manufacturer did not in fact rely on the patented invention for the design of its product, the manufacturer is in a far weaker negotiating position. The patentee can credibly seek far greater royalties, very likely backed up with the threat of shutting down the manufacturer if the Court indeed finds the patent valid and infringed and grants injunctive relief. The manufacturer could go back and redesign its product, but to do so (a) could well require a major redesign effort and/or cause a significant disruption to production, (b) would still leave potential liability for any products sold after the patent issued before the redesigned products are available for sale, and (c) could present compatibility problems with other products or between different versions of this product. In other words, for all of these reasons, the manufacturer is highly susceptible to hold-up by the patentee. I submit that this “hold-up” problem is very real today, and that both patent and antitrust policy makers should regard hold-up as a problem of first-order significance in the years ahead.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 7 of 32
+
+---
+
+The hold-up problem is worst in industries where hundreds if not thousands of patents, some already issued, others pending, can potentially read on a given product. In these industries, the danger that a manufacturer will “step on a land mine” is all too real. The result will be that some companies avoid the mine field altogether, i.e., refrain from introducing certain products for fear of hold-up. Other companies will lose their corporate legs, i.e., will be forced to pay royalties on patents that they could easily have invented around at an earlier stage, had they merely been aware that such a patent either existed or was pending. Of course, ultimately (the expected value of) these royalties must be reflected in the price of final goods.
+
+In short, with multiple overlapping patents, and under a system in which patent applications are secret and patents slow to issue (relative to the speed of new product introduction), we have a volatile mix of two powerful types of “transaction costs” that can burden innovation: (1) the complements problem, the solution of which requires coordination, perhaps large-scale coordination; and (2) the hold-up problem, which is quite resistant to solution in the absence of either (a) better information at an earlier stage about patents likely to issue, and/or (b) the ability of interested parties to challenge patents at the PTO before they have issued and are given some presumption of validity by the Courts.
+
+Clearly, these concerns form the basis for a serious discussion about reform of the patent system. $^7$ However, my intention in this paper is to explore how private companies can best navigate the patent system we currently have, and how our antitrust laws can be enforced in a way that is sensitive to the transaction costs associated with our current patent system. I see relatively little that private companies can do to overcome the hold-up problem without reform of the patent system itself. But there is quite a bit they can do to solve the complements problem, which itself is greatly exacerbated by the hold-up problem.
+
+## C. Overlapping Patents and Business Strategy in Practice
+
+To solve the complements problem generally, and to cut through the patent thicket specifically, requires coordination among rights holders. Such coordination itself faces two types of obstacles. First, there are inevitably coordination costs that must be overcome. Second,
+
+$^{7}$ For a thoughtful discussion of possible reforms at the Patent and Trademark Office, see Merges (1999).
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 8 of 32
+
+---
+
+antitrust sensitivities are invariably heightened when companies in the same or related lines of business combine their assets, jointly set fees of any sort, or even talk directly with one another. Because such coordination may involve the elimination of competition, we have a complex interaction between private and public interests. Even as coordination between rights holders is critical, from a public-policy perspective we cannot presume that private deals are in the public interest. Antitrust authorities will legitimately want to know whether consumers are helped or harmed by any arrangement; injured parties may seek redress under the antitrust laws or by alleging patent misuse.
+
+## 1. Cross-Licenses
+
+Cross-licenses commonly are negotiated when each of two companies has patents that may read on the other's products or processes. Rather than blocking each other and going to court or ceasing production, the two enter into a cross-license. Especially with a royalty-free cross-license, each firm is then free to compete, both in designing its products without fear of infringement and in pricing its products without the burden of a per-unit royalty due to the other. Thus, cross-licenses can solve the complements problem, at least among two firms, and thus be highly pro-competitive.
+
+A cross-license is simply an agreement between two companies that grants each the right to practice the other's patents. Cross-licenses may or may not involve fixed fees or running royalties; running royalties can in principle run in one direction or both. Cross-licenses may involve various field-of-use restrictions or geographic restrictions. Cross-licenses may involve some but not all relevant patents held by either party; “ carve-outs ” are not uncommon. And cross-licenses, like regular licenses, may be confined to patents issued (or pending) as of the date of the license, or they may include patents to be granted through a certain time in the future.
+
+## 2. Patent Pools and Package Licenses
+
+When two or more companies control patents necessary to make a given product, and when at least some actual or potential manufacturers may not themselves hold any such patents, a patent pool or a package license can be the natural solution to the complements problem. Under a patent pool, an entire group of patents is licensed in a package, either by one of the patent holders or by a new entity established for this purpose, usually to anyone willing to pay the associated
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 9 of 32
+
+---
+
+royalties. Under a package license, two or more patent holders agree to the terms on which they will jointly license their complementary patents and divide up the proceeds. A nice example of a patent pool is the Manufacturers Aircraft Association formed in 1917 to license a number of patents necessary for the production of airplanes, patents controlled by The Wright-Martin Aircraft Corporation, the Curtiss Aeroplane & Motor Corporation, and others. 8 I discuss below some more recent patent pools that have been used to help establish compatibility standards.
+
+## 3. Cooperative Standard Setting
+
+The need to solve the complements problem tends to especially great in the context of standard setting. For example, when the International Telecommunications Union (ITU) establishes a new standard for fax transmissions or modem protocols, the participants are loathe to agree to a standard that can be controlled by any single firm through its patents. Thus, standard-setting organizations like the ITU or the American National Standards Institute (ANSI) typically require that participants agree to license all patents essential to compliance with any standard on “fair, reasonable, and non-discriminatory” terms. Rules such as this are explicitly intended to reduce or eliminate any “hold-up” problems. However, it is well to note that many standard-setting organizations are wary of sanctioning any specific agreement regarding the magnitude of licensing terms for fear of antitrust liability, as such agreements might be construed as “price fixing.” Perversely, by leaving the precise licensing terms vague, this caution can in fact lead to ex post hold-up by particular rights holders, contrary both to the goal of enabling innovation and to consumers' interests.
+
+The case in which multiple firms control patents essential to a standard fits well with the formal economic analysis described above. In essence, any manufacturer seeking to product a compliant product must obtain a license from each rights holder to avoid facing an infringement action. Inventing around is typically impractical, as it would preclude the manufacturer from claiming that its products are “ compliant ” and thus assuring consumers that they are fully compatible with the prevailing standard. Thus, standard setting very often has especially strong elements of both the complements problem and the hold-up problem.
+
+8 See Klein (1997) for a further description of this pool and how it operated. In this case, the Assistant Secretary of the Navy, Franklin D. Roosevelt, had to lean on the industry to form to pool and help enable wartime production of aircraft.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 10 of 32
+
+---
+
+## 4. Settlements of Patent Disputes
+
+Cross-licenses (or simply licenses) are a most common way in which companies resolve patent disputes. But other forms of settlement arise, two of which I touch on below. First, I discuss acquisitions, in which one firms simply acquires the other, thereby resolving the dispute and assembling the various intellectual property rights within a single company. Second, I comment on cash payments in exchange for exit, a strategy whereby one company pays the other company to exit the market, and thus to drop its challenge to the first company's patent. In each of these cases, legitimate questions arise as to whether any particular private agreement truly is in the public interest.
+
+## D. Antitrust Limits
+
+As I have indicated, many of the business solutions to the complements problem and the hold-up problem raise antitrust issues. Quite generally, agreements among companies that either do compete, or might compete, directly with each other raise antitrust warning flags. For each business form, I consider below its antitrust treatment.
+
+Generally speaking, one can imagine two rather different approaches that antitrust might take to firms' efforts to coordinate to solve the complements problem. One approach is to ask whether the agreement in question leads to more competition than would occur without that agreement. This is the approach advocated in the Department of Justice and Federal Trade Commission Antitrust Guidelines for the Licensing of Intellectual Property, which state in §3.1 that:
+
+However, antitrust concerns may arise when a licensing arrangement harms competition among entities that would have been actual or likely potential competitors in a relevant market in the absence of the license (entities in a “ horizontal relationship ” ).
+
+Another, quite different approach, would be to ask whether the agreement in question is the most competitive agreement possible. Put differently, one could ask whether a given agreement is the least restrictive alternative that is workable in the sense of solving the legitimate business problem faced, such as unblocking patent positions. Clearly, this latter standard, which does not reflect current antitrust enforcement policy according to the Guidelines, would be far tougher on all forms of cooperation among patent and copyright holders.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 11 of 32
+
+---
+
+## III. Cross Licensing
+
+### A. Cross Licenses and Design Freedom
+
+Cross-licenses are the preferred means by which large companies clear blocking patent positions amongst themselves. Based in part on work I have done on behalf of Intel, I can report that broad cross-licenses are the norm in markets for the design and manufacture of microprocessors. 9 For example, Intel has entered into a number of broad cross-licenses with other major industry participants, such as IBM, under which most of each company's vast patent portfolio is licensed to the other. Furthermore, the companies generally agree to grant licenses to each other for patents that will be issued several years into the future, typically for the lifetime of the cross-licensing agreement. Often, these cross licenses involving no running royalties, although they may involve “balancing payments” at the outset to reflect differences in the strength of the two companies' patent portfolios as reflected in a “patent pageant,” and/or the vulnerability of each to an infringement action by the other. For example, Hewlett-Packard and Xerox recently announced a cross-license that settled their outstanding patent disputes.
+
+From the perspective of competition policy, cross-licenses of this sort are quite attractive. The traditional concern with cross-licenses among competitors is that running royalties will be used as a device to elevate prices and effectuate a cartel; see Katz and Shapiro (1985) . Clearly, such concerns do not apply to licenses that involve small or no running royalties, but rather have fixed up-front payments. Another concern is that the granting of licenses to future patents will reduce each company's incentive to innovate because its rival will be able to imitate its improvements. $^10$ While correct in theory, it is clear, at least in the case of semiconductors and no doubt more widely, that this concern is dwarfed by the benefits arising when each firm enjoys enhanced design freedom by virtue of its access to the other firm's patent portfolio. There is little doubt that these broad cross licenses permit the more efficient use of engineers (arguably the
+
+9 See Hall and Ham (1999) and Grindley and Teece (1997) for a additional studies of licensing practices in the semiconductor industry.
+
+10 This concern about discouraging innovation also arises with respect to grantbacks, under which one company agrees to license its future patents in exchange for rights to use an existing patent held by another company. See Gilbert and Shapiro (1997) for a further discussion of grantbacks.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 12 of 32
+
+---
+
+resource that governs the rate of innovation in the semiconductor industry), better products, and faster product-design cycles. In other words, when IBM and Intel sign a forward-looking cross license, each is enabled to innovate more quickly and more effectively without fear that the other will hold it up by asserting a patent that it has unintentionally infringed. And neither firm is really all that worried that the other will actually copy its products, just because the other has a license to most of its patents. Of course, the impressive rate of innovation in the semiconductor industry in the presence of a web of such cross-licenses offers direct empirical support for the view that these cross-licenses promote rather than stifle innovation.
+
+## B. Intel's Policy of "IP For IP"
+
+Despite all of these benefits, the Federal Trade Commission attacked Intel's crosslicensing practices in 1998. $^11$ One key episode behind the FTC's complaint involved Intel's conduct when faced with a lawsuit by Intergraph, a workstation manufacturer, asserting that Intel's microprocessors infringed on certain patents held by Intergraph. Of course, lawsuits like Intergraph's are a necessary part of (the “ threat point ” behind) any cross-licensing negotiation: if one party is not happy with the terms offered by the other, it always has the option of initiating patent litigation. In response to Intergraph's infringement action against Intel, Intel withdrew its own intellectual property from Intergraph by suing Intergraph for infringement of Intel's patents and by withdrawing the supply of Intel trade secrets to Intergraph, trade secrets which Intergraph valued highly for the purposes of designing systems built on Intel chips.
+
+Evidently viewing Intel's conduct as “ unfair, ” the FTC attempted to fashion an antitrust case against Intel based on this conduct, along with a similar response by Intel to a lawsuit initiated by Digital Equipment Corporation. 12 The FTC action against Intel sharply exposed the fact that the FTC and Intel had fundamentally different views about the impact of the conduct at issue. The FTC saw Intel as using its existing monopoly power to fortify its position by lowering its royalty costs per chip and potentially offering superior products by incorporating technologies patented by others. Intel viewed itself as engaging in a defensive exercise which was a necessary
+
+11 In the Matter of Intel Corporation, Docket No. 9288, Complaint filed June 8, 1998. The Complaint is available at http://www.ftc.gov/os/1998/9806/intelfin.cmp.htm. I was retained by Intel to work on this matter.
+
+12 For one well-informed articulation of the theory underlying the FTC's position, see Baker (1999).
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 13 of 32
+
+---
+
+aspect of cross-licensing, namely trading intellectual property for intellectual property (“IP for IP”) and withdrawing its own intellectual property when faced with a frontal assault on its core product line in the form on an infringement action seeking injunctive relief. Intel, well aware what a juicy target it posed, believed it had every right to protect itself from hold-up, and certainly no duty to give special treatment in the form of Intel trade secrets and advance product samples to a company attempting to hold it up.
+
+The problem for the FTC was that the conduct at issue, especially with respect to Intergraph, was directed at a customer of Intel's, not a competitor. Brushing aside concerns about hold-up, and playing down the important role of cross-licenses in the semiconductor industry, the FTC found no “ business justification ” for Intel's conduct, and thus was prepared to presume that the conduct was anticompetitive without actually studying the impact of the conduct on Intel's competitors. In fact, Intel's true rivals in microprocessor design and manufacturing (such as AMD, Motorola, Sun, or IBM) were either not subject to the conduct at issue (since they were not Intel customers at all and thus not recipients of the Intel trade secrets at issue), or had ongoing cross-licenses with Intel under which the litigation triggering these episodes would simply not occur in the first place.
+
+Fortunately, a compromise was reached and a settlement agreed to between the FTC and Intel. 13 In essence, Intel agreed not to withdraw product information needed by its customers to build systems based on soon-to-be-released Intel chips. (Presumably, this promise provides some benefit to Intel by assuring its customers that they will not be held up once relying on Intel for their new systems.) But Intel is not obligated to continue to provide trade secrets on products farther out on their roadmap (i.e., products that will not be introduced for a year or two) to customers suing Intel, and Intel was not obligated to provide any trade secrets to a company suing Intel and seeking a court injunction to shut down Intel's microprocessor business.
+
+The Intel situation also exposes the interplay between government enforcement of the antitrust laws and private antitrust actions. Even while the FTC was investigating Intel, bringing a complaint against Intel, and ultimately settling with Intel, Intergraph was engaged in its own antitrust and patent battle with Intel. Intergraph won a resounding victory in the first round of the
+
+13 For more information on the settlement between the FTC and Intel, see http://www.ftc.gov/os/1999/9903/d09288intelagreement.htm.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 14 of 32
+
+---
+
+that battle, in which the District Court judge in Alabama issued a searing anti-Intel opinion ruling, among other things, that Intel's microprocessors and associated trade secrets were “essential facilities” under antitrust laws, thus imposing a duty on Intel to sell its microprocessors to Intergraph and to make its trade secrets available to Intergraph, Intergraph's lawsuit against Intel notwithstanding. This opinion was based on strands of antitrust law that require dominant companies to deal with their rivals, especially if the dominant firm has established an ongoing course of dealing with rivals in the past. $^14$
+
+Ultimately, however, Intel was vindicated. The District Court judge later ruled that Intel was not in fact infringing on Intergraph's patents. And, most significantly, the Court of Appeals for the Federal Circuit vacated the District Court's antitrust and essential facility opinion. 15 In a strongly worded and sweeping opinion, the appeals court ruled that Intel's conduct did not violate the antitrust laws because it was not directed at a competitor and indeed could have no adverse impact on competition in the market where Intel was alleged to have monopoly power, namely the market for microprocessors, in which Intergraph did not compete. The FTC's efforts to fashion an antitrust case out of Intel's conduct look even more dubious now in the light of this subsequent decision by the Court of Appeals.
+
+The Intel episode is closely related to another ongoing debate regarding the intersection between intellectual property rights and antitrust law: can a company violate the antitrust laws simply by refusing to license its patents, or by refusing to sell patented items, to its rivals? Most commentators have said for some time that a refusal to license patents cannot in and of itself constitute an antitrust violation. However, the Supreme Court has signaled that unilateral refusals to sell can indeed constitute antitrust violations, especially if a company has established an ongoing course of dealing with its rivals. 16 The precise conditions under which a refusal to license a patent (or to sell patented items) could constitute an antitrust violation has remained unclear.
+
+14 The key recent Supreme Court case here is Aspen Skiing Company vs. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985), although the essential facilities doctrine goes back to the case of U.S. vs. Terminal Railroad Association of St. Louis, 224 U.S. 383 (1912).
+
+15 Intergraph Corporation v. Intel Corporation, United States Court of Appeals for the Federal Circuit, 98-1308, Decided November 5, 1999, Judge Newman writing the opinion for the Court.
+
+16 The classic cites are Otter Tail Power Co. vs. U.S., 410 U.S. 366 (1973) (duty to sell wholesale electric power to a retail competitor) and Aspen Skiing Company vs. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985), (duty to continue to offer a joint lift ticket with a rival ski slope).
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 15 of 32
+
+---
+
+Most observers were stunned when the Ninth Circuit Court of Appeals ruled in 1997 that Kodak was liable for refusing to sell patented spare parts for its machines to independent service organizations seeking to compete against Kodak in the business of servicing Kodak copiers and micrographics equipment. As the Court acknowledged, this was the first time a unilateral refusal to sell a patented item had been judged to be an antitrust violation. $^17$ Just recently, the Court of Appeals for the Federal Circuit came to a very different conclusion, ruling that a company's unilateral decision not to license a patent (or sell a patented item) could never in and of itself constitute an antitrust violation. $^18$ Hopefully, the Supreme Court will resolve this significant split among the Circuit Courts and clarify that unilateral refusals to license patents are immune from antitrust challenge.
+
+Intel's practices, and those of other firms who require “ grantbacks ” of relevant patents in exchange for a license to key enabling patents, copyrights, or trade secrets, raises further interesting questions about the role of “ self help ” in the digital economy. 19 One view of such business strategies cum legal regimes is that they are a welcome effort by leading firms to establish a type of “ litigation-free ” zone likely to favor innovation and get around some of the current difficulties with our patent system and the patent thicket it causes. A less favorable view is that these arrangements represent efforts by powerful firms to establish private legal regimes that favor themselves and make it more difficult for upstarts to challenge the dominance of current market leaders. Is a cross-licensing policy of “ IP for IP ” a beneficial way to cut through the patent thicket, or a strong-arm tactic by a dominant firm that enjoys powerful patent rights and seeks access to others' intellectual property in exchange?
+
+17 The Court set up a tortured standard under which a company's decision to refuse to license its patent was “presumptively valid,” but could be overcome by evidence that the company's intent was anticompetitive. Of course, asserting intellectual property rights against a would-be rival is typically “anti-competitive” in the sense of trying to eliminate a competitor (or at least earn royalties from the competitor, which add to the competitor's costs), so this test is not in fact workable. Amazingly, the Court said that Kodak would be justified in refusing to sell patented parts if its intent was to earn a return on its R&D investment required to design and manufacture those parts, but not if its intent was to eliminate competitors who rely on those very patented parts. I testified on behalf of Kodak in this case.
+
+18 United States Court of Appeals for the Federal Circuit, 99-1323, In Re Independent Service Organizations Antitrust Litigation, CSU, et. al. v. Xerox Corporation, Decided February 17, 2000, Judge Mayer writing the opinion.
+
+19 For a discussion of self-help focusing on copyright holders, see Dam (1998).
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 16 of 32
+
+---
+
+## IV. Patent Pools
+
+A patent pool involves a single entity (either a new entity or one of the original patent holders) that licenses the patents of two or more companies to third parties as a package. In many respects, a patent pool (much like a package license) is the purest solution to the complements problem described above and analyzed in the Appendix. Indeed, licensees may well welcome such a pool, both for the convenience of “one-stop shopping” and because a subset of the required patents may be of little or no value by themselves. Thus, from the licensee's perspective, licensing the entire package is simpler and avoids the danger of paying for some patent rights that turn out to be useless without other complementary rights.
+
+### A. Essential Patents vs. Rival Patents
+
+The Department of Justice has clearly articulated its policy towards patent pools/package licensing in a trio of business review letters regarding an MPEG patent pool and two DVD patent pools. The essence of this approach, which precisely mirrors the economic principles articulated above, is that inclusion of truly complementary patents in a patent pool is desirable and procompetitive, but assembly of substitute or rival patents in a pool can eliminate competition and lead to elevated license fees. Put differently, the key distinction in forming a patent pool is that between “blocking” or “essential” patents, which properly belong in the pool, and “substitute” or “rival” patents, which may need to remain separate.
+
+In the MPEG case, 20 the Department approved the creation of a pool of patents necessary to enable manufacturers to meet the MPEG-2 video compression technology. This pool, encompassing patents from Fujitsu, General Instrument, Lucent, Matsushita, Mitsubishi, Philips, Scientific-Atlanta, Sony, and Columbia University, permits “one-stop shopping” for makers of televisions, digital video disks and players, telecommunications equipment as well as cable, satellite, and broadcast television services. To support their formation of a patent pool, these nine patent holders conducted an extensive search to identify all patents essential to the MPEG-2 standard and include them in the pool. The licensing agent for the pool, MPEG LA, will employ an independent patent expert to determine whether a patent in the pool is in fact essential, and
+
+20 See the June 26, 1997 press release, http://www.usdoj.gov/atr/public/press_releases/1997/1173.htm.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 17 of 32
+
+---
+
+whether other patents as well are essential and thus suitable for inclusion in the pool. As stated by the Department, “the use of the independent-expert mechanism will help ensure that the portfolio will contain only patents that are truly essential to the MPEG-2 standard, weeding out patents that are competitive alternatives to each other.”
+
+In the first Digital Versatile Disk (DVD) case, 21 the Department approved a proposal by Philips, Sony, and Pioneer to jointly license patents necessary to make discs and players that comply with the DVD-Video and DVD-ROM standards. Again, only essential patents are to be included in the joint licensing program. As with the earlier CD licensing program of Sony and Philips, licenses will be offered by Philips, in this case on behalf of all three firms. Again, an independent patent expert will be employed to ensure that the license only conveys the rights to essential patents. As stated by the Department, “the expert will help ensure that the patent pool does not combine patents that would otherwise be competing with each other.” The Department subsequently approved a second joint licensing scheme relating to the DVD-Video and DVDROM standards, 22 this one including patents held by Toshiba (the licensing entity), Hitachi, Matsushita, Mitsubishi, Time Warner, and Victor Company of Japan. Note that the effect of these two patent pools appears to be to reduce but not eliminate the complements problem, since there remain two separate pools, not just one: “two-stop shopping,” it would appear.
+
+## B. A Patent Pool Created to Resolve Claims of Blocking Patents
+
+In contrast to the Department of Justice's approval of these three patent pools, the Federal Trade Commission in March 1998 challenged a patent pool formed by Summit Technology, Inc. and VisX, Inc. two firms that manufacture and market lasers to perform a new, and increasingly popular, vision correcting eye surgery, photorefractive keratectomy. 23 According to the FTC: “Instead of competing with each other, the firms placed their competing patents in a patent pool and share the proceeds each and every time a Summit or VISX laser is used.” The FTC was ostensibly following the same principles employed by the Justice Department, namely to permit the assembly of complementary or essential patents, but not rival patents, into a pool. According
+
+21 See the December 17, 1998 press release, http://www.usdoj.gov/atr/public/press_releases/1998/2120.htm.
+
+22 See the June 10, 1999 business review letter, http://www.usdoj.gov/atr/public/press_releases/1999/2484.htm.
+
+23 See the March 24, 1998 press release, http://www.ftc.gov/opa/1998/9803/eye.htm.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 18 of 32
+
+---
+
+to the FTC, the two companies agreed not to license their patents independently. However, the companies in this case argued vigorously that they did indeed have mutually blocking patents, making their pool, Pillar Point Partners, pro-competitive. In August 1998 the two companies settled with the FTC and agreed to lift any restrictions on each other regarding the licensing of their patents; ultimately, their patent pool was dissolved. 24
+
+The Summit and VisX case raises a number of very interesting an tricky issues regarding patent pools and joint licensing programs. First, if Summit and VisX reasonably believed that their patents blocked each other at the time they formed the pool, was that sufficient to justify the formation of a pool? How hard were they required to look into the validity of each other's claims before agreeing to form Pillar Point Partners? Second, if each firm believed it could, at considerable expense, delay, and risk, invent around the other's patents, should the two firms be prohibited from forming a pool and rather forced to attempt to invent around each other's patents, under the view that consumer might thereby enjoy the benefits of direct competition (although the product might be delayed, or never introduced, in the absence of the pool)? Third, is there competitive harm in placing some potentially rival patents into the pool, assuming that each party in fact controls valid blocking patents, making some type of pool pro-competitive? Fourth, can the pool be attacked on antitrust grounds based on the argument that a less restrictive alternative, namely a cross-license, would have achieved the same legitimate purposes and created additional competition? If so, does it matter in this assessment that Summit and VisX agreed that the pool would license their patents to third parties, something that a cross-license would not permit, unless it contained rather unusual sublicensing rights?
+
+## V. Cooperative Standard Setting
+
+Blocking patents are especially common in the context of standard-setting: once a standard is picked, any patents (or copyrights) necessary to comply with that standard become truly essential. If the standard becomes popular, each such patent can confer significant market
+
+24 For a description of the settlement, see the August 21, 1998 press release, http://www.ftc.gov/opa/1998/9808/sumvisx.htm. Despite this settlement, the FTC continued to pursue VisX for allegedly acquiring a key patent by inequitable conduct and fraud by omission on the U.S. Patent and Trademark Office. However, an administrative law judge subsequently dismissed this complaint; see the June 4, 1999 press release, http://www.ftc.gov/opa/1999/9906/visx.htm.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 19 of 32
+
+---
+
+power on its owner, and the standard itself is subject to “ hold-up ” if these patent holders are not somehow obligated to license their patents on “ reasonable ” terms. As noted above, for precisely this reason, standard-setting bodies require participants to license any essential patents on reasonable terms as a quid pro quo before adopting any standards. $^25$
+
+Fortunately, antitrust concerns have not prevented a great many cooperative standardsetting efforts from proceeding forward. Some participants go so far as to say that much of he innovation taking place now in the telecommunications, Internet, and computer areas is standardsbased. Indeed, even the fiercest enemies often team up in the software industry to promote new standards. Back in 1997, Microsoft and Netscape, two companies hardly known as cozy partners, agreed to include compatible versions of Virtual Reality Modeling Language (developed by Silicon Graphics) in their browsers. This agreement was expected to make it far easier for consumers to view 3-D images on the Web. Earlier, Microsoft agreed to support the Open Profiling Standard, which permits users of personal computers to control what personal information is disclosed to a particular Web site, and which had previously been advanced by Netscape, along with Firefly Network, Inc. and Verisign Inc.
+
+But neither is cooperative standard setting immune from antitrust scrutiny. In the consumer electronics area, for example, the Justice Department investigated Sony, Philips, and others regarding the establishment of the CD standard in the 1980s. Cooperative efforts to set optical disc standards have also been challenged in private antitrust cases, on the theory that agreements to adhere to a standard are an unreasonable restraint of trade:
+
+[d]efendants have agreed, combined, and conspired to eliminate competition... by agreeing not to compete in the design of formats for compact discs and compact disc players, and by instead agreeing to establish, and establishing, a common format and design... 26
+
+25 Note that these rules can create the perverse incentive for patent holders to assert that at least some of their patents are not in fact essential, but perhaps merely extremely helpful, in complying with the standard. By this device, a patent holder can in principle either refuse to license its patent to others (especially once the standard has become established, and perhaps for a patent that issued after the standard is established) or seek something more than “fair and reasonable” royalties. Of course, whether the terms “fair and reasonable” are evaluated on an ex ante or ex post basis is not precisely clear, although the terms would have little force if applied only on an ex post basis.
+
+26 "Second Amended Complaint," Electronics Texas, Inc., et al. v. Pioneer Electronic Corp. et al. Eastern District of Texas, Case No. 4:95 CV 229, filed August 2, 1996 at 12.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 20 of 32
+
+---
+
+Does cooperation lead to efficient standardization, increased competition, and additional consumer benefits? Or is cooperative standard setting a means for firms collectively to stifle competition, to the detriment of consumers and firms not included in the standard-setting group? Answering these questions and evaluating the limits that should be placed on cooperative standard-setting efforts require an analysis of the competitive effects of such cooperation in comparison with some reasonable but-for world. Inevitably, an antitrust analysis of cooperative standard-setting involves an assessment of how the market would likely evolve without the cooperation. One possibility is that multiple, incompatible products would prevail in the market, if not for the cooperation. Another possibility is that the market would eventually tip to a single product, even without cooperation. Even in this latter case, an initial industrywide standard can have significant efficiency and welfare consequences, for three reasons: (1) cooperation may lock in a different product design than would emerge from competition; (2) cooperation may eliminate a standards war waged prior to tipping; and (3) cooperation is likely to enable multiple firms to supply the industry-standard product, whereas a standards war may lead to a single, proprietary product.
+
+## A. The Costs and Benefits of Compatibility and Standards
+
+There are significant benefits associated with achieving compatibility. These include:
+
+- • Successful Launching of a Bandwagon or Network
+• Greater Realization of Network Effects
+• Protecting Buyers from Stranding
+• Enabling Competition Within an Open Standard
+Likewise, standardization and compatibility can impose very real costs on consumers:
+
+- • Constraints on Variety and Innovation
+• Loss of Ex Ante Competition to Win the Market
+• Proprietary Control Over a Closed Standard
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 21 of 32
+
+---
+
+## B. Legal Treatment of Cooperative Standard Setting
+
+I now look more closely at the intellectual property issues that arise specifically in the context of standard setting, where the participants typically agree to license their patents on “fair, reasonable, and non-discriminatory” terms.
+
+Firms are sometimes accused of hiding intellectual property rights until after the proprietary technology has been embedded in a formal standard. I view this issue primarily as one of contract law. Standard-setting groups typically have provisions in their charters compelling participants either to reveal all relevant intellectual property rights or to commit to licensing any intellectual property rights embedded in the standard on “reasonable” terms. 27 Clearly, these rules help control the “hold-up” problem. In some cases, however, the precise requirements imposed by a standard-setting group may be unclear. In these circumstances, if the standard affects nonparticipants, including consumers, there is a public interest in clarifying the duties imposed on participants in a fashion that promotes rather than stifles competition.
+
+The question of whether firms should be allowed, or even encouraged, to set standards cooperatively is part of the broader issue of collaboration among competitors, a storied area within antitrust law. Most of the case law deals with quality and performance standards rather than compatibility standards . $^28$ Existing cases also have tended to focus on the standard-setting process itself, rather than the outcomes of cooperative standard setting.
+
+Antitrust liability has been found for participants in a standard-setting process who abuse that process to exclude competitors from the market. One leading case is Allied Tube & Conduit Corp. v. Indian Head, Inc., 486 U.S. 492 (1988), in which the Supreme Court affirmed a jury verdict against a group of manufacturers of steel conduit for electrical cable. These manufacturers conspired to block an amendment of the National Electric Code that would have permitted the use of plastic conduit. They achieved this by “packing” the annual meeting of the National Fire Protection Association, whose model code is widely adopted by state and local
+
+27 Note that a company might profit from refusing to participate in the standard-setting process, in the hope that the resulting standard will nonetheless (perhaps inadvertently) infringe on the company's patent. Then the company would not be obligated to license its blocking patent on fair and reasonable terms, if at all. This would at least create the possibility that the company in question could “hijack” the standard and make it proprietary once it became established.
+
+$^{28}$ See Anton Yao (1995) for a more complete discussion of the legal treatment of performance standards.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 22 of 32
+
+---
+
+governments. The other leading case is American Society of Mechanical Engineers v. Hydrolevel Corp., 456 U.S. 556 (1982), in which the Supreme Court affirmed an antitrust judgment against a trade association. In this case, the chairman of an association subcommittee offered an “unofficial” ruling that plaintiff’s product was unsafe, and this ruling was used by plaintiff’s rival (who enjoyed representation on the sub-committee) to discourage customers from buying plaintiff’s product.
+
+Antitrust risks associated with excluding a rival from the market appear to be less of a problem for an “open” standard, but could arise if the companies promoting the standard block others from adhering to the standard or seek royalties from outsiders. The DOJ business review letters regarding the MPEG-2, DVD-Video, and DVD-ROM standards are excellent illustrations of how the enforcement agencies can successfully handle intellectual property in the standard-setting context.
+
+As the Supreme Court has noted, “ Agreement on a product standard is, after all, implicitly an agreement not to manufacture, distribute, or purchase certain types of products. ” 29 To date, this type of reasoning has not been used to impose per se liability on software standard-setting activities. Indeed, I know of no successful antitrust challenges to cooperation to set compatibility standards. The closest case of which I am aware is Addamax Corporation v. Open Software Foundation, Inc., 888 F. Supp. 274 (1995). In Addamax , the District Court refused to grant summary judgment on behalf of the Open Software Foundation, an industry consortium formed to develop a platform-independent version of the UNIX operating system. OSF conducted a bidding to select a supplier of security software. After failing to be selected, Addamax brought antitrust claims against OSF, Hewlett Packard, and Digital Equipment Corporation, asserting that OSF had chosen the winner not based on the merits but to favor specific companies and technologies. The Addamax case looks problematic, inasmuch as the primary purpose of OSF was to permit its members to team up to offer stronger competition against the leading UNIX vendors, Sun Microsystems and AT & T, and there was no evidence suggesting that OSF's failure to pick Addamax was based on its members desire to control the market in which Addamax itself operated.
+
+29 Allied Tube & Conduct Corp. v. Indian Head, Inc., 486 U.S. 492, 500 (1988).
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 23 of 32
+
+---
+
+Ultimately, the antitrust risks faced by companies who are trying to set compatibility standards appear to be relatively minor as long as the scope of the agreement truly is limited to standard setting and steers clear of distribution, marketing, and pricing. While the law has typically looked for integration and risk-sharing among collaborators in order to classify cooperation as a joint venture and escape per se condemnation, these are not very helpful useful screens for standard-setting activities. The essence of cooperative standard setting is not the sharing of risks associated with specific investments, or the integration of operations, but rather the contribution of complementary intellectual property rights and the expression of unified support to ignite positive feedback for a new technology.
+
+The limits imposed by public policy in the area of compatibility standards remain unclear. The most specific statement by the antitrust enforcement agencies can be found in a recent FTC Staff Report. 30 The Staff Report recognized a need for clarification in this area:
+
+“the time has come for a significant effort to rationalize, simplify, and articulate in one document the antitrust standards that federal enforcers will apply in assessing collaborations among competitors. This effort should be directed at drafting and promulgating ‘competitor collaboration guidelines’ that would be applicable to a wide variety of industry settings and flexible enough to apply sensibly as industries continue rapidly to innovate and evolve.” 31
+
+Since that call for action, the FTC has conducted Joint Venture Hearings, and the Commission and the Antitrust Division issued on October 1, 1999 a draft of new “ Antitrust Guidelines for Collaborations Among Competitors ” (available at either Agency's web site).
+
+## C. Hidden Patents and Hold-Up in Standard Setting
+
+A number of disputes have surfaced recently that illustrate the thorny problems associated with “hidden” patent rights that were later exerted against established standards. 32
+
+30 Federal Trade Commission. "Anticipating the 21st Century: Competition Policy in the New High-Tech Global Marketplace," Chapter 9, "Networks and Standards," (June 1996).
+
+31 Ibid, Chapter 10, "Joint Ventures," (June 1996) at 17.
+
+32 There are many more examples of disputes involving “hidden” patent rights and standard setting, including: Wang vs. Mitsubishi; Microsoft and Cascading Style Sheets; and ETSI and Third-Generation Mobile Telephones.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 24 of 32
+
+---
+
+## 1. Dell Computer and the VESA VL-bus Standard
+
+The leading U.S. example of this type of antitrust action is the FTC's consent agreement with Dell Computer Corporation, announced in November 1995. Although the case involved computer hardware, it is important for the software community as well. The assertion was that Dell threatened to exercise undisclosed patent rights against computer companies adopting the VL-bus standard, a mechanism to transfer data instructions between the computer's CPU and its peripherals such as the hard disk drive or the display screen. The VL-bus was used in 486 chips, but it has now been supplanted by the PCI bus. According to the FTC,
+
+“During the standard-setting process, VESA [Video Electronics Standard Association] asked its members to certify whether they had any patents, trademarks, or copyrights that conflicted with the proposed VL-bus standard; Dell certified that it had no such intellectual property rights. After VESA adopted the standard -- based in part, on Dell's certification -- Dell sought to enforce its patent against firms planning to follow the standard.” 33
+
+There were two controversial issues surrounding this consent decree: (a) the FTC did not assert that Dell acquired market power, and indeed the VL-bus never was successful; and (b) the FTC did not assert that Dell intentionally misled VESA. My analysis suggests that anticompetitive harm is unlikely to arise in the absence of significant market power and that the competitive effects are not dependent on Dell's intentions.
+
+## 2. Motorola and the ITU V.34 Modem Standard
+
+Another good example of how competition can be affected when standard-setting organizations impose ambiguous duties on participants is the case of Motorola and the V.34 modem standard adopted by the International Telecommunications Union. Motorola agreed to license its patents essential to the standard case to all comers on “fair, reasonable, and nondiscriminatory terms.” 34 Once the standard was in place, Motorola then made offers that some industry participants did not regard as meeting this obligation. Litigation ensued between Rockwell and Motorola, in part over the question of whether “reasonable” terms should mean: (a)
+
+33 See http://www.ftc.gov/opa/9606/dell2.htm.
+
+34 I served as an expert in this matter retained by Rockwell; the views stated here do not necessarily reflect those of any party to the case.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 25 of 32
+
+---
+
+the terms that Motorola could have obtained ex ante, in competition with other technology that could have been placed in the standard; or (b) the terms that Motorola could extract ex post, given that the standard is set and Motorola's patents are essential to that standard.
+
+These issues are best dealt with by the standard-setting bodies, or standard-setting participants, either by making more explicit the duties imposed on participants, or by encouraging ex ante competition among different holders of intellectual property rights to get their property into the standard. Unfortunately, antitrust concerns have led at least some of these bodies to steer clear of such ex ante competition, on the grounds that their job is merely to set technical standards, not to get involved in “ prices, ” including the terms on which intellectual property will be made available to other participants. The ironic result has been to embolden some companies to seek substantial royalties after participating in formal standard setting activities.
+
+## VI. Settlements of Patent Disputes
+
+Cross-licenses and patent pools can be ways to settle intellectual property disputes. For example, the Summit and VisX patent pool discussed above, Pillar Point Partners, was essentially a settlement of a patent dispute between Summit and VisX.
+
+Generally speaking, antitrust authorities have legitimate concerns that parties will settle their intellectual property disputes in ways that stifle competition. As a matter of economic theory, there is no reason to expect the two parties' collective interests in settlement, and especially in the form of any settlement they adopt, to coincide with the public interest, which includes consumer interests. So, while the law surely welcomes the settlement of disputes generally, and does not seek to force parties to litigate to the death, some settlements can be anticompetitive. Based on this general view, Assistant Attorney General for Antitrust Joel Klein recently suggested (see Klein (1997)) that parties notify the Justice Department of certain settlements that they enter into, much as parties are required to notify the Justice Department and the FTC in advance of their intention to merge.
+
+Firms are quite creative in crafting settlements to intellectual property disputes, and by no means restrict their attention to cross-licenses and patent pools. For example, one tried and true method of settling a dispute is for the companies involved simply to merge. However, the antitrust authorities are well aware that such mergers can themselves eliminate competition, and
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 26 of 32
+
+---
+
+they will view such mergers with skepticism if there is a good change that the two parties will in fact be capable of competing against each other, their patent claims notwithstanding. A good example of such a merger that was modified in response to FTC concerns was the proposed merger of Boston Scientific and CVIS in the area of imaging catheters. 35 An interesting twist in such cases is that the parties' posturing in court, where they each have an incentive to assert that they are not infringing on the other's patents, provides direct ammunition to the FTC or DOJ to assert that the two companies could indeed compete independently if not for the merger.
+
+A second method that companies can use to settle a patent dispute is for one company to simply pay the other company to drop its claims and exit the market. Such agreements raise obvious antitrust concerns, because an incumbent firm may be willing to pay handsomely to eliminate a potential competitor and avoid the risk of having its patent challenged, especially if no equally effective challenger is likely to arrive on the scene any time soon. The losers in such deals can easily be subsequent would-be entrants (if the patent were struck down) or consumers (who would benefit from a finding that the patent at issue is invalid or not infringed). Put differently, a settlement can generate negative externalities, either to other firms or to consumers, and thus there is a legitimate role of the Courts and the antitrust enforcement agencies to oversee such settlements.
+
+One class of settlements that are suspicious on their face is that involving agreements between incumbent manufacturers of branded pharmaceuticals and would-be rivals who seek to offer generic competition by challenging the validity of the patents underlying the branded product's dominant position. It has been reported recently that the FTC is considering challenging several such settlements. 36 These cases have the interesting twist resulting from the fact that certain generic manufacturers can gain preferential rights to enter the market before others are permitted to do so. As a result, the branded manufacturer may be able to stall
+
+35 See the May 3, 1995 press release, http://www.ftc.gov/opa/1995/9505/boscvis.htm. The recently proposed merger of Gemstar and TV Guide is another example of a merger/settlement that raises antitrust issues.
+
+36 One episode under investigation involves Abbott Laboratories, Novartis's Geneva Pharmaceuticals unit, and the popular hypertension drug, Hytrin. Another episode involves Aventis (the new company formed from the merger of Hoechst and Rhone-Poulenc), Andrx, and the heart drug Cardizem CD. Abbott reportedly agreed to pay Geneva $4.5 million per month to delay the launch of a generic version of Hytrin. Abbott asserts that its agreement with Geneva is “in accordance with all laws.” See the Wall Street Journal , February 7, 2000, “FTC Panel Backs Suit Against Abbott, Novartis on Deal for Hypertension Drug,” p. B20.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 27 of 32
+
+---
+
+competition by entering into a suitable agreement with the uniquely-placed generic manufacturer, knowing that subsequent rivals will face some delay. In order to identify and prevent any anticompetitive agreements of this nature, the FTC has asked that the FDA require companies to notify the FDA of any such settlements and make that information available to the FTC for its review.
+
+## VII. Conclusions
+
+Our current patent system is causing a potentially dangerous situation in several fields, including biotechnology, semiconductors, computer software, and e-commerce in which a wouldbe entrepreneur or innovator may face a barrage of infringement actions that it must overcome to bring its product or service to market. In other words, we are in danger of creasing significant transactions costs for those seeking to commercialize new technology based on multiple patents, overlapping rights, and hold-up problems. Under these circumstances, it is fair to ask whether the pendulum has swung too far in the direction of strong patent rights, ranging from the standards used at the Patent and Trademark Office for approving patent applications, to the secrecy of such applications, to the presumption afforded by the courts to patent validity, to the right of patent holders to seek injunctive relief by insisting that infringing firms cease production of the offending products.
+
+Under these circumstances, we can ill afford to further raise transactions costs by making it difficult patentees possessing complementary and potentially blocking patents to coordinate to engage in cross-licensing, package licensing, or to form patent pools. Yet antitrust law can potentially play such a counterproductive role, especially since antitrust jurisprudence starts with a hostility towards cooperation among horizontal rivals.
+
+So far, the Department of Justice has displayed a keen understanding of the need for those holding complementary rights to coordinate in the licensing of those rights, but the Federal Trade Commission has exhibited less restraint, and arguably is making it more difficult for firms to engage in cross-licenses, to offer package licenses, or to form pro-competitive patent pools. Many of these issues are likely to be extremely important in the near future, especially with the rise of standard-setting as an essential part of the process by which new technologies are commercialized.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 28 of 32
+
+---
+
+## References
+
+Anton, James and Dennis Yao, (1995), "Standard-Setting Consortia, Antitrust, and HighTechnology Industries," Antitrust Law Journal, vol. 64, pp. 247-265.
+
+Baker, Jonathan B., (1999), "Promoting Innovation Competition Through the Aspen/Kodak Rule," George Mason Law Review, vol. 7, pp. 495-521.
+
+Balto, David, (1999), "Networks and Exclusivity: Antitrust Analysis to Promote Network Competition," George Mason Law Review, vol. 7, pp. 523-576.
+
+Cohen, Wesley M., Richard R. Nelson and John Walsh, (1997) "Appropriability Conditions and Why Firms Patent and Why They Do Not in the American Manufacturing Sector."
+
+Dam, Kenneth, (1998), "Self-Help in the Digital Jungle," John M. Olin Law & Economics Working Paper No. 59, University of Chicago Law School, August.
+
+Farrell, Joseph and Michael Katz, (1998), "The Effects of Antitrust and Intellectual Property Law on Compatibility and Innovation," Antitrust Bulletin, 1998.
+
+Federal Trade Commission, "Competition Policy in the New High-Tech Global Marketplace," Staff Report, May 1996.
+
+Gilbert, Richard, and Carl Shapiro, (1997), " Antitrust Issues in the Licensing of Intellectual Property: The Nine No-No's Meet the Nineties," Brookings Papers on Economics: Microeconomics, pp. 283-336.
+
+Grindley, Peter, and David J. Teece (1997), "Managing Intellectual Capital: Licensing and CrossLicensing in Semiconductors and Electronics," California Management Review, vol. 39, no. 2, pp. 1-34.
+
+Hall, Bronwyn, and Rose Marie Ham, (1999), "The Patent Paradox Revisited: Firm Strategy and Patenting in the U.S. Semiconductor Industry."
+
+Heller, M.A. and R. S. Eisenberg, (1998), "Can Patents Deter Innovation? The Anticommons in Biomedical Research," Science, 280, 698-701.
+
+Katz, Michael and Carl Shapiro, (1985), "On the Licensing of Innovations," Rand Journal of Economics, Winter.
+
+Katz, Michael and Carl Shapiro, (1994), "Systems Competition and Network Effects," Journal of Economic Perspectives, vol. 8, no. 2, pp. 93-115.
+
+Klein, Joel I., (1997), “Cross-Licensing and Antitrust Law,” available at http://www.usdoj.gov/atr/public/speeches/1123.htm.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 29 of 32
+
+---
+
+Kortum, S. and J. Lerner, (1998), "Stronger Protection or Technological Revolution: What is Behind the Recent Surge in Patenting?"
+
+Lemley, Mark and David McGowan, (1998), "Legal Implications of Network Economic Effects," California Law Review, vol. 86, pp. 481-611.
+
+Merges, Robert P., (1999), "As Many as Six Impossible Patents Before Breakfast: Property Rights for Business Concepts and Patent System Reform," Berkeley Technology Law Journal, vol. 14, no. 2, pp. 577-615.
+
+Shapiro, Carl, (1989), "Theories of Oligopoly Behavior," in Handbook of Industrial Organization, R. Schmalensee and R. Willig, eds., Elsevier Science Publishers, pp. 330414.
+
+Shapiro, Carl, (1996a), "Antitrust in Network Industries," available at http://www.usdoj.gov/atr/public/speeches/shapir.mar.
+
+Shapiro, Carl, (1999), "Exclusivity in Network Industries," George Mason Law Review, vol. 7, pp. 673-684.
+
+Shapiro, Carl, and Hal R. Varian, (1998), Information Rules: A Strategic Guide to the Network Economy, Harvard Business School Press.
+
+U.S. Department of Justice and Federal Trade Commission, (1995) Antitrust Guidelines for the Licensing of Intellectual Property," April.
+
+U.S. Department of Justice and Federal Trade Commission, (1999) Antitrust Guidelines for Collaborations Among Competitors, ” (Draft), October.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 30 of 32
+
+---
+
+## Technical Appendix
+
+Here I show that prices can be well above monopoly levels if multiple firms have critical patents, all of which read on a single product. More precisely, if $N$ firms each control a patent that is essential for the production of a given product, and if these $N$ firms independently set their licensing fees, the resulting markup on that product is $N$ times the monopoly markup.
+
+Suppose that $N$ firms, $i=1,...N$ , each own a patent that is essential to the production of a given product. For simplicity, let us think of there being a competitive industry that produces this product, buying and assembling the necessary components from each of these $N$ firms. For this purpose we can think of firm $i$ either a setting a license fee for the use of its patent, or as setting a price at which it will sell its essential component to the competitive assembly industry; the theory is identical either way.
+
+The cost to firm $i$ per unit (for making and selling its component or for licensing its patent to assemblers) is denoted by $c_i$ . The price of component $i$ (or the license fee charged by firm $i$ ) is denoted by $p_i$ . The price of the product itself is denoted by $p$ . In addition to paying royalties (or buying components), the assembly firms incur an assembly cost per unit equal to $\alpha$ . Competition at the “assembly” level insures that $p = \alpha + \sum\limits_{i=1}^N p_i$ .
+
+Demand for the product in question is denoted by $D(p)$ . The (absolute value of the) elasticity of demand is given by $\varepsilon \equiv -\frac{D^{\prime}(p) p}{D(p)}$ . In general, $\varepsilon$ will vary with $p$ .
+
+I assume that the $N$ firms set their component prices, equivalently their license fees, independently and non-cooperatively. In other words, I look for the Nash Equilibrium in the prices $p_1,...,p_N$ . The profits for firm $i$ are given by
+
+$$\pi_{i}=D(p)\left(p_{i}-c_{i}\right) .$$
+
+The first-order condition for firm $i$ is given by
+
+$$\frac{d\pi_{i}}{dp_{i}}=D(p)+D^{\prime}(p)(p_{i}-c_{i})=0.$$
+
+Adding up across all $i$ gives
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 31 of 32
+
+---
+
+$$D(p) N+D^{\prime}(p) \sum_{i=1}^{N}\left(p_{i}-c_{i}\right)=0 .$$
+
+which can be re-written as
+
+$$\sum_{i=1}^{N} \frac{\left(p_{i}-c_{i}\right)}{p}=-\frac{D(p)}{p D^{\prime}(p)} N .$$
+
+Using the definition of the elasticity of demand, and the fact that $p=\alpha+\sum _{i=1}^{N}p_{i}$, we have
+
+$$\frac{p-\left(\alpha+\sum_{i=1}^{N} c_{i}\right)}{p}=\frac{N}{\varepsilon} .\quad (1)$$
+
+In other words, the percentage markup over cost for the product in question is equal to $N$ times the inverse of the elasticity of demand. In contrast, the standard monopoly markup rule would be
+
+$$\frac{p-\left(\alpha+\sum_{i=1}^{N} c_{i}\right)}{p}=\frac{1}{\varepsilon} .\quad (2)$$
+
+The markup with $N$ independent firms controlling key patents is equal to $N$ times the monopoly markup.
+
+It can be shown that the combined profits of the $N$ firms under independent pricing is lower than would be earned by a monopolist selling all $N$ components. This implies that the firms have an incentive to coordinate their pricing. A package license for all $N$ components would lead to higher (combined) profits and lower prices for consumers.
+
+Shapiro on Patent Thicket, NBER Conference on Innovation Policy and the Economy, Page 32 of 32
+

processed/00005_W4233445588/document.md

@@ -0,0 +1,495 @@
+Journal of Economic Perspectives-Volume 19, Number 2- Spring 2005-Pages 75-98
+
+# Probabilistic Patents
+
+Mark A. Lemley and Carl Shapiro
+
+For many years, economists typically conceptualized patents as well-defined property rights giving their owners either a monopoly over some market or at least a significant competitive advantage in that market due to control over a product improvement or a low-cost method of production (Nordhaus, 1969; Reinganum, 1989) . Once a patent was issued, this approach tended to assume that the patent was valid, that it granted a right of definite scope, and that users of the patented technology respected that right or were forced by courts to do so. Treating patents as well-defined rights to exclude rivals has permitted economists to focus on important and complex relationships among patents, innovation, competition and the diffusion of technology.
+
+More recently, however, scholars and policymakers have begun to look more closely at the empirical evidence regarding the issuance of patents and patent litigation. Nearly 200,000 patents are issued every year after a very limited examination process. Most issued patents turn out to have little or no commercial significance, which is one reason that only 1.5 percent of patents are ever litigated, and only 0.1 percent of patents are ever litigated to trial. Given these uncertainties, economists have increasingly recognized that a patent does not confer upon its owner the right to exclude but rather a right to try to exclude by asserting the patent in court (Shapiro, 2003a) . When a patent holder asserts its patent against an alleged infringer, the patent holder is rolling the dice. If the patent is found invalid, the property right will have evaporated.
+
+■ Mark A. Lemley is the William H. Neukom Professor of Law, Stanford Law School, Stanford, California, and of counsel, Keker & Van Nest LLP, San Francisco, California. Carl Shapiro is the Transamerica Professor of Business Strategy, Haas School of Business, and Professor of Economics, University of California at Berkeley, Berkeley, California.
+
+---
+
+76 Journal of Economic Perspectives
+
+The risk that a patent will be declared invalid is substantial. Roughly half of all litigated patents are found to be invalid, including some of great commercial significance. For example, Chiron's patent on monoclonal antibodies specific to breast cancer antigens was invalidated by a jury in 2002 in a suit in which Chiron had sought over $1 billion in damages from Genentech. A decision by the U.S. Court of Appeals for the Federal Circuit to invalidate an Eli Lilly patent on Prozac in 2000, less than two years before the patent was set to expire, caused Lilly's stock price to drop 31 percent in a day.
+
+Virtually all property rights contain some element of uncertainty. The owner of real estate may find that the title to that property is flawed; title insurance exists to deal with this risk. The (careless) owner of a trademark may find that its mark has been used so widely as to become a generic term, thus losing trademark protection. But the uncertainty associated with patents is especially striking, and indeed is fundamental to understanding the effects of patents on innovation and competition. There are two fundamental dimensions of uncertainty: 1) uncertainty about the commercial significance of the invention being patented, and 2) uncertainty about the validity and scope of the legal right being granted. Uncertainty about commercial significance is critical when studying the process by which patents are issued. Uncertainty about validity and scope are critical when studying the enforcement and litigation of patents.
+
+This article explores the economics of probabilistic patents. We begin with a brief description of the system by which patents are issued and litigated in the United States. We then discuss uncertainty regarding commercial significance and how patent applicants and would-be licensees hedge against that uncertainty. We turn next to proposals to reform the system by strengthening the examination process through which patents are granted. This literature emphasizes that expending the resources necessary to increase the certainty of issued patents across-theboard may not make economic sense given the vast number of patent applications and the small number that end up being commercially important. More effective are patent reforms that focus resources on patent applications that are likely to be of commercial significance.
+
+We then explore the incentives of patent holders and alleged infringers to settle their disputes rather than litigate them to completion. Indeed, virtually every patent licensing and cross-licensing agreement can be seen as the settlement of a patent dispute. However, the frequency or form of such private settlements may not serve the public interest, because litigating patent disputes to completion tends to generate positive externalities, by clarifying the limits of patent protection if the patent is upheld or encouraging wider use of the innovation if the patent is invalidated. While settlements of patent litigation between actual or potential rivals are normal and generally desirable, they also are agreements between competitors that can limit competition. In fact, a recent flurry of court cases have recognized the impact of uncertainty while scrutinizing the antitrust implications of patent settlements.
+
+---
+
+Mark A. Lemley and Carl Shapiro 77
+
+## Patent Prosecution and Patent Litigation in the United States
+
+Patenting is big business in the United States and throughout the world. $^1$ Inventors file over 350,000 patent applications a year with the U.S. Patent and Trademark Office (PTO), a number that has grown steadily (U.S. PTO Annual Report, 2003), and spend over $ 5 billion a year just on the process of obtaining those patents (Lemley, 2001). The PTO grants nearly 200,000 new patents a year, a number that has roughly doubled over the past 15 years (U.S. PTO Annual Report, 2003, Table 6).
+
+Patents are rewards for those who have contributed to economic growth through their inventions. Any resulting market power enjoyed by a patent holder is typically considered a social cost that is necessary to stimulate innovation and provide a return on R&D expenditures. When patents are granted covering technologies that were already known or were obvious, the resulting patents cause social costs without offsetting benefits. Therefore, the lower the quality of patents—that is, the greater is the fraction of patents improperly issued—the less efficient the patent system is at stimulating innovation. Complaints about patent quality are not new, but they have grown louder in recent years (Federal Trade Commission (FTC), 2003; National Academies of Science (NAS), 2004).
+
+### Patent Applications and Patent Prosecution
+
+The scope of a patent is defined by its “claims.” Examiners at the Patent and Trademark Office look at the claims in each patent application to see if the invention described in the claims meets the statutory requirements for patentability; the key requirements are that the invention be novel and not obvious. The PTO tests for novelty by looking for “prior art,” typically as described in other patents, in publications and in existing products. The applicant is required to disclose relevant prior art of which it is aware, but not to conduct a thorough search for prior art. The PTO also grants a patent only if it believes that the invention would not have been obvious “to a person of ordinary skill in the art” at the time the application was filed.
+
+The applicant is required to disclose its invention in the patent application, which will then be made public when the patent is issued (or 18 months after the application is filed if the applicant also seeks patents outside the United States, as most do). The disclosure must be sufficient so that those skilled in the art can make and use the invention without undue experimentation.
+
+Inventors regularly file patent applications without any clear idea of whether the invention will be a commercial success and in some instances whether the category of invention is even patentable at all. In part, this dynamic is driven by the strong incentives to file applications early. Inventors who commercialize a product,
+
+1 More detailed descriptions of the operation of the U.S. patent system are available from many sources. See, for example, National Academies of Science (2004, Appendix A, "A Patent Primer") or Schechter and Thomas (2003).
+
+---
+
+78 Journal of Economic Perspectives
+
+publish a paper or disclose an idea to the public have only a year in the United States to get a patent application on file (35 U.S.C. § 102(b)), or the innovation passes into the public domain. Those who wish protection abroad have even stronger incentives to file quickly, both because Europe lacks a one-year grace period and because if more than one inventor claims the same invention, every country but the United States will give the patent to the first person to file an application, rather than the first to invent . In theory, these policies encourage prompt filing of patents and increase the disclosure of ideas. They also help reduce the risk that firms will make investments in good faith and only find out later that they have infringed on the patent of another firm. $^2$ But the incentives for early filing also mean that many inventors file patent applications before they have any good sense of whether their inventions have commercial significance. For example, inventors in the fields of pharmaceuticals and biotechnology file applications on a large number of drugs or therapies before they know whether those drugs will be safe and effective.
+
+Moreover, the scope of patentable subject matter has been expanding over time (Gallini, 2002). The Federal Circuit Court of Appeals added software to the list of patentable inventions in the 1980s and 1990s, and “business methods” in 1998. $^3$ Because the scope of patentable subject matter can change over time, inventors may file patent applications even in areas that are not currently eligible for patent protection, simply to hedge their bets.
+
+Once an application is filed, the patent applicant negotiates with the patent examiner over the allowability and scope of the claims. The burden is on the Patent and Trademark Office to provide a reason not to issue a patent sought by an applicant. If the applicant is dissatisfied with the claims allowed by the patent examiner, the applicant can file a continuation application even after receiving a patent and thus continue to seek a patent with broader claims. The applicant can add new matter to the continuation application and claim the invention with additional, new elements, albeit with a later date of invention. Remarkably, there is no limit to the number of times the patentee can seek continuation of a patent application (Lemley and Moore, 2004) . Applicants are even allowed to amend their
+
+2 Prior to 1999, U.S. patent applications were not disclosed prior to the issuance of the patent. Some patent applications remained hidden from public view for years, even decades, while the PTO considered those applications. These so-called “submarine patents” allowed patent holders to engage in significant opportunism, in part by strategically slowing down the PTO process and in part by amending claims to capture new products introduced into the market well after the patent application was initially filed. Submarine patents are less of a concern since the law was changed in 1999 to require the disclosure of most U.S. patent applications 18 months after they are filed.
+
+5 Perhaps the best-known patent on a “business method” is Amazon’s 1999 patent no. 5,960,411. This “one-click” patent protects Amazon’s online ordering system, which allows consumers to make purchases with a single mouse click. Amazon generated a great deal of attention by asserting this patent against its most direct online rival, Barnes & Noble. For more background, see Lerner (2002) and his discussion of the case of State Street Bank & Trust Co. v. Signature Financial Group, Inc. (149 F.3d 1368 [1998]), in which the Federal Circuit explicitly rejected the notion that “business methods” could not be patented. The State Street case involved a software program that was used to value mutual funds.
+
+---
+
+Probabilistic Patents 79
+
+applications to capture products that are appearing in the market, so long as they (arguably) stay within the bounds of the invention described in the initial application, which can be broad and rather vague. Furthermore, applicants who are not content with the examiner's decision have appeal rights.
+
+While the examination process at the Patent and Trademark Office takes nearly three years on average (Allison and Lemley, 2000) , a patent examiner spends only 18 hours per application on average during those three years reading the application, searching for and reading prior art, comparing the prior art to the application, writing one or more provisional rejections, reviewing responses and amendments, often conducting an interview with the applicant's attorney and writing a notice of allowance (Lemley, 2001; FTC, 2003) . Even with this quick look, the PTO recently reported a backlog of more than 750,000 patent applications (U.S. PTO Annual Report, 2003, Table 5) . Further, legal scholars who have studied the patent prosecution process have pointed to structural problems that encourage the PTO to grant patents of doubtful quality, including high examiner turnover and an incentive system that rewards examiners for allowing but not for rejecting applications (Merges, 1999; Thomas, 2001) . As a result, the overwhelming majority of patent applications in the United States, perhaps 85 percent, ultimately result in an issued patent — far more than in Europe and Japan (Quillen, Webster and Eichman, 2003; NAS, 2004) . $^4$
+
+## Patent Litigation and Damages
+
+Patent lawsuits take place in the federal courts, usually in front of a jury. Appeals of patent decisions go to the Federal Circuit Court of Appeals, a specialized appeals court for patent cases established in the early 1980s. Many observers believe that the creation of the Federal Circuit has led to decisions more favorable to patent holders (Dreyfuss, 1989; Kortum and Lerner, 1999).
+
+The vast majority of patents are never asserted in litigation. Only 1.5 percent of all patents are ever litigated, and only 0.1 percent are litigated to trial (Lanjouw and Schankerman, 2001; Lemley, 2001) , though litigation rates vary by industry and reach as high as 6 percent in biotechnology (Lerner, 1995) . Great care must therefore be taken when interpreting data from any sample of litigated patent cases. The patents involved in litigation are those that are important enough commercially to justify the costs of litigation and for which the parties were unable to reach a mutually attractive settlement.
+
+4 One cannot simply divide the number of issued patents into the number of applications to obtain the grant rate, for several reasons. First, because the number of applications increases from year to year, and because it takes almost three years on average for the PTO to issue a patent, the proper comparison would be between the number of patents issued in a given year and the number of applications filed three years earlier. Second, over a quarter of all U.S. patent applications are so-called “continuation” applications, in which patentees return to the PTO in an effort to obtain the same patents that had been denied earlier (Graham and Mowery, 2002; Lemley and Moore, 2004) . To complicate matters further, continuations can sometimes result in multiple patents. Quillen, Webster and Eichman have done a careful study controlling for these variables, and they determine that the grant rate is 85 percent on the most plausible set of assumptions.
+
+---
+
+80 Journal of Economic Perspectives
+
+When patents are litigated, substantial uncertainty arises. Defendants in patent cases typically claim that the patent is invalid, usually based on the existence of prior art not found by the Patent and Trademark Office. However, patents are afforded a presumption of validity; to have a patent declared invalid requires “clear and convincing evidence.” Defendants also usually claim that they do not infringe the patent, even if it is valid. Of patents litigated to a final determination (appeal, trial, or summary judgment), 46 percent are held invalid (Allison and Lemley, 1998; see also Moore, 2000) . To some extent, this uncertainty results because disputes that are litigated to judgment are those for which the outcome is unclear, so that the parties differ significantly in their beliefs about their prospects for winning (Priest and Klein, 1984; Cooter and Rubinfeld, 1989) . $^5$ But that principle cannot explain all of the uncertainty or the variation in patentee win rates over time and by court (Allison and Lemley, 1998; Chien and Lemley, 2003) .
+
+In some cases, even when patents are held valid, they are found to be not infringed or are deemed unenforceable (Moore, 2000) . In some areas, particularly the process of determining the meaning of patent claims, the Federal Circuit Court of Appeals reverses district court judgments approximately one-third of the time (Chu, 2001; Moore, 2002, 2004) .
+
+A patent holder who wins an infringement suit can obtain an injunction preventing the infringing party from practicing the patent, which may force the infringing party to withdraw its products from the market. (In rare cases, the patent holder can also obtain a preliminary injunction, forcing the alleged infringer to cease using the patented technology during the patent litigation.) The infringing party might, however, be able to invent around the patent and stay in the market, albeit with higher costs or a less attractive product. A victorious patent holder also can seek damages from past infringement, either in lost profits or reasonable royalties. If the infringement is found to be “willful,” the infringing party may be forced to pay three times the actual damages.
+
+## Patents as Lottery Tickets
+
+Once issued, a patent remains in force until 20 years after the patent application was originally filed. To keep a patent in force, the patent holder must pay maintenance fees, ranging from several hundred to a few thousand dollars, at the end of the third, seventh and eleventh years. Between 55 and 67 percent of issued U.S. patents lapse for failure to pay maintenance fees before the end of their term (Moore, 2004; Lemley, 2001), which indicates that these patents are of little value to their owners. The distribution of value of patents appears to be highly skewed, with the top 1 percent of patents more than a thousand times as valuable as the median patent (Allison, Lemley, Moore and Trunkey, 2004; Pakes, 1986; Schanker-
+
+5 The magnitude of the stakes, asymmetry in the stakes and selection effects all play a role in determining observed litigation outcomes. Marco (2004) attempts to estimate these selection effects.
+
+---
+
+Mark A. Lemley and Carl Shapiro 81
+
+man and Pakes, 1986; Lanjouw and Schankerman, 1999) . Many patents are virtually worthless, either because they cover technology that is not commercially important, because they are impossible to enforce effectively, or because they are very unlikely to hold up if litigated and thus cannot be asserted effectively. A small number of patents are of enormous economic significance.
+
+Why do inventors file for many patents that turn out to have little or no value? Surely part of the reason is that patent applicants do not know which patents will be valuable and which will be worthless (Scherer, 2001; Denton and Heald, 2004) . But other explanations have been offered: a failure to understand the value of patents (Rivette and Kline, 2000) ; the use of patents to obtain financing and boost market valuation (Lemley, 2000; Hall, Jaffe and Trajtenberg, 2005) ; the use of patents as signaling mechanisms (Long, 2002) ; and the “defensive” use of patents to deter others from suing (Hall and Ziedonis, 2001; Lemley, 2001) . Even individually weak patents might have value as part of a large patent portfolio, because the portfolio can be licensed as a block or can serve to deter lawsuits (Parchomovsky and Wagner, 2004) .
+
+Many patent applications, and indeed patents themselves, are like lottery tickets. Inventors who are uncertain of the commercial significance of their ideas seek to patent many of them anyway, knowing that most of the resulting patents will turn out to be worthless, but hoping that a few will pay off big-time (Scherer, 2001). Just as people flock to stores to buy lottery tickets when the grand prize grows large, patent applicants have found ways to improve their chances of gaining patent protection in areas they consider promising. Two of the most common practices used by patentees to increase their chances of winning the patent lottery are continuations and a proliferation of closely related patents.
+
+Patent continuations, as discussed earlier, stem from the rather remarkable rule in U.S. patent law that patent applicants are free to try again and again, without limit, to persuade the Patent and Trademark Office to grant them a patent (Lemley and Moore, 2004) . They can even obtain a patent and then continue prosecution on a related application, as a hedge against the possibility that the market will change in a way that renders the first patent obsolete, or that the first patent is invalid based on prior art not cited to the PTO. Continuations are a large and growing part of patent practice, accounting for more than a quarter of all applications now filed (Graham and Mowery, 2004) . In some industries, notably biotechnology and pharmaceuticals, firms typically keep a continuation application pending during the entire lifetime of the original patent (Lemley and Moore, 2004) .
+
+Patent owners also improve their chance of winning the patent lottery by filing multiple patents on closely related technologies, thereby increasing the chance that their patents will cover technology that becomes widely adopted by market participants. In a number of key industries, particularly semiconductors (Hall and Ziedonis, 2001) and computer software (Bessen and Hunt, 2004) , companies file numerous patent applications on related components that are integrated into a single functional product. The result is a “patent thicket,” in which hundreds of
+
+---
+
+82 Journal of Economic Perspectives
+
+patents can apply to a single product (Shapiro, 2001; FTC, 2003). If the holder of a large patent portfolio asserts its patents against another company and claims that the other company is infringing dozens or even hundreds of its patents, the target company faces a very complex and costly undertaking if it chooses to fight all of those patent infringement claims in court, knowing that it has to win all or nearly all of the individual patent cases to avoid paying significant royalties or even being enjoined from selling its product (Parchomovsky and Wagner, 2004).
+
+Both continuation practices and patent thickets can have negative consequences on other firms in the market. A competitor who designs around an issued patent — a legal activity that patent policy actively encourages (Conigliaro, Greenberg and Lemley, 2001)— cannot know whether the patentee has a continuation application waiting in the wings with claims that can be drafted to cover the design-around. Indeed, some unscrupulous patentees intentionally delay the issuance of their patents to take other firms by surprise, increasing their royalty rates once companies operating in the industry have made irreversible investments (Graham and Mowery, 2004; Lemley and Moore, 2004) . Patent reformers have suggested limits on the use of continuations and courts have adopted doctrines designed to limit their abuse (FTC, 2003) .
+
+Similarly, patent thickets can have deleterious effects on both competition and innovation. One way to cut through the patent thicket is for incumbents with extensive patent portfolios to enter into broad cross-licenses (that is, exchanges of roughly symmetric patent positions) to “clear” the thicket. However, new entrants who lack large patent portfolios may be at a major disadvantage in this situation because they have no patents to trade. Without such cross-licenses, the result is inefficient “royalty stacking,” in which a manufacturer without its own patent portfolio must pay royalties to a number of separate companies. $^6$ Defensive patenting is a natural, even inevitable, strategy in industries with patent thickets, but defensive patenting itself can increase the density of the thicket.
+
+We do not mean to suggest that patent applicants invariably have little sense of the commercial significance of their inventions. To the contrary, applicants appear to have considerable private information at an early stage about the likely value of at least some of their patents. Allison, Lemley, Moore and Trunkey (2004) find that the most significant predictors of ultimate value observable to researchers are the industry, the number of prior art references, the number of claims in the patent and the time invested in prosecution of the patent. They conclude that patentees spend more time and energy along these dimensions when they believe a patent is
+
+6 This is a classic instance of the “Cournot complements” problem (a form of double marginalization), which is known to lead to inefficiently high prices that can even exceed the monopoly level (Cournot, 1838; Shapiro, 2001) . Heller and Eisenberg (1998) have termed a related problem the “tragedy of the anticommons.” Based on survey data, Walsh, Arora and Cohen (2003) question whether the anticommons problem has actually interfered with production in biomedical research, the area in which Heller and Eisenberg apply their theory. But among the ways they find the industry has avoided anticommons problems is to invalidate patents and to ignore them—approaches that do not deny that patents create problems in these industries.
+
+---
+
+Probabilistic Patents 83
+
+more valuable. They quote one general counsel at a software company as saying “of the 600 patents we file a year, we pretty much know which 20 we have to have and which 580 it would be nice to have.” For the remaining majority, the lottery effect comes into play.
+
+## Reforming the System of Granting Patents
+
+The patent system involves a quid pro quo . If you are the first to come up with a novel and nonobvious invention, and if you are prepared to disclose the workings of that invention to the public, in exchange you can receive exclusive rights to practice that invention for a limited time. However, when patents are improperly issued for rights that are not novel, or are “obvious,” the public suffers without justification by paying supracompetitive prices. There is widespread and growing concern that the Patent and Trademark Office issues far too many “questionable” patents that are unlikely to be found valid based on a thorough review of the sort one sees in patent litigation. Can the system can be designed to work better at reasonable cost?
+
+Any reform of the patent system must account for the fact that patent applicants typically have superior information to the Patent and Trademark Office about likely commercial significance. There are good reasons to doubt the efficiency of a system for granting patents under which 1) patents differ greatly in their commercial significance and value; 2) patent applicants are uncertain about the value of their ideas, but have far superior information to examiners; 3) patent applicants often have superior information as well about prior art, but are under no obligation to conduct a search for the relevant prior art; 4) patent applicants can persist repeatedly through the continuation process in seeking to have certain claims accepted by patent examiners; 5) the burden of proof falls upon the PTO to explain why a patent application will not be granted; and 6) patent examiners are faced with a flood of applications and have little time to devote to each one. These problems are likely to be most pronounced in areas where technology is changing rapidly. Thus, the system is skewed toward the grant of patents of dubious objective validity, based on a brief, inconclusive process, which are then potentially subject to later disputes with other firms in which legal fees can easily run into millions of dollars for both sides (American Intellectual Property Law Association, 2003). Uncertainty and asymmetric information are endemic in such a patent system.
+
+A number of scholars and policymakers have proposed reforms designed to reduce the number of improperly issued patents without causing genuine innovators to be denied patents. Two prominent recent examples come from the Federal Trade Commission (2003) and the National Academy of Sciences (2004). One common proposal is to hire more patent examiners and allow them to devote more time to the review of selected patent applications. As part of this effort, examiners could be encouraged to seek more information from patent applicants. Another common proposal is to establish a more effective opposition system in which
+
+---
+
+84 Journal of Economic Perspectives
+
+interested third parties could challenge the validity of an issued patent before an administrative patent board. Still another proposal is to raise the legal standard for nonobviousness, making it more difficult for applicants describing marginal improvements to obtain patent protection. We now consider some of these proposed reforms. $^7$
+
+As a starting point for thinking about patent reform, to what level of certainty should the system aspire when a patent is granted? It cannot be economically efficient to determine the validity of patents with anything approaching certainty during the application process. Indeed, Lemley (2001) has argued that the PTO is “rationally ignorant” of the actual validity of a patent, because extra resources devoted to determining the validity of a patent are largely wasted in the 95 percent of the cases in which the patent is neither litigated nor licensed for a royalty. However, others have argued that the benefits of avoiding some likely invalid patents are sufficiently great that society should spend the more money to weed out more bad patents (Gallini, 2002; Kesan and Ghosh, 2004) .
+
+Two key questions arise in considering how much effort society should put into examining patent applications. First, how effective would increased examination be at weeding out objectively bad patents without improperly denying patents for true innovations? Second, how great are the costs imposed on third parties when patents are issued that would not have been issued with more careful examination?
+
+Some improvements in the patent examination system, like better hiring and retention practices to improve the experience and qualifications of patent examiners and altering any institutional incentives that encourage examiners to grant doubtful patents, might improve the examination process at very low cost (Merges, 1999; Thomas, 2001) . However, other changes will require significant new expenditures. Devoting substantially more resources to patent examination is more likely to be efficient if those additional resources can be focused on the patents whose validity will turn out to be commercially significant. For example, following the findings of Allison, Lemley, Moore and Trunkey (2004) mentioned earlier, the patent examiners might focus greater attention on patent applications that have features correlated with greater ultimate value — like those with more claims and more prior art references. A simpler approach, which the PTO is already following to some degree, is to focus greater resources in areas of new or rapidly changing technology where the PTO has more difficulty identifying prior art, such as software and business methods. However, doing this may simply encourage patent applicants to avoid those fields (Allison and Hunter, 2004) .
+
+An alternative means of focusing attention on particular patents is to engage the incentives and information either of the party seeking the patent or parties opposed to the patent (Jaffe and Lerner, 2004) . For example, patent applicants could be required to conduct a full search for prior art. New procedures could be
+
+7 For a more complete analysis of the reforms proposed by the FTC (2003) and the NAS (2004), see the Summer 2004 Special Issue of the Berkeley Technology Law Journal devoted to these reforms, including Shapiro (2004).
+
+---
+
+Mark A. Lemley and Carl Shapiro 85
+
+established to encourage third parties to challenge a patent without entering into full-blown patent litigation. Europe and Japan already have an opposition procedure, whereby third parties can challenge issued patents in a streamlined manner. Work by Harhoff, Scherer and Vopel (2003) and others has found that the European patent opposition system is effective in identifying important patents. $^8$ However, any opposition system requiring the active participation by third parties to challenge patents is subject to a free-riding problem: any third party that challenges a patent will gain only a slice of the social benefit if a patent is overturned and thus will lack sufficient incentive to challenge that patent vigorously, even if the patent is highly questionable. This general problem applies as well to the existing system of patent litigation, as discussed below (Farrell and Merges, 2004).
+
+An alternative, more novel approach is to let patent applicants select either the normal, brief examination process, which would lead to a Standard Patent if the application were approved, or a more rigorous application process, which would lead to Super Patent if the application were approved. For such a system to work, the courts would have to give less weight to a Standard Patent than to a Super Patent. 9 Patent systems based on self-selection have the attractive feature that they do not require that the Patent and Trademark Office be able to determine which technologies are most likely to be commercially significant. Indeed, one can imagine a myriad of possibilities if one thinks of the process of issuing patents in terms of designing a mechanism that can issue a variety of property rights with different levels of strength based on the process and the level of resources devoted to different patent applications.
+
+## Patent Litigation Uncertainty and Reforms
+
+Even if the process for granting patents is improved, when a patent does enter litigation, considerable uncertainty will continue to exist about its validity and scope. The grounds for invalidating patents vary by industry, but objections related to prior art—obviousness, novelty and statutory bars—are the most common ground (Allison and Lemley, 1998). The meaning of patent claim terms—called “claim construction”—is hotly debated in virtually every patent case, and courts have found ambiguity even in such innocuous terms as “a,” “or,” “to” and “when.” Even once the meaning of the patent claims has been determined, the “doctrine of equivalents” can sometimes permit the patentee to expand its rights beyond the
+
+8 Under current U.S. law, re-examinations of patents that have been granted involving third-party participation have been permitted since 1999, but remain extremely rare. According to the FTC (2003, chapter 1, p. 27) report, the inter partes re-examination process had been used only four times in the first three and a half years of its operation. Apparently, challengers find the opposition process unattractive, in large part because the outcome of the opposition is binding in subsequent litigation (Janis, 2000).
+
+9 Doug Lichtman has suggested such an approach. Australia's system of "petty patents" shares some features with this proposal.
+
+---
+
+86 Journal of Economic Perspectives
+
+literal protection of the patent (Gallini, 2002) . A final source of uncertainty is the doctrine of “inequitable conduct,” in which patents may be rendered unenforceable if the patentee deceived or omitted to state information to the PTO during prosecution. While relatively few patents are held unenforceable for inequitable conduct (Moore, 2000) , allegations of unenforceability are ubiquitous, and the Federal Circuit has described the growth of such claims as “an absolute plague” ( Burlington v. Dayco , 849 F.2d 1418, 1421 [1988] ).
+
+Reform efforts focused on patent litigation have taken two very different approaches. Some reform efforts focus on reducing litigation uncertainty. For example, some reform proposals would combine a stronger process of patent examination with a stronger legal presumption of patent validity (Kesan and Banik, 2000) . Similarly, the Federal Circuit has been preoccupied over the last several years with placing limits on the “doctrine of equivalents” to restrict the ability of a patent holder to expand the coverage of the patent. A number of commentators have also proposed eliminating or limiting the doctrine of “willful infringement” under which an infringer must pay enhanced damages if it intentionally infringed a patent (FTC, 2003; NAS, 2004; Lemley and Tangri, 2003; Powers and Carlson, 2001) . Under current law, “willful infringement” occurs only when an infringer is aware of the patent and believes the patent is valid and believes that its conduct infringes. The law has developed a complex set of rules requiring the company to obtain an attorney's opinion as to the proper scope and validity of the patent; reliance in good faith on that opinion will insulate a defendant from liability for willfulness. Reform proposals suggest that given the real uncertainty as to scope and validity of patents, the doctrine of willful infringement should be abolished or at least modified to put it on a more objective footing.
+
+Other reform proposals push in the opposite direction—acknowledging that the scope and validity of patent rights are uncertain and ensuring that the law accurately reflects that uncertainty. Thus, the FTC (2003), Lemley (2001) and others have proposed legislation or judicial action to lower the burden of proof that a challenger must meet to invalidate a patent. Currently, a challenger must establish invalidity based on “clear and convincing evidence.” The FTC recommends a lower standard based on “the preponderance of the evidence.” Although it does not support the FTC's recommendation, even the American Intellectual Property Law Association (2004) has proposed scaling back the application of the presumption of validity through judicial interpretation. Reducing the presumption of validity could also work in tandem with the Super Patents idea— Super Patents would get a strong presumption of validity, but ordinary patents would not. $^10$
+
+10 A few scholars have even suggested increasing uncertainty for patents whose validity is not in doubt. Ayres and Klemperer (1999) observe that the ratio of marginal profits to deadweight loss grows large as price approaches the monopoly price, and they argue that uncertain or probabilistic patent rights can therefore confer most of the private benefits of a certain right to exclude at a fraction of the social costs. They illustrate their general point using a model in which a) the patent holder would not have the right
+
+---
+
+Probabilistic Patents 87
+
+In deciding among reform proposals, it is worth remembering that reducing litigation uncertainty is not, in and of itself, a goal in designing the patent system. Some uncertainty is an inevitable part of any system involving litigation. Furthermore, litigation over patent validity could be eliminated entirely simply by making the PTO's validity determinations final — but few would advocate such a course. In studying the uncertainty surrounding the patent system, we ultimately are interested not in that uncertainty per se , but rather in the effects of the patent system, and its uncertainty, on innovation, inventors, competition and consumers.
+
+## Private Incentives to Challenge Patents
+
+If court challenges to commercially significant but questionable patents were fast and cheap, then improperly issued patents might have little market impact, because they would quickly be challenged and overturned. Unfortunately, the patent litigation process does not work in anything approaching this idealized fashion.
+
+The main problem with the litigation system can be demonstrated with an example. Suppose that widgets are supplied in a competitive market consisting of ten identical firms, each with constant marginal cost of $ 40 per unit. For simplicity, suppose that demand for widgets is linear, given by $P = 100 - Q$ , where $P$ is price and $Q$ is quantity. The resulting competitive price of widgets is $ 40, and 60 widgets are sold. Now suppose that a new method of producing widgets is developed which lowers the production cost to $ 30 per unit. If this method is freely available to all producers, the price of widgets will fall to $ 30 per unit, and quantity will rise to 70. Suppose, however, that a patent is issued covering this new method of production. Again for simplicity, suppose that the patent is held by a firm that does not produce widgets. The owner of this patent selects a royalty rate, $R$ , in dollars per widget, at which it will license its patent to all widget manufacturers.
+
+Begin with the case in which the patent is unquestionably valid. In this case, widget manufacturers have a choice to make: continue producing at $40 per unit or take a license, produce at $30 per unit using the new and superior method of production, but pay royalties. From the perspective of the patent holder, setting the optimal royalty rate is a problem of monopoly pricing. No one will pay a royalty greater than $10 per unit, since that is the cost savings associated with the patented process. Indeed, the monopoly outcome is for the patent holder to choose a royalty
+
+to stop a rival from infringing, but rather could only seek compensatory damages after the patent expired, and b) even these damages would only be awarded with some probability. While we agree that conventional patent policy is “inefficient at the margin” in the sense they describe, it is not clear to us how their proposal could plausibly be made operational.
+
+---
+
+88 Journal of Economic Perspectives
+
+$R^* = 10$ . $^11$ The patent holder receives royalty payments of $10 per unit for 60 units, appropriating the entire cost savings associated with its invention. There is a deadweight loss of $50 resulting from the fact that only 60 units are produced, not 70, which would be socially optimal given the actual (social) cost of production of $30 per widget. $^12$
+
+Now change the story to reflect probabilistic real-world patents. Suppose that the ten widget producers are aghast that this patent was issued and insist that the patented method was obvious to someone skilled in art at the time of the patent application. Indeed, we may imagine that, following the publication of some basic research in the public domain, the widget producers soon learned how to apply those research findings to their production methods and thus lower their costs by $ 10 per unit. Perhaps they even made significant, technology-specific capital investments. Meanwhile, unknown to them, someone outside the industry had filed for a patent covering this technology and convinced the PTO that its application met the novelty and nonobviousness tests. For simplicity, let us suppose that the patent at issue is highly questionable: there is only a 20 percent chance that the patent would be found valid if tested in court. However, litigation involves some cost, $ C. Will any of the widget manufacturers challenge the patent?
+
+If any individual widget maker considers challenging the patent, it must consider two outcomes. If the patent is upheld, then the firm has spent $ C and gained nothing. If the patent is overturned, then all firms will be relieved of any royalty obligations, and the price of widgets will fall to $ 30. $^13$ Invalidating the patent is a public good that benefits consumers of widgets, but not any one widget manufacturer or even widget manufacturers collectively. Thus, no individual widget maker can recover the $ C litigation costs. Accepting the $ 10 royalty is a dominant strategy for each widget manufacturer in this setting. In the end, consumers end up paying $ 10 per widget to the patent holder even if the patent should never have been issued. Furthermore, the prospect of the prize of $ 600 in royalties to the patent holder will encourage rent-seeking behavior by patent applicants.
+
+What is driving this striking result that even a weak patent can command royalties approaching those of an ironclad patent covering the same claims? The key insight is that invalidating a patent generates significant positive externalities, and activities that generate positive externalities are undersupplied. There are very strong reasons to believe that challenges to patents are undersupplied (Gilbert, 2004; Farrell and Merges, 2004) . In practice, this means that companies accused of infringing will tend to settle patent disputes—for example, by paying royalties—
+
+$^{11}$ For a royalty rate $R$ between zero and ten, the price of widgets will be $30+R$, and the quantity produced will be $70-R$. The patent owner's licensing revenues will be $R(70-R)$. This expression is strictly increasing in $R$ for $R$ between zero and ten.
+
+12 The $50 represents ten units not produced, with an average foregone social surplus of $5 per unit. Each unit would have cost $30 to produce, and their average value to consumers would have been $35.
+
+15 The key case is the Supreme Court' s decision in Blonder-Tongue Lab v. University of Illinois Found (402 U.S. 313 [1971]), under which an alleged infringer can prevent an infringement suit if the patent claim asserted against it has been declared invalid in another case.
+
+---
+
+Mark A. Lemley and Carl Shapiro 89
+
+rather than litigating. It may also lead companies to pool their patents inefficiently and share royalties rather than engage in patent litigation (Gilbert, 2004; Choi, 2003) . In a more general setting, Farrell and Shapiro (2005) show that the royalties commanded by the owner of a probabilistic patent can easily be disproportionate to the strength of the patent and are highest when a single licensee's profits are very sensitive to its own costs but not sensitive to the level of industry-wide costs.
+
+We do not mean to suggest that our simple example provides anything approaching a complete analysis of this problem, nor that the royalties commanded by a patent are generally unrelated to the strength of that patent. In the example just given, the widget manufacturers had no incentive at all to challenge patents because their position is symmetrical and competition is perfect. In the real world, participants in most industries have rents stemming from imperfections in competition, specific capital investments, oligopoly, product differentiation and brand value and nonconstant marginal cost curves. As a result, most accused infringers will have some incentive to challenge the validity of a patent, but that incentive will be suboptimal. Whether an accused infringer has sufficient incentive to mount an effective challenge will depend on a number of factors: how significant are those rents; how readily they will be dissipated to competitors; how much money is at stake based on past infringement; and what is the relationship between litigation expenditures and success at trial? A company sued for $ 1 billion in royalties will likely have the incentive to pay $ 5 or $ 10 million in legal fees even if competitors will also benefit substantially; a company sued for $ 10 million in royalties may well not have strong incentives to defend the suit. In short, the simple example just presented is a polar case. But more general analyses show that serious problems arise in relying on private parties to challenge questionable patents.
+
+Just starting from the example sketched above, a number of questions spring to mind. How would the analysis change if the widget manufacturers produced differentiated products? How would the analysis change if the widget manufacturers had made specific capital investments and produced subject to increasing marginal cost, at least in the short run? How would the outcome change if existing manufacturers could coordinate their decisions to challenge the patent (while still being prohibited from colluding on price)? How would the analysis change if only a few widget makers existed and they acted as oligopolists? Can downstream consumers, the ultimate beneficiaries of a successful patent challenge in this model, effectively band together to fight the patent? On the other hand, could the patent holder fight back by credibly threatening to charge a higher royalty to anyone challenging its patent than to those who agree to pay royalties without a fight? Farrell and Merges (2004) offer insightful further discussion of these issues. They emphasize two basic reasons why individual firms accused of patent infringement have suboptimal incentives to challenge the patents asserted against them: 1) the public good problem — the fact that rivals to the allegedly infringing firm will benefit from a finding that the patent is invalid (or that its claims should be read narrowly); and 2) the pass-through problem — the fact that higher uniform royalty costs are passed through in the form of higher prices, thus muting the incentives
+
+---
+
+90 Journal of Economic Perspectives
+
+of alleged infringers to avoid paying such uniform royalties. $^14$ Farrell and Shapiro (2005) ask more generally how holders of weak patents will structure their licensing agreements to induce licensees to accept licenses rather than to challenge their patents.
+
+Since invalidating a patent provides a public good, typically to the benefit of competitors and consumers, one can naturally consider policies to overcome this public-good problem. One standard approach for dealing with public goods is to reward or subsidize those who contribute to the public good. For example, one might subsidize those who successfully challenge patents by instituting a bounty system (Thomas, 2001; Miller, 2004) . An alternative reward would be to give certain exclusive rights to the party who successfully challenges a patent. The HatchWaxman Act, which grants a limited period of exclusivity to the first generic supplier to challenge a pharmaceutical patent, has this flavor. These incentives are counterbalanced by the fact that a company that initiates a post-grant opposition signals to the patent holder that it may be infringing the patent. In the words of industry participants, a firm that initiates a post-grant opposition effectively “ paints a big target on its back. ”
+
+A second standard approach is for the government to supply the public good. The government can and does challenge some issued patents when the PTO re-examines a patent, perhaps in response to third party complaints or information. However, such government-led re-examinations are currently rare. One way around this problem would be to empower a government agency to challenge patents based on information provided by interested industry participants, even ones who remain anonymous. The Federal Trade Commission, with its consumer-oriented mission, might be suitable for this task, perhaps working in conjunction with other agencies that could contribute their technical expertise in different areas of science and technology. Another approach is to encourage public interest organizations to challenge suspect patents. Two such organizations, the Electronic Frontier Foundation and the Public Patent Foundation, have begun to file administrative challenges to patents.
+
+A final approach is to impose restrictions on the agreements that litigants can reach to settle patent cases, in order to prevent agreements that harm competition. We discuss such restrictions in the next section.
+
+14 Farrell and Merges (2004) also assert that the outcome of patent litigation tends to be tilted toward the party spending more on litigation, which tends to be the party with the most at stake. Therefore, if the patent holder has much more at stake than does any individual alleged infringer, even if litigation occurs, the outcome may be tilted in favor of the patent holder. While this may well be true as a general matter, we strongly suspect that expenditures on litigation are subject to diminishing returns, so such differences may be of little significance in high-stakes cases. A firm spending $ 1 million litigating a patent case will likely do much better than a firm that spends only $ 100,000, but it does not follow that a firm spending $ 20 million will do much better than a firm spending $ 15 million.
+
+---
+
+Probabilistic Patents 91
+
+## Antitrust Limits on Patent Settlements
+
+Of the 1.5 percent of all patents that are litigated, some 95 percent of the cases end in settlements rather than verdicts. This figure does not count all of the settlements that happen without suit ever being filed. The prevalence of actual and potential competitors entering into patent settlement agreements that restrict competition raises important questions about possible antitrust violations.
+
+Some antitrust limits on the settlements of patent disputes between rivals are unquestionably needed. Consider an incumbent monopolist who faces the threat of entry from a single potential entrant. The monopolist owns a patent that it is asserting against the potential entrant. For simplicity, assume that both parties agree that if fully litigated, the patent will be found valid with probability P. Suppose that the patent at issue is valuable only in this market, and that no other potential entrants exist, so no public-goods problem arises with invalidating this patent. However, consumers are affected by the presence or absence of competition, so externalities remain associated with the decision to litigate or settle and the terms on which a settlement occurs. There is no reason to assume that bargaining between the monopolist and the potential entrant to maximize their joint profits will lead to a socially optimal settlement.
+
+Indeed, the incumbent monopolist and the potential entrant will quite probably achieve an anticompetitive settlement, at least in the absence of antitrust rules limiting the manner in which they can resolve their dispute. As long as monopoly profits are greater than joint duopoly profits, the monopolist and the entrant will have an incentive to negotiate in a way that leads to the monopoly level of output and the monopoly price. In comparison with litigation, such a settlement would deprive consumers of the competition that would arise if the patent were declared invalid, which would occur with probability $1-P$ . For a relatively weak patent, consumers can be significantly harmed by an agreement between the incumbent monopolist and the potential entrant that maximizes their joint profits. In the limiting case as $P$ approaches zero, a weak patent can be used as a fig leaf to cover an agreement not to compete.
+
+The expected joint profits of the incumbent monopolist and potential entrant are higher from a clever settlement agreement than from litigation. Expected profits from settlement depend upon the terms of the settlement. But consider an easy way for the parties to settle and achieve full monopoly profits: the incumbent can pay the potential entrant not to enter the market. 15 Intuitively, settlement leads to higher joint profits for two reasons: settlement eliminates the chance that
+
+$^{15}$ How big a payment is needed? By litigating the entrant, $E$ can earn expected profits of $(1-P) \times \pi_E$ $-C_D$ where $\pi_E$ represents $E$ 's profits and $C_E$ represents $E$ 's litigation costs. The incumbent $M$ can induce $E$ to agree not to enter the market by paying $E$ an amount $F > (1-P) \times \pi_E - C_E$ . Is there a mutually agreeable level of $F$ ? Yes. Including the payment of $F$ , $M$ earns $\pi_M - F$ , where $\pi_M$ represents the incumbent's profits if there is no entry. In contrast, by litigating, $M$ would earn expected profits of $P$ * $\pi_M + (1-P) \times \pi_1 - C_M$ , where $\pi_1$ represents the incumbent's profits if entry occurs. Settling is superior to litigating for $M$ if and only if $\pi_M - F > P \times \pi_M + (1-P) \times \pi_J - C_M$ , which can be written as $F < (\pi_M$
+
+---
+
+92 Journal of Economic Perspectives
+
+profit-dissipating competition will break out if the patent is proven invalid, and it avoids litigation costs. $^16$ A particularly corrosive form of settlement occurs when the litigants collude to ask the court to vacate a decision it has already rendered, often one holding a patent invalid or construing its scope narrowly. A published court decision on the validity or scope of a patent is a public good, and courts should refuse to vacate their opinions upon settlement. Nonetheless, judges sometimes do vacate earlier decisions at the request of the litigating parties.
+
+Of course, it is blatantly illegal for a monopolist to pay its sole potential competitor to stay off the market. Nor could such a payment be hidden in the form of an acquisition: under U.S. antitrust law, a monopolist is not allowed to acquire its sole potential entrant. But is this same payment anticompetitive in the context of a patent settlement? Does the answer depend upon the strength of the patent?
+
+The courts have been grappling with this issue, and more generally the antitrust limits on patent settlements, in a series of cases over the past five years (Hovenkamp, Janis and Lemley, 2003, 2004b; Cotter, 2003) . Many of these cases have arisen from settlements between incumbent pharmaceutical manufacturers and potential generic competitors who they have accused of patent infringement, largely because the Hatch-Waxman legislation regulates entry into the market and therefore gives a pharmaceutical patentee who settles with a generic some power to keep out all competitors, not just one (FTC, 2002; Morse, 2002) .
+
+The courts have devoted most of their attention to settlements involving payments from incumbents to would-be generic suppliers, known as “ reverse payments ” because they flow from the patent holder to the challenger, in contrast to conventional licensing payments that challengers make to patent holders. Courts have come to different conclusions on the legality of such reverse or “ exclusion payments. ” For example, Abbott, the maker of Hytrin, a very successful drug used to treat hypertension and enlarged prostate, was faced with potential generic entry by Geneva, which Abbott accused of infringing its patent. Under their 1998 settlement, Abbott agreed to pay Geneva $ 4.5 million per month for some period of time in exchange for which Geneva agreed not to enter the market. The Eleventh Circuit found that this agreement was not per se illegal and instructed the
+
+$-\pi _{j})\times (1-P)+C_{M}$. There are mutually attractive levels of F if and only if $(1-P)^{*}\pi _{E}-C_{E}<(\pi _{M}-\pi _{j})\times (1-P)+C_{M}$. Rearranging, this is equivalent to
+
+$$\left(\pi_{M}-\pi_{f}-\pi_{E}\right) \times(1-P)+C_{M}+C_{E}>0 .$$
+
+Assuming $\pi_M > \pi_I + \pi_E$ i.e., that competition dissipates total profits, this inequality must be satisfied. $^16$ These dangers are not limited to the case of a single potential entrant. With multiple potential entrants, the firms still benefit as a group if they can agree to licensing terms that support the monopoly outcome. However, the presence of multiple potential licensees requires much more complex multiparty negotiations. We do not mean to suggest that the monopoly outcome necessarily can be achieved in all cases. Among other things, it requires significant barriers to entry in order to work. See Farrell and Shapiro (2005) for further analysis along these lines.
+
+---
+
+Mark A. Lemley and Carl Shapiro 93
+
+District Court to reconsider the case. $^17$ A variant of this pattern arose in a case involving Schering-Plough, the maker of the prescription drug K-Dur 20, which is used to treat low potassium. Schering-Plough entered into an agreement with Upsher-Smith, a potential generic competitor, which involved a payment from Schering-Plough to Upsher-Smith and an agreement by Upsher-Smith not to enter the market before a specified date. The Federal Trade Commission has found this agreement to be anticompetitive and held that agreements of this sort are presumptively unlawful. The Eleventh Circuit reversed. The courts have yet to establish a clear and uniform approach to these cases. $^18$
+
+Previous work by the authors and others has argued that patent settlements involving payments in excess of avoided litigation costs by incumbent patent holders to potential entrants accused of infringing should be presumed to be anticompetitive (Hovenkamp, Janis and Lemley, 2003; Shapiro, 2003a, 2003b; Cotter, 2003, 2004; Leffler and Leffler, 2002, 2003) . The likely effect of such payments is to delay entry, either in comparison with the outcome of litigation over the patent or in comparison with a settlement not involving these payments. $^19$ Neither competition nor innovation is promoted by allowing the owner of a weak patent to pay would-be challengers to refrain entirely from competing during the lifetime of the patent.
+
+Defenders of such agreements, and some courts, have argued that the antitrust laws should not be used to weaken patent protection, and that the patent at issue might well not be invalid (Schildkraut, 2004; Crane, 2004) . We agree that a patent holder does not violate the antitrust laws by excluding rivals who have been proven to infringe a valid patent; the essence of a patent is a right to exclude others from practicing the patented invention. But a patent does not give its owner the right to exclude rivals who are allegedly infringing, at least not without a court order. Payments from patent holders to alleged infringers in exchange for their agreement to stay off the market therefore go beyond the patent grant and exclude allegedly infringing competition, to the detriment of consumers (Shapiro, 2003b) . If a
+
+17 Valley Drug Co. v. Geneva Pharmaceuticals, Inc., 344 F. 3 rd 1294 (11th Cir. 2003). On remand, the district court once again found the agreement illegal. Shapiro has served as an expert witness for Kaiser, a purchaser of Hytrin who claims to be injured by Abbott's agreement with Geneva.
+
+18 The decision of the Federal Trade Commission in this case, In the Matter of Schering-Plough Corporation, et al., is available at <http://www.ftc.gov>. In contrast to the Abbott-Geneva agreement just noted, the Sixth Circuit found a similar agreement between Hoescht Marion Roussel and Andrx regarding the drug Cardizem CD to be per se illegal (In re Cardizem Antitrust Litigation, 3322 F. 3 rd 896, Sixth Circuit, 2003). In yet another case, the District Court ruled that a similar agreement involving Cipro was not per se illegal because the strength of the patent must be considered as part of the antitrust analysis (In re Ciprofloxacin Hydrochloride Antitrust Litigation, 261 F. Supp. 2 nd 188, E.D.N.Y. 2003). This court later found the Cipro agreement to be lawful.
+
+19 Willig and Bigelow (2004) argue that in cases involving negotiated entry dates, such reverse payments can be procompetitive. They thus oppose a per se rule prohibiting such payments. Under the standard we favor, such payments would be presumptively anticompetitive, giving the settling parties the chance to demonstrate in a given case that justifications such as those offered by Willig and Bigelow — based on risk aversion, imperfect capital markets and asymmetric information — apply with sufficient force to overcome the presumption.
+
+---
+
+94 Journal of Economic Perspectives
+
+patent holder believes that it properly has the right to exclude an alleged infringer, the patent holder can seek a preliminary injunction forcing the alleged infringer from the market. Such preliminary injunctions may be granted if the patent holder is likely to prevail and would suffer irreparable harm from the allegedly infringing competition, perhaps because the alleged infringer would not be able fully to compensate the patent owner for damages if the patent is later found to be valid and infringed. Indeed, in the pharmaceutical context, in which every reverse payment case so far has arisen, the law provides for an automatic preliminary injunction for 30 months; reverse payments are an effort to extend the period of exclusion beyond that point without having to litigate the patent. Importantly, courts will deny a preliminary injunction if serious questions exist about the validity of the patent, so the patentee's decision to settle with a reverse payment rather than to seek an injunction provides some evidence about the perceived strength of the patent. In addition, we show in our previous articles that one often can infer a certain degree of patent weakness from the fact that the patent owner is paying to avoid the risk that its patent will be found invalid, especially if the payment is large. Therefore, large reverse payments are inconsistent with a claim by the patent holder that its patent very likely would be found valid if litigated.
+
+Patent settlements take many forms and can raise a variety of antitrust issues even when they do not involve reverse payments from patent holders to alleged infringers to stay off the market. Virtually every licensing agreement can be seen as the settlement of a potential patent dispute. In other cases, patent disputes are settled through mergers or acquisitions or by forming patent pools or engaging in cross-licensing. Some pharmaceutical companies have settled with would-be generic suppliers by negotiating a date at which the generic supplier could enter the market; Hovenkamp, Janis and Lemley (2003) have argued that such agreements are reasonable so long as they do not include a reverse payment. Patent applicants also settle interference disputes in which each claims to be the first to have come up with an invention; such settlements will typically involve a payment from one side to the other because the side conceding priority is giving up potential ownership of the patent (Hovenkamp, Janis and Lemley, 2004a) .
+
+Most settlements are quite reasonable in competitive terms, and thus we face the complex question of identifying those few settlements that are anticompetitive. Shapiro (2003a) proposed a framework for establishing the antitrust limits on patent settlements based on the following principle: the settlement cannot lead to lower expected consumer surplus than would arise from ongoing litigation. This rule is similar in character to existing antitrust rules relating to mergers and to licensing agreements, as reflected in the Horizontal Merger Guidelines and the Antitrust Guidelines for the Licensing of Intellectual Property issued by the Federal Trade Commission and the U.S. Department of Justice. The rule proposed and analyzed by Shapiro also respects the rights of patent holders, while preventing companies from using the cover of patent settlements to engage in cartel-like agreements. For better or worse, applying this rule typically requires some assessment of the strength of the relevant patents, either directly or by inference. In the “ reverse payment ”
+
+---
+
+Probabilistic Patents 95
+
+cases described above, the presence of a substantial payment from patentee to accused infringer may well imply that the patentee paid for a reduction in competition in comparison with ongoing litigation. In contrast, this same framework implies that simple licensing agreements cannot be presumed anticompetitive: royalties will reflect the underlying strength of the patent, since no licensee will burden itself with high royalties unless the patent is indeed likely to be found valid and infringed. Shapiro (2003a) studies a variety of other, more complex types of settlements, including mergers and patent pools as well as negotiated entry dates.
+
+## Conclusion
+
+The patent system does not grant an absolute right to inventors to exclude others from practicing their inventions, as many economic models assume. Rather, the patent system gives the patent holder a right to try to exclude others by asserting its patent against them in court. The actual scope of a patent right, and even whether the right will withstand litigation at all, are uncertain and contingent questions. This uncertainty is not an accident or mistake. Rather, it is an inherent part of our patent system, an accommodation to the hundreds of thousands of patent applications filed each year, the inability of third parties to participate effectively in determining whether a patent should issue, and the fact that for the vast majority of issued patents, scope and validity are of little or no commercial significance. Modeling patents as probabilistic rights requires us to rethink reform of the patent granting process, patent opposition procedures, our approach to patent litigation, the efficacy of litigation as a means of invalidating patents that were improperly issued and antitrust policy toward the settlement of patent lawsuits.
+
+- ■ We thank Rochelle Dreyfuss, Joe Farrell, Richard Gilbert, Rose Hagan, James Hines, Paul
+Klemperer, Josh Lerner, Kevin Outterson, Timothy Taylor and Michael Waldman for their
+input, including comments on an earlier draft.
+## References
+
+Allison, John R. and Starling Hunter. 2004. “On Experimental Patent Reform and the Failure to Fix a Non-Problem.” Working paper.
+
+Allison, John R. and Mark A. Lemley. 1998. "Empirical Analysis of the Validity of Litigated Patents." American Intellectual Property Law Association Quarterly Journal. 26:3, pp. 185-275.
+
+Allison, John R. and Mark A. Lemley. 2000.
+
+"Who's Patenting What? An Empirical Exploration of Patent Prosecution." Vanderbilt Law Review. 53:6, pp. 2099-174.
+
+Allison, John R., Mark A. Lemley, Kimberly A. Moore and R. Derek Trunkey. 2004. "Valuable Patents." Georgetown Law Journal. 92:3, pp. 435-80.
+
+American Intellectual Property Law Association. 2003. Report of Annual Economic Sur-
+
+---
+
+96 Journal of Economic Perspectives
+
+vey. Arlington, Va.: American Intellectual Property Law Association.
+
+American Intellectual Property Law Association. 2004. Response to the October 2003 FTC Report; Available at <http://www.aipla.org>.
+
+Ayres, Ian and Paul Klemperer. 1999. "Limiting Patentees' Market Power Without Reducing Innovation Incentives: The Perverse Benefits of Uncertainty and Non-Injunctive Remedies." Michigan Law Review. 97:4, pp. 985-1033.
+
+Bessen, James and Robert M. Hunt. 2004. "An Empirical Look at Software Patents." Working Paper No. 03-17; Available at <http://www. researchoninnovation.org/swpat.pdf>.
+
+Chien, Colleen V. and Mark A. Lemley. 2003. "Are the U.S. Patent Priority Rules Really Necessary?" Hastings Law Journal. 54:5, pp. 1299-333.
+
+Choi, Jay Pil. 2003. "Patent Pools and CrossLicensing in the Shadow of Patent Litigation." CESifo Working Paper No. 1070.
+
+Chu, Christian A. 2001. "Empirical Analysis of the Federal Circuit's Claim Construction Trends." Berkeley Technology Law Journal. 16:3, pp. 1075-164.
+
+Conigliaro, Matthew, Andrew Greenberg and Mark A. Lemley. 2001. "Foreseeability in Patent Law." Berkeley Technology Law Journal. 16:3, pp. 1045-074.
+
+Cooter, Robert and Daniel Rubinfeld. 1989. "Economic Analysis of Legal Disputes and their Resolution." Journal of Economic Literature. 27:4, pp. 1067-097.
+
+Cotter, Thomas F. 2003. "Refining the ‘Presumptive Illegality’ Approach to Settlements of Patent Disputes Involving Reverse Payments." Minnesota Law Review. 87:6, pp. 1789-816.
+
+Cotter, Thomas F. 2004. "Antitrust Implications of Patent Settlements Involving Reverse Payments: Defending a Rebuttable Presumption of Illegality in Light of Some Recent Scholarship." Antitrust Law Journal. 71:3, pp. 1069-097.
+
+Cournot, A. A. 1838. Researches into the Mathematical Principles of the Theory of Wealth. English Edition, New York: Kelley.
+
+Crane, Daniel A. 2004. "Ease Over Accuracy in Assessing Patent Settlements." Minnesota Law Review. 88:3, pp. 698-711.
+
+Denton, F. Russell and Paul J. Heald. 2004. “Random Walks, Non-Cooperative Games and the Complex Mathematics of Patent Pricing.” Rutgers Law Review. 55:4, pp. 1175–288.
+
+Dreyfuss, Rochelle Cooper. 1989. "The Federal Circuit: A Case Study in Specialized Courts." New York University Law Review. 64:1, pp. 1-74.
+
+Farrell, Joseph and Robert P. Merges. 2004. “Incentives to Challenge and Defend Patents: Why Litigation Won't Reliably Fix Patent Office
+
+Errors and Why Administrative Patent Review Might Help." Berkeley Technology Law Journal. 19:3, pp. 943-70.
+
+Farrell, Joseph and Carl Shapiro. 2005. "How Strong Are Weak Patents?" Working paper, University of California at Berkeley.
+
+Federal Trade Commission (FTC). 2002. “Generic Drug Entry Prior to Patent Expiration: An FTC Study.” Available at <http://www.ftc.gov/ os/2002/07/genericdrugstudy.pdf>.
+
+Federal Trade Commission (FTC). 2003. "To Promote Innovation: A Proper Balance of Competition and Patent Law and Policy." Available at <http://www.ftc.gov/opa/2003/10/ cpreport.htm>.
+
+Gallini, Nancy T. 2002. "The Economics of Patents: Lessons From Recent U.S. Patent Reform." Journal of Economic Perspectives. 16:2, pp. 131-54.
+
+Gilbert, Richard J. 2004. "Antitrust for Patent Pools: A Century of Policy Evolution." Stanford Technology Law Review. April; Available at <http://stlr.stanford.edu/STLR/Core_Page/>.
+
+Graham, Stuart and David Mowery. 2004. “Submarines in Software? Continuations in US Software Patenting in the 1980s and 1990s.” Economics of Innovation and New Technology. 13:5, pp. 443–56.
+
+Hall, Bronwyn H. and Rosemarie Ziedonis. 2001. "The Patent Paradox Revisited: An Empirical Study of Patenting in the U.S. Semiconductor Industry, 1979–1995." RAND Journal of Economics. 32:1, pp. 101–28.
+
+Hall, Bronwyn H., Adam B. Jaffe and Manuel Trajtenberg. 2005. "Market Value and Patent Citations." RAND Journal of Economics. 36:1, forthcoming.
+
+Harhoff, Dietmar and Markus Reitzig. 2004. “Determinants of Opposition Against EPO Patent Grants—The Case of Biotechnology and Pharmaceuticals.” International Journal of Industrial Organization. 22:4, pp. 443–80.
+
+Harhoff, Dietmar, Frederic M. Scherer and Katrin Vopel. 2003. "Citations, Family Size, Opposition and the Value of Patent Rights." Research Policy. 328, pp. 1343-363.
+
+Heller, Michael and Rebecca Eisenberg. 1998. “Can Patents Deter Innovation? The Anticommons in Biomedical Research.” Science. 280:5364, pp. 698–701.
+
+Hovenkamp, Herbert, Mark D. Janis and Mark A. Lemley. 2003. "Anticompetitive Settlement of Intellectual Property Disputes." Minnesota Law Review. 87:6, pp. 1719-766.
+
+Hovenkamp, Herbert, Mark D. Janis and Mark A. Lemley. 2004a. IP and Antitrust: An Analysis of
+
+---
+
+Mark A. Lemley and Carl Shapiro 97
+
+Antitrust Principles Applied to Intellectual Property Law. New York: Aspen Publishers.
+
+Hovenkamp, Herbert, Mark D. Janis and Mark A. Lemley. 2004b. "Balancing Ease and Accuracy in Assessing Pharmaceutical Exclusion Payments.: Minnesota Law Review. 88:3, pp. 712-21.
+
+Jaffe, Adam B. and Josh Lerner. 2004. Innovation and Its Discontents: How Our Broken Patent System is Endangering Innovation and Progress, and What to Do About It. Princeton, N.J.: Princeton University Press.
+
+Kaplow, Louis. 1984. "The Patent-Antitrust Intersection: A Reappraisal." Harvard Law Review. June, 97, pp. 1813-889.
+
+Kesan, Jay P. and Marc Banik. 2000. “Patents as Incomplete Contracts: Aligning Incentives for R&D Investment With Incentives to Disclose Prior Art.” Washington University Journal of Law and Policy . 2:23, pp. 23–54.
+
+Kesan, Jay P. and Shubha Ghosh. 2004. "What Do Patents Purchase? In Search of Optimal Ignorance in the Patent Office." Houston Law Review. 40, pp. 1219-264.
+
+Kortum, Samuel and Joshua Lerner. 1999. "Stronger Protection or Technological Revolution: What is Behind the Recent Surge in Patenting?" Research Policy. 28:1, pp. 1–22.
+
+Lanjouw, Jean O. and Mark Schankermann. 2001. "Characteristics of Patent Litigation: A Window on Competition." RAND Journal of Economics. 32:1, pp. 129-51.
+
+Lanjouw, Jean O. and Mark Schankerman. 2004. "Patent Quality and Research Productivity: Measuring Innovation with Multiple Indicators." Economic Journal. 114:495, pp. 441-65.
+
+Lanjouw, Jean O., Ariel Pakes and Jonathan Putnam. 1998. "How to Count Patents and Value Intellectual Property: Uses of Patent Renewal and Application Data." Journal of Industrial Economics. 44:4, pp. 405-33.
+
+Leffler, Keith and Cristofer Leffler. 2002. "Want to Pay a Competitor to Exit the Market? Settle a Patent Infringement Case." Economics Committee Newsletter. 2:1, pp. 26-35.
+
+Leffler, Keith and Cristofer Leffler. 2003. "The Probabilistic Nature of Patent Rights: In Response to Kevin McDonald." Antitrust. Summer, pp. 77-83.
+
+Lemley, Mark A. 2000. "Reconceiving Patents in the Age of Venture Capital." Journal of Small and Emerging Business Law. 4:1, pp. 137-48.
+
+Lemley, Mark A. 2001. "Rational Ignorance at the Patent Office." Northwestern University Law Review. 95:4, pp. 1497-532.
+
+Lemley, Mark A. and Kimberly A. Moore. 2004. "Ending Abuse of Patent Continuations." Boston University Law Review. 84:1, pp. 63-123.
+
+Lemley, Mark A. and Ragesh K. Tangri. 2003. “Ending Patent Law's Willfulness Game.” Berkeley Technology Law Journal. 18:4, pp. 1085–126.
+
+Lerner, Joshua. 1995. "Patenting in the Shadow of Competitors." Journal of Law and Economics. 38:2, pp. 463-95.
+
+Lerner, Joshua. 2002. "Where Does State Street Lead? A First Look at Finance Patents, 19712000." Journal of Finance. 57:2, pp. 901-30.
+
+Long, Clarisa. 2002. "Patent Signals." University of Chicago Law Review. 69:2, pp. 625-79.
+
+Marco, Alan C. 2004. "The Selection Effects (and Lack Thereof) in Patent Litigation: Evidence from Trials." Topics in Economic Analysis and Policy. 4:1, article 21.
+
+Merges, Robert P. 1999. "As Many As Six Impossible Patents Before Breakfast: Property Rights for Business Concepts and Patent System Reform." Berkeley Technology Law Journal. 14:2, pp. 577-616.
+
+Miller, Joseph Scott. 2004. "Building a Better Bounty: Litigation-Stage Rewards for Defeating Patents." Berkeley Technology Law Journal. 19:2, pp. 667-739.
+
+Moore, Kimberly A. 2000. "Judges, Juries and Patent Cases—An Empirical Peek Inside the Black Box." Michigan Law Review. 99:2, pp. 365409.
+
+Moore, Kimberly A. 2002. "Are District Court Judges Equipped to Handle Patent Cases?" Harvard Journal of Law and Technology. 15:1, pp. 1-39.
+
+Moore, Kimberly A. 2004. "Worthless Patents." Available at <http://mason.gmu.edu/ ~kamoore>.
+
+Morse, M. Howard. 2002. "Settlement of Intellectual Property Disputes in the Pharmaceutical and Medical Device Industries: Antitrust Rules." George Mason Law Review. 10:3, pp. 359– 405.
+
+National Academies of Science (NAS). 2003. Patents in the Knowledge-Based Economy . Wesley M. Cohen and Stephen A. Merrill, eds. Washington, D.C.: National Academies Press.
+
+National Academies of Science. 2004. "A Patent System for the 21" Century." Available at (http://www.nap.edu/books/0309089107/ html).
+
+Nordhaus, William. 1969. Invention, Growth and Welfare: A Theoretical Treatment of Technological Change. Cambridge, Mass.: MIT Press.
+
+Pakes, Ariel. 1986. "Patents as Options: Some Estimates of the Value of Holding European Patent Stocks." Econometrica. 54:4, pp. 755-784.
+
+Parchomovsky, Gideon and R. Polk Wagner. 2004. "Patent Portfolios." Working paper, University of Pennsylvania Law School.
+
+Powers, Matthew D. and Stephen C. Carlson.
+
+---
+
+98 Journal of Economic Perspectives
+
+2001. "The Evolution and Impact of the Doctrine of Willful Patent Infringement." Syracuse Law Review. 51:1, pp. 53-113.
+
+Priest, George L. and Benjamin Klein. 1984. "The Selection of Disputes for Litigation." Journal of Legal Studies. 13:1, pp. 1-55.
+
+Quillen, Cecil D., Jr., Ogden H. Webster and Richard Eichmann. 2003. "Continuing Patent Applications and Performance of the U.S. Patent and Trademark Office—Extended." Federal Circuit Bar Journal. 12:1, pp. 35-55.
+
+Reinganum, Jennifer. 1989. "The Timing of Innovation: Research, Development and Diffusion," in The Handbook of Industrial Organization . Robert Willig and Richard Schmalensee, eds. Amsterdam: North-Holland, pp. 849–908.
+
+Rivette, Kevin and David Kline. 2000. Rembrandts in the Attic: Unlocking the Hidden Value of Patents. Boston: Harvard Business School Press.
+
+Schankerman, Mark and Ariel Pakes. 1986. "Estimates of the Value of Patent Rights in European Countries During the Post-1950 Period." Economic Journal. 96:127, pp. 1052-076.
+
+Schechter, Roger E. and John R. Thomas. 2003. Intellectual Property: The Law of Copyrights, Patents and Trademarks. St. Paul, Minnesota: Thomson West.
+
+Scherer, F. M. 2001. "The Innovation Lottery," in Expanding the Boundaries of Intellectual Property. Rochelle Dreyfuss et al., eds. Oxford: Oxford University Press, pp. 3-21.
+
+Schildkraut, Marc G. 2004. "Patent-Splitting Settlements and the Reverse Payment Fallacy." Antitrust Law Journal. 71:3, pp. 1033-068.
+
+Segal, Ilya. 1999. "Contracting with Externalities." Quarterly Journal of Economics. 104:2, pp. 337-88.
+
+Shapiro, Carl. 2001. "Navigating the Patent Thicket: Cross-Licenses, Patent Pools, and Standard-Setting." in NBER Innovation Policy and the Economy. Adam Jaffe, Joshua Lerner and Scott Stern, eds. Cambridge, Mass.: MIT Press, pp. 119-50.
+
+Shapiro, Carl. 2003a. "Antitrust Limits to Patent Settlements." RAND Journal of Economics. 34:2, pp. 391-411.
+
+Shapiro, Carl. 2003b. "Antitrust Analysis of Patent Settlements Between Rivals." Antitrust. Summer, pp. 70-77.
+
+Shapiro, Carl. 2004. "Economic Critique of the FTC's Proposals for Patent Reform." Berkeley Technology Law Journal. 19:23, pp. 943-70.
+
+Thomas, John R. 2001. "Collusion and Collective Action in the Patent System: A Proposal for Patent Bounties." University of Illinois Law Review. 2001:1, pp. 305-53.
+
+U.S. Patent and Trademark Office. 2003. "Annual Report for 2003." Available at <http:// www.uspto.gov/web/offices/com/annual/ 2003/2003annualreport.pdf>.
+
+U.S. Patent and Trademark Office. 2003. “The 21st Century Strategic Plan 4.” February 3; Available at <http://www.uspto.gov/web/offices/ com/strat21/index.htm>.
+
+Walsh, John P., Ashish Arora and Wesley M. Cohen. 2003. “Effects of Research Tool Patents and Licensing on Biomedical Innovation,” in Patents in the Knowledge-Based Economy . Wesley M. Cohen and Stephen A. Merrill, eds. Washington, D.C.: National Research Council, pp. 285–340.
+
+Willig, Robert D. and John P. Bigelow. 2004. “Antitrust Policy Toward Agreements That Settle Patent Litigation.” Antitrust Bulletin. 49:3, forthcoming.
+

processed/00008_W3125878831/document.md

@@ -0,0 +1,506 @@
+Journal of Economic Perspectives-Volume 27, Number 1-Winter 2013-Pages 23-44
+
+# Patents and Innovation: Evidence from Economic History
+
+Petra Moser
+
+W hat is the optimal system of intellectual property rights to encourage innovation? In the most basic theoretical models, patents pose a tradeoff between the social benefits from stronger incentives for invention and losses in consumer welfare as a result of monopoly pricing (Nordhaus 1969) . But providing stronger patents for early generations of inventors may also weaken incentives to invest in research and development for later generations (for example, Scotchmer 1991 in this journal), so that the overall effects of stronger patents on innovation are difficult to predict. Negative incentive effects are particularly severe if the boundaries of intellectual property are poorly defined, so that later generations of inventors place themselves at risk of ruinous litigation. Litigation risks are exacerbated when incumbents build “ thickets ” of strategic patents that cover little innovative progress and instead serve as a legal weapon to protect incumbents' profits (Shapiro 2001; Hall and Ziedonis 2001) . Recent patent wars over smart phones and tablet computers have moved these issues to the forefront of policy debates, but the underlying tensions are substantially more general. Empirical analyses that exploit a wealth of historical datasets and exogenous variation, when done carefully, can help to improve our understanding of these tensions and inform contemporary patent policy.
+
+Empirical analyses of historical data have emphasized the role of patent laws in creating incentives to invent, promoting innovation, and encouraging economic growth (for example, Khan and Sokoloff 1993; Lamoreaux and Sokoloff 1999; Khan 2005). In the absence of economy-wide data on the quantity of innovations, patent
+
+■ Petra Moser is a Fellow at the Center for Advanced Studies in the Behavioral Sciences and Assistant Professor of Economics, Stanford University, Stanford, California, and Faculty Research Fellow, National Bureau of Economic Research, Cambridge, Massachusetts.
+
+http://dx.doi.org/10.1257/jep.27.1.23.
+
+doi=10.1257/jep.27.1.23
+
+---
+
+24 Journal of Economic Perspectives
+
+counts have become the standard measure of innovation (for example, Schmookler 1962, 1966; Sokoloff 1988; Moser and Voena 2012), fueled in part by the creation of National Bureau of Economic Research dataset of US patents and citations between 1976 and 2002 (Hall, Jaffe, and Trajtenberg 2001), and more recently by the availability of historical patent data since 1920 through a collaboration between the US Patent and Trademark Office and Google Patents.
+
+Patent data may, however, fail to capture innovation that occurs outside of the patent system — for example, in countries without patent laws or in industries in which inventors rely on alternative mechanisms to protect their intellectual property. In fact, survey data for the late twentieth century indicate that commercial research and development labs in most industries deem alternative mechanisms, such as secrecy and lead-time (being the first firm to offer a new product) to be more effective than patents (Levin, Klevorick, Nelson, and Winter 1987; Cohen, Nelson, and Walsh 2000) . Historical accounts also indicate that innovation often occurs independently of patents as a result of knowledge sharing (Allen 1983; Nuvolari 2004; Thomson 2009) or cultural attitudes that encourage risk taking (Landes 1969) and scientific experimentation (Mokyr 2009) .
+
+Historical events — including a series of prominent technology exhibitions that started with the 1851 Crystal Palace world's fair in London — have created rich archival records on innovation within and outside of the patent system, which offer opportunities to measure the share and the characteristics of innovations that occur outside of the patent system. Data on exhibits and prizes that international juries awarded to the most innovative exhibits make it possible to examine innovation in countries without patent laws, and thus to exploit a large amount of credibly exogenous variation in patent laws to investigate the effects of patent laws on innovation. Patent laws that were in force in the mid-nineteenth century had largely been adopted ad hoc according to idiosyncratic allegiances of national rulers (Penrose 1951, p. 13) and before interest groups from individual industries had learned to lobby for stronger patents. Scientific breakthroughs that reduced the effectiveness of alternative mechanisms to protect intellectual property created exogenous shifts towards patenting, which make it possible to examine the role that patents play, for example, in the diffusion of ideas. Historical events, such as the creation of the first patent pool in 1856 and the compulsory licensing of enemy-owned US patents as a result of World War I, create opportunities to examine the effects of policies that strengthen or weaken the monopoly power of patents.
+
+To use historical evidence to guide patent policies today, one must carefully compare historical and modern institutions, political conditions, and changes in the technological characteristics of industries over time. Empirical evidence from economic history, however, can help to inform important policy questions that have proven difficult to answer with modern data. For example, does the existence of strong patent laws encourage innovation? What proportion of innovations is patented? Is this share constant across industries and over time? How does patenting affect the diffusion of knowledge? How effective are prominent mechanisms, such
+
+---
+
+Petra Moser 25
+
+as patent pools and compulsory licensing, that have been proposed to address problems with the patent system? 1
+
+## Have Patent Laws Increased the Rate of Innovation?
+
+In 1474, the Venetian Republic began to offer exclusive rights to inventors and entrepreneurs who had invented or brought new technologies to Venice. Intended to attract skilled artisans, the Republic's rudimentary patent system was copied by other European rulers to promote economic development and, more frequently, to reward political and financial support (David 1994, p. 134; Boldrin and Levin 2008, p. 43–44). In 1623, Britain's Statute of Monopolies transferred the right of granting monopolies from King James I to Parliament. North and Thomas (1973) argue that this shift, which replaced a royal prerogative to sell monopolies by a legal property rights in ideas, played a critical role in encouraging Britain's Industrial Revolution. The first article of the US Constitution instructed Congress to “promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” This provision established the foundation for the world's first modern patent system, which Khan and Sokoloff (1998, 2001) argue was instrumental in encouraging technological progress and economic growth in the United States.
+
+Recent interpretations, however, contend that patents played no major role in encouraging technological development and economic growth during Britain’s Industrial Revolution (Clark 2006; Mokyr 2009; Allen 2009). Mokyr (2009), for example, emphasizes the importance of a shift towards science-based experimentation during the Enlightenment in setting the stage for Europe’s Industrial Revolution. Alternative accounts of US innovation have emphasized the importance of relative factor prices, and in particular, the high costs of labor relative to the abundance of natural resources, as an impetus for mechanization, and for the development of a specifically American system of manufacturing (Rothbarth 1946; Habbakuk 1962; Rosenberg 1963, 1969, 1972; Hounshell 1985).
+
+Historical variation in patent laws in the nineteenth century—when some countries had not yet adopted patent laws while other abolished them for political reasons—offers unique opportunities to investigate the effects of patent laws on innovation. Switzerland, for example, had no patents until the country adopted a rudimentary patent system in 1888 and switched towards a full-fledged system in 1907 (Schiff 1971). Denmark provided limited patent protection for up to five years
+
+1 In addition to patents, innovation policy includes other types of intellectual property rights, such as copyrights, which protect books, music, and software. National governments have also begun to increasingly use prizes as an alternative mechanism to encourage innovation. More generally, the ability to attract high-skilled scientists and workers is likely to be a key factor in determining rates of innovation. Economic history also offers rich opportunities to explore the effectiveness of these alternative mechanisms (see for example Li, MacGarvie, and Moser 2012; Moser, Voena, and Waldinger 2011; Moser and Nicholas 2012).
+
+---
+
+26 Journal of Economic Perspectives
+
+in 1874, but waited until 1894 to enact an official patent law (Agnew 1874, p. 430; Boult 1895, p. 136). The Netherlands abolished its patent system in 1869 after a political victory of the free trade movement, which reflected a common view of patents as a form of protectionism and rejected them as a restriction on trade (Schiff 1971). Even for countries with patent laws, the strength of patents was far from uniform. In 1876, for example, patents in Denmark and Greece expired after five years, while patents in other countries lasted for a minimum of twelve years (Lerner 2000). Inventors around the world were also heavily dependent on domestic patent laws because patenting abroad was prohibitively expensive and — until the Paris Convention of 1883 — national patent systems discriminated heavily against foreign patentees (Bilir, Moser, and Talis 2011).
+
+Analyses of technologies that were exhibited at nineteenth-century world's fairs exploit such variation to examine differences in innovation for countries with and without patent laws. Exhibition catalogues, which guided visitors through the vast grounds of nineteenth- and early twentieth-century technology fairs, list all exhibits. Collecting these data and matching them with reports on prize-winning innovations, as well as with patent data and with geographic information, makes it possible to examine the number and the characteristics of innovations that occurred inside and outside of the patent system, which has been difficult to accomplish using patent counts as the standard indicator of innovation.
+
+Exhibition data are available for the Crystal Palace Exhibition in London in 1851, the American Centennial Exhibition in Philadelphia in 1876, the World's Columbian Exhibition in Chicago in 1893, and the Panama-Pacific International Exposition in San Francisco in 1915. In 1851, the Crystal Palace, a 1,848-foot long greenhouse of cast iron and glass, was the largest enclosed space on earth; it housed 17,062 exhibitors from 40 countries. At a time when London had fewer than two million inhabitants, more than six million entry tickets were sold for the Crystal Palace. In 1876, visitors at the US Centennial Exhibition would have had to walk more than the distance of a marathon to see 30,864 exhibitors from 35 countries; almost ten million people visited the fair (Kroker 1975, p. 146). In 1893, the World's Columbian Exposition covered 717 acres of land and water in Jackson Park by Lake Michigan; it attracted 27.5 million visitors. In 1915, San Francisco's Marina and Presidio was converted to a fairground; it welcomed 30,000 exhibitors from 32 countries and 19 million visitors.
+
+Analyses of the 1851 and 1876 exhibits reveal a perhaps surprising amount of high-quality innovations in countries without patent laws. In 1851, Switzerland and Denmark contributed 110 exhibits per million people, compared with a mean of 55 and a median of 36 per million people for all countries (Moser 2005). Swiss exhibits were also more likely to win prizes for exceptional novelty and usefulness. In 1851, 43 percent of Swiss exhibits won a prize, compared with a mean of 35 percent and a median of 33 percent for all countries. In 1876, Switzerland contributed 168 exhibits per million in population, compared with a mean of 87 and a median of 61 for all countries (Moser and Zimring 2012). The Netherlands — which had abolished patents in 1869 — won more prizes per
+
+---
+
+Patents and Innovation: Evidence from Economic History 27
+
+exhibit than any other country, with 86 percent, compared with a mean of 46 and a median of 45 percent for all countries.
+
+The world's fair data also indicate that only a small share of innovations were patented, calling into question the role of intellectual property rights in encouraging Britain's Industrial Revolution. In 1851, 11 percent of British exhibits were patented. These results are consistent with historical accounts, which emphasize the importance of cultural factors (Clark 2006; Mokyr 2009) as well as systems of collective invention without patents. For example, improvements in Cornish steam engines (Nuvolari 2004) and in blast furnaces in Cleveland's iron industry in the United Kingdom were shared freely within a system of collective invention (Allen 1983) in which patenting was rare. $^2$
+
+Data on prize-winning British exhibits help to shed light on the interaction between the quality of inventions and inventors' decision to use patents. Existing theoretical models indicate that firms may decide to keep important innovations secret because patents require disclosure, which is risky if patents are ineffective at blocking competitors from using a patented invention (Anton and Yao 2004; Horstmann, MacDonald, and Slivinski 1985) . Exhibition data, however, indicate that high-quality innovations are slightly more likely to be patented: In 1851, 15 percent of British exhibits that won prizes for exceptional usefulness and quality were patented, compared with 11 percent of average-quality exhibits.
+
+Exhibition data on the share of innovations without patents make it possible to examine how the characteristics of patent institutions influence inventors' use of patents. Khan and Sokoloff (1998, 2001, in this journal) have credited the design and low costs of patenting under the US system with encouraging technical progress and economic growth through the “democratization” of invention. In the mid nineteenth century, British inventors faced a drawn-out and expensive process, with exorbitant legal fees and bribes (MacLeod 1988, p. 76) in addition to official fees of $ 37,000 (in 2000 US dollars, Lerner 2000). $^3$ By comparison, US inventors could mail in their applications and paid only $ 618 in fees (in 2000 US dollars, Lerner 2000). Patenting rates, however, were only slightly higher for US compared with British exhibits—at 15 compared with 11 percent (Moser 2012, p. 54).
+
+US courts have also always been more likely to uphold the patent rights of early generations of inventors, while British courts tended to be more anti-patent (Dutton 1984; Khan 2005). This pro-patent bias may, however, have discouraged US rates of innovation as early as the mid nineteenth century, anticipating problems with the current system (Bessen and Meurer 2008). In 1846, for example, the US Patent and Trademark Office issued patent 4,750 to Elias Howe for an Improvement in Sewing Machines. Howe's patent was broad enough to cover most commercially viable
+
+2 Inventions within systems of collective invention were predominantly incremental (or micro-, rather than macro-inventions, Mokyr 1990), which Landes (1969, p. 92) argues "were probably more important in the long run than the major inventions that have been remembered in history books."
+
+3 Reforms of the British and other European patent systems during the "Patent Controversy" (1855–1873) may have been triggered by the Crystal Palace exhibition and the unexpected quality of US innovations (Machlup and Penrose 1950; Rosenberg 1969, p. 2).
+
+---
+
+28 Journal of Economic Perspectives
+
+sewing machines at the time. Like a twenty-first century “patent troll,” Howe used his patent to threaten litigation instead of commercializing his invention. In 1852, a District Court upheld Howe’s patent, and he began to collect license fees of $25 per machine, roughly one-fifth the average price of a sewing machine (Lampe and Moser 2012b). Then other firms sued based on their own patents, and production came to a near halt in the 1851–1856 “sewing machine wars” (Bissell 1999, p. 84). By 1867, Howe had received $2 million in license fees (Parton 1867) roughly $27.8 million in 2011 dollars (converted using the GDP deflator, based on data from Officer and Williamson 2011).
+
+## Did the Creation of Plant Patents in 1930 Encourage Innovation?
+
+Throughout the early twentieth century, living organisms such as livestock, bacteria, and plants could not be patented. After World War I, however, concerns about food security motivated the creation of intellectual property rights for plants that propagate asexually (through roots rather than seeds) in the US Plant Patent Act of 1930. Breeders of food crops had argued that, in the absence of effective alternative mechanisms, they were heavily dependent on patent rights to recover large development costs. The Stark Brothers Nursery, for example, had built a large cage, armed with a burglar alarm, to prevent competitors from stealing cuttings of the first Golden Delicious apple tree, as shown in Figure 1. By creating plant patents, Congress hoped to encourage domestic innovation and the development of a domestic US plant breeding industry.
+
+Nearly half of all US plant patents between 1930 and 1970, however, were for roses, suggesting that the 1930 legislation may have missed its target of establishing food security (Moser and Rhode 2012, pp. 418–420). Anecdotal evidence indicates that the creation of plant patents may have facilitated the development of a research-based US rose breeding industry. Similar to pharmaceutical research and development today, it took up to twelve years to develop a new rose, and fewer than one in 1,000 seedlings typically proved commercially successful (Robb 1964, p. 389; Stewart 2007, p. 131). Once a new rose had been developed, it was easy for competitors to copy and propagate through cuttings, so that original breeders could not rely on secrecy or being first to recuperate their costs of research and development. Until World War II, US nurseries had depended on imported nursery stock from Europe, but in the 1940s, roughly a decade after the Plant Patent Act, commercial nurseries, which account for the majority of plant patents, began to build mass hybridization programs for roses.
+
+Data on registrations of newly created roses between 1916 and 1970, as an alternative measure of innovation, however, suggest that the effect of plant patents was limited. Registration data suggest that US breeders created fewer new roses after 1931. Moreover, less than 20 percent of new rose varieties registered after 1930 were patented (Moser and Rhode 2012, pp. 429–434). In fact, information on lineage indicates that most roses that are commercially successful today descended from
+
+---
+
+Petra Moser 29
+
+Figure 1
+
+A Cage that Stark Brothers Nursery Built around Its Golden Delicious Apple Tree
+
+![Figure](figures/figure_002.png)
+
+Source: Image from Rossman (1930, p. 395), reproduced in Moser and Rhode (2012, p. 415). Note: The cage was built around the Stark Brother's Golden Delicious tree to prevent competitors from stealing shoots of the tree; it was equipped with an alarm.
+
+the breeding efforts of public sector plant scientists that preceded the creation of plant patents. Furthermore, historical records suggest that the US rose industry received a boost when World War II cut off rose supplies from European competitors and US breeders began to produce their own nursery stock based on licensed European roses.
+
+## Patents, Secrecy, and the Direction of Technical Change
+
+Exhibition data also indicate that the share of innovations that inventors chose to patent varied strongly across industries. For example, fewer than 5 percent of Britain's chemical exhibits in 1851, 10 percent of scientific instruments, and 8 percent of exhibits in food processing were patented, compared with 20 percent of manufacturing machinery (Moser 2012). Remarkably, US and British inventors appear to have relied on patents—and avoided patents—in the same industries despite vast differences between the British and the American patent system. Historical accounts suggest that variation in the effectiveness of secrecy, as an alternative
+
+---
+
+30 Journal of Economic Perspectives
+
+to patents, was instrumental in determining variation in the use of patents. Secrecy was an effective mechanism to protect mid-nineteenth-century improvements in chemicals because science had not yet evolved enough to allow competitors to reverse engineer them. Given the crude analytical tools of the time, valuable dyes such as indigo and madder red proved impervious to industrial espionage until the late nineteenth century (Haber 1958, p. 83). Secrecy was also effective in protecting improvements in the production of scientific instruments, such as the rectangular prisms of Swiss glassmaker T. Daguet of Soleure and the optical instruments of Danish makers (Berichterstattungs-Kommission, vol 1, 1853, pp. 813–19, 930–41). Watchmakers in the Swiss Valleé de Joux maintained tight secrecy surrounding an improved mechanism to measure minutes by agreeing not to take apprentices between 1823 and 1840 ( Jaquet and Chapuis 1945, p. 165).
+
+But if inventors' dependence on patents varies across industries, patent laws may influence the direction of technical change (Moser 2005) : In countries without patent laws, inventors depend entirely on secrecy, lead time, and other alternatives to patents in protecting their intellectual property. As a result, investments in research and development may be most attractive in industries in which secrecy can effectively guarantee exclusive rights long enough to allow inventors to recoup their investments. In countries with patent laws, inventors can use legal protection to establish exclusivity in any industry, so factors other than the effectiveness of secrecy determine the direction of technical change.
+
+Cross-country comparisons of exhibition data confirm that innovation in countries without patent laws focused on a narrow set of industries in which secrecy was effective. At the Crystal Palace, one-fourth of exhibits from countries without patent laws were scientific instruments, compared with one-seventh of exhibits from other countries (Moser 2005). Countries without patent laws also had larger shares of innovations in textiles, especially dyes, and in food processing.
+
+In food processing, the history of margarine illustrates the effectiveness of secrecy relative to patents. The French chemist Mège Mouriès, for example, believed his invention to be protected by a patent, and disclosed the process of producing margarine from suet to two Dutch entrepreneurs, Jurgens and van den Bergh. Jurgens and van den Bergh began to manufacture margarine in 1871—two years after the Netherlands had abolished patent laws in response to a victory of the freetrade movement. After a falling out, van den Bergh kept his improvements secret, and Jurgens was unable to reverse engineer the superior taste of van den Bergh margarine (which allowed for its commercialization) until 1905 (Schiff 1971).
+
+More generally, the share of Dutch innovations in food processing experienced a marked increase after the Netherlands abolished patents in 1869. In 1851, 11 percent of exhibits from the Netherlands were related to food processing. In 1876, 37 percent of Dutch exhibits, including a disproportionate amount of awardwinners, originated from this industry (Moser 2005). Many other innovations in the field, including milk chocolate, baby foods, and ready-made soups, were made in Switzerland and the Netherlands when neither country offered patents (Schiff 1971, pp. 52–58).
+
+---
+
+Patents and Innovation: Evidence from Economic History 31
+
+Survey data from the late twentieth century indicate that the relative effectiveness of secrecy and patents continued to vary strongly across industries. For example, respondents from 634 American research and development labs in 1983 (Levin, Klevorick, Nelson, and Winter 1987) and from 1,478 American firms in 1994 (Cohen, Nelson, and Walsh 2000) report that secrecy is more effective than patents as a mechanism to protect intellectual property in most industries. Harhoff and Hoisl (2006) present comparable evidence for European countries. Only for pharmaceuticals and chemical inventions are patents consistently rated as an effective mechanism to protect intellectual property today. Compared with midnineteenth-century reports, which emphasize the effectiveness of secrecy to protect chemical inventions, these results indicate that the effectiveness of secrecy varies not only across industries, but also over time.
+
+Scientific breakthroughs, which lowered the effectiveness of secrecy, may be one important factor that determines inventors' propensity to patent. In chemicals, for example, analytical advances such as August Kekulé's model of the benzene ring in 1865 and Dmitrii Mendeleev's publication of the periodic table in 1869, transformed chemical analysis in the second half of the nineteenth century. As a result of these advances, it became much riskier to protect chemicals through secrecy (Haber 1958, p. 81). At the same time, these analytical advances had no effects on innovations in machinery, which had always been easy to copy.
+
+In Moser (2012) , I exploit this differential shift to examine the effects of exogenous changes in the effectiveness of secrecy on inventors' propensity to patent. Difference-in-differences comparisons reveal a significant shift towards patenting in response to analytical advances: In 1851 and 1876, 0 and 5 percent of US chemical innovations were patented, respectively. In 1893 and 1915, 19 and 20 percent of US chemical innovations were patented, respectively. During the same time, patenting rates in manufacturing machinery — an industry in which secrecy was always ineffective — stayed roughly constant between 44 and 49 percent (Moser 2012, pp. 62–67). These results suggest that scientific breakthroughs, such as the publication of the periodic table in the nineteenth century or the decoding of the human genome today, may not only affect the speed of innovation but also increase inventors' dependency on patents.
+
+## Patent Laws and the Diffusion of Innovation
+
+This science-driven shift towards patenting makes it possible to explore whether patent rights encourage the geographic diffusion of innovative activity, which in turn has important consequences for cumulative innovation and economic growth. Analyses of patent laws typically focus on incentive effects and have largely ignored diffusion, even though disclosure and teaching a new set of firms about the “mysteries” of more advanced technologies was an important goal of early patent systems (David 1994). In fact patents are often considered as a mechanism
+
+---
+
+32 Journal of Economic Perspectives
+
+to prevent rather than encourage the diffusion of patented ideas. As Abramovitz (1989, pp. 39–40) wrote:
+
+[T]here is a need to balance the potential private rewards of innovation, which are the incentive for private investment, against the social interest in spreading knowledge and encouraging its widespread and rapid commercial application. The first element calls for protecting the private investor in an exclusive right to exploit the new knowledge he has gained. The second calls for limiting that exclusive privilege to permit diffusion and to support the competitive investments of rivals.
+
+Lamoreaux and Sokoloff (1999) , however, link the increase in US patenting in the late nineteenth century with the emergence of professional patent agents, whose role was to facilitate the trade in patented ideas. The case of Mège Mouriès (the unfortunate inventor of margarine) suggests that inventors may be more willing to disclose technical information to competitors if they feel protected by a patent. In another example from early nineteenth-century England, the UK iron founder Robert Ransome began to advertise his plough-shares to all ironmongers in Norwich and 50 outlets in East Anglia after he received a patent in 1803 (MacLeod 1988, p. 100) . By contrast, inventors have fiercely guarded knowledge from spreading to people outside their social network in the absence of intellectual property. For example, silk weavers in seventeenth-century Bologna hanged Ugolino Menzani for sharing the knowledge of a new silk twisting machine with Venetian weavers (Belfanti 2004, p. 581) , and mechanics in the nineteenth-century Pennsylvania cotton industry relied on family relations to exchange technical knowledge (Wallace 1986, pp. 211–46) .
+
+In Moser (2011) , I exploit the shift towards patenting in the nineteenth-century chemicals industry to explore whether patenting may, in fact encourage the diffusion of innovative activity: by creating intellectual property rights in ideas, patents may encourage inventors to disseminate knowledge of patented inventions, which in turn facilitates cumulative innovation and learning by doing. $^4$ A geographic analysis of exhibition data confirms that the shift towards patenting in chemicals was followed by a significant weakening in the geographic localization of inventive activity in chemicals. This decline in geographic concentration cannot be explained by changes in the localization of production; data from decennial census records for 1840 to 1920 indicate that the localization of chemical production remained relatively stable after 1876. Measuring changes in the diffusion of innovations by a geographic Herfindahl–Hirschmann index and using 1876 as a baseline, geographic concentration decreased by more than 70 percent for chemicals after 1876, compared with roughly 25 percent for manufacturing machinery. Differencein-differences regressions, which compare changes after 1876 in the geographic
+
+$^{4}$ See Scotchmer (1991) for a survey of the literature on cumulative innovation.
+
+---
+
+Petra Moser 33
+
+concentration of innovations in chemicals and manufacturing machinery, indicate that a 1 percent increase in the share of patented innovations was associated with a 1.3 percent decrease in localization.
+
+Thus, the sum of the historical evidence from exhibition data, plant patents, and other sources indicates that patent laws may influence the direction of technological change and help to encourage the diffusion of knowledge, even though patent laws do not appear to be a necessary or sufficient condition for higher rates of innovation.
+
+## Mechanisms to Modify Patent Laws: Patent Pools
+
+How can economic policy modify existing patent systems to make them more effective? A major problem with any patent system lies in the difficulty of defining the boundaries of the technology space that is covered by a patent. As a result, patent examiners may issue patents that cover overlapping areas of the technology space, such that two or more firms own blocking patents for the same technology. This in turn leads to infringement litigation, which impedes the production of new technologies and may discourage innovation.
+
+Patent pools, which allow a group of firms to combine their patents, have emerged as a prominent mechanism to resolve blocking patents and prevent or resolve patent wars. In the 1990s, four pools formed in the information technology industry: the MPEG-2 pool, the 3G platform, and two DVD pools (Merges 2001). More recently, Google launched an open-source video format pool to counter MPEG LA's pool for the H.264 video coding standard, and MPEG LA has announced plans for a pool to cover kits for diagnostic genetic testing.
+
+Although patent pools may weaken the intensity of competition, as they allow a group of firms to combine their individually held patents, regulators and courts have allowed pools, arguing, “In a case involving blocking patents, such an arrangement is the only reasonable method for making the invention available to the public” ( International Mfg. Co. v. Landon , 336 F.2d 723, 729 [9th Cir. 1964]). Another argument in favor of pools is that, at least in theory, pools that combine complementary patents may reduce license fees for outside firms as they eliminate “n-marginalization,” which occurs when firms that own patents for parts of a product charge license fees that are too high compared with the profit-maximizing fee for the complete product (Lerner and Tirole 2004; Shapiro 2001, p. 134).
+
+This positive view of patent pools is consistent with the early history of a pool that formed in the US aircraft industry to encourage the production of planes during World War I. In 1917, patent litigation between the Orville and Wilbur Wright Company and their competitor, the Curtiss Company, had brought the US production of planes to a halt. A committee under Franklin Roosevelt, then Assistant Secretary of the Navy, recommended that Wright and Curtiss form a patent pool. After the pool had formed, US output of aircraft increased from 83 in 1916 to 11,950 in 1918 (Bittlingmayer 1988; Stubbs 2002). The aircraft pool remained in
+
+---
+
+34 Journal of Economic Perspectives
+
+effect until 1975, when the US Department of Justice decided to dissolve the pool, arguing that it had “ lessened competition in research and development ” ( Federal Register 40(142), July 23, 1975, p. 30848). This decision exemplifies the tension between the potential benefits and costs of patent pools.
+
+In theoretical models, the predicted effects of patent pools on innovation are ambiguous. The prospect of a pool may motivate firms to enter a race to patent the technologies that will form the pool; this race could be productive, or it may be socially wasteful if it encourages duplicative research and strategic patenting (Dequiedt and Versaevel 2012) . The creation of a pool may also encourage investments in research and development by reducing litigation risks for members and thereby increasing expected profits from research and development (Shapiro 2001) , but it may also lead pool members to cut their own investments in research and development because they hope to be able to free-ride on the investments of other members (Vaughn 1956, p. 67) . Incentives to free-ride are particularly strong for pools that include “grant-back provisions,” which require members to offer all new patents to the pool, and innovative members may abandon the pool to protect their patents (Aoki and Nagaoka 2004) . Grant-back provisions may, however, also encourage innovation by reducing the potential for hold-up (Lerner, Strojwas, and Tirole 2007) .
+
+Empirical evidence on the effects of modern pools on innovation is limited so far. Qualitative evidence indicates that innovation increased in response to a pool for CDs, but declined in response to a pool for disk drives (Flamm 2012). In the open source software industry, the creation of a pool was followed by a modest increase in the number of new open source software products per year for technology fields in which IBM contributed patents to the pool (Ceccagnoli, Forman, and Wen 2012). $^5$
+
+Economic history offers opportunities to investigate pools across a broad range of industries and regulatory settings (Gilbert 2004) , starting with the first pool in US history, the Sewing Machine Combination (1856–1877) . This pool shared key characteristics of pools that are predicted to encourage innovation today: It combined nine complementary patents, which were necessary to build a commercially viable sewing machine, and it resolved the sewing machine patent war between Elias Howe, the Singer Company, and two other manufacturers, which had delayed commercialization. Litigation data confirm that the creation of a pool lowered litigation risks for members (Lampe and Moser 2010, p. 900) . The pool also reduced license fees from $ 25 for Howe's patent to $ 5 for the bundle of patents for members and $ 15 for outside firms, confirming theoretical predictions.
+
+Patenting, however, declined after the pool formed and only increased again after the pool dissolved in 1877 (Lampe and Moser 2010, p. 913) . A comparison with the British sewing machine industry, which had no patent pool, suggests that
+
+5 Earlier empirical analyses have focused on the determinants of pool participation (Layne–Farrar and Lerner 2010) and on rules that govern interactions between pool members (Lerner, Strojwas, and Tirole 2007) .
+
+---
+
+Patents and Innovation: Evidence from Economic History 35
+
+Figure 2
+
+Share of Sewing Machine Patents in All Patents: United States versus Britain
+
+![Figure](figures/figure_002.png)
+
+Source: Lampe and Moser (2010). Notes: US patents granted in USPTO main class 112 ("sewing") and British patents from A Cradle of Inventions: British Patents from 1617 to 1894. Series excludes patents for attachments, tables, and stands.
+
+this decline in innovation was a purely American phenomenon, as we can see in Figure 2 . In Britain, sewing machine patents continued to increase gradually as a share of all British patents until the early 1874 and experienced no increase after 1877.
+
+To investigate whether this decline in patenting reflected a decline in innovation, we collected additional data on objective improvements in the performance of sewing machines. Articles on sewing machines in nineteenth-century magazines, such as the Scientific American and the Ladies' Home Journal suggest that the key characteristics that consumers valued in a sewing machine were low weight, little noise, and most importantly, a high speed of sewing, measured as the number of stitches per minute that a machine could perform. Data on improvements in sewing speed, which we collected from company records and trade journals in the Smithsonian Institution Library, and shown in Figure 3 , indicate that improvements slowed soon after the pool had been established and did not recover until it had dissolved (Lampe and Moser 2010, pp. 916–17).
+
+Whether these results are generalizable to other industries and modern pools is an open question. The unambiguous decline in innovation for sewing machines, however, highlights the need for additional empirical — and theoretical — analyses to guide antitrust policy towards pools. Theoretical models of effects on price are
+
+---
+
+36 Journal of Economic Perspectives
+
+Figure 3
+
+Stitches per Minute
+
+![Figure](figures/figure_002.png)
+
+Sources: Figure from Lampe and Moser (2010) . Data from the Scientific American (1846–1869), exhibition catalogues, such as the “United States Commissioners Report to the Universal Exposition in Paris,” “The Report of the Twenty-seventh Exhibition of American Manufactures, Held in the City of Philadelphia,” ads in contemporary trade publications, including “The Textile American;” and historical industry analysis, such as Uniting the Tailors: Trade Unionism amongst the Tailoring Workers of London and Leeds, 1870–1939. Notes: Figure 3 plots improvements in sewing speed based on data collected from company records and trade journals in the Smithsonian Institution Library. The solid line plots a fourth-order polynomial trend.
+
+well developed (Shapiro 2001; Lerner and Tirole 2004) , but effects of patent pools on innovation are equally important and less well understood. Existing theoretical models also focus almost exclusively on member firms, but ignore effects on outside firms. Patent data, however, indicate that outside firms produced the large majority of patents across industries (Lampe and Moser 2012a) , suggesting that their response to the creation of a pool is essential to understanding the welfare effects of pools.
+
+A better understanding of the mechanism by which pools influence the rate and direction of innovation is particularly important as the use of pools expands into innovative research fields with high social value, such as biochemistry, medicines, or energy. The case of the sewing machine industry suggests that the creation of a pool may soften the intensity of competition for member firms, which tend to be larger and more established, at the expense of outside firms, which tend to be smaller and younger than pool members. For example, the sewing machine pool appears to have exacerbated litigation risks for outside firms, even as it reduced
+
+---
+
+Petra Moser 37
+
+such risks for members (Lampe and Moser 2010, p. 907) . The pool also created differential license fees that favored pool members, even though it reduced license fees (as theory predicts). Current antitrust guidelines allow pools to charge differential license fees, unless they have been shown to have direct anticompetitive effects. The experience of the sewing machine pool, however, indicates that differential license fees — which make it harder for outside firms to offer the pool technology at a competitive price — diverted the research investments of outside firms towards technologically inferior substitutes for the pool technologies (Lampe and Moser 2012b) . This finding suggests that — in the absence of effective regulation — patent pools may influence not only levels, but also the direction of technical change.
+
+## Compulsory Licensing
+
+An alternative mechanism to modify patent systems is compulsory licensing, which weakens the monopoly power of patents by licensing them to competing firms without the consent of patent owners. This policy has moved to the forefront of international trade debates, as international treaties, such as the Agreement on Trade Related Aspects of Intellectual Property Rights (TRIPS) have strengthened foreign-owned patents in developing countries, reducing access to life-saving drugs and other essential innovations (Deardorff 1992; Grossman and Lai 2004; Chaudhuri, Goldberg, and Jia 2006). To address this issue, Article 31 of TRIPS allows national governments to issue compulsory licenses of foreign-owned patents in cases of national emergencies. The World Trade Organization Doha Declaration of 2001 (WT/MIN(01)/DEC/1, Art. 5.b) further specifies that national governments have “the freedom to determine the grounds upon which such licenses are granted.” Thailand and Brazil, for example, have used compulsory licensing to procure antiretroviral drugs for millions of patients with HIV/AIDS, and India has used the threat of compulsory licensing to procure vaccines for swine flu (Kremer 2002; Galvão 2002; Gostin 2006; Steinbrook 2007).
+
+Immediate access to foreign-owned inventions may, however, come at the cost of discouraging domestic invention in the licensing country if it displaces domestic research and development. But compulsory licensing may also encourage domestic research and development that is complementary to foreignowned inventions, and the ability to produce foreign-owned inventions may create opportunities for cumulative innovation (Scotchmer 1991) and learning by doing (Arrow 1962). As a result, the effects of compulsory licensing on domestic invention are theoretically ambiguous. Empirical analyses are complicated by the fact that governments are more likely to use compulsory licensing if demand for foreign-owned inventions is high and if domestic production capacities are advanced enough to produce them; both factors may increase domestic invention irrespective of compulsory licensing.
+
+An episode of compulsory licensing under the US Trading with the Enemy Act (TWEA) as a result of World War I creates a unique opportunity to identify the
+
+---
+
+38 Journal of Economic Perspectives
+
+effects of compulsory licensing on invention. Passed on November 17, 1917, the TWEA was intended to “dislodge the hostile Hun within our gates” and to place all enemy property “beyond the control or influence of its former owners, where it cannot eventually yield aid or comfort to the enemy” (US Office of Alien Property Custodian 1919, p. 13 and 17). In March 28, 1918, the TWEA was amended to grant the Alien Property Custodian, Mitchell Palmer, the power to sell enemy property, including all enemy-owned patents “as though he were the owner thereof” (US Office of Alien Property Custodian 1919, p. 22). By February 22, 1919, Palmer announced that “practically all known enemy property in the United States has been taken over by me” (US Office of Alien Property Custodian 1919, p. 7). In 1919, the US Chemical Foundation began to issue nonexclusive licenses of enemy-owned patents to US firms.
+
+In Moser and Voena (2012) , we exploit this event to examine the effects of compulsory licensing on the patenting activity of US inventors in organic chemistry. Baseline estimates compare changes after 1918 in patent issues per year for 336 technologies with compulsory licensing, with changes for a control group of 7,248 technologies without licensing. Methodologically, the analysis takes advantage of the detailed classification system of the US Patent and Trademark Office to distinguish narrowly defined technologies (measured at the level of subclasses) that were differentially affected by compulsory licensing. Technology fixed effects (at the level of subclasses) and year fixed effects, as well as technology-specific trends make it possible to control for variation in the inventors' use of patents across technologies and over time. The difference-in-differences analyses comparing narrowly defined technologies (at a unit of analysis much below the industry level) make it possible to control for unobservable factors, such as improvements in education, the creation of protectionist tariffs, or the temporary absence of German competitors during the war, which may have encouraged US invention across all types of chemical technologies regardless of compulsory licensing.
+
+Baseline estimates indicate a 20 percent increase in domestic patenting in response to compulsory licensing (Moser and Voena 2012, p. 404). Estimates of time-varying effects indicate that this increase set in with a lag of eight to nine years and remained large and statistically significant throughout the 1930s (Moser and Voena 2012, p. 409).
+
+These results suggest that compulsory licensing may help to increase innovation in the licensing countries, even though this increase occurs with some delay if the licensing country lags behind the technology frontier. At the time of the Trading with the Enemy Act, the United States lagged behind Germany in the field of organic chemistry and needed “time to learn” (Arora and Rosenberg 1998, p. 79), even though other branches of US chemical invention were well-developed. For example, the hopes of duplicating German dyes seemed slim for US firms in 1919. Du Pont’s initial runs of indigo (which had been developed and patented by the German chemical firm BASF) turned out green (Hounshell and Smith 1988, p. 90). Similarly, countries such as Brazil and India, which are technologically advanced in many fields, seek to license foreign technologies in fields where
+
+---
+
+Patents and Innovation: Evidence from Economic History 39
+
+domestic invention is weak, and may require some time to catch up to the frontier in these fields.
+
+Learning from patent documents is particularly difficult if information in patent documents is incomplete or obscure. The German BASF, for example, had “effectively bulwarked its discovery [of the Haber–Bosch process of nitrogen fixation] with strong, broad patents which detailed meticulously the apparatus, temperatures and pressures, but cleverly avoided particulars as to the catalysts employed or their preparation" (Haynes 1945, pp. 86–87). “A prolonged learning experience was necessary [for US firms] to understand the two sides of catalysis, the chemical side and the engineering and design side” (Mowery and Rosenberg 1998, p. 75).
+
+In the case of compulsory licensing, these problems are exacerbated because licensees typically cannot access the uncodified knowledge that is embodied in skilled workers and scientists who developed the original improvement. Thus the US Winthrop Chemical Company, which had acquired all of the German company Bayer's production machinery in addition to its patents “could not figure out how to make the sixty-three drugs that were supposed to be [its] stock-in-trade . . . The former German supervisors having been jailed or deported, nobody knew how to run the machines; . . . the patents, which were supposed to specify manufacturing processes, were marvels of obfuscation” (Mann and Plummer 1991, pp. 52–53).
+
+Domestically, regulators have used compulsory licensing as a remedy to restore competition in industries that have become dominated by a small group of firms. For example, Scherer (1977, pp. 47–48) estimates that the US Federal Trade Commission and the US Department of Justice had made thousands of patents available by 1977, in industries ranging from glassware (in the 1946 breakup of the Hartford Empire pool) to copy machines (in the 1975 decision against Xerox). As a mechanism to address anticompetitive patenting behavior in domestic markets, compulsory licensing is expected to increase overall welfare by encouraging competition (Tandon 1982; Gilbert and Shapiro 1990). Survey results and case studies suggest that compulsory licencing may not provoke dramatic changes in rates of patenting and innovation (for example, Scherer 1977, Chien 2003), but more systematic empirical analyses are needed.
+
+## Conclusions
+
+Critics of the current patent system argue that a shift towards the strategic use of patents as a “sword” to hold up competitors and extract license fees threatens the effectiveness of patents as a means to encourage innovation (for example, Duhigg and Lohr 2012). The underlying problems with this system, however, may be much broader, and understanding them is critical to the design of patent policies. As early as the 1850s, patentees who did not produce anything were able to hold up entire industries because they had been issued broad patents that had been affirmed in court.
+
+---
+
+40 Journal of Economic Perspectives
+
+Historical evidence suggests that in countries with patent laws, the majority of innovations occur outside of the patent system. Countries without patent laws have produced as many innovations as countries with patent laws during some time periods, and their innovations have been of comparable quality. Even in countries with relatively modern patent laws, such as the mid-nineteenth-century United States, most inventors avoided patents and relied on alternative mechanisms when these were feasible. Secrecy emerged as a key mechanism to protect intellectual property. The effectiveness of secrecy relative to patents varies with the technological characteristics of innovations across industries and over time. In industries where secrecy was effective, inventors were less likely to use patents. Advances in scientific analysis, which lowered the effectiveness of secrecy, increased inventors' dependency on patents.
+
+Incorporating these basic facts changes the predicted effects of patent laws on innovation. If a substantial share of innovation occurs outside of the patent system, policies that implement even the most drastic shifts towards stronger patents may fail to encourage innovation. If inventors' dependence on patent protection varies across industries, implementing stronger patent rights may alter the direction of technical change. If property rights in ideas encourage inventors to publicize technical information, a shift towards patenting may encourage the diffusion of knowledge.
+
+History also offers a laboratory in which researchers can explore the effectiveness of alternative remedies to problems with the current patent system. For example, patent pools, which allow competing firms to combine their patents, have been proposed as a mechanism to resolve litigation risks as a result of overlapping patent grants, when more than one firm owns patents for the same technology. Historical evidence, however, indicates that pools may discourage and divert research and development by outside firms if the pools create differential litigation risks and licensing schemes that favor their members. Another prominent mechanism is compulsory licensing, which allows competitors to produce patented inventions without the consent of the patent owners. Historical evidence suggests that this policy may encourage innovation by allowing a new set of firms to produce a patented technology, and possibly by increasing competition to improve the technology.
+
+Overall, the weight of the existing historical evidence suggests that patent policies, which grant strong intellectual property rights to early generations of inventors, may discourage innovation. On the contrary, policies that encourage the diffusion of ideas and modify patent laws to facilitate entry and encourage competition may be an effective mechanism to encourage innovation. Carefully executed historical analyses can help to shed further light on these pressing issues of patent policy.
+
+■ I wish to thank David Autor, Eric Hilt, Ryan Lampe, Stephanie Lee, Xing Li, Joel Mokyr, Hoan Nguyen, John List, Paul Rhode, Chang-Tai Hseih, Carlos Serrano, Timothy Taylor, Joel Watson, and especially Gavin Wright for helpful suggestions, and the National Science Foundation for support through NSF Grant SES0921859 and CAREER Grant 1151180.
+
+---
+
+Petra Moser 41
+
+## References
+
+Abramovitz, Moses. 1989. Thinking about Growth: And Other Essays on Economic Growth and Welfare. Cambridge University Press.
+
+Agnew, Ficher W. 1874. The Law and Practice relating to Letters Patent for Inventions: Together with Notices of the Patent Laws in Force in the Principal Foreign States and in the Colonies. London: Wildy and Sons.
+
+Allen, Robert C. 1983. "Collective Invention." Journal of Economic Behavior and Organization 4(1): 1-24.
+
+Allen, Robert C. 2009. The British Industrial Revolution in Global Perspective. University of Chicago Press.
+
+Anton, James J., and Dennis A. Yao. 2004. “Little Patents and Big Secrets: Managing Intellectual Property.” RAND Journal of Economics 35(1): 1–22.
+
+Aoki, Reiko, and Sadao Nagaoka. 2004. "The Consortium Standard and Patent Pools." Hi-Stat Discussion Paper 32.
+
+Arora, A., and N. Rosenberg. 1998. "Chemicals: A US success story." In Chemicals and Long-Term Economic Growth, edited by A. Arora, R. Landau, and N. Rosenberg, 71-102. New York: Wiley.
+
+Arrow, Kenneth J. 1962. "The Economic Implications of Learning by Doing." Review of Economic Studies 29(3): 155-73.
+
+Belfanti, Carlo. 2004. "Guilds, Patents, and the Circulation of Technical Knowledge: Northern Italy during the Early Modern Age." Technology and Culture 45(3): 569-89.
+
+Berichterstattungs-Kommission der Deutschen Zollvereins-Regierungen. 1852–53. Aatlicher Bericht über die Industrie-Austellung aller Völker zu London im Jahre 1851 , Vols I–III. Berlin, Prussia: Verlag der Deckerschen Geheimen Ober-Hofbuchdruckerei.
+
+Bessen, James, and Michael J. Meurer. 2008. Patent Failure: How Judges, Bureaucrats, and Lawyers Put Innovators at Risk. Princeton University Press.
+
+Bilir, L. Kamran, Petra Moser, and Irina Talis. 2011. "Do Patent Treaties Encourage Technology Transfer? Evidence from the Paris Convention" http://ssrn.com/abstract=1893052.
+
+Bissell, Don. 1999. The First Conglomerate: 145 Years of the Singer Sewing Machine Company. Brunswick, ME: Audenreed Press.
+
+Bittlingmayer, George. 1988. "Property Rights, Progress, and the Aircraft Patent Agreement." Journal of Law and Economics 31(1): 227-48.
+
+Boldrin, Michele, and David K. Levine. 2008. Against Intellectual Monopoly . Cambridge University Press.
+
+Boult, Alfred J. 1895. Digest of British and Foreign
+
+Patent Laws, 2nd edition. London: Boult, Wade & Kilburn.
+
+Ceccagnoli, Marco, Christopher Forman, and Wen Wen. 2012. “Patent Pools, Thickets, and the Open Source Software Entry by Start-up Firms.” http://conference.nber.org/confer/2012/IPKE /forman.pdf.
+
+Chaudhuri, Shubham, Pinelopi K. Goldberg, and Panle Gia. 2006. "Estimating the Effects of Global Patent Protection in Pharmaceuticals: A Case Study of Quinolones in India." American Economic Review 96(5): 1477-1514.
+
+Chien, Colleen. 2003. "Cheap Drugs at What Price to Innovation: Does the Compulsory Licensing of Pharmaceuticals Hurt Innovation." Berkeley Technology Law Journal 18(3):853.
+
+Clark, Gregory. 2006. A Farewell to Alms: A Brief Economic History of the World. Princeton University Press.
+
+Cohen, Wesley M., Richard R. Nelson, and John P. Walsh. 2000. "Protecting Their Intellectual Assets: Appropriability Conditions and Why U.S. Manufacturing Firms Patent (or Not)." NBER Working Paper 7552.
+
+David, Paul A. 1994. "The Evolution of Intellectual Property Institutions." In Economics in a Changing World , Vol. 1: System Transformation: Eastern and Western Assessments , edited by A. Aganbegyan, O. Bogomolov, and M. Kaser. London: MacMillan.
+
+Deardorff, Alan, V. 1992. "Welfare Benefits of Global Patent Protection." Economica 59(233): 35-51.
+
+Department of Justice, and Federal Trade Commission. 1995. "Antitrust Guidelines for the Licensing of Intellectual Property."
+
+Department of Justice, and Federal Trade Commission. 2007. Antitrust Enforcement and Intellectual Property Rights: Promoting Innovation and Competition.
+
+Dequiedt, Vianney, and Bruno Versaevel. 2012. “Patent Pools and Dynamic R&D Incentives.” Documents De Travail No. 07-03. http://papers .ssrn.com/sol3/papers.cfm?abstract_id=988303.
+
+Duhigg, Charles, and Steve Lohr. 2012. “The Patent, Used as a Sword.” New York Times , October 7, p. A1. http://www.nytimes.com/2012/10/08 /technology/patent-wars-among-tech-giants-can -stifle-competition.html?smid=pl-share.
+
+Dutton, Harold I. 1984. The Patent System and Inventive Activity during the Industrial Revolution, 1750–1852. Manchester, UK: Manchester University Press.
+
+Federal Register. 1975. July 23, 40(142): 30848.
+
+---
+
+42 Journal of Economic Perspectives
+
+Flamm, Kenneth. 2012. "A Tale of Two Standards: Patent Pools and Innovation in the Optical Disk Drive Industry." http://conference.nber.org /confer/2012/IPKE/flamm.pdf.
+
+Galvão, Jane. 2002. "Access to Antiretroviral Drugs in Brazil." Lancet 360(9348): 1862-65.
+
+Gilbert, Richard J. 2004. “Antitrust for Patent Pool: A Century of Policy Evolution.” Stanford Technology Law Review , April 28. http://stlr.stanford .edu/2004/04/antitrust-for-patent-pools/.
+
+Gilbert, Richard J., and Carl Shapiro. 1990. “Optimal Patent Length and Breadth.” RAND Journal of Economics 21(1): 106-112.
+
+Gostin, Lawrence O. 2006. "Medical Countermeasures for Pandemic Influenza: Ethics and the Law." JAMA 295(5): 554–56.
+
+Grossman, Gene M., and Edwin L.-C Lai. 2004. “International Protection of Intellectual Property.” American Economic Review 94(5): 1635–53.
+
+Habakkuk, H. J. 1962. American and British Technology in the Nineteenth Century: The Search for Labour-Saving Inventions. Cambridge University Press.
+
+Haber, Ludwig Fritz. 1958. The Chemical Industry during the Nineteenth Century: A Study of the Economic Aspect of Applied Chemistry in Europe and North America. Oxford, UK: Oxford University Press.
+
+Hall, Bronwyn H., Adam B. Jaffe, and Manuel Trajtenberg. 2001. "The NBER Patent Citation Data File: Lessons, Insights and Methodological Tools." NBER Working Paper 8498.
+
+Hall, Bronwyn H., and Rosemarie Ham Ziedonis. 2001. “The Patent Paradox Revisited: An Empirical Study of Patenting in the US Semiconductor Industry, 1979–1995.” RAND Journal of Economics 32(1): 101–128.
+
+Harhoff, Dietmar, and Karin Hoisl. 2006. “Institutionalized Incentives for Ingenuity—Patent Value and the German Employees’ Inventions Act.” Research Policy 36(8): 1143–62.
+
+Haynes, Williams. 1945. American Chemical Industry—The World War I Period: 1912–1922. New York: D. Van Nostrand Company.
+
+Horstmann, Ignatius J., Glenn MacDonald, and Alan Slivinski. 1985. "Patents as Information Transfer Mechanisms: To Patent or (Maybe) Not to Patent." Journal of Political Economy 93(5): 837-58.
+
+Hounshell, David. 1985. From the American System to Mass Production, 1800–1932: The Development of Manufacturing Technology in the United States. Baltimore, MD: Johns Hopkins University Press.
+
+Hounshell, David A., and John Kenly Smith. 1988. Science and Corporate Strategy: Du Pont R&D, 1902–1980. Cambridge University Press.
+
+Jaquet, Eugène, and Chapuis, Alfred. 1945. La Montre Suisse. Bale and Olten: Editions Urs Graf.
+
+Khan, B. Zorina. 2005. The Democratization of Invention: Patents and Copyrights in American Economic Development, 1790–1920. Cambridge University Press.
+
+Khan, B. Zorina, and Kenneth L. Sokoloff. 1993. "Schemes of Practical Utility: Entrepreneurship and Innovation among 'Great Inventors' in the United States, 1790–1865." Journal of Economic History 53(2): 289–307.
+
+Khan, B. Zorina, and Kenneth L. Sokoloff. 1998. "Patent Institutions, Industrial Organization and Early Technological Change: Britain and the United States, 1790–1850." In Technological Revolutions in Europe , edited by M. Bergand K. Bruland, Cheltenham, 292–313. UK: Edward Elgar.
+
+Khan, B. Zorina, and Kenneth L. Sokoloff. 2001. “History Lessons: The Early Development of Intellectual Property Institutions in the United States.” Journal of Economic Perspectives 15(3): 233–46
+
+Kremer, Michael. 2002. "Pharmaceuticals and the Developing World." Journal of Economic Perspectives 16(4): 67–90.
+
+Kretschmer, Winfried. 1999. Geschichte der Weltausstellungen. Frankfurt: Campus Verlag.
+
+Kroker, Evelyn. 1975. Die Weltausstellungen im 19. Jahrhundert: Industrieller Leistungsnachweis, Konkurrenzverhalten und Kommunikationsfunktion unter Berücksichtigung der Montanindustrie des Ruhrgebietes zwischen 1851 und 1880. Göttingen, Germany: Vandhoeck Ruprecht.
+
+Lamoreaux, Naomi R., and Kenneth L. Sokoloff. 1999. "Inventors, Firms, and the Market for Technology in the Late Nineteenth and Early Twentieth Centuries." In Learning by Doing in Markets, Firms and Countries, edited by Naomi R. Lamoreaux, Daniel M. G. Raff, and Peter Temin P. University of Chicago Press.
+
+Lampe, Ryan, and Petra Moser. 2010. “Do Patent Pools Encourage Innovation? Evidence from the Nineteenth-Century Sewing Machine Industry.” Journal of Economic History 70(4): 898–920.
+
+Lampe, Ryan, and Petra Moser. 2012a. “Patent Pools and Innovation? Evidence from 20 U.S. Industries under the New Deal.” Stanford Law and Economics Olin Working Paper No. 417. http:// ssrn.com/abstract=1967246.
+
+Lampe, Ryan, and Petra Moser. 2012b. “Patent Pools and Innovation in Substitutes—Evidence from the 19th Century Sewing Machine Industry” http://ssrn.com/abstract=1468062.
+
+Landes, David S. 1969. The Unbound Prometheus. Technological Change and Industrial Development in Western Europe from 1750 to the Present. Cambridge University Press.
+
+Layne-Farrar, Anne, and Josh Lerner. 2010. "To Join or Not to Join: Examining Patent Pool
+
+---
+
+Patents and Innovation: Evidence from Economic History 43
+
+Participation and Rent Sharing Rules." International Journal of Industrial Organization 29(2): 294–303.
+
+Lerner, Josh. 2000. "150 Years of Patent Protection." NBER Working Paper 7478.
+
+Lerner, Josh, Marcin Strojwas, and Jean Tirole. 2007. "The Design of Patent Pools: The Determinants of Licensing Rules." RAND Journal of Economics 38(3): 610-25.
+
+Lerner, Josh, and Jean Tirole. 2004. "Efficient Patent Pools." American Economic Review 94(3): 691-711.
+
+Levin, Richard C., Alvin K. Klevorick, Richard R. Nelson, and Sidney G. Winter. 1987. "Appropriating the Returns from Industrial Research and Development." Brookings Papers on Economic Activity no. 3 (Special Issue on Microeconomics): 783-832.
+
+Li, Xing, Megan MacGarvie and Petra Moser. 2012. "Dead Poets' Property—The Copyright Act of 1814 and the Price of Books in the Romantic Period." http://ssrn.com/abstract=2170447.
+
+Machlup, Fritz, and Edith Penrose. 1950. "The Patent Controversy in the Nineteenth Century." Journal of Economic History 10(1): 1–29.
+
+MacLeod, Christine. 1988. Inventing the Industrial Revolution: The English Patent System, 1660-1800. Cambridge University Press.
+
+Mann, Charles C., and Mark L. Plummer. 1991. The Aspirin Wars: Money, Medicine, and 100 Years of Rampant Competition. New York: Knopf.
+
+Merges, Robert. 2001. "Institutions for Intellectual Property Exchange: The Case of Patent Pools." In Intellectual Products: Novel Claims to Protection and their Boundaries, edited by Rochelle Dreyfuss. Oxford University Press.
+
+Mokyr, Joel. 1990. The Lever of Riches: Technological Creativity and Economic Progress . New York: Oxford University Press.
+
+Mokyr, Joel. 2009. The Enlightened Economy: An Economic History of Britain, 1700–1850. New Haven, CT: Yale University Press.
+
+Moser, Petra. 2005. "How Do Patent Laws Influence Innovation? Evidence from NineteenthCentury World's Fairs." American Economic Review 95(4): 1215-36.
+
+Moser, Petra. 2011. "Do Patents Weaken the Localization of Innovations? Evidence from World's Fairs." Journal of Economic History 71(2): 363-82.
+
+Moser, Petra. 2012. "Innovation without Patents: Evidence from World's Fairs." Journal of Law and Economics 55(1): 43-74.
+
+Moser, Petra, and Tom Nicholas. 2012. “Publicity is Prize-less. Non-monetary Awards as a Mechanism to Encourage Patenting.” http://ssrn .com/abstract=1910220.
+
+Moser, Petra, and Paul W. Rhode. 2012. “Did Plant Patents Create the American Rose?” Chap. 8 in The Rate and Direction of Inventive Activity Revisited, edited by Joshua Lerner and Scott Stern. University of Chicago Press.
+
+Moser, Petra, and Alessandra Voena. 2012. "Compulsory Licensing: Evidence from the Trading-with-the-Enemy-Act." American Economic Review 102(1): 396-427.
+
+Moser, Petra, Alessandra Voena, and Fabian Waldinger. 2011. “German Jewish Émigrés and U.S. Invention.” March 21. Available at SSRN: http:// ssrn.com/abstract=1910247 or http://dx.doi.org /10.2139/ssrn.1910247.
+
+Moser, Petra, and Assaf Zimring. 2012. “Patent Laws, Market Size and Innovation.” Earlier version available at SSRN: http://ssrn.com/abstract =2010136.
+
+Mowery, David C. and Nathan Rosenberg. 1998. Paths of Innovation: Technological Change in 20th-Century America . Cambridge University Press.
+
+Nordhaus, William D. 1969. "An Economic Theory of Technological Change." American Economic Review 59(2): 18-28.
+
+North, Douglass C., and Robert P. Thomas. 1973. The Rise of the Western World: A New Economic History. Cambridge University Press.
+
+Nuvolari, Alessandro. 2004. "Collective Invention during the British Industrial Revolution: The Case of the Cornish Pumping Engine." Cambridge Journal of Economics 28(3): 347–63.
+
+Officer, Lawrence, and Sam Williamson. 2011. “Measures of Worth.” MeasuringWorth.com, www .measuringworth.com/worthmeasures.php.
+
+Parton, James. 1867. "History of the Sewing Machine." Atlantic Monthly, May.
+
+Penrose, Edith Tilton. 1951. The Economics of the International Patent System. Baltimore, MD: Johns Hopkins University Press.
+
+Robb, Harry C., Jr. 1964. "Plant Patents." In Encyclopaedia of Patent Practice and Invention Management, edited by Robert Calvert, pp. 641-55. New York: Reinhold.
+
+Rosenberg, Nathan. 1963. "Technological Change in the Machine Tool Industry, 1840-1910." Journal of Economic History 23(4): 414-43.
+
+Rosenberg, Nathan. 1969. The American System of Manufactures: The Report of the Committee on the Machinery of the United States 1855, and the Special Reports of George Wallis and Joseph Whitworth, 1854. Edinburgh University Press.
+
+Rosenberg, Nathan. 1972. Technology and American Economic Growth. New York: Harper.
+
+Rossman, Joseph. 1930. "The Planter Breeder Becomes an Inventor." Science News-Letter 18, December 20, pp. 349-95.
+
+---
+
+44 Journal of Economic Perspectives
+
+Rothbarth, Edwin. 1946. "Causes of the Superior Efficiency of U.S.A. Industry Compared with British Industry." Economic Journal 56: 383-90.
+
+Scherer, Frederic. M. 1977. The Economic Effects of Patent Compulsory Licensing. New York University, Graduate School of Business Administration, Center for the Study of Financial Institutions.
+
+Schiff, Eric. 1971. Industrialization without National Patents: The Netherlands 1868–1912, Switzerland 1850–1907. Princeton University Press.
+
+Schmookler, Jacob. 1962. "Economic Sources of Inventive Activity." Journal of Economic History 22(1): 1-20.
+
+Schmookler, Jacob. 1966. Invention and Economic Growth. Cambridge: Harvard University Press.
+
+Scotchmer, Suzanne. 1991. "Standing on the Shoulders of Giants: Cumulative Research and the Patent Law." Journal of Economic Perspectives 5(1): 29-41.
+
+Shapiro, Carl. 2001. "Navigating the Patent Thicket: Cross Licenses, Patent Pools, and Standard Setting." Chap. 4 in Innovation Policy and the Economy, Vol. 1, edited by A. Jaffe, J. Lerner, and S. Stern. Cambridge, MA: MIT Press.
+
+Sokoloff, Kenneth. 1988. "Inventive Activity in Early Industrial America: Evidence from Patent Records, 1790-1846." Journal of Economic History 48(4): 813-50.
+
+Steinbrook, Robert. 2007. "Thailand and the
+
+Compulsory Licensing of Efavirenz." New England Journal of Medicine 356(6): 544-46.
+
+Stewart, Amy. 2007. Flower Confidential: The Good, the Bad, and the Beautiful in the Business of Flowers. Chapel Hill, NC: Algonquin Books of Chapel Hill.
+
+Stubbs, Kevin D. 2002. Race to the Front: The Material Foundation of Coalition Strategy in the Great War, 1914–1918. Westport, CT: Greenwood Publishing Group.
+
+Tandon, Pankaj. 1982. "Optimal Patents with Compulsory Licensing." Journal of Political Economy 90(3): 470-86.
+
+Thomson, Ross. 2009. Structures of Change in the Mechanical Age: Technological Innovation in the United States, 1790–1865. John Hopkins University Press.
+
+US Office of Alien Property Custodian. 1919. “Alien Custodian Report: A Detailed Report by the Alien Property Custodian of All Proceedings Had by Him under the Trading with the Enemy Act during the Calendar Year 1918 and to the Close of Business on February 15, 1919.” Washington: Government Printing Office.
+
+Vaughan, Floyd L. 1956. The United States Patent System: Legal and Economic Conflicts in American Patent History. Norman, OK: University of Oklahoma Press.
+
+Wallace, Anthony F. C. 1986. Rockdale: The Growth of an American Village in the Early Industrial Revolution. New York: Knopf.
+

processed/00016_W2168654179/document.md

@@ -0,0 +1,769 @@
+2
+
+# Intellectual Property: When Is It the Best Incentive System?
+
+Nancy Gallini, University of Toronto Suzanne Scotchmer, University of California, Berkeley
+
+## Executive Summary
+
+Intellectual property is not the only mechanism used in the American economy for rewarding R&D. Prizes and contract research of various types are also common. Given the current controversies that swirl around intellectual property policies, we review the economic reasoning that supports patent and other intellectual property over the alternatives. For those economic environments where intellectual property is justified, we review some of the arguments for why it is designed as it is. We focus particularly on the issue of how broad awards should be and how much protection should go to the original inventor (as opposed to those who subsequently improve the invention). We emphasize that the ideal design of an intellectual property system depends on the ease with which rightsholders can enter into licensing and other contractual arrangements involving these rights.
+
+## I. Introduction
+
+Intellectual property is the foundation of the modern information economy. It fuels the software, life sciences, and computer industries, and pervades most other products we consume. Although most inventors consider it essential, it is currently under attack by some academics and policymakers. One complaint is that intellectual property rewards inventors beyond what is necessary to spur innovation. Another is that intellectual property is a drag to innovation, rather than a spur, since it prevents inventions from being used efficiently, especially in creating further innovations. A third complaint is that some inventions should not be protected at all but, instead, be supported by public sponsors.
+
+Controversies over what should constitute intellectual property swirl around business methods, computer software, research tools in the biomedical industry, and genetic sequences. However this is not new; controversies have swirled around every new technology in the
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+52
+
+Gallini and Schotchmer
+
+twentieth century. A sampler might include the question of whether player piano rolls should receive copyright protection, whether purification of chemical compounds constitutes “invention” for purposes of patent law, and whether mathematical algorithms such as public key encryption should be patentable. Technologies that fall outside the subject matter of patents and copyrights have sometimes received sui generis protections, such as computer chips under the Semiconductor Chip Protection Act.
+
+For all these technologies, the same questions arise: Are there natural market forces that protect inventors so that formal protections or other incentives are not necessary? If not, is intellectual property the best incentive system, or would the technology more appropriately be developed by a public sponsor and offered freely in the public domain? How should intellectual property be designed so as to minimize deadweight loss due to monopoly pricing without undermining incentives to innovate?
+
+Our objective in this paper is to review what economists have said about incentive schemes to promote R&D including intellectual property. While we focus on environments in which other forms of protection are not available, we note that other protections can obviate the need for any formal reward system. For example, encryption offers the potential to protect digitally distributed products such as music, movies, and software, even in the absence of intellectual property (National Research Council 2000). In the realm of databases, for which formal protections have been mandated in Europe and proposed in the U.S. Congress, vendors are protecting their data with both clever business strategies and technology (Maurer 1999, Maurer and Scotchmer 1999). In markets with network effects, there may be natural barriers to entry, so that a vendor may capture the entire market even without formal protection (Farrell 1995). And, of course, trade secrecy can be an important protection, especially when firms devise clever nondisclosure agreements that enable them to license without leaking the secret to unauthorized users (e.g., see Anton and Yao 1994). In some of these examples, the alternative protection involves social costs that could be avoided by formal intellectual property. But if not, the case for intellectual property may be weak.
+
+In section II , we compare intellectual property with alternative incentive schemes. Without losing the thread of the paper, the reader who is only interested in the design of intellectual property (as opposed to other incentive schemes) can skip the last three subsections of
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c
+
+---
+
+Intellectual Property
+
+53
+
+section II. In section III we review optimal design issues for intellectual property, especially the question of patent breadth, and in section IV we turn to the special problems that arise when innovation is cumulative. In section V, we summarize the arguments for and against intellectual property. We comment on whether the design recommendations of economists can actually be implemented, and argue that intellectual property regimes should be designed so that the subject matter of each one has relatively homogeneous needs for protection.
+
+## II. Alternative Mechanisms for Rewarding Innovation
+
+Competitive markets may not be conducive to innovation, for a reason that was well articulated by Arrow (1962). Inventions are information, and information is a public good. An invention such as a wireless palmtop is a combination of tangible embodiments and an intangible idea, as well as information about how to manufacture it. Typically, both the information and the tangible embodiments are costly to the inventor, but only the tangible components are costly to a rival. Without some sort of protection or reward, the inventor will therefore be at a market disadvantage relative to rivals, and will possibly be dissuaded from investing.
+
+Arrow explained why some incentive scheme is needed, but not which scheme. Many schemes have been used in practice. In the seventeenth century, for example, a prize was offered in France for developing a workable water turbine (Reynolds 1983, p. 338) . For about a century in the same era, a prize was outstanding for developing a method to calculate longitude at sea (Sobel 1995) . In the modern era, R & D is sponsored to a large extent by government grants. According to the National Science Foundation (2000) , in 1998 about 30 % of U.S. research was funded by the Federal government. These examples raise the following question: In what environments are there better incentive schemes than intellectual property?
+
+We shall use the term intellectual property (IP) to mean an exclusive right to market an invention for a fixed time period. It includes copyrights, patents, plant patents, protection under the Plant Variety Protection Act, and other sui generis types of protection. By a prize we mean a payment funded out of general revenue that is made to a researcher conditional on delivering a specified invention. Prizes can either be tailored individually to firms, depending on their efficiency characteristics, or can be offered symmetrically to any firm that wants to compete,
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+54
+
+Gallini and Schotchmer
+
+just as a patent is. By procurement , we mean a mechanism to solve the problem of getting an invention at minimum cost, in a timely manner, or otherwise efficiently (e.g., Laffont and Tirole 1986, 1987). A simple procurement mechanism would be an auction for the right to be paid when the invention is delivered.
+
+A form of procurement commonly used in government-sponsored research appears, on its face, to be a fixed-price contract. For example, the National Institutes of Health give funding in advance for projects that are described in the proposals. Funds are not withheld if the output is not delivered, since the idea of the contract is to pay costs as they accrue. If such funding were a one-time event for each researcher, researchers might be inclined to “take the money and run.” This moral hazard problem is overcome because future grants are contingent on previous success. The linkage between previous success and future funding seems even more specific in the case of the National Science Foundation. Fixed-price contracts thus operate much like prizes, with the wrinkle that a researcher must convince the sponsor in advance that his output might be worthy of a prize. For this purpose, his reputation might suffice, and in some cases, much of the research has already been completed.
+
+We begin our analysis with a benchmark. When both the costs and values of innovations are publicly observable to both firms and a public sponsor, IP is not the best incentive scheme. A better scheme is for a public sponsor to choose the projects with the largest net social benefits, and pay for them on delivery, using funds from general revenue. With IP, projects are funded out of monopoly profits. Monopoly pricing is equivalent to taxing a single market, which is generally thought to impose greater deadweight loss than the broad-based taxation that generates general revenue. Thus, to justify intellectual property, there must be some type of asymmetric information about the costs and benefits of research programs.
+
+We first make some comparative remarks about intellectual property, prizes, and procurement contracts. These remarks are much in the spirit of Wright (1983), who gave the first formal treatment of how asymmetric information should inform our choice among incentive mechanisms. In the subsections that follow, we then show that these three mechanisms can generally be improved upon. $^1$
+
+IP has an obvious defect as well as obvious virtues. The defect is the deadweight loss due to monopoly pricing. The virtues are several. Most importantly, if the costs and benefits of R&D investments are
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+55
+
+known only to firms, and not to government sponsors, firms will use their superior knowledge to screen investments. A sponsor does not need to decide in advance which investments are meritorious. An investor knows that he will be punished by the market if he does not invest wisely. Another obvious virtue is that the prospect of valuable IP might elicit higher levels of effort than those generally associated with sponsored research. For example, much has been made of the human genome project, whose completion was accelerated by a private firm hoping to win IP rights on gene sequences. Finally, an IP system imposes the costs of an invention on its users. In other incentive mechanisms, the costs are borne more generally by taxpayers. Taxpayers might rightfully revolt if asked to bear the costs of developing, say, computer games.
+
+Lest these advantages of IP be overstated, however, we note that prizes have many of the same virtues. If an investment's prospective value is known to the sponsor (or defined by the sponsor, as in the case of military wares), the sponsor can screen projects himself. A prize system then seems superior to IP. It avoids deadweight loss, and can be as good as IP at motivating effort.
+
+Moreover, IP will not work as an incentive mechanism unless third parties can observe at least some aspects of value. A rightholder must be able to defend his right against potential infringers. He must be able to prove in court that his intellectual property meets the standard for protection, and that an alleged infringer is marketing a product that falls within the breadth of his claims. Aspects of the invention's value must therefore be observable ex post, although typically at the high cost of litigation and discovery.
+
+The ex post observability requirement will typically impose less cost under an IP system than under a prize system. Under an IP system, the costs of discovery are incurred only if there is litigation. In contrast, for a prize, costs would have to be incurred for every invention in order for the sponsor to set a payment commensurate with the value. $^2$ Therefore, our distinction is not really between observability and nonobservability, but rather on whether the value is known to the sponsor without incurring cost. The most natural example is when the sponsor defines the value of the invention himself, as in military procurement.
+
+Recently the World Health Organization and the World Bank have suggested prizes for developing vaccines that would not be developed or might not be widely enough distributed under a system of
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+56
+
+Gallini and Schotchmer
+
+proprietary rights. The problems are great: how to assess whether a vaccine merits a prize; how to ensure that the prizes are not given prematurely before higher-quality vaccines are brought forward; how to ensure that the prizes are actually given, when it is easy to manufacture reasons to withhold them. Prizes can be organized so that worthy projects need not be identified in advance, but administering the prize then becomes particularly burdensome. The problems are particularly acute where innovation is cumulative. See Kremer (2000) for a thoughtful and detailed analysis of how such a system might work.
+
+Unlike IP, a procurement contract would typically not be offered to all comers. Instead there would be a negotiation phase in which the procurement officer tries to sort out which firm(s) are more efficient, and only offers the “ prize ” to those firms. A mechanism that allows such flexibility is more effective by definition than a prize offered to all comers. As for prizes, the sponsor must identify worthy projects. For traditional government procurement, such as for fighter jets, this is automatic. For medical research, the sponsor may solicit open-ended proposals, which entails administrative cost. In addition, the negotiation required for procurement might be politically infeasible as well as costly.
+
+In the next subsections, we investigate optimal incentive mechanisms in specific research environments, with a view toward understanding how optimal mechanisms relate to IP, prizes, and simple reimbursements. We focus on environments in which no alternative mechanisms for protection (private or market) are available, and on single inventions that do not lead to future innovations. Following Scotchmer (1999b) , we stylize the allocation problem as having three facets, which are intertwined. The first is the decision problem: should a project be undertaken? The second is the delegation problem: by which firms, or how many, and at what rates of investment? The third is the funding problem: Can the deadweight loss of monopoly pricing be avoided?
+
+## The Problem of Aggregating Information
+
+To solve the decision, delegation, and funding problems jointly, all the information that is decentralized among firms may have to be aggregated. IP, prizes, and simple procurement mechanisms such as fixed-price contracts and auctions cannot aggregate information, and are therefore flawed at the outset.
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+57
+
+To see this, consider a well-defined project, such as finding an AIDS vaccine or developing supersonic transport. Suppose that there are two potential researchers, $i=1,2$ , and that each researcher $i$ has an efficiency parameter $c_i$ for this project, interpreted as the cost of success. The product will have a common value $v$ regardless of which firm develops it, and each firm has a signal $v_i$ of this value. The underlying value will typically be determined by the extent of demand or anything else that affects monopoly profit and social welfare. Because each $v_i$ is a noisy signal of an underlying common value, it is natural to suppose that the signals $\{v_1,v_2\}$ are correlated. It is less obvious whether the cost parameters $\{c_1,c_2\}$ would be correlated. We shall assume that they are independent draws from a known distribution.
+
+To make an efficient investment decision, each firm would like to know the other firm's signal. For example, a firm with a low signal of value, $v_1$ = $L$ , might invest if it knew the other firm had a high signal of value, $v_2$ = $H$ , but not otherwise. But neither the value nor its best estimate is known ex ante to other firm, since neither can observe the other's signal.
+
+The importance of aggregating information is revealed in the following special case in which both the costs and the signals take on binary values: $c_\textbf{f} \boldsymbol{\epsilon}\{l,h\},~v_l \boldsymbol{\epsilon}\{L,H\}$ . Suppose that the first-best, full-information rule for allocative efficiency is that the project should be undertaken unless (1) both firms have high costs, regardless of the signals of value or (2) both firms have low signals of value, regardless of costs. The project should be undertaken by a single firm if (3) at least one firm has low cost and at least one firm has a high signal of value, or (4) both firms have high cost and both have high signals of value.
+
+Suppose $(c_1,v_1)=(l,L)$ . Firm 1 should invest if $(c_2,v_2)=(h,H)$ but not if $(c_2,v_2)=(h,L)$ . Without knowing firm 2's information, firm 1 could not make an efficient decision. Such could be the case under a patent system. Firm 1 may fail to invest because it is pessimistic about value $(v_1=L)$ , and firm 2 may fail to invest because its costs are too high $(c_2=h)$ . If the firms could share their information, firm 1 would invest based on firm 2's propitious information about the market. To some extent, the firms should be able to learn each other's private information by observing each other's investments. However, even if the firms know each other's costs, they might get stuck in an inefficient, but self-reinforcing, equilibrium where each invests because the other is investing, and each incorrectly thinks the other has a high signal of value (or vice versa) (Minehart and Scotchmer 1999) . When the firms have
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+58
+
+Gallini and Schotchmer
+
+different, unobservable costs, the difficulties of making inferences from investment behavior are compounded. A firm that invests could be doing so either because it has low cost or because it has very propitious private information about the market. The observing firm cannot distinguish between these two cases.
+
+Neither IP nor prizes nor simple procurement mechanisms (e.g., auctions) can cope with the problem of aggregating information. Scotchmer (1999b) describes a procurement mechanism that bears little resemblance to auctions, prizes, or IP, but can achieve as good an outcome as when the signals of value are known, provided the firms' signals of value are correlated. $^3$ While the mechanism described will delegate efficiently, it may not be realistic given the constraints of government procurement. The mechanism might entail payments from firms to the government, or payments to firms that are not asked to invest. Such payments would be difficult to enforce.
+
+The problems with the efficient procurement mechanism may explain the use of prizes, IP, and simple procurement mechanisms, but, under the conditions presented in this example, no one has studied their relative merits as second-best mechanisms. In order to identify the relative merits of the simple schemes and other more realistic mechanisms, we now consider the decision and delegation problems separately.
+
+## The Delegation Problem
+
+We isolate the problem of optimal delegation by assuming that the sponsor already knows the optimal decision, namely, to invest. That is, the sponsor knows the value of the project and that it exceeds the cost of delivery, but it does not know which firm(s) is (are) more costefficient. Optimal delegation has two components: choosing the most efficient firm or group of firms, and motivating the firm(s) to invest at efficient rates.
+
+If the sponsor faced only a problem of selecting the more efficient firm(s), then the delegation problem would be easy to solve, e.g., by auctioning the right to invest. In contrast, IP and prizes could lead to inefficiency. If the market has room for only one firm, there is no reason to suppose that the lower-cost firm will be the entrant, especially when the relative efficiencies of the firms are not publicly observable.
+
+But even an auction will not perform well when there is also a problem of inciting the right amount of effort, so that the invention is deliv-
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+59
+
+ered in a timely manner. The appropriate rate of progress is key to the economics of R&D How much additional cost should be tolerated in return for a higher rate of progress?
+
+A firm's willingness to accelerate invention at higher total cost depends on the prize it will receive, conditional on delivering the product. Thus the size of the prize determines the rate of investment. However the optimal size of prize (and the optimal rate of investment) depend both on the researcher's “efficiency” and on his efficiency relative to other firms. For an inefficient firm, the optimal rate of investment may be zero if it is possible to delegate to a more efficient firm, but positive if the other firm is even less efficient. Thus, the problem is to tailor prizes both to the firms' individual efficiencies and to their relative efficiencies.
+
+Gandal and Scotchmer (1993) study this problem, and show that the sponsor should offer a menu of options with both fixed fees and firm-specific prizes. $^4$ The menu serves two purposes: it gets the firms to reveal their relative efficiencies, and, once the contracts are awarded, it gets the firms to invest at the efficient rates. The difficulty is in the coordination: each firm's efficient rate depends on both firms' efficiency parameters. A simple patent or prize system, where the IP or prizes are not tailored to the firms' relative efficiency, will not ensure that only the most efficient firm(s) invest, or at the efficient rates. And a simple fixed-price contract may not create incentives to invest fast enough, even if the contract is auctioned to the more efficient firm.
+
+The message here is that, even when the value of the prospective invention is known prior to the investments, optimal procurement requires a mixture of prizes and fixed payments, rather than a pure prize system, a patent system, or an auction. Simple mechanisms can be resurrected as best in very simple contexts. An auction performs well when the only issue is to choose the most efficient firm, but there is no issue of eliciting the right amount of effort. A simple prize performs well when there is a single firm qualified to undertake the research. If the prize is set equal to the social value, the firm will have the same objective function as society and will invest efficiently. Since the best simple mechanisms are different for different simple contexts, it is no surprise that complicated research environments with several firms call for mechanisms that combine instruments.
+
+In the next section we focus on the optimal decision problem, assuming that the value of the innovation is unknown. In order to avoid the
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+60
+
+Gallini and Schotchmer
+
+problem of optimal delegation, we also assume there is a single potential researcher.
+
+## The Decision Problem
+
+We have just pointed out that if there is a single firm qualified for the research program, the optimal mechanism is a prize set equal to the social value. The firm's private incentives are then aligned with social incentives. However, to set such a prize, the sponsor must know the social value in advance or observe it ex post. Since IP automatically reflects the social value, at least to some extent, IP looks like an attractive alternative to a prize when the social value is unobservable. We now investigate whether this justification for IP holds up.
+
+Kremer (1998) proposes a system to create a prize equal to the social value, even when the sponsor cannot observe it in advance. His proposal involves IP, but avoids deadweight loss by turning a patent into a prize. He proposes that the patent authority take possession of the patent, and auction it to the highest bidder, assuming that every firm can observe the value ex post. The rules of the auction are that with very small probability the patent will actually be sold to the highest bidder, and otherwise the invention will be put in the public domain. Firms will bid the true value, hence revealing it. The social value is estimated from the revealed private value, and the inventor receives a prize equal to the social value, paid out of general revenue. He will thus invest if the social value exceeds his cost, as is efficient. $^5$
+
+Another scheme to avoid deadweight loss is proposed by DeLaat (1996) . To illustrate his idea in a very simple model, suppose that a potential R & D project is described by a pair ( c,v ), where c is the cost, which is observable to the sponsor, and v is the value, which is not. But if the cost c is observable to the sponsor, he can ask the researcher to report the prospective value v , and then give a fixed-price contract to reimburse the cost c if and only if the prospective value exceeds the cost. Since the researcher earns zero profit whatever he reports (he is only reimbursed the cost), he will report the value truthfully to the sponsor, who will make the efficient decision whether to invest. Thus, IP is unnecessary.
+
+But this scheme only seems credible if (contrary to the premise) the value of the invention is observable ex post, or if the sponsor can verify that the researcher is investing exactly as he promised (as deLaat assumes explicitly). $^6$ If not, the researcher could use the contract money
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+61
+
+for other purposes and deliver a shoddy product; there is a disabling problem of moral hazard, which IP could overcome.
+
+Nevertheless, we can conclude from the arguments of Kremer and deLaat that if either cost or value is truly observable to a sponsor, there may be a better mechanism than IP. Consistent with this view, Scotchmer (1999a) justified patents by assuming that the cost and value are both unobservable. A similar interpretation can be made for Cornelli and Schankerman (1999) . The latter present a model where the value of an invention is endogenous to the firm's investment effort, which, in turn, depends on an unobservable efficiency parameter. In effect, neither cost nor value is observable to the sponsor. Thus it is hard to see how any mechanism short of IP could be effective. Since the value of the patent increases with the value of the invention, a patent system gives the firm at least some incentive to spend more resources to create a product of greater value. Cornelli and Schankerman show how this incentive can be increased by using a patent renewal system.
+
+The patent renewal system is a menu of options (F,T) , where F is a payment from the patentholder to the sponsor and T is a patent life. 7 The fee F increases with the patent life, and might start out negative (a subsidy). The patentee can then “buy” a longer patent life by paying renewal fees. The value of the patent automatically increases with the value of the invention, but increases more for higher-value inventions, since those are the ones that will be renewed in return for fees. Thus the incentive to develop higher-value products is compounded.
+
+Scotchmer (1999a) derives the renewal system as a multidimensional screening mechanism for ideas, (c,v) , where both are unobservable. Again, it is the higher-value ideas that will be renewed the longest, compounding their value. Thus the cost $c$ that firms are willing to bear may go up faster than linearly with the value of the innovation, $v$ .
+
+As mentioned, the renewal system could start with subsidies, which are then reduced as firms pay fees in return for a longer patent life. Subsidies are advocated by Shavell and van Ypersele (1998) on grounds that they are a more efficient way to reward innovators than IP. Subsidizing low-value innovations allows the protection on high-value innovations to be shorter (thus reducing the deadweight loss), without jeopardizing incentives to innovate.
+
+The problem with subsidies, of course, is that they may be exploited by opportunistic firms, which could collect the subsidy and either not invest or produce something worthless. To avoid this problem, subsidies, like their close kin, prizes, must be contingent on some aspect of
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+62
+
+Gallini and Schotchmer
+
+the resulting invention, such as its value. Thus it seems reasonable to suppose that subsidy schemes will not be used if the invention's value or success cannot be verified ex post. But then we have a contradiction. If subsidies are possible, it must be because some aspect of value is observable ex post. If so, IP should not be used at all, since prizes (rewards, fixed-price contracts) dominate. IP and prizes can serve the same screening function, and can motivate firms to the same levels of effort, but prizes avoid the deadweight loss. Consistent with this caveat, renewal schemes seen in practice do not provide for subsidies. (See Calandrillo (1998) for a broader set of criticisms of subsidies.)
+
+In conclusion, IP can be justified in two ways. First, it can be justified as a screening mechanism to encourage investment in high-value projects, which may also have high cost. Second, it can be justified as a means to increase the rate at which firms invest, either to increase value or to accelerate progress. Without a means to link prizes to social value, there is no alternative to achieve these results. These virtues of IP should be weighed against the aggregation problems described earlier when more than one firm is capable of the research.
+
+Assuming that, in a second-best analysis, IP would prevail, we now ask how the right should be designed. We have already discussed the benefits of a renewal system. But how broad and long should protection be?
+
+## III. Optimal Design: The Case of a Single Innovation
+
+Perhaps the most influential work on patent design was that of Nordhaus (1969), who explained why patents (or other IP) should have finite length. If the sole concern is to encourage innovation, then IP should last forever. And if the sole concern is to avoid deadweight loss that occurs through proprietary prices, then IP should not exist at all. A finite length of protection balances these two concerns. Longer protection would encourage more innovation, but only by prolonging the deadweight loss on inventions that would be made anyway.
+
+Nordhaus's simple framework spawned a large literature on the design of IP with consideration of patent races, imitation by rivals, technology licensing, and how the design question changes when technology is cumulative. In this section we focus on the design question of breadth (also called scope ), which occupied considerable journal space in the 1990s. In the next section we turn to sequential or cumulative research, where breadth plays a different role.
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+63
+
+We begin with Gilbert and Shapiro (1990; GS), who introduced the notion of patent scope into the Nordhaus analysis. They define patent scope as the price $p$ that the innovator is able to charge for the product that embodies the innovation. Thus a patent policy is (T,p) , where $T$ is the patent life. While such a definition is far removed from what a court might use, the analysis that arises from using it is still informative, as discussed below. Maximizing social surplus over all combinations (T,p) that yield enough revenue to cover the cost of research, GS find that optimal patent length is infinite, with the patent scope set at the level that just covers R & D investment. $^8$ That is, the optimal design is for the patent to be narrow and long.
+
+Gallini (1992) reversed this design conclusion in a model where patent breadth determines the ease of entry into the protected market. She defined scope technologically, as the cost $K$ that rivals must incur to imitate the invention without infringement. Thus a patent policy is a pair ( T,K ). The lower price that results from narrow scope arises from rivals' attempts to “invent around” the patent, rather than from some type of regulatory or antitrust action, as assumed by GS. In contrast to GS, the innovator's profit does not strictly increase with patent life, since a long patent life will encourage imitation (hence competition) before the patent expires. An increase in patent life provides incentives for wasteful imitation but not for productive innovation.
+
+For a given imitation cost $K$ , a sufficiently long patent will attract imitators, resulting in oligopoly pricing instead of monopoly pricing. Conversely, for a given patent life $T$ , a sufficiently narrow scope will attract entrants. Patent life and scope are complementary in that both instruments must be increased or reduced to achieve most efficiently the required reimbursement to the innovator.
+
+With imitation, the social cost of a patent may have two components: deadweight loss and the cost of imitation. The optimal patent policy minimizes these costs. Gallini shows that the optimal design is to avoid entry entirely by making the patent broad and short, in contrast to that proposed in GS. That is, the patent should be just long enough to generate the required revenue for the monopolist patentholder, and broad enough to prevent imitation.
+
+However, this reversal depends on an assumption about licensing (or, rather, its absence). In the Gallini model, if the patent is too long or too narrow, the innovator is assumed to sit back passively and watch imitators erode her market share. Maurer and Scotchmer (1998; MS) point out that the duplicative waste could be avoided voluntarily
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+64
+
+Gallini and Schotchmer
+
+through licensing rather than by adjusting patent policy, which can again reverse the optimal design. Whatever the market outcome without licensing, the innovator and potential entrants can achieve the same market outcome (price and number of entrants) through a licensing agreement with appropriate royalties and other fees. Since both the innovator and potential entrants can jointly save the imitation costs, they prefer licensing to imitation. The innovator can do even better by fine-tuning the number of entrants.
+
+An important point of agreement among GS, Gallini, and MS is that a narrow patent reduces market price. However, their arguments differ. GS have in mind some sort of regulatory mechanism; Gallini argues that the price reduction will occur through duplicative entry; and MS argue that the price reduction will occur through licensing to prevent duplication. In addition, the analyses of social cost differ, leading to different prescriptions about optimal length and breadth. Since GS do not recognize imitation costs, they simply ask whether the deadweight loss of monopoly pricing is smaller with a long patent and low price, or with a short patent and high price. Gallini argues that if the social cost includes the cost of imitation, the optimal policy should be aimed at avoiding it. MS argue that the imitation costs will not be borne in practice if licensing is available, so that the GS type of analysis is restored.
+
+It is worthwhile expanding on why licensing will lower the market price, by considering what would happen if there were a single potential entrant. The latter situation was analyzed by Gallini (1984) , who first pointed out that licensing can prevent entry. With a single potential entrant (or a fixed number), the optimal licensing strategy is to sustain the profit-maximizing (monopoly) price with high royalties, and to share the revenues by using other fees. The licensor has an incentive to keep the market price high regardless of the cost of imitation. In contrast, in the argument above, the licensor is worried about imitation by non licensees as well as by licensees; there is always an unlicensed potential entrant. The patentholder commits to a low market price precisely to reduce the attractiveness of entry by nonlicensees, who can be numerous and unidentifiable ex ante. This point stresses the significance of potential entry to the welfare analysis of licensing and, therefore, to the optimal design of IP.
+
+The foregoing discussion shows that private contracting can dramatically alter the optimal design of patents, and that public and private instruments may be complementary in reducing social costs. Patent scope governs the market price in the proprietary market, and licens-
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c
+
+---
+
+Intellectual Property
+
+65
+
+ing prevents wasteful imitation. In this environment where goods are homogeneous, licensing determines the design of patent policy: If licensing is available, a case can be made for narrow and long patents; if licensing is not available, the analysis points to patents that are broad and short.
+
+Licensing may not occur for a variety of reasons, in which case we need a more thorough investigation of the relative merits of the GS and the Gallini arguments, in broader economic environments than they address. Such an analysis has been provided by Denicolò (1996) . He explains that narrow (and long) or broad (and short) patents depend on the concavity or convexity, respectively, of the relationship between social welfare and postinnovation profit. Situations in which relatively short broad patents are optimal include costly imitation; Cournot duopoly with constant marginal costs; and horizontally differentiated firms and linear transportation costs, as in Klemperer (1990) .
+
+We now turn to cumulative innovation in which subsequent research activity is directed toward the development of improvements or applications of a previous innovation.
+
+## IV. Optimal Design: The Case of Cumulative Innovation
+
+In the above discussion, IP is designed for isolated innovations that may be imitated. In reality, research is cumulative. Innovations build upon each other, and subsequent research activity is directed toward improvements or applications of previous discoveries. This fact changes the problem of patent design in interesting and complex ways.
+
+The first and most fundamental complexity, articulated by Scotchmer (1991) , is that early innovators lay a foundation for later innovations. The later innovations could not be made without the earlier ones. So that the first innovator has enough incentive to invest, she should be given some claim on profit of the later innovations; otherwise, early innovators could be underrewarded for the social value they create. This is particularly evident in the case of a research tool for which all the social value resides in the innovations it facilitates. If the innovator could not profit from the later products, she would have no incentive to create the tool. The incentive problems are particularly vexed in the case of creative destruction , discussed by Schumpeter (1943) : an innovator's descendants can actually become the instruments of his destruction.
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+66
+
+Gallini and Schotchmer
+
+The Schumpeterian perspective highlights an important problem that arises in the cumulative context: that of dividing the profit between innovators in a way that respects their costs. If, for example, only one pot of money is available for distribution between two innovators and most is allocated to the first firm, the second inventor's incentive for research is reduced, and vice versa. Green and Scotchmer (1995) argue that because of the difficulties in dividing profit, patent lives will have to be longer than if the whole sequence of innovations occurred in a single firm. Ex ante licensing—licensing before investments are made—is a way of mimicking the latter outcome. As in the case of a single invention, the availability of private contracting influences the optimal patent scope when innovation is cumulative (see below).
+
+Cumulativeness changes the design instruments that are relevant to the length of protection. The statutory life can be irrelevant when a noninfringing substitute, such as an improvement, can displace a protected product. What matters is the effective life, that is, the time until the noninfringing substitute appears (Scotchmer 1991; O'Donoghue, Scotchmer, and Thisse 1998 (OST)). The effective life is determined by patent scope or leading breadth , which is interpreted as the minimum quality improvement that avoids infringement. As in the case of costly imitation discussed above, the effectiveness of patent life as an instrument for R & D may be limited when subsequent innovation can undermine profitability.
+
+Finally, cumulativeness makes a third instrument—the minimum standard for protection, or minimum inventive step—relevant to the optimal design of IP. For copyrighted works the standard for protection is low (as is the breadth of protection), while for patents, the patentability standard (or novelty requirement) can be quite stringent. In our discussion of isolated inventions above, we assumed that the invention was protectable, since there would be no incentives to innovate if there were no IP or other incentive instruments. But in the cumulative context, patentability on second-generation inventions is less essential, since an innovation can be protected by an exclusive license on a previous patent it infringes, rather than by its own patent. Leading breadth and the standard for patentability together determine the level of “forward protection” each innovation has.
+
+Several arguments favoring both weak and strict standards for IP protection have been advanced. Scotchmer and Green (1990) argue with caution for a weak standard (a weak “ novelty requirement ” ), so that firms are encouraged to disclose every small bit of progress. While
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+67
+
+these disclosures could speed up invention by giving a technological boost to competitors, Scotchmer and Green warn that the weak novelty requirement could also encourage firms to choose trade secrecy over patents. In contrast, a tightening of the standards for patentability can encourage firms to be more ambitious in the improvements they attempt to develop (O’Donoghue 1998) or can direct their investments toward more socially useful inventions (Eswaran and Gallini 1996) . Even when the standard for protection does not reorient research efforts, it can affect the division of profit among sequential researchers. Scotchmer (1996) argues that the strictest novelty requirement (no protection) on second-generation products would tilt the joint profit of a sequence of innovations in favor of earlier innovators without jeopardizing second-generation advances. A second-generation product can be protected by an exclusive license on the infringed patent of the earlier generation. Denicolò (2000a, b) makes a case for a patent policy with a weak patentability standard and narrow leading breadth. In a model in which firms race for the first- and second-generation patents, he shows that tilting profits in favor of earlier innovators might only encourage a socially wasteful patent race at the stage of basic research and underinvestment in the second stage.
+
+Although the complexities of cumulativeness seem to defy clear, unqualified design implications, one lesson is clear: The optimal design of IP depends importantly on the ease with which rights holders can contract around conflicts in rights. Contracting is especially relevant to the question of breadth, which determines the likelihood that a follow-on innovation will infringe a prior patent.
+
+A danger of IP that has been debated from its inception to the present (see Machlup and Penrose 1950) is that IP can stifle innovation and slow progress. Merges and Nelson (1990) link this danger to breadth, using examples from the aircraft, radio, and pharmaceutical industries to argue for narrow patents. An earlier example concerned steam engines. James Watt refused to license his patents for improvement, with the result that there was a flood of pent-up invention when his patents expired in 1800 (Derry and Williams 1993, p. 324).
+
+In contrast, Kitch (1977) argues that broad patents are socially beneficial precisely because they stimulate further developments. Scotchmer (1991) and Green and Scotchmer (1995) take the same point of view, but focus on how ex ante contracting affects division of profit. With ex ante contracting, the role of breadth is not to determine whether subsequent products are made (they will be made if they add to joint profit), but rather to determine how the profit is divided. This
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+68
+
+Gallini and Schotchmer
+
+theme is also carried forward in later papers, e.g., Merges (1998, 1999), Scotchmer (1996), Lemley (1997) (who compares how copyright and patent doctrines respectively treat the possibility of blocking), O'Donoghue, Scotchmer, and Thisse (1998), and Schankerman and Scotchmer (2001). Matutes, Regibeau, and Rockett (1996) and Chang (1995) argue for broad patents even without assuming that ex ante contracts can be made.
+
+To some extent, broad patents are also supported by the arguments of O'Donoghue, Scotchmer, and Thisse (1998), who study breadth in a model with with an infinite sequence of improved products (quality ladder). If patents are relatively narrow, the effective life of each patent ends when a noninfringing improvement arrives, and is thus endogenous. But if the patent is broad, then the statutory life is also the effective life: Every subsequent innovation on the quality ladder infringes during the statutory life and must be marketed under license. To achieve the same rate of progress under both regimes, the effective patent life with a narrow patent must be longer than the (effective) statutory life with a broad patent. Broad, short patents are more efficient at rewarding innovators along the quality ladder, because less of the total profit in the system accrues to high-value innovations that would be made in any case, and more goes to the innovators who need additional incentives.
+
+Thus, with some caution, we can extract from the literature a case for broad (and short) patents. Broad patents can serve the public interest by preventing duplication of R & D costs, facilitating the development of second-generation products, and protecting early innovators who lay a foundation for later innovators. However, these benefits disappear if licensing fails. Heller and Eisenberg (1998) argue that licensing will likely fail when researchers must negotiate multiple licenses, as now occurs in the biomedical industry. Mazzoleni and Nelson (1998) caution that these transaction costs may limit the use of contracts for coordinating innovations that follow from a broad patent.
+
+Another problem with licensing is that it can lessen competition both in innovation markets 9 and in product markets. It thus raises antitrust issues, even in the simpler context where there is no cumulative aspect. One of the difficult issues is determining whether an ex ante merger of research activities through licensing is efficient or inefficient from a social perspective. On the efficiency side, ex ante licensing can enable firms to avoid duplicated costs and to delegate efficiently, much as discussed in section II . But on the inefficiency side, ex ante licensing can retard progress, e.g., by nullifying the acceleration that would other-
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+69
+
+wise come from a patent race. See Gilbert and Sunshine (1995) for a discussion of these issues.$^{10}$
+
+The cumulative context raises another issue. Above we focused on the salutary effects of licensing, namely that ex ante licensing can ensure investment in infringing follow-on products that would add to joint profit. Turning this argument on its head, licensing can stifle noninfringing follow-on products that would detract from joint profit. Gallini and Winter (1985) analyze a situation where a potential competitor is licensed ex ante in order to dissuade him from investing in a noninfringing cost reduction that would have lowered prices in the market. Such licensing clearly reduces product market competition relative to what the Congress apparently intended in designing patent law. If such licensing occurred ex post to prevent production of the costreducing innovation after it had been developed, it would presumably be an antitrust violation. Chang (1995) analyzes precisely that type of ex post collusion and advocates a strict antitrust rule against collusion. For a discussion of how principles of competition policy might be formulated to distinguish ex ante licensing that is procompetitive from that which is anticompetitive, see Scotchmer (1998).
+
+Besen and Maskin (2000) argue that if firms do not license in a way that takes full advantage of their IP, e.g., because of antitrust restrictions, then licensing may reduce industry profits below those available without licensing, and the broad patents that support such licensing are counterproductive. In this sense, Besen and Makin's paper is consistent with the above observation that impediments to contracting may strengthen the case for narrow patents.
+
+In light of these qualifications, what conclusions can we make for patent design in the cumulative context? One interpretation is that, when research is cumulative, relatively broad patents may be efficient if ex ante contracting is available. However we prefer to be cautious; the jury is still out.
+
+What is conclusive is the importance of private contracting. Whether property rights are helpful or counterproductive in encouraging innovation depends on the ease with which innovators can enter into agreements for rearranging and exercising those rights, as constrained by the rules of antitrust law.
+
+## V. Conclusions
+
+In the past two decades, academic interest in the economics and law of intellectual property has exploded. The renewed interest has been
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+70
+
+Gallini and Schotchmer
+
+fueled by controversies surrounding new technologies, by international agreements, and by changes in the nature of protection, e.g., see Mazzoleni and Nelson (1998). It is generally thought that IP rights have been strengthened, but there is also evidence that some forms of IP, in particular, patents, have previously been ineffective (Cohen et al. 2000). Contrary to the apparent intent, strengthening of IP is thought by some commentators to impede research rather than to promote it (Heller and Eisenberg 1998). In this environment, economists have had much to say about both the optimal design of IP and the advisability of substituting other incentive mechanisms.
+
+Although it comes as no surprise that a property system has defects, we hope we have illuminated some offsetting virtues, and some circumstances where other mechanisms, such as prizes, fixed-price contracts, and auctions can dominate. Our main conclusions on the effectiveness of intellectual property are that:
+
+- 1. IP is probably the best mechanism for screening projects when value
+and cost are not observable by the sponsor, since the private value of IP
+reflects the social value, and firms automatically compare some meas-
+ure of value with the cost of innovation. In addition, IP encourages
+firms to accelerate progress, since the reward is conditional on success.
+Prizes could serve the same purposes if the size of the prize could be
+linked to the social value but without the deadweight loss of monopoly
+pricing.
+2. Neither IP nor prizes can aggregate the information that is decen-
+tralized among firms, and neither will be completely effective at dele-
+gating research effort efficiently. A procurement system that restricts
+prizes to certain firms, or differentiates prizes according to firms' rela-
+tive efficiencies, can improve on a simple prize system or patent sys-
+tem, but then there must be an ex ante negotiation to select the favored
+firms.
+For circumstances where IP is justified, we asked how the property right should be designed. Every IP regime has provisions on length, breadth, and the standard for protection. The economics literature on design of IP concerns the appropriate choice of these provisions. The optimal length, breadth, and standard for protection depend on the economic environment, e.g., the shape of the demand curve, the rate at which improvements to existing technologies are developed, or the relative costs of sequential innovators.
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+71
+
+How much flexibility is there in designing IP rights differently for different economic environments? In fact, there is a lot of flexibility. Different IP regimes are targeted at different subject matter, and the subject matter is an important defining aspect of the IP regime. Copyright has traditionally been targeted at literature, other printed matter, and art. Patents have traditionally been targeted at manufactured items. The subject matters of sui generis laws typically have been very specific, e.g., the Plant Patent Act, the Plant Variety Protection Act, the Semiconductor Chip Protection Act, and the proposed database legislation.
+
+The IP regimes that cover different subject matter are noticeably varied in the three important features: length, breadth, and standard for protection. On the matter of length, copyrights last essentially forever, patents last 20 years, and chip protection lasts 10 years. On the matter of breadth, copyright protection is restrained by fair use exemptions and by the fact that the underlying “ideas” are not protected; patents have the doctrine of equivalents; and copying of chips is allowed for some uses but not others. We thus believe that it is incorrect to criticize the economic design arguments on grounds that, in IP, “one size fits all.” While we do not think it would be appropriate to define new IP regimes for every small category of technology, we wish to emphasize that the Congress can exercise as much flexibility as it wishes, and that courts also have some flexibility.
+
+Each IP regime should cover subject matter with similar needs for protection, especially if heterogeneous needs cannot be remedied by courts. Many controversies arise because of heterogeneity within IP regimes. For example, business methods probably do not need the strong protection provided by the Patent Act, even though such protection is appropriate for other patentable subject matter. A new regime could have been created for business methods, but protection under the Patent Act could alternatively be weakened through the courts' interpretation of novelty and nonobviousness.
+
+Finally, there are the design recommendations themselves. We have not been specific in this review about the exact ways in which length, breadth, and standards for protection should reflect the economic environments, and refer the reader to the underlying papers for more detail. Instead, we have emphasized a message of a different sort: the optimal design of the property right should depend on whether firms contract with others for the use of their protected innovations. With fluid contracting, policies that otherwise would be inefficient may be
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+72
+
+Gallini and Schotchmer
+
+optimal. For example, licensing can avoid wasteful imitation, making an otherwise inefficient narrow patent optimal. In the cumulative context, there is a danger that broad patents will inhibit future innovators from making product improvements. But with contracting, the patentholder can profit from, instead of being threatened by, new improved products, and will ensure that they arise even if infringing. The most striking message of the literature is that IP and private instruments may be complementary in reducing social costs from an overreaching or insufficient protection regime.
+
+However, contracting also has the potential to undermine competition in ways that were not anticipated or approved by the Congress when designing IP. Contracting that we have not covered includes cross-licensing and patent pools. We have also not discussed joint ventures and other alliances for avoiding litigation, duplicated efforts, and holdups. A recurring theme, especially evident in these contexts, is that despite the efficiencies that contracting can ensure, contracting may also facilitate anticompetitive behavior. See Hall and Ziedonis (2001), Shapiro (2000), Denicolò (2000a, b). To understand whether the property system is too strong, too weak, or necessary at all requires us to understand the incentives for contracting, and its potential anticompetitive consequences.
+
+## Notes
+
+This article will also appear in Legal Orderings and Economic Institutions, F. Cafaggi, A. Nicita and U. Pagano, eds., Routledge Studies in Political Economy (2002).
+
+- 1. For example, in the environment discussed by Wright (1983) , none of the three mecha-
+nisms is optimal. The first best can be achieved with a mechanism similar to the one men-
+tioned in note 3 below.
+2. Prizes might also require enforcement. John Harrison's longitude prize was delayed
+for decades while the prize committee attempted to prove that astronomical solutions
+were superior to his clock. Harrison eventually sought redress in Parliament, and was
+partially rewarded.
+3. She suggests a two-part procedure. The sponsor first asks the firms to reveal their in-
+formation on value and then, if warranted, employs the best procurement mechanism to
+delegate to the least-cost firm(s). Following Cremer and McLean (1988) , it is costless to
+get the firms to reveal their correlated information on value. They are asked to report
+their signals of value, and then rewarded if they agree and punished if they disagree.
+Due to the correlation, an equilibrium is to report truthfully, and the payments can be
+chosen so that each firm makes zero expected profit.
+4. A related problem is studied by Bhattacharya et al. (1998) . Instead of assuming that
+firms have different efficiency parameters, they assume that firms have different knowl-
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+73
+
+edge about the cost of achieving an innovation. If the knowledge is revealed, then all firms have the same cost. Their mechanisms also use payments conditional on delivery (prizes).
+
+5. We caution, however, that the Kremer scheme is efficient only if there is a single researcher. A prize equal to the social value could easily attract other firms to a race in which the firms overinvest (Loury 1979). Not only is there a problem of overinvesting, but inefficient firms as well as efficient firms may invest. This is the problem avoided by the more complex procurement mechanism discussed above, where prizes are tailored to the firms' relative efficiency in order to make sure that the investment effort is undertaken by the more efficient ones.
+
+6. In deLaat's model, the sponsor chooses the “size of the invention,” which is observable, given the firm's report of the market size (value), which is unobservable to the sponsor. DeLaat assumes that the sponsor can verify which invention is made, but not the market conditions (e.g., demand) for the invention.
+
+7. For recent empirical investigations of how firms exercise their option to renew, and implications for the values of innovations, see Lanjouw (1997) and Schankerman (1998).
+
+8. The intuition for this result can be found in the familiar economic principle that underlies Ramsey pricing. Ramsey pricing solves the problem of maximizing consumer surplus in multiple markets subject to the constraint that revenues cover cost. The solution is to set prices below monopoly prices so that the markup of price in each market is inversely proportional to the elasticity of demand in each market. In the patent problem, the different time periods are parallel to different markets, and since the demands are assumed to be constant over time, the markup of price over cost in each period is identical.
+
+9. See the U.S. Department of Justice-Federal Trade Commission Antitrust Guidelines for the Licensing of Intellectual Property (1995) for a discussion of innovation markets.
+
+10. One of the thorny questions that arise is whether competition policy should view licensing practices more leniently than otherwise if incentives to innovate are at stake. See Gallini and Trebilcock (1998) for a discussion of this issue.
+
+## References
+
+Anton, J., and D. Yao. 1994. "Expropriation and Inventions—Appropriable Rents in the Absence of Property Rights." American Economic Review 84(1): 191-209.
+
+Arrow, K. J. 1962. "Economic Welfare and the Allocation of Resources for Invention," in Universities-National Bureau of Economic Research Conference. Series, The Rate and Direction of Economic Activities: Economic and Social Factors. Princeton: Princeton University Press.
+
+Besen, James, and Eric Maskin. 2000, January. "Sequential Innovation, Patents and Imitation." Working Paper no. 00-01. MIT.
+
+S. Battacharya, D'Aspremont, C., and L.-A. Gerard-Varet. 1998. "Knowledge as a Public Good: Efficient Sharing and Incentives for Development Effort." Journal of Mathematical Economics 30: 389-404.
+
+Calandrillo, Steve. 1998. "An Economic Analysis of Intellectual Property Rights: Justifications and Problems of Exclusive Rights, Incentives to Generate Information, and
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+74
+
+Gallini and Schotchmer
+
+the Alternative of a Government-Run Reward System." Fordham Intellectual Property Media and Entertainment Law Journal 9: 301-348.
+
+Cohen, Wesley M., Richard R. Nelson and John P. Walsh (2000), "Protecting their Intellectual Assets: Appropriability Conditions and Why U.S. Manufacturing Firms Patent (or Not)", National Bureau of Economic Research Working Paper 7552.
+
+Chang, H. 1995. "Patent Scope, Antitrust Policy, and Cumulative Innovation." RAND Journal of Economics 26: 34-57.
+
+Cornelli, F., and Mark S. Schankerman. 1999. "Patent Renewals and R&D Incentives." RAND Journal of Economics 30(2).
+
+Cremer, J., and R. P. McLean. 1988. "Full Extraction of the Surplus in Bayesian and Dominant Strategy Auctions." Econometrica 56(6): 1247-57.
+
+De Laat, Eric A. A. 1996. "Patents or Prizes: Monopolistic R&D and Asymmetric Information." International Journal of Industrial Organization 15(3): 369-90.
+
+Denicolò, V. 1996. "Patent Races and Optimal Patent Breadth and Length." Journal of Industrial Economics 44: 249-265.
+
+Denicolò, V. 2000a. "Two Stage Patent Races and Patent Policy." RAND Journal of Economics 31: 488-501.
+
+Denicolò, V. 2000b. "Combination of Competing Patents and Antitrust Policy." Working paper. Department of Economics, University of Bologna.
+
+Derry, T. K., and Williams, Trevor I. 1993. A Short History of Technology: From Earliest Times to A.D. 1900. Mineola, NY: Dover.
+
+Eswaran, M. and N. Gallini. 1996. "Patent Policy and the Direction of Technological Change," RAND Journal of Economics 27(4): 722-746.
+
+Farrell, J. 1995. "Arguments for Weaker Intellectual Property Protection in Network Industries." In Brian Kahin ed., Standards Policy for Information Infrastructure. 368-377. Cambridge, MA: MIT Press.
+
+Gallini N. T. 1984. "Deterrence by Market Sharing: A Strategic Incentive for Licensing." American Economic Review 74: 931-41.
+
+Gallini, N. T. 1992. "Patent Length and Breadth with Costly Imitation." RAND Journal of Economics 44: 52-63.
+
+Gallini, N., and M. Trebilcock. 1998. "Intellectual Property Rights and Competition Policy—A Framework for the Analysis of Economic and Legal Issues." In R. Anderson and N. Gallini, eds., Competition Policy and Intellectual Property Rights in the Knowledge-Based Economy. Industry Canada Research Series, University of Calgary Press: 17-64.
+
+Gallini, N. T., and R. Winter. 1985. "Licensing in the Theory of Innovation." RAND Journal of Economics 16: 237-52.
+
+Gandal, N., and S. Scotchmer. 1993. "Coordinating Research through Research Joint Ventures." Journal of Public Economics, 51: 173-93.
+
+Gilbert, R., and C. Shapiro. 1990. "Optimal Patent Length and Breadth." RAND Journal of Economics 21: 106-12.
+
+Gilbert, R., and G. C. Sunshine. 1995. "Incorporating Dynamic Efficiency Concerns in Merger Analysis: The Use of Innovation Markets." Antitrust Law Journal 63: 569-602.
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+75
+
+Green, Jerry, and Suzanne Scotchmer. 1995. "On the Division of Profit in Sequential Innovation." RAND Journal of Economics, 26: 20-33.
+
+Hall, B. H. and R. Ham Ziedonis. 2001. "The Patent Paradox Revisited: An Empirical Study of Patenting in the U.S. Semiconductor Industry, 1979-1995," RAND Journal of Economics 32(1): 101-128.
+
+Heller, Michael A., and R. S. Eisenberg. 1998. "Can Patents Deter Innovation? The Anticommons in Biomedical Research." Science 280: 698-701.
+
+Kitch, E. W. 1977. "The Nature and Function of the Patent System." Journal of Law and Economics 20: 265-90.
+
+Klemperer, P. 1990. "How Broad Should the Scope of Patent Protection Be?" RAND Journal of Economics 21: 113-30.
+
+Kremer, Michael. 1989. "Patent Buyouts: A Mechanism for Encouraging Innovation." Quarterly Journal of Economics 113: 1137-67.
+
+Kremer, Michael. 2000. "Creating Markets for New Vaccines." Conference Paper, Innovation Policy and the Economy. Cambridge, MA: National Bureau of Economic Research.
+
+Laffont, J.-J., and J. Tirole. 1986. "Using Cost Observation to Regulate Firms." Journal of Political Economy 94: 614-41.
+
+Laffont, J.-J., and J. Tirole. 1987. "Auctioning Incentive Contracts." Journal of Political Economy 95: 921-937.
+
+Lanjouw, J. 1998. "Patent Protection in the Shadow of Litigation: Simulation Estimations of Patent Value." Review of Economic Studies. Vol 65, 671-710
+
+Lemley, M. A. 1997. "The Economics of Improvement in Intellectual Property." Texas Law Review 75: 989.
+
+Loury, G. C. 1979. "Market Structure and Innovation." Quarterly Journal of Economics 93: 395-410.
+
+Machlup, Fritz, and Edith Penrose. 1950. "The Patent Controversy in the Nineteenth Century." Journal of Economic History 10(1): 2.
+
+Matutes, C., P. Regibeau, and K. E. Rockett. 1996. "Optimal Patent Protection and the Diffusion of Innovation." Rand Journal of Economics 27: 60-83.
+
+Maurer, S. 1999. "Raw Knowledge: Protecting Technical Databases for Science & Industry." Report commissioned by the National Academy of Sciences. www.nap.edu/html/proceedings_sci_tech/appC.html
+
+Maurer, S. M., and S. Scotchmer. 1998. "The Independent Invention Defense in Intellectual Property." John M. Olin Working Paper no. 98-11. Boalt School of Law, University of California, Berkeley.
+
+Maurer, S. M., and S. Scotchmer. 1999. "Database Protection: Is it Broken and Should We Fix it?" Science 284: 1129-1130.
+
+Mazzoleni, Roberto, and Richard R. Nelson. 1998. "The Benefits and Costs of Strong Patent Protection: A Contribution to the Current Debate." Research Policy 27: 273-84.
+
+Merges, R. P. 1998. "Antitrust Review of Patent Acquisitions: Property Rights, Firm Boundaries, and Organization." In Robert D. Anderson and Nancy T. Gallini, eds., Com-
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+76
+
+Gallini and Schotchmer
+
+petition Policy and Intellectual Property Rights in the Knowledge-Based Economy. Calgary: University of Calgary Press.
+
+Merges, R. P. 1999. "Intellectual Property Rights, Input Markets and the Value of Intangible Assets." Working Paper. Berkeley: University of California.
+
+Merges, R. P., and R. R. Nelson. 1990. "On the Complex Economics of Patent Scope." Columbia Law Review 90(4): 839-916.
+
+Minehart, D., and S. Scotchmer. 1999. "Ex Post Regret and the Decentralized Sharing of Information." Games and Economic Behavior 27(1): 114-131.
+
+National Research Council. 2000. The Digital Dilemma: Intellectual Property in the Information Age. Committee on Intellectual Property Rights and the Emerging Information Infrastructure, Computer Science and Telecommunications Board. Washington DC: National Academy Press.
+
+National Science Foundation. 2000. Science and Engineering Indicators 2000. www.nsf.gov/sbe/srs/seind00/frames.htm
+
+Nordhaus, W. 1969. Invention, Growth and Welfare: A Theoretical Treatment of Technological Change. Cambridge, MA: MIT Press.
+
+O'Donoghue, T. 1998. "A Patentability Requirement for Sequential Innovation." Rand Journal of Economics 29: 654-67.
+
+O'Donoghue, T., S. Scotchmer, and J. F. Thisse. 1998. "Patent Breadth, Patent Length and the Pace of Technological Progress." Journal of Economics and Management Strategy 7: 1-32.
+
+Reynolds, T. 1983. Stronger than a Hundred Men: A History of the Vertical Water Wheel. Baltimore: Johns Hopkins Press.
+
+Schankerman, M. 1998. "How Valuable is Patent Protection? Estimates by Technology Field." RAND Journal of Economics 29: 77-107.
+
+Schankerman, M., and S. Scotchmer. 2001. "Damages and Injunctions in Protecting Intellectual Property." RAND Journal of Economics 32: 199-220.
+
+Schumpeter, J. 1943. Capitalism, Socialism and Democracy. London: Unwin University Books.
+
+Scotchmer, S. 1991. "Standing on the Shoulders of Giants: Cumulative Research and the Patent Law." Symposium on Intellectual Property Law. Journal of Economic Perspectives, 5(1): 29-41.
+
+Scotchmer, S. 1996. "Protecting Early Innovators: Should Second-Generation Products Be Patentable?" RAND Journal of Economics 27: 322-31.
+
+Scotchmer, S. 1998. "R&D Joint Ventures and other Cooperative Arrangements." In Competition Policy and Intellectual Property Rights in the Knowledge-Based Economy, R. Anderson and N. Gallini, eds. Industry Canada Research Series. University of Calgary Press: 203-22.
+
+Scotchmer, S. 1999a. "On the Optimality of the Patent Renewal System." RAND Journal of Economics 30, 181-196.
+
+Scotchmer, S. 1999b. "Delegating Effort in a Common Value Project." IBER Working Paper no. E99-266. Berkeley: University of California.
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+
+---
+
+Intellectual Property
+
+77
+
+Scotchmer, S., and J. Green 1990. "Novelty and Disclosure in Patent Law." RAND Journal of Economics 21: 131-46.
+
+Shapiro, C. 2000. "Navigating the Patent Thicket: Cross Licenses, Patent Pools, and Standard-Setting." Working Paper. Berkeley: University of California Competition Policy Center. Working Paper No. CPC00-11. Now published in Innovation Policy and the Economy, Vol. 1, MIT Press, 2001. Adam B. Jaffe, Josh Lerner, and Scott Stern, eds.
+
+Shavell, S., and T. van Ypserle 1998. "Rewards versus Intellectual Property Rights." Olin Discussion Paper no. 246. Harvard University.
+
+Sobel, D. 1995. Longitude: The True Story of a Lone Genius Who Solved the Greatest Scientific Problem of His Time. New York: Walker.
+
+U.S. Department of Justice and Federal Trade Commission. 1995. U.S. Department of Justice-Federal Trade Commission Antitrust Guidelines for The Licensing of Intellectual Property.
+
+Wright, B. 1983. "The Economics of Invention Incentives: Patents, Prizes and Research Contracts." American Economic Review, 73, 691-707.
+
+This content downloaded from 138.186.079.110 on July 25, 2016 21:02:24 PM
+
+
+All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c)
+
+
+1
+

processed/00019_W2152132348/document.md

@@ -0,0 +1,719 @@
+Massachusetts Institute of Technology Department of Economics Working Paper Series Harvard University Department of Economics Working Paper Series Working Paper No. Forthcoming
+
+# Sequential Innovation, Patents, and Imitation
+
+James Bessen and Eric Maskin
+
+Working Paper 00-01 January 2000
+
+Room E52-251 50 Memorial Drive Cambridge, MA 02142
+
+This paper can be downloaded without charge from the Social Science Research Network Paper Collection at http://papers.ssrn.com/paper.taf?abstract_id=206189
+
+---
+
+# Sequential Innovation, Patents, and Imitation*
+
+by
+
+James Bessen and Eric Maskin © 1997, 1999 Working Paper 11/99
+
+Abstract: How could such industries as software, semiconductors, and computers have been so innovative despite historically weak patent protection? We argue that if innovation is both sequential and complementary — as it certainly has been in those industries — competition can increase firms' future profits thus offsetting short-term dissipation of rents. A simple model also shows that in such a dynamic industry, patent protection may reduce overall innovation and social welfare. The natural experiment that occurred when patent protection was extended to software in the 1980's provides a test of this model. Standard arguments would predict that R & D intensity and productivity should have increased among patenting firms. Consistent with our model, however, these increases did not occur. Other evidence supporting our model includes a distinctive pattern of cross-licensing in these industries and a positive relationship between rates of innovation and firm entry.
+
+James Bessen Research on Innovation jbessen@researchoninnovation.org
+
+Eric Maskin Harvard University and Massachusetts Institute of Technology emaskin@mit.edu
+
+*This research was supported in part by a grant from the NSF.
+
+---
+
+2 - Sequential innovation, patents and imitation - 11/99
+
+## 1. Introduction
+
+The standard economic rationale for patents is to protect potential innovators from imitation and thereby give them the incentive to incur the cost of innovation. Conventional wisdom holds that, unless would-be competitors are constrained from imitating an invention, the inventor may not reap enough profit to cover that cost. Thus, even if the social benefit of invention exceeds the cost, the potential innovator without patent protection may decide against innovating altogether. $^1$
+
+Yet, interestingly, some of the most innovative industries today — software, computers, and semiconductors — have historically had weak patent protection and have experienced rapid imitation of their products. 2 Defenders of patents may counter that, had stronger intellectual property protection been available, these industries would have been even more dynamic. But we will argue that the evidence suggests otherwise.
+
+In fact, the software industry in the United States was subjected to a revealing natural experiment in the 1980's. Through a sequence of court decisions, patent protection for computer programs was significantly strengthened. We will show that, far from unleashing a flurry of new innovative activity, these stronger property rights ushered in a period of stagnant, if not declining, R&D among those industries and firms that patented most.
+
+We maintain, furthermore, that there was nothing paradoxical about this outcome. For industries like software or computers, there is actually good reason to believe that imitation promotes innovation and that strong patents (long patents of broad scope) inhibit it. Society might be well served if such industries had only limited intellectual property protection. Moreover, many firms might genuinely welcome competition and the prospect of being imitated. $^3$
+
+This is because these are industries in which innovation is both sequential and complementary . By “sequential,” we mean that each successive invention builds on the preceding one—in the way that Windows built on DOS. And by “complementary,” we mean that each potential innovator takes a somewhat different research line and thereby enhances the overall probability that a particular goal
+
+1 This is not the only justification for patents, as we will discuss in Section 2. But it is the rationale most commonly advanced. Also, patents provide a spillover benefit through their disclosure requirements. We focus here, as does the traditional literature, on the capacity of patents to block competitors.
+
+2Software was routinely considered excluded from patent protection until court decisions in the late 80's. Semiconductor and computer patent enforcement was very uneven until the organization of the Federal Circuit court in 1982. Both areas contend with substantial problems of prior art [Aharonian, OTA, 1992], and some experts contend that as many as 90% of semiconductor patents are not truly novel and therefore invalid [Taylor and Silberston, 1973]. These problems make consistent enforcement difficult.
+
+Surveys of managers in semiconductors and computers consistently report that patents only weakly protect and incent innovation. Levin et al [1987] found that patents were rated weak at protecting the returns to innovation, far behind the protection gained through lead time and learning curve advantages. Patents in electronics industries were estimated to increase imitation costs by only 7% [Mansfield, Schwartz and Wagner] and 7 - 15% [Levin, et al]. Taylor and Silberston [1973] found that very little R&D was performed to take advantage of patent rights.
+
+As one might expect, diffusion and imitation are rampant in these industries. Firms formally and informally exchange information themselves, employees frequently move from one firm to another, and spin-off firms are common. More important, imitation lags are brief. Tilton [1971] estimated that the time from initial discovery to commercial imitation in Japan averaged just over one year in semiconductors in the 60's. In software, imitation lags are sometimes shorter.
+
+3 When IBM announced its first personal computer in 1981, Apple Computer, then the leading personal computer-maker, responded with full-page newspaper ads headlined "Welcome IBM. Seriously." A high tech cliché is that competition "expands the market."
+
+---
+
+3 - Sequential innovation, patents and imitation - 11/99
+
+is reached within a given time. $^4$ Undoubtedly, the many different approaches taken to voicerecognition software, for example, hastened the availability of commercially viable packages.
+
+The sequential and complementary nature of innovation is widely recognized, especially in high tech industries. Gort and Klepper [1982] found an average of 19 sequential improvements to 23 major innovations and many more uncounted improvements. Analysis of many innovations has found that most of the productivity gain is achieved via improvements to the original innovation (see for example, Enos [1962] ). Scotchmer [1996] , Green and Scotchmer [1995] , and Chang [1995] have studied the theoretical implications of sequential innovation. A variety of empirical studies have found strong evidence of innovative complementarities, including Jaffe [1986] and Henderson and Cockburn [1996] (see a review of the literature in Griliches [1992] ). Spence [1984] and Levin and Reiss [1988] have studied the effects of spillover complementarities in a nonsequential environment.
+
+A firm that patents its product in a world of sequential and complementary innovation can prevent its competitors from using that product (or sufficiently similar ideas) to develop further innovations. And because these competitors may have valuable ideas not available to the original firm about how to achieve such innovations, the patent may therefore slow down the pace of invention.
+
+Of course, patent traditionalists have a counter-argument to this train of logic too: if a jealously guarded patent seriously interferes with innovative activity, the patent-holder can simply license it (thereby allowing innovation to occur). After all, presumably he could capture the value added of such innovation by an appropriately chosen licensing fee (or so the argument goes). The flaw in this argument is that licensing creates competition and, as we will show in Section 3 , the profitdissipating effect of this competition may well outweigh the additional profit created by greater innovation. In such a case, the patent-holder would choose not to license, even though society would suffer from that decision.
+
+But whether or not patent protection is available, a firm may well be better off if other firms imitate and compete with it. Although imitation reduces the firm's current profit, it raises the probability of further innovation and thereby improves the prospect that this firm will make another profitable discovery later on.
+
+In short, when innovation is sequential and complementary, standard reasoning about patents and imitation may get turned on its head. Imitation becomes a spur to innovation, while strong patents become an impediment.
+
+We will proceed as follows. In Section 2 we will review the static model that underlies the traditional justification for patents. We show that, besides helping to ensure that innovative firms cover their costs, patents can also encourage innovative activity on the part of firms that otherwise would be more inclined to imitation. Then in Section 3 we turn to a dynamic model and delineate the circumstances in which patents inhibit innovation and firms are made better off by being initiated. Finally, in Section 4 we discuss the evidence for the pertinence of this dynamic model to high tech industries.
+
+4 One example of complementarity treated in the literature is that of information spillovers, following Arrow [1962] . But we do not limit ourselves to this example.
+
+---
+
+4 - Sequential innovation, patents and imitation - 11/99
+
+## 2. The static model
+
+We consider an industry consisting of two ( ex ante symmetric) firms. $^5$ Each firm can undertake R&D to discover and develop an innovation with expected (social) value $v$ . The cost of R&D is $c$ and, if a single firm incurs that cost, the probability of successful innovation is $p$ . $^6$
+
+We assume that if a firm is not copied, it can capture the social value of its innovation. $^7$ However, the other firm, unless prevented by patent protection, can costlessly develop an imitation. And, if it does so, the innovating firm obtains only a fraction $s$ ( $s < \frac{1}{2}$ ) of the value of $v$ (to simplify matters, we assume that the imitating firm captures this same fraction $s$ ). $^8$
+
+We suppose that the expected social value of undertaking R&D is positive:
+
+$$(1)\quad p\cdot  v-c>0.$$
+
+However, the net payoff from R\&D to a firm that anticipates being imitated is only $s \cdot  p \cdot  v - c$. And, even if (1) holds, we may well have
+
+$$(2)\quad s\cdot  p\cdot  v-c<0.$$
+
+The combination of (1) and (2) represents the classic incentive failure that the patent system is meant to address. Formula (2) captures the idea that, without patent protection, a firm cannot make money on R\&D investment when its potential innovation is imitable. A patent proscribes imitation and, therefore, guarantees an innovator the full net social return on R\&D expenditure ( $p \cdot  v - c$ ). Formula (1) then tells us that, as long as society gains from R\&D investment, so will the innovator himself.
+
+But even in a setting where (2) does not hold — so that R & D remains profitable despite imitation — patents may well serve a useful purpose. This is because they can encourage several firms to go after the same innovation. Typically, different firms have different ideas about how to solve a particular technological problem. Therefore, increasing the number of firms in pursuit of a solution raises the probability that someone will succeed; this is what we called “ complementarity ” in the introduction.
+
+5 Limiting the model to two firms is a matter of expositional convenience only; all our results extend to three or more firms. Indeed, we will see below that our arguments about the inefficiency of patents and the drawback of mergers as a corrective to that inefficiency become stronger with more than two firms.
+
+6 Our setting in this section is static but, if it is viewed as the reduced form of a dynamic model, $p$ could also be interpreted as the discount factor corresponding to the time lag to innovation.
+
+7 This is, of course, a strong assumption. However, the incentive failure and inefficiency from monopoly that arise when it is not satisfied are already well understood. The assumption is a simple way to abstract from these familiar distortions.
+
+8 The assumptions that imitation is costless and that an imitator enjoys the same share of profit as the innovator (i.e., the innovator has no first-mover advantage) are, of course, unrealistic. By invoking them, however, we are strengthening the case for patents. This will bolster our argument in Section 3 where we point out the shortcomings of the patent system.
+
+---
+
+5 - Sequential innovation, patents and imitation - 11/99
+
+We model the complementarity by assuming that, if both firms undertake R&D, the event that firm 1 is successful is statistically independent of that of firm 2.9 Hence, the total probability of successful innovation with two firms investing is
+
+$$1-(1-p)^{2}=2p-p^{2},$$
+
+and the social net benefit of this investment is
+
+$$(3)\quad \left(2p-p^{2}\right)v-2c\ .$$
+
+Formula (3) implies that the marginal social benefit of the second firm's R\&D expenditure is $(2p-p^2)v-2c-(pv-c)$. Let us suppose this is positive, i.e.,
+
+$$(4)\quad \left(p-p^{2}\right) v-c>0\ .$$
+
+Now, in the absence of patent protection, a firm's expected profit if both firms undertake R&D is
+
+$$(5)\quad s\left(2 p-p^{2}\right) v-c$$
+
+($2p-p^2$ is the probability of successful innovation, after which each firm enjoys a payoff of $sv$). But even if (5) is positive (and (4) implies that it will be, provided that $s$ is not too much smaller than 1⁄2), an equilibrium in which both firms invest may not be sustainable. Indeed, if one firm undertakes R\&D, the other will do so too only if (5) exceeds
+
+$$(6)\quad p\cdot  s\cdot  v,$$
+
+the expected payoff from free-riding on the innovative firm's R&D activity. That is, even when (5) is positive, only one firm will invest if
+
+$$(7)\quad s(p-p^{2})v-c<0.$$
+
+With the prospect of patent protection, by contrast, a firm undertaking R&D can expect a payoff of
+
+$$(8)\quad \frac{1}{2}(2p-p^{2})v-c,$$
+
+if the other invests as well (we assume that even if both firms make the discovery, only one obtains the patent, $^10$ and that each has an equal chance of this). Note that (8) is positive as long as (4) holds. Thus, if it is socially desirable for a second firm to invest, patent protection will induce it to do so, whereas, without such protection, it might well merely imitate. Patents accomplish more,
+
+9 More realistically, the techniques available to each firm might be correlated to some degree, leading to correlation between the two firms' chances of success. We assume statistical independence (no correlation) only for convenience.
+
+10 This gets at the idea that patents have breadth, and so a patent-holder can hold up the implementation of other firms' discoveries that are similar, but not identical, to his own.
+
+---
+
+6 - Sequential innovation, patents and imitation - 11/99
+
+therefore, than merely protecting innovators from imitation. Indeed, they create a risk of overinvestment: (8) could be positive even if (4) fails to hold. $^11$
+
+We can summarize our observations in the following result (which is also illustrated in Figure 1):
+
+Proposition 1: In the above (static) model, patent protection gives rise to at least as much R&D investment 12 (and hence innovation) as in either (a) a regime without patent protection or (b) the social optimum.
+
+Observe that the possible over-investment in R & D induced by patents could, in principle, be avoided if there were no complementarities of research across firms. Specifically, one could imagine awarding a firm an ex ante patent, e.g., the right to develop a vaccine against a particular disease. 13 Such protection would, of course, serve to deter additional firms from attempting to develop the innovation in question. And this would be efficient, provided that these other firms would not enhance the probability or speed of development, i.e., provided that they conferred no complementarity.
+
+Notice that there is no room for patent licensing in this static model. A patent-holder derives a payoff $v$ . Were he to license to his competitor, the total industry payoff would be $2sv$ , which is less than $v$ .
+
+This simple static model thus captures basic results of patent race models such as Loury [1979] and Dasgupta and Stiglitz [1980] . It also captures aspects of static models that consider spillover complementarities such as Spence [1984] , which emphasizes the socially redundant R & D that can occur under patents. These results, however, change sharply when dynamic considerations are introduced.
+
+## 3. Dynamic model
+
+Let us now enrich the model to accommodate a sequence of potential innovations, each of which builds on its immediate predecessor. More specifically, we will suppose that, for a firm to have a realistic chance of developing the innovation of the next generation, it must have market experience with that of the current generation. That is, it must be operating at the “state of the art.” The idea behind this assumption is that any new product or process would spring directly from those of the current generation, and so “hands on” experience with the latter is a prerequisite for innovation. For example, the firms surveyed by Levin et al [1987] (in more than one hundred manufacturing industries) reported that typically only a few other firms are capable of duplicating their innovations and that learning curves, lead time and sales and service efforts provide significant obstacles to imitation.
+
+11 The possibility that patents can give rise to excessive spending on R&D is well known from the patent-race literature; see Dasgupta and Stiglitz [1980] and Loury [1979].
+
+12 And possibly strictly more.
+
+$^{13}$ Wright [1983] similarly explores patent awards and government contracts.
+
+---
+
+7 - Sequential innovation, patents and imitation - 11/99
+
+Formally, consider an infinite sequence of potential innovations, each of which makes the previous generation obsolete and has incremental value $v$ over that generation. Given that the generation $t$ innovation has already been developed, a firm that incurs R & D cost $c$ and participates in the generation $t$ market has positive probability of discovering the generation $t+1$ innovation. If there is just one firm attempting to innovate, this probability is $p$ . If this firm undertakes R & D in every generation, the expected number of innovations is
+
+$$(9)\quad p(1-p)+2p^{2}(1-p)+3p^{3}(1-p)+...=\frac{p}{1-p},$$
+
+where the first term on the left-hand side of (9) is the expectation of one innovation, the second term is the expectation of two, etc. If an innovation fails to occur in any given generation, no further innovation is possible. The firm's overall expected profit and the social net benefit of R&D are both given by
+
+$$(10)\quad -c+p(1-p)(v-c)+2p^{2}(1-p)(v-c)...=-c+\frac{p}{1-p}(v-c)=\frac{p\cdot  v-c}{1-p},$$
+
+since we continue to assume that a monopolist can capture all the social surplus.
+
+If, for each innovation, there are two firms investing in R&D and participating in the market, then, given the same complementarity as in the static model, the expected number of innovations is
+
+$$\left(2 p-p^{2}\right)(1-p)^{2}+2\left(2 p-p^{2}\right)^{2}(1-p)^{2}+...=\frac{2 p-p^{2}}{(1-p)^{2}},$$
+
+and the social surplus contributed by the second firm is
+
+$$(11)\quad \frac{(2p-p^{2})v-2c}{(1-p)^{2}}-\frac{pv-c}{1-p}=\frac{pv-(1+p)c}{(1-p)^{2}}.$$
+
+Let us retain the assumption that, if one firm imitates the other's innovation, each enjoys a share $s$ of the total gross value. We first explore the nature of equilibrium when there is no patent protection.
+
+If one firm attempts to innovate in each generation, while the other only imitates, the innovator's overall expected profit is
+
+$$(12)\quad \frac{psv-c}{1-p}.$$
+
+However, the other firm will choose to imitate only if
+
+$$(13)\quad \frac{s(2p-p^{2})v-c}{(1-p)^{2}}<\frac{p sv}{1-p}.$$
+
+---
+
+8 - Sequential innovation, patents and imitation - 11/99
+
+The left-hand side of (13) is the firm's expected profit when both firms undertake R&D in every generation. The right-hand side is its profit if it always imitates. Formula (13) simplifies to
+
+$$(14)\quad p\cdot  s\cdot  v<c.^{14}$$
+
+Observe that (12) and (14) imply that neither firm will invest if (14) holds, and that both will invest if it fails to hold.
+
+Comparing (14) with its static-model counterpart (7) , we see that there is a greater incentive for a second firm to invest in the dynamic setting than in the static model (notice that this is true only for the second firm; the incentives for R & D are exactly the same in either model for the first firm). This is because the second firm's R & D in the dynamic model raises the probability, not only of the next innovation, but of subsequent innovations, and this is advantageous to the firm even if, subsequently, it merely imitates the first firm.
+
+Next we consider what happens under a patent system, assuming that each successive innovation is patentable and, for simplicity, each patent applies only to a single innovation. Although subsequent innovations are not covered by a given patent, each patent still allows patent holders to block entry to the subsequent markets — firms must participate in the original market in order to succeed with a follow-on product. That is, a patent conveys a hold-up right over subsequent innovations.
+
+With patents, at least one firm is willing to invest provided that (10) is positive, i.e.,
+
+$$(15)\quad \frac{v}{c}>\frac{1}{p}.$$
+
+Indeed, both firms will be willing to “ race ” to the first innovation (with perhaps the loser then dropping out altogether) provided that
+
+$$(16)\quad \frac{v}{c}>\frac{2-p^{2}}{2p-p^{2}}.$$
+
+But from (11), it is not efficient for both firms to invest in R\&D if
+
+$$(17)\quad \frac{1+p}{p}>\frac{v}{c}.$$
+
+And since the left-hand side of (17) exceeds the right-hand side of (16), we obtain, for this firstgeneration innovation, the standard result of excessive R&D.
+
+14 We would obtain the same formula if we compared the left-hand side of (13) with the profit from one-time imitation followed by R&D investment in subsequent generations.
+
+---
+
+9 - Sequential innovation, patents and imitation - 11/99
+
+Matters are quite different, however, in subsequent generations. Indeed, both firms will continue to undertake R & D after the first generation only if patent-holders are prepared to license their innovations. The patent-holder can find licensing profitable only if the joint profits of both firms exceed the monopoly profits the patent-holder could obtain without licensing. But this might not be so. The total value created will be larger with licensing because of complementarities, but competition may dissipate a share of this value to consumers, resulting in lower joint profits.
+
+Let us assume that licensing fees take the form of lump-sum payments. $^15$ In this case, product market competition would be unaffected by the patent and license. Thus, ignoring fees, each firm would earn $sv$ , and so a licensee would expect to earn
+
+$$(18)\quad sv+\frac{s(2p-p^{2})v-c}{(1-p)^{2}},$$
+
+where the first term in (18) is the current profit from competing against the patent-holder and the second term is expected future profit (assuming that all subsequent innovations are licensed, so that both firms can profitably undertake R & D). Total joint profits are
+
+$$(19)\quad 2sv+\frac{2s(2p-p^{2})v-2c}{(1-p)^{2}}=\frac{2sv-2c}{(1-p)^{2}}$$
+
+and the patent holder could extract this amount by charging a fee equal to (18).
+
+Thus, the license-holder will license if and only if
+
+$$(20)\quad \frac{v-c}{1-p}<2sv+\frac{2s(2p-p^{2})v-2c}{(1-p)^{2}},$$
+
+where the left-hand side of (20) is (monopoly) profit without licensing and the right-hand side is duopoly profit plus the licensing fee. Formula (20) can be rewritten as
+
+$$(21)\quad \frac{v}{c}>\frac{1+p}{2s-1+p}.$$
+
+Thus unlike the static model, licensing could be to a patent-holder's advantage in this dynamic setting.
+
+But for $s$ sufficiently small, we have
+
+15 This assumption would make sense, say, if the licensor could not readily monitor the extent to which the licensee's output made use of the licensed item. For example, it may not be clear how important a patented idea is for the source code of a software product. This would seriously interfere with setting fees as a function of output. Even if the licensor could monitor the licensee's output, it might well not be able to monitor the licensee's costs, in which case it would optimally set a fee that with positive probability was higher than the licensee would accept (in the event that the licensee's costs turned out to be high). In this setting with uncertainty about costs, we would obtain the same conclusion as below: too little licensing and hence too little innovation.
+
+---
+
+10 - Sequential innovation, patents and imitation - 11/99
+
+$$(22)\quad \frac{1+p}{p}<\frac{1}{sp}<\frac{1+p}{2s-1+p},^{16}$$
+
+where, from (11) , $\dfrac{1+p}{p}$ is the smallest value of $\dfrac{v}{c}$ for which it is efficient for both firms to invest, and, from (14) , $\dfrac{1}{sp}$ is the smallest value for which both firms will choose to invest in the absence of a patent system. Then innovations in the range $\dfrac{1}{sp} < \dfrac{v}{c} < \dfrac{1+p}{2s-1+p}$ would be pursued in a market without patents, but would not be licensed with patents. Thus, whether or not patents are available, the market will engender too little R & D. But, with sufficiently vigorous competition ( $s$ sufficiently small), (22) implies that a market with patents will, in general, be less efficient than one without, because patent-holders will not license sufficiently often (and so the pace of innovation will be even slower than in an economy without patents).
+
+We have stressed the case where (21) fails, and so patent licensing does not suffice to ensure adequate innovation. But even when (21) holds - the case in which innovations are sufficiently “ important ”- there is still an important difference between the static and dynamic models. Whereas a firm would invariably shun competition and imitation by the other firms in the former setting, it may welcome them in the latter. To see this, notice that in the static model, a firm expects payoff $pv-c$ without competition, but at most $\frac{1}{2}(2p-p^2)v-c$ (the payoff would be even lower were patents not available) when it has a rival. Hence, competition always makes a firm worse off. In the dynamic mode, however, a lone firm's payoff is given by (10) , whereas, when (21) holds, its expected payoff, whether or not patents are available, is
+
+$$\frac{s(2p-p^{2})v-c}{(1-p)^{2}},$$
+
+which is greater than (10). Hence, the firm's profit is actually enhanced by imitation and competition in this case, because of the greater rate of innovation that they foster. We summarize our findings (also illustrated in Figure 2 ) as follows:
+
+Proposition 2: After the first-generation innovation, patent protection gives rise to efficient R & D (and the absence of patent protection gives rise to insufficient R & D) if and only if it is socially optimal for just one firm to invest. When having more than one firm undertake R & D is efficient, however, a regime without patents induces an R & D investment level (and hence a pace of innovation) that (although still too low) is, in general, more efficient than one with patent protection (provided that competition is sufficiently intense). Moreover, if innovations are sufficiently important ( $v/c$ is sufficiently big), not only the social but the private return to R & D (i.e., a firm's profit) is enhanced by competition and imitation.
+
+$^{16}$ The left-hand inequality in (22) always holds; the right-hand inequality holds if $s<\frac{1}{2+p}$.
+
+---
+
+11 - Sequential innovation, patents and imitation - 11/99
+
+Remark: Although it is true that patent protection is efficient in the case where it is socially desirable to have only one firm invest, the “specialness” of this case is perhaps masked by our assumption that there are only two firms. Had we assumed half a dozen firms instead, the case in which exactly one innovating firm is optimal would appear much more restrictive. Moreover, this case also is presumably most relevant for innovations that are only “marginal.” For “important” innovations (high returns), society would want multiple firms to be in the race. Finally, our assumptions of costless imitation and no first-mover advantage (see footnote 8 ) also stack the deck in favor of patents. With more realistic assumptions, patents might not confer an advantage even when one firm is efficient.
+
+In our model, regimes with and without patents both give rise to too little innovation. One may ask, therefore, whether there is some superior third alternative. In fact, in our model a merger between the two competing firms would presumably permit their complementary ideas to flourish and of course would eliminate cutthroat rivalry altogether. Thus, the merger provides a theoretical solution to the R & D incentive problem. However, it is not likely to do so in practice, although it is often attempted in the software and computer industries. Indeed, the conclusion in our model that merger is optimal rests heavily on our simplifying (and heroic) assumption that monopoly creates no distortion. Here again our two-firm model may be misleading. In a more realistic model with a larger number of rivals, merger of all of them would become correspondingly more difficult and distortionary.
+
+## 4. Empirical evidence of dynamic innovation
+
+We present three pieces of evidence suggesting that the dynamic rather than the static model applies to innovative industries: 1.) The pattern of cross-licensing in high tech industries, 2.) The positive correlation between rates of innovation and rates of firm entry, and, 3.) The natural experiment in patent protection of software.
+
+### Cross-licensing of entire patent portfolios
+
+The distinctive nature of patent licensing in several high tech industries is difficult to reconcile with traditional intellectual property models. Under static models of innovation, firms offer patent licenses to competitors only under restrictive conditions. Specifically, in the static model presented here, patent holders would never offer licenses to competing firms. However, this model assumes a “ drastic ” product innovation. Other models that allow for “ non-drastic ” process improvements find that under some product market conditions, sufficiently minor innovations would be offered to competing firms. of these models, however, countenance situations where firms license their entire patent portfolios for a given field to direct competitors.
+
+Yet cross-licenses covering entire fields were common between competitors during the first several decades of the semiconductor and computer industries. In the semiconductor industry during 1975 alone, 34 cross-licensing (and second source) agreements between rival firms were announced [Webbink, p. 99] . These licenses covered whole portfolios of patents related to an entire technical field rather than individual inventions, and many covered future as well as existing patents.
+
+Moreover, these cross-licenses were not intended as barriers to entry as has sometimes been the case in other industries [Scherer, 1980, see also Bittlingmayer, 1988] . A survey of company executives concluded “...all company officials interviewed agreed that no company had ever been
+
+---
+
+12 - Sequential innovation, patents and imitation - 11/99
+
+prevented from entering the semiconductor business because of patents, nor had any company ever been refused a patent license [Webbink, p. 100].”
+
+Indeed, this cross-licensing is fully consistent with the dynamic model of innovation. 17 In the first instance, the cross-licenses relieve competing firms of holdup problems. The dynamic model suggests that patents provide initial inventors with holdup rights on subsequent inventions. Survey evidence from the semiconductor industry [Hall and Ham, 1999] and from all industries [Cohen, et al., 1998] indicates that firms view these holdup rights as a major feature of patents. Under the static model, a patent could also holdup other innovations, if, for instance, patent grants were excessively broad. Although overly broad patents and uncertain enforcement may exacerbate holdup problems, industry participants are clear that typical semiconductor innovations draw on hundreds of previous developments and that it is primarily sequential innovation that generates opportunities for holdup [Grindley and Teece, 1997, Hall and Ham, 1999] .
+
+For this reason, high tech cross-licensing displays a dynamic concern with possible future benefits that is at odds with the static model. For example, although an established firm's portfolio might have been sufficient to prevent the entry of a startup competitor, most established firms apparently recognized that startups might contribute significantly to future industry innovations. Hence according to Webbink's survey of semiconductor firms, cross-licenses were readily offered to new firms (often at very modest royalties), contrary to the static model.
+
+Moreover, licensing contracts typically cover future developments. $^18$ For example, when Robert Noyce founded Intel in the late 60's, he obtained cross-licenses from Texas Instruments, IBM and Fairchild that gave them access to future patents to the year 1999. The typical semiconductor cross-license today provides access to patents developed five years into the future. In many cases, the licenses provide use of covered patents for the entire term of the patent (now 20 years) [Grindley and Teece, 1997] .
+
+This dynamic concern for future developments guided licensing behavior from the beginning of the semiconductor industry. If ever an innovative firm could have gathered large monopoly rents from its patents, Bell Laboratories could have done so from its basic transistor patents. Instead, Western Electric (in charge of licensing the patents) aggressively offered the patents to all comers at a low 2 % royalty beginning in 1953, dropping this royalty entirely in 1956. 19 The vice president for electronic component development explained the logic:
+
+We realized that if this thing [the transistor] was as big as we thought, we couldn't keep it to ourselves and we couldn't make all the technical contributions. It was to our interest to spread it around. If you cast your bread on the water, sometimes it comes back angel food cake [Tilton, p75-6].
+
+17 Fershtman and Kamien [1992] model cross-licensing for situations where two complementary technologies are required to make a product. The logic is similar to the one suggested by our dynamic model in that firms make potential (but unknown and therefore uncontractable) complementary contributions to future products. The difference is that Fershtman and Kamien describe the exchange of specific patents , whereas in the semiconductor industry, consistent with the dynamic model, firms exchanged whole portfolios covering an entire field.
+
+18 Note that in our dynamic model a market with an agreement that all future patents will be licensed is equivalent to one in which there are no patents at all.
+
+19After 1956 Western Electric agreed to offer these patents royalty-free as part of an antitrust consent decree. However, industry observers and AT&T executives have commented that AT&T was willing to do this in any case.
+
+---
+
+13 - Sequential innovation, patents and imitation - 11/99
+
+Such behavior illustrates a strong concern with potential future benefits that goes well beyond static considerations of holdup. When different firms make complementary and sequential technical contributions (the dynamic model), collaborative forms of licensing may be privately advantageous.
+
+During the 1980's, industry norms began to erode substantially as some established firms chose to milk their patent portfolios for royalties rather than use them for new developments. Also, changes in the legal regime made it easier for patent holders to sue and win, most notably the organization of the Federal Circuit court for patent appeals in 1983 [Angel, 1994] . This diminished crosslicensing, especially to new firms, e.g., Intel has sued sub-licensees of its original cross-licenses. But these developments do not alter the significance of semiconductor cross-licensing as an example of firms acting on dynamic concerns about future innovation.
+
+## Innovation and firm entry
+
+The relationship between innovation and firm entry provides another test of the two models.
+
+In the static model of intellectual property, innovation incentives depend on the patent holder's ability to extract monopoly rents. The magnitude of these rents depends on product market conditions. Rents will be greatest when the patent holder enjoys a complete monopoly; rents will generally be inferior when other firms can enter the product market, even if they produce only imperfect substitutes. $^20$ Thus a high rate of firm entry is often taken as prima facie evidence of insufficient appropriability.
+
+One important source of variation in firm entry occurs over the product life cycle of an innovation and is related to patent protection. From studies by Gort and Klepper [1982] , we know that entry conditions change over the life of major innovations. Initially, a single firm will typically enjoy a monopoly, often supported by an initial set of patents. When these patents expire (or entry is allowed for other reasons), firms freely enter. Over time additional innovations are made to improve the product. As some firms build large portfolios of in-force patents on these improvements, entry once again becomes more difficult. Combined with maturing demand, new firms stop entering and less successful firms exit, resulting eventually in a stable industry. $^21$
+
+If the static model holds, innovation rates should be greatest when entry is most limited and innovation should decline when large numbers of new firms enter. Also, markets that experience greater entry should, in general, exhibit lower rates of innovation. By contrast, if the dynamic model holds, innovation rates might well be greater during phases of high entry and also for products experiencing high entry. Entrants may bring complementary knowledge that increases industry prospects for successful innovation.
+
+The evidence squarely supports the dynamic model. Gort and Klepper assembled information on innovation rates and entry rates for 23 major new products from the transistor to the zipper. They
+
+20 For example, consider the case where an innovator holds a patent on an improvement to a base product. The rents on the improvement will be greatest when the innovating firm has a monopoly on the base product as well. In general, the innovating firm will not realize the same rents from the improvement if entrants can freely offer an unimproved base product as a substitute—the unimproved version of the product can be offered at a lower price, tending to dissipate some of the rents.
+
+$^{21}$ This exposition highlights the role of patents in changing conditions for entry; Gort and Klepper provide a broader explanation.
+
+---
+
+14 - Sequential innovation, patents and imitation - 11/99
+
+divide these data into five phases of each product's life-cycle, the phases defined by net entry behavior. The first phase is the monopoly stage (or near-monopoly in a few cases), the second exhibits positive net entry, in the third, entrants roughly balance exiters, the fourth has negative net entry, and the fifth exhibits rough stability again. For each phase they report the annual rate of entry and of innovation. Innovation rates are further divided into rates for major innovations and for minor innovations.
+
+Table 1 and Figure 3 show the means (weighted by duration) of the annual rates of innovation for each phase for both major innovations and total innovations. As can be seen, neither the initial monopoly phase nor the final phases — both periods when entry is most constrained — have particularly high rates of innovation. The highest rates of innovation appear instead during the second and third phases, during and immediately following the period of greatest firm entry.
+
+This result can be explored more formally with a Poisson model of the innovation count data. For the $i$ th product during a phase of duration $\Delta t_i$ , we assume that the hazard for an innovation, $\lambda_i$ , is an exponential function of the net entry rate of firms, $n_i$ :
+
+$$\lambda_{i}=\Delta t_{i} \cdot  e^{\alpha+\beta n_{i}} .$$
+
+The probability that the number of innovations during this period is $y$ is
+
+$$P(Y=y)=\frac{e^{-\lambda_{i}}\cdot \lambda_{i}^{y}}{y!}.$$
+
+It is possible that changes in both the rate of innovations and the rate of firm entry result from exogenous changes in technological opportunities. That is, firms may choose to enter when opportunities to innovate are greater. In this case, the independent variable would be correlated with the error term. To correct for this, we perform an instrumental variables estimation.
+
+We begin, however, with a straightforward maximum likelihood estimation of this simple model displayed in column 1 of Table 2 , both for all innovations (top) and for only major innovations (bottom). The results show a significant positive relationship between entry and innovation.
+
+The Poisson regression model assumes that the variance of $y$ equals the mean. But this will not be the case if there are stochastic errors in addition to the Poisson sampling error [see Hausman, Hall and Griliches, 1984, Cameron and Trivedi, 1986] . To allow for possible “ over-dispersion, ” we also performed the negative binomial regression described in Cameron and Trivedi. These regressions are shown in column 2. The coefficients are quite similar, positive and highly significant. A likelihood ratio test indicates that over-dispersion does exist, hence the negative binomial model is preferred.
+
+For comparison we also perform a nonlinear least squares estimation in column 3. This method is consistent, but not efficient for our model, and so we estimate standard errors using White's heteroscedastic-consistent method [1980] . Results are similar, but in the estimation on major innovations the coefficient for entry is significant only at the 5 % level.
+
+---
+
+15 - Sequential innovation, patents and imitation - 11/99
+
+To correct for possible endogeneity in the independent variable we instrument the rate of firm entry with two variables. Desirable instruments should be correlated with the rate of entry, but uncorrelated with changes in technological opportunity. The first instrument is the average number of firms in the industry over the entire product life. 22 Although technological opportunity may influence the gross rate of entry, the exit process is independent, and the equilibrium number of firms over all phases would seem to be determined by market size and structure independently of opportunity. The second instrument is a simple dummy flag that takes the value of 1 during the initial monopoly phase and 0 otherwise. The initial monopoly period, if it exists, is presumed to result from an original set of strong, broad patents and should thus be independent of technological opportunity as well. 23
+
+Estimates using these instruments are shown in Column 4. Again, coefficients are similar and the coefficient on firm entry is significantly positive. $^24$
+
+The marginal effect of firm entry is not large—roughly 30 or 40 entrant firms correspond to one innovation. But this does not alter the interpretation. Firm entry does not decrease innovation as suggested by the static model; instead, new firms increase the rate of innovation.
+
+## The Natural Economic Experiment in Software
+
+The semiconductor, computer and software industries have historically experienced high levels of innovation despite weak patent protection. This suggests that the dynamic model is applicable, but, by itself, this evidence is not conclusive. Although these industries have been innovative without strong patent protection, perhaps they would have been far more innovative with strong protection; perhaps these industries offer many technological possibilities, but only the most highly profitable possibilities are realized under weak patent protection.
+
+Fortunately, this alternative explanation can be tested. The patent courts subjected the software industry to a natural economic experiment during the 1980's. 25 Before this time, patent protection for innovations was very limited; instead, innovations were protected by copyright. This meant practically that direct copying of a software product was prohibited, but that copying the ideas and concepts embodied in software was not. Market entry therefore required significant investment in development, but entry could not be barred.
+
+A series of court decisions in the early 1980's had the effect of extending patent protection to many software ideas. Consequently the number of patents issued annually covering software grew
+
+$^{22}$ These supplementary data were graciously provided by Steven Klepper.
+
+23 If initial monopolies do not arise from patent protection, then the static model would be irrelevant in any case. Note further that since we are instrumenting a nonlinear least squares estimation, the instruments apply to the pseudo-regressors of a linearized model, not to the rate of entry directly. To correspond to the form of the pseudo-regressors, the instruments were multiplied by the duration of the phase. Also, terms were included using the square of the average number of firms, the phase duration and a constant.
+
+24 A Hausman specification test could not reject the hypothesis that the simple nonlinear least squares estimator was consistent. The instrumental variables estimation was repeated using only the second instrument, the initial patent flag, plus phase duration and a constant. Results were positive with an even higher coefficient. Generally similar although sometimes less significant results were also obtained including industry dummies and performing a fixed effects analysis conditioning on the product sums [Hausman, Hall and Griliches, 1984] .
+
+$^{25}$ Some other natural experiments involving the extension of patent protection are [Scherer and Weisburst, 1995] and [Challu, 1995].
+
+---
+
+16 - Sequential innovation, patents and imitation - 11/99
+
+exponentially from the mid-80's to about 7,000 in 1995 (see Figure 4 ). Within the software industry, this has sometimes been described as a case of "fixing what ain't broke." Advocates counter, arguing along the lines of the static model, that increased patent protection should increase software innovativeness [USPTO, 1994] .
+
+If the static model is correct, then the extension of patent protection should have produced a sharp increase in R&D spending among those firms and industries applying for patents. This should have subsequently been followed by an increase in productivity growth. The changes should be measurable and large after controlling for other, possibly offsetting changes.
+
+According to the static model, R & D should increase with patent protection because firms can profitably pursue R & D projects that yield smaller returns, that is, projects with lower values of $\nu$ /c . This can be seen as follows. As shown in Figure 1 , projects with low values of $\nu$ /c , labeled “ insufficient innovation ” under no patents, become feasible for a greater number of firms with the extension of patent protection. Assume a stationary distribution of R & D opportunities ranked by $\nu$ /c such that $F\big(\nu/c\big)$ is the cumulative R & D spending required to invest in all opportunities with a return less than $\nu$ /c . For simplicity assume $F$ is concave. For the case of no patents, designate the entry threshold value of $\nu$ /c for one firm as $T_1^n$ and the threshold for two firms as $T_2^n$ (the values of these are given in Figure 1 ). Then all opportunities with returns between $x$ and $x+dx$ such that $T_1^n \leq  x < T_2^n$ will consume in total $dF(x)$ R & D dollars and will generate expected value of $p \cdot  x \cdot  dF(x)$ . Opportunities where $T_2^n \leq  x$ will require $2dF(x)$ R & D dollars (two firms investing) and will generate expected value of $p(2-p) \cdot  x \cdot  dF(x)$ . Then the average value of $\nu$ /c for the industry is the ratio of total value $\nu$ to total R & D investment,
+
+$$A_{n}=\frac{p \cdot  \int_{T_{1}^{n}}^{T_{2}^{n}} x \cdot  d F(x)+p(2-p) \cdot  \int_{T_{2}^{n}}^{\infty} x \cdot  d F(x)}{1 \cdot  \int_{T_{1}^{n}}^{T_{2}^{n}} d F(x)+2 \cdot  \int_{T_{2}^{n}}^{\infty} d F(x)}=\frac{p \int_{T_{1}^{n}}^{\infty} x \cdot  d F(x)+p(1-p) \int_{T_{2}^{n}}^{\infty} x \cdot  d F(x)}{\int_{T_{1}^{n}}^{\infty} d F(x)+\int_{T_{2}^{n}}^{\infty} d F(x)} .$$
+
+With patents, the equivalent thresholds are $T_1^P$ and $T_2^P$ and the corresponding average value of $v/c$ is $A_p$ . Now examining the table in Figure 1 , it is true that $T_1^P < T_2^P < T_1^n < T_2^n$ . Using this, it is straightforward to show that $A_p < A_n$ .
+
+In other words, the average value of v/c should decrease for industries with the extension of patent protection. This logic can be readily extended to cases with more than two firms. Also, allowing each firm to have an equal chance of being an early-mover for any R&D opportunity means that the average value of v/c should also decrease for firms, or alternatively, the average value of c/v should increase.
+
+For empirical analysis it is useful to note two aspects of this predicted change. First, since productivity is increasing in these industries (see below), the net social value $v$ will increase at least as fast as output. Therefore, an increase in $c$ / $v$ implies an increase in the ratio of R & D spending to output. In other words, the extension of patents should cause an increase in relative (to output)
+
+---
+
+17 - Sequential innovation, patents and imitation - 11/99
+
+R&D spending. Relative R&D spending is a more useful measure than absolute R&D spending given the changing composition of industries as firms acquire, divest, startup and discontinue product lines and industries grow.
+
+Second, $v$ represents a discounted stream of future values. Typically the increase in value (and the associated increase in output) associated with an innovation will follow the expenditure of R & D only after considerable delay. For this reason, we should expect the ratio of R & D to output to increase quite rapidly upon the extension of patent protection and subsequently level off.
+
+In contrast, if these industries follow the dynamic model, then we should expect the imposition of patents to lead to a reduction of R&D and productivity growth. This reduction should arise as patent holders stake out exclusive claims in different research areas inhibiting complementary innovation.
+
+Unlike the static model, however, we should expect this transition to be quite gradual for two reasons. First, any patent holder faces a large body of well-established prior art and possibly competing claims. In such an environment, a patent portfolio capable of fencing off an area of research can be built up only gradually. In fact, there has been relatively little software patent litigation so far and the companies with large portfolios are only just beginning to pursue software patent claims [Business Week, 1997]. $^26$
+
+Second, some of the most innovative firms may be reluctant to aggressively pursue patent claims. As we have seen above, although static firms will be better off with patent protection, dynamic firms may actually be better off without it. Thus the most innovative firms might seek to maintain industry norms of cooperation rather than to aggressively exert all patent rights. In this case, these norms will deteriorate slowly, and so problems of exclusive development may appear slowly. In fact, support for cooperative norms appears strong among many innovative software companies — senior executives from companies such as Microsoft, Sun and Oracle have expressed a general reluctance to pursue patent litigation and view their patenting activity as primarily defensive [PC Magazine, 1997, USPTO, 1994].
+
+Indeed, pure software companies as a whole have not applied for many patents. A naive view might expect software patents to be obtained predominately by firms in the computer programming and data processing industry (SIC 737). In fact, the largest software patentees are in the computer hardware and telecommunications industries—industries which sell software products and also incorporate software in hardware products. The top 10 U. S. firms obtaining software patents in 1995 are listed in Table 3. The top ranked pure software firm in 1995 was Microsoft (rank 24) with 39 software patents. 27
+
+To summarize, if the static model holds, relative R & D spending should have increased sharply, followed by productivity. If the dynamic model applies, relative R & D spending and productivity would not have increased and might exhibit a mild decrease.
+
+26 Also, given the incompleteness of patent portfolios, much of this activity is directed not toward exclusive control of a market, but toward extracting royalties. Nevertheless, excessive royalties may limit complementary activity at the margin.
+
+27The software patent series used in this analysis were developed by Greg Aharonian of the Internet Patent News Service. The criteria for software patents include not only the USTPO patent class, but also detailed examination of the specification, claims and abstract.
+
+---
+
+18 - Sequential innovation, patents and imitation - 11/99
+
+We examine these changes among three different samples of firms:
+
+- 1.) The top 10 U.S. software patentees in 1995, accounting for 35 % of the software patents issued
+to U.S. companies in that year,
+2.) The industry groupings for computer hardware and programming services in the NSF R & D
+survey (company R & D funds for SIC 357 and part 737 and 871) [NSF, 1996, 1997], and,
+3.) The grouping of computer, telecommunications and electronic components (SIC 357, 365-7) in
+the NBER R & D Masterfile [Hall, 1988], a listing of publicly traded U.S. firms. $^28$
+For the first and last samples, the R&D and sales measures are global. For the NSF sample, the R&D measures are domestic only and we measure R&D intensity using the NSF figures for sales for SIC 357 and 737.29
+
+We initially explore R & D spending relative to sales (R & D intensity) rather than output. The trend in these measures is shown in Figure 5 . The late 80's display a leveling off and possibly a reversal of an upward trend in research intensity over the previous decade. There does not appear to be so much as a 10 % increase in R & D intensity among the firms and industries obtaining software patents. 30
+
+There could be two sorts of offsetting changes: 1.) Technological opportunities may have simultaneously fallen abruptly, and, 2.) The cost of performing R & D could have simultaneously risen sharply. A decline in technological opportunity seems at odds with the continued rapid growth and rapid innovation in these industries. Bronwyn Hall [1993] performs an econometric analysis on the same NBER dataset and finds that the output elasticity of R & D did not fall during the 1980's, but instead rose. 31
+
+Hall also presents evidence that the general costs of performing R & D did not rise sharply. If R & D costs had increased overall, offsetting an erstwhile increase in R & D spending, then the R & D intensity of other industries should have fallen . Figure 6 presents ratios of R & D intensity of software-related industries to the R & D intensity of the entire manufacturing sector. As can be seen, the relative R & D intensity of software-related industries fell over this time period. Thus not only
+
+28 Data for the top 10 firms was obtained from annual reports, 10-Ks and the NBER Masterfile. The series for AT&T was based on consolidated figures including NCR, the computer company which was purchased by AT&T in 1991. For this reason we use only the top 9 firms prior to 1991, although the difference is not significant. Both the NSF samples and the NBER R&D Masterfile are firm-based surveys where all of the R&D and sales of a firm are assigned to the SIC category of the firm's major product line. Thus our measures are diluted by non-software R&D and non-software output. Nevertheless, as long as software development constitutes a substantial portion of R&D, then we should expect to see a significant increase in R&D intensity.
+
+29 Note that beginning in 1985 FASB required that a portion of software development expense should be capitalized, hence reported R&D includes directly expensed items plus the amortization expense of capitalized software. The introduction of this change may have had a slight distortionary effect on reported R&D, tending to delay a portion of expenditures. The effect of this accounting change was to spread the impact of any sharp changes in R&D spending over two or three years. This effect was temporary, significant largely for pure software firms and of relatively brief duration (software is typically amortized over three years or less). This was not a substantial factor for the 10 largest software patentees and, based on this, would not seem to be a major factor for industry measures either.
+
+$^{30}$ The NSF series becomes erratic after 1992 as the result of sample changes and as some firms were re-classified into different industries.
+
+31 The increase was concentrated among smaller public firms as large firms apparently lost productivity switching from mainframe technology to microcomputers. But overall technological opportunity did not decline.
+
+---
+
+19 - Sequential innovation, patents and imitation - 11/99
+
+did these industries fail to show a large increase in relative R&D spending, but they lagged behind the rest of the manufacturing sector over this period.
+
+It is possible, however, that R & D spending relative to sales may understate R & D relative to output because of price effects. That is, as firms gain monopoly power with patents, prices may rise, inflating the sales figure in the denominator. To consider this possibility, Figure 7 displays the ratio of real R & D to output where R & D has been deflated using the NBER R & D deflator and sales have been deflated by a shipments-weighted index derived from the NBER Productivity Database for the industries involved. As can be seen, R & D relative to output exhibits a significant decline during the late 1980's. Perhaps prices have been mis-measured for the computer industry. However, it seems unlikely that any measurement error could be so large as to mask major price increases. Hence this evidence is hard to reconcile with the static model.
+
+Bronwyn Hall has suggested [1993] that competition may have hit the large mainframe firms in the industry especially hard as new firms entered the computer industry in the early 1980's. Consequently the response of the large firms (and by implication industry averages) might not be representative of firms in the industry as a whole. To consider this possibility, we examined two sub-samples from the software-related firms in the NBER R & D Masterfile: a balanced panel of 49 small firms and an unbalanced panel of new public firms. 32 Figure 8 shows the R & D intensity of these panels compared to the performance of the top 9 software patentees. As can be seen, R & D spending diverged between these groups during the early 1980's, consistent with Hall's interpretation, but these groups did not increase relative R & D spending either during the late 1980's in response to software patents.
+
+Thus the extension of patent protection to software did not generate a relative increase in R&D spending as predicted by the static model; instead, consistent with the dynamic model, R&D spending seems to have remained roughly steady or to have declined. Not surprisingly, these industries did not demonstrate increased productivity growth as a result of the patent bonanza, as seen in Figure 9 . Although multi-factor productivity may have fallen for reasons related to the transition from mainframes to microcomputers, there is no evidence of any underlying productivity increase commensurate with the increase in patents.
+
+Of course, these industries have remained innovative and productivity growth is still positive. This does not, however, contradict the dynamic model; rather, the negative effects of the patent extension may not be felt for some time as industry cooperative norms continue and as litigation remains limited. The bill for this experiment has not yet come due.
+
+In summary, the initial high level of innovation in software, mixed industry support for patents, and an apparent gradual slowdown of R & D intensity all suggest the applicability of the dynamic model. This conclusion is also supported by the distinctive pattern of cross-licensing in the semiconductor and computer industries and by the more general positive relationship between innovation and firm entry. All of this evidence is difficult to reconcile with the traditional static model of intellectual property.
+
+32 The small firms are all those existing in 1980 and 1990 with fewer than 1,000 employees in 1980. The new firms are defined as firms that first enter the NBER R&D Masterfile after 1973 and have fewer than 5,000 employees their first recorded year. Conversations with Compustat confirmed that this procedure was likely to screen out most spin-offs, re-organizations and listing changes. New firms were dropped from the panel after 8 years.
+
+---
+
+20 - Sequential innovation, patents and imitation - 11/99
+
+## Conclusion
+
+Intellectual property appears to be one of those areas where results that seem secure in the context of a static model are overturned in a dynamic model. Imitation invariably inhibits innovation in a static world; in a dynamic world, imitators can provide benefit to both the original innovator and to society as a whole. Patents preserve innovation incentives in a static world; in a dynamic world, firms may have plenty of incentive to innovate without patents and patents may constrict complementary innovation.
+
+This suggests a cautionary note regarding intellectual property protection. The reflexive view that “ stronger is always better ” is incorrect; rather a balanced approach is required. The ideal patent policy limits “ knock-off ” imitation, but allows developers who make similar, but potentially valuable complementary contributions. In this sense, copyright protection for software programs (which has gone through its own evolution over the last decade) may have achieved a better balance than patent protection. In particular, industry participants complain that software patents have been too broad and too obvious, leading to holdup problems [USTPO]. Also in this regard, patent systems that limit patent breadth, such as the Japanese system, may offer a better balance. Thus our model suggests another, different rationale for narrow patent breadth than the recent economic literature on this subject.
+
+## Bibliography
+
+AHARONIAN, G. 1995. Patent News Service.
+
+ANGEL, DAVID P. 1994. Restructuring for Innovation: the remaking of the US semiconductor industry, , Guilford, NY.
+
+ARROW, K. 1962. "Economic Welfare and the Allocation of Resources for Inventions," The Rate and Direction of Inventive Activity, Princeton Univ.
+
+BARTELSMAN, E. AND GRAY, W., 1996. "The NBER Manufacturing Productivity Database", NBER Technical Working Paper, no. 205.
+
+BITTLINGMAYER, G. 1988. "Property rights, progress, and the aircraft patent agreement," Journal of Law and Economics, v. 31, p. 227.
+
+CAMERON, A. AND TRIVEDI, P. 1986. "Econometric models based on count data: comparisons and applications of some estimators and tests," Journal of Applied Econometrics, v. 1, p. 29.
+
+CHALLU, P. 1995. "Effects of the monopolistic patenting of medicine in Italy since 1978," International Journal of Technology Management, v. 10, p. 237.
+
+CHANG, H. 1995. "Patent scope, antitrust policy and cumulative innovation," RAND Journal of Economics, v. 26, no.1, p. 34.
+
+COHEN, W., GOTO, A., NAGATA, A, NELSON, R., AND WALSH, J. 1998. "R&D Spillovers, Patents and the Incentives to Innovate in Japan and the United States," unpublished working paper.
+
+Dasgupta, P. and Maskin, E. 1987. "The simple economics of research portfolios," The Economic Journal, v. 97, p. 581.
+
+DASGUPTA, P. AND STIGLITZ, J. 1980. "Uncertainty, Industrial Structure, and the Speed of R&D," Bell Journal of Economics, 11, p. 1.
+
+ENOS, J. 1962. Petroleum Progress and Profits: a history of process innovation, MIT Press.
+
+F ERSHTMAN , C. AND K AMIEN , M. 1992. “Cross licensing of complementary technologies,” International Journal of Industrial Organization, v. 10, p. 329.
+
+---
+
+21 - Sequential innovation, patents and imitation - 11/99
+
+GREEN, J. AND SCOTCHMER, S. 1995. "On the division of profit in sequential innovation," RAND Journal of Economics, v. 26, no.1, p. 20.
+
+GORT, M. AND KLEPPER, S. 1982. "Time paths in the diffusion of product innovations," The Economic Journal, v. 92, p. 630.
+
+GRILICHES, Z. 1992. "The Search for R&D Spillovers," Scandinavian Journal of Economics, v. 94, supplement, p. 29.
+
+GRINDLEY, P. AND TEECE, D. 1997. "Managing intellectual capital: licensing and cross-licensing in semiconductors and electronics," California Management Review, v. 39, no. 2, p. 8.
+
+HALL, B. 1988. "The manufacturing sector master file: 1959-1987," NBER Working Paper Series, No. 3366, p. .
+
+HALL, B. 1993. "Industrial research during the 1980s: Did the rate of return fall?," Brookings Papers: Microeconomics 2, , p. 289.
+
+HALL, B. AND HAM, R. 1999. "The patent paradox revisited: determinants of patenting in the US semiconductor industry, 1980-94," NBER Working Paper Seris, no. 7062.
+
+HAUSMAN, J., HALL, B., AND GRILICHES, Z. 1984. "Econometric models for count data with an application to the patents-R&D relationship," Econometrica, v. 52, no. 4, p. 909.
+
+HENDERSON, R. AND COCKBURN, I. 1996. "Scale, scope and spillovers: the determinants of research productivity in drug discovery," RAND Journal of Economics, v. 27, no. 1, p. 32.
+
+JAFFE, A. 1986. "Technological opportunity and spillovers of R&D: Evidence from firms' patents, profits and market value," American Economic Review, v. 76, no. 5, p. 984.
+
+LEE, T. AND WILDE, L. 1980. "Market Structure and Innovation: A Reformulation," Quarterly Journal of Economic, 94, p. 429.
+
+LEVIN, R. 1982. "The Semiconductor Industry," Government and Technical Progress: A cross-industry analysis, Nelson, R. ed., Pergamon, NY.
+
+LEVIN, R. AND REISS, P. 1988. "Cost-reducing and demand-creating R&D with Spillovers," RAND Journal of Economics, v. 19, no. 4, p. 538.
+
+LEVIN, R., KLEVORICK, A., NELSON, R. AND WINTER, S. 1987. "Appropriating the Returns from Industrial Research and Development," Brookings Papers on Economic Activity [3], , p. 783.
+
+LOURY, G. 1979. "Market Structure and Innovation," Quarterly Journal of Economic, 93, p. 395.
+
+MANSFIELD, E., SCHWARTZ, M. AND WAGNER, S. 1981. "Imitation Costs and Patents: An empirical study," Economic Journal, v. 91, p. 907.
+
+MYHRVOLD, N. 1997. "Interview," PC Magazine, Sept. 9, 97, p. 30.
+
+NATIONAL SCIENCE FOUNDATION. Annual. Research and Development in Industry.
+
+NATIONAL SCIENCE FOUNDATION. Annual. Science and Engineering Indicators.
+
+NORDHAUS, W. 1969. Invention, Growth and Welfare: A Theoretical Treatment of Technological Change, MIT, Cambridge.
+
+OFFICE OF TECHNOLOGY ASSESSMENT 1992. Finding a Balance: computer software, intellectual property and the challenge of technological change, Congress of the US.
+
+SAGER, I. 1997. "Big Blue is out to collar software scofflaws," Business Week, March 17, 1997.
+
+SCHERER, F. 1980. Industrial market structure and economic performance, , Rand McNally.
+
+SCHERER, F. AND WEISBURST, S. 1995. "Economic effects of strengthening pharmaceutical patent protection in Italy," International Review of Industrial Property and Copyright Law, p. 1009.
+
+SCOTCHMER, S. 1996. "Protecting early innovators: should second-generation products be patentable?," RAND Journal of Economics, v. 27, no. 2, p. 322.
+
+SPENCE, M. 1984. "Cost reduction, competition, and industry performance," Econometrica, v. 52, no. 1, p. 101.
+
+---
+
+22 - Sequential innovation, patents and imitation - 11/99
+
+TAYLOR, C. T. AND SILBERSTON, Z. A. 1973. The economic impact of the patent system: a study of the British experience, Cambridge Univ. p. 294.
+
+TILTON, J. 1971. International Diffusion of Technology: The Case of Semiconductors, , Brookings Inst.
+
+USPTO 1994. "Hearings on Software Patent Protection," United States Patent and Trademark Office, JanuaryFebruary 1994.
+
+WEBINK, D. 1977. "The semiconductor industry : a survey of structure, conduct, and performance : staff report to the Federal Trade Commission," Federal Trade Commission.
+
+WHITE, H. 1980. "A heteroscedasticity-consistent covariance matrix estimator and a direct test for heteroscedacticity," Econometrica, v. 48, p. 817.
+
+Table 1. Weighted means of annual rates of innovation by phase.
+
+<table><tr><td></td><td>Phase 1</td><td>Phase 2</td><td>Phase 3</td><td>Phase 4</td><td>Phase 5</td></tr><tr><td>Rate of all innovations</td><td>0.39</td><td>0.57</td><td>0.62</td><td>0.36</td><td>0.43</td></tr><tr><td>Rate of major innovations</td><td>0.19</td><td>0.29</td><td>0.22</td><td>0.18</td><td>0.22</td></tr><tr><td>Rate of net firm entry</td><td>0.22</td><td>5.05</td><td>-0.07</td><td>-4.97</td><td>0.16</td></tr></table>
+
+
+Source: Gort and Klepper, 1982
+
+---
+
+23 - Sequential innovation, patents and imitation - 11/99
+
+Table 2. Regressions on innovation counts.
+
+<table><tr><td>Regression</td><td>1</td><td>2</td><td>3</td><td>4</td></tr><tr><td>Regression method</td><td>Poisson</td><td>Negative Binomial</td><td>Nonlinear least squares</td><td>Nonlinear least squares instrumental variables</td></tr><tr><td>Dependent variable</td><td colspan="4">Total number of innovations</td></tr><tr><td>Coefficient of firm net entry rate</td><td>.046* (.007)</td><td>.034* (.003)</td><td>.052* (.014)</td><td>.062* (.009)</td></tr><tr><td>Constant</td><td>-.820* (.052)</td><td>-.694* (.105)</td><td>-.934* (.366)</td><td>-.943* (.120)</td></tr><tr><td> $\theta$  in NEGBIN II</td><td>--</td><td>1.373* (.216)</td><td>--</td><td>--</td></tr><tr><td> $R^{2}$ </td><td>.43</td><td>.37</td><td>.44</td><td>.43</td></tr><tr><td>Dependent variable</td><td colspan="4">Number of major innovations</td></tr><tr><td>Coefficient of firm net entry rate</td><td>.041* (.011)</td><td>.035* (.003)</td><td>.045 (.022)</td><td>.046* (.015)</td></tr><tr><td>Constant</td><td>-1.547* (.074)</td><td>-1.444* (.105)</td><td>-1.668* (.261)</td><td>-1.666* (.132)</td></tr><tr><td> $\theta$  in NEGBIN II</td><td>--</td><td>1.560* (.216)</td><td>--</td><td>--</td></tr><tr><td> $R^{2}$ </td><td>.35</td><td>.30</td><td>.36</td><td>.36</td></tr></table>
+
+
+*Significant at the 1 % level. Asymptotic standard errors in parentheses, using White heteroscedasticity consistent standard errors for nonlinear regressions. Regressions cover 418 total innovations, 200 of these rated as major innovations, during 77 product phases over 887 product-years. Data are from Gort and Klepper [1982] . Instruments include a flag indicating initial monopoly phase, the average number of firms over the entire product life cycle, the square of the average number of firms, all multiplied by the duration of the phase, phase duration, and a constant. A Hausman specification test between the third and fourth columns does not reject the hypothesis that the third column is consistent (P = .346 for all innovations and P = .917 for major innovations). NLLS-IV using only the monopoly flag times phase duration, duration and a constant as instruments generates significant and even larger coefficients on net entry.
+
+---
+
+24 - Sequential innovation, patents and imitation - 11/99
+
+Table 3. Top 10 Software Patentees, 1995
+
+<table><tr><td>Firm</td><td>Software Patents Issued 1995</td><td>Total Utility Patents Issued 1995</td><td>R&amp;D Spending 1994 (millions)</td></tr><tr><td>International Business Machines</td><td>503</td><td>1383</td><td>3,382</td></tr><tr><td>AT&amp;T</td><td>185</td><td>638</td><td>3,110</td></tr><tr><td>Motorola</td><td>157</td><td>1012</td><td>1,860</td></tr><tr><td>Xerox (including Fuji Xerox)</td><td>121</td><td>551</td><td>895</td></tr><tr><td>Hewlett Packard</td><td>89</td><td>470</td><td>2,027</td></tr><tr><td>Digital Equipment</td><td>80</td><td>189</td><td>1,301</td></tr><tr><td>General Electric</td><td>59.</td><td>758</td><td>1,176</td></tr><tr><td>Apple Computer</td><td>57.</td><td>129</td><td>564</td></tr><tr><td>Ford Motor Co.</td><td>53.</td><td>334</td><td>5,214</td></tr><tr><td>Eastman Kodak</td><td>49.</td><td>772</td><td>859</td></tr></table>
+
+
+Sources: PATNEWS, USPTO, annual reports
+
+---
+
+25 - Sequential innovation, patents and imitation - 11/99
+
+Figure 1.Number of firms innovating in static model.
+
+![Figure](figures/figure_002.png)
+
+<table><tr><td>Innovating firms</td><td>Social Optimum</td><td>No Patents</td><td>Patents</td></tr><tr><td>One firm</td><td>$\frac{v}{c} > \frac{1}{p}$</td><td>$\frac{v}{c} > \frac{1}{ps}$</td><td>$\frac{v}{c} > \frac{1}{p}$</td></tr><tr><td>Two firms</td><td>$\frac{v}{c} > \frac{1}{p(1-p)}$</td><td>$\frac{v}{c} > \frac{1}{ps(1-p)}$</td><td>$\frac{v}{c} > \frac{1}{p(1-p/2)}$</td></tr></table>
+
+
+---
+
+26 - Sequential innovation, patents and imitation - 11/99
+
+Figure 2. Number of innovating firms (after first stage) in dynamic model.
+
+![Figure](figures/figure_002.png)
+
+<table><tr><td>Innovating firms</td><td>Social Optimum</td><td>No Patents</td><td>Patents</td></tr><tr><td>One firm</td><td>$\frac{v}{c} > \frac{1}{p}$</td><td></td><td>$\frac{v}{c} > \frac{1}{p}$</td></tr><tr><td>Two firms</td><td>$\frac{v}{c} > \frac{1+p}{p}$</td><td>$\frac{v}{c} > \frac{1}{ps}$</td><td>$\frac{v}{c} > \frac{1+p}{2s-1+p}$</td></tr></table>
+
+
+---
+
+27 - Sequential innovation, patents and imitation - 11/99
+
+Figure 3. Innovation rates during product life-cycle phases
+
+![Figure](figures/figure_002.png)
+
+---
+
+28 - Sequential innovation, patents and imitation - 11/99
+
+Figure 4. U. S. Software Patents Issued
+
+![Figure](figures/figure_002.png)
+
+Source: Internet PATNEWS service.
+
+---
+
+29 - Sequential innovation, patents and imitation - 11/99
+
+Figure 5. R&D Intensity for Software-related industries and firms
+
+![Figure](figures/figure_002.png)
+
+Sources: NSF Research and Development in Industry, Science and Engineering Indicators NBER R&D Masterfile, Annual reports.
+
+NBER series includes SIC 357, 365, 366, 367. NSF series includes SIC 357, and after 1986 part 737 and part 871. NSF series includes sample changes and hence is not directly comparable from year to year. Top firms come from Patent News Service rankings of software patents issued in 1995.
+
+---
+
+30 - Sequential innovation, patents and imitation - 11/99
+
+Figure 6. Relative R&D Intensity
+
+![Figure](figures/figure_002.png)
+
+Sources: NSF Research and Development in Industry, Science and Engineering Indicators NBER R&D Masterfile, Annual reports.
+
+Relative intensity is the ratio of R&D spending to output divided by that ratio of R&D spending to output for the entire manufacturing sector. NBER series includes SIC 357, 365, 366, 367.
+
+NSF series includes SIC 357, and after 1986 part 737 and part 871.
+
+Top firms come from Patent News Service rankings of software patents issued in 1995 relative to the NBER series for all manufacturing.
+
+---
+
+31 - Sequential innovation, patents and imitation - 11/99
+
+Figure 7. Real R&D / Real Output (Deflated R&D Intensity)
+
+![Figure](figures/figure_002.png)
+
+Sources: NBER R&D Masterfile, Annual reports.
+
+NBER series includes SIC 357, 365, 366, 367. R&D is deflated using NBER R&D deflator. Sales are deflated using a shipments-weighted mean for these industries from the NBER Productivity Database.
+
+---
+
+32 - Sequential innovation, patents and imitation - 11/99
+
+Figure 8. R&D Intensity for Small and New Firms
+
+![Figure](figures/figure_002.png)
+
+Source: NBER R&D Master File, Annual Reports
+
+“Small firms” is a balanced panel of 49 firms from the software-related industries found in the NBER Master File in both 1980 and 1990 that had fewer than 1,000 employees in 1980. “New firms” are firms from the NBER Master File in software-related industries that first appear after 1973 and that had fewer than 5,000 employees their first recorded year. The unbalanced panel of new firms includes only the first eight years that a firm appears in the file.
+
+---
+
+33 - Sequential innovation, patents and imitation - 11/99
+
+Figure 9. Total Factor Productivity Growth of Software-related Manufacturing Industries
+
+![Figure](figures/figure_002.png)
+
+Source: NBER Manufacturing Productivity Database, Output-weighted aggregates of Bartelsman & Gray [1996] calculations for 4-digit industries. Moving averages are over three years.
+

processed/00036_W2090138223/document.md

@@ -0,0 +1,579 @@
+Oxford Review of Economic Policy, Volume 23, Number 4, 2007, pp.568-587
+
+# Patents and patent policy
+
+Bronwyn H. Hall*
+
+Abstract A patent is the legal right of an inventor to exclude others from making or using a particular invention. This right is sometimes termed an `intellectual property right' and is viewed as an incentive for innovation. This article surveys the evidence on patent effectiveness in encouraging innovation and reviews the current controversies in patent policy.
+
+Key words: patents, intellectual property, incentives
+
+JEL classification: K11, L4, 034
+
+## I. Introduction
+
+An inevitable consequence of the growth of the `knowledge' or `information' economy is the increased importance of instruments designed to protect the property rights associated with these intangibles. Chief among the formal means of such protection is the patent, defined as the legal right of an inventor to exclude others from making or using a particular invention. This right is customarily limited in time, to 20 years from the date of application submission in most countries. The principle behind the modern patent is that an inventor is allowed a limited amount of time to exclude others from supplying or using an invention in order to encourage inventive activity by preventing immediate imitation. In return, the inventor is required to make the description and implementation of the invention public rather than keeping it secret, allowing others to build more easily on the knowledge contained in his invention.
+
+Given the increased salience of intellectual property rights (IPR) broadly and especially of patents to firms competing in the knowledge economy, it is perhaps not surprising
+
+*University of California at Berkeley and University of Maastricht, e-mail: bhhall@econ.berkeley.edu
+
+I am grateful to two anonymous referees for helpful comments.
+
+doi: 10.1093/icb/grm037
+
+© The Author 2007. Published by Oxford University Press. For permissions please e-mail: journals.permissions@oxfordjournals.org
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+569
+
+that this increase has been accompanied by increased criticism and controversy over the functioning of patent systems throughout the world. In fact, a number of government and quasi-governmental agencies have issued reports calling for patent reform in the very recent past, in some cases as part of a broader agenda of IPR reform. $^1$ The recommendations offered have ranged from procedural reforms specific to individual jurisdictions, all the way to calls for reductions in the enforcement of the TRIPS (Trade-related Aspects of Intellectual Property Rights) agreement in developing countries. $^2$ They include recommendations that patent and IPR policy operate in tandem with competition policy, that impediments to basic scientific research using patented inputs be reduced, and that the patent systems in individual countries and regions work towards harmonization in order to reduce the transactions costs of the system.
+
+This article sets the stage for a discussion of current trends in patenting and patent policy with a brief history and guide to the current operation of patent systems around the world. It then reviews the economic rationale for patents and the evidence that patents provide or do not provide appropriate incentives for innovation. I then discuss why the system has been under strain recently and review the arguments for various policy reforms. $^3$
+
+## II. A brief history and guide
+
+Patents have a long history, although some of the earliest patents are simply the grant of a legal monopoly in a particular good, rather than protection of an invention from imitation. Early examples of technology-related patents are: Brunelleschi's patent on a boat designed to carry marble up the Arno River, issued by the Florentine government in 1421; the Venetian patent law of 1474; and various patent monopolies granted by the English crown between the fifteenth and seventeenth centuries. The modern patent, which requires a working model or written description of an invention, dates from the eighteenth century, first in Britain (1718) and then in the United States (1790), followed closely by France (in both the latter two cases one of the consequences of a revolution). $^4$ Many other Continental European countries introduced patents during the 19th century, as did Japan (JPO, 2006) and India (James, 2007). During the twentieth century, the use of patent systems became almost universal and the signing of the TRIPS agreement has ensured that all countries who are members of the World Trade Organization (WTO) will have at least a minimal level of patent protection.
+
+In 1883 the Paris Convention for the Protection of Industrial Property guaranteed national treatment of patent applicants from any country that was a party to it. Its most important provision gave applicants who were nationals or residents of one member state the right to
+
+1 For the UK, see Gowers (2006). For the USA, see Federal Trade Commission (FTC, 2003); National Academies’ Board on Science, Technology, and Economic Policy (2004); the Reply to the National Academies Report by the American Intellectual Property Law Association (2004); Maskus (2006). Elsewhere, see Commission on Intellectual Property Rights (2002); Japanese Government (2004); Danish Board of Technology (2005); and IBM (2006). For a set of papers reporting recent economic research on patents, see OECD (2003).
+
+2 For example, in February 2007, US Congressman Henry Waxman called on Novartis not to challenge India’s denial of a patent on a new cancer drug, Glivec, as a violation of TRIPS ( MIP News , 19 February 2007, available at http://www.managingip.com/default.asp?page=8)
+
+3 For an excellent contemporary survey of the impact of changes in IPR systems on management and strategy, see Ziedonis (2007).
+
+$^{4}$ Ladas and Parry (2003). See also the EPO and USPTO websites (EPO, 2007a, and USPTO, 2007).
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+570
+
+Bronwyn H. Hall
+
+file an application in their own country and then, as long as an application was filed in another country that was a member of the treaty within a specified time (now 12 months), to have the date of filing in the home country count as the effective filing date in that other country (the ‘priority date’). This is an important feature of the patent system, as it enables worldwide priority to be obtained for an invention originating in any one country, in addition to ensuring that in principle all inventors are treated equally by the system, regardless of the country from which they come.
+
+Although the process for granting a patent varies slightly according to the jurisdiction for which protection is desired, the adoption of the TRIPS agreement in 1995 ensures that it is approximately the same everywhere in the world. This agreement requires its member countries to make patent protection available for any product or process invention in any field of technology with only a few specified exceptions. It also requires them to make the term of protection available for a period of not less than 20 years from the date of filing the patent application.
+
+The World Intellectual Property Organization (WIPO) has almost 200 member states and lists an equivalent number of national patent offices and industrial property offices on its website. In general, the patent right extends only within the border of the jurisdiction that has granted it (usually but not always a country). An important exception to the one country – one state rule is the European system, where it is possible to file a patent application at the European Patent Office (EPO) that will become a set of national patent rights in several European countries at the time of issue (EPO, 2006). A similar situation exists with respect to the African Regional Intellectual Property Organization (ARIPO). The exact number and choice of countries is under control of the applicant. Patents granted by the EPO have the same legal status as patents granted by the various national offices that are party to the European Patent Convention (EPC).
+
+The Patent Cooperation Treaty (PCT) came into existence in 1978, and now has 133 countries as contracting signatories. Any resident or national of a contracting state of the PCT may file an international application under the PCT that specifies the office which should conduct the search. The PCT application serves as an application filed in each designated contracting state. However, in order to obtain patent protection in a particular state, a patent needs to be granted by that state to the claimed invention contained in the international application. The advantage of a PCT application is that fewer searches need be conducted and the process is therefore less expensive. Thus, although application and search are to some extent standardized across offices, grants are not. In fact, 87 per cent of the PCT applications go to one of three patent offices for search: those in the United States, Europe, and Japan (WIPO, 2007). Most of the other systems rely on these offices for the search process and follow them in a number of other areas. Therefore, much of what follows focuses on these three major systems.
+
+Many patent offices have a provision for challenging patents following their issue. In the United States, any third party may request re-examination of a patent during its lifetime, although for various reasons related to potential subsequent litigation this opportunity is rarely taken up. $^5$ In Europe and Japan, robust patent opposition systems with limited time frames operate, and these systems are often used by rival firms as an alternative to more expensive litigation (Hall et al. , 2003; Harhoff and Reitzig, 2004) . In Europe this avenue of
+
+5 Users of this procedure are stopped from raising any issues of validity that might have been raised in reexamination in subsequent litigation. In practice, fewer than 1 per cent of US patents are re-examined, and almost half of those re-exams are requested by the patent-holder or the USPTO itself (Graham et al., 2003).
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+571
+
+challenge is particularly attractive because it is the last opportunity to attack the validity of a patent at the Europe-wide level rather than in individual national courts.
+
+Patents are valuable only if they can be enforced and this fact has a number of implications for their use. First, the ability of the courts to reach the `correct' verdict with respect to infringement and validity will matter; in situations or jurisdictions where there is a great deal of uncertainty about the outcome, and even if both parties agree as to the merits of the case, it may be worthwhile for one or both of them to pursue the issue further or, in some cases, to reach a private financial settlement to avoid a random outcome in the courts. $^6$ Second, the costs of litigation will matter: where parties with deep pockets can threaten those with less access to financial resources, or where the opportunity cost of paying attention to a patent suit is high, as in small entrepreneurial firms. On the other hand, smaller parties with little to lose can also hold up firms with large sunk investments at risk (Hall and Ziedonis, 2001) . Finally, the threat of litigation may discourage firms from even entering certain areas, thus providing a disincentive rather than an incentive for R & D. Lerner (1995) documented this phenomenon for biotechnology.
+
+The degree to which these kinds of threats matter depends to a great extent on the costs and extent of litigation, both of which tend to be higher in the United States than in many other countries. However, there are signs that concerns about litigation cost have been increasing elsewhere, notably in Europe where there is active debate over proposals to reduce enforcement costs by creating a supranational patent court system of some kind. At the time of writing, the current proposal, which is the European Patent Litigation Agreement (EPLA), has not yet been ratified by enough countries to make it legally binding. EPO (2006) gives the rationale for such an agreement along with some estimates of patent litigation cost in Europe and EPO (2007 b ) reports on the current state of all legislative initiatives in Europe.
+
+Research on patent litigation is difficult because of the data collection problem (it frequently requires accessing the records of courts in several different jurisdictions), but in recent years there has been series of studies of US patent litigation (Moore, 2000; Lanjouw and Schankerman, 2001; Bessen and Meurer, 2005) and at least one of the German system (Cremers, 2004) . All of these studies document the fact that litigated patents tend to be the more valuable patents, as one might have expected. The US studies also show that only about 5 per cent of such suits go to trial, with the remainder being settled before going to trial. They also show that whether patent litigation has increased depends on whether it is measured in aggregate or per patent. That is, the increase in patent litigation has roughly paralleled the increase in patenting, at least in the United States, although there is some indication that the litigation rate has risen in the very recent past (Bessen and Meurer, 2005) .
+
+## III. Do patents encourage innovation?
+
+The economic view of patents is that they offer a bargain between society and the inventor: in return for a limited period of exclusivity, the inventor agrees to make his invention public rather than keeping it secret. Two questions immediately arise from this: first, what is the optimal design of such a policy instrument? and, second, are patents effective at accomplishing this task? Much of recent economic research on patents has been directed to attempting to answer these questions and has found that the trade-off between a ‘short-term monopoly’ and
+
+$^{6}$ See Farrell and Shapiro (2007) for detailed models of this process.
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+572
+
+Bronwyn H. Hall
+
+Table 1: The patent system trade-offs
+
+<table><tr><td>Effects on:</td><td>Benefit</td><td>Cost</td></tr><tr><td>Innovation</td><td>creates an incentive for R&amp;D; promotes the diffusion of ideas</td><td>mpedes the combination of new ideas and inventions; raises transaction costs</td></tr><tr><td>Competition</td><td>facilitates entry of new small firms with limited assets; allows trading of inventive knowledge, markets for technology</td><td>creates short-term monopolies, which may become long-term in network industries</td></tr></table>
+
+
+`innovation incentive' is much more complex than the usual textbook treatment, making the optimal design problem extremely difficult.
+
+Table 1 presents a framework for thinking about the costs and benefits of the patent system in two dimensions, innovation and competition. The plain boxes represent the traditional trade-off between negative effects for competition arising from market power and the positive effects for innovation, while the shaded boxes represent newer views of the ways in which patents might actually discourage innovation but enable and encourage competition.
+
+Thus this table suggests that, in addition to the familiar arguments that patents increase innovation via incentive effects and diffusion, and decrease competition because they create temporary monopolies, there are offsetting effects in both cases, effects that have become more apparent in recent years. $^7$ These offsetting effects are the tendency of patents to increase the costs of subsequent innovators, especially when these innovators need to combine inventions from many sources, as well as the fact that patents may help competition by facilitating the vertical disintegration of knowledge-intensive industries and helping new entrants. The next two sections of the article review the evidence, both theoretical and empirical, on the effectiveness of the current patent systems in achieving their goals. This is followed by a discussion of the theoretical literature on optimal design of innovation incentive systems.
+
+## (i) Theoretical evidence
+
+As is often the case with models that admit the complexity of the world, the theoretical literature in this area produces ambiguous results with respect to incentives provided by patents. In the simplest case, where a patent corresponds to a single product and knowledge is not particularly cumulative, it is clear that patents will encourage innovation. Offering individuals the short-term right to exclude others from practising an invention provides the inventors with the opportunity to earn rents or supranormal profits when they innovate that
+
+7 This is not to say that these effects have gone completely unrecognized in the past. Consider the following quotation from a sugar manufacturer in Great Britain during the nineteenth century: `In the manufacture with which I am connected—the sugar trade—there are somewhere like 300 or 400 patents. Now, how are we to know all these 400 patents? How are we to manage continually, in the natural process of making improvements in manufacture, to know which of these patents we are at any time conflicting with? So far as I know, we are not violating any patent; but really, if we are to be exceedingly earnest in the question, probably we would require to have a highly paid clerk in London continually analysing the various patents; and every year, by the multiplication of patents, this difficulty is becoming more formidable.’ (R. A. Macfie, quoted in `Is the Granting of Patents for Inventions Conducive to the Interests of Trade?', Transactions of the National Association for the Promotion of Social Science (George W. Hastings, ed., 1865, p. 666).)
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+573
+
+are higher than those they would earn if there were immediate free entry into imitation of their invention.
+
+The early theoretical industrial organization literature on patent races even seemed to suggest that patents produced too much innovation (Wright, 1983; Reinganum, 1989) . However, models that incorporate the cumulative nature of innovation or dispersed ownership of the patents required for producing a new good yield more ambiguous results (Judd, 1985; Scotchmer 1991; Bessen and Maskin, 2006) . A modest theoretical literature pioneered by Scotchmer (1991) has developed that analyses a number of such models. Two main cases and their variations have been considered: those where two innovative stages are required to produce the product (the `research tool' case) and those where there is a sequence of products, each of which is an improvement over the previous product (the cumulative or `quality ladder' case). In either case, it is possible that any particular invention uses one or more other inventions as input, or is an input to one or more future inventions. This type of analysis has increased in importance because of the complexity of modern technology and also because of growth in patent use in sectors that traditionally had regarded patent protection as relatively unimportant. Briefly described, the new setting is one where a single product involves hundreds of patents, and where one innovation builds directly on many others. Neither feature is really new, but both have assumed increasing importance in a number of technology areas, such as information technology and biotechnology.
+
+When development of an innovative product requires multiple patent inputs, Heller and Eisenberg (1998) have argued forcefully that the licensing solution may fail because of transactions costs if a large number of patent-holders are involved. One consequence of this fragmentation threat may be increased defensive patenting by the product developer in order to be able to threaten a counter suit if attacked. Empirical evidence for this proposition has been provided by Ziedonis (2004) in the context of the semiconductor industry.
+
+A recent paper by Bessen and Maskin (2006) develops a model of sequential (cumulative) and complementary innovation in a differentiated product setting and analyses the effect of patent protection in two settings, non-sequential and infinitely sequential (that is, where each invention builds on the preceding invention). The results are somewhat complex but intuitive: in the static non-sequential case, having patents generally yields higher welfare than not having them. But in the fully sequential case, the equilibrium without patents has higher welfare and innovation than the equilibrium with patents if the upper tail of the distribution of innovation values is sufficiently thick. They show that the two commonly used distributions for innovation value, lognormal and Pareto, satisfy this condition. $^8$ They also show that, in some cases, even the original innovator may benefit from the absence of patents in the dynamic case, because he receives spillovers from follow-on innovation. Empirical evidence that suggests some positive value to this kind of spillover has been offered recently by Belenzon (2006) using sequences of patent citations to measure knowledge flows in and out of the firm.
+
+### (ii) Empirical evidence
+
+The question posed at the beginning of this section has also proved difficult to answer empirically, largely because of the absence of real experiments, at least since the nineteenth century. Some researchers have looked at historical eras when there were changes to the
+
+8 Grabowski and Vernon (1994), Hall et al. (2005), and Harhoff et al. (1999) provide evidence on the distribution of innovation and patent value.
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+574
+
+Bronwyn H. Hall
+
+system and examined the consequences for subsequent innovative activity, measured either by patenting in a jurisdiction not affected by the changes to the system or by invention counts obtained independently (Lerner, 2002; Moser, 2005) . A second widely used approach is to survey firm managers, asking about their patent use and the use of patents in their industry (Mansfield, 1986; Levin et al. , 1987; Cohen et al. , 2002; Arundel, 2003) . Using these kinds of survey data matched to R & D spending and innovation outcomes, more econometric model-based approaches have been pursued by Baldwin et al. (2000), Arora et al. (2003), and Bloom et al. (2005), among others.
+
+A few conclusions have emerged from this body of work. First, introducing or strengthening a patent system (lengthening the patent term, broadening subject matter coverage or available scope, improving enforcement) unambiguously results in an increase in patenting and also in the use of patents as a tool of firm strategy (Hall and Ziedonis, 2001; Lerner, 2002) . This is to be expected, but it is much less clear that these changes result in an increase in innovative activity (Lerner 2002; Baldwin et al. , 2000) , although they may redirect such activity toward things that are patentable and away from those than can be kept secret within the firm (Moser, 2005) . Sakakibara and Branstetter (2001) studied the effects of expanding patent scope by allowing multiple claims in Japan in 1988 and found that this change to the patent system had a very small effect on R & D activity in Japanese firms.
+
+Exceptions to this conclusion are two studies based on cross-country data: Park and Ginarte (1997) and Kanwar and Evenson (2003) . Park and Ginarte (1997) use aggregate data for 60 countries during 1960 – 90 and an index of the strength of IPR (subject matter coverage, term length, etc.) which they developed. Using a simultaneous equations model of economic growth, investment, schooling, and R & D investment, they found that the strength of IPR was positively associated with investment and R & D investment in countries with above median income but not for the less-developed countries. IPR had no independent effect on growth above and beyond that contributed by investment and R & D. However, Park and Ginarte (1997) also show that the strength of IPR in high-income countries (but not in low-income countries) can be predicted by prior R & D intensity, which raises some questions about the simultaneity of IP protection and a country's orientation towards R & D and innovation. That is, it is possible that the demand for IP protection increases when a large share of the industrial base is engaged in innovative activities.
+
+The study by Kanwar and Evenson (2003) looks at the variation across country in R&D spending as a function of the Ginarte–Park index over the 1981–95 period and finds similar results, with stronger IP protection related to higher R&D intensity. Although well done in many respects, this study makes no attempt to explore the potential endogeneity of the relationship, nor does it control for the level of development of the countries, which arguably drives both R&D and the development of IP institutions.
+
+In contrast to the two results cited above, Qian (2007) performs a similar analysis for pharmaceutical patents in 85 countries over the period 1978 – 99, but using matched sampling and also fixed country-effect estimators. She finds that national patent protection does not stimulate domestic innovation activities, except at higher development levels, and that above a certain level of patent protection, innovation activities are actually reduced.
+
+A third finding from the empirical literature is that if there is an increase in innovation due to patents, it is likely to be centred in the pharmaceutical, biotechnology, and medical instrument areas, and possibly specialty chemicals. This conclusion relies mostly on survey evidence from a number of countries which shows rather conclusively that patents are not among the important means to appropriate returns to innovation, except perhaps in pharmaceuticals, medical devices, and some specialty chemicals (Mansfield, 1986; Levin et al. , 1987; Arora
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+575
+
+et al. , 2001; Cohen et al. , 2002) . Using a structural model that combines survey responses with accounting data on R & D, Arora et al. (2003) found that increasing the patent premium, which they define as the difference in pay-offs to patented and unpatented inventions, can be expected to increase R & D in most manufacturing sectors, with the greatest increase in medical instruments, followed by biotechnology and pharmaceuticals.
+
+Fourth and finally, the existence and strength of the patent system affects the organization of industry, by allowing trade in knowledge, which facilitates the vertical disintegration of knowledge-based industries and the entry of new firms that possess only intangible assets (Hall and Ziedonis, 2001; Arora et al. , 2003; Arora and Merges, 2004) . The argument is that, by creating a strong property right for the intangible asset, the patent system enables activities that formerly had to be kept within the firm because of secrecy and contracting problems to move out into separate entities. Although limited, research in this area supports this conclusion in the chemical and semiconductor industries.
+
+Thus the bottom line from the empirical evidence is that the patent system provides clear incentives for innovation in only a few sectors, but that firms and industries do respond to its presence, both by making use of the system and by sometimes tailoring their innovative strategies to its presence. As Edith Penrose said in 1951 when speaking to the same question,
+
+If national patent laws did not exist, it would be difficult to make a conclusive case for introducing them; but the fact that they do exist shifts the burden of proof and it is equally difficult to make a really conclusive case for abolishing them. (Penrose, 1951)
+
+One possible interpretation of this remark is that history matters, in the sense that industrial organization and firms adapt to the institutional regime in which they operate and changing this regime, whatever it is, involves substantial short-term costs that may not be outweighed by the long-term benefits.
+
+## IV. Optimal patent design
+
+As has been noted by many others (for example, Wright, 1983; Shavell and van Ypserele, 2001) , there are alternative means of providing incentives for innovation, such as prizes or research contracts, but these tend to be of limited value when the goal of the inventive activity is unknown or hard to identify ex ante . See Scotchmer (2005, ch. 2) for an excellent discussion of the alternative incentive systems.
+
+The seminal work on optimal patent design was Nordhaus (1969) , which considered two policy instruments: the length of the patent term and the breadth of the patent, i.e. the range or scope of the inventions covered. The broader the scope of a patent, the larger the number of competing products and processes that will infringe the patent, and the larger the market power of the patentholder. The greater the length of a patent, the longer the period over which the firm can earn monopoly profits. Later work by Gilbert and Shapiro (1990) and Klemperer (1990) built on and extended his method of analysis. Unfortunately, even though all three sets of authors simplified the problem by assuming that a patent corresponds to a product and that there is no uncertainty, the welfare conclusions still turn on assumptions about the nature of the product market (how competitive it is) and the existence of close substitutes for the patented product. The principal conclusion from this line of work is that optimal patent design is likely to depend on the nature of the product market and the technology, which is inconsistent with long-standing practice and policy in most patent systems. Historically, the
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+576
+
+Bronwyn H. Hall
+
+only important exception to the homogeneous treatment of technologies is the extreme case of excluding some of them (such as pharmaceutical products, medical practices, or disembodied software, at various times and in various countries) completely from the system.
+
+The design of patent-based incentive systems has now been extended in two directions: the first looks at the sequential innovation problem and the second focuses on incentivecompatible schemes for eliciting the private information of inventors about value. Green and Scotchmer (1995) analyse the case of sequential innovation with two innovators, showing that it is difficult, if not impossible, to set incentives at the correct level for both the first and subsequent innovators. In general, they find that ex ante agreements between the two innovators have the potential to enhance innovation, and that in the case where the first technology is basic, ex ante agreements are welfare enhancing because they allow profit to be transferred to the first innovator. This in turn compensates him for the shortened patent life and allows the overall patent life to be kept to a minimum given the need for both innovators to cover their costs. Of course, the difficulty with this analytic set-up is the fact that subsequent innovators can rarely be identified at the time the first innovation is undertaken.
+
+O’Donoghue et al. (1998) studied the length – breadth trade-off in the sequential innovation case with licensing. In this case, follow-on innovations can shorten the effective life of a patent, so length and breadth are interrelated rather than being two separate policy tools. They find that very broad but finite-lived patents improve innovation and diffusion, but that long and narrow patents (which are easily displaced by similar but not identical inventions) can lower R & D costs by encouraging effort toward larger innovative steps.
+
+A promising line of work in this area is the development of incentive-compatible patent renewal or compulsory licensing mechanisms designed to elicit information from inventors about the approximate real value of their innovation. These models assume both that there is uncertainty ex ante about value and the R & D effort that will be undertaken and that the inventor has more information about value ex post . Cornelli and Schankerman (1999) show that in this case it is optimal for the government to offer firms a menu of patent lives and associated lump-sum patent fees. When there is no post-patent learning by the firm, this mechanism is equivalent to offering a schedule of annual renewal fees. Even in the usual case where the value of the invention is revealed over time, renewal fees may be a useful mechanism for encouraging the release of inventions into the public domain. Using simulation, Cornelli and Schankerman show that existing patent renewal fee schedules rise much too slowly with patent life given the likely distribution of invention value.
+
+Hopenhayn and Mitchell (2001) showed that such a scheme may be dominated by a menu that trades off the breadth and length of a patent, with the inventors of more fertile innovations choosing to receive broader but shorter patents. However, because of the difficulty of producing a workable definition of breadth in practice, this idea is far more difficult to implement than a set of menus involving only renewal fees and patent terms. Recently, Hopenhayn et al. (2007) have developed a more complex information-revelation scheme in the setting where innovation is cumulative.
+
+## V. Recent patenting trends
+
+### (i) Policy changes
+
+As with (almost) all governmental institutions, patent systems have evolved and continue to evolve in ways that are ultimately driven by forces related both to a perception of increased
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+577
+
+global competition, especially in knowledge-intensive sectors, and also to technological change itself. Such changes as the expansion of subject-matter coverage, strengthening of enforcement systems, and the encouragement of patenting by upstream actors can all be seen as driven by these forces. As is often the case with innovation, it is also true that many of the changes in patenting strategy observed around the world have originated in the United States and then diffused elsewhere, both via imitation and also via the process of intergovernmental negotiation. This section of the paper discusses these trends and their implications.
+
+Unfortunately (from the perspective of optimal policy), many of the changes in patent policy in the United States during the past two decades have been as a result of court decisions, especially those of the Court of Appeals of the Federal Circuit (CAFC), and to a lesser extent by the Supreme Court. Addressed as they are to the features of individual cases, these decisions do not always consider the broader policy implications as they set precedents. As a result of a series of court decisions by these bodies, the subject matter eligible for patenting has been extended to new technologies (biotechnology), technologies not previously subject to patent protection (business methods, software), and to upstream scientific research tools, materials, and discoveries (Madey v. Duke, 2002). The rights of patentholders vis-à-vis alleged infringers have been strengthened by such decisions as Polaroid v. Kodak (1986/1991), which yielded a major damage award to Polaroid and shut down the instant camera business of Kodak. However, recent decisions by the Supreme Court have reversed this trend to some extent. $^9$
+
+Of course, in many ways these court decisions were the consequence of legislative changes in 1982, during which the CAFC was created, and the position of patent holders strengthened by a number of procedural changes in the courts. In a comparison of appeals cases from 1953 to 1978 and from 1982 to 1990, the share of District Court decisions finding validity and infringement that were upheld by the higher court increased from 62 to 90 per cent. Decisions of invalidity and no infringement were reversed 12 per cent of the time before the Federal Circuit's creation and 18 per cent afterwards (Lerner, 1995). Moreover, the rate of preliminary injunctions increased dramatically (Lanjouw and Lerner, 1998). In a recent study using modern time-series methods and taking account of the sequence of appeals decisions, Henry and Turner (2006) found robust evidence that the CAFC was less likely to invalidate patents than earlier courts, but no less likely to affirm or reverse infringement findings. Some commentators have argued that a specialized patent court is more likely to be ‘captured’ by the patent bar and those whose interests are served by strong patents of any kind, and that the potential for this outcome should be borne in mind when considering creation of a similar court for European patents.
+
+The early 1980s was also the period when the well-known Bayh $-$ Dole Act passed into the law, leading to the emergence of new players,such as universities and public research institutions, as well as an increase in activity at institutions that had already been patenting some of their research results. Although the impact and importance of this Act remain controversial in the United States (see, for example, Mowery et al. , 2004) , it has been seen as a model by many other countries who are anxious to improve their record of commercializing university research and a number of them have experimented with similar changes (Geuna and Nesta, 2006) . A full discussion of the issues associated with the patenting of university
+
+9 KSR International v Teleflex Inc. (No. 04-1350) 119 Fed. Appx 282, on non-obviousness, and eBay Inc, et al. v. MercExchange, L. L. C. (No. 05-130) 401 F. 3d 1323, on the four-factor test for injunctions.
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+578
+
+Bronwyn H. Hall
+
+research outputs is beyond the scope of this article and the reader is referred to Siegel et al. (2007, this issue).
+
+As discussed earlier, the economic view of patents is that they represent a trade-off between the costs of granting limited market power to firms versus the benefits of encouraging innovation, in much the same way that competition policy is designed to maintain a balance between the costs of increased market power from concentration and the benefits of scale efficiencies. It should come as no surprise, then, that patent policy and competition policy are occasionally misaligned, and increasingly need to be looked at simultaneously.
+
+In the area of competition policy, from 1981 onwards there has been a marked evolution in the attitudes of the US Justice Department’s Antitrust Division and the Federal Trade Commission (FTC) towards business conduct involving patents, resulting in a much more nuanced and pro-patent position (FTC, 2003). Following changes to practice in 1981 and 1988, in 1995 the Justice Department and the FTC jointly issued Antitrust Guidelines for the Licensing of Intellectual Property, declaring that ‘the Agencies do not presume that intellectual property creates market power in the antitrust context’ and intellectual property licensing is ‘generally pro-competitive’. Similarly, new licensing guidelines were issued by the European Commission in 2004 that provide a safe harbour for many licensing agreements, especially those involving firms with low market shares. For useful comparisons of the two sets of provisions and a discussion of the differences, see Gilbert (2004) and Delrahim (2004). The Korean and Japanese Fair Trade Commissions also have what amounts to a ‘rule of reason’ approach to the regulation of such agreements.
+
+Taken together, these changes add up to a considerable strengthening of patent holder rights and a broadening of the reach of the patent system. As I summarize in the next section, the response to these changes on the part of private firms has been dramatic.
+
+### (ii) Strategic response
+
+The most obvious response to these changes in the US patent system was the increase in patenting across many sectors, leading to a doubling of US patent applications and grants during the 10-year period between 1992 and 2002. In Hall (2005), I used a simple time-series analysis to show that the time series of aggregate patent applications in the USA displayed a structural break in 1984, with the annual growth rate increasing from zero to over 6 per cent. Such a growth rate will produce a doubling in 12 years. I also showed that most of the growth was due to increased patenting by firms in the information and communication technology (ICT) sectors, which is consistent with the view that much of it is for defensive reasons (Arora et al. , 2001; Hall and Ziedonis, 2001; Hicks et al. , 2001) . At the same time, the contribution of increased university and public research institution patenting to growth was relatively small, not because such patenting did not increase, but because the share remains small. From a regional perspective, over half the growth was due to inventors in the United States, one-third to those in Asia, and the small remainder to inventors in Europe. Thus the growth in patenting at the USPTO through 2002 was driven by the behaviour of the ICT sector in the USA and Asia.
+
+Recent data from WIPO confirm that this growth trend is worldwide (WIPO, 2007) . Applications at the EPO tripled in the 20 years between 1985 and 2005, while those at the Japanese Patent Office grew somewhat earlier, at 10 per cent per annum between 1980 and 1985 and then 50 per cent between 1985 and 2005. Figures 1 and 2 show that applications worldwide have doubled between 1994 and 2005, whereas grants have only grown by 50 per
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+579
+
+Figure 1: Worldwide patent filings
+
+![Figure](figures/figure_003.png)
+
+Source: WIPO Statistics Database.
+
+Figure 2: Patents granted worldwide
+
+![Figure](figures/figure_006.png)
+
+Source: WIPO Statistics Database.
+
+cent during the same period, probably because of increased backlogs in patent offices around the world.
+
+A number of other behavioural changes have accompanied this increase in patenting: slightly higher renewal rates, more frequent assertion of patents, a doubling of US District Court patent suits between 1988 and 2001, and some evidence that the probability of a suit per patent has increased recently (Bessen and Meurer, 2005) . The complexity of patents in terms of number of claims and citations of prior art has grown, and patentees tend to invest more in the processes of application and examination. In testimony before Congress, the current director of the USPTO, Jon Dudas, reported that more than 100,000 of the 355,000 patent applications filed in 2004 were continuations of applications that had been previously reviewed by an examiner. He also reported on the problem of ‘super-sized’ applications submitted by a minority of applicants (7 per cent of the applications account for 25 per cent of the claims examined; some are submitted on CD-ROMs with thousands of claims).
+
+The super-size phenomenon has also been documented elsewhere by van Zeebroeck et al. (2006) , who report that applications of over 1,000 pages are now frequently filed at the EPO and other patent offices around the world, and several applications have even reached 100,000
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+580
+
+Bronwyn H. Hall
+
+pages or up to 20,000 claims. They looked at the sources of the surge in the number and size of patent applications and conclude that a major factor driving the growth is the diffusion of the US patent application model via the PCT. Nagaoka (2006) reported that the number of claims per patent application in Japan tripled between 1990 and 2003.
+
+In addition, many critics have argued that the sheer volume of patent applications threatens to degrade the quality of issued patents, or lengthen the backlog, or both. On the backlog there is no doubt. In April 2005, Dudas reported that pendency in data-processing technologies stood at three years and growing, and that without intervention, the current backlog of applications awaiting first review could double from 500,000 to a million in the next 5 years. The EPO grant lags were already long and are getting longer. During the 6 years between 1998 and 2006, the waiting time for examination at the Japanese Patent Office increased from 19 to 26 months (Nagaoka, 2006) .
+
+Finding hard evidence of a decline in quality is more difficult, although a number of legal scholars and practitioners have been vocal on the subject (for example, Hunt, 1999; Barton, 2000; Kingston, 2001; Lunney, 2001; Jaffe and Lerner, 2004) . There are several reasons to believe that quality (especially the application of the non-obviousness and novelty criteria of patentability) has suffered as the number of applications has grown. First, the number of patent examiners has not kept pace with the increase in workload represented by the increased number and growing complexity of the applications. Second, in the United States there does seem to have been a dilution of the application of the non-obviousness standard in biotechnology (due to court decisions) and some limitations on applying it properly to business-method patent applications, for, among other reasons, the absence of adequate written prior-art documents. Recent changes in the treatment of genomic and business-method applications were introduced at the USPTO (the second pair of eyes for business-method patents and the requirement of a specific application or use for a new gene sequence) that resulted in a slowing down of patent grants in those fields, suggesting that the previous bar may have been set too low.
+
+During the US FTC/Department of Justice hearings on the patent system and antitrust policy in 2002, a number of industry representatives expressed concerns about the difficulty of negotiating the patent thicket in their area and the risk of being `held up' ex post by a patent on a technology that was only a small component of their product. This complaint was heard largely from those in the complex product industries (the ICT sector), such as Robert Barr, then Vice-president for Intellectual Property and Worldwide Patent Counsel at Cisco Corporation. He described two types of problems faced by firms in the sector: the first being the large stockpiling of patents necessary as a defensive measure against others in the industry, and the second the threat posed by small entities that have nothing at risk themselves and may not even be producers. On the first, Barr says the following:
+
+My observation is that patents have not been a positive force in stimulating innovation at Cisco. . . . Everything we have done to create new products would have been done even if we could not obtain patents on the innovations and inventions contained in these products. . . . The only practical response to this problem of unintentional and sometimes unavoidable patent infringement is to file hundreds of patents each year ourselves, so that we can have something to bring to the table in cross-licensing negotiations. . . . The time and money we spend on patent filings, prosecution, and
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+581
+
+maintenance, litigation and licensing could be better spent on product development and research leading to more innovation. $^10$
+
+On the second problem (that of being attacked ex post by a small entity that does not face much risk of infringement itself):
+
+stockpiling patents does not really solve the problem of unintentional patent infringement through independent development. If we are accused of infringement by a patent holder who does not make and sell products, or who sells in much smaller volume than we do, our patents do not have sufficient value to the other party to deter a lawsuit or reduce the amount of money demanded by the other company. 11
+
+The first of the problems Barr describes is clearly a case of mutually assured destruction that leaves the firms in question no better (and no worse) off than if they were not accumulating massive numbers of patents for defensive purposes, and yet at the same time is a very costly strategy. Increasing the administrative costs of patents to firms or reforms within the industry itself to discourage this behaviour would seem to be the obvious solution, since it would be in the interest of all firms involved to reduce spending on this activity. However, the second problem is more controversial: the small entities that assert patents in this way may have legitimate claims to ownership of some of the technology in a large firm's product. Some observers have even questioned how common this kind of patent assertion is. Nevertheless, the IT industry in general has been very concerned about these kinds of threats and their consequences for the incentives to invest in complex technologies that might potentially incorporate a piece of technology which leads to a dispute that cannot be resolved by cross-licensing.
+
+Concern over patent hold-up is especially relevant in the case of standards technologies, which are difficult to invent around and which are very important in this sector (Shapiro, 2001) . This problem has caught the attention of many firms worldwide because of the highly publicized and controversial litigation over DRAM standards initiated by Rambus against firms such as Hynix, Samsung, and Micron Technology. The litigation is over patents whose applications were filed while Rambus was participating in the standards-setting committee (JEDEC). The EPO overturned one of the relevant patents in an opposition case and the US FTC initiated a successful complaint over their behaviour in this case, but some of the cases are still in the courts. $^12$ Nagaoka (2006) reported that the practice of applying for patents after a standard has been chosen is widespread: using US and worldwide patent family data, he showed that 50 per cent of the essential patents covering the MPEG2, DVD, and W-CDMA standards were applied for after the specifications were set.
+
+The final area where changes in patenting practice and IP management have raised concern in policy circles is the increased patenting of `research tools' and the consequences of this. Walsh et al. (2003) interviewed some 70 players in the biotechnology research area and found that, by and large, IP in biotechnology is being managed relatively successfully. Because of increasing patent assertion and the extension of patentability to life-forms and gene sequences, the associated costs of research are somewhat higher and research can sometimes be slowed, but it is rarely blocked altogether. There are, however, occasional cases of restricted access
+
+$^{10}$ Barr (2002, pp. 675-7).
+
+11 Barr (2002, pp. 679-80).
+
+12 MIP Weekly News, 15 February 2004 (EPO) and 8 August 2006 (FTC), available at http://www.managingip.com
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+582
+
+Bronwyn H. Hall
+
+to foundational discoveries and to some diagnostic genetic tests. A number of `working solutions' have evolved, including negotiated licences and royalty payments. Patents are also circumvented by inventing around them, using substitute research tools, and locating research activity offshore. Institutional responses include the National Institutes of Health guidelines encouraging research grantees to facilitate access to patented research tools and the steps taken by several research organizations to place results in the public domain, where they become patent-defeating prior art.
+
+## VI. The TRIPS controversy
+
+The TRIPS agreement that was incorporated into the charter of the WTO in 1994 essentially sets uniform patent standards throughout the membership of the WTO, although in the case of developing countries it allows considerable delay in adopting these standards. 13 The agreement was negotiated without a great deal of economic input, and the idea of uniform standards has proved very controversial among economists. The most obvious argument against the principle behind TRIPS is that the trade-off between the costs of market power granted by a patent and the benefits of the innovation incentive is not likely to be uniform across different development levels, so that a one-size-fits-all rule is far from optimal. A secondary argument is that for a small less-developed country, the fixed cost of operating a patent office takes resources and trained personnel away from more productive activities. Analysis is also complicated by the fact that the optimal policy needs to be determined in an open-economy setting, where costs are incurred locally but benefits may spill globally.
+
+Grossman and Lai (2004) analysed a simple version of the problem, with only two countries (developed, or North, and less developed, or South). In a non-cooperative equilibrium, North will choose a stronger patent policy under reasonable conditions (North’s market is at least as large as South’s and its R & D capability is greater). The global welfare optimum has multiple outcomes, but, without transfers, the outcomes which favour the North are those with longer patents in the South. Using a similar set-up, Scotchmer (2004) shows that harmonization generally strengthens IP protection in all countries, and that because there are no similar institutions to manage the benefits of public support for innovation, the tendency is towards too much protection and too little public spending with respect to a (global) social optimum.
+
+A useful survey of the empirical evidence on the incentive effects of patents in the developing-country context is provided by Branstetter (2004) . He finds that there is little evidence of a strong incentive effect for domestic innovators, but some acceleration in the diffusion of advanced technology by multinational firms (who feel more protected in the presence of a functioning patent system). Although this result suggests that spillover effects from foreign direct investment may be enhanced, it also means that the monopoly rent from a given innovation has been enhanced, essentially promoting innovation in the developed world rather than the developing world.
+
+There is also the previously cited evidence of Qian (2007) with respect to pharmaceuticals, which shows that the incentive effects of patents depend on the development level of the country. The historical experience of the USA in the early nineteenth century (when national treatment was not available to foreign inventors) and Taiwan in the twentieth century (where
+
+13 Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, Agreement on Trade-related Aspects of Intellectual Property Rights, 15 April 1994, 33 I.L.M. (1994) 81.
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+583
+
+patents were unimportant until the mid-1980s, after a certain level of industrial development had been reached) confirms that where technology imitation is an important tool of catching up, it may be optimal to wait before introducing strong patent protection.
+
+## VII. Conclusion
+
+This article has offered a brief tour of patent systems, past, present, and immediate future. As one reflects on the results of economic research in this area, one is struck by the not inconsiderable tension between what we know about patents as an innovation incentive and the general thrust of contemporary patent policy. This tension is visible in two main areas: the expansion of patenting activity in several dimensions (subject matter coverage, type of patenting entity, and geographic) and its `one-size-fits-all' nature.
+
+Thus, although the research I have surveyed does not indicate an overwhelmingly important role for patents in encouraging innovation except in a few sectors, there is no doubt that the general policy stance of many governments is to encourage firms, individuals, and research organizations to learn about the patent system and apply for patents. To cite one example, the European Union has had an initiative in place for several years designed to educate small and medium-sized enterprises in the use of the patent system. A number of countries have adopted measures to encourage patenting and technology licensing by universities and governmental research organizations, modelling them on the well-known Bayh–Dole Act in the United States (Geuna and Nesta, 2006) . And, of course, the TRIPS agreement clearly was intended to raise the level of patenting activity in developing countries by guaranteeing a minimum level of protection throughout the global economy. At the level of the individual agent, this is undoubtedly the right strategy, even if social welfare is not necessarily increased by a massive increase in patenting activity.
+
+Given the importance of knowledge assets in modern-day economies, some of the increase in patenting and its importance that we have observed is obviously due to increased inventive activity. It is unlikely that patent systems will disappear or be completely replaced by alternative incentive systems for eliciting this kind of activity. Nevertheless, one has to ask whether the marginal scientist/engineer is best employed doing R & D or examining patents. That is, would society as a whole be better off if the inventive step requirement for a patent were raised, leading to fewer patents but without discouraging real innovations?
+
+The evidence surveyed here also leads inexorably to the conclusion that a significant problem for policy-makers is the heterogeneity of responses to the system, a heterogeneity that is firmly grounded in the heterogeneity of technology and its development. The debate presently taking place in the United States over patent reform highlights the problem: pharmaceutical firms, among others, find that the present system works well for them and are opposed to any changes designed to improve its operation for firms in `complex' technology industries such as telecommunications and computing. ICT firms, on the other hand, seem to view the system as a necessary evil, requiring costly investments in patent portfolio building for defensive purpose while using other methods to secure returns to their own innovations. Many of these firms support reforms to the system that are designed to mitigate the problems which arise when a product contains many minor inventions and relies on a number of standards that may be covered by patents.
+
+In spite of the fact that the cost and nature of invention clearly varies across technological field, patent law and international agreements essentially agree that the same level of patent protection should be available for all technologies without discrimination. However, as Burk
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+584
+
+Bronwyn H. Hall
+
+and Lemley (2002) point out, application of the law by patent offices and the courts can lead to substantial variation across technologies in practice. Given the likely response of firms and inventors to attempts to define statutory differences in treatment across technologies, this type of flexibility is perhaps the best we can do. But the problem of heterogeneity and whether there is something that can be done about it remains a subject for future policy research.
+
+## References
+
+Arora, A., and Merges, R. P. (2004), 'Specialized Supply Firms, Property Rights, and Firm Boundaries', Industrial and Corporate Change, 13, 451-75.
+
+— Ceccagnoli, M., and Cohen, W. (2003), 'R&D and the Patent Premium', Working Paper No. 9431, Cambridge, MA, NBER.
+
+— Fosfuri, A., and Gambardella, A. (2001), Markets for Technology: The Economics of Innovation and Corporate Strategy, Cambridge, MA, MIT Press.
+
+Arundel, A. (2003), Patents in the Knowledge-based Economy: Report of the Know Survey, Maastricht, MERIT, University of Maastricht.
+
+Baldwin, J. R., Hanl, P., and Sabourin, D. (2000), ‘Determinants of Innovative Activity in Canadian Manufacturing Firms: The Role of Intellectual Property Rights’, Working Paper No. 122, Ottawa, Statistics Canada.
+
+Barr, R. (2002), Testimony at the US Federal Trade Commission/Department of Justice Hearings on Competition and Intellectual Property Law and Policy in the Knowledge-Based Economy, University of California at Berkeley, 28 February.
+
+Barton, J. H. (2000), 'Reforming the Patent System', Science, 287, 1933-4.
+
+Belenzon, S. (2006), ‘Knowledge Flow and Sequential Innovation: Implications for Technology Diffusion, R&D and Market Value’, Nuffield College, Oxford University, available at http://www.economics.ox.ac.uk/Research/wp/pdf/paper259.pdf
+
+Bessen, J., and Maskin, E. (2006), 'Sequential Innovation, Patents, and Imitation', Working Paper, Boston University School of Law and Princeton University.
+
+— Meurer, M. (2005), ‘The Patent Litigation Explosion’, Law and Economics Working Paper No. 05-18, Boston University School of Law.
+
+Bloom, N., Van Reenen, J., and Schankerman, M. (2005), 'Identifying Technology Spillovers and Product Market Rivalry', Discussion Paper No. 3916, London, CEPR.
+
+Branstetter, L. G. (2004), 'Do Stronger Patents Induce More Local Innovation?', Journal of International Law, 7(2), 359-70.
+
+Burk, D., and Lemley, M. (2002), 'Is Patent Law Technology-specific?', Berkeley Technology Law Journal, 17, 1155-206.
+
+Cohen, W. M., Goto, A., Nagata, A., Nelson, R. R., and Walsh, J. P. (2002), 'R&D Spillovers, Patents and the Incentives to Innovate in Japan and the United States', Research Policy, 31, 1349-67.
+
+Commission on Intellectual Property Rights (2002), Integrating Intellectual Property Rights and Development Policy, London, Commission on Intellectual Property Rights, September.
+
+Cornelli, F., and Schankerman, M. (1999), 'Patent Renewals and R&D Incentives', Rand Journal of Economics, 30(2), 197-213.
+
+Cremers, K. (2004), 'Determinants of Patent Litigation in Germany', Discussion Paper No. 04-72, Mannheim, Centre for European Economic Research (ZEW).
+
+Danish Board of Technology (2005), Recommendations for a Patent System of the Future, report by a working group under the Danish Board of Technology, June.
+
+Delrahim, M. (2004), ‘US and EU Approaches to the Antitrust Analysis of Intellectual Property Licensing: Observations from the Enforcement Perspective’, US Department of Justice, available at http://www.usdoj.gov/atr/public/speeches/203228.pdf
+
+EPO (2006), ‘Assessment of the Impact of the European Patent Litigation Agreement (EPLA) on Litigation of European Patents’, European Patent Office, available at http://www.european-patentoffice.org/epo/epla/pdf/impact_assessment_2006_02_v1.pdf
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+585
+
+EPO (2007a), 'Patent System', European Patent Office, available at http://www.epo.org/topics/patentsystem.html (accessed 3 December 2007).
+
+— (2007b), ‘Legislative Initiatives’, available at http://www.epo.org/patents/law/legislative-initiatives.html (accessed 3 December 2007).
+
+Farrell, J., and Shapiro, C. (2007), 'How Strong are Weak Patents?', paper presented at the Conference on the Economics of Competition and Innovation, October, University of California at Berkeley.
+
+FTC (2003), To Promote Innovation—The Proper Balance of Competition and Patent Law and Policy, Washington, DC, Federal Trade Commission, October.
+
+Geuna, A., and Nesta, L. J. J. (2006), 'University Patenting and its Effects on Academic Research: The Emerging European Evidence', Research Policy, 35(6), 790-807.
+
+Gilbert, R. J. (2004), 'Converging Doctrines? US and EU Antitrust Policy for the Licensing of Intellectual Property', Antitrust, 19(1), 51-7.
+
+- Shapiro, C. (1990), 'Optimal Patent Length and Breadth', RAND Journal of Economics, 21, 106-12.
+
+Gowers, A. (2006), Gowers Review of Intellectual Property, London, HMSO, December.
+
+Grabowski, H., and Vernon, J. M. (1994), 'Returns to R&D on New Drug Introductions in the 1980s', Journal of Health Economics, 13, 383-406.
+
+Graham, S. J. H., Hall, B. H., Harhoff, D., and Mowery, D. C. (2003), ‘Post-issue Patent *Quality Control’’: A Comparative Study of US Patent Re-examinations and European Patent Oppositions’, in W. Cohen and S. Merrill (eds), Intellectual Property in the Knowledge-Based Economy, . Washington, DC, National Academies Press.
+
+Green, J., and Scotchmer, S. (1995), 'On the Division of Profit in Sequential Innovation', Rand Journal of Economics, 26, 20-33.
+
+Grossman, G. M., and Lai, E. L.-C. (2004), 'International Protection of Intellectual Property', American Economic Review, 94(5), 1635-53.
+
+Hall, B. H. (2005), 'Exploring the Patent Explosion', Journal of Technology Transfer, 30, 35-48.
+
+— Ziedonis, R. H. (2001), 'The Patent Paradox Revisited: An Empirical Study of Patenting in the US Semiconductor Industry, 1979-1995', RAND Journal of Economics, 32, 101-28.
+
+— Jaffe, A. B., and Trajtenberg, M. (2005), ‘Market Value and Patent Citations’, RAND Journal of Economics, 36, 16-38.
+
+— Graham, S., Harhoff, D., and Mowery, D. C. (2003), ‘Prospects for Improving US Patent Quality via Postgrant Opposition’, Innovation Policy and the Economy, 4, 115–43.
+
+Harhoff, D., and Reitzig, M. (2004), 'Determinants of Opposition against EPO Patent Grants: The Case of Biotechnology and Pharmaceuticals', International Journal of Industrial Organization, 22, 443-80.
+
+— Narin, F., Scherer, F. M., and Vopel, K. (1999), 'Citation Frequency and the Value of Patented Inventions', Review of Economics and Statistics, 81, 511-15.
+
+Heller, M. A., and Eisenberg, R. S. (1998), 'Can Patents Deter Innovation? The Anticommons in Biomedical Research', Science, 280, 698-701.
+
+Henry, M., and Turner, J. L. (2006), 'The Court of Appeals for the Federal Circuit's Impact on Patent Litigation', Journal of Legal Studies, 35, 85-117.
+
+Hicks, D., Breitzman, T., Olivastro, D., and Hamilton, K. (2001), 'The Changing Composition of Innovative Activity in the US—A Portrait Based on Patent Analysis', Research Policy, 30, 681.
+
+Hopenhayn, H., and Mitchell, M. (2001), ‘Innovation Variety and Patent Breadth’, Rand Journal of Economics, 32(1), 152-66.
+
+— Llobet, G., and Mitchell, M. (2007), ‘Rewarding Sequential Innovators: Prizes, Patents, and Buyouts’, Journal of Political Economy, 114 (6), 1041–68.
+
+Hunt, R. M. (1999), 'Nonobviousness and the Incentive to Innovate: An Economic Analysis of Intellectual Property Reform', Federal Reserve Bank Working Paper No. 99-3, Federal Reserve Bank of Philadelphia, Economic Research Division.
+
+IBM (2006), Building a New IP Marketplace, A Global Innovation Outlook Report, September.
+
+Jaffe, A. B., and Lerner, J. (2004), Innovation and Its Discontents: How Our Broken Patent System is Endangering Innovation and Progress, and What to Do About It, Princeton, NJ, Princeton University Press.
+
+James, T. C. (2007), ‘Patent Protection in India: Policy Trends’, presentation to the Conference on Intellectual Property Rights, Globalisation and Related Issues, Delhi School of Economics, India, November.
+
+Japanese Government (2004), Intellectual Property Strategic Program, May.
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+586
+
+Bronwyn H. Hall
+
+JPO (2006), 'A History of System of Industrial Property Rights', Japanese Patent Office, available at http://www.deux.jpo.go.jp/cgi/search.cgi?query=history&lang=en&root=short (accessed 19 December 2006).
+
+Judd, K. L. (1985), 'On the Performance of Patents', Econometrica, 53, 567-95.
+
+Kanwar, S., and Evenson, R. (2003), 'Does Intellectual Property Protection Spur Technical Change?', Oxford Economic Papers, 55, 235-64.
+
+Kingston, W. (2001), 'Innovation Needs Patents Reform', Research Policy, 30(3), 403-23.
+
+Klemperer, P. (1990), 'How Broad should the Scope of Patent Protection Be?', RAND Journal of Economics, 21, 113-30.
+
+Ladas and Parry LLP (2003), 'A Brief History of the Patent Law of the United States', available at http://www.ladas.com/patents/uspatenthistory.html (accessed 19 December 2006).
+
+Lanjouw, J. O., and Lerner, J. (1998), 'The Enforcement of Intellectual Property Rights: A Survey of the Empirical Literature', Annales d'Economie et de Statistique, 49-50, 223-46.
+
+— Schankerman, M. (2001), 'Characteristics of Patent Litigation: A Window on Competition', RAND Journal of Economics, 32, 129-51.
+
+Lerner, J. (1995), 'Patenting in the Shadow of Competitors', Journal of Law and Economics, 38, 463-95.
+
+- (2002), 'Patent Policy Shifts and Innovation over 150 Years', American Economic Review, 92, 221-5.
+
+Levin, R. C., Klevorick, A. K., Nelson, R. R., and Winter, S. G. (1987), 'Appropriating the Returns from Industrial Research and Development', Brookings Papers on Economic Activity, 3, 783-831.
+
+Lunney, G. S., Jr (2001), 'e-Obviousness', Michigan Telecommunications Technology Law Review, 7, 363-422.
+
+MIP News (various dates), Managing Intellectual Property News, available at http://www.managingip.com
+
+Mansfield, E. (1986), 'Patents and Innovation: An Empirical Study', Management Science, 32(2), 173-81.
+
+Maskus, K. (2006), Reforming US Patent Policy, Council on Foreign Relations, November.
+
+Moore, K. A. (2000), 'Judges, Juries, and Patent Cases—An Empirical Peek Inside the Black Box', Michigan Law Review, 99(281), 365-409.
+
+Moser, P. (2005), 'How do Patent Laws Influence Innovation? Evidence from Nineteenth-century World Fairs', American Economic Review, 95, 1214-36.
+
+Mowery, D. C., Nelson, R. R., Sampat, B. N., and Ziedonis, A. A. (2004), Ivory Tower and Industrial Innovation: University-Industry Technology Transfer Before and After the Bayh-Dole Act, Stanford, CA, Stanford University Press.
+
+Nagaoka, S. (2006), ‘Reform of Patent System in Japan and Challenges’, paper presented at the Conference on 21st Century Innovation Systems for Japan and the United States: Lessons from a Decade of Change, Tokyo, Institute of Innovation Research, Hitotsubashi University.
+
+National Academies' Board on Science, Technology, and Economic Policy (2004), A Patent System for the 21st Century.
+
+Nordhaus, W. (1969), Invention, Growth and Welfare: A Theoretical Treatment of Technological Change, Cambridge, MA, MIT Press.
+
+O'Donoghue, T., Scotchmer, S., and Thisse, J.-F. (1998), 'Patent Breadth, Patent Life, and the Pace of Technological Progress', Journal of Economics and Management Strategy, 7(1), 1-32.
+
+OECD (2003), Patents, Innovation and Economic Performance, OECD Conference Proceedings, Paris, Organization for Economic Cooperation and Development.
+
+Park, W. G., and Ginarte, J. C. (1997), 'Intellectual Property Rights and Economic Growth', Contemporary Economic Policy, 15, 51-61.
+
+Penrose, E. T. (1951), The Economics of the International Patent System, Baltimore, MD, Johns Hopkins University Press.
+
+Qian, Y. (2007), 'Do Additional National Patent Laws Stimulate Domestic Innovation in a Global Patenting Environment', Review of Economics and Statistics, forthcoming.
+
+Reinganum, J. F. (1989), 'The Timing of Innovation: Research, Development, and Diffusion', in R. Schmalensee and R. D. Willig (eds), Handbook of Industrial Organization, vol. 1, Amsterdam, NorthHolland.
+
+Sakakibara, M., and Branstetter, L. (2001), 'Do Stronger Patents Induce more Innovation? Evidence from the 1988 Japanese Patent Law Reforms', RAND Journal of Economics, 32, 77-100.
+
+Scotchmer, S. (1991), 'Standing on the Shoulders of Giants', Journal of Economic Perspectives, 5(1), 29-41.
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+
+---
+
+Patents and patent policy
+
+587
+
+Scotchmer, S. (2004), 'The Political Economy of Intellectual Property Treaties', Journal of Law, Economics, and Organization, 20, 415-37.
+
+— (2005), Innovation and Incentives, Cambridge, MA, MIT Press.
+
+Shapiro, C. (2001), 'Navigating the Patent Thicket: Cross Licenses, Patent Pools, Standard-setting', Innovation Policy and the Economy, 1, 119-50.
+
+Shavell, S., and van Ypserele, T. (2001), 'Rewards versus Intellectual Property Rights', Journal of Law and Economics, 44, 525-47.
+
+USPTO (2007), United States Patent and Trademark Office, available at http://www.uspto.gov (accessed 4 January 2007).
+
+Van Zeebroeck, N., and Van Pottelsberghe de la Potterie, B. (2007), 'Filing Strategies and Patent Value', paper presented at the Second Annual Meeting of EPIP, Lund, Sweden.
+
+—— Guellec, D. (2006), ‘Claiming More: The Increased Voluminosity of Patent Applications and Its Determinants’, CEPR Working Paper No. 5971, London.
+
+Walsh, J. P., Arora, A., and Cohen, W. M. (2003), 'Effects of Research Tool Patenting and Licensing on Biomedical Innovation', in W. M. Cohen and S. A. Merrill (eds), Patents in the Knowledge-based Economy, Washington, DC, National Academies Press, 285-340.
+
+WIPO (2007), 'Patent Statistics', World Intellectual Property Organization, available at http://www.wipo.int/ ipstats/en/statistics/patents/ (accessed 30 January 2007).
+
+Wright, B. D. (1983), 'The Economics of Invention Incentives: Patents, Prizes, and Research Contracts', American Economic Review, 73, 691-707.
+
+Ziedonis, R. H. (2004), 'Don't Fence Me In: Fragmented Markets for Technology and the Patent Acquisition Strategies of Firms', Management Science, 50, 804-20.
+
+— (2007), ‘Intellectual Property and Innovation’, forthcoming in S. Scott (ed.), Blackwell Handbook on Technology and Innovation Management, Oxford, Blackwell.
+
+Downloaded from http://oxrep.oxfordjournals.org/ at Stanford University on July 26, 2012
+

processed/00039_W2722564865/document.md

@@ -0,0 +1,504 @@
+10
+
+RIETI Discussion Paper Series 17-E-066
+
+Trends and Priority Shifts in Artificial Intelligence Technology Invention: A global patent analysis
+
+FUJII Hidemichi Nagasaki University MANAGI Shunsuke RIETI
+
+![Figure](figures/figure_004.png)
+
+The Research Institute of Economy, Trade and Industry
+
+http://www.rieti.go.jp/en/
+
+---
+
+RIETI Discussion Paper Series 17-E-066
+
+
+May2017
+
+
+1
+
+# Trends and Priority Shifts in Artificial Intelligence Technology Invention: A global patent  analysis1
+
+FUJII Hidemichi2, MANAGI Shunsuke3
+
+## Abstract
+
+Artificial intelligence (AI) technology can play a critical role in economic development, resource conservation, and environmental protection by increasing efficiency. This study is the first to apply a decomposition framework to clarify the determinants of AI technology invention. Exploiting data from the World Intellectual Property Organization, this study clarifies the determining factors that contribute to AI technology patent publications based on technology type. Consisting of 13,567 AI technology patents for the 2000-2016 period, our worldwide dataset includes patent publication data from the United States, Japan, China, Europe, and the Patent Cooperation Treaty (PCT). We find that priority has shifted from biological- and knowledge-based models to specific mathematical models and other AI technologies, particularly in the United States and Japan. Our technology type and country comparison shows that the characteristics of AI technology patent publication differ among companies and countries.
+
+Keywords : Artificial intelligence, Patent decomposition analysis, Research and development strategy, Biological model, Knowledge based modeling
+
+JEL classification: O30, O34, L00
+
+RIETI Discussion Papers Series aims at widely disseminating research results in the form of professional papers, thereby stimulating lively discussion. The views expressed in the papers are solely those of the author(s), and neither represent those of the organization to which the author(s) belong(s) nor the Research Institute of Economy, Trade and Industry.
+
+↑ This study is conducted as a part of the “Economics of Artificial Intelligence” project undertaken at the Research Institute of Economy, Trade and Industry (RIETI). The author is grateful for helpful comments and suggestions by Makoto Yano (RIETI), Hiroshi Ohashi (University of Tokyo), Masayuki Morikawa (RIETI), and Discussion Paper seminar participants at RIETI.
+
+2 Graduate School of Fisheries and Environmental Sciences, Nagasaki University, 1-14 Bunkyo-machi, Nagasaki 852-8521, Japan
+
+3 Urban Institute & School of Engineering, Kyushu University, Japan; Research Institute of Economy, Trade and Industry (RIETI)
+
+1
+
+---
+
+## 1. Introduction
+
+By enhancing and creating efficiencies, artificial intelligence (AI) technology can play a critical role in improving social well-being in areas as diverse as economic development, precision medicine, public welfare, and environmental protection (National Science and Technology Council, 2016; Parkes and Wellman, 2015) . AI technology will significantly contribute to increases in human welfare across a wide range of sectors, including transportation, service robotics, healthcare, education, low-resource communities, public safety, employment, and entertainment (Stone et al., 2016) . According to (Teactica, 2015) , market opportunities for AI systems will increase from $ 202.5 million in 2015 to $ 11.1 billion by 2024.
+
+The global importance of AI technology has been growing. The 2015 Strategy for American Innovation established nine high-priority research areas, including the BRAIN initiative (Insel et al., 2013) and the Precision Medicine initiative (Hodson, 2016) . Additionally, AI is listed as a high-priority technology for a super-smart society in Japan's 5th Science and Technology Basic Plan (2016 to 2020) . These research and development (R&D) strategies focus on the expansion of the AI business market and are intended to improve international market competitiveness.
+
+Generally, patents grant inventors hold exclusive rights to protect their knowledge and technology from the competitor (Maresh et al., 2016) . This action is also observed in A.I.
+
+2
+
+---
+
+technology patenting. According to Bajpai (2016), IBM, world largest A.I. patent inventor, has projected that one billion consumers will be reached via Watson by end of 2017, primarily through its partnerships across companies. This report also pointed out sufficient A.I. patenting gives IBM a lead against companies which have just begun to work in the field of A.I. Thus, A.I. patenting is key corporate R&D strategy to have business collaboration with protecting intellectual property and to gain the position in rapid growing market.
+
+Against the backdrop of this acceleration of AI technology development across the globe, the number of patents granted has rapidly increased. Figure 1 shows the number of AI patents granted by application country and technology type and reveals that the number of AI patents has increased more than threefold, i.e., from 708 items in 2012 to 2,888 items in 2016. In particular, AI patents granted in the U.S. increased by 1,628 items during this period (Figure 1(a)), which represents approximately 75 % of the AI patent increase worldwide.
+
+<Figure 1 about here>
+
+As shown in Figure 1 (b), the patent share of each AI technology type changed from 2012 to 2016. In 2012, biological and knowledge-based models were the leaders in patented AI technologies. However, from 2012 to 2016, the number of patents granted for specific mathematical models and other AI technologies rapidly increased, doubling from 2015 to 2016.
+
+3
+
+---
+
+These two figures show the short-term trend of AI patenting based on country and technology. However, Figure 1 is not able to reveal why the number of patents granted for each AI technology type changed.
+
+To clarify why the trend in AI patenting changed, the R & D strategies of the inventors must be clarified. This strategy is the key driver of technology development (Fujii and Managi, 2016). Notably, not all AI technologies contribute equally to improved economic performance. Certain AI technologies directly contribute to profits by creating new products and services, whereas others contribute indirectly and only minimally. Therefore, the incentives for companies to invent AI technologies vary depending on the type of technology considered. A determinant analysis of inventions that focuses on the characteristics of each type of AI technology is important if we are to suggest effective policies to encourage R & D in such technology.
+
+This study is the first to use a decomposition framework to clarify the determinants of AI technology invention. Application of the decomposition method to technological invention was proposed by Fujii and Managi (2016) . This application can clarify the main factors that promote innovation, whether they are the corporate priority given to specific technology inventions or the result of the scale effect of R & D activities. This new decomposition application is useful to policymakers because it yields a better understanding of policies at both the macro and micro levels that can effectively encourage the invention of specific AI
+
+4
+
+---
+
+technology inventions (such as biological, knowledge-based, and specific mathematical models).
+
+The objective of this study is to clarify the determining factors that contribute to AI technology patent publications based on technology type. We also discuss how inventors' industrial characteristics and national R & D policy affect the invention of AI technology by country using company-level data. The novel contribution of this research is to clarify the primary driver of AI patent granting using a patent decomposition framework and the log mean Divisia index analysis. By combining these two approaches, we can distinguish the contribution of the R & D strategy factor and the R & D scale factor. Many previous studies have focused exclusively on the number of patent publications, which is affected by both the priority given to inventions and the scale of research activity (Cecere et al., 2014; Lybbert and Zolas, 2014; Park, 2014; Roper and Hewitt-Dundas, 2015) . This study attempts to derive a pure measure of the priority given to inventions from patent publication data by controlling for the scale effect (Fujii, 2016) .
+
+To consider the characteristics of each AI technology, we followed the practice of the USPTO and the Japan patent office (JPO) (Appendix 1) and divided the patent data into the following four AI technology groups: (1) biological models, (2) knowledge-based models, (3) specific mathematical models, and (4) other AI technology models.
+
+5
+
+---
+
+## 2. Methodology
+
+We apply a decomposition analysis framework to clarify the changing factors involved in granting AI technology patents. We use the following three indicators to decompose the AI technology patents granted: the priority of a specific AI technology (PRIORITY), the importance of AI technology among all patents granted (AItech), and the scale of R & D activity (SCALE).
+
+We define the PRIORITY indicator as the number of specific AI patents granted divided by the total number of AI patents granted, thus providing the share of specific AI patents granted among total AI patents. This indicator can be increased if the number of specific AI patents granted increases more quickly than the total number of AI patents granted, thus indicating that inventors are concentrating their research resources on specific types of AI technology inventions. Inventors are prioritizing specific AI technology types over other types when PRIORITY is increased.
+
+Similarly, the AItech indicator is defined as the total number of AI patents granted divided by the total number of patents granted, which indicates the share of total AI patents out of total patents. This indicator can be increased if the number of total AI patents granted increases more quickly than the number of total patents granted, thus indicating that inventors are concentrating their research resources on AI technology inventions. Inventors are prioritizing the invention of AI technology over other types of technology when AItech is
+
+6
+
+---
+
+increased.
+
+The SCALE indicator is defined as the total number of patents granted and thus represents the scale of R & D activities. Generally, active R & D efforts promote the invention of new technologies. Thus, the total number of patents granted reflects the level of active R & D efforts. Additionally, R & D activities in companies depend on corporate financial circumstances because the number patents granted is associated with the cost of researcher salaries, the operating cost of experimental materials, and the cost of applying for and registering patents. For example, following the financial crisis caused by the collapse of Lehman Brothers, the number of patents granted decreased (Fujii and Managi, 2016) . Thus, companies with serious financial difficulties decreased their R & D activities to reduce their bankruptcy risk. This decrease in R & D activities led to a decrease in the number of new patents granted, including those related to AI technologies. Therefore, the scale of R & D activity is an important factor in understanding why the number of AI patents granted has changed. SCALE increase as the total number of patents granted increases. The number of patents granted for AI technology would be increased by an increase in overall R & D activities if the SCALE score increased.
+
+Here, we introduce the decomposition approach using the biological model-based technology patent group as a specific type of AI patent granted (Table 1). The number of biological model-based technology patents granted (BIOLOGICAL) is decomposed using the total AI patents granted (AIpatent) and total patents granted (TOTAL), as in equation (1).
+
+7
+
+---
+
+## <Table 1 about here>
+
+$$BIOLOGICAL=\frac{ BIOLOGICAL }{ Alpatent } \times \frac{ Alpatent }{ TOTAL } \times  TOTAL = PRIORITY  \times  Altech  \times  SCALE \quad (1)$$
+
+We consider the change in biological model-based patents granted from year t-1 (BIOLOGICAL 1-1 ) to year t (BIOLOGICAL 1 ). Using equation (1), the growth ratio of biological model-based patents granted can be represented as follows:
+
+$$\frac{BIOLOGICAL^{t}}{BIOLOGICAL^{t-1}}=\frac{PRIORITY^{t}}{PRIORITY^{t-1}} \times \frac{Altech^{t}}{Altech^{t-1}} \times \frac{SCALE^{t}}{SCALE^{t-1}}\quad (2)$$
+
+We transform equation (2) into a natural logarithmic function and thus obtain equation (3). Notably, zero values in the dataset cause problems in the formulation of the decomposition due to the properties of logarithmic functions. To solve this problem, the logarithmic mean Divisia index (LMDI) literature suggests replacing zero values with a small positive number (Ang and Liu, 2007).
+
+$$lnBIOLOGICAL^{f}-lnBIOLOGICAL^{t-1}=\ln \left(\frac{PRIORITY^{t}}{PRIORITY^{t-1}}\right)+\ln \left(\frac{Altech^{t}}{Altech^{t-1}}\right)+\ln \left(\frac{SCALE^{t}}{SCALE^{t-1}}\right)(3)$$
+
+Multiplying both sides of equation (3) by $\omega_i^t = (\textsc{BIOLOGICAL}^t - \textsc{BIOLOGICAL}^{t-1})/$ (InBIOLOGICAL $^t - \ln$ BIOLOGICAL $^{t-1}$ ) yields equation (4) as follows.
+
+$$\begin{split}
+BIOLOGICAL^{t}-BIOLOGICAL^{t-1}=&\angleBIOLOGICAL^{t,t-1}\\
+&=\omega_{i}^{t}\ln\left(\frac{PRIORITY^{t}}{PRIORITY^{t-1}}\right)+\omega_{i}^{t}\ln\left(\frac{Altech^{t}}{Altech^{t-1}}\right)+\omega_{i}^{t}\ln\left(\frac{SCALE^{t}}{SCALE^{t-1}}\right)
+\end{split}\quad (4)$$
+
+Therefore, changes in the number of patents granted for biological model-based technologies ( $\angle$ BIOLOGICAL) are decomposed by changes in PRIORITY (first term), Altech
+
+8
+
+---
+
+(second term) and SCALE (third term). The term $\omega_i^t$ operates as an additive weight for the estimated number of patents granted for biological model-based technologies. This decomposition technique was developed by Ang et al. (1998) , and is termed the LMDI.
+
+The novel aspect of this research is that it clarifies the R & D strategies of companies using LMI analysis. Many previous studies have focused exclusively on the number of patents granted, which is affected by prioritizing certain invention types and the scale of research activity. This study attempts to derive the pure priority of inventions from patentgranted data by controlling for the scale effect. Fujii (2016) applies a decomposition framework to patent data analysis using the two factors of priority and scale. In this study, we propose an approach developed to distinguish the priority change of specific AI technology and that of total AI technology.
+
+## 3. Data and results
+
+### 3.1 Data
+
+We used patent-granted data from PATENTSCOPE, which is provided by the World Intellectual Property Organization (WIPO). The PATENTSCOPE database covers more than 56 million patents granted. We specified AI certain technology patents based on the patent clarification provided by both the USPTO and JPO (Appendix 1). We collected the patentgranted data on 7 February 2017 from the PATENTSCOPE database.
+
+9
+
+---
+
+As explained in Table 1 , this study focuses on four AI technology types: (1) biological model-based technology (BIOLOGICAL), (2) knowledge-based model technology (KNOWLEDGE), (3) specific mathematical model-based technology (MATHEMATICAL), and (4) other AI technology (OTHER). Following Fujii (2016) and Fujii and Managi (2016) we use only the primary IPC code and the primary applicant name to construct the patent dataset to avoid double-counting patent data.
+
+### 3.2 Comparative analysis of AI patents granted
+
+### 3.2.1 When was a specific AI technology patent invented and where?
+
+Table 2 represents the change of AI patents granted by type of technology at each patent office. Table 2 shows that the composition of patent-granted shares differs among countries. The knowledge-based model represents more than half of the total number of AI patents granted by the USPTO, whereas the biological model is the major technology type granted by the SIPO and JPO. Another finding is that the share of the specific mathematical model is only 1.7 % in the JPO, which is extremely low compared with that of other patent offices. This outcome occurs because Japanese AI researchers primarily focus on android technology-based R & D (and not mathematical elements), which represents the core AI technology (The Japan News, 2017). PCT, EPO, and the patent offices of other countries exhibit similar trends with respect to the technology share pattern of AI patent publications.
+
+10
+
+---
+
+<Table 2 about here>
+
+Next, we consider the numerical change of AI patents granted. As shown in Table 2, all the patent offices except the JPO published the largest number of AI patents from 2015 to 2016. Notably, the number of patents granted more than doubled at the USPTO, SIPO, and PCT. However, the average number of patents granted per year at the JPO was the largest from 2005 to 2009 for the biological model and from 2010 to 2014 for the knowledge-based model.
+
+One interpretation of this result is that the Japanese market is less attractive for AI technology application. Most AI technology services are strongly related to big data collection through the internet (such as social network systems, credit card payments, and sensors). Because of concerns among its residents, Japan is strict regarding the use of private information for business (Kawasaki, 2015) . The business barrier regarding big data collection and use minimizes the incentive to obtain AI patents in Japan. In the U.S., the government has established rules and regulations regarding the use of private information as big data (Hardy and Maurushat, 2017; Manyika et al., 2011) . Additionally, there are large governmental R & D expenditures for AI technology innovation in the U.S., which is another strong incentive for AI technology development (National Science and Technology Council, 2016) .
+
+### 3.2.2 Who invented which AI technology patent?
+
+11
+
+---
+
+Table 3 lists the top 30 applicants for AI patents granted worldwide. The bottom rows represent the number of patents granted to universities in the U.S., China, and Japan. As shown in Table 3, IBM is the world's leading recipient for AI patents granted. Additionally, of the top 30 grantees for AI patents granted, 18 applicants are U.S. companies, 8 applicants are Japanese companies, and 4 applicants are companies from other countries. Notably, Chinese companies and universities are not listed among the top 30 countries evaluated for the 2000-2016 period, which implies that AI patents granted in China are obtained by many applicants.
+
+<Table 3 about here>
+
+Next, we discuss the composition of the AI technology patent share for each applicant. Table 3 indicates that the patent portfolio of AI technology varies among applicants. Qualcomm and BRAIN corporation garnered the highest share for the biological model. However, SAP and Cognitive Scale had the largest share for the knowledge-based model. D-wave obtained 92 % of other AI patents, an outcome that represents a completely different trend from other companies. Notably, the companies listed in the top half of the list obtained patents in a wide range of AI technology areas.
+
+According to Table 3 , a large proportion of the AI patents granted to Chinese and Japanese universities were for technology based on the biological model. This trend differs
+
+12
+
+---
+
+from that found for U.S. universities. In addition, U.S. universities have obtained a large proportion of patents for AI technology that uses a knowledge-based model. This trend resembles that found for the composition of patents granted by the USPTO (Table 2). One interpretation of these results is that U.S. universities have an advantage with respect to accessing and analyzing big data. Generally, U.S. universities have more opportunity to collaborate with U.S. companies, which control big data for AI technology development. Access to big data is a key factor in developing a knowledge-based analysis (Gu et al., 2017) . Additionally, U.S. universities more successfully train students with substantial analytical talent in graduate school (Manyika et al., 2011) than the universities of other countries. Human resources for big data analysis are another important driver of technology development on the knowledge-based model in the U.S.
+
+### 3.2.3 Who invented a patented AI technology and where?
+
+In this section, we discuss the distribution of AI patent applications by applicant. As shown in Table 4 , most U.S. companies have a large share of AI patent invention according to the USPTO data. By contrast, the share of patented inventions of U.S. companies from the JPO and SIPO is small. With the exception of NTT, non-U.S. companies have more than a 16 % patent share at the USPTO. Specifically, Samsung has 61 % of all AI patents issued by the USPTO. Surprisingly, four of the eight Japanese companies were granted more AI patents by the USPTO than by the JPO. These results imply that Japanese companies have strong
+
+13
+
+---
+
+incentives to obtain AI patents from the USPTO, while there is less incentive for U.S. companies to obtain AI patents from the JPO. This result is consistent with the interpretation that big data use creates an advantage for the U.S. market with respect to AI technology application.
+
+<Table 4 about here>
+
+Based on Table 4 , universities clearly tend to apply for AI patents at domestic patent offices. In particular, 98 % of the AI patents obtained by Chinese universities were granted by the SIPO, with a low number granted by other patent offices. By contrast, U.S. and Japanese universities apply for AI patents at the PCT in addition to at their domestic patent offices. Notably, approximately 45 % of the AI technology patents granted in China were obtained by Chinese universities. This outcome is unique. In other countries, private companies are the primary patent applicants. This trend is also observed in other technological fields (e.g., nanotechnology (Huang and Wu, 2012) and aquaculture technology) (Fujii et al., 2017) . Fong et al. (2015) note that “China’s National Medium and Long Term Science and Technology Development Planning (2006–2020)” significantly improved Chinese university technology transfer. Additionally, the same scholars conclude that economic incentives and royalties are key factors incentivizing the increase in the number of patented inventions at Chinese
+
+14
+
+---
+
+universities.
+
+Moreover, the Chinese government sets a high priority on AI technology development in China's 13 th Five-Year Plan (2016-2020). Under this plan, research institutes and universities are encouraged to invent new AI technologies. These governmental targets can be considered the key factor driving the increase in the priority placed on AI technology in China.
+
+### 3.3 Patent decomposition analysis
+
+Figure 2 shows the results of a decomposition analysis for four specific AI technology patents granted at all the patent offices listed in Appendix 2. Because the AI patent trend changes beginning in 2012 (Figure 1), we divided the decomposition analysis results into two periods (the first period runs from 2000 to 2011, and the second period from 2012 to 2016). The plotted point in red indicates the change in the number of specific patents granted, and the bar chart shows the effects of each decomposed factor on the number of patents granted related to specific AI technologies. The sum of the bars is equivalent to the value of the plotted point. The figure shows the differences in the driving factors for patents granted based on the type of AI technology. Detailed results from the decomposition analysis are provided in Appendices 3-6.
+
+<Figure 2 about here>
+
+15
+
+---
+
+Figure 2 shows that the number of patents granted for technology based on the biological and knowledge-based models increased during the first period. However, the priority of specific AI technology affects these two technology types differently. As shown in Figure 2 , during the first period, the relative priority of the biological model was negative, whereas that of the knowledge-based model was positive. This result implies that the priority of AI technology patent invention shifted from the biological model to the knowledge-based model over the first period. The number of patents granted for the other two technology types did not change significantly during the first period, which indicates that these two technologies were treated as less important than technologies based on the biological and knowledge-based models during that period.
+
+Based on the results for the second period, the number of patents granted substantially increased for all four AI technologies. In addition, the priority of specific technologies shifted from the biological and knowledge-based models to the specific mathematical model and other AI model during the second period. Specifically, the number of patents granted for other AI technology was 624 items during the second period, which is more than that for the biological model (565 items) and close to that for the knowledge-based model (693 items) (see red points in Figure 2 ).
+
+There are two main reasons why the number of patents granted for other AI technology
+
+16
+
+---
+
+increased during the second period. First, technology based on the biological and knowledgebased models that use big data have garnered attention in the business market in recent years as the technology demand for high information-processing capabilities became stronger. This technology demand exerted a strong incentive to invent the quantum computer, which is categorized as other AI technology (Lloyd et al., 2016) .
+
+Second, the range of AI technology became broader and more complex during the second period, which makes it difficult to categorize AI technology patents into the major technology groups, such as the biological model or the knowledge-based model. Here, we examine the breakdown for AI technology patents whose primary IPC code is G06N99 to investigate the primary driver of other AI technology patent growth. There are 611 items whose IPC is solely G06N99 that are not registered using a secondary IPC code. The number of patents whose second IPC is G06N5/02, G06N5/04, or G06N7/00 total 83, 129, and 106 items, respectively. These IPC codes are included under the knowledge-based model and specific mathematical models (Appendix 1). Thus, other AI technology patents includes patent items that are strongly related to knowledge-based or specific mathematical models.
+
+Patent items categorized as other AI technology and registered with a second IPC code of G06F17/30 (information retrieval; therefore, database structures) include 86 items from 2000 to 2015 and 77 items published after 2015. This technology group contributes to creating data labeling and tagging to achieve more efficient machine learning. For example, IBM listed
+
+17
+
+---
+
+“Labeling of data for machine learning (US20150356457)”, and Microsoft listed “Metadata tag description generation” in publication number US20160358096. These AI patents involve crucial technology that is used to rapidly analyze big data collected by social network services and sensors for the Internet of Things. However, there is no appropriate IPC clarification for these technologies, and such patent items are registered under other AI technology.
+
+Table 5 shows the patent decomposition analysis results by patent office. The table shows that the main contributors of patents granted are mixed in the first period among the patent offices. However, AI technology commonly contributed as the main driver of patent invention for the biological and knowledge-based models during the second period at all the patent offices. Additionally, the priority of a specific technology increased other AI technology at all the patent offices during the second period. Based on these two findings, the R & D priority of AI technology became stronger during the second period — particularly for other AI technology — at the five patent offices. In addition, the priority of specific mathematical models increased at the USPTO and JPO but decreased at the EPO. Thus, the priority of specific technology commonly contributed to other AI technology invention at the five patent offices and the specific mathematical model at the USPTO and JPO.
+
+<Table 5 about here>
+
+18
+
+---
+
+The key point of the results for the first period is that the scale change of R & D activity contributes at the SIPO for all four AI technology types. One interpretation of this trend is that the Chinese patent application law revisions in 2001 and 2009 simplified patent applications for domestic companies that used the subsidy program (Dang and Motohashi, 2015). Hu et al. (2017) noted that a rapid patent application increase at the SIPO was caused by external factors, such as the revision of the patent law and a new subsidy system and not by internal factors (e.g., R & D priority changes and human resources for R & D). Thus, the revision of the Chinese patent application system contributed to expanding R & D activities (e.g., patent applications) at the SIPO, which increased the number of patents for AI inventions.
+
+At the USPTO, the priority of AI technology increased the number of patents granted for the four AI technology types during both the first and second periods. This result was observed only at the USPTO and indicates that AI patent invention behavior in the U.S. is unique and successfully incentivized by U.S. governmental policies (National Science and Technology Council, 2016; Taylor, 2016). Additionally, the contribution of the priority of specific technology shifted from the knowledge-based model to the specific mathematical model and other AI technology at the USPTO and JPO during the second period. Nonetheless, the priority of specific technology negatively affected the specific mathematical models at the SIPO, PCT, and EPO.
+
+19
+
+---
+
+## 4. Summary and conclusions
+
+This study examined the trend and priority change of AI technology using patentgranted data from 2000 to 2016. We focused on the following four technology types: (1) biological model, (2) knowledge-based model, (3) specific mathematical model, and (4) other AI technology. Employing a patent decomposition analysis framework, we clarified the trends and priority changes for patent inventions for these four technology types. The main results are summarized as follows.
+
+First, AI technology patents were primarily granted at the USPTO to private U.S. companies, and in particular to IBM, Microsoft, and Qualcomm. Additionally, many U.S. companies primarily applied for patents at the USPTO and have little share at other patent offices. Non-U.S. companies also focused on obtaining patents from the USPTO in addition to their domestic patent offices. These results show that the U.S. market is attractive for both U.S. and non-U.S. companies. However, other countries' markets are not particularly attractive for U.S. companies.
+
+Second, universities are key AI technology inventors in the U.S. and China. Specifically, 45 % of AI technology patents granted at the SIPO were obtained by Chinese universities. Other important findings include that 98 % of the AI technology patents granted to Chinese universities were registered at SIPO, whereas U.S. and Japanese universities received 72 % and
+
+20
+
+---
+
+78% of their relevant patents from domestic patent offices, respectively.
+
+Finally, we find that the relative priority of R & D shifted from the biological and knowledge-based models to specific mathematical models and other AI technology, particularly in the U.S. and Japan. Additionally, R & D priority characteristics vary among the patent offices and AI technology types. These results imply that the international framework for AI technology development should be considered for effective R & D policy construction.
+
+21
+
+---
+
+## References
+
+Ang, B.W., Liu, N., 2007. Handling zero values in the logarithmic mean Divisia index decomposition approach. Energy Policy 35, 238-46.
+
+Ang, B.W., Zhang, F.Q., Choi, K.H., 1998. Factorizing changes in energy and environmental indicators through decomposition. Energy 23, 489-95.
+
+Bajpai, P. (2016) IBM's Astonishing Patent Strategy Shows Where It's Going Next. News Headlines, Technology, Stocks in Nasdaq. http://www.nasdaq.com/article/ibmsastonishing-patent-strategy-shows-where-its-going-next-cm725790 (accessed 28th April, 2017)
+
+Cecere, G., Corrocher, N., Gossart, C., Ozman, M., 2014. Technological pervasiveness and variety of innovators in Green ICT: A patent-based analysis. Research Policy 43, 1827-39.
+
+Dang, J., Motohashi, K., 2015. Patent statistics: A good indicator for innovation in China? Patent subsidy program impacts on patent quality. China Economic Review 35, 13755.
+
+Fong, P.S.W., Chang, X., Chen, Q., 2015. Faculty patent assignment in the Chinese mainland: Evidence from the top 35 patent application universities. The Journal of Technology Transfer, 1-27. doi:10.1007/s10961-015-9434-7
+
+Fujii, H., 2016. Decomposition analysis of green chemical technology inventions from 1971
+
+22
+
+---
+
+to 2010 in Japan. Journal of Cleaner Production 112, 4835-43.
+
+Fujii, H., Managi, S., 2016. Research and development strategy for environmental technology in Japan: A comparative study of the private and public sectors. Technological Forecasting and Social Change 112, 293-302.
+
+Fujii, H., Sakakura, Y., Soyano, K., Matsushita, Y., Hagiwara, A., 2017. Research and development strategy of fisheries technology innovation. Nagasaki university.
+
+Gu, D., Li, J., Li, X., Liang, C., 2017. Visualizing the knowledge structure and evolution of big data research in healthcare informatics. International Journal of Medical Informatics 98, 22-32.
+
+Hardy, K., Maurushat, A., 2017. Opening up government data for Big Data analysis and public benefit. Computer Law and Security Review 33, 30-7.
+
+Hodson, R., 2016. Precision medicine. Nature 537, S49.
+
+Hu, A.G.Z., Zhang, P., Zhao, L., 2017. China as number one? Evidence from China's most recent patenting surge. Journal of Development Economics 124, 107-19.
+
+Huang, C., Wu, Y., 2012. State-led technological development: A case of China's nanotechnology development. World Development 40, 970-82.
+
+Insel, T.R., Landis, S.C., Collins, F.S., 2013. The NIH brain initiative. Science 340, 687-8.
+
+Kawasaki, S., 2015. The challenges of transportation/traffic statistics in Japan and directions for the future. IATSS Research 39, 1-8.
+
+23
+
+---
+
+Lloyd, S., Garnerone, S., Zanardi, P., 2016. Quantum algorithms for topological and geometric analysis of data. Nature Communications 7, 10138.
+
+Lybbert, T.J., Zolas, N.J., 2014. Getting patents and economic data to speak to each other: An ‘Algorithmic Links with Probabilities’ approach for joint analyses of patenting and economic activity. Research Policy 43, 530-42.
+
+Manyika, J., Chui, M., Brown, B., Bughin, J., Dobbs, R., Roxburgh, C., Byers, A.H., McKinsey Global Institute, 2011. Big Data: The Next Frontier for Innovation, Competition, and Productivity. McKinsey, Incorporated.
+
+Maresch, D., Fink, M., Harms, R. When patents matter: The impact of competition and patent age on the performance contribution of intellectual property rights protection. Technovation 57-58, 14-20.
+
+National Science and Technology Council, 2016. The National Artificial Intelligence Research and Development Strategic Plan. CreateSpace Independent Publishing Platform, Washington, D.C.
+
+Park, J.Y., 2014. The evolution of waste into a resource: Examining innovation in technologies reusing coal combustion by-products using patent data. Research Policy 43, 1816-26.
+
+Parkes, D.C., Wellman, M.P., 2015. Economic reasoning and artificial intelligence. Science 349, 267.
+
+24
+
+---
+
+Roper, S., Hewitt-Dundas, N., 2015. Knowledge stocks, knowledge flows and innovation: Evidence from matched patents and innovation panel data. Research Policy 44, 132740.
+
+Stone, P., Brooks, R., Brynjolfsson, E., Calo, R., Etzioni, O., Hager, G., Hirschberg, J., Kalyanakrishnan, S., Kamar, E., Kraus, S., Leyton-Brown, K., Parkes, D., Press, W., Saxenian, A., Shah, J., Tambe, M., Teller, A., 2016. Artificial intelligence and life in 2030. one hundred year study on artificial intelligence: Report of the 2015-2016 Study Panel. Stanford University, Stanford, CA.
+
+Taylor, R.D., 2016. The next stage of U.S. communications policy: The emerging embedded infosphere. Telecommunications Policy. doi: 10.1016/j.telpol.2016.11.007.
+
+Teactica, 2015. Artificial Intelligence for Enterprise Applications to Reach $11.1 Billion in Market Value by 2024.
+
+The Japan News, 2017. Japan's Critically Late Start in AI Research (from http://www.thejapan-news.com/news/article/0003514495).
+
+25
+
+---
+
+Table 1. Description of AI technology patent group
+
+<table><tr><td>Patent group</td><td>Description of patent group [IPC code]</td></tr><tr><td>Biological model</td><td>Computer systems based on biological models, including neural network models, genetic models, architectures, physical realization, learning methods, biomolecular computers, and artificial life [IPC=G06N3/00].</td></tr><tr><td>Knowledge-based model</td><td>Computer systems that utilize knowledge-based models, including knowledge engineering, knowledge acquisition, extracting rules from data, and inference methods or devices [IPC=G06N5/00].</td></tr><tr><td>Specific mathematical model</td><td>Computer systems based on specific mathematical models, including fuzzy logic, physical realization, chaos models or non-linear system models, and probabilistic networks [G06/N7/00].</td></tr><tr><td>Other AI technology</td><td>Subject matter not provided for previously described groups of the G06N3 subclass, including quantum computers, learning machines, and molecular computers [G06/N99/00].</td></tr></table>
+
+
+Source: USPTO Class 706 Data processing: Artificial intelligence. Report on FY2014 Trend survey of patent application technology: Artificial intelligence (2016) https://www.jpo.go.jp/shiryou/pdf/gidou-houkoku/26_21.pdf.
+
+26
+
+---
+
+Table 2. Data description of AI technology patents granted (item)
+
+<table><tr><td rowspan="2">Patent office</td><td rowspan="2">AI technology type</td><td rowspan="2">2000-2016</td><td rowspan="2">Share</td><td colspan="4">Yearly average number of patents granted</td></tr><tr><td>2000-2004</td><td>2005-2009</td><td>2010-2014</td><td>2015-2016</td></tr><tr><td rowspan="4">USPTO</td><td>Biological</td><td>1,455</td><td>19.9%</td><td>44</td><td>63</td><td>80</td><td>259</td></tr><tr><td>Knowledge</td><td>4,152</td><td>56.9%</td><td>50</td><td>166</td><td>320</td><td>738</td></tr><tr><td>Mathematical</td><td>672</td><td>9.2%</td><td>9</td><td>18</td><td>30</td><td>194</td></tr><tr><td>Other</td><td>1,019</td><td>14.0%</td><td>1</td><td>3</td><td>57</td><td>359</td></tr><tr><td rowspan="4">SIPO</td><td>Biological</td><td>1,184</td><td>73.7%</td><td>9</td><td>32</td><td>103</td><td>232</td></tr><tr><td>Knowledge</td><td>219</td><td>13.6%</td><td>1</td><td>9</td><td>16</td><td>45</td></tr><tr><td>Mathematical</td><td>114</td><td>7.1%</td><td>4</td><td>6</td><td>8</td><td>14</td></tr><tr><td>Other</td><td>90</td><td>5.6%</td><td>1</td><td>7</td><td>4</td><td>15</td></tr><tr><td rowspan="4">JPO</td><td>Biological</td><td>679</td><td>56.4%</td><td>12</td><td>62</td><td>46</td><td>40</td></tr><tr><td>Knowledge</td><td>410</td><td>34.1%</td><td>4</td><td>31</td><td>37</td><td>25</td></tr><tr><td>Mathematical</td><td>21</td><td>1.7%</td><td>1</td><td>1</td><td>1</td><td>5</td></tr><tr><td>Other</td><td>94</td><td>7.8%</td><td>1</td><td>4</td><td>3</td><td>26</td></tr><tr><td rowspan="4">PCT</td><td>Biological</td><td>723</td><td>46.3%</td><td>35</td><td>30</td><td>38</td><td>104</td></tr><tr><td>Knowledge</td><td>480</td><td>30.7%</td><td>18</td><td>26</td><td>29</td><td>58</td></tr><tr><td>Mathematical</td><td>114</td><td>7.3%</td><td>2</td><td>5</td><td>9</td><td>16</td></tr><tr><td>Other</td><td>244</td><td>15.6%</td><td>7</td><td>8</td><td>12</td><td>56</td></tr><tr><td rowspan="4">EPO</td><td>Biological</td><td>452</td><td>44.0%</td><td>26</td><td>24</td><td>24</td><td>42</td></tr><tr><td>Knowledge</td><td>306</td><td>29.8%</td><td>8</td><td>22</td><td>20</td><td>26</td></tr><tr><td>Mathematical</td><td>106</td><td>10.3%</td><td>2</td><td>5</td><td>10</td><td>12</td></tr><tr><td>Other</td><td>164</td><td>16.0%</td><td>7</td><td>8</td><td>10</td><td>22</td></tr><tr><td rowspan="4">Other patent</td><td>Biological</td><td>434</td><td>49.9%</td><td>25</td><td>13</td><td>31</td><td>45</td></tr><tr><td>Knowledge</td><td>220</td><td>25.3%</td><td>10</td><td>8</td><td>17</td><td>23</td></tr><tr><td>office</td><td>Mathematical</td><td>135</td><td>15.5%</td><td>4</td><td>3</td><td>13</td><td>16</td></tr><tr><td></td><td>Other</td><td>80</td><td>9.2%</td><td>3</td><td>4</td><td>5</td><td>9</td></tr></table>
+
+
+Source: Author estimate using IPC code in Appendix 1 and PATSTAT database. Note: USPTO: United States Patent and Trademark Office; SIPO: State Intellectual Property Office of The People's Republic of China; JPO: Japan Patent Office; PCT: Patent Cooperation Treaty; EPO is European Patent Office.
+
+27
+
+---
+
+Table 3. Number of AI patents granted and technology portfolios: 2000 to 2016
+
+<table><tr><td rowspan="2">Rank</td><td rowspan="2">Applicant name</td><td rowspan="2">Country</td><td rowspan="2">Total patents</td><td colspan="4">Patent portfolio of AI technology</td></tr><tr><td>Biological</td><td>Knowledge</td><td>Mathematical</td><td>Other</td></tr><tr><td>1</td><td>IBM</td><td>USA</td><td>1,057</td><td>22%</td><td>56%</td><td>8%</td><td>14%</td></tr><tr><td>2</td><td>Microsoft</td><td>USA</td><td>466</td><td>22%</td><td>44%</td><td>9%</td><td>24%</td></tr><tr><td>3</td><td>Qualcomm</td><td>USA</td><td>450</td><td>83%</td><td>7%</td><td>3%</td><td>7%</td></tr><tr><td>4</td><td>NEC</td><td>Japan</td><td>255</td><td>23%</td><td>49%</td><td>8%</td><td>20%</td></tr><tr><td>5</td><td>Sony</td><td>Japan</td><td>212</td><td>51%</td><td>33%</td><td>6%</td><td>10%</td></tr><tr><td>6</td><td>Google</td><td>USA</td><td>195</td><td>41%</td><td>36%</td><td>7%</td><td>17%</td></tr><tr><td>7</td><td>Siemens</td><td>Germany</td><td>192</td><td>54%</td><td>31%</td><td>10%</td><td>5%</td></tr><tr><td>8</td><td>Fujitsu</td><td>Japan</td><td>154</td><td>27%</td><td>60%</td><td>9%</td><td>4%</td></tr><tr><td>9</td><td>Samsung</td><td>Korea</td><td>119</td><td>56%</td><td>28%</td><td>3%</td><td>13%</td></tr><tr><td>10</td><td>NTT</td><td>Japan</td><td>94</td><td>35%</td><td>49%</td><td>0%</td><td>16%</td></tr><tr><td>11</td><td>Hewlett-Packard</td><td>USA</td><td>93</td><td>22%</td><td>44%</td><td>4%</td><td>30%</td></tr><tr><td>12</td><td>Yahoo</td><td>USA</td><td>88</td><td>14%</td><td>57%</td><td>16%</td><td>14%</td></tr><tr><td>13</td><td>Toshiba</td><td>Japan</td><td>86</td><td>22%</td><td>57%</td><td>7%</td><td>14%</td></tr><tr><td>14</td><td>D-wave</td><td>Canada</td><td>77</td><td>1%</td><td>4%</td><td>3%</td><td>92%</td></tr><tr><td>15</td><td>Hitachi</td><td>Japan</td><td>69</td><td>20%</td><td>38%</td><td>12%</td><td>30%</td></tr><tr><td>15</td><td>SAP</td><td>USA</td><td>69</td><td>23%</td><td>70%</td><td>6%</td><td>1%</td></tr><tr><td>17</td><td>Canon</td><td>Japan</td><td>68</td><td>59%</td><td>28%</td><td>3%</td><td>10%</td></tr><tr><td>18</td><td>Xerox</td><td>USA</td><td>62</td><td>15%</td><td>45%</td><td>18%</td><td>23%</td></tr><tr><td>19</td><td>GE</td><td>USA</td><td>59</td><td>14%</td><td>59%</td><td>22%</td><td>5%</td></tr><tr><td>20</td><td>Mitsubishi Electric</td><td>Japan</td><td>53</td><td>49%</td><td>43%</td><td>2%</td><td>6%</td></tr><tr><td>21</td><td>Honeywell</td><td>USA</td><td>49</td><td>24%</td><td>51%</td><td>22%</td><td>2%</td></tr><tr><td>22</td><td>Boeing</td><td>USA</td><td>48</td><td>31%</td><td>60%</td><td>4%</td><td>4%</td></tr><tr><td>23</td><td>Cisco</td><td>USA</td><td>47</td><td>15%</td><td>38%</td><td>0%</td><td>47%</td></tr><tr><td>23</td><td>Oracle</td><td>USA</td><td>47</td><td>17%</td><td>55%</td><td>9%</td><td>19%</td></tr><tr><td>25</td><td>British Telecomm</td><td>UK</td><td>44</td><td>41%</td><td>57%</td><td>2%</td><td>0%</td></tr><tr><td>26</td><td>Intel</td><td>USA</td><td>43</td><td>35%</td><td>51%</td><td>5%</td><td>9%</td></tr><tr><td>27</td><td>Amazon</td><td>USA</td><td>41</td><td>15%</td><td>39%</td><td>2%</td><td>44%</td></tr><tr><td>28</td><td>Brain Corporation</td><td>USA</td><td>40</td><td>80%</td><td>15%</td><td>3%</td><td>3%</td></tr><tr><td>28</td><td>Cognitive scale</td><td>USA</td><td>40</td><td>0%</td><td>88%</td><td>0%</td><td>13%</td></tr><tr><td rowspan="7">28</td><td>Facebook</td><td>USA</td><td>40</td><td>0%</td><td>40%</td><td>13%</td><td>48%</td></tr><tr><td>University total</td><td>World</td><td>1,177</td><td>69%</td><td>19%</td><td>6%</td><td>6%</td></tr><tr><td>U.S. university</td><td>USA</td><td>241</td><td>41%</td><td>38%</td><td>7%</td><td>14%</td></tr><tr><td>Chinese university</td><td>China</td><td>725</td><td>82%</td><td>10%</td><td>5%</td><td>3%</td></tr><tr><td>Japanese university</td><td>Japan</td><td>93</td><td>83%</td><td>15%</td><td>1%</td><td>1%</td></tr></table>
+
+
+Source: Author estimate using IPC code in Appendix 1 and PATSTAT database.
+
+28
+
+---
+
+Table 4. Distribution of country or organization of AI patents granted from 2000 to 2016
+
+<table><tr><td>Rank</td><td>Applicant name</td><td>Country</td><td>USPTO</td><td>SIPO</td><td>JPO</td><td>PCT</td><td>EPO</td><td>Other</td></tr><tr><td>1</td><td>IBM</td><td>USA</td><td>90\2</td><td>Microsoft</td><td>USA</td><td>74\3</td><td>Qualcomm</td><td>USA</td><td>32\4</td><td>NEC</td><td>Japan</td><td>36\5</td><td>Sony</td><td>Japan</td><td>32\6</td><td>Google</td><td>USA</td><td>75\7</td><td>Siemens</td><td>Germany</td><td>28\8</td><td>Fujitsu</td><td>Japan</td><td>42\9</td><td>Samsung</td><td>Korea</td><td>61\10</td><td>NTT</td><td>Japan</td><td>0\11</td><td>Hewlett-Packard</td><td>USA</td><td>71\12</td><td>Yahoo</td><td>USA</td><td>86\13</td><td>Toshiba</td><td>Japan</td><td>45\14</td><td>D-wave</td><td>Canada</td><td>39\15</td><td>Hitachi</td><td>Japan</td><td>30\16</td><td>SAP</td><td>USA</td><td>74\17</td><td>Canon</td><td>Japan</td><td>47\18</td><td>Xerox</td><td>USA</td><td>89\19</td><td>GE</td><td>USA</td><td>85\20</td><td>Mitsubishi Electric</td><td>Japan</td><td>25\21</td><td>Honeywell</td><td>USA</td><td>39\22</td><td>Boeing</td><td>USA</td><td>60\23</td><td>Cisco</td><td>USA</td><td>85\24</td><td>Oracle</td><td>USA</td><td>100\25</td><td>British Telecomm</td><td>UK</td><td>16\26</td><td>Intel</td><td>USA</td><td>67\27</td><td>Amazon</td><td>USA</td><td>80\28</td><td>Brain Corporation</td><td>USA</td><td>80\28</td><td>Cognitive scale</td><td>USA</td><td>100\28</td><td>Facebook</td><td>USA</td><td>100\</td><td>University total</td><td>World</td><td>20\U.S. university</td><td>USA</td><td>72\Chinese university</td><td>China Japan</td><td>1\</td><td>Japanese university</td><td>Japan</td><td>1\</td></tr></table>
+
+
+Source: Author's estimation using the IPC code in Appendix 1 and PATSTAT database. Note: USPTO: United States Patent and Trademark Office; SIPO: State Intellectual Property Office of The People's Republic of China; JPO: Japan Patent Office; PCT: Patent Cooperation Treaty; EPO: European Patent Office.
+
+29
+
+---
+
+Table 5. Results of decomposition analysis by patent office: 2000 to 2016
+
+<table><tr><td rowspan="2">Specific technology</td><td rowspan="2">Patent office</td><td colspan="4">Change from 2000 to 2012</td><td colspan="4">Change from 2012 to 2016</td></tr><tr><td> $\triangle$  Specific technology patent</td><td>Priority (specific)</td><td>Priority (AI)</td><td>Scale</td><td> $\angle$  Specific technology patent</td><td>Priority (specific)</td><td>Priority (AI)</td><td>Scale</td></tr><tr><td rowspan="5">Biological model</td><td>USPTO</td><td>39</td><td>-73.2</td><td>85.2</td><td>26.9</td><td>260</td><td>-10.6</td><td>286.0</td><td>-15.3</td></tr><tr><td>SIPO</td><td>89</td><td>22.9</td><td>-9.1</td><td>75.1</td><td>183</td><td>-14.3</td><td>109.6</td><td>87.7</td></tr><tr><td>JPO</td><td>39</td><td>-17.9</td><td>73.5</td><td>-16.6</td><td>-4</td><td>-12.1</td><td>28.3</td><td>-20.3</td></tr><tr><td>PCT</td><td>-3</td><td>-16.4</td><td>-8.5</td><td>21.9</td><td>74</td><td>-10.9</td><td>75.8</td><td>9.2</td></tr><tr><td>EPO</td><td>19</td><td>-17.3</td><td>26.0</td><td>10.4</td><td>29</td><td>5.1</td><td>29.6</td><td>-5.7</td></tr><tr><td rowspan="5">Knowledge-based model</td><td>USPTO</td><td>239</td><td>85.2</td><td>123.8</td><td>30.0</td><td>612</td><td>-292.6</td><td>938.3</td><td>-33.8</td></tr><tr><td>SIPO</td><td>13</td><td>0.0</td><td>2.6</td><td>10.3</td><td>36</td><td>1.7</td><td>19.0</td><td>15.3</td></tr><tr><td>JPO</td><td>40</td><td>38.2</td><td>12.3</td><td>-10.4</td><td>-19</td><td>-25.1</td><td>18.2</td><td>-12.1</td></tr><tr><td>PCT</td><td>11</td><td>5.7</td><td>-10.0</td><td>15.3</td><td>48</td><td>-7.6</td><td>48.9</td><td>6.7</td></tr><tr><td>EPO</td><td>11</td><td>7.4</td><td>-1.6</td><td>5.2</td><td>12</td><td>-4.5</td><td>18.8</td><td>-2.3</td></tr><tr><td rowspan="5">Specific</td><td>USPTO</td><td>15</td><td>-6.9</td><td>14.3</td><td>7.5</td><td>265</td><td>122.6</td><td>156.8</td><td>-14.4</td></tr><tr><td>SIPO</td><td>4</td><td>-10.6</td><td>3.6</td><td>11.0</td><td>10</td><td>-0.5</td><td>5.8</td><td>4.7</td></tr><tr><td>mathematical</td><td>JPO</td><td>0</td><td>-0.1</td><td>0.3</td><td>-0.2</td><td>8</td><td>7.6</td><td>1.4</td><td>-1.0</td></tr><tr><td>model</td><td>PCT</td><td>2</td><td>2.0</td><td>-1.3</td><td>1.2</td><td>15</td><td>-1.0</td><td>14.3</td><td>1.7</td></tr><tr><td></td><td>EPO</td><td>12</td><td>12.2</td><td>-1.4</td><td>1.3</td><td>-1</td><td>-8.6</td><td>9.3</td><td>-1.7</td></tr><tr><td rowspan="5">Other AI technology</td><td>USPTO</td><td>10</td><td>4.6</td><td>4.4</td><td>1.0</td><td>491</td><td>191.5</td><td>326.9</td><td>-27.4</td></tr><tr><td>SIPO</td><td>3</td><td>-5.5</td><td>-0.7</td><td>9.2</td><td>21</td><td>14.7</td><td>3.5</td><td>2.8</td></tr><tr><td>JPO</td><td>2</td><td>-3.1</td><td>6.4</td><td>-1.3</td><td>34</td><td>30.7</td><td>11.4</td><td>-8.2</td></tr><tr><td>PCT</td><td>9</td><td>5.7</td><td>-1.9</td><td>5.2</td><td>54</td><td>21.1</td><td>29.6</td><td>3.3</td></tr><tr><td>EPO</td><td>8</td><td>4.7</td><td>0.3</td><td>3.0</td><td>21</td><td>9.3</td><td>14.5</td><td>-2.8</td></tr></table>
+
+
+Note: USPTO: United States Patent and Trademark Office; SIPO: State Intellectual Property Office of The People's Republic of China; JPO: Japan Patent Office; PCT: Patent Cooperation Treaty; EPO: European Patent Office.
+
+30
+
+---
+
+![Figure](figures/figure_000.png)
+
+Figure 1a. Number of AI patents granted by country
+
+![Figure](figures/figure_002.png)
+
+Figure 1b. Number of AI patents granted by technology
+
+Figure. 1. Trend of AI patents granted: 2000 to 2016 (number of items)
+
+Source: Author estimate using IPC code in Appendix 1 and PATSTAT database.
+
+Note: USPTO: United States Patent and Trademark Office; SIPO: State Intellectual Property Office of The People's Republic of China; JPO: Japan Patent Office; PCT: Patent Cooperation Treaty; EPO: European Patent Office
+
+31
+
+---
+
+![Figure](figures/figure_000.png)
+
+Figure. 2. Results of patent decomposition analysis (number of items)
+
+Note: The vertical axis is standardized by setting the number of changes in patents granted in 2000 and 2012 to zero.
+
+## Supplementary Information
+
+Appendix 1. International patent clarification related to AI technologies
+
+32
+
+---
+
+<table><tr><td>IPC</td><td>Technology group</td><td>Description</td></tr><tr><td>G06N 3/00</td><td>Biological model</td><td>Computer systems based on biological models</td></tr><tr><td>G06N 3/02</td><td>Biological model</td><td>Using neural network models</td></tr><tr><td>G06N 3/04</td><td>Biological model</td><td>Architectures</td></tr><tr><td>G06N 3/06</td><td>Biological model</td><td>Physical realization</td></tr><tr><td>G06N 3/063</td><td>Biological model</td><td>Using electronic means</td></tr><tr><td>G06N 3/067</td><td>Biological model</td><td>Using optical means</td></tr><tr><td>G06N 3/08</td><td>Biological model</td><td>Learning methods</td></tr><tr><td>G06N 3/10</td><td>Biological model</td><td>Simulation on general-purpose computers</td></tr><tr><td>G06N 3/12</td><td>Biological model</td><td>Using genetic models</td></tr><tr><td>G06N 5/00</td><td>Knowledge-based model</td><td>Computer systems utilizing knowledge-based models</td></tr><tr><td>G06N 5/02</td><td>Knowledge-based model</td><td>Knowledge representation</td></tr><tr><td>G06N 5/04</td><td>Knowledge-based model</td><td>Inference methods or devices</td></tr><tr><td>G06N 7/00</td><td>Specific mathematical model</td><td>Computer systems based on specific mathematical models</td></tr><tr><td>G06N 7/02</td><td>Specific mathematical model</td><td>Using fuzzy logic</td></tr><tr><td>G06N 7/04</td><td>Specific mathematical model</td><td>Physical realization</td></tr><tr><td>G06N 7/06</td><td>Specific mathematical model</td><td>Simulation on general-purpose computers</td></tr><tr><td>G06N 7/08</td><td>Specific mathematical model</td><td>Using chaos models or non-linear system models</td></tr><tr><td>G06N 99/00</td><td>Other AI technology</td><td>Subject matter not provided for in other groups of this subclass</td></tr></table>
+
+
+Source: USPTO Class 706 Data processing: Artificial intelligence.
+
+Report on FY2014 Trend survey of patent application technology: Artificial intelligence (2016) https://www.jpo.go.jp/shiryou/pdf/gidou-houkoku/26_21.pdf.
+
+33
+
+---
+
+Appendix 2. Patent data collection period in Patentscope database by country
+
+<table><tr><td>Country</td><td>Data collection period</td><td>Country</td><td>Data collection period</td></tr><tr><td>PCT</td><td>20.10.1978–16.12.2016</td><td>Jordan</td><td>31.12.1899–16.03.2016</td></tr><tr><td>Argentina</td><td>12.02.1965–24.03.2016</td><td>Kenya</td><td>12.05.1996–01.02.2011</td></tr><tr><td>Bahrain</td><td>10.03.1957–29.09.2005</td><td>Mexico</td><td>02.12.1991–03.06.2016</td></tr><tr><td>Brazil</td><td>26.04.1972–07.09.2016</td><td>Morocco</td><td>07.07.1977–01.06.2016</td></tr><tr><td>Canada</td><td>12.08.1869–27.11.2016</td><td>Nicaragua</td><td>23.12.1982–07.02.2014</td></tr><tr><td>Chile</td><td>20.04.2000–28.05.2016</td><td>Panama</td><td>10.03.1990–30.10.2013</td></tr><tr><td>China</td><td>05.01.1989–10.11.2016</td><td>Peru</td><td>22.02.1989–01.01.2016</td></tr><tr><td>Colombia</td><td>14.02.1995–01.07.2016</td><td>Portugal</td><td>24.06.1967–01.07.2016</td></tr><tr><td>Costa Rica</td><td>03.10.0108–30.07.2016</td><td>Republic of Korea</td><td>27.07.1979–01.09.2016</td></tr><tr><td>Cuba</td><td>13.03.1968–01.09.2016</td><td>Russian Federation</td><td>16.02.1993–28.11.2016</td></tr><tr><td>Dominican Rep.</td><td>11.05.1964–01.07.2016</td><td>Russian Federation (USSR data)</td><td>01.03.1919–28.11.2016</td></tr><tr><td>Ecuador</td><td>02.10.1990–01.05.2015</td><td>Singapore</td><td>29.11.1995–31.08.2016</td></tr><tr><td>Egypt</td><td>02.01.2002–31.10.2014</td><td>South Africa</td><td>27.01.1983–30.07.2015</td></tr><tr><td>El Salvador</td><td>11.03.1970–25.06.2016</td><td>Spain</td><td>15.03.1827–17.08.2016</td></tr><tr><td>Estonia</td><td>18.10.1994–16.11.2016</td><td>Tunisia</td><td>01.01.1999–31.03.2016</td></tr><tr><td>Eurasian Patent Office</td><td>02.07.1996–01.09.2016</td><td>United Arab Emirates</td><td>02.07.2002–03.01.2013</td></tr><tr><td>Germany</td><td>03.07.1877–25.11.2016</td><td>United Kingdom</td><td>05.07.1782–13.10.2016</td></tr><tr><td>Germany (DDR data)</td><td>15.06.1951–23.04.1999</td><td>United States of America</td><td>01.08.1790–14.12.2016</td></tr><tr><td>Guatemala</td><td>31.12.1961–18.06.2016</td><td>Uruguay</td><td>17.08.1990–30.07.2016</td></tr><tr><td>Honduras</td><td>30.01.2004–28.04.2015</td><td>Viet Nam</td><td>26.05.1997–27.04.2010</td></tr><tr><td>Israel</td><td>02.01.1900–01.11.2016</td><td>ARIPO</td><td>04.07.1985–29.07.2008</td></tr><tr><td>Japan</td><td>09.01.1993–11.11.2016</td><td>European Patent Office</td><td>21.12.1978–24.11.2016</td></tr></table>
+
+
+Source: World Intellectual Property Organization, National Collections–Data Coverage. https://patentscope.wipo.int/search/en/help/data_coverage.jsf
+
+34
+
+---
+
+![Figure](figures/figure_000.png)
+
+Appendix 3. Decomposition analysis of patents granted for the biological model (number of items)
+
+Note: The vertical axis is standardized by setting the number of changes in patents granted in 2000 to zero.
+
+35
+
+---
+
+![Figure](figures/figure_000.png)
+
+Appendix 4. Decomposition analysis of patents granted for the knowledge-based model (number of items)
+
+Note: The vertical axis is standardized by setting the number of changes in patents granted in 2000 to zero.
+
+36
+
+---
+
+![Figure](figures/figure_000.png)
+
+Appendix 5. Decomposition analysis of patents granted for specific mathematical model (number of items)
+
+Note: The vertical axis is standardized by setting the number of changes in patents granted in 2000 to zero.
+
+37
+
+---
+
+![Figure](figures/figure_000.png)
+
+Appendix 6. Decomposition analysis of patents granted for other AI technology (number of items)
+
+Note: The vertical axis is standardized by setting the number of changes in patents granted in 2000 to zero.
+
+38
+

processed/00043_W2152794177/document.md

@@ -0,0 +1,308 @@
+# Patent Overlay Mapping: Visualizing Technological Distance
+
+Luciano Kay Center for Nanotechnology in Society, University of California Santa Barbara, 2327 Girvetz Hall, Santa Barbara, CA 93106-2150. E-mail: luciano.kay@ucsb.edu Nils Newman Intelligent Information Services Corporation, P.O. Box 77691, Atlanta, GA 30357. E-mail: newman@iisco.com Jan Youtie Enterprise Innovation Institute & School of Public Policy, 75 Fifth Street NW, Suite 300, Atlanta, GA 30308. E-mail: jan.youtie@innovate.gatech.edu Alan L. Porter School of Public Policy, Georgia Institute of Technology, Atlanta GA & Search Technologies, Atlanta, GA 30332-0345. E-mail: alan.porter@isye.gatech.edu Ismael Rafols Ingenio (CSIC-UPV), Universitat Politècnica de València, CPI, 8E 4rt, Accés J, Camí de Vera s/n, València 46022, Spain & SPRU - Science and Technology Policy Research, University of Sussex, Brighton, UK. E-mail: i.rafols@ingenio.upv.es
+
+This paper presents a new global patent map that represents all technological categories and a method to locate patent data of individual organizations and technological fields on the global map. This overlay map technique may support competitive intelligence and policy decision making. The global patent map is based on similarities in citing-to-cited relationships between categories of the International Patent Classification (IPC) of European Patent Office (EPO) patents from 2000 to 2006. This patent data set, extracted from the PATSTAT database, includes 760,000 patent records in 466 IPC-based categories. We compare the global patent maps derived from this categorization to related efforts of other global patent maps. The paper overlays the nanotechnology-related patenting activities of two companies and two different nanotechnology subfields on the global patent map. The exercise shows the potential of patent overlay maps to visualize technological areas and potentially support decision making. Furthermore, this study shows that IPC categories that are similar to one another based on citingto-cited patterns (and thus close in the global patent map) are not necessarily in the same hierarchical IPC branch,
+
+thereby revealing new relationships between technologies that are classified as pertaining to different (and sometimes distant) subject areas in the IPC scheme.
+
+## Introduction
+
+The visualization of knowledge or technological landscapes has been a prominent part of publication and patent analyses since their origins (Hinze, Reiss, & Schmoch, 1997; Small, 1973) . However, only in the past decade have improvements in computational power and algorithms allowed the creation of large maps covering a full database, the so-called global maps of science (see overviews by Klavans & Boyack, 2009; Rafols, Porter, & Leydesdorff, 2010) . $^1$ These science maps or scientograms are the visualization of the relations among areas of science using network analysis algorithms.
+
+Visualization procedures for science maps have generally been used to explore and visually identify scientific
+
+Received July 1, 2013; revised September 26, 2013; accepted October 2, 2013
+
+© 2014 ASIS&T • Published online 7 May 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/asi.23146
+
+1 Lately there has been a proliferation of global maps (see, e.g., Bollen et al., 2009; Boyack, Börner, & Klavans, 2009; Boyack, Klavans, & Börner, 2005; Janssens, Zhang, Moor, & Glänzel, 2009; Leydesdorff & Rafols, 2009; Moya-Anegon et al., 2004; Moya-Anegón, Vargas-Quesada, Chinchilla-Rodríguez, Corera-Álvarez, & Herrero-Solana, 2007; Rosvall & Bergstrom, 2010) .
+
+JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY, 65(12):2432-2443, 2014
+
+---
+
+frontiers, grasp the extent and evolution of scientific domains, and analyze the frontiers of scientific research change (Van den Besselaar & Leydesdorff, 1996) . Science mapping efforts have also been used to inspire crossdisciplinary discussion to find ways to communicate scientific progress (see, e.g., Mapping Science at http:// www.scimaps.org/). Although science maps cannot replace other methodological approaches to data analysis, visual thinking can help to interpret and find meaning in complex data by transforming abstract and intangible data sets into something visible and concrete (Chen, 2003) . Diverse approaches can be used to create visualizations.
+
+The purpose of this paper is twofold: first, to present the results of a global patent map and, second, to introduce the “ overlay map ” technique to locate the relative technological position of an organization's patent activity to support competitive intelligence and policy decision making. This research draws on the concept of technological distance to interpret linkages among technologies and elaborate a method for a meaningful visualization of technological landscapes.
+
+This visualization approach is a logical extension of the experience with science overlay maps. It draws closely on our previous work on science mapping (Rafols et al., 2010) and opens up new avenues for understanding patent landscapes, which markedly differ from scientific landscapes. The need for the development of tools to benchmark and capture the temporal change of organizational innovation activities, or patterns of technological change, motivates this work. More generally, this new approach accompanies the broader change from hierarchical, structured knowledge in science and technology (i.e., with subdisciplines and specialties that match departmental structures) to a web of ways of knowing resulting from changing social contracts (Gibbons et al., 1994) and increasing institutional hybridity (Etzkowitz & Leydesdorff, 2000) and dissonance between epistemic and social structures. Our paper shows that in many instances, technological similarity based on citing-tocited references is not the same as the hierarchical structures used to organize patented knowledge.
+
+To exemplify the kind of analytical support offered by this approach, this paper illustrates the application of patent overlay maps to benchmark the nanotechnology-related patenting activities of two companies and to reveal the core structure of patenting activities in two different nanotechnology subfields. Nanotechnology is an umbrella term referring to a diverse set of emerging technologies that improve or enable materials, devices, and systems using novel properties resulting from the engineering and assembly of matter at extremely small scales. At the nanoscale, scientific discoveries have unveiled novel properties that offer the potential for applications in a wide array of market segments such as energy, pharmaceuticals, and semiconductors. With a range of potential applications, nanotechnology is anticipated to have significant business and economic impacts in future years. Our previous work illustrated how science overlay maps help to provide a better understanding of the
+
+characteristics and evolution of the nanotechnology field and its subfields (see, e.g., Porter & Youtie, 2009; Rafols & Meyer, 2010).
+
+This paper is organized as follows: The next section reviews and discusses the concept of technological distance and the analysis of patent literature. The Implementation section presents the methodological approach. Test and Preliminary Results presents preliminary outputs based on the application of patent overlay maps to general patent data sets and the analysis of company patent portfolios and technological fields. Finally, the Conclusion discusses the advantages and drawbacks of the method and elaborates on the next steps and future of patent mapping. The paper also includes information to access supplementary material made available by the authors online as detailed in the Appendix.
+
+## Technological Distance and Its Operationalization
+
+Technological distance, or the extent to which a set of patents reflects different types of technologies, is a key characteristic in being able to visualize innovative opportunities (Breschi, Lissoni, & Malerba, 2003) . Patent documents that reference other patents in similar technology areas have been suggested to offer incremental opportunities to advance an area, whereas patent documents that refer across diverse categories may offer the potential for radical innovation (Olsson, 2004) . Technological distance is often proxied by patent categories, with patents in a given patent category being considered more similar to one another than to those in other patent categories (Jaffe, 1986; Kauffman, Lobo, & Macready, 2000) . For example, Franz (2009) uses patent citations between U.S. patent categories and assigns weights to a patent citing another patent in a different category to reflect a larger technological distance. Hinze et al. (1997) look at co-assignment of multiple International Patent Classification (IPC) categories as a measure of the distance between 30 technological fields. A challenge in relying on patent classifications is that, as technology changes, technology-oriented applications may draw from patents in different hierarchical categories, and subsequently lead to further diversity in patents that cite patents in these categories. Hinze et al.'s (1997) contribution was important because it established that the global map of patents is similar for different countries (United States, Japan, Germany) and for different time periods (1982 – 1985, 1986 – 1989, and 1990 – 1993). Given such stability, one can then think of this stable structure as a basemap over which to compare the technological distribution of specific organizations in the same way that we may compare the distribution of different plant species or multinationals over the world map.
+
+This investigation draws on the concept of technological distance and proposes an alternative approach to relying on administrative patent categories using patent mapping techniques to visualize technological landscapes based on similarity as indicated through citing-to-cited relationships. A
+
+JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014 2433 DOI: 10.1002/asi
+
+---
+
+patent map is a symbolic representation of technological fields that are associated with relevant themes. Technological fields are positioned in the map so that similar fields are situated nearby and dissimilar components are situated at a distance. The map is constructed from a similarity matrix based on citing-to-cited patents (i.e., a matrix that reflects similarities among IPC categories in how patents cite each other). The similarity measures are calculated from correlation functions among fields according to citations among patent categories. This multidimensional matrix is projected onto a two-dimensional space. Visual output provides for flexibility in interpreting the multidimensional relationships among the patent categories. In addition, this approach allows the user to “ overlay” subsets of patent data — representing different types of technological fields, institutions, or geographical regions — to understand the particular technological thrusts and areas of concentration of these entities (Rafols et al., 2010) .
+
+Recently, other scholars have pursued a similar patent record-level approach to create global maps of technology that characterizes the proximity and dependency of technological areas (see, e.g., preliminary work in Boyack & Klavans [2008] , and related approaches by Schoen et al. [2012] , or Leydesdorff, Kushnir, & Rafols [2012] ). $^2$ Those efforts have also sought to use the maps to benchmark industrial corporations in order to inform corporate and policy decision making. The differences with the approach presented in this paper are primarily related to the definition of categories (which yields different numbers and compositions of technology groups) and the relationships among them (generally based on citation-based co-occurrence of IPC categories, which yields maps with different structures). The Boyack and Klavans (2008) work is based on Class (3-digit) level IPC categories, and Leydesdorff et al. (2012) include Class (3-digit) and Subclass (4-digit) analyses based on USPTO data rather than EPO. These IPC-based approaches work with the existing classification system, which is a product of patent office history, regardless of the intensive quantity of patents in certain categories. For example, categories such as A61 (“Medical or Veterinary Science”) have a large quantity of patents, whereas categories such as A42B (“Hats”) have few. This uneven distribution of patents limits visualization ability if using the native classification system as is. The contribution of this work is the development of a patent mapping approach based on IPC categories that corrects this uneven patent distribution as explained below. The Schoen et al. (2012) patent map is based on technology-based categorization that combines different IPC branches. As was the case in science maps (Klavans & Boyack, 2009; Rafols & Leydesdorff, 2009) , it is important to compare the results of diverse global patent maps using different classification and visualization algorithms to test the robustness of patterns observed. Without significant consensus on the shape and
+
+relative position of categories, global maps are meaningless as stable landscapes needed to compare organizational or technological subsets.
+
+The approach used in this paper draws on learning from the authors' prior work on science mapping, particularly the trade-off between sufficient detail and not too much detail to be easily visualized by the user. The challenges faced when developing this kind of patent map include gathering patent data in appropriate quantity to create meaningful maps and the choice of an equivalent to citation patterns (because citations may not be functionally equivalent to journal citations) and an equivalent to Web of Science Categories (previously known as ISI Subject Categories), for which IPC categories may not be suitable analogs. Using IPC categories from patent documents also involves specific challenges, such as deciding on the appropriate level of analysis to obtain satisfactory results. This latter point is related to the IPC classification scheme that offers Sections, Classes, Subclasses, and Groups from which to choose. Although the Subclass (i.e., 4-digit IPC) level seems appropriate because of the degree of detail in subject matter definitions, it suffers a “ population ” problem related to the significant variation of the number of patents classified in each IPC Subclass, which is likely to lead to underrepresented technologies in maps. Some Subclasses have several hundred thousand patents, whereas others have only a few hundred. Thus, a more appropriate grouping of IPC categories is needed to more evenly represent the number of patents across the patent system.
+
+## Implementation
+
+This global patent map is based on citing-to-cited relationships among the IPCs of European Patent Office (EPO) patents from 2000 to 2006. This period was chosen because of its stability with respect to IPC 7 categories. IPC 7, at the time we conducted this study, represented the longest period of stable classification, as IPC 8 was just rolling out and could potentially add and/or modify categories. Future work would involve comparing patent overlay maps based on IPC 7 and IPC 8, but first the project team needed to make sure it could produce a mapping process with a stable set of categories. The data set containing IPCs' relationships, extracted from the fall 2010 PATSTAT database version, represents more than 760,000 patent records in more than 400 IPC categories. This data range begins with patent EP0968708 (which was published in January 2000) and ends with patent EP1737233 (published in December 2006.) An analysis with this kind of coverage benefits from a relative stability of Version 7 of the patent classification system maintained during the 2000 – 2006 time period.
+
+In this approach, the process of data gathering and preprocessing involves, first, going through each patent record to collect all the instances of IPC categories in the data set and, second, solving the aforementioned “population” problem. The proposed solution for patent categories with relatively few patents is to fold the IPC category up into the next highest
+
+2 Thomson Reuters also has a patent visualization capability, Aureka, but it is a local rather than a global mapping application.
+
+2434 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014 DOI: 10.1002/asi
+
+---
+
+TABLE 1. Data preprocessing to group IPC categories, selected examples.
+
+<table><tr><td>Original IPC in data set</td><td>Catchwords</td><td>Original record count</td></tr><tr><td>A61B Authors' process</td><td>Diagnosis; Surgery; Identification splits this out into:</td><td>25,808</td></tr><tr><td>A61B 5/00</td><td>Measuring for diagnostic purposes</td><td>1,415</td></tr><tr><td>A61B 17/00</td><td>Surgical instruments, devices or methods, e.g. tourniquets</td><td>1,493</td></tr><tr><td>A61B 19/00</td><td>Instruments, implements or accessories for surgery or diagnosis not covered by any of the groups</td><td>1,444</td></tr><tr><td>and a remainder: A61B</td><td></td><td></td></tr></table>
+
+
+Note. "Each IPC with an instance count greater than 1,000 was kept in its original state.
+
+
+bEach IPC with an instance count less than 1,000 was folded up to the next highest level until the count exceeded 1,000 or the class level was reached.
+
+level of aggregation to create relatively similar sized categories. This solution comprises the following three rules: (a) for IPC categories with large population, use the smallest Subgroup level; (b) for small population IPC categories, aggregate up to General Group level, Subclass, or Class; and (c) establish a floor cutoff and drop small aggregated populations. As a result, IPC categories with instance counts greater than 1,000 in the data set were kept in their original state. Those categories with instance counts less than 1,000 were folded up to the next highest level until the count exceeded 1,000 or the Class level was reached. During the folding, any other IPC categories with counts exceeding 1,000 in the same branch were left out of the folding count. If at the Class level (i.e., 3-digit), the population was less than 1,000, the IPC code was dropped for being too small to map. Table 1 illustrates this approach for the 4-digit IPC class A61B.
+
+This preprocessing (in which the roll-up heuristics were performed through a compiled code written in C++) yields IPC categories at the Class, Subclass, Main Group, and Subgroup levels, with levels that ensure broadly similar numbers (i.e., within two orders of magnitude) of patents across categories. Although we keep referring to these categories as “ IPC categories, ” they are not the standard IPC categories since they have a mixed hierarchical composition. The smallest categories in the data set have 1,000 patents, with this bottom threshold chosen to yield a sufficient count for statistical analyses. The largest category—A61K (defined as “ Preparations for Medical, Dental, or Toilet Purposes ” ) even with subtracting 16 seven-digit IPCs with more than 1,000 patents each—still has more than 85,000 patents. The initial implementation actually involved testing several cutoff values (e.g., 700, 1,000, and 1,500 records) that yielded different numbers of IPC categories. The cutoff at 1,000 was deemed suitable for this analysis, as it seems to provide a sensible compromise between accuracy of the fields, and readability in the map. This choice produces 466
+
+IPC categories that are mapped to a thesaurus for data preprocessing. Out of these categories, 44 categories (representing 2.78 million patents) remain at the Class (3-digit) level, 297 categories (representing 29.11 million patents) remain at the Subclass (4-digit) level, 56 at categories at the Main Group level (representing 5.10 million patents), and 69 at the Subgroup level (representing 4.75 million patents) (Table 2).
+
+TABLE 2. Number of categories and patents obtained with the multilevel aggregation process.
+
+<table><tr><td>Level in classification  $^{2}$ </td><td># Categories</td><td>Mean # apps  $^{b}$ </td><td> $\%$  of apps  $^{c}$ </td></tr><tr><td>Class (3 digit)</td><td>44</td><td>63,280</td><td>6.7%</td></tr><tr><td>Subclass (4 digit)</td><td>297</td><td>97,997</td><td>69.7%</td></tr><tr><td>Main Group (7 digit, \00)</td><td>56</td><td>91,144</td><td>12.2%</td></tr><tr><td>Subgroup (7 digit, \#\#)</td><td>69</td><td>68,781</td><td>11.4%</td></tr><tr><td>Total</td><td>466</td><td>89,569</td><td>100.0%</td></tr></table>
+
+
+Note. $^{1}$ See www.wipo.int for more information about these levels.
+
+b Mean number of patent applications.
+
+c Share of patent applications in the data set.
+
+The next step involves extracting from PATSTAT the patents cited by the target records. The IPCs of those patents are mapped to the 466 IPC categories. Some of the patents cited by those in our IPC 7 data set were published under previous categorization systems; however, this spillover does not lead to any problems from a categorization standpoint because the IPC integrates prior categorizations into more recent versions. The result of this data collection allows the creation of a table containing, in each row, sets of Patent Number, IPC Number, Cited Patent Number, and Cited IPC Number. This data table has been further processed and saved in an appropriate file format for the next step using the software Pajek. This software also helped to create the global map and individual overlay maps for examples of companies and technological fields.
+
+The final data processing steps involve generating a cosine similarity matrix among citing IPC categories (using conventional cosine similarity normalized by the square root of the squared sum) and then factor analysis of the IPC categories (following the method used in global science maps by Leydesdorff & Rafols [2009] ). A factor analysis of the citing-to-cited matrix among IPC categories is then used to consolidate the 466 categories into 35 “ macro patent categories. ” No distinction was made between primary and second classifications and all citing-to-cited relationships were counted equally (i.e., without fractional counting). We tested different factor solutions from 10 to 40. The 35-factor solution appeared to provide a sensible and convenient classification of the IPC categories. These 35 factors form the basis for color-coding the 466 categories that are represented in visualizations. The list of 35 factors is available in Supplementary File 1 (see details in the Appendix). The visualizations also require converting IPC codes to succinct text labels, which we did by shortening lengthy IPC definitions. Therefore, labels may not fully capture all the technologies within a category. These IPC category labels were then used
+
+JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014 2435 DOI: 10.1002/asi
+
+---
+
+![Figure](figures/figure_000.png)
+
+FIG. 1. Full patent map of 466 technology categories and 35 technological areas. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
+
+as a basis for creating descriptors for each factor as shown in the maps (next section).
+
+The creation of patent overlay maps using a wide range of IPC-based categories requires consideration of the classification system of reference. This research draws on the IPC 7 classification system that, compared to previous versions, includes class codes such as B82B that are relevant to the nanotechnology domain. The IPC 7 system is also more stable than the more recent IPC 8, but still received some updates during the time period relevant to this study, including the addition of the B82B technology classification. Those updates do not affect the structure of the maps because the newly added classifications represent a small number of patents (i.e., below our cutoff value) and do not affect the map-based analyses because patent records in newly added classifications are generally assigned to other technology categories as well. $^3$ Future developments of these maps will require updating the thesaurus developed to match the 466 categories of the global patent maps.
+
+## Test and Preliminary Results
+
+### The Global Patent Map
+
+The full map of patents shows all 466 categories in a Kamada-Kawai layout (using Pajek) that represents
+
+technological distances and groups of technologies in each of the 35 factors or technological areas shown with the same color (Figure 1 ). Label- and color-related settings were adjusted to produce a reasonably clear map and facilitate its examination. The map suggests three broad dimensions of patenting interrelationships based on the overall position of technological areas. The left side of the map represents bio-related patents, including food, medicine, and biology. The lower right part of the map includes semiconductor, electronics, and information & communications technologies (ICT). The upper right portion of the map is primarily comprised of automotive and metal-mechanical-related technology groups.
+
+### Difference Between Hierarchy and Similarity
+
+A closer look shows that the structure of the map reflects technological relationships across the hierarchical administrative boundaries of the subject matter specifications in the IPC scheme. While counts of IPC sections (i.e., the first letter of IPC codes, A, B, C, D, E, F, G, H) are commonly used as a measure of technological distance in patents, the 35 technological areas that are derived from cross-citations in our patent map often span multiple sections. For instance, the Vehicles area includes six different sections, and the Heating and Cooling, Construction, and Metals areas include five different sections. Textiles, Lighting, Semiconductors, and Chem and Polymers include four different sections. Eleven technological areas (Measurement, Domestic
+
+3 The analysis shows that only 0.2 % of the patents of Samsung and 2.6 % of the patents of Dupont that are solely assigned to the B82B class are not represented in the maps.
+
+2436 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014 DOI: 10.1002/asi
+
+---
+
+![Figure](figures/figure_000.png)
+
+FIG. 2. Full patent map of 466 technology categories colored according to eight 1-digit IPC classes. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
+
+Appliances, Plastics and Wheels, Photolithography, Optics, Copying and Printing, Catalysis and Separation, Lab equipment, Cosmetics and Med Chem, Biologics, Drugs, and Med Chem) include three different sections. Ten areas (Turbines and Engines, Machine Tools, Furnace, Electric Power, Info Transmission, Data Commerce, Med Instruments, Combustion Engines, Telephone Comm, TV, Imaging & Comm) have two different sections. Only Medical Devices, Food, Recording, Computing, and Radio Communication areas encompass a single section (further details on this are available in Supplementary File 1).
+
+This difference between hierarchy and similarity can be observed by comparing Figure 1 with the same map with the nodes colored according to the eight major IPC sections (Figure 2 ). This observation is strong evidence that the IPC classification on its own is not an appropriate framework to investigate technological diversity without taking account of technological distance.
+
+We can further elucidate the classification underlying our map by relating these categories to those in another prominent patent map. The map developed by Leydesdorff and colleagues uses Classes (3-digit) and Subclasses (4-digit) (refer to http://www.leydesdorff.net/ipcmaps/). In contrast, our map uses more detailed categorizations to disaggregate some of the patent groupings into more fine-grained analyzable components. By way of example, Leydesdorff's Classbased (3-digit) map has a single node representing “ Medical or Veterinary Science ” (IPC A61) on the bottom left side of
+
+the map. However, because A61 includes a large share (more than 20 % ) of patents, our map disaggregates this “ supernode ” into 53 different nodes which end up in five different medical/veterinary science related clusters or technological areas: (a) Drugs, Med. Chemistry, (b) Biologics, (c) Cosmetics and Med. Chemistry, (d) Medical Instruments, and (e) Medical Devices. Each of these areas is made up of categories that come from different sections and are classified at different levels. Table 3 illustrates how these multiple levels coexist for the case of the technological area that we labeled “ Biologics. ” This area includes categories at the Class level ( “ Agriculture, ” A01), Subclass ( “ Peptides, ” C07K), Main Group ( “ Peptides, medical, ” A61K 38/00), and Subgroup ( “ Recombinant DNA, ” C12N 15/09).
+
+Given our map's ability to present disaggregated categories, we are able to show 35 technological areas or clusters versus only five in the Leydesdorff map. This more disaggregated clustering enables differentiation of the patent portfolios of a company engaged in cosmetics patenting from one engaged in drug development and from yet another engaged in medical instrument development. In conclusion, we believe that the multilevel method of classification proposed here achieves a more accurate description than a straightforward use of IPC classes at the Class or Subclass level.
+
+A major problem in these comparisons is that in the case of patents, unlike the map of science, where there has been a preestablished conventional understanding of disciplines,
+
+JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014 2437 DOI: 10.1002/asi
+
+---
+
+TABLE 3. List of categories for the technological area "Biologics."
+
+<table><tr><td>Category label</td><td>IPC number</td><td># Apps</td></tr><tr><td>Agriculture</td><td>A01</td><td>45,126</td></tr><tr><td>Animal husbandry</td><td>A01K</td><td>14,548</td></tr><tr><td>Peptides, medical</td><td>A61K 38/00</td><td>482,120</td></tr><tr><td>Antigens</td><td>A61K 39/00</td><td>20,010</td></tr><tr><td>Antibodies</td><td>A61K 39/395</td><td>47,662</td></tr><tr><td>Gene therapy</td><td>A61K 48/00</td><td>15,899</td></tr><tr><td>Saccharides</td><td>C07H 21/00</td><td>14,578</td></tr><tr><td>Peptides, compounds</td><td>C07K</td><td>58,219</td></tr><tr><td>Peptides from humans</td><td>C07K 14/435</td><td>43,462</td></tr><tr><td>Peptides from animals</td><td>C07K 14/47</td><td>14,602</td></tr><tr><td>Immunoglobulins</td><td>C07K 16/18</td><td>27,481</td></tr><tr><td>Extractions from organisms</td><td>C12N</td><td>26,627</td></tr><tr><td>Modified fung</td><td>C12N 1/15</td><td>47,884</td></tr><tr><td>Modified yeasts</td><td>C12N 1/19</td><td>32,469</td></tr><tr><td>Cellulose processes</td><td>C12N 1/21</td><td>13,631</td></tr><tr><td>Virus transformed cells</td><td>C12N 5/10</td><td>10,402</td></tr><tr><td>Recombinant DNA</td><td>C12N 15/09</td><td>21,345</td></tr><tr><td>Genes encoding animal proteins</td><td>C12N 15/12</td><td>25,010</td></tr><tr><td>Fermentation for food</td><td>C12P</td><td>29,202</td></tr><tr><td>Testing, microorganisms</td><td>C12Q</td><td>18,442</td></tr><tr><td>Testing, nucleic acids</td><td>C12Q 1/68</td><td>22,731</td></tr><tr><td>Bacteriology</td><td>C12R</td><td>48,984</td></tr><tr><td>Measuring biological material</td><td>G01N 33/50</td><td>517,367</td></tr><tr><td>Immunoassay</td><td>G01N 33/53</td><td>43,835</td></tr><tr><td>Measuring using proteins, amino acids, lipids</td><td>G01N 33/68</td><td>119,957</td></tr></table>
+
+
+Note. An equivalent table for each of the 35 technological areas can be found in Supporting File 1, under the tab “Label and Count Table” (Available at: http://www.sussex.ac.uk/Users/ir28/patmap/KaySupplementary1.xls).
+
+it is not clear how groups of technologies can be interpreted. This problem is compounded by the heterogeneous nature of the patent classes, which includes materials (e.g., “ Alloys, ” C22C), devices (e.g., “ Machines and engines, ” F01), and products (e.g., “ Ships, ” B63). This conceptual diversity is observed within the technological classes derived from the factor analysis. For example, the area of “ Turbines and engines ” includes “ Turbines ” (F01D), “ Jet propulsion ” (F02K), “ Aircraft equipment ” (B64D), and “ Airplanes and helicopters ” (B64D) — elements from distinct branches of the IPC classification. These four subclasses obviously co-occur but rather than being similar they likely co-occur because they are embedded and/or complementary. And the area of “ Lightning ” includes “ Basic electric elements ” (H01), Lighting (F21), “ Vehicle signaling ” (B60Q), and “ Specialized equipment used in roads ” (E01F) — which again are complementary pieces of technology at different levels of aggregation. This difficulty we are facing is not simply a problem of classification, but a conundrum due to the multiple meanings and scales that the technology concept may take (Arthur, 2010) .
+
+Awareness of the conceptual heterogeneity of nodes or elements in the map raises the issue of whether the maps show “similarity” between categories as we have assumed, or other properties such as co-occurrence and complementarity. For example, patents of metals and automobiles are related not because these categories are similar but because
+
+automobiles are often made of metals. Also, plastics and metals may co-occur simply because they are materials that are used in similar products such as buckets and automobiles, not because they are similar. $^4$ This issue suggests that the interpretation of the patent map should be ontologically flexible. In other words, when interpreting it, one should take into account that both the elements and the relations may have different meanings.
+
+## Comparison of Map Structures
+
+The overall structure of our map appears to be consistent with previous technological maps based on patents that used different algorithms for aggregating IPC categories (Boyack & Klavans, 2008; Hinze et al., 1997; Leydesdorff et al., 2012; Schoen et al., 2012) . Hinze's map offers the most straightforward comparison given that it only has 35 categories. The similarity in position to our map with the technological areas in the extreme ends of the network is quite striking. For example, in the bioscience pole, categories related with food, drugs, and biotechnology have the same relative position. In the ICT pole as well, optics, audiovisual technologies, and telecommunications also keep the relative position. “ Semiconductors, ” however, occupies a more central position in our map than in Hinze's, resembling the map developed by Boyack but not the one by Leydesdorff. In the vehicle/mechanical pole, one can relate Hinze's categories to our technological areas (engines, turbines, mechanical elements, transport), but it is not possible to compare the relative positions of the two maps due to the lack of comparable labels in the categories. It is worth stressing that Hinze's visualization is based in multidimensional scaling, whereas the one presented in Figure 1 is achieved with the Kamada-Kawai algorithm; hence, the similarities arise from factors other than the particular layouts used in each case.
+
+Based on these similarities between Hinze's and Figure 1 , we are confident that our map captures the main axis in the broad relative position of technologies. The map that differs the most, among those inspected, is the one by Schoen et al. (2012) , which nevertheless still partially captures the three axes mentioned. The difference in Schoen's map might reflect a different layout algorithm rather than substantive differences in the relations among technologies. In sum, further research is needed to better understand the relative structure of technologies, and ascertain whether the structure observed is robust and stable (as it was surprisingly found in the map of science, see Klavans & Boyack [2009] ), or whether it is susceptible to differences stemming from the use of different (still equally valid) presentation angles.
+
+## Interconnectedness
+
+Another interesting feature of the global patent map is the high level of interconnectedness of most of the 35
+
+$^{4}$ We thank Antoine Schoen for this point.
+
+2438 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014 DOI: 10.1002/asi
+
+---
+
+technological areas. This can be observed not only in many connections among technology groups within each technological area, as shown by the densest areas of the map, but also across them. Some exceptions are areas such as Food, Drugs & Med Chem, Biologics, TV Imaging & Comm, Cosm & Med Chem, and Radio & Comm that form more uniform clusters of technology groups (i.e., they appear as clusters of nodes of the same color) (Figure 1 ). Another notable feature is the short distance between technologies in a handful of groups such as Drugs & Med Chem and Biologics, as shown by denser areas and darker lines in the left-hand side of the maps. The sparse areas of the map are those associated with technological areas that comprise fewer technology categories include, for example, Electric Power, Lighting, and Recording.
+
+### Patent Overlay Maps
+
+Based on the global patent map, patent overlay maps allow, for example, benchmarking of companies and specific technological fields. To illustrate and test the application of patent map overlays, two corporate data sets of nanotechnology patent applications have been created for Samsung and DuPont, and two nanotechnology subfield data sets have been created for Nano-Biosensors and Graphene nanotechnology applications, using data from the Georgia Tech Global Nanotechnology databases in the same time period (2000–2006).
+
+The visual examination of maps shows nanotechnology development foci that vary across companies (even for those in similar industry sectors) and different patenting activity levels for the studied period. The two overlays presented herein appear diversified and encompass a number of technological areas. The patent overlay created for Samsung, for example, shows activity concentrated in semiconductors and optics, with a notable level of patenting activity across other areas as well (Figure 3a). The company also has some prominent activity in technological areas broadly defined as Catalysis & Separation, Photolithography, and Chemistry & Polymers. The focus of DuPont (Figure 3b), on the other hand, is more on Drugs, Medicine & Chemistry, Chemistry & Polymers, and Biologics. This company seems to have a portfolio of patent applications that is even more diversified, but it also is less active in terms of patenting activity, than Samsung.
+
+The application of patent overlays to the analysis of technological subfields can also help provide a better understanding of technologies involved in the development of these subfields and relationships between them and with the patent portfolio of companies. Yet while the patent maps applied to companies reflect the result of a corporate strategy implemented by a single organization, patent maps applied to technological fields reflect the aggregation of activities of multiple (and usually numerous) organizations in the same or different sectors.
+
+In the application of patent overlay maps to nanotechnology, technological developments in nanobiosensors are
+
+focused on categories such as Laboratory Equipment, Semiconductors, and Biologics (Figure 4a). The subfield of Graphene, a more recent domain that was recognized with the 2010 Nobel Prize in Physics, presents lower activity levels with a diversified focus on Catalysis & Separation, Chemistry & Polymers, Semiconductors and Optics, among others (Figure 4b).
+
+## Conclusion
+
+This paper presents preliminary results of a new patent visualization tool with potential to support competitive intelligence and policy decision making, following a method successfully used in science overlay mapping (Rafols et al., 2010) . The approach involves a two-step visualization process. First, we build a global map that shows the technological distance among patent categories using citing-tocited information for 7 years of EPO data. Second, we overlay the patenting activity of specific organizations or in specific technological fields over the fixed “ backbone ” of the patent map. The aim of this superposition or overlay is to help understand the patent portfolio of an organization in the context of the overall technological landscape.
+
+The approach offers distinctive visualization capability with parsimony. In contrast to prior IPC-based global patent maps, this approach recombines IPC categories to reflect a finer distribution of patents. Thus, it enables improved differentiation ability in categories with a large amount of patenting activity such as “ Medical or Veterinary Science ” (IPC A61).
+
+The definition of categories and its implementation using a thesaurus to match IPC categories facilitates replication by helping to trace back individual categories to verify results and make improvements. Nevertheless, these maps are only reliable to the extent that assignation of patents to IPC categories is accurate and meaningful. Since patent assignation to IPCs may not always be accurate, a large set of patents may be required to ensure that the portfolio of patents shown in an overlay map can be trusted to convey the patenting activities of an organization represented (in the case of science maps, this was estimated to be above 1,500 publications for high-resolution accuracy, and above 100 publications for lower resolution) (see Appendix 1 in Rafols et al., 2010) .
+
+One of the most interesting findings is that IPC categories that are close to one another in the patent map are not necessarily in the same hierarchical IPC branch. This finding reveals new patterns of relationships among technologies that pertain to different (and sometimes distant) subject areas in the IPC classification. The finding suggests that technological distance is not always well proxied by relying on the IPC administrative structure, for example, by assuming that a set of patents represents substantial technological distance because the set references different IPC sections. This paper shows that patents in certain technology areas tend to cite multiple and diverse IPC sections. For example, the Drugs & Medicine and Biologics dimensions include
+
+JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014 2439 DOI: 10.1002/asi
+
+---
+
+![Figure](figures/figure_000.png)
+
+FIG. 3. Patent overlays applied to company benchmarking. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
+
+2440 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014 DOI: 10.1002/asi
+
+---
+
+![Figure](figures/figure_000.png)
+
+FIG. 4. Patent overlays applied to field mapping. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
+
+JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014 2441 DOI: 10.1002/asi
+
+---
+
+various drug-related Subclasses in IPC Class A61, but they also include several chemistry compound Subclasses in IPC Class C07; traditional measures would assume that technologies in these dimensions are distant because they include two different sections (sections A and C), but our network map shows that technologies in these two sections are closely interrelated, inasmuch as the patents in these Subclasses tend to cite one another. An improved measure of technological distance would take into consideration patent citation or co-occurrence characteristics.
+
+Potential applications of patent overlay maps include organizational and regional/country benchmarking (e.g., for the examination of competitive positions), exploration of potential collaborations, and general analysis of technological changes over time. For example, the comparison of maps over time can reveal new patterns of relationships among categories that might help to understand the emergence of new fields and the extent of their impact. Patent maps may also reveal relatively unexplored technological areas that are more central to other technologies or highlight denser areas with more technological interdependency that may form platforms for the emergence of future technology applications (such as the Drugs & Medicine and Biologics categories in the maps shown in this paper). Most of these explorations may require greater granularity for such analysis and policy decision making (except in the case of large firms with extensive patent portfolios, such as the examples of Samsung and DuPont illustrated). This need for granularity is a challenge that faces all global maps. Future work would enable greater ability to drill-down in certain areas, as well as to compare different global maps — for example, maps based on IPC 8 with maps based on IPC 7 version — but a stable global map is required as an initial base for such an effort.
+
+Ongoing work has sought to overcome some issues found in the development of the original patent overlay maps. Among the most important issues is the coverage of the thesaurus developed to match 466 IPC categories based on the main patent data set. Although this data set covers a wide range of IPC categories, the resulting thesaurus still does not match a number of IPC categories in the data sets created for patent overlay maps. This kind of issue varies across patent overlay data sets and may represent a significant proportion of the patent records in certain cases. This is, however, a problem that can be solved in future implementations by creating a new thesaurus based on a larger data set that covers more than 7 years of patent activity.
+
+The next steps in this research thrust include updates of the basemap based on the current version of the PATSTAT database and use of the most recent IPC classification, version 8, and eventually the Cooperative Patent Classification (CPC). Refining the patent database to focus only on patent grants (it currently includes applications as well as grants) is one path for future work, while another is to develop a patent map for patents from other patent authorities besides EPO. In addition, the stability of the patent maps could be tested with the segmentation of maps by year or
+
+year ranges. The backbone patent map in this paper should be compared with results from other global patent mapping efforts to determine the extent of consistency between these maps. Although we have presented some preliminary comparisons, a more rigorous and systematic approach for comparing these maps and categorizations is needed (see, e.g., Klavans & Boyack, 2009; Rafols & Leydesdorff, 2009) . Potential future research includes the analysis of connections between patent maps and science maps, with particular focus on technological fields with strong science links, 5 as well as classifications based on full clustering of an entire database rather than a subset (as recently done in science with more than 10 million records by Waltman & van Eck, 2012) .
+
+## Acknowledgments
+
+We thank Kevin Boyack, Loet Leydesdorff, and Antoine Schoen for open and fruitful discussions about this paper. This research was undertaken largely at Georgia Tech drawing on support from the U.S. National Science Foundation (NSF) through the Center for Nanotechnology in Society (Arizona State University; Award No. 0531194); and NSF Award No. 1064146 (“Revealing Innovation Pathways: Hybrid Science Maps for Technology Assessment and Foresight”). Part of this research was also undertaken in collaboration with the Center for Nanotechnology in Society, University of California Santa Barbara (NSF Awards No. 0938099 and No. 0531184). The findings and observations contained in this paper are those of the authors and do not necessarily reflect the views of the US National Science Foundation.
+
+## References
+
+Arthur, W.B. (2010). The nature of technology. London: Penguin.
+
+Bollen, J., Van de Sompel, H., Hagberg, A., Bettencourt, L., Chute, R., Rodriguez, M.A., & Balakireva, L. (2009). Clickstream data yields highresolution maps of science. PLoS ONE, 4(3), e4803.
+
+Boyack, K.W., Börner, K., & Klavans, R. (2009). Mapping the structure and evolution of chemistry research. Scientometrics, 79(1), 45-60.
+
+Boyack, K.W., & Klavans, R. (2008). Measuring science–technology interaction using rare inventor–author names. Journal of Informetrics, 2, 173–182.
+
+Boyack, K.W., Klavans, R., & Börner, K. (2005). Mapping the backbone of science. Scientometrics, 64, 351-374.
+
+Breschi, S., Lissoni, F., & Malerba, F. (2003). Knowledge-relatedness in firm technological diversification. Research Policy, 32, 69-87.
+
+Chen, C. (2003). Mapping scientific frontiers: The quest for knowledge visualization. London: Springer.
+
+Etzkowitz, H., & Leydesdorff, L. (2000). The dynamics of innovation: From National Systems and “Mode 2” to a Triple Helix of universityindustry-government relations. Research Policy, 29, 109–123.
+
+Franz, J.S. (2009). Constructing technological distances from US patent data. IEEE Xplore Online Proceedings, Atlanta Conference on Science, Technology and Innovation.
+
+5See, for example, http://www.mapofscience.com for an overlay of patents in the map of science carried out by Boyack and Klavans (unpublished). Accessed September 23, 2013.
+
+2442 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014 DOI: 10.1002/asi
+
+---
+
+Gibbons, M., Limoges, C., Nowotny, H., Schwartzman, S., Scott, P., & Trow, M. (1994). The new production of knowledge. The dynamics of science and research in contemporary societies. London: SAGE.
+
+Hinze, S., Reiss, T., & Schmoch, U. (1997). Statistical analysis on the distance between fields of technology. Paper presented at the Innovation Systems and European Integration (ISE), Targeted Socio-Economic Research Program, 4th Framework Program of the European Commission (DGXII), Karlsruhe, Germany. Retrieved from http://www.isi .fraunhofer.de/isi-media/docs/isi-publ/1997/isi97b81/technology-fieldsdiastance.pdf?WSESSIONID=5712ff2ca5ffcf0d9590afc8ef7e1486.
+
+Jaffe, A. (1986). Technological opportunities and spillovers of R&D Evidence from firms' patents, profits, and market value. American Economic Review, 76(5), 984–1001.
+
+Janssens, F., Zhang, L., Moor, B.D., & Glänzel, W. (2009). Hybrid clustering for validation and improvement of subject-classification schemes. Information Processing & Management, 45(6), 683-702.
+
+Kauffman, S., Lobo, J., & Macready, W.G. (2000). Optimal search on a technology landscape. Journal of Economic Behaviour and Organization, 43, 141-166.
+
+Klavans, R., & Boyack, K.W. (2009). Toward a consensus map of science. Journal of the American Society for Information Science and Technology, 60(3), 455-476.
+
+Leydesdorff, L., Kushnir, D., & Rafols, I. (2012). Interactive overlay maps for US patent (USPTO) data based on International Patent Classification (IPC). Scientometrics. DOI: 10.1007/s11192-012-0923-2.
+
+Leydesdorff, L., & Rafols, I. (2009). A global map of science based on the ISI subject categories. Journal of the American Society for Information Science and Technology, 60(2), 348-362.
+
+Moya-Anegón, F., Vargas-Quesada, B., Chinchilla-Rodríguez, Z., Corera-Álvarez, E., & Herrero-Solana, V. (2007). Visualizing the marrow of science. Journal of the American Society for Information Science and Technology, 58(14), 2167–2179.
+
+Moya-Anegon, F., Vargas-Quesada, B., Herrero-Solana, V., ChinchillaRodriguez, Z., Corera-Alvarez, E., & Munoz-Fernandez, F.J. (2004). A new technique for building maps of large scientific domains based on the cocitation of classes and categories. Scientometrics, 61(1), 129-145.
+
+Olsson, O. (2004). Technological opportunity and growth. Journal of Economic Growth, 10(1), 35-57.
+
+Porter, A.L., & Youtie, J. (2009). Where does nanotechnology belong in the map of science? Nature Nanotechnology, 4, 534-536.
+
+Rafols, I., & L., Leydesdorff, L. (2009). Content-based and algorithmic classifications of journals: Perspectives on the dynamics of scientific communication and indexer effects. Journal of the American Society for Information Science and Technology, 60(9), 1823-1835.
+
+Rafols, I., & Meyer, M. (2010). Diversity and network coherence as indicators of interdisciplinarity: Case studies in bionanoscience. Scientometrics, 82(2), 263–287.
+
+Rafols, I., Porter, A.L., & Leydesdorff, L. (2010). Science overlay maps: A new tool for research policy and library management. Journal of the American Society for Information Science and Technology, 61(9), 18711887.
+
+Rosvall, M., & Bergstrom, C.T. (2010). Mapping change in large networks. PLoS ONE, 5(1), e8694.
+
+Schoen, A., Villard, L., Laurens, P., Cointet, J.-P., Heimeriks, G., & Alkemade, F. (2012). The network structure of technological developments: Technological distance as a walk on the technology map. Presented at the STI Indicators Conference 2012, Montréal, Canada.
+
+Small, H. (1973). Citing-to-cited in the scientific literature: A new measure of the relationship between two documents. Journal of the American Society for Information Science and Technology, 24(4), 265-269.
+
+Van den Besselaar, P., & Leydesdorff, L. (1996). Mapping change in scientific specialties: A scientometric reconstruction of the development of artificial intelligence. Journal of the American Society for Information Science and Technology, 46(6), 415-436.
+
+Waltman, L., & van Eck, N.J. (2012). A new methodology for constructing a publication-level classification system of science. Journal of the American Society for Information Science and Technology, 63(12), 2378– 2392.
+
+## Supporting Information
+
+Additional Supporting Information may be found in the online version of this article at the publisher's web-site:
+
+Appendix. Supplementary Materials.
+
+JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014 2443 DOI: 10.1002/asi
+

processed/00043_W2152794177/metadata.json

@@ -0,0 +1,1589 @@
+{
+  "source_file": "/tmp/tmpkpv2wg8f/00043_W2152794177_Patent_overlay_mapping_Visualizing_technological_distance.pdf",
+  "total_pages": 12,
+  "pages": [
+    {
+      "page_number": 1,
+      "elements": [
+        {
+          "bbox": [
+            61,
+            67,
+            422,
+            117
+          ],
+          "label": "sec_0",
+          "reading_order": 0,
+          "text": "Patent Overlay Mapping: Visualizing\nTechnological Distance"
+        },
+        {
+          "bbox": [
+            62,
+            194,
+            621,
+            453
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "Luciano Kay\nCenter for Nanotechnology in Society, University of California Santa Barbara, 2327 Girvetz Hall, Santa\nBarbara, CA 93106-2150. E-mail: luciano.kay@ucsb.edu\nNils Newman\nIntelligent Information Services Corporation, P.O. Box 77691, Atlanta, GA 30357. E-mail: newman@iisco.com\nJan Youtie\nEnterprise Innovation Institute & School of Public Policy, 75 Fifth Street NW, Suite 300, Atlanta, GA 30308.\nE-mail: jan.youtie@innovate.gatech.edu\nAlan L. Porter\nSchool of Public Policy, Georgia Institute of Technology, Atlanta GA & Search Technologies, Atlanta, GA\n30332-0345. E-mail: alan.porter@isye.gatech.edu\nIsmael Rafols\nIngenio (CSIC-UPV), Universitat Politècnica de València, CPI, 8E 4rt, Accés J, Camí de Vera s/n, València\n46022, Spain & SPRU - Science and Technology Policy Research, University of Sussex, Brighton, UK.\nE-mail: i.rafols@ingenio.upv.es"
+        },
+        {
+          "bbox": [
+            62,
+            491,
+            340,
+            745
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "This paper presents a new global patent map that repre-\nsents all technological categories and a method to locate\npatent data of individual organizations and technological\nfields on the global map. This overlay map technique may\nsupport competitive intelligence and policy decision\nmaking. The global patent map is based on similarities in\nciting-to-cited relationships between categories of the\nInternational Patent Classification (IPC) of European\nPatent Office (EPO) patents from 2000 to 2006. This\npatent data set, extracted from the PATSTAT database,\nincludes 760,000 patent records in 466 IPC-based catego-\nries. We compare the global patent maps derived from\nthis categorization to related efforts of other global patent\nmaps. The paper overlays the nanotechnology-related\npatenting activities of two companies and two different\nnanotechnology subfields on the global patent map. The\nexercise shows the potential of patent overlay maps to\nvisualize technological areas and potentially support\ndecision making. Furthermore, this study shows that IPC\ncategories that are similar to one another based on citing-\nto-cited patterns (and thus close in the global patent map)\nare not necessarily in the same hierarchical IPC branch,"
+        },
+        {
+          "bbox": [
+            353,
+            491,
+            631,
+            530
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "thereby revealing new relationships between technolo-\ngies that are classified as pertaining to different (and\nsometimes distant) subject areas in the IPC scheme."
+        },
+        {
+          "bbox": [
+            353,
+            546,
+            424,
+            561
+          ],
+          "label": "sec_1",
+          "reading_order": 4,
+          "text": "Introduction"
+        },
+        {
+          "bbox": [
+            353,
+            566,
+            631,
+            717
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "The visualization of knowledge or technological land-\nscapes has been a prominent part of publication and patent\nanalyses since their origins (Hinze, Reiss, & Schmoch,\n1997; Small, 1973) . However, only in the past decade have\nimprovements in computational power and algorithms\nallowed the creation of large maps covering a full database,\nthe so-called global maps of science (see overviews by\nKlavans & Boyack, 2009; Rafols, Porter, & Leydesdorff,\n2010) . $^1$ These science maps or scientograms are the visual-\nization of the relations among areas of science using\nnetwork analysis algorithms."
+        },
+        {
+          "bbox": [
+            353,
+            717,
+            630,
+            745
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "Visualization procedures for science maps have generally\nbeen used to explore and visually identify scientific"
+        },
+        {
+          "bbox": [
+            62,
+            768,
+            338,
+            793
+          ],
+          "label": "foot",
+          "reading_order": 7,
+          "text": "Received July 1, 2013; revised September 26, 2013; accepted October 2,\n2013"
+        },
+        {
+          "bbox": [
+            62,
+            801,
+            338,
+            825
+          ],
+          "label": "foot",
+          "reading_order": 8,
+          "text": "© 2014 ASIS&T • Published online 7 May 2014 in Wiley Online Library\n(wileyonlinelibrary.com). DOI: 10.1002/asi.23146"
+        },
+        {
+          "bbox": [
+            353,
+            755,
+            630,
+            824
+          ],
+          "label": "fnote",
+          "reading_order": 9,
+          "text": "1 Lately there has been a proliferation of global maps (see, e.g., Bollen\net al., 2009; Boyack, Börner, & Klavans, 2009; Boyack, Klavans, & Börner,\n2005; Janssens, Zhang, Moor, & Glänzel, 2009; Leydesdorff & Rafols,\n2009; Moya-Anegon et al., 2004; Moya-Anegón, Vargas-Quesada,\nChinchilla-Rodríguez, Corera-Álvarez, & Herrero-Solana, 2007; Rosvall &\nBergstrom, 2010) ."
+        },
+        {
+          "bbox": [
+            62,
+            845,
+            532,
+            859
+          ],
+          "label": "foot",
+          "reading_order": 10,
+          "text": "JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY, 65(12):2432-2443, 2014"
+        }
+      ]
+    },
+    {
+      "page_number": 2,
+      "elements": [
+        {
+          "bbox": [
+            63,
+            49,
+            340,
+            215
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "frontiers, grasp the extent and evolution of scientific\ndomains, and analyze the frontiers of scientific research\nchange (Van den Besselaar & Leydesdorff, 1996) . Science\nmapping efforts have also been used to inspire cross-\ndisciplinary discussion to find ways to communicate scien-\ntific progress (see, e.g., Mapping Science at http://\nwww.scimaps.org/). Although science maps cannot replace\nother methodological approaches to data analysis, visual\nthinking can help to interpret and find meaning in complex\ndata by transforming abstract and intangible data sets into\nsomething visible and concrete (Chen, 2003) . Diverse\napproaches can be used to create visualizations."
+        },
+        {
+          "bbox": [
+            63,
+            215,
+            340,
+            337
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "The purpose of this paper is twofold: first, to present the\nresults of a global patent map and, second, to introduce the\n“ overlay map ” technique to locate the relative technological\nposition of an organization's patent activity to support com-\npetitive intelligence and policy decision making. This\nresearch draws on the concept of technological distance to\ninterpret linkages among technologies and elaborate a\nmethod for a meaningful visualization of technological\nlandscapes."
+        },
+        {
+          "bbox": [
+            63,
+            337,
+            340,
+            595
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "This visualization approach is a logical extension of the\nexperience with science overlay maps. It draws closely on\nour previous work on science mapping (Rafols et al., 2010)\nand opens up new avenues for understanding patent land-\nscapes, which markedly differ from scientific landscapes.\nThe need for the development of tools to benchmark and\ncapture the temporal change of organizational innovation\nactivities, or patterns of technological change, motivates this\nwork. More generally, this new approach accompanies the\nbroader change from hierarchical, structured knowledge in\nscience and technology (i.e., with subdisciplines and spe-\ncialties that match departmental structures) to a web of ways\nof knowing resulting from changing social contracts\n(Gibbons et al., 1994) and increasing institutional hybridity\n(Etzkowitz & Leydesdorff, 2000) and dissonance between\nepistemic and social structures. Our paper shows that in\nmany instances, technological similarity based on citing-to-\ncited references is not the same as the hierarchical structures\nused to organize patented knowledge."
+        },
+        {
+          "bbox": [
+            63,
+            595,
+            340,
+            826
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "To exemplify the kind of analytical support offered by\nthis approach, this paper illustrates the application of patent\noverlay maps to benchmark the nanotechnology-related pat-\nenting activities of two companies and to reveal the core\nstructure of patenting activities in two different nanotech-\nnology subfields. Nanotechnology is an umbrella term refer-\nring to a diverse set of emerging technologies that improve\nor enable materials, devices, and systems using novel prop-\nerties resulting from the engineering and assembly of matter\nat extremely small scales. At the nanoscale, scientific dis-\ncoveries have unveiled novel properties that offer the poten-\ntial for applications in a wide array of market segments\nsuch as energy, pharmaceuticals, and semiconductors. With\na range of potential applications, nanotechnology is antici-\npated to have significant business and economic impacts in\nfuture years. Our previous work illustrated how science\noverlay maps help to provide a better understanding of the"
+        },
+        {
+          "bbox": [
+            354,
+            49,
+            630,
+            91
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "characteristics and evolution of the nanotechnology field\nand its subfields (see, e.g., Porter & Youtie, 2009; Rafols &\nMeyer, 2010)."
+        },
+        {
+          "bbox": [
+            354,
+            91,
+            631,
+            270
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "This paper is organized as follows: The next section\nreviews and discusses the concept of technological distance\nand the analysis of patent literature. The Implementation\nsection presents the methodological approach. Test and\nPreliminary Results presents preliminary outputs based on\nthe application of patent overlay maps to general patent\ndata sets and the analysis of company patent portfolios and\ntechnological fields. Finally, the Conclusion discusses the\nadvantages and drawbacks of the method and elaborates on\nthe next steps and future of patent mapping. The paper also\nincludes information to access supplementary material\nmade available by the authors online as detailed in the\nAppendix."
+        },
+        {
+          "bbox": [
+            354,
+            287,
+            628,
+            304
+          ],
+          "label": "sec_1",
+          "reading_order": 6,
+          "text": "Technological Distance and Its Operationalization"
+        },
+        {
+          "bbox": [
+            354,
+            308,
+            631,
+            758
+          ],
+          "label": "para",
+          "reading_order": 7,
+          "text": "Technological distance, or the extent to which a set of\npatents reflects different types of technologies, is a key char-\nacteristic in being able to visualize innovative opportunities\n(Breschi, Lissoni, & Malerba, 2003) . Patent documents\nthat reference other patents in similar technology areas have\nbeen suggested to offer incremental opportunities to\nadvance an area, whereas patent documents that refer across\ndiverse categories may offer the potential for radical inno-\nvation (Olsson, 2004) . Technological distance is often\nproxied by patent categories, with patents in a given patent\ncategory being considered more similar to one another than\nto those in other patent categories (Jaffe, 1986; Kauffman,\nLobo, & Macready, 2000) . For example, Franz (2009) uses\npatent citations between U.S. patent categories and assigns\nweights to a patent citing another patent in a different\ncategory to reflect a larger technological distance. Hinze\net al. (1997) look at co-assignment of multiple International\nPatent Classification (IPC) categories as a measure of the\ndistance between 30 technological fields. A challenge in\nrelying on patent classifications is that, as technology\nchanges, technology-oriented applications may draw from\npatents in different hierarchical categories, and subsequently\nlead to further diversity in patents that cite patents in these\ncategories. Hinze et al.'s (1997) contribution was important\nbecause it established that the global map of patents is\nsimilar for different countries (United States, Japan,\nGermany) and for different time periods (1982 – 1985, 1986 –\n1989, and 1990 – 1993). Given such stability, one can then\nthink of this stable structure as a basemap over which to\ncompare the technological distribution of specific organiza-\ntions in the same way that we may compare the distribution\nof different plant species or multinationals over the world\nmap."
+        },
+        {
+          "bbox": [
+            354,
+            758,
+            631,
+            826
+          ],
+          "label": "para",
+          "reading_order": 8,
+          "text": "This investigation draws on the concept of technological\ndistance and proposes an alternative approach to relying on\nadministrative patent categories using patent mapping tech-\nniques to visualize technological landscapes based on simi-\nlarity as indicated through citing-to-cited relationships. A"
+        },
+        {
+          "bbox": [
+            152,
+            846,
+            628,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 9,
+          "text": "JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014\n2433\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 3,
+      "elements": [
+        {
+          "bbox": [
+            62,
+            49,
+            339,
+            296
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "patent map is a symbolic representation of technological\nfields that are associated with relevant themes. Technologi-\ncal fields are positioned in the map so that similar fields are\nsituated nearby and dissimilar components are situated at a\ndistance. The map is constructed from a similarity matrix\nbased on citing-to-cited patents (i.e., a matrix that reflects\nsimilarities among IPC categories in how patents cite each\nother). The similarity measures are calculated from correla-\ntion functions among fields according to citations among\npatent categories. This multidimensional matrix is projected\nonto a two-dimensional space. Visual output provides for\nflexibility in interpreting the multidimensional relationships\namong the patent categories. In addition, this approach\nallows the user to “ overlay” subsets of patent data —\nrepresenting different types of technological fields, institu-\ntions, or geographical regions — to understand the particular\ntechnological thrusts and areas of concentration of these\nentities (Rafols et al., 2010) ."
+        },
+        {
+          "bbox": [
+            62,
+            296,
+            340,
+            772
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "Recently, other scholars have pursued a similar patent\nrecord-level approach to create global maps of technology\nthat characterizes the proximity and dependency of techno-\nlogical areas (see, e.g., preliminary work in Boyack &\nKlavans [2008] , and related approaches by Schoen et al.\n[2012] , or Leydesdorff, Kushnir, & Rafols [2012] ). $^2$ Those\nefforts have also sought to use the maps to benchmark indus-\ntrial corporations in order to inform corporate and policy\ndecision making. The differences with the approach pre-\nsented in this paper are primarily related to the definition of\ncategories (which yields different numbers and compositions\nof technology groups) and the relationships among them\n(generally based on citation-based co-occurrence of IPC\ncategories, which yields maps with different structures). The\nBoyack and Klavans (2008) work is based on Class (3-digit)\nlevel IPC categories, and Leydesdorff et al. (2012) include\nClass (3-digit) and Subclass (4-digit) analyses based on\nUSPTO data rather than EPO. These IPC-based approaches\nwork with the existing classification system, which is a\nproduct of patent office history, regardless of the intensive\nquantity of patents in certain categories. For example, cat-\negories such as A61 (“Medical or Veterinary Science”) have\na large quantity of patents, whereas categories such as A42B\n(“Hats”) have few. This uneven distribution of patents limits\nvisualization ability if using the native classification system\nas is. The contribution of this work is the development of a\npatent mapping approach based on IPC categories that cor-\nrects this uneven patent distribution as explained below. The\nSchoen et al. (2012) patent map is based on technology-based\ncategorization that combines different IPC branches. As was\nthe case in science maps (Klavans & Boyack, 2009; Rafols &\nLeydesdorff, 2009) , it is important to compare the results of\ndiverse global patent maps using different classification and\nvisualization algorithms to test the robustness of patterns\nobserved. Without significant consensus on the shape and"
+        },
+        {
+          "bbox": [
+            353,
+            50,
+            630,
+            91
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "relative position of categories, global maps are meaningless\nas stable landscapes needed to compare organizational or\ntechnological subsets."
+        },
+        {
+          "bbox": [
+            353,
+            91,
+            631,
+            445
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "The approach used in this paper draws on learning from\nthe authors' prior work on science mapping, particularly the\ntrade-off between sufficient detail and not too much detail to\nbe easily visualized by the user. The challenges faced when\ndeveloping this kind of patent map include gathering patent\ndata in appropriate quantity to create meaningful maps and\nthe choice of an equivalent to citation patterns (because\ncitations may not be functionally equivalent to journal cita-\ntions) and an equivalent to Web of Science Categories (pre-\nviously known as ISI Subject Categories), for which IPC\ncategories may not be suitable analogs. Using IPC categories\nfrom patent documents also involves specific challenges,\nsuch as deciding on the appropriate level of analysis to obtain\nsatisfactory results. This latter point is related to the IPC\nclassification scheme that offers Sections, Classes, Sub-\nclasses, and Groups from which to choose. Although the\nSubclass (i.e., 4-digit IPC) level seems appropriate because\nof the degree of detail in subject matter definitions, it suffers\na “ population ” problem related to the significant variation of\nthe number of patents classified in each IPC Subclass, which\nis likely to lead to underrepresented technologies in maps.\nSome Subclasses have several hundred thousand patents,\nwhereas others have only a few hundred. Thus, a more\nappropriate grouping of IPC categories is needed to more\nevenly represent the number of patents across the patent\nsystem."
+        },
+        {
+          "bbox": [
+            353,
+            463,
+            442,
+            479
+          ],
+          "label": "sec_1",
+          "reading_order": 4,
+          "text": "Implementation"
+        },
+        {
+          "bbox": [
+            353,
+            485,
+            631,
+            744
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "This global patent map is based on citing-to-cited rela-\ntionships among the IPCs of European Patent Office (EPO)\npatents from 2000 to 2006. This period was chosen because\nof its stability with respect to IPC 7 categories. IPC 7, at the\ntime we conducted this study, represented the longest period\nof stable classification, as IPC 8 was just rolling out and\ncould potentially add and/or modify categories. Future work\nwould involve comparing patent overlay maps based on IPC\n7 and IPC 8, but first the project team needed to make sure\nit could produce a mapping process with a stable set of\ncategories. The data set containing IPCs' relationships,\nextracted from the fall 2010 PATSTAT database version,\nrepresents more than 760,000 patent records in more than\n400 IPC categories. This data range begins with patent\nEP0968708 (which was published in January 2000) and\nends with patent EP1737233 (published in December 2006.)\nAn analysis with this kind of coverage benefits from a rela-\ntive stability of Version 7 of the patent classification system\nmaintained during the 2000 – 2006 time period."
+        },
+        {
+          "bbox": [
+            353,
+            744,
+            630,
+            826
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "In this approach, the process of data gathering and prepro-\ncessing involves, first, going through each patent record to\ncollect all the instances of IPC categories in the data set and,\nsecond, solving the aforementioned “population” problem.\nThe proposed solution for patent categories with relatively\nfew patents is to fold the IPC category up into the next highest"
+        },
+        {
+          "bbox": [
+            62,
+            801,
+            338,
+            825
+          ],
+          "label": "fnote",
+          "reading_order": 7,
+          "text": "2 Thomson Reuters also has a patent visualization capability, Aureka, but\nit is a local rather than a global mapping application."
+        },
+        {
+          "bbox": [
+            63,
+            846,
+            540,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 8,
+          "text": "2434 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 4,
+      "elements": [
+        {
+          "bbox": [
+            63,
+            50,
+            339,
+            74
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "TABLE 1. Data preprocessing to group IPC categories, selected\nexamples."
+        },
+        {
+          "bbox": [
+            63,
+            77,
+            337,
+            245
+          ],
+          "label": "tab",
+          "reading_order": 1,
+          "text": "<table><tr><td>Original IPC in data set</td><td>Catchwords</td><td>Original record count</td></tr><tr><td>A61B Authors' process</td><td>Diagnosis; Surgery; Identification splits this out into:</td><td>25,808</td></tr><tr><td>A61B 5/00</td><td>Measuring for diagnostic purposes</td><td>1,415</td></tr><tr><td>A61B 17/00</td><td>Surgical instruments, devices or methods, e.g. tourniquets</td><td>1,493</td></tr><tr><td>A61B 19/00</td><td>Instruments, implements or accessories for surgery or diagnosis not covered by any of the groups</td><td>1,444</td></tr><tr><td>and a remainder: A61B</td><td></td><td></td></tr></table>"
+        },
+        {
+          "bbox": [
+            63,
+            248,
+            339,
+            297
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Note. \"Each IPC with an instance count greater than 1,000 was kept in its\noriginal state.\n\n\nbEach IPC with an instance count less than 1,000 was folded up to the next\nhighest level until the count exceeded 1,000 or the class level was reached."
+        },
+        {
+          "bbox": [
+            63,
+            335,
+            340,
+            554
+          ],
+          "label": "half_para",
+          "reading_order": 3,
+          "text": "level of aggregation to create relatively similar sized catego-\nries. This solution comprises the following three rules: (a) for\nIPC categories with large population, use the smallest Sub-\ngroup level; (b) for small population IPC categories, aggre-\ngate up to General Group level, Subclass, or Class; and (c)\nestablish a floor cutoff and drop small aggregated popula-\ntions. As a result, IPC categories with instance counts greater\nthan 1,000 in the data set were kept in their original state.\nThose categories with instance counts less than 1,000 were\nfolded up to the next highest level until the count exceeded\n1,000 or the Class level was reached. During the folding, any\nother IPC categories with counts exceeding 1,000 in the same\nbranch were left out of the folding count. If at the Class level\n(i.e., 3-digit), the population was less than 1,000, the IPC\ncode was dropped for being too small to map. Table 1 illus-\ntrates this approach for the 4-digit IPC class A61B."
+        },
+        {
+          "bbox": [
+            63,
+            554,
+            340,
+            826
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "This preprocessing (in which the roll-up heuristics were\nperformed through a compiled code written in C++) yields\nIPC categories at the Class, Subclass, Main Group, and\nSubgroup levels, with levels that ensure broadly similar\nnumbers (i.e., within two orders of magnitude) of patents\nacross categories. Although we keep referring to these cat-\negories as “ IPC categories, ” they are not the standard IPC\ncategories since they have a mixed hierarchical composition.\nThe smallest categories in the data set have 1,000 patents,\nwith this bottom threshold chosen to yield a sufficient count\nfor statistical analyses. The largest category—A61K\n(defined as “ Preparations for Medical, Dental, or Toilet Pur-\nposes ” ) even with subtracting 16 seven-digit IPCs with more\nthan 1,000 patents each—still has more than 85,000 patents.\nThe initial implementation actually involved testing several\ncutoff values (e.g., 700, 1,000, and 1,500 records) that\nyielded different numbers of IPC categories. The cutoff at\n1,000 was deemed suitable for this analysis, as it seems to\nprovide a sensible compromise between accuracy of the\nfields, and readability in the map. This choice produces 466"
+        },
+        {
+          "bbox": [
+            354,
+            240,
+            631,
+            337
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "IPC categories that are mapped to a thesaurus for data pre-\nprocessing. Out of these categories, 44 categories (represent-\ning 2.78 million patents) remain at the Class (3-digit) level,\n297 categories (representing 29.11 million patents) remain at\nthe Subclass (4-digit) level, 56 at categories at the Main\nGroup level (representing 5.10 million patents), and 69 at the\nSubgroup level (representing 4.75 million patents) (Table 2)."
+        },
+        {
+          "bbox": [
+            355,
+            50,
+            630,
+            74
+          ],
+          "label": "cap",
+          "reading_order": 6,
+          "text": "TABLE 2. Number of categories and patents obtained with the multilevel\naggregation process."
+        },
+        {
+          "bbox": [
+            356,
+            77,
+            628,
+            165
+          ],
+          "label": "tab",
+          "reading_order": 7,
+          "text": "<table><tr><td>Level in classification  $^{2}$ </td><td># Categories</td><td>Mean # apps  $^{b}$ </td><td> $\\%$  of apps  $^{c}$ </td></tr><tr><td>Class (3 digit)</td><td>44</td><td>63,280</td><td>6.7%</td></tr><tr><td>Subclass (4 digit)</td><td>297</td><td>97,997</td><td>69.7%</td></tr><tr><td>Main Group (7 digit, \\00)</td><td>56</td><td>91,144</td><td>12.2%</td></tr><tr><td>Subgroup (7 digit, \\#\\#)</td><td>69</td><td>68,781</td><td>11.4%</td></tr><tr><td>Total</td><td>466</td><td>89,569</td><td>100.0%</td></tr></table>"
+        },
+        {
+          "bbox": [
+            354,
+            168,
+            613,
+            205
+          ],
+          "label": "anno",
+          "reading_order": 8,
+          "text": "Note. $^{1}$ See www.wipo.int for more information about these levels.\n\nb Mean number of patent applications.\n\nc Share of patent applications in the data set."
+        },
+        {
+          "bbox": [
+            354,
+            336,
+            631,
+            540
+          ],
+          "label": "para",
+          "reading_order": 9,
+          "text": "The next step involves extracting from PATSTAT the\npatents cited by the target records. The IPCs of those patents\nare mapped to the 466 IPC categories. Some of the patents\ncited by those in our IPC 7 data set were published under\nprevious categorization systems; however, this spillover\ndoes not lead to any problems from a categorization stand-\npoint because the IPC integrates prior categorizations into\nmore recent versions. The result of this data collection\nallows the creation of a table containing, in each row, sets of\nPatent Number, IPC Number, Cited Patent Number, and\nCited IPC Number. This data table has been further pro-\ncessed and saved in an appropriate file format for the next\nstep using the software Pajek. This software also helped to\ncreate the global map and individual overlay maps for\nexamples of companies and technological fields."
+        },
+        {
+          "bbox": [
+            354,
+            540,
+            631,
+            826
+          ],
+          "label": "para",
+          "reading_order": 10,
+          "text": "The final data processing steps involve generating a\ncosine similarity matrix among citing IPC categories (using\nconventional cosine similarity normalized by the square root\nof the squared sum) and then factor analysis of the IPC\ncategories (following the method used in global science\nmaps by Leydesdorff & Rafols [2009] ). A factor analysis of\nthe citing-to-cited matrix among IPC categories is then used\nto consolidate the 466 categories into 35 “ macro patent\ncategories. ” No distinction was made between primary and\nsecond classifications and all citing-to-cited relationships\nwere counted equally (i.e., without fractional counting). We\ntested different factor solutions from 10 to 40. The 35-factor\nsolution appeared to provide a sensible and convenient clas-\nsification of the IPC categories. These 35 factors form the\nbasis for color-coding the 466 categories that are represented\nin visualizations. The list of 35 factors is available in Supple-\nmentary File 1 (see details in the Appendix). The visualiza-\ntions also require converting IPC codes to succinct text\nlabels, which we did by shortening lengthy IPC definitions.\nTherefore, labels may not fully capture all the technologies\nwithin a category. These IPC category labels were then used"
+        },
+        {
+          "bbox": [
+            152,
+            846,
+            628,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 11,
+          "text": "JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014\n2435\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 5,
+      "elements": [
+        {
+          "bbox": [
+            63,
+            49,
+            631,
+            376
+          ],
+          "label": "fig",
+          "reading_order": 0,
+          "text": "![Figure](figures/figure_000.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_000.png"
+        },
+        {
+          "bbox": [
+            62,
+            386,
+            631,
+            411
+          ],
+          "label": "cap",
+          "reading_order": 1,
+          "text": "FIG. 1. Full patent map of 466 technology categories and 35 technological areas. [Color figure can be viewed in the online issue, which is available at\nwileyonlinelibrary.com.]"
+        },
+        {
+          "bbox": [
+            62,
+            434,
+            339,
+            462
+          ],
+          "label": "half_para",
+          "reading_order": 2,
+          "text": "as a basis for creating descriptors for each factor as shown in\nthe maps (next section)."
+        },
+        {
+          "bbox": [
+            62,
+            462,
+            340,
+            694
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "The creation of patent overlay maps using a wide range of\nIPC-based categories requires consideration of the classifi-\ncation system of reference. This research draws on the IPC\n7 classification system that, compared to previous versions,\nincludes class codes such as B82B that are relevant to the\nnanotechnology domain. The IPC 7 system is also more\nstable than the more recent IPC 8, but still received some\nupdates during the time period relevant to this study, includ-\ning the addition of the B82B technology classification.\nThose updates do not affect the structure of the maps\nbecause the newly added classifications represent a small\nnumber of patents (i.e., below our cutoff value) and do not\naffect the map-based analyses because patent records in\nnewly added classifications are generally assigned to other\ntechnology categories as well. $^3$ Future developments of\nthese maps will require updating the thesaurus developed to\nmatch the 466 categories of the global patent maps."
+        },
+        {
+          "bbox": [
+            62,
+            709,
+            223,
+            726
+          ],
+          "label": "sec_1",
+          "reading_order": 4,
+          "text": "Test and Preliminary Results"
+        },
+        {
+          "bbox": [
+            63,
+            730,
+            176,
+            746
+          ],
+          "label": "sec_2",
+          "reading_order": 5,
+          "text": "The Global Patent Map"
+        },
+        {
+          "bbox": [
+            63,
+            751,
+            339,
+            780
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "The full map of patents shows all 466 categories in\na Kamada-Kawai layout (using Pajek) that represents"
+        },
+        {
+          "bbox": [
+            353,
+            434,
+            631,
+            613
+          ],
+          "label": "para",
+          "reading_order": 7,
+          "text": "technological distances and groups of technologies in each\nof the 35 factors or technological areas shown with the same\ncolor (Figure 1 ). Label- and color-related settings were\nadjusted to produce a reasonably clear map and facilitate its\nexamination. The map suggests three broad dimensions of\npatenting interrelationships based on the overall position of\ntechnological areas. The left side of the map represents\nbio-related patents, including food, medicine, and biology.\nThe lower right part of the map includes semiconductor,\nelectronics, and information & communications technolo-\ngies (ICT). The upper right portion of the map is primarily\ncomprised of automotive and metal-mechanical-related\ntechnology groups."
+        },
+        {
+          "bbox": [
+            353,
+            626,
+            566,
+            643
+          ],
+          "label": "sec_2",
+          "reading_order": 8,
+          "text": "Difference Between Hierarchy and Similarity"
+        },
+        {
+          "bbox": [
+            353,
+            648,
+            631,
+            825
+          ],
+          "label": "para",
+          "reading_order": 9,
+          "text": "A closer look shows that the structure of the map reflects\ntechnological relationships across the hierarchical adminis-\ntrative boundaries of the subject matter specifications in the\nIPC scheme. While counts of IPC sections (i.e., the first\nletter of IPC codes, A, B, C, D, E, F, G, H) are commonly\nused as a measure of technological distance in patents, the\n35 technological areas that are derived from cross-citations\nin our patent map often span multiple sections. For instance,\nthe Vehicles area includes six different sections, and\nthe Heating and Cooling, Construction, and Metals areas\ninclude five different sections. Textiles, Lighting, Semicon-\nductors, and Chem and Polymers include four different sec-\ntions. Eleven technological areas (Measurement, Domestic"
+        },
+        {
+          "bbox": [
+            63,
+            790,
+            339,
+            824
+          ],
+          "label": "fnote",
+          "reading_order": 10,
+          "text": "3 The analysis shows that only 0.2 % of the patents of Samsung and 2.6 %\nof the patents of Dupont that are solely assigned to the B82B class are not\nrepresented in the maps."
+        },
+        {
+          "bbox": [
+            63,
+            846,
+            540,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 11,
+          "text": "2436 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 6,
+      "elements": [
+        {
+          "bbox": [
+            62,
+            48,
+            631,
+            376
+          ],
+          "label": "fig",
+          "reading_order": 0,
+          "text": "![Figure](figures/figure_000.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_000.png"
+        },
+        {
+          "bbox": [
+            62,
+            385,
+            631,
+            410
+          ],
+          "label": "cap",
+          "reading_order": 1,
+          "text": "FIG. 2. Full patent map of 466 technology categories colored according to eight 1-digit IPC classes. [Color figure can be viewed in the online issue, which\nis available at wileyonlinelibrary.com.]"
+        },
+        {
+          "bbox": [
+            63,
+            444,
+            340,
+            595
+          ],
+          "label": "half_para",
+          "reading_order": 2,
+          "text": "Appliances, Plastics and Wheels, Photolithography, Optics,\nCopying and Printing, Catalysis and Separation, Lab equip-\nment, Cosmetics and Med Chem, Biologics, Drugs, and\nMed Chem) include three different sections. Ten areas (Tur-\nbines and Engines, Machine Tools, Furnace, Electric Power,\nInfo Transmission, Data Commerce, Med Instruments,\nCombustion Engines, Telephone Comm, TV, Imaging &\nComm) have two different sections. Only Medical Devices,\nFood, Recording, Computing, and Radio Communication\nareas encompass a single section (further details on this are\navailable in Supplementary File 1)."
+        },
+        {
+          "bbox": [
+            63,
+            595,
+            340,
+            690
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "This difference between hierarchy and similarity can be\nobserved by comparing Figure 1 with the same map with the\nnodes colored according to the eight major IPC sections\n(Figure 2 ). This observation is strong evidence that the IPC\nclassification on its own is not an appropriate framework to\ninvestigate technological diversity without taking account of\ntechnological distance."
+        },
+        {
+          "bbox": [
+            63,
+            690,
+            340,
+            826
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "We can further elucidate the classification underlying our\nmap by relating these categories to those in another promi-\nnent patent map. The map developed by Leydesdorff and\ncolleagues uses Classes (3-digit) and Subclasses (4-digit)\n(refer to http://www.leydesdorff.net/ipcmaps/). In contrast,\nour map uses more detailed categorizations to disaggregate\nsome of the patent groupings into more fine-grained analyz-\nable components. By way of example, Leydesdorff's Class-\nbased (3-digit) map has a single node representing “ Medical\nor Veterinary Science ” (IPC A61) on the bottom left side of"
+        },
+        {
+          "bbox": [
+            354,
+            444,
+            631,
+            635
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "the map. However, because A61 includes a large share (more\nthan 20 % ) of patents, our map disaggregates this “ super-\nnode ” into 53 different nodes which end up in five different\nmedical/veterinary science related clusters or technological\nareas: (a) Drugs, Med. Chemistry, (b) Biologics, (c) Cos-\nmetics and Med. Chemistry, (d) Medical Instruments, and\n(e) Medical Devices. Each of these areas is made up of\ncategories that come from different sections and are classi-\nfied at different levels. Table 3 illustrates how these multiple\nlevels coexist for the case of the technological area that\nwe labeled “ Biologics. ” This area includes categories at the\nClass level ( “ Agriculture, ” A01), Subclass ( “ Peptides, ”\nC07K), Main Group ( “ Peptides, medical, ” A61K 38/00), and\nSubgroup ( “ Recombinant DNA, ” C12N 15/09)."
+        },
+        {
+          "bbox": [
+            354,
+            635,
+            631,
+            785
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "Given our map's ability to present disaggregated catego-\nries, we are able to show 35 technological areas or clusters\nversus only five in the Leydesdorff map. This more disag-\ngregated clustering enables differentiation of the patent port-\nfolios of a company engaged in cosmetics patenting from\none engaged in drug development and from yet another\nengaged in medical instrument development. In conclusion,\nwe believe that the multilevel method of classification\nproposed here achieves a more accurate description than a\nstraightforward use of IPC classes at the Class or Subclass\nlevel."
+        },
+        {
+          "bbox": [
+            354,
+            785,
+            631,
+            826
+          ],
+          "label": "para",
+          "reading_order": 7,
+          "text": "A major problem in these comparisons is that in the case\nof patents, unlike the map of science, where there has been\na preestablished conventional understanding of disciplines,"
+        },
+        {
+          "bbox": [
+            152,
+            846,
+            628,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 8,
+          "text": "JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014\n2437\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 7,
+      "elements": [
+        {
+          "bbox": [
+            63,
+            50,
+            320,
+            64
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "TABLE 3. List of categories for the technological area \"Biologics.\""
+        },
+        {
+          "bbox": [
+            63,
+            65,
+            337,
+            380
+          ],
+          "label": "tab",
+          "reading_order": 1,
+          "text": "<table><tr><td>Category label</td><td>IPC number</td><td># Apps</td></tr><tr><td>Agriculture</td><td>A01</td><td>45,126</td></tr><tr><td>Animal husbandry</td><td>A01K</td><td>14,548</td></tr><tr><td>Peptides, medical</td><td>A61K 38/00</td><td>482,120</td></tr><tr><td>Antigens</td><td>A61K 39/00</td><td>20,010</td></tr><tr><td>Antibodies</td><td>A61K 39/395</td><td>47,662</td></tr><tr><td>Gene therapy</td><td>A61K 48/00</td><td>15,899</td></tr><tr><td>Saccharides</td><td>C07H 21/00</td><td>14,578</td></tr><tr><td>Peptides, compounds</td><td>C07K</td><td>58,219</td></tr><tr><td>Peptides from humans</td><td>C07K 14/435</td><td>43,462</td></tr><tr><td>Peptides from animals</td><td>C07K 14/47</td><td>14,602</td></tr><tr><td>Immunoglobulins</td><td>C07K 16/18</td><td>27,481</td></tr><tr><td>Extractions from organisms</td><td>C12N</td><td>26,627</td></tr><tr><td>Modified fung</td><td>C12N 1/15</td><td>47,884</td></tr><tr><td>Modified yeasts</td><td>C12N 1/19</td><td>32,469</td></tr><tr><td>Cellulose processes</td><td>C12N 1/21</td><td>13,631</td></tr><tr><td>Virus transformed cells</td><td>C12N 5/10</td><td>10,402</td></tr><tr><td>Recombinant DNA</td><td>C12N 15/09</td><td>21,345</td></tr><tr><td>Genes encoding animal proteins</td><td>C12N 15/12</td><td>25,010</td></tr><tr><td>Fermentation for food</td><td>C12P</td><td>29,202</td></tr><tr><td>Testing, microorganisms</td><td>C12Q</td><td>18,442</td></tr><tr><td>Testing, nucleic acids</td><td>C12Q 1/68</td><td>22,731</td></tr><tr><td>Bacteriology</td><td>C12R</td><td>48,984</td></tr><tr><td>Measuring biological material</td><td>G01N 33/50</td><td>517,367</td></tr><tr><td>Immunoassay</td><td>G01N 33/53</td><td>43,835</td></tr><tr><td>Measuring using proteins, amino acids, lipids</td><td>G01N 33/68</td><td>119,957</td></tr></table>"
+        },
+        {
+          "bbox": [
+            63,
+            385,
+            338,
+            420
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Note. An equivalent table for each of the 35 technological areas can be\nfound in Supporting File 1, under the tab “Label and Count Table” (Available\nat: http://www.sussex.ac.uk/Users/ir28/patmap/KaySupplementary1.xls)."
+        },
+        {
+          "bbox": [
+            62,
+            458,
+            340,
+            745
+          ],
+          "label": "half_para",
+          "reading_order": 3,
+          "text": "it is not clear how groups of technologies can be interpreted.\nThis problem is compounded by the heterogeneous nature of\nthe patent classes, which includes materials (e.g., “ Alloys, ”\nC22C), devices (e.g., “ Machines and engines, ” F01), and\nproducts (e.g., “ Ships, ” B63). This conceptual diversity is\nobserved within the technological classes derived from the\nfactor analysis. For example, the area of “ Turbines and\nengines ” includes “ Turbines ” (F01D), “ Jet propulsion ”\n(F02K), “ Aircraft equipment ” (B64D), and “ Airplanes and\nhelicopters ” (B64D) — elements from distinct branches of\nthe IPC classification. These four subclasses obviously\nco-occur but rather than being similar they likely co-occur\nbecause they are embedded and/or complementary. And the\narea of “ Lightning ” includes “ Basic electric elements ”\n(H01), Lighting (F21), “ Vehicle signaling ” (B60Q), and\n“ Specialized equipment used in roads ” (E01F) — which\nagain are complementary pieces of technology at different\nlevels of aggregation. This difficulty we are facing is not\nsimply a problem of classification, but a conundrum due to\nthe multiple meanings and scales that the technology\nconcept may take (Arthur, 2010) ."
+        },
+        {
+          "bbox": [
+            62,
+            745,
+            339,
+            825
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "Awareness of the conceptual heterogeneity of nodes or\nelements in the map raises the issue of whether the maps\nshow “similarity” between categories as we have assumed,\nor other properties such as co-occurrence and complemen-\ntarity. For example, patents of metals and automobiles are\nrelated not because these categories are similar but because"
+        },
+        {
+          "bbox": [
+            353,
+            49,
+            631,
+            160
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "automobiles are often made of metals. Also, plastics and\nmetals may co-occur simply because they are materials that\nare used in similar products such as buckets and automo-\nbiles, not because they are similar. $^4$ This issue suggests that\nthe interpretation of the patent map should be ontologically\nflexible. In other words, when interpreting it, one should\ntake into account that both the elements and the relations\nmay have different meanings."
+        },
+        {
+          "bbox": [
+            354,
+            179,
+            501,
+            195
+          ],
+          "label": "sec_1",
+          "reading_order": 6,
+          "text": "Comparison of Map Structures"
+        },
+        {
+          "bbox": [
+            353,
+            200,
+            631,
+            540
+          ],
+          "label": "para",
+          "reading_order": 7,
+          "text": "The overall structure of our map appears to be consistent\nwith previous technological maps based on patents that used\ndifferent algorithms for aggregating IPC categories (Boyack\n& Klavans, 2008; Hinze et al., 1997; Leydesdorff et al.,\n2012; Schoen et al., 2012) . Hinze's map offers the most\nstraightforward comparison given that it only has 35 catego-\nries. The similarity in position to our map with the techno-\nlogical areas in the extreme ends of the network is quite\nstriking. For example, in the bioscience pole, categories\nrelated with food, drugs, and biotechnology have the same\nrelative position. In the ICT pole as well, optics, audiovisual\ntechnologies, and telecommunications also keep the relative\nposition. “ Semiconductors, ” however, occupies a more\ncentral position in our map than in Hinze's, resembling the\nmap developed by Boyack but not the one by Leydesdorff.\nIn the vehicle/mechanical pole, one can relate Hinze's\ncategories to our technological areas (engines, turbines,\nmechanical elements, transport), but it is not possible to\ncompare the relative positions of the two maps due to the\nlack of comparable labels in the categories. It is worth stress-\ning that Hinze's visualization is based in multidimensional\nscaling, whereas the one presented in Figure 1 is achieved\nwith the Kamada-Kawai algorithm; hence, the similarities\narise from factors other than the particular layouts used in\neach case."
+        },
+        {
+          "bbox": [
+            353,
+            540,
+            631,
+            732
+          ],
+          "label": "para",
+          "reading_order": 8,
+          "text": "Based on these similarities between Hinze's and\nFigure 1 , we are confident that our map captures the main\naxis in the broad relative position of technologies. The map\nthat differs the most, among those inspected, is the one by\nSchoen et al. (2012) , which nevertheless still partially cap-\ntures the three axes mentioned. The difference in Schoen's\nmap might reflect a different layout algorithm rather than\nsubstantive differences in the relations among technologies.\nIn sum, further research is needed to better understand the\nrelative structure of technologies, and ascertain whether the\nstructure observed is robust and stable (as it was surprisingly\nfound in the map of science, see Klavans & Boyack [2009] ),\nor whether it is susceptible to differences stemming from the\nuse of different (still equally valid) presentation angles."
+        },
+        {
+          "bbox": [
+            354,
+            750,
+            445,
+            765
+          ],
+          "label": "sec_1",
+          "reading_order": 9,
+          "text": "Interconnectedness"
+        },
+        {
+          "bbox": [
+            353,
+            770,
+            630,
+            800
+          ],
+          "label": "para",
+          "reading_order": 10,
+          "text": "Another interesting feature of the global patent map is\nthe high level of interconnectedness of most of the 35"
+        },
+        {
+          "bbox": [
+            367,
+            812,
+            522,
+            824
+          ],
+          "label": "fnote",
+          "reading_order": 11,
+          "text": "$^{4}$ We thank Antoine Schoen for this point."
+        },
+        {
+          "bbox": [
+            63,
+            846,
+            540,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 12,
+          "text": "2438 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 8,
+      "elements": [
+        {
+          "bbox": [
+            63,
+            49,
+            340,
+            255
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "technological areas. This can be observed not only in many\nconnections among technology groups within each techno-\nlogical area, as shown by the densest areas of the map, but\nalso across them. Some exceptions are areas such as Food,\nDrugs & Med Chem, Biologics, TV Imaging & Comm,\nCosm & Med Chem, and Radio & Comm that form more\nuniform clusters of technology groups (i.e., they appear as\nclusters of nodes of the same color) (Figure 1 ). Another\nnotable feature is the short distance between technologies in\na handful of groups such as Drugs & Med Chem and Bio-\nlogics, as shown by denser areas and darker lines in the\nleft-hand side of the maps. The sparse areas of the map are\nthose associated with technological areas that comprise\nfewer technology categories include, for example, Electric\nPower, Lighting, and Recording."
+        },
+        {
+          "bbox": [
+            63,
+            274,
+            165,
+            290
+          ],
+          "label": "sec_2",
+          "reading_order": 1,
+          "text": "Patent Overlay Maps"
+        },
+        {
+          "bbox": [
+            63,
+            296,
+            340,
+            432
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "Based on the global patent map, patent overlay maps\nallow, for example, benchmarking of companies and specific\ntechnological fields. To illustrate and test the application of\npatent map overlays, two corporate data sets of nanotech-\nnology patent applications have been created for Samsung\nand DuPont, and two nanotechnology subfield data sets have\nbeen created for Nano-Biosensors and Graphene nanotech-\nnology applications, using data from the Georgia Tech\nGlobal Nanotechnology databases in the same time period\n(2000–2006)."
+        },
+        {
+          "bbox": [
+            63,
+            432,
+            340,
+            662
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "The visual examination of maps shows nanotechnology\ndevelopment foci that vary across companies (even for those\nin similar industry sectors) and different patenting activity\nlevels for the studied period. The two overlays presented\nherein appear diversified and encompass a number of tech-\nnological areas. The patent overlay created for Samsung, for\nexample, shows activity concentrated in semiconductors and\noptics, with a notable level of patenting activity across other\nareas as well (Figure 3a). The company also has some\nprominent activity in technological areas broadly defined as\nCatalysis & Separation, Photolithography, and Chemistry &\nPolymers. The focus of DuPont (Figure 3b), on the other\nhand, is more on Drugs, Medicine & Chemistry, Chemistry\n& Polymers, and Biologics. This company seems to have a\nportfolio of patent applications that is even more diversified,\nbut it also is less active in terms of patenting activity, than\nSamsung."
+        },
+        {
+          "bbox": [
+            63,
+            662,
+            340,
+            798
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "The application of patent overlays to the analysis of tech-\nnological subfields can also help provide a better under-\nstanding of technologies involved in the development of\nthese subfields and relationships between them and with the\npatent portfolio of companies. Yet while the patent maps\napplied to companies reflect the result of a corporate strategy\nimplemented by a single organization, patent maps applied\nto technological fields reflect the aggregation of activities of\nmultiple (and usually numerous) organizations in the same\nor different sectors."
+        },
+        {
+          "bbox": [
+            63,
+            798,
+            340,
+            826
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "In the application of patent overlay maps to nanotechnol-\nogy, technological developments in nanobiosensors are"
+        },
+        {
+          "bbox": [
+            353,
+            49,
+            631,
+            148
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "focused on categories such as Laboratory Equipment, Semi-\nconductors, and Biologics (Figure 4a). The subfield of Gra-\nphene, a more recent domain that was recognized with the\n2010 Nobel Prize in Physics, presents lower activity levels\nwith a diversified focus on Catalysis & Separation, Chem-\nistry & Polymers, Semiconductors and Optics, among others\n(Figure 4b)."
+        },
+        {
+          "bbox": [
+            354,
+            165,
+            420,
+            181
+          ],
+          "label": "sec_1",
+          "reading_order": 7,
+          "text": "Conclusion"
+        },
+        {
+          "bbox": [
+            354,
+            186,
+            631,
+            364
+          ],
+          "label": "para",
+          "reading_order": 8,
+          "text": "This paper presents preliminary results of a new patent\nvisualization tool with potential to support competitive intel-\nligence and policy decision making, following a method\nsuccessfully used in science overlay mapping (Rafols et al.,\n2010) . The approach involves a two-step visualization\nprocess. First, we build a global map that shows the techno-\nlogical distance among patent categories using citing-to-\ncited information for 7 years of EPO data. Second, we\noverlay the patenting activity of specific organizations or in\nspecific technological fields over the fixed “ backbone ” of the\npatent map. The aim of this superposition or overlay is to\nhelp understand the patent portfolio of an organization in the\ncontext of the overall technological landscape."
+        },
+        {
+          "bbox": [
+            354,
+            364,
+            631,
+            459
+          ],
+          "label": "para",
+          "reading_order": 9,
+          "text": "The approach offers distinctive visualization capability\nwith parsimony. In contrast to prior IPC-based global patent\nmaps, this approach recombines IPC categories to reflect a\nfiner distribution of patents. Thus, it enables improved dif-\nferentiation ability in categories with a large amount of\npatenting activity such as “ Medical or Veterinary Science ”\n(IPC A61)."
+        },
+        {
+          "bbox": [
+            354,
+            459,
+            631,
+            649
+          ],
+          "label": "para",
+          "reading_order": 10,
+          "text": "The definition of categories and its implementation using\na thesaurus to match IPC categories facilitates replication by\nhelping to trace back individual categories to verify results\nand make improvements. Nevertheless, these maps are only\nreliable to the extent that assignation of patents to IPC cat-\negories is accurate and meaningful. Since patent assignation\nto IPCs may not always be accurate, a large set of patents\nmay be required to ensure that the portfolio of patents shown\nin an overlay map can be trusted to convey the patenting\nactivities of an organization represented (in the case of\nscience maps, this was estimated to be above 1,500 publi-\ncations for high-resolution accuracy, and above 100 publi-\ncations for lower resolution) (see Appendix 1 in Rafols\net al., 2010) ."
+        },
+        {
+          "bbox": [
+            354,
+            649,
+            631,
+            826
+          ],
+          "label": "para",
+          "reading_order": 11,
+          "text": "One of the most interesting findings is that IPC categories\nthat are close to one another in the patent map are not\nnecessarily in the same hierarchical IPC branch. This finding\nreveals new patterns of relationships among technologies\nthat pertain to different (and sometimes distant) subject\nareas in the IPC classification. The finding suggests that\ntechnological distance is not always well proxied by relying\non the IPC administrative structure, for example, by assum-\ning that a set of patents represents substantial technological\ndistance because the set references different IPC sections.\nThis paper shows that patents in certain technology areas\ntend to cite multiple and diverse IPC sections. For example,\nthe Drugs & Medicine and Biologics dimensions include"
+        },
+        {
+          "bbox": [
+            152,
+            846,
+            628,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 12,
+          "text": "JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014\n2439\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 9,
+      "elements": [
+        {
+          "bbox": [
+            63,
+            46,
+            631,
+            785
+          ],
+          "label": "fig",
+          "reading_order": 0,
+          "text": "![Figure](figures/figure_000.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_000.png"
+        },
+        {
+          "bbox": [
+            62,
+            792,
+            630,
+            807
+          ],
+          "label": "cap",
+          "reading_order": 1,
+          "text": "FIG. 3. Patent overlays applied to company benchmarking. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]"
+        },
+        {
+          "bbox": [
+            63,
+            846,
+            540,
+            869
+          ],
+          "label": "foot",
+          "reading_order": 2,
+          "text": "2440 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 10,
+      "elements": [
+        {
+          "bbox": [
+            63,
+            49,
+            627,
+            783
+          ],
+          "label": "fig",
+          "reading_order": 0,
+          "text": "![Figure](figures/figure_000.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_000.png"
+        },
+        {
+          "bbox": [
+            77,
+            792,
+            615,
+            807
+          ],
+          "label": "cap",
+          "reading_order": 1,
+          "text": "FIG. 4. Patent overlays applied to field mapping. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]"
+        },
+        {
+          "bbox": [
+            152,
+            845,
+            627,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 2,
+          "text": "JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014\n2441\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 11,
+      "elements": [
+        {
+          "bbox": [
+            62,
+            49,
+            339,
+            186
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "various drug-related Subclasses in IPC Class A61, but they\nalso include several chemistry compound Subclasses in IPC\nClass C07; traditional measures would assume that tech-\nnologies in these dimensions are distant because they\ninclude two different sections (sections A and C), but our\nnetwork map shows that technologies in these two sections\nare closely interrelated, inasmuch as the patents in these\nSubclasses tend to cite one another. An improved measure of\ntechnological distance would take into consideration patent\ncitation or co-occurrence characteristics."
+        },
+        {
+          "bbox": [
+            62,
+            186,
+            339,
+            513
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "Potential applications of patent overlay maps include\norganizational and regional/country benchmarking (e.g., for\nthe examination of competitive positions), exploration of\npotential collaborations, and general analysis of technologi-\ncal changes over time. For example, the comparison of maps\nover time can reveal new patterns of relationships among\ncategories that might help to understand the emergence of\nnew fields and the extent of their impact. Patent maps may\nalso reveal relatively unexplored technological areas that are\nmore central to other technologies or highlight denser areas\nwith more technological interdependency that may form\nplatforms for the emergence of future technology applica-\ntions (such as the Drugs & Medicine and Biologics catego-\nries in the maps shown in this paper). Most of these\nexplorations may require greater granularity for such analy-\nsis and policy decision making (except in the case of large\nfirms with extensive patent portfolios, such as the examples\nof Samsung and DuPont illustrated). This need for granular-\nity is a challenge that faces all global maps. Future work\nwould enable greater ability to drill-down in certain areas, as\nwell as to compare different global maps — for example,\nmaps based on IPC 8 with maps based on IPC 7 version —\nbut a stable global map is required as an initial base for such\nan effort."
+        },
+        {
+          "bbox": [
+            62,
+            513,
+            339,
+            690
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "Ongoing work has sought to overcome some issues found\nin the development of the original patent overlay maps.\nAmong the most important issues is the coverage of the\nthesaurus developed to match 466 IPC categories based on\nthe main patent data set. Although this data set covers a wide\nrange of IPC categories, the resulting thesaurus still does not\nmatch a number of IPC categories in the data sets created for\npatent overlay maps. This kind of issue varies across patent\noverlay data sets and may represent a significant proportion\nof the patent records in certain cases. This is, however, a\nproblem that can be solved in future implementations by\ncreating a new thesaurus based on a larger data set that\ncovers more than 7 years of patent activity."
+        },
+        {
+          "bbox": [
+            62,
+            690,
+            339,
+            826
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "The next steps in this research thrust include updates of\nthe basemap based on the current version of the PATSTAT\ndatabase and use of the most recent IPC classification,\nversion 8, and eventually the Cooperative Patent Classifica-\ntion (CPC). Refining the patent database to focus only on\npatent grants (it currently includes applications as well as\ngrants) is one path for future work, while another is to\ndevelop a patent map for patents from other patent authori-\nties besides EPO. In addition, the stability of the patent maps\ncould be tested with the segmentation of maps by year or"
+        },
+        {
+          "bbox": [
+            353,
+            49,
+            631,
+            242
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "year ranges. The backbone patent map in this paper should\nbe compared with results from other global patent mapping\nefforts to determine the extent of consistency between these\nmaps. Although we have presented some preliminary com-\nparisons, a more rigorous and systematic approach for com-\nparing these maps and categorizations is needed (see, e.g.,\nKlavans & Boyack, 2009; Rafols & Leydesdorff, 2009) .\nPotential future research includes the analysis of connec-\ntions between patent maps and science maps, with particular\nfocus on technological fields with strong science links, 5 as\nwell as classifications based on full clustering of an entire\ndatabase rather than a subset (as recently done in science\nwith more than 10 million records by Waltman & van Eck,\n2012) ."
+        },
+        {
+          "bbox": [
+            354,
+            262,
+            458,
+            279
+          ],
+          "label": "sec_1",
+          "reading_order": 5,
+          "text": "Acknowledgments"
+        },
+        {
+          "bbox": [
+            353,
+            284,
+            631,
+            489
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "We thank Kevin Boyack, Loet Leydesdorff, and Antoine\nSchoen for open and fruitful discussions about this paper.\nThis research was undertaken largely at Georgia Tech\ndrawing on support from the U.S. National Science Foun-\ndation (NSF) through the Center for Nanotechnology in\nSociety (Arizona State University; Award No. 0531194);\nand NSF Award No. 1064146 (“Revealing Innovation Path-\nways: Hybrid Science Maps for Technology Assessment and\nForesight”). Part of this research was also undertaken in\ncollaboration with the Center for Nanotechnology in\nSociety, University of California Santa Barbara (NSF\nAwards No. 0938099 and No. 0531184). The findings and\nobservations contained in this paper are those of the authors\nand do not necessarily reflect the views of the US National\nScience Foundation."
+        },
+        {
+          "bbox": [
+            353,
+            510,
+            419,
+            526
+          ],
+          "label": "sec_1",
+          "reading_order": 7,
+          "text": "References"
+        },
+        {
+          "bbox": [
+            355,
+            532,
+            599,
+            543
+          ],
+          "label": "reference",
+          "reading_order": 8,
+          "text": "Arthur, W.B. (2010). The nature of technology. London: Penguin."
+        },
+        {
+          "bbox": [
+            355,
+            543,
+            628,
+            577
+          ],
+          "label": "reference",
+          "reading_order": 9,
+          "text": "Bollen, J., Van de Sompel, H., Hagberg, A., Bettencourt, L., Chute, R.,\nRodriguez, M.A., & Balakireva, L. (2009). Clickstream data yields high-\nresolution maps of science. PLoS ONE, 4(3), e4803."
+        },
+        {
+          "bbox": [
+            355,
+            577,
+            628,
+            600
+          ],
+          "label": "reference",
+          "reading_order": 10,
+          "text": "Boyack, K.W., Börner, K., & Klavans, R. (2009). Mapping the structure and\nevolution of chemistry research. Scientometrics, 79(1), 45-60."
+        },
+        {
+          "bbox": [
+            355,
+            600,
+            628,
+            633
+          ],
+          "label": "reference",
+          "reading_order": 11,
+          "text": "Boyack, K.W., & Klavans, R. (2008). Measuring science–technology inter-\naction using rare inventor–author names. Journal of Informetrics, 2,\n173–182."
+        },
+        {
+          "bbox": [
+            355,
+            633,
+            630,
+            656
+          ],
+          "label": "reference",
+          "reading_order": 12,
+          "text": "Boyack, K.W., Klavans, R., & Börner, K. (2005). Mapping the backbone of\nscience. Scientometrics, 64, 351-374."
+        },
+        {
+          "bbox": [
+            355,
+            657,
+            628,
+            679
+          ],
+          "label": "reference",
+          "reading_order": 13,
+          "text": "Breschi, S., Lissoni, F., & Malerba, F. (2003). Knowledge-relatedness in\nfirm technological diversification. Research Policy, 32, 69-87."
+        },
+        {
+          "bbox": [
+            355,
+            679,
+            628,
+            703
+          ],
+          "label": "reference",
+          "reading_order": 14,
+          "text": "Chen, C. (2003). Mapping scientific frontiers: The quest for knowledge\nvisualization. London: Springer."
+        },
+        {
+          "bbox": [
+            355,
+            703,
+            628,
+            736
+          ],
+          "label": "reference",
+          "reading_order": 15,
+          "text": "Etzkowitz, H., & Leydesdorff, L. (2000). The dynamics of innovation:\nFrom National Systems and “Mode 2” to a Triple Helix of university-\nindustry-government relations. Research Policy, 29, 109–123."
+        },
+        {
+          "bbox": [
+            355,
+            736,
+            628,
+            770
+          ],
+          "label": "reference",
+          "reading_order": 16,
+          "text": "Franz, J.S. (2009). Constructing technological distances from US patent\ndata. IEEE Xplore Online Proceedings, Atlanta Conference on Science,\nTechnology and Innovation."
+        },
+        {
+          "bbox": [
+            353,
+            789,
+            630,
+            825
+          ],
+          "label": "fnote",
+          "reading_order": 17,
+          "text": "5See, for example, http://www.mapofscience.com for an overlay of\npatents in the map of science carried out by Boyack and Klavans (unpub-\nlished). Accessed September 23, 2013."
+        },
+        {
+          "bbox": [
+            63,
+            846,
+            540,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 18,
+          "text": "2442 JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY-December 2014\nDOI: 10.1002/asi"
+        }
+      ]
+    },
+    {
+      "page_number": 12,
+      "elements": [
+        {
+          "bbox": [
+            63,
+            51,
+            339,
+            85
+          ],
+          "label": "reference",
+          "reading_order": 0,
+          "text": "Gibbons, M., Limoges, C., Nowotny, H., Schwartzman, S., Scott, P., &\nTrow, M. (1994). The new production of knowledge. The dynamics of\nscience and research in contemporary societies. London: SAGE."
+        },
+        {
+          "bbox": [
+            63,
+            85,
+            339,
+            165
+          ],
+          "label": "reference",
+          "reading_order": 1,
+          "text": "Hinze, S., Reiss, T., & Schmoch, U. (1997). Statistical analysis on the\ndistance between fields of technology. Paper presented at the Innovation\nSystems and European Integration (ISE), Targeted Socio-Economic\nResearch Program, 4th Framework Program of the European Commis-\nsion (DGXII), Karlsruhe, Germany. Retrieved from http://www.isi\n.fraunhofer.de/isi-media/docs/isi-publ/1997/isi97b81/technology-fields-\ndiastance.pdf?WSESSIONID=5712ff2ca5ffcf0d9590afc8ef7e1486."
+        },
+        {
+          "bbox": [
+            63,
+            165,
+            339,
+            198
+          ],
+          "label": "reference",
+          "reading_order": 2,
+          "text": "Jaffe, A. (1986). Technological opportunities and spillovers of R&D Evi-\ndence from firms' patents, profits, and market value. American Eco-\nnomic Review, 76(5), 984–1001."
+        },
+        {
+          "bbox": [
+            63,
+            198,
+            339,
+            232
+          ],
+          "label": "reference",
+          "reading_order": 3,
+          "text": "Janssens, F., Zhang, L., Moor, B.D., & Glänzel, W. (2009). Hybrid clus-\ntering for validation and improvement of subject-classification schemes.\nInformation Processing & Management, 45(6), 683-702."
+        },
+        {
+          "bbox": [
+            63,
+            232,
+            339,
+            265
+          ],
+          "label": "reference",
+          "reading_order": 4,
+          "text": "Kauffman, S., Lobo, J., & Macready, W.G. (2000). Optimal search on a\ntechnology landscape. Journal of Economic Behaviour and Organization,\n43, 141-166."
+        },
+        {
+          "bbox": [
+            63,
+            265,
+            339,
+            301
+          ],
+          "label": "reference",
+          "reading_order": 5,
+          "text": "Klavans, R., & Boyack, K.W. (2009). Toward a consensus map of science.\nJournal of the American Society for Information Science and Technol-\nogy, 60(3), 455-476."
+        },
+        {
+          "bbox": [
+            63,
+            301,
+            339,
+            334
+          ],
+          "label": "reference",
+          "reading_order": 6,
+          "text": "Leydesdorff, L., Kushnir, D., & Rafols, I. (2012). Interactive overlay maps\nfor US patent (USPTO) data based on International Patent Classification\n(IPC). Scientometrics. DOI: 10.1007/s11192-012-0923-2."
+        },
+        {
+          "bbox": [
+            63,
+            334,
+            339,
+            369
+          ],
+          "label": "reference",
+          "reading_order": 7,
+          "text": "Leydesdorff, L., & Rafols, I. (2009). A global map of science based on the\nISI subject categories. Journal of the American Society for Information\nScience and Technology, 60(2), 348-362."
+        },
+        {
+          "bbox": [
+            63,
+            369,
+            339,
+            413
+          ],
+          "label": "reference",
+          "reading_order": 8,
+          "text": "Moya-Anegón, F., Vargas-Quesada, B., Chinchilla-Rodríguez, Z.,\nCorera-Álvarez, E., & Herrero-Solana, V. (2007). Visualizing the\nmarrow of science. Journal of the American Society for Information\nScience and Technology, 58(14), 2167–2179."
+        },
+        {
+          "bbox": [
+            63,
+            413,
+            339,
+            458
+          ],
+          "label": "reference",
+          "reading_order": 9,
+          "text": "Moya-Anegon, F., Vargas-Quesada, B., Herrero-Solana, V., Chinchilla-\nRodriguez, Z., Corera-Alvarez, E., & Munoz-Fernandez, F.J. (2004). A\nnew technique for building maps of large scientific domains based on the\ncocitation of classes and categories. Scientometrics, 61(1), 129-145."
+        },
+        {
+          "bbox": [
+            63,
+            458,
+            339,
+            481
+          ],
+          "label": "reference",
+          "reading_order": 10,
+          "text": "Olsson, O. (2004). Technological opportunity and growth. Journal of Eco-\nnomic Growth, 10(1), 35-57."
+        },
+        {
+          "bbox": [
+            355,
+            51,
+            630,
+            74
+          ],
+          "label": "reference",
+          "reading_order": 11,
+          "text": "Porter, A.L., & Youtie, J. (2009). Where does nanotechnology belong in the\nmap of science? Nature Nanotechnology, 4, 534-536."
+        },
+        {
+          "bbox": [
+            355,
+            74,
+            630,
+            119
+          ],
+          "label": "reference",
+          "reading_order": 12,
+          "text": "Rafols, I., & L., Leydesdorff, L. (2009). Content-based and algorithmic\nclassifications of journals: Perspectives on the dynamics of scientific\ncommunication and indexer effects. Journal of the American Society for\nInformation Science and Technology, 60(9), 1823-1835."
+        },
+        {
+          "bbox": [
+            355,
+            119,
+            630,
+            153
+          ],
+          "label": "reference",
+          "reading_order": 13,
+          "text": "Rafols, I., & Meyer, M. (2010). Diversity and network coherence as indi-\ncators of interdisciplinarity: Case studies in bionanoscience. Scientomet-\nrics, 82(2), 263–287."
+        },
+        {
+          "bbox": [
+            355,
+            153,
+            630,
+            197
+          ],
+          "label": "reference",
+          "reading_order": 14,
+          "text": "Rafols, I., Porter, A.L., & Leydesdorff, L. (2010). Science overlay maps: A\nnew tool for research policy and library management. Journal of the\nAmerican Society for Information Science and Technology, 61(9), 1871-\n1887."
+        },
+        {
+          "bbox": [
+            355,
+            198,
+            630,
+            221
+          ],
+          "label": "reference",
+          "reading_order": 15,
+          "text": "Rosvall, M., & Bergstrom, C.T. (2010). Mapping change in large networks.\nPLoS ONE, 5(1), e8694."
+        },
+        {
+          "bbox": [
+            355,
+            221,
+            630,
+            266
+          ],
+          "label": "reference",
+          "reading_order": 16,
+          "text": "Schoen, A., Villard, L., Laurens, P., Cointet, J.-P., Heimeriks, G., &\nAlkemade, F. (2012). The network structure of technological develop-\nments: Technological distance as a walk on the technology map. Pre-\nsented at the STI Indicators Conference 2012, Montréal, Canada."
+        },
+        {
+          "bbox": [
+            355,
+            266,
+            630,
+            301
+          ],
+          "label": "reference",
+          "reading_order": 17,
+          "text": "Small, H. (1973). Citing-to-cited in the scientific literature: A new measure\nof the relationship between two documents. Journal of the American\nSociety for Information Science and Technology, 24(4), 265-269."
+        },
+        {
+          "bbox": [
+            355,
+            301,
+            630,
+            346
+          ],
+          "label": "reference",
+          "reading_order": 18,
+          "text": "Van den Besselaar, P., & Leydesdorff, L. (1996). Mapping change in sci-\nentific specialties: A scientometric reconstruction of the development of\nartificial intelligence. Journal of the American Society for Information\nScience and Technology, 46(6), 415-436."
+        },
+        {
+          "bbox": [
+            355,
+            346,
+            630,
+            391
+          ],
+          "label": "reference",
+          "reading_order": 19,
+          "text": "Waltman, L., & van Eck, N.J. (2012). A new methodology for constructing\na publication-level classification system of science. Journal of the Ameri-\ncan Society for Information Science and Technology, 63(12), 2378–\n2392."
+        },
+        {
+          "bbox": [
+            354,
+            412,
+            485,
+            429
+          ],
+          "label": "sec_1",
+          "reading_order": 20,
+          "text": "Supporting Information"
+        },
+        {
+          "bbox": [
+            354,
+            433,
+            631,
+            462
+          ],
+          "label": "para",
+          "reading_order": 21,
+          "text": "Additional Supporting Information may be found in the\nonline version of this article at the publisher's web-site:"
+        },
+        {
+          "bbox": [
+            354,
+            467,
+            529,
+            483
+          ],
+          "label": "para",
+          "reading_order": 22,
+          "text": "Appendix. Supplementary Materials."
+        },
+        {
+          "bbox": [
+            152,
+            846,
+            630,
+            868
+          ],
+          "label": "foot",
+          "reading_order": 23,
+          "text": "JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY—December 2014\n2443\nDOI: 10.1002/asi"
+        }
+      ]
+    }
+  ],
+  "metadata": {}
+}

processed/00056_W3124826676/document.md

@@ -0,0 +1,345 @@
+U.S. Patent and Trademark Office
+
+
+OFFICE OF CHIEF ECONOMIST
+
+
+Economic Working Paper Series
+
+What is the Probability of Receiving a US Patent?
+
+Michael Carley, Economist Deepak Hegde, Assistant Professor Alan Marco, Economist
+
+USPTO Economic Working Paper No. 2013-2 Original version: December 2013 Current version: January 2014
+
+The views expressed are those of the individual authors and do not necessarily reflect official positions of the Office of Chief Economist or the U. S. Patent and Trademark Office. USPTO Economic Working Papers are preliminary research being shared in a timely manner with the public in order to stimulate discussion, scholarly debate, and critical comment. For more information about the USPTO’s Office of Chief Economist, visit www.uspto.gov/economics.
+
+UNITED STATES PATENT AND TRADEMARK OFFICE
+
+Uspto
+
+Electronic copy available at: http://ssrn.com/abstract-2849631
+
+---
+
+# What is the Probability of Receiving a US Patent?*
+
+Michael Carley† Deepak Hegde‡,§ Alan Marco**
+
+January 10, 2013
+
+## Abstract
+
+We follow the prosecution histories of the 2.15 million new patent applications filed at the US Patent and Trademark Office between 1996 and 2005 to calculate patent allowance rates. 55.8 % of the applications emerged as patents without using continuation procedures to spawn related applications. The success rate of applications decreased substantially from 1996 to 2005, particularly for applications in the “Drugs and Medical Instruments” and “Computers and Communications” fields. Applications filed by large firms are more likely to emerge as patents than those filed by small firms. We discuss the implications of our findings for inventors, policy makers, and social scientists who use successful patent applications as indicators of innovation.
+
+USPTO Economics Working Paper No. 2013-2. Available at: http://ssrn.com/abstract=2367149 Keywords: JEL Classification Numbers: O34
+
+* Authors are listed in alphabetical order. The authors thank Paul D'Agostino, Charles Eloshway, Kira Fabrizio, Dan Hunter, Jenny Kuan, Joanne Oxley, Gregory Mills, Binta Robinson, Patrick Ross, Robert Seamans, and Arvids Ziedonis for helpful discussions and suggestions. A.M. is Acting Chief Economist, M.C. is Economist, and D.H. is a visiting scholar at the USPTO. D.H.'s research was funded by the 20122013 Kauffman Faculty Fellowship. The analysis and recommendations in this manuscript, however, are solely the authors' and do not necessarily reflect the views of either the USPTO or the Kauffman Foundation.
+
+1 United States Patent and Trademark Office
+
+* New York University
+
+$^{3}$ Corresponding author; email: dhegde@stern.nyu.edu
+
+** United States Patent and Trademark Office
+
+2
+
+Electronic copy available at: http://ssrn.com/abstract-2849631
+
+---
+
+## 1 Introduction
+
+Inventors choose among different appropriability mechanisms, such as patents, copyrights, trademarks, and trade-secrecy, to protect their inventions based on their relative costs and benefits (Cohen, et al 2000) . A key element of the inventors' costbenefit calculus is the expectation that their patent application will succeed. However, little information exists on the historical rates at which patent applications are granted in the US. This paucity of information about the probability of getting patents impairs inventors' decisions regarding their choice of appropriability mechanisms; it also afflicts policy debates on the rigor of patent examination and abuses of the US patent system (see National Academy of Sciences 2001, Jaffe and Lerner 2004, Bessen and Meurer 2008) . For example, writing with economist Gary Becker, Judge Posner recently opined that “the problem of patent trolls is a function in part of the promiscuity with which the patent office has issued patents...” (Posner 2013) .
+
+The calculation of patent allowance rates, while simple on the surface, is complicated by several aspects of the patent examination process. First, US patent applications that are rejected after examination by the patent office can spawn closely related but “new” applications (called “continuations”) that are hard to track but may finally emerge as patents. Second, the US patent office publishes information on the outcomes of examination only for the applications that are published (after patent grant for applications filed before November 29, 2000 and after 18-months from application date for applications filed on or after November 29, 2000 that are still pending at 18 months with some exceptions, See 35 USC 122). Third, applicants alter and narrow the claims of their applications during the examination process. Thus, the allowance of some patentable claims within an application is not the same as the allowance of an application as it was filed, and should be taken into account in any discussion of allowance rates.
+
+What is the probability that a patent application filed at the US Patent and Trademark Office (PTO, “the agency,” or “the office”) emerges as a patent? Our objective here is to establish some facts related to this question by analyzing unique application-level data available internally at the PTO. The data tracks each of the 2.15 million new utility patent applications filed at the PTO between 1996 and 2005. These applications represent the population of “progenitor applications,” that is, applications unrelated to any previously filed US applications. We track the applications from the date they entered the Office through June 30, 2013, by which time 99.8 % of the progenitor applications had exited the system as a patent or were abandoned. This allows us to link each progenitor application to its children applications (subsequent applications spawned by the progenitor applications through the use of various continuation procedures) and to accurately estimate the probability of allowance without the limitations of previous studies based on
+
+3
+
+Electronic copy available at: http://ssrn.com/abstract-2849631
+
+---
+
+partial samples of published applications ( e.g. , Lemley and Sampat 2008) or exit cohorts ( e.g. , Quillen and Webster 2001, 2009) . 1,2
+
+In order to capture the complexity of the examination process, we calculate three measures of patent allowance rates: (i) first action allowance rate , or the proportion of progenitor applications that are allowed without further examination; (ii) progenitor allowance rate (or simply, allowance rate), or the proportion of progenitor applications that are allowed and patented without using any continuation procedure, and (iii) family allowance rate , or the proportion of progenitor applications that produce at least one patent, including the outcomes of applications that emerge from the progenitors through the use of continuation procedures.
+
+## 2 The US patent examination process
+
+We simplify our description of the patent examination procedures and rules, and discuss only the most salient events relevant to our objective. $^3$ Accordingly, Figure 1 presents a stylized version of the US patent examination process, using data for the 1996-2005 filing cohorts of progenitor applications. Each application is queued for examination when the application is docketed to an examiner. Applications that are incomplete or not accompanied by the appropriate fees within the grace period are considered abandoned and not docketed to an examiner. The first significant correspondence that an applicant receives from the office is called a “first action on the merits” (or simply “first action”). The first action includes a search report with a listing of relevant prior art that supports the examiner's decision of either allowance or non-final rejection. The office allowed 11.4 % of the progenitor applications at first action and delivered a non-final rejection decision for 86.4 % of the applications, with the remaining 2.3 % being abandoned prior to first action. 36.1 % of the progenitor applications were allowed after one or more rounds of amendments and negotiations with the examiner, but prior to a final rejection. 14.5 % abandoned between non-final rejection and final rejection. 38.7 % received a final rejection.
+
+1 Data on unpublished applications are not made available to the public to protect the intellectual property of patent applicants who may choose to abandon their applications prior to 18-month publication date. If unpublished applications are more likely to be abandoned, allowance rates calculated based on publicly available data (i.e., published applications) will be biased upwards. However, we will make available on request, detailed instructions on how to obtain data on published patent applications.
+
+2 The careful work of Lemley and Sampat (2008) attempts to overcome some of these problems by tracking a sample of applications through the Patent Application Information Retrieval (PAIR) system, but their analysis is based on a sample of published applications. The small sample of 9,960 applications filed in January 2001 they examine also limits their study's scope.
+
+3 The USPTO's official patent application filing guide provides a more comprehensive description of the rules and procedures. See http://www.uspto.gov/patents/resources/types/utility.jsp
+
+4
+
+---
+
+Figure 1 here
+
+For most applications, prosecution at the office ends with patent allowance (and patent issue) or with abandonment. Applications are considered abandoned if the applicant does not respond to the examiner's decision within the stipulated deadlines, or if the applicant expressly requests abandonment. Hence, there is no formally decisive rejection from the Office—only abandonments that result from applicants' actions and non-actions. Applicants can continue to submit amended applications with additional material to persuade the examiner even after receiving a final rejection. 9.3\% of applications received a final rejection and were allowed after the applicant responded with after final amendments and supporting material. Further, applicants can formally appeal a final rejection by submitting an appeal to the erstwhile Board of Patent Appeals and Interferences.7.5\% (2.9/38.7) of the final rejections were subject to appeals and 41.4\% (1.2/2.9) of the appeals resulted in allowances. 2.7\% of applications were abandoned after allowance; thus, overall, 55.8\% of progenitor applications emerged as patents without the use of continuation procedures.
+
+## Continuation procedures
+
+Applicants can continue prosecution after receiving a final rejection (or even after they receive an allowance), by using various continuation procedures at the PTO. Some scholars have blamed the procedures for several abuses of the patent system including submarine patents, long pending patent applications, and low-quality patents ( e.g. , Lemley and Moore 2004) , while others have pointed out that they may help applicants revise their applications to reflect the developments to their inventions while the applications are under examination (Hegde, et al, 2009) .
+
+31 % of the progenitor applications utilized some form of continuation procedure. These continuation procedures at the office can be of two broad types: non-serialized continuations and serialized continuations. 4 Non-serialized continuations do not receive a new serial number and are immediately docked to the same examiner that prosecuted the progenitor (the progenitor application is counted as abandoned in many official statistics and examiner production metrics). Requests for Continued Examination (RCEs, instituted in 1999) are by far the most common type of non-serialized continuations and applicants may file an RCE multiple times during prosecution. 5 19.5 % of the progenitors
+
+4 Serialized continuations can be exercised at any point during patent examination. Non-serialized continuations may only be used after particular events in prosecution-typically after final rejection.
+
+5 There have been several incarnations of non-serialized continuations, including Continued Prosecution Applications (CPAs), Rule 129 continuations (R129s), and File Wrapper Continuations (FWCs). The most recent incarnation (and by far the most prevalent) is the Request for Continued Examination (RCEs). Throughout this section, we refer collectively to all these non-serialized continuations as RCEs. Until
+
+5
+
+---
+
+filed at least one RCE. Of the applications that moved from non-final rejection to final rejection, 38.7 % filed an RCE. Thus, if one includes allowances of the non-serialized continuation applications that emerged from the progenitors, the allowance rate jumps from 55.8 % to 69.2 % .
+
+In contrast to non-serialized continuations, serialized continuations are treated as new applications; they receive a new serial number and are docketed to examiners based on the classification of the new application. There are three types of serialized continuations. Applicants may file for a simple continuation (CON) of a parent application to receive the benefit of the parent's priority date so long as the CON limits itself to the specification described in the parent. Applicants can use the “Continuation-In-Part” (CIP) to introduce new subject matter to an existing application. Alternatively, if two or more independent and distinct inventions are claimed in one application, the Office may require the application to be restricted to one of the inventions, and the applicant may file a divisional (DIV) application. Serialized continuations receive the priority date of the progenitor, with the exception of new matter added in CIPs, so long as the progenitor is pending when the serialized continuation is filed. The progenitor does not have to be abandoned following a serialized continuation. The parent and child may proceed through the examination process in parallel, and a single progenitor can produce a chain of several serialized continuations resulting in multiple patents, thus complicating the calculation of allowance rates for progenitors. 15.8 % of the progenitor applications gave birth to at least one serialized continuation as of June 30, 2013.Overall, 71.2 % of progenitor applications resulted in the issue of at least one patent after counting the allowances of applications emerging from (serialized and non-serialized) continuation procedures.
+
+## Figure 2 here
+
+Figure 2 plots the three allowance rates by the entry year of the progenitors. The figure shows that the probability of allowance is substantially lower for the recent cohort of applications. The striking decline in first action allowance rates and progenitor allowance rates is unlikely to be due to censoring since the mean first-action pendency for applications filed during the period was 21.1 months and total pendency was 29.1 months (first-action pendency refers to the time between application date and first-action date; total pendency refers to the time between application date and disposal date; Hegde 2012 reports pendency statistics at the PTO between 1991 and 2010). Although less than 1 % of the progenitor applications in our study were pending to date, a larger proportion
+
+November 2009, RCEs were put on the “amended docket,” which meant that the examiner had to respond within two months. Since that time, RCEs have gone on the “special new docket,” meaning that the examiner has more discretion as to when to respond (similar to newly docketed applications).
+
+6
+
+---
+
+of abandoned progenitors have continuation applications that are still pending, thus potentially biasing our family allowance rates downward for the later years. We account for this by calculating the maximum possible family allowance rate that would occur if all pending applications were to eventually issue. This upper bound is represented by dashed lines in Figure 2 . This correction demonstrates that for the 1996-2005 cohorts, the average family allowance rate could at most be 72.3 % (as compared to the rate of 71.2 % based on disposals observed to date) and the decline in allowance rates between 1996 and 2005 is quite robust. 6
+
+Our interviews with patent experts at the USPTO suggested at least three possible explanations for the decline. First, the financial market bust in March 2000 and the following period of tighter financial constraints may have forced some inventors to abandon their patent applications. Second, the PTO introduced several procedures in 2000 to augment the quality of patent prosecution (e.g., the “second pair of eyes” system which subjected certain applications to mandatory assessment by more than one examiner before being allowed), which may have decreased the probability of patent allowance. Finally, the number of pending applications, first-action pendency, and total pendency all steadily increased during the period of our study. Longer pendencies have been shown to be correlated with more abandonments, thus lowering the observed allowance rates (Mitra-Kahn, Marco et al 2013) . 7 Of course, establishing the causal effects of these and other potential influences on allowance rates is difficult, and requires separating out the effects of changes in the USPTO from changes in the propensity of applicants to abandon their applications. We thus defer a careful examination of the determinants of allowance rates for future research.
+
+## 3 Allowance rates across technology fields
+
+It is well known that patent value varies across industries (Cohen et al 2000) . Inventors in discrete-product industries, such as chemicals and pharmaceuticals, tend to use patents to preclude imitation by rivals, while those in complex product industries such as electronics and computers amass patents to enhance their bargaining power in cross-
+
+6 The effect of censoring is more pronounced for more recent cohorts, increasing sharply after 2005, thus validating our choice of 2005 as the cut-off year for our study. Figure A1 of the Supplementary Appendix presents the lower and upper bounds for each of the three allowance rates for 1991 to 2010. As the window between filing and observation shrinks, the observed allowance rates will fall to 0 % and the hypothetical maximum for each allowance rate will approach 100 % .
+
+7 Table A1 of the Supplementary Appendix presents the correlation between our allowance rate measures and the percent change in GDP from the previous year, the number of applications pending in the year of filing and the total pendency for applications disposed in the year of filing. All three allowance rates are strongly, negatively correlated with pendency and the number of pending applications and are moderately, positively correlated with the percent change in GDP.
+
+7
+
+---
+
+licensing negotiations (Hall and Ziedonis 2001). Inventors in different industries also appear to pursue different strategies during the patent examination process, including in their use of CONs (Hegde, et al 2008), and judicial decisions (for example, the State Street decision in 1998 to validate patenting of business method patents, or the recent Myriad decision invalidating patenting of DNA segments) affect the standards of patentability for some technological fields, while leaving the standards unchanged for others. $^8$
+
+Figure 3 here
+
+Figure 3 displays the probability of patent allowance for the patent technology categories defined by Hall, Jaffe, and Trajtenberg (2001) . 9 Applications in Drugs and Medical Instruments have the lowest average allowance rates (allowance rate of 42.8 % ) and applications in the Electrical and Electronics sectors enjoy the highest allowance rates (allowance rate of 66.6 % ). In the Computers and Communication sector, which includes a large majority of the controversial software and business method patents, allowance rates are relatively lower (allowance rate of 49.8 % ). Applicants appear to use continuation procedures more in the sectors with lower allowance rates (44.1 % of the progenitor applications used at least one of the continuation procedures in the Drugs and Medical sector; see Table A3 of the Supplementary Appendix). The decline in allowance rates is particularly striking for Drugs and Medical Instrument patents and for Computers and Communication patents (see also Graham and Vishnubhakat 2013) . In these sectors, both first-action allowance rate and progenitor allowance rates declined by more than 50 % (Figures A2-A4 of the Appendix compare sectoral trends for the three allowance rates).
+
+## 4 Allowance rates across inventor types
+
+Small inventors play an important role in the US innovation system and the Office seeks to lower their costs of patenting by charging discounted (50%-75%) examination fees. $^10$ Like small entities, foreign inventors may also find it difficult to access the legal resources required to enhance their chances of receiving patents. Does the probability of patent issue differ for different applicant types? To answer this, we identified patent
+
+$^{8}$ See 149 F.3d 1368 (Fed. Cir. 1998) and 569 U.S. 12-398 (2013), respectively.
+
+9 Hall, Jaffe, and Trajtenberg (2001) create a mapping from US Patent Classification (USPC) to six technology categories for issued utility patent applications. The data were updated in 2006. We apply the 2006 mapping to all progenitor applications in our dataset in order to treat abandoned and issued applications similarly. Child applications are assigned to the same technology category as the progenitor application.
+
+10 USPTO fees are discounted by 50% for applicants and patentees who qualify as “Small Entities” by having fewer than 500 employees (37 CFR 1.27). For exact patent examination fees, see: http://www.uspto.gov/web/offices/ac/qs/ope/fee031913.htm
+
+8
+
+---
+
+applications as belonging to foreign inventors if the primary inventor on the application was located abroad, and identified small-inventors as those that qualified for the USPTO's small-entity discounts. Large foreign inventors accounted for 39 % , large U.S. inventors 31.1 % , small foreign inventors 9.6 % , and small U.S. inventors 20.1 % of our 2.15 Million progenitor applications.
+
+## Figure 4 here
+
+Figure 4 reveals that large foreign inventors enjoy the highest progenitor and family allowance rates (60.5 % and 77 % respectively), followed by large US inventors (57 % and 75.2 % ). US small inventors have the lowest rates of patent allowance, particularly family allowance rates. Foreign applicants and small inventors are less likely to use continuation applications (Table A4 of the Supplementary Appendix reports the percentage of progenitor applications that used the different types of continuations by applicant type). The differences in allowance rates across the different applicant types appear more substantial in some fields (e.g. Computers and Communications) than others (Table A5 of the Supplementary Appendix reports the allowance rates for the different applicant types across technology fields).
+
+These numbers should be interpreted with caution. The lower allowance rates for US small applicants could reflect either their higher propensity to abandon their applications during the examination process, or differences in the nature of inventions and subject matter covered by their applications. Similarly, large foreign inventors may enjoy higher allowance rates either because they choose to seek protection in the US for their most important inventions, or because they are more likely to have access to the legal resources required to maximize the probability of allowance.
+
+## 5 Concluding thoughts
+
+Our analysis of progenitor applications filed between 1996 and 2005 uncovers several interesting facts that counter conventional wisdom. We find that the first action allowance rate for patent applications is just 11.4 % . Only 55.8 % of progenitor applications eventually emerge as patents after several rounds of amendments. The family allowance rate, which accounts for the use of continuation procedures by progenitor applications, is just 71.2 % . The probability of patent issue declined during the period of our study: starting at nearly 70 % for the applications filed in 1996, progenitor allowance rates declined to 40 % for the 2005 cohort (even accounting for censoring issues as shown in Figure A1). Applications in the “ Drugs and Medical Instruments ” fields are least likely to be successful and applications in the “ Electrical and Electronics ” fields are most likely to be successful. Allowance rates declined sharply for
+
+9
+
+---
+
+applications filed between 1996 and 2005 in the “Drugs and Medical Instruments” and “Computers and Communication” fields. Allowance rates are lower across all technology sectors for small firms.
+
+What are the implications of these findings? Many scholars have interpreted patent allowance rates, typically incorrectly calculated, as reflecting examination quality alone, and argued that the high allowance rates at the PTO indicate low examination quality ( e.g. , Quillen and Webster 2001, 2009) . Our findings challenge the conventional wisdom that the PTO allows nearly all of the applications it receives, and rubber stamps applications without scrutiny. We also find no evidence for the claims that the PTO is becoming more lenient in granting patents. To the extent that some inventors invest in preparing US patent applications, based on assumptions about the probability of being successful, our findings help correct their “priors,” and thus make more informed decisions about their investments.
+
+Scholars in economics and management widely use the number of successful patent applications as a proxy for the innovation intensity of firms, industries, and even nations. To the extent that at least some of these differences are shaped by systematic differences in the probability of patent allowance across types of inventors, technological fields, and time, as we have documented, scholars need to account for factors underpinning these differences before drawing conclusions about the rate of innovation based on simple counts of successful patent applications.
+
+Our study suggests that patent allowance rates should be interpreted with caution by policy makers. Allowance rates are the product of an “opt out” system for applicants: thus, the rates are driven not only by the office’s rejection of applications, but applicants’ willingness to continue the prosecution of their applications. Accordingly, the rates may reflect the influence of several variables including the patentability of the subject matter claimed in the applications (which varies across technological fields), the rigor of the patent examination process, the time taken for examination at the PTO, judicial decisions about valid subject matter, and applicants’ access to the resources required to keep their applications alive. Some of these variables could be uncorrelated with the rigor of the examination process. Hence, economists should investigate the factors underlying the fluctuations in allowance rates, and be aware of the infeasibility of defining an “optimal” allowance rate before recommending changes to the examination system based on observed rates. Just as having a lenient process that rubber stamps applications without scrutiny can impose costs on our innovation system, an allowance rate that is “too low” may deter inventors, particularly those that cannot engage in costly negotiations with patent examiners, from seeking patents, or worse still, investing in innovation.
+
+10
+
+---
+
+## References
+
+Bessen, J. & Meurer, M.J. 2008. Patent Failure: How Judges, Bureaucrats, and Lawyers Put Innovators at Risk (Princeton University Press, 2008).
+
+Cohen W. M., Nelson, R.R., & Walsh J.P., 2000. Protecting Their Intellectual Assets: Appropriability Conditions and Why US Manufacturing Firms Patent (or Not), National Bureau of Economic Research Working Paper W7552.
+
+Graham, S., & Vishnubhakat, S. 2013. Of Smart Phone Wars and Software Patents. Journal of Economic Perspectives 27 (1): 67-86.
+
+Hall, B. H., Jaffe, A. B., & Trajtenberg, M. 2001. The NBER Patent Citations Data File: Lessons, Insights), National Bureau of Economic Research Working Paper W8498.
+
+Hall, B. H. & Ziedonis. R. 2001. The Patent Paradox Revisited: An Empirical Study of Patenting in the US Semiconductor Industry, 1979–1995, RAND Journal of Economics, 32:1, p 101–28.
+
+Hegde, D., Mowery, D.C. & Graham, S.J. 2009. Pioneering inventors or thicket-builders: which firms use continuations in patenting?, Management Science 55, 1214–1226 (2009).
+
+Jaffe, A.B. & Lerner, J. 2004. Innovation and Its Discontents: How Our Broken Patent System is Endangering Innovation and Progress, and What to Do About It (Princeton University Press, 2004).
+
+Lemley, M.A. & Sampat, B.N. 2008. Is the Patent Office a Rubber Stamp? Emory Law J. 58, 415-427.
+
+Lemley M. A. & Moore K. 2004. Ending Abuse of Patent Continuations, Boston University Law Review, 84(1), p 63-123.
+
+Mitra-Kahn, B., Marco, A., et al., 2013, “Patent backlogs, inventories, and pendency: An international framework,” UK IPO & PTO joint report, http://www.ipo.gov.uk/ipresearch-uspatlog-201306.pdf
+
+11
+
+---
+
+National Academy of Sciences 2004 Committee on Intellectual Property Rights in the Knowledge-Based Economy, National Research Council. A Patent System for the 21st Century, National Academies Press, Washington, DC, 2004.
+
+Posner, Richard. 2013. Patent Trolls. The Becker-Posner Blog, dated 07/21/2013, Accessed from http://www.becker-posner-blog.com/2013/07/patenttrollsposner.html on 08/03/2013
+
+Quillen, C. D. & Webster. O. H. 2001. Continuing Patent Applications and Performance of the US Patent Office, Federal Circuit Bar Journal, 1, p 1-21.
+
+Quillen, C. D. & Webster. O. H. 2009. Continuing Patent Applications and Performance of the US Patent Office—One More Time, Federal Circuit Bar Journal, 18 (13), p 379-404.
+
+12
+
+---
+
+## Figures and Tables
+
+Figure 1: The US Patent Examination Process
+
+![Figure](figures/figure_002.png)
+
+Notes: Figure is a simplified representation of the US patent examination process and shows the key intermediate and final outcomes, as of June 30, 2013, for the 2.15 million applications filed for the first time (“progenitor” applications) at the PTO between 1996 and 2005. The percentage indicated at each transition-state reflects the percentage of the total progenitor applications that reached the state. First-action allowance rate refers to the proportion of progenitor applications that are allowed without amendment; Progenitor allowance rate refers to the proportion of progenitor applications that were eventually allowed and patented without using continuation processes; Family allowance rate refers to the proportion of progenitor applications that produce at least one patent, including the allowances of continuation applications that emerge from the progenitors. Abandonments and allowances may not sum to 100 % due to rounding.
+
+13
+
+---
+
+Figure 2: Trends in allowance rates, 1996-2005
+
+![Figure](figures/figure_001.png)
+
+Notes: Figure shows trends in the three types of allowance rates for the 2.15 million applications filed at the PTO for the first time between 1996 and 2005. 18,270 of the 2.15 million applications were pending as of June 30, 2013 and the dotted lines (for the first-action allowance rate and progenitor allowance rate) represent the corresponding rates if all the pending applications are, in fact, allowed. Thus, they represent the theoretical upper-bound for the allowance rates. For progenitor applications that produced continuation applications which are still pending, we calculate the maximum possible family allowance rate for each progenitor cohort by assuming that every pending continuation application produced by the progenitors will eventually be allowed. This maximum possible family allowance rate is represented by the corresponding dashed line.
+
+14
+
+---
+
+Figure 3: Allowance rates by patent technology fields (for patent applications filed between 1996 and 2005)
+
+![Figure](figures/figure_001.png)
+
+Notes: Figure shows the three types of allowance rates for applications filed at the PTO for the first time between 1996 and 2005, across the six NBER patent technology fields.
+
+15
+
+---
+
+Figure 4: Allowance rates by inventor type (for patent applications filed between 1996 and 2005)
+
+![Figure](figures/figure_001.png)
+
+Notes: Figure shows the three types of allowance rates for applications filed at the USPTO for the first time between 1996 and 2005, across the four inventor types.
+
+16
+
+---
+
+## Appendix. Supplementary statistics
+
+Table A1: Correlations between allowance rates and environmental covariates, 1996-2005
+
+<table><tr><td></td><td>(A)</td><td>(B)</td><td>(C)</td><td>(D)</td><td>(E)</td></tr><tr><td>(A) First Action Allowance Rate</td><td></td><td></td><td></td><td></td><td></td></tr><tr><td>(B) Progenitor Allowance Rate</td><td>0.949</td><td></td><td></td><td></td><td></td></tr><tr><td>(C) Family Allowance Rate</td><td>0.950</td><td>0.998</td><td></td><td></td><td></td></tr><tr><td>(D)Percent Change in Real GDP</td><td>0.352</td><td>0.482</td><td>0.515</td><td></td><td></td></tr><tr><td>(E)Total Pending Applications</td><td>-0.925</td><td>-0.994</td><td>-0.992</td><td>-0.505</td><td></td></tr><tr><td>(F)Total Pendency</td><td>-0.925</td><td>-0.967</td><td>-0.963</td><td>-0.349</td><td>0.971</td></tr></table>
+
+
+Note: Table shows contemporaneous correlations between allowance rates and potential environmental determinants of allowance rates (all variables are measured annually, for each year between 1996 and 2005). Total pending applications refer to the stock of patent applications filed, and in the examination process for the given year. Total pendency refers to the average time, in months, between patent application date and patent disposal date during the entry year of the progenitor applications in our study.
+
+17
+
+---
+
+Table A2: Progenitor applications and related continuation applications, 1996-2005
+
+<table><tr><td rowspan="2">Year</td><td rowspan="2">Applications</td><td colspan="4">Serialized Continuations</td><td rowspan="2">Non-serialized Continuations (RCEs)</td><td rowspan="2">Either Continuation</td></tr><tr><td>CON</td><td>CIP</td><td>DIV</td><td>Any</td></tr><tr><td>1996</td><td>146,260</td><td>6.9\%</td><td>5.6\%</td><td>6.5\%</td><td>17.7\%</td><td>11.2\%</td><td>24.9\%</td></tr><tr><td>1997</td><td>166,232</td><td>5.8\%</td><td>5.3\%</td><td>6.7\%</td><td>16.5\%</td><td>12.1\%</td><td>25.6\%</td></tr><tr><td>1998</td><td>182,717</td><td>6.3\%</td><td>5.0\%</td><td>6.8\%</td><td>16.9\%</td><td>13.4\%</td><td>26.9\%</td></tr><tr><td>1999</td><td>197,704</td><td>6.9\%</td><td>5.0\%</td><td>6.9\%</td><td>17.5\%</td><td>14.5\%</td><td>28.3\%</td></tr><tr><td>2000</td><td>222,480</td><td>7.1\%</td><td>4.8\%</td><td>6.5\%</td><td>17.2\%</td><td>15.7\%</td><td>29.0\%</td></tr><tr><td>2001</td><td>232,668</td><td>7.1\%</td><td>4.4\%</td><td>6.5\%</td><td>16.9\%</td><td>17.4\%</td><td>30.3\%</td></tr><tr><td>2002</td><td>233,246</td><td>6.7\%</td><td>4.4\%</td><td>6.1\%</td><td>16.1\%</td><td>19.7\%</td><td>31.5\%</td></tr><tr><td>2003</td><td>235,861</td><td>6.3\%</td><td>4.1\%</td><td>5.1\%</td><td>14.6\%</td><td>24.1\%</td><td>33.7\%</td></tr><tr><td>2004</td><td>250,338</td><td>6.3\%</td><td>3.4\%</td><td>4.9\%</td><td>13.7\%</td><td>27.3\%</td><td>35.6\%</td></tr><tr><td>2005</td><td>278,160</td><td>6.5\%</td><td>2.7\%</td><td>4.7\%</td><td>13.2\%</td><td>29.2\%</td><td>37.1\%</td></tr></table>
+
+
+Note: Table shows the number of progenitor applications filed in the corresponding year, and the percentage of the applications from each cohort that produced the different types of continuations.
+
+18
+
+---
+
+Table A3: The use of Continuation applications across technology fields, 1996-2005
+
+<table><tr><td rowspan="2">Technology Field</td><td rowspan="2">Applications</td><td colspan="4">Serialized Continuations</td><td rowspan="2">Non-serialized Continuations (RCEs)</td><td rowspan="2">Either Continuation</td></tr><tr><td>CON</td><td>CIP</td><td>DIV</td><td>Any</td></tr><tr><td>Chemical</td><td>245,150</td><td>6.0\%</td><td>5.3\%</td><td>9.2\%</td><td>19.1\%</td><td>18.2\%</td><td>32.8\%</td></tr><tr><td>Drugs \</td><td>Medical</td><td>227,936</td><td>12.8\%</td><td>8.2\%</td><td>10.0\%</td><td>28.2\%</td><td>24.5\%</td><td>44.1\%</td></tr><tr><td>Computers \</td><td>Comm.</td><td>611,046</td><td>8.3\%</td><td>3.2\%</td><td>3.6\%</td><td>14.1\%</td><td>26.7\%</td><td>36.0\%</td></tr><tr><td>Electrical \</td><td>Electronic</td><td>402,401</td><td>4.7\%</td><td>3.0\%</td><td>7.7\%</td><td>14.5\%</td><td>16.4\%</td><td>27.5\%</td></tr><tr><td>Mechanical</td><td>311,040</td><td>3.9\%</td><td>3.8\%</td><td>4.9\%</td><td>11.9\%</td><td>13.2\%</td><td>22.7\%</td></tr><tr><td>Others</td><td>348,093</td><td>4.6\%</td><td>5.2\%</td><td>4.2\%</td><td>13.2\%</td><td>13.4\%</td><td>23.7\%</td></tr></table>
+
+
+Note: Table shows the number of progenitor applications filed in each NBER patent technology field (between 1996 and 2005), and the percentage of the applications that produced the different types of continuations.
+
+19
+
+---
+
+Table A4: The use of Continuation applications across applicant types, 1996-2005
+
+<table><tr><td rowspan="2">Applicant Type</td><td rowspan="2">Applications</td><td colspan="4">Serialized Continuations</td><td rowspan="2">Non-serialized Continuations</td><td rowspan="2">Either Continuation</td></tr><tr><td>CON</td><td>CIP</td><td>DIV</td><td>Any</td></tr><tr><td>Large Foreign</td><td>838,210</td><td>4.4\%</td><td>1.3\%</td><td>5.9\%</td><td>11.2\%</td><td>21.1\%</td><td>29.1\%</td></tr><tr><td>Small Foreign</td><td>207,460</td><td>3.7\%</td><td>3.7\%</td><td>2.9\%</td><td>9.7\%</td><td>12.1\%</td><td>19.3\%</td></tr><tr><td>Large US</td><td>668,527</td><td>9.2\%</td><td>5.2\%</td><td>7.6\%</td><td>20.4\%</td><td>23.0\%</td><td>37.6\%</td></tr><tr><td>Small US</td><td>431,469</td><td>8.2\%</td><td>9.2\%</td><td>5.0\%</td><td>20.5\%</td><td>14.3\%</td><td>30.0\%</td></tr></table>
+
+
+Note: Table shows the number of progenitor applications filed by each applicant type (between 1996 and 2005), and the percentage of each type's applications that produced the different types of continuations.
+
+20
+
+---
+
+Table A5: Allowance Rates across applicant types and technology fields, 1996-2005
+
+<table><tr><td>Technology Field</td><td>Applicant Type</td><td>Applications</td><td>First Action</td><td>Progenitor</td><td>Family</td></tr><tr><td rowspan="4">Chemical</td><td>Large Foreign</td><td>112,598</td><td>11.0\%</td><td>59.6\%</td><td>75.4\%</td></tr><tr><td></td><td>Large US</td><td>76,595</td><td>11.3\%</td><td>57.2\%</td><td>74.1\%</td></tr><tr><td></td><td>Small Foreign</td><td>20,245</td><td>11.6\%</td><td>52.9\%</td><td>64.4\%</td></tr><tr><td></td><td>Small US</td><td>35,712</td><td>9.7\%</td><td>52.4\%</td><td>65.8\%</td></tr><tr><td rowspan="4">Computers \&amp; Commun.</td><td>Large Foreign</td><td>244,453</td><td>11.7\%</td><td>54.5\%</td><td>74.0\%</td></tr><tr><td></td><td>Large US</td><td>251,253</td><td>8.9\%</td><td>51.8\%</td><td>74.1\%</td></tr><tr><td></td><td>Small Foreign</td><td>32,847</td><td>9.6\%</td><td>37.7\%</td><td>48.9\%</td></tr><tr><td></td><td>Small US</td><td>82,493</td><td>6.4\%</td><td>34.5\%</td><td>49.6\%</td></tr><tr><td rowspan="4">Drugs \&amp; Medical</td><td>Large Foreign</td><td>62,142</td><td>5.3\%</td><td>45.0\%</td><td>63.6\%</td></tr><tr><td></td><td>Large US</td><td>69,632</td><td>6.0\%</td><td>43.1\%</td><td>62.7\%</td></tr><tr><td></td><td>Small Foreign</td><td>27,372</td><td>5.7\%</td><td>39.9\%</td><td>55.4\%</td></tr><tr><td></td><td>Small US</td><td>68,790</td><td>5.6\%</td><td>41.5\%</td><td>58.3\%</td></tr><tr><td rowspan="4">Electrical \&amp; Electronics</td><td>Large Foreign</td><td>204,125</td><td>15.5\%</td><td>67.7\%</td><td>83.3\%</td></tr><tr><td></td><td>Large US</td><td>122,529</td><td>14.2\%</td><td>69.3\%</td><td>84.5\%</td></tr><tr><td></td><td>Small Foreign</td><td>30,489</td><td>17.0\%</td><td>57.7\%</td><td>65.2\%</td></tr><tr><td></td><td>Small US</td><td>45,258</td><td>13.1\%</td><td>60.0\%</td><td>71.1\%</td></tr><tr><td rowspan="4">Mechanical</td><td>Large Foreign</td><td>128,328</td><td>15.1\%</td><td>68.8\%</td><td>82.1\%</td></tr><tr><td></td><td>Large US</td><td>74,681</td><td>14.1\%</td><td>67.2\%</td><td>80.5\%</td></tr><tr><td></td><td>Small Foreign</td><td>40,274</td><td>15.8\%</td><td>56.2\%</td><td>63.7\%</td></tr><tr><td></td><td>Small US</td><td>67,757</td><td>12.0\%</td><td>57.1\%</td><td>65.9\%</td></tr><tr><td rowspan="4">Others</td><td>Large Foreign</td><td>86,564</td><td>11.3\%</td><td>60.7\%</td><td>74.6\%</td></tr><tr><td></td><td>Large US</td><td>73,837</td><td>9.9\%</td><td>56.5\%</td><td>71.9\%</td></tr><tr><td></td><td>Small Foreign</td><td>56,233</td><td>13.5\%</td><td>51.1\%</td><td>57.7\%</td></tr><tr><td></td><td>Small US</td><td>131,459</td><td>9.5\%</td><td>49.3\%</td><td>57.4\%</td></tr></table>
+
+
+Note: Table shows the number of progenitor applications filed in each of the six NBER patent technology fields by each applicant type (between 1996 and 2005), and the percentage of each type's applications that produced the different types of continuations.
+
+21
+
+---
+
+Figure A1: Trends in allowance rates with adjustments for censoring, 1991-2010
+
+![Figure](figures/figure_001.png)
+
+Notes: Figure shows trends in the three types of allowance rates for the 4.2 million applications filed at the PTO for the first time between 1991 and 2010. A significant number of applications filed after 2005 were pending as of June 30, 2013 and the dotted lines (for the first-action allowance rate and progenitor allowance rate) represent the corresponding rates if all the pending applications are, in fact, allowed. Thus, they represent the theoretical upperbound for the allowance rates (a vast majority of the applications filed for the first time in 2010 were past the first action, but still pending at the office: if all of these pending applications were to issue, then the progenitor allowance rate for 2008 applications would be around 68 % ). For progenitor applications that produced continuation applications that are still pending, we calculate the maximum possible family allowance rate by assuming that every pending continuation application produced by the progenitors will eventually be allowed. This maximum possible family allowance rate is represented by the corresponding dashed line.
+
+22
+
+---
+
+Figure A2: Trends in First Action Allowance Rate by Technology Field, 1996-2005
+
+![Figure](figures/figure_001.png)
+
+Figure A3: Trends in Progenitor Allowance Rate by Technology Field, 1996-2005
+
+![Figure](figures/figure_003.png)
+
+23
+
+---
+
+Figure A4: Trends in Family Allowance Rate by Technology Field, 1996-2005
+
+![Figure](figures/figure_001.png)
+
+24
+

processed/00056_W3124826676/metadata.json

@@ -0,0 +1,1708 @@
+{
+  "source_file": "/tmp/tmpajgyhya6/00056_W3124826676_What_Is_the_Probability_of_Receiving_a_US_Patent.pdf",
+  "total_pages": 24,
+  "pages": [
+    {
+      "page_number": 1,
+      "elements": [
+        {
+          "bbox": [
+            97,
+            82,
+            310,
+            137
+          ],
+          "label": "para",
+          "reading_order": 0,
+          "text": "U.S. Patent and Trademark Office\n\n\nOFFICE OF CHIEF ECONOMIST\n\n\nEconomic Working Paper Series"
+        },
+        {
+          "bbox": [
+            133,
+            265,
+            578,
+            294
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "What is the Probability of Receiving a US Patent?"
+        },
+        {
+          "bbox": [
+            235,
+            411,
+            458,
+            462
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "Michael Carley, Economist\nDeepak Hegde, Assistant Professor\nAlan Marco, Economist"
+        },
+        {
+          "bbox": [
+            220,
+            605,
+            492,
+            660
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "USPTO Economic Working Paper No. 2013-2\nOriginal version: December 2013\nCurrent version: January 2014"
+        },
+        {
+          "bbox": [
+            78,
+            676,
+            615,
+            740
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "The views expressed are those of the individual authors and do not necessarily reflect official positions of the Office of\nChief Economist or the U. S. Patent and Trademark Office. USPTO Economic Working Papers are preliminary research\nbeing shared in a timely manner with the public in order to stimulate discussion, scholarly debate, and critical comment.\nFor more information about the USPTO’s Office of Chief Economist, visit www.uspto.gov/economics."
+        },
+        {
+          "bbox": [
+            524,
+            788,
+            642,
+            850
+          ],
+          "label": "foot",
+          "reading_order": 5,
+          "text": "UNITED STATES\nPATENT AND TRADEMARK OFFICE\n\nUspto"
+        },
+        {
+          "bbox": [
+            176,
+            876,
+            518,
+            893
+          ],
+          "label": "foot",
+          "reading_order": 6,
+          "text": "Electronic copy available at: http://ssrn.com/abstract-2849631"
+        }
+      ]
+    },
+    {
+      "page_number": 2,
+      "elements": [
+        {
+          "bbox": [
+            135,
+            121,
+            555,
+            148
+          ],
+          "label": "sec_0",
+          "reading_order": 0,
+          "text": "What is the Probability of Receiving a US Patent?*"
+        },
+        {
+          "bbox": [
+            300,
+            163,
+            390,
+            214
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "Michael Carley†\nDeepak Hegde‡,§\nAlan Marco**"
+        },
+        {
+          "bbox": [
+            298,
+            230,
+            395,
+            248
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "January 10, 2013"
+        },
+        {
+          "bbox": [
+            320,
+            280,
+            372,
+            296
+          ],
+          "label": "sec_1",
+          "reading_order": 3,
+          "text": "Abstract"
+        },
+        {
+          "bbox": [
+            119,
+            307,
+            574,
+            439
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "We follow the prosecution histories of the 2.15 million new patent applications filed at\nthe US Patent and Trademark Office between 1996 and 2005 to calculate patent\nallowance rates. 55.8 % of the applications emerged as patents without using continuation\nprocedures to spawn related applications. The success rate of applications decreased\nsubstantially from 1996 to 2005, particularly for applications in the “Drugs and Medical\nInstruments” and “Computers and Communications” fields. Applications filed by large\nfirms are more likely to emerge as patents than those filed by small firms. We discuss the\nimplications of our findings for inventors, policy makers, and social scientists who use\nsuccessful patent applications as indicators of innovation."
+        },
+        {
+          "bbox": [
+            119,
+            452,
+            363,
+            520
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "USPTO Economics Working Paper No. 2013-2.\nAvailable at: http://ssrn.com/abstract=2367149\nKeywords:\nJEL Classification Numbers: O34"
+        },
+        {
+          "bbox": [
+            119,
+            655,
+            571,
+            728
+          ],
+          "label": "fnote",
+          "reading_order": 6,
+          "text": "* Authors are listed in alphabetical order. The authors thank Paul D'Agostino, Charles Eloshway, Kira\nFabrizio, Dan Hunter, Jenny Kuan, Joanne Oxley, Gregory Mills, Binta Robinson, Patrick Ross, Robert\nSeamans, and Arvids Ziedonis for helpful discussions and suggestions. A.M. is Acting Chief Economist,\nM.C. is Economist, and D.H. is a visiting scholar at the USPTO. D.H.'s research was funded by the 2012-\n2013 Kauffman Faculty Fellowship. The analysis and recommendations in this manuscript, however, are\nsolely the authors' and do not necessarily reflect the views of either the USPTO or the Kauffman Foundation."
+        },
+        {
+          "bbox": [
+            119,
+            728,
+            306,
+            739
+          ],
+          "label": "fnote",
+          "reading_order": 7,
+          "text": "1 United States Patent and Trademark Office"
+        },
+        {
+          "bbox": [
+            120,
+            739,
+            218,
+            751
+          ],
+          "label": "fnote",
+          "reading_order": 8,
+          "text": "* New York University"
+        },
+        {
+          "bbox": [
+            120,
+            751,
+            345,
+            763
+          ],
+          "label": "fnote",
+          "reading_order": 9,
+          "text": "$^{3}$ Corresponding author; email: dhegde@stern.nyu.edu"
+        },
+        {
+          "bbox": [
+            120,
+            763,
+            309,
+            774
+          ],
+          "label": "fnote",
+          "reading_order": 10,
+          "text": "** United States Patent and Trademark Office"
+        },
+        {
+          "bbox": [
+            340,
+            840,
+            351,
+            855
+          ],
+          "label": "foot",
+          "reading_order": 11,
+          "text": "2"
+        },
+        {
+          "bbox": [
+            176,
+            876,
+            518,
+            893
+          ],
+          "label": "foot",
+          "reading_order": 12,
+          "text": "Electronic copy available at: http://ssrn.com/abstract-2849631"
+        }
+      ]
+    },
+    {
+      "page_number": 3,
+      "elements": [
+        {
+          "bbox": [
+            119,
+            138,
+            251,
+            157
+          ],
+          "label": "sec_1",
+          "reading_order": 0,
+          "text": "1 Introduction"
+        },
+        {
+          "bbox": [
+            119,
+            158,
+            572,
+            357
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "Inventors choose among different appropriability mechanisms, such as patents,\ncopyrights, trademarks, and trade-secrecy, to protect their inventions based on their\nrelative costs and benefits (Cohen, et al 2000) . A key element of the inventors' cost-\nbenefit calculus is the expectation that their patent application will succeed. However,\nlittle information exists on the historical rates at which patent applications are granted in\nthe US. This paucity of information about the probability of getting patents impairs\ninventors' decisions regarding their choice of appropriability mechanisms; it also afflicts\npolicy debates on the rigor of patent examination and abuses of the US patent system (see\nNational Academy of Sciences 2001, Jaffe and Lerner 2004, Bessen and Meurer 2008) .\nFor example, writing with economist Gary Becker, Judge Posner recently opined that\n“the problem of patent trolls is a function in part of the promiscuity with which the patent\noffice has issued patents...” (Posner 2013) ."
+        },
+        {
+          "bbox": [
+            119,
+            366,
+            569,
+            565
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "The calculation of patent allowance rates, while simple on the surface, is complicated by\nseveral aspects of the patent examination process. First, US patent applications that are\nrejected after examination by the patent office can spawn closely related but “new”\napplications (called “continuations”) that are hard to track but may finally emerge as\npatents. Second, the US patent office publishes information on the outcomes of\nexamination only for the applications that are published (after patent grant for\napplications filed before November 29, 2000 and after 18-months from application date\nfor applications filed on or after November 29, 2000 that are still pending at 18 months\nwith some exceptions, See 35 USC 122). Third, applicants alter and narrow the claims of\ntheir applications during the examination process. Thus, the allowance of some\npatentable claims within an application is not the same as the allowance of an application\nas it was filed, and should be taken into account in any discussion of allowance rates."
+        },
+        {
+          "bbox": [
+            119,
+            575,
+            572,
+            774
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "What is the probability that a patent application filed at the US Patent and Trademark\nOffice (PTO, “the agency,” or “the office”) emerges as a patent? Our objective here is to\nestablish some facts related to this question by analyzing unique application-level data\navailable internally at the PTO. The data tracks each of the 2.15 million new utility patent\napplications filed at the PTO between 1996 and 2005. These applications represent the\npopulation of “progenitor applications,” that is, applications unrelated to any previously\nfiled US applications. We track the applications from the date they entered the Office\nthrough June 30, 2013, by which time 99.8 % of the progenitor applications had exited the\nsystem as a patent or were abandoned. This allows us to link each progenitor application\nto its children applications (subsequent applications spawned by the progenitor\napplications through the use of various continuation procedures) and to accurately\nestimate the probability of allowance without the limitations of previous studies based on"
+        },
+        {
+          "bbox": [
+            340,
+            840,
+            351,
+            856
+          ],
+          "label": "foot",
+          "reading_order": 4,
+          "text": "3"
+        },
+        {
+          "bbox": [
+            176,
+            876,
+            519,
+            893
+          ],
+          "label": "foot",
+          "reading_order": 5,
+          "text": "Electronic copy available at: http://ssrn.com/abstract-2849631"
+        }
+      ]
+    },
+    {
+      "page_number": 4,
+      "elements": [
+        {
+          "bbox": [
+            119,
+            122,
+            566,
+            155
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "partial samples of published applications ( e.g. , Lemley and Sampat 2008) or exit cohorts\n( e.g. , Quillen and Webster 2001, 2009) . 1,2"
+        },
+        {
+          "bbox": [
+            119,
+            167,
+            564,
+            298
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "In order to capture the complexity of the examination process, we calculate three\nmeasures of patent allowance rates: (i) first action allowance rate , or the proportion of\nprogenitor applications that are allowed without further examination; (ii) progenitor\nallowance rate (or simply, allowance rate), or the proportion of progenitor applications\nthat are allowed and patented without using any continuation procedure, and (iii) family\nallowance rate , or the proportion of progenitor applications that produce at least one\npatent, including the outcomes of applications that emerge from the progenitors through\nthe use of continuation procedures."
+        },
+        {
+          "bbox": [
+            119,
+            324,
+            422,
+            346
+          ],
+          "label": "sec_1",
+          "reading_order": 2,
+          "text": "2 The US patent examination process"
+        },
+        {
+          "bbox": [
+            119,
+            347,
+            573,
+            609
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "We simplify our description of the patent examination procedures and rules, and discuss\nonly the most salient events relevant to our objective. $^3$ Accordingly, Figure 1 presents a\nstylized version of the US patent examination process, using data for the 1996-2005 filing\ncohorts of progenitor applications. Each application is queued for examination when the\napplication is docketed to an examiner. Applications that are incomplete or not\naccompanied by the appropriate fees within the grace period are considered abandoned\nand not docketed to an examiner. The first significant correspondence that an applicant\nreceives from the office is called a “first action on the merits” (or simply “first action”).\nThe first action includes a search report with a listing of relevant prior art that supports\nthe examiner's decision of either allowance or non-final rejection. The office allowed\n11.4 % of the progenitor applications at first action and delivered a non-final rejection\ndecision for 86.4 % of the applications, with the remaining 2.3 % being abandoned prior to\nfirst action. 36.1 % of the progenitor applications were allowed after one or more rounds\nof amendments and negotiations with the examiner, but prior to a final rejection. 14.5 %\nabandoned between non-final rejection and final rejection. 38.7 % received a final\nrejection."
+        },
+        {
+          "bbox": [
+            120,
+            635,
+            566,
+            700
+          ],
+          "label": "fnote",
+          "reading_order": 4,
+          "text": "1 Data on unpublished applications are not made available to the public to protect the intellectual property of\npatent applicants who may choose to abandon their applications prior to 18-month publication date. If\nunpublished applications are more likely to be abandoned, allowance rates calculated based on publicly\navailable data (i.e., published applications) will be biased upwards. However, we will make available on\nrequest, detailed instructions on how to obtain data on published patent applications."
+        },
+        {
+          "bbox": [
+            120,
+            700,
+            570,
+            749
+          ],
+          "label": "fnote",
+          "reading_order": 5,
+          "text": "2 The careful work of Lemley and Sampat (2008) attempts to overcome some of these problems by tracking a\nsample of applications through the Patent Application Information Retrieval (PAIR) system, but their\nanalysis is based on a sample of published applications. The small sample of 9,960 applications filed in\nJanuary 2001 they examine also limits their study's scope."
+        },
+        {
+          "bbox": [
+            120,
+            749,
+            552,
+            775
+          ],
+          "label": "fnote",
+          "reading_order": 6,
+          "text": "3 The USPTO's official patent application filing guide provides a more comprehensive description of the\nrules and procedures. See http://www.uspto.gov/patents/resources/types/utility.jsp"
+        },
+        {
+          "bbox": [
+            340,
+            827,
+            351,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 7,
+          "text": "4"
+        }
+      ]
+    },
+    {
+      "page_number": 5,
+      "elements": [
+        {
+          "bbox": [
+            308,
+            121,
+            385,
+            139
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Figure 1 here"
+        },
+        {
+          "bbox": [
+            119,
+            151,
+            573,
+            382
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "For most applications, prosecution at the office ends with patent allowance (and patent\nissue) or with abandonment. Applications are considered abandoned if the applicant does\nnot respond to the examiner's decision within the stipulated deadlines, or if the applicant\nexpressly requests abandonment. Hence, there is no formally decisive rejection from the\nOffice—only abandonments that result from applicants' actions and non-actions.\nApplicants can continue to submit amended applications with additional material to\npersuade the examiner even after receiving a final rejection. 9.3\\% of applications\nreceived a final rejection and were allowed after the applicant responded with after final\namendments and supporting material. Further, applicants can formally appeal a final\nrejection by submitting an appeal to the erstwhile Board of Patent Appeals and\nInterferences.7.5\\% (2.9/38.7) of the final rejections were subject to appeals and 41.4\\%\n(1.2/2.9) of the appeals resulted in allowances. 2.7\\% of applications were abandoned after\nallowance; thus, overall, 55.8\\% of progenitor applications emerged as patents without the\nuse of continuation procedures."
+        },
+        {
+          "bbox": [
+            120,
+            393,
+            264,
+            411
+          ],
+          "label": "sec_1",
+          "reading_order": 2,
+          "text": "Continuation procedures"
+        },
+        {
+          "bbox": [
+            119,
+            425,
+            558,
+            541
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "Applicants can continue prosecution after receiving a final rejection (or even after they\nreceive an allowance), by using various continuation procedures at the PTO. Some\nscholars have blamed the procedures for several abuses of the patent system including\nsubmarine patents, long pending patent applications, and low-quality patents ( e.g. ,\nLemley and Moore 2004) , while others have pointed out that they may help applicants\nrevise their applications to reflect the developments to their inventions while the\napplications are under examination (Hegde, et al, 2009) ."
+        },
+        {
+          "bbox": [
+            119,
+            551,
+            570,
+            685
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "31 % of the progenitor applications utilized some form of continuation procedure. These\ncontinuation procedures at the office can be of two broad types: non-serialized\ncontinuations and serialized continuations. 4 Non-serialized continuations do not receive\na new serial number and are immediately docked to the same examiner that prosecuted\nthe progenitor (the progenitor application is counted as abandoned in many official\nstatistics and examiner production metrics). Requests for Continued Examination (RCEs,\ninstituted in 1999) are by far the most common type of non-serialized continuations and\napplicants may file an RCE multiple times during prosecution. 5 19.5 % of the progenitors"
+        },
+        {
+          "bbox": [
+            120,
+            697,
+            532,
+            726
+          ],
+          "label": "fnote",
+          "reading_order": 5,
+          "text": "4 Serialized continuations can be exercised at any point during patent examination. Non-serialized\ncontinuations may only be used after particular events in prosecution-typically after final rejection."
+        },
+        {
+          "bbox": [
+            120,
+            726,
+            559,
+            775
+          ],
+          "label": "fnote",
+          "reading_order": 6,
+          "text": "5 There have been several incarnations of non-serialized continuations, including Continued Prosecution\nApplications (CPAs), Rule 129 continuations (R129s), and File Wrapper Continuations (FWCs). The most\nrecent incarnation (and by far the most prevalent) is the Request for Continued Examination (RCEs).\nThroughout this section, we refer collectively to all these non-serialized continuations as RCEs. Until"
+        },
+        {
+          "bbox": [
+            340,
+            826,
+            351,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 7,
+          "text": "5"
+        }
+      ]
+    },
+    {
+      "page_number": 6,
+      "elements": [
+        {
+          "bbox": [
+            119,
+            122,
+            555,
+            187
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "filed at least one RCE. Of the applications that moved from non-final rejection to final\nrejection, 38.7 % filed an RCE. Thus, if one includes allowances of the non-serialized\ncontinuation applications that emerged from the progenitors, the allowance rate jumps\nfrom 55.8 % to 69.2 % ."
+        },
+        {
+          "bbox": [
+            119,
+            199,
+            573,
+            528
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "In contrast to non-serialized continuations, serialized continuations are treated as new\napplications; they receive a new serial number and are docketed to examiners based on\nthe classification of the new application. There are three types of serialized continuations.\nApplicants may file for a simple continuation (CON) of a parent application to receive\nthe benefit of the parent's priority date so long as the CON limits itself to the\nspecification described in the parent. Applicants can use the “Continuation-In-Part” (CIP)\nto introduce new subject matter to an existing application. Alternatively, if two or more\nindependent and distinct inventions are claimed in one application, the Office may\nrequire the application to be restricted to one of the inventions, and the applicant may file\na divisional (DIV) application. Serialized continuations receive the priority date of the\nprogenitor, with the exception of new matter added in CIPs, so long as the progenitor is\npending when the serialized continuation is filed. The progenitor does not have to be\nabandoned following a serialized continuation. The parent and child may proceed through\nthe examination process in parallel, and a single progenitor can produce a chain of\nseveral serialized continuations resulting in multiple patents, thus complicating the\ncalculation of allowance rates for progenitors. 15.8 % of the progenitor applications gave\nbirth to at least one serialized continuation as of June 30, 2013.Overall, 71.2 % of\nprogenitor applications resulted in the issue of at least one patent after counting the\nallowances of applications emerging from (serialized and non-serialized) continuation\nprocedures."
+        },
+        {
+          "bbox": [
+            308,
+            538,
+            385,
+            557
+          ],
+          "label": "sec_1",
+          "reading_order": 2,
+          "text": "Figure 2 here"
+        },
+        {
+          "bbox": [
+            119,
+            569,
+            569,
+            719
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "Figure 2 plots the three allowance rates by the entry year of the progenitors. The figure\nshows that the probability of allowance is substantially lower for the recent cohort of\napplications. The striking decline in first action allowance rates and progenitor\nallowance rates is unlikely to be due to censoring since the mean first-action pendency\nfor applications filed during the period was 21.1 months and total pendency was 29.1\nmonths (first-action pendency refers to the time between application date and first-action\ndate; total pendency refers to the time between application date and disposal date; Hegde\n2012 reports pendency statistics at the PTO between 1991 and 2010). Although less than\n1 % of the progenitor applications in our study were pending to date, a larger proportion"
+        },
+        {
+          "bbox": [
+            120,
+            738,
+            570,
+            776
+          ],
+          "label": "foot",
+          "reading_order": 4,
+          "text": "November 2009, RCEs were put on the “amended docket,” which meant that the examiner had to respond\nwithin two months. Since that time, RCEs have gone on the “special new docket,” meaning that the examiner\nhas more discretion as to when to respond (similar to newly docketed applications)."
+        },
+        {
+          "bbox": [
+            340,
+            826,
+            352,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 5,
+          "text": "6"
+        }
+      ]
+    },
+    {
+      "page_number": 7,
+      "elements": [
+        {
+          "bbox": [
+            119,
+            122,
+            572,
+            255
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "of abandoned progenitors have continuation applications that are still pending, thus\npotentially biasing our family allowance rates downward for the later years. We account\nfor this by calculating the maximum possible family allowance rate that would occur if\nall pending applications were to eventually issue. This upper bound is represented by\ndashed lines in Figure 2 . This correction demonstrates that for the 1996-2005 cohorts, the\naverage family allowance rate could at most be 72.3 % (as compared to the rate of 71.2 %\nbased on disposals observed to date) and the decline in allowance rates between 1996 and\n2005 is quite robust. 6"
+        },
+        {
+          "bbox": [
+            119,
+            264,
+            572,
+            512
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "Our interviews with patent experts at the USPTO suggested at least three possible\nexplanations for the decline. First, the financial market bust in March 2000 and the\nfollowing period of tighter financial constraints may have forced some inventors to\nabandon their patent applications. Second, the PTO introduced several procedures in\n2000 to augment the quality of patent prosecution (e.g., the “second pair of eyes” system\nwhich subjected certain applications to mandatory assessment by more than one examiner\nbefore being allowed), which may have decreased the probability of patent allowance.\nFinally, the number of pending applications, first-action pendency, and total pendency all\nsteadily increased during the period of our study. Longer pendencies have been shown to\nbe correlated with more abandonments, thus lowering the observed allowance rates\n(Mitra-Kahn, Marco et al 2013) . 7 Of course, establishing the causal effects of these and\nother potential influences on allowance rates is difficult, and requires separating out the\neffects of changes in the USPTO from changes in the propensity of applicants to abandon\ntheir applications. We thus defer a careful examination of the determinants of allowance\nrates for future research."
+        },
+        {
+          "bbox": [
+            119,
+            538,
+            436,
+            559
+          ],
+          "label": "sec_1",
+          "reading_order": 2,
+          "text": "3 Allowance rates across technology fields"
+        },
+        {
+          "bbox": [
+            119,
+            560,
+            572,
+            626
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "It is well known that patent value varies across industries (Cohen et al 2000) . Inventors in\ndiscrete-product industries, such as chemicals and pharmaceuticals, tend to use patents to\npreclude imitation by rivals, while those in complex product industries such as\nelectronics and computers amass patents to enhance their bargaining power in cross-"
+        },
+        {
+          "bbox": [
+            120,
+            650,
+            564,
+            713
+          ],
+          "label": "fnote",
+          "reading_order": 4,
+          "text": "6 The effect of censoring is more pronounced for more recent cohorts, increasing sharply after 2005, thus\nvalidating our choice of 2005 as the cut-off year for our study. Figure A1 of the Supplementary Appendix\npresents the lower and upper bounds for each of the three allowance rates for 1991 to 2010. As the window\nbetween filing and observation shrinks, the observed allowance rates will fall to 0 % and the hypothetical\nmaximum for each allowance rate will approach 100 % ."
+        },
+        {
+          "bbox": [
+            120,
+            713,
+            572,
+            775
+          ],
+          "label": "fnote",
+          "reading_order": 5,
+          "text": "7 Table A1 of the Supplementary Appendix presents the correlation between our allowance rate measures and\nthe percent change in GDP from the previous year, the number of applications pending in the year of filing\nand the total pendency for applications disposed in the year of filing. All three allowance rates are strongly,\nnegatively correlated with pendency and the number of pending applications and are moderately, positively\ncorrelated with the percent change in GDP."
+        },
+        {
+          "bbox": [
+            340,
+            826,
+            351,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 6,
+          "text": "7"
+        }
+      ]
+    },
+    {
+      "page_number": 8,
+      "elements": [
+        {
+          "bbox": [
+            119,
+            122,
+            563,
+            237
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "licensing negotiations (Hall and Ziedonis 2001). Inventors in different industries also\nappear to pursue different strategies during the patent examination process, including in\ntheir use of CONs (Hegde, et al 2008), and judicial decisions (for example, the State\nStreet decision in 1998 to validate patenting of business method patents, or the recent\nMyriad decision invalidating patenting of DNA segments) affect the standards of\npatentability for some technological fields, while leaving the standards unchanged for\nothers. $^8$"
+        },
+        {
+          "bbox": [
+            308,
+            247,
+            385,
+            264
+          ],
+          "label": "cap",
+          "reading_order": 1,
+          "text": "Figure 3 here"
+        },
+        {
+          "bbox": [
+            119,
+            278,
+            570,
+            525
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "Figure 3 displays the probability of patent allowance for the patent technology categories\ndefined by Hall, Jaffe, and Trajtenberg (2001) . 9 Applications in Drugs and Medical\nInstruments have the lowest average allowance rates (allowance rate of 42.8 % ) and\napplications in the Electrical and Electronics sectors enjoy the highest allowance rates\n(allowance rate of 66.6 % ). In the Computers and Communication sector, which includes\na large majority of the controversial software and business method patents, allowance\nrates are relatively lower (allowance rate of 49.8 % ). Applicants appear to use\ncontinuation procedures more in the sectors with lower allowance rates (44.1 % of the\nprogenitor applications used at least one of the continuation procedures in the Drugs and\nMedical sector; see Table A3 of the Supplementary Appendix). The decline in allowance\nrates is particularly striking for Drugs and Medical Instrument patents and for Computers\nand Communication patents (see also Graham and Vishnubhakat 2013) . In these sectors,\nboth first-action allowance rate and progenitor allowance rates declined by more than\n50 % (Figures A2-A4 of the Appendix compare sectoral trends for the three allowance\nrates)."
+        },
+        {
+          "bbox": [
+            119,
+            552,
+            420,
+            573
+          ],
+          "label": "sec_1",
+          "reading_order": 3,
+          "text": "4 Allowance rates across inventor types"
+        },
+        {
+          "bbox": [
+            119,
+            573,
+            568,
+            657
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "Small inventors play an important role in the US innovation system and the Office seeks\nto lower their costs of patenting by charging discounted (50%-75%) examination fees. $^10$\nLike small entities, foreign inventors may also find it difficult to access the legal\nresources required to enhance their chances of receiving patents. Does the probability of\npatent issue differ for different applicant types? To answer this, we identified patent"
+        },
+        {
+          "bbox": [
+            120,
+            668,
+            448,
+            685
+          ],
+          "label": "fnote",
+          "reading_order": 5,
+          "text": "$^{8}$ See 149 F.3d 1368 (Fed. Cir. 1998) and 569 U.S. 12-398 (2013), respectively."
+        },
+        {
+          "bbox": [
+            120,
+            685,
+            569,
+            735
+          ],
+          "label": "fnote",
+          "reading_order": 6,
+          "text": "9 Hall, Jaffe, and Trajtenberg (2001) create a mapping from US Patent Classification (USPC) to six\ntechnology categories for issued utility patent applications. The data were updated in 2006. We apply the\n2006 mapping to all progenitor applications in our dataset in order to treat abandoned and issued applications\nsimilarly. Child applications are assigned to the same technology category as the progenitor application."
+        },
+        {
+          "bbox": [
+            120,
+            735,
+            572,
+            775
+          ],
+          "label": "fnote",
+          "reading_order": 7,
+          "text": "10 USPTO fees are discounted by 50% for applicants and patentees who qualify as “Small Entities” by having\nfewer than 500 employees (37 CFR 1.27). For exact patent examination fees, see:\nhttp://www.uspto.gov/web/offices/ac/qs/ope/fee031913.htm"
+        },
+        {
+          "bbox": [
+            340,
+            826,
+            351,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 8,
+          "text": "8"
+        }
+      ]
+    },
+    {
+      "page_number": 9,
+      "elements": [
+        {
+          "bbox": [
+            119,
+            122,
+            563,
+            204
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "applications as belonging to foreign inventors if the primary inventor on the application\nwas located abroad, and identified small-inventors as those that qualified for the\nUSPTO's small-entity discounts. Large foreign inventors accounted for 39 % , large U.S.\ninventors 31.1 % , small foreign inventors 9.6 % , and small U.S. inventors 20.1 % of our\n2.15 Million progenitor applications."
+        },
+        {
+          "bbox": [
+            308,
+            215,
+            385,
+            232
+          ],
+          "label": "sec_1",
+          "reading_order": 1,
+          "text": "Figure 4 here"
+        },
+        {
+          "bbox": [
+            119,
+            245,
+            570,
+            410
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "Figure 4 reveals that large foreign inventors enjoy the highest progenitor and family\nallowance rates (60.5 % and 77 % respectively), followed by large US inventors (57 % and\n75.2 % ). US small inventors have the lowest rates of patent allowance, particularly family\nallowance rates. Foreign applicants and small inventors are less likely to use\ncontinuation applications (Table A4 of the Supplementary Appendix reports the\npercentage of progenitor applications that used the different types of continuations by\napplicant type). The differences in allowance rates across the different applicant types\nappear more substantial in some fields (e.g. Computers and Communications) than others\n(Table A5 of the Supplementary Appendix reports the allowance rates for the different\napplicant types across technology fields)."
+        },
+        {
+          "bbox": [
+            119,
+            421,
+            569,
+            537
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "These numbers should be interpreted with caution. The lower allowance rates for US\nsmall applicants could reflect either their higher propensity to abandon their applications\nduring the examination process, or differences in the nature of inventions and subject\nmatter covered by their applications. Similarly, large foreign inventors may enjoy higher\nallowance rates either because they choose to seek protection in the US for their most\nimportant inventions, or because they are more likely to have access to the legal\nresources required to maximize the probability of allowance."
+        },
+        {
+          "bbox": [
+            119,
+            563,
+            307,
+            583
+          ],
+          "label": "sec_1",
+          "reading_order": 4,
+          "text": "5 Concluding thoughts"
+        },
+        {
+          "bbox": [
+            119,
+            584,
+            571,
+            766
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "Our analysis of progenitor applications filed between 1996 and 2005 uncovers several\ninteresting facts that counter conventional wisdom. We find that the first action\nallowance rate for patent applications is just 11.4 % . Only 55.8 % of progenitor\napplications eventually emerge as patents after several rounds of amendments. The\nfamily allowance rate, which accounts for the use of continuation procedures by\nprogenitor applications, is just 71.2 % . The probability of patent issue declined during\nthe period of our study: starting at nearly 70 % for the applications filed in 1996,\nprogenitor allowance rates declined to 40 % for the 2005 cohort (even accounting for\ncensoring issues as shown in Figure A1). Applications in the “ Drugs and Medical\nInstruments ” fields are least likely to be successful and applications in the “ Electrical and\nElectronics ” fields are most likely to be successful. Allowance rates declined sharply for"
+        },
+        {
+          "bbox": [
+            340,
+            826,
+            351,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 6,
+          "text": "9"
+        }
+      ]
+    },
+    {
+      "page_number": 10,
+      "elements": [
+        {
+          "bbox": [
+            119,
+            122,
+            557,
+            172
+          ],
+          "label": "half_para",
+          "reading_order": 0,
+          "text": "applications filed between 1996 and 2005 in the “Drugs and Medical Instruments” and\n“Computers and Communication” fields. Allowance rates are lower across all\ntechnology sectors for small firms."
+        },
+        {
+          "bbox": [
+            119,
+            182,
+            571,
+            346
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "What are the implications of these findings? Many scholars have interpreted patent\nallowance rates, typically incorrectly calculated, as reflecting examination quality alone,\nand argued that the high allowance rates at the PTO indicate low examination quality\n( e.g. , Quillen and Webster 2001, 2009) . Our findings challenge the conventional wisdom\nthat the PTO allows nearly all of the applications it receives, and rubber stamps\napplications without scrutiny. We also find no evidence for the claims that the PTO is\nbecoming more lenient in granting patents. To the extent that some inventors invest in\npreparing US patent applications, based on assumptions about the probability of being\nsuccessful, our findings help correct their “priors,” and thus make more informed\ndecisions about their investments."
+        },
+        {
+          "bbox": [
+            119,
+            358,
+            568,
+            474
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "Scholars in economics and management widely use the number of successful patent\napplications as a proxy for the innovation intensity of firms, industries, and even nations.\nTo the extent that at least some of these differences are shaped by systematic differences\nin the probability of patent allowance across types of inventors, technological fields, and\ntime, as we have documented, scholars need to account for factors underpinning these\ndifferences before drawing conclusions about the rate of innovation based on simple\ncounts of successful patent applications."
+        },
+        {
+          "bbox": [
+            119,
+            485,
+            569,
+            750
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "Our study suggests that patent allowance rates should be interpreted with caution by\npolicy makers. Allowance rates are the product of an “opt out” system for applicants:\nthus, the rates are driven not only by the office’s rejection of applications, but applicants’\nwillingness to continue the prosecution of their applications. Accordingly, the rates may\nreflect the influence of several variables including the patentability of the subject matter\nclaimed in the applications (which varies across technological fields), the rigor of the\npatent examination process, the time taken for examination at the PTO, judicial decisions\nabout valid subject matter, and applicants’ access to the resources required to keep their\napplications alive. Some of these variables could be uncorrelated with the rigor of the\nexamination process. Hence, economists should investigate the factors underlying the\nfluctuations in allowance rates, and be aware of the infeasibility of defining an “optimal”\nallowance rate before recommending changes to the examination system based on\nobserved rates. Just as having a lenient process that rubber stamps applications without\nscrutiny can impose costs on our innovation system, an allowance rate that is “too low”\nmay deter inventors, particularly those that cannot engage in costly negotiations with\npatent examiners, from seeking patents, or worse still, investing in innovation."
+        },
+        {
+          "bbox": [
+            338,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 4,
+          "text": "10"
+        }
+      ]
+    },
+    {
+      "page_number": 11,
+      "elements": [
+        {
+          "bbox": [
+            119,
+            122,
+            199,
+            142
+          ],
+          "label": "sec_1",
+          "reading_order": 0,
+          "text": "References"
+        },
+        {
+          "bbox": [
+            120,
+            144,
+            568,
+            183
+          ],
+          "label": "reference",
+          "reading_order": 1,
+          "text": "Bessen, J. & Meurer, M.J. 2008. Patent Failure: How Judges, Bureaucrats, and Lawyers\nPut Innovators at Risk (Princeton University Press, 2008)."
+        },
+        {
+          "bbox": [
+            120,
+            188,
+            551,
+            251
+          ],
+          "label": "reference",
+          "reading_order": 2,
+          "text": "Cohen W. M., Nelson, R.R., & Walsh J.P., 2000. Protecting Their Intellectual Assets:\nAppropriability Conditions and Why US Manufacturing Firms Patent (or Not),\nNational Bureau of Economic Research Working Paper W7552."
+        },
+        {
+          "bbox": [
+            120,
+            257,
+            540,
+            297
+          ],
+          "label": "reference",
+          "reading_order": 3,
+          "text": "Graham, S., & Vishnubhakat, S. 2013. Of Smart Phone Wars and Software Patents.\nJournal of Economic Perspectives 27 (1): 67-86."
+        },
+        {
+          "bbox": [
+            120,
+            304,
+            564,
+            343
+          ],
+          "label": "reference",
+          "reading_order": 4,
+          "text": "Hall, B. H., Jaffe, A. B., & Trajtenberg, M. 2001. The NBER Patent Citations Data File:\nLessons, Insights), National Bureau of Economic Research Working Paper W8498."
+        },
+        {
+          "bbox": [
+            120,
+            349,
+            558,
+            412
+          ],
+          "label": "reference",
+          "reading_order": 5,
+          "text": "Hall, B. H. & Ziedonis. R. 2001. The Patent Paradox Revisited: An Empirical Study of\nPatenting in the US Semiconductor Industry, 1979–1995, RAND Journal of\nEconomics, 32:1, p 101–28."
+        },
+        {
+          "bbox": [
+            120,
+            418,
+            569,
+            480
+          ],
+          "label": "reference",
+          "reading_order": 6,
+          "text": "Hegde, D., Mowery, D.C. & Graham, S.J. 2009. Pioneering inventors or thicket-builders:\nwhich firms use continuations in patenting?, Management Science 55, 1214–1226\n(2009)."
+        },
+        {
+          "bbox": [
+            119,
+            487,
+            558,
+            548
+          ],
+          "label": "reference",
+          "reading_order": 7,
+          "text": "Jaffe, A.B. & Lerner, J. 2004. Innovation and Its Discontents: How Our Broken Patent\nSystem is Endangering Innovation and Progress, and What to Do About It\n(Princeton University Press, 2004)."
+        },
+        {
+          "bbox": [
+            120,
+            556,
+            569,
+            594
+          ],
+          "label": "reference",
+          "reading_order": 8,
+          "text": "Lemley, M.A. & Sampat, B.N. 2008. Is the Patent Office a Rubber Stamp? Emory Law J.\n58, 415-427."
+        },
+        {
+          "bbox": [
+            120,
+            601,
+            525,
+            641
+          ],
+          "label": "reference",
+          "reading_order": 9,
+          "text": "Lemley M. A. & Moore K. 2004. Ending Abuse of Patent Continuations, Boston\nUniversity Law Review, 84(1), p 63-123."
+        },
+        {
+          "bbox": [
+            120,
+            648,
+            568,
+            710
+          ],
+          "label": "reference",
+          "reading_order": 10,
+          "text": "Mitra-Kahn, B., Marco, A., et al., 2013, “Patent backlogs, inventories, and pendency: An\ninternational framework,” UK IPO & PTO joint report,\nhttp://www.ipo.gov.uk/ipresearch-uspatlog-201306.pdf"
+        },
+        {
+          "bbox": [
+            338,
+            826,
+            354,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 11,
+          "text": "11"
+        }
+      ]
+    },
+    {
+      "page_number": 12,
+      "elements": [
+        {
+          "bbox": [
+            119,
+            124,
+            553,
+            184
+          ],
+          "label": "reference",
+          "reading_order": 0,
+          "text": "National Academy of Sciences 2004 Committee on Intellectual Property Rights in the\nKnowledge-Based Economy, National Research Council. A Patent System for the\n21st Century, National Academies Press, Washington, DC, 2004."
+        },
+        {
+          "bbox": [
+            120,
+            191,
+            529,
+            252
+          ],
+          "label": "reference",
+          "reading_order": 1,
+          "text": "Posner, Richard. 2013. Patent Trolls. The Becker-Posner Blog, dated 07/21/2013,\nAccessed from http://www.becker-posner-blog.com/2013/07/patent-\ntrollsposner.html on 08/03/2013"
+        },
+        {
+          "bbox": [
+            120,
+            261,
+            563,
+            298
+          ],
+          "label": "reference",
+          "reading_order": 2,
+          "text": "Quillen, C. D. & Webster. O. H. 2001. Continuing Patent Applications and Performance\nof the US Patent Office, Federal Circuit Bar Journal, 1, p 1-21."
+        },
+        {
+          "bbox": [
+            121,
+            306,
+            563,
+            365
+          ],
+          "label": "reference",
+          "reading_order": 3,
+          "text": "Quillen, C. D. & Webster. O. H. 2009. Continuing Patent Applications and Performance\nof the US Patent Office—One More Time, Federal Circuit Bar Journal, 18 (13), p\n379-404."
+        },
+        {
+          "bbox": [
+            338,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 4,
+          "text": "12"
+        }
+      ]
+    },
+    {
+      "page_number": 13,
+      "elements": [
+        {
+          "bbox": [
+            79,
+            61,
+            214,
+            82
+          ],
+          "label": "sec_1",
+          "reading_order": 0,
+          "text": "Figures and Tables"
+        },
+        {
+          "bbox": [
+            79,
+            93,
+            334,
+            111
+          ],
+          "label": "cap",
+          "reading_order": 1,
+          "text": "Figure 1: The US Patent Examination Process"
+        },
+        {
+          "bbox": [
+            136,
+            120,
+            555,
+            696
+          ],
+          "label": "fig",
+          "reading_order": 2,
+          "text": "![Figure](figures/figure_002.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_002.png"
+        },
+        {
+          "bbox": [
+            79,
+            705,
+            614,
+            824
+          ],
+          "label": "anno",
+          "reading_order": 3,
+          "text": "Notes: Figure is a simplified representation of the US patent examination process and shows the key intermediate\nand final outcomes, as of June 30, 2013, for the 2.15 million applications filed for the first time (“progenitor”\napplications) at the PTO between 1996 and 2005. The percentage indicated at each transition-state reflects the\npercentage of the total progenitor applications that reached the state. First-action allowance rate refers to the\nproportion of progenitor applications that are allowed without amendment; Progenitor allowance rate refers to the\nproportion of progenitor applications that were eventually allowed and patented without using continuation\nprocesses; Family allowance rate refers to the proportion of progenitor applications that produce at least one patent,\nincluding the allowances of continuation applications that emerge from the progenitors. Abandonments and\nallowances may not sum to 100 % due to rounding."
+        },
+        {
+          "bbox": [
+            338,
+            840,
+            355,
+            856
+          ],
+          "label": "foot",
+          "reading_order": 4,
+          "text": "13"
+        }
+      ]
+    },
+    {
+      "page_number": 14,
+      "elements": [
+        {
+          "bbox": [
+            79,
+            87,
+            341,
+            106
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Figure 2: Trends in allowance rates, 1996-2005"
+        },
+        {
+          "bbox": [
+            125,
+            147,
+            560,
+            419
+          ],
+          "label": "fig",
+          "reading_order": 1,
+          "text": "![Figure](figures/figure_001.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_001.png"
+        },
+        {
+          "bbox": [
+            79,
+            435,
+            615,
+            556
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Notes: Figure shows trends in the three types of allowance rates for the 2.15 million applications filed at the PTO\nfor the first time between 1996 and 2005. 18,270 of the 2.15 million applications were pending as of June 30, 2013\nand the dotted lines (for the first-action allowance rate and progenitor allowance rate) represent the corresponding\nrates if all the pending applications are, in fact, allowed. Thus, they represent the theoretical upper-bound for the\nallowance rates. For progenitor applications that produced continuation applications which are still pending, we\ncalculate the maximum possible family allowance rate for each progenitor cohort by assuming that every pending\ncontinuation application produced by the progenitors will eventually be allowed. This maximum possible family\nallowance rate is represented by the corresponding dashed line."
+        },
+        {
+          "bbox": [
+            339,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 3,
+          "text": "14"
+        }
+      ]
+    },
+    {
+      "page_number": 15,
+      "elements": [
+        {
+          "bbox": [
+            78,
+            61,
+            614,
+            97
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Figure 3: Allowance rates by patent technology fields (for patent applications filed between 1996\nand 2005)"
+        },
+        {
+          "bbox": [
+            84,
+            109,
+            604,
+            417
+          ],
+          "label": "fig",
+          "reading_order": 1,
+          "text": "![Figure](figures/figure_001.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_001.png"
+        },
+        {
+          "bbox": [
+            78,
+            432,
+            615,
+            462
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Notes: Figure shows the three types of allowance rates for applications filed at the PTO for the first time between\n1996 and 2005, across the six NBER patent technology fields."
+        },
+        {
+          "bbox": [
+            338,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 3,
+          "text": "15"
+        }
+      ]
+    },
+    {
+      "page_number": 16,
+      "elements": [
+        {
+          "bbox": [
+            79,
+            60,
+            569,
+            79
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Figure 4: Allowance rates by inventor type (for patent applications filed between 1996 and 2005)"
+        },
+        {
+          "bbox": [
+            84,
+            90,
+            536,
+            389
+          ],
+          "label": "fig",
+          "reading_order": 1,
+          "text": "![Figure](figures/figure_001.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_001.png"
+        },
+        {
+          "bbox": [
+            79,
+            405,
+            614,
+            436
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Notes: Figure shows the three types of allowance rates for applications filed at the USPTO for the first time\nbetween 1996 and 2005, across the four inventor types."
+        },
+        {
+          "bbox": [
+            338,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 3,
+          "text": "16"
+        }
+      ]
+    },
+    {
+      "page_number": 17,
+      "elements": [
+        {
+          "bbox": [
+            79,
+            61,
+            324,
+            82
+          ],
+          "label": "sec_1",
+          "reading_order": 0,
+          "text": "Appendix. Supplementary statistics"
+        },
+        {
+          "bbox": [
+            79,
+            93,
+            578,
+            111
+          ],
+          "label": "cap",
+          "reading_order": 1,
+          "text": "Table A1: Correlations between allowance rates and environmental covariates, 1996-2005"
+        },
+        {
+          "bbox": [
+            113,
+            122,
+            580,
+            231
+          ],
+          "label": "tab",
+          "reading_order": 2,
+          "text": "<table><tr><td></td><td>(A)</td><td>(B)</td><td>(C)</td><td>(D)</td><td>(E)</td></tr><tr><td>(A) First Action Allowance Rate</td><td></td><td></td><td></td><td></td><td></td></tr><tr><td>(B) Progenitor Allowance Rate</td><td>0.949</td><td></td><td></td><td></td><td></td></tr><tr><td>(C) Family Allowance Rate</td><td>0.950</td><td>0.998</td><td></td><td></td><td></td></tr><tr><td>(D)Percent Change in Real GDP</td><td>0.352</td><td>0.482</td><td>0.515</td><td></td><td></td></tr><tr><td>(E)Total Pending Applications</td><td>-0.925</td><td>-0.994</td><td>-0.992</td><td>-0.505</td><td></td></tr><tr><td>(F)Total Pendency</td><td>-0.925</td><td>-0.967</td><td>-0.963</td><td>-0.349</td><td>0.971</td></tr></table>"
+        },
+        {
+          "bbox": [
+            78,
+            258,
+            614,
+            334
+          ],
+          "label": "anno",
+          "reading_order": 3,
+          "text": "Note: Table shows contemporaneous correlations between allowance rates and potential environmental determinants\nof allowance rates (all variables are measured annually, for each year between 1996 and 2005). Total pending\napplications refer to the stock of patent applications filed, and in the examination process for the given year. Total\npendency refers to the average time, in months, between patent application date and patent disposal date during the\nentry year of the progenitor applications in our study."
+        },
+        {
+          "bbox": [
+            338,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 4,
+          "text": "17"
+        }
+      ]
+    },
+    {
+      "page_number": 18,
+      "elements": [
+        {
+          "bbox": [
+            79,
+            61,
+            538,
+            80
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Table A2: Progenitor applications and related continuation applications, 1996-2005"
+        },
+        {
+          "bbox": [
+            90,
+            90,
+            594,
+            303
+          ],
+          "label": "tab",
+          "reading_order": 1,
+          "text": "<table><tr><td rowspan=\"2\">Year</td><td rowspan=\"2\">Applications</td><td colspan=\"4\">Serialized Continuations</td><td rowspan=\"2\">Non-serialized Continuations (RCEs)</td><td rowspan=\"2\">Either Continuation</td></tr><tr><td>CON</td><td>CIP</td><td>DIV</td><td>Any</td></tr><tr><td>1996</td><td>146,260</td><td>6.9\\%</td><td>5.6\\%</td><td>6.5\\%</td><td>17.7\\%</td><td>11.2\\%</td><td>24.9\\%</td></tr><tr><td>1997</td><td>166,232</td><td>5.8\\%</td><td>5.3\\%</td><td>6.7\\%</td><td>16.5\\%</td><td>12.1\\%</td><td>25.6\\%</td></tr><tr><td>1998</td><td>182,717</td><td>6.3\\%</td><td>5.0\\%</td><td>6.8\\%</td><td>16.9\\%</td><td>13.4\\%</td><td>26.9\\%</td></tr><tr><td>1999</td><td>197,704</td><td>6.9\\%</td><td>5.0\\%</td><td>6.9\\%</td><td>17.5\\%</td><td>14.5\\%</td><td>28.3\\%</td></tr><tr><td>2000</td><td>222,480</td><td>7.1\\%</td><td>4.8\\%</td><td>6.5\\%</td><td>17.2\\%</td><td>15.7\\%</td><td>29.0\\%</td></tr><tr><td>2001</td><td>232,668</td><td>7.1\\%</td><td>4.4\\%</td><td>6.5\\%</td><td>16.9\\%</td><td>17.4\\%</td><td>30.3\\%</td></tr><tr><td>2002</td><td>233,246</td><td>6.7\\%</td><td>4.4\\%</td><td>6.1\\%</td><td>16.1\\%</td><td>19.7\\%</td><td>31.5\\%</td></tr><tr><td>2003</td><td>235,861</td><td>6.3\\%</td><td>4.1\\%</td><td>5.1\\%</td><td>14.6\\%</td><td>24.1\\%</td><td>33.7\\%</td></tr><tr><td>2004</td><td>250,338</td><td>6.3\\%</td><td>3.4\\%</td><td>4.9\\%</td><td>13.7\\%</td><td>27.3\\%</td><td>35.6\\%</td></tr><tr><td>2005</td><td>278,160</td><td>6.5\\%</td><td>2.7\\%</td><td>4.7\\%</td><td>13.2\\%</td><td>29.2\\%</td><td>37.1\\%</td></tr></table>"
+        },
+        {
+          "bbox": [
+            79,
+            330,
+            614,
+            361
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Note: Table shows the number of progenitor applications filed in the corresponding year, and the percentage of the\napplications from each cohort that produced the different types of continuations."
+        },
+        {
+          "bbox": [
+            339,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 3,
+          "text": "18"
+        }
+      ]
+    },
+    {
+      "page_number": 19,
+      "elements": [
+        {
+          "bbox": [
+            78,
+            61,
+            543,
+            80
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Table A3: The use of Continuation applications across technology fields, 1996-2005"
+        },
+        {
+          "bbox": [
+            60,
+            90,
+            633,
+            237
+          ],
+          "label": "tab",
+          "reading_order": 1,
+          "text": "<table><tr><td rowspan=\"2\">Technology Field</td><td rowspan=\"2\">Applications</td><td colspan=\"4\">Serialized Continuations</td><td rowspan=\"2\">Non-serialized Continuations (RCEs)</td><td rowspan=\"2\">Either Continuation</td></tr><tr><td>CON</td><td>CIP</td><td>DIV</td><td>Any</td></tr><tr><td>Chemical</td><td>245,150</td><td>6.0\\%</td><td>5.3\\%</td><td>9.2\\%</td><td>19.1\\%</td><td>18.2\\%</td><td>32.8\\%</td></tr><tr><td>Drugs \\</td><td>Medical</td><td>227,936</td><td>12.8\\%</td><td>8.2\\%</td><td>10.0\\%</td><td>28.2\\%</td><td>24.5\\%</td><td>44.1\\%</td></tr><tr><td>Computers \\</td><td>Comm.</td><td>611,046</td><td>8.3\\%</td><td>3.2\\%</td><td>3.6\\%</td><td>14.1\\%</td><td>26.7\\%</td><td>36.0\\%</td></tr><tr><td>Electrical \\</td><td>Electronic</td><td>402,401</td><td>4.7\\%</td><td>3.0\\%</td><td>7.7\\%</td><td>14.5\\%</td><td>16.4\\%</td><td>27.5\\%</td></tr><tr><td>Mechanical</td><td>311,040</td><td>3.9\\%</td><td>3.8\\%</td><td>4.9\\%</td><td>11.9\\%</td><td>13.2\\%</td><td>22.7\\%</td></tr><tr><td>Others</td><td>348,093</td><td>4.6\\%</td><td>5.2\\%</td><td>4.2\\%</td><td>13.2\\%</td><td>13.4\\%</td><td>23.7\\%</td></tr></table>"
+        },
+        {
+          "bbox": [
+            78,
+            263,
+            613,
+            294
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Note: Table shows the number of progenitor applications filed in each NBER patent technology field (between 1996\nand 2005), and the percentage of the applications that produced the different types of continuations."
+        },
+        {
+          "bbox": [
+            338,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 3,
+          "text": "19"
+        }
+      ]
+    },
+    {
+      "page_number": 20,
+      "elements": [
+        {
+          "bbox": [
+            79,
+            61,
+            493,
+            79
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Table A4: The use of Continuation applications across applicant types, 1996-2005"
+        },
+        {
+          "bbox": [
+            80,
+            88,
+            611,
+            196
+          ],
+          "label": "tab",
+          "reading_order": 1,
+          "text": "<table><tr><td rowspan=\"2\">Applicant Type</td><td rowspan=\"2\">Applications</td><td colspan=\"4\">Serialized Continuations</td><td rowspan=\"2\">Non-serialized Continuations</td><td rowspan=\"2\">Either Continuation</td></tr><tr><td>CON</td><td>CIP</td><td>DIV</td><td>Any</td></tr><tr><td>Large Foreign</td><td>838,210</td><td>4.4\\%</td><td>1.3\\%</td><td>5.9\\%</td><td>11.2\\%</td><td>21.1\\%</td><td>29.1\\%</td></tr><tr><td>Small Foreign</td><td>207,460</td><td>3.7\\%</td><td>3.7\\%</td><td>2.9\\%</td><td>9.7\\%</td><td>12.1\\%</td><td>19.3\\%</td></tr><tr><td>Large US</td><td>668,527</td><td>9.2\\%</td><td>5.2\\%</td><td>7.6\\%</td><td>20.4\\%</td><td>23.0\\%</td><td>37.6\\%</td></tr><tr><td>Small US</td><td>431,469</td><td>8.2\\%</td><td>9.2\\%</td><td>5.0\\%</td><td>20.5\\%</td><td>14.3\\%</td><td>30.0\\%</td></tr></table>"
+        },
+        {
+          "bbox": [
+            79,
+            223,
+            614,
+            255
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Note: Table shows the number of progenitor applications filed by each applicant type (between 1996 and 2005), and\nthe percentage of each type's applications that produced the different types of continuations."
+        },
+        {
+          "bbox": [
+            337,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 3,
+          "text": "20"
+        }
+      ]
+    },
+    {
+      "page_number": 21,
+      "elements": [
+        {
+          "bbox": [
+            79,
+            61,
+            505,
+            79
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Table A5: Allowance Rates across applicant types and technology fields, 1996-2005"
+        },
+        {
+          "bbox": [
+            88,
+            88,
+            604,
+            510
+          ],
+          "label": "tab",
+          "reading_order": 1,
+          "text": "<table><tr><td>Technology Field</td><td>Applicant Type</td><td>Applications</td><td>First Action</td><td>Progenitor</td><td>Family</td></tr><tr><td rowspan=\"4\">Chemical</td><td>Large Foreign</td><td>112,598</td><td>11.0\\%</td><td>59.6\\%</td><td>75.4\\%</td></tr><tr><td></td><td>Large US</td><td>76,595</td><td>11.3\\%</td><td>57.2\\%</td><td>74.1\\%</td></tr><tr><td></td><td>Small Foreign</td><td>20,245</td><td>11.6\\%</td><td>52.9\\%</td><td>64.4\\%</td></tr><tr><td></td><td>Small US</td><td>35,712</td><td>9.7\\%</td><td>52.4\\%</td><td>65.8\\%</td></tr><tr><td rowspan=\"4\">Computers \\&amp; Commun.</td><td>Large Foreign</td><td>244,453</td><td>11.7\\%</td><td>54.5\\%</td><td>74.0\\%</td></tr><tr><td></td><td>Large US</td><td>251,253</td><td>8.9\\%</td><td>51.8\\%</td><td>74.1\\%</td></tr><tr><td></td><td>Small Foreign</td><td>32,847</td><td>9.6\\%</td><td>37.7\\%</td><td>48.9\\%</td></tr><tr><td></td><td>Small US</td><td>82,493</td><td>6.4\\%</td><td>34.5\\%</td><td>49.6\\%</td></tr><tr><td rowspan=\"4\">Drugs \\&amp; Medical</td><td>Large Foreign</td><td>62,142</td><td>5.3\\%</td><td>45.0\\%</td><td>63.6\\%</td></tr><tr><td></td><td>Large US</td><td>69,632</td><td>6.0\\%</td><td>43.1\\%</td><td>62.7\\%</td></tr><tr><td></td><td>Small Foreign</td><td>27,372</td><td>5.7\\%</td><td>39.9\\%</td><td>55.4\\%</td></tr><tr><td></td><td>Small US</td><td>68,790</td><td>5.6\\%</td><td>41.5\\%</td><td>58.3\\%</td></tr><tr><td rowspan=\"4\">Electrical \\&amp; Electronics</td><td>Large Foreign</td><td>204,125</td><td>15.5\\%</td><td>67.7\\%</td><td>83.3\\%</td></tr><tr><td></td><td>Large US</td><td>122,529</td><td>14.2\\%</td><td>69.3\\%</td><td>84.5\\%</td></tr><tr><td></td><td>Small Foreign</td><td>30,489</td><td>17.0\\%</td><td>57.7\\%</td><td>65.2\\%</td></tr><tr><td></td><td>Small US</td><td>45,258</td><td>13.1\\%</td><td>60.0\\%</td><td>71.1\\%</td></tr><tr><td rowspan=\"4\">Mechanical</td><td>Large Foreign</td><td>128,328</td><td>15.1\\%</td><td>68.8\\%</td><td>82.1\\%</td></tr><tr><td></td><td>Large US</td><td>74,681</td><td>14.1\\%</td><td>67.2\\%</td><td>80.5\\%</td></tr><tr><td></td><td>Small Foreign</td><td>40,274</td><td>15.8\\%</td><td>56.2\\%</td><td>63.7\\%</td></tr><tr><td></td><td>Small US</td><td>67,757</td><td>12.0\\%</td><td>57.1\\%</td><td>65.9\\%</td></tr><tr><td rowspan=\"4\">Others</td><td>Large Foreign</td><td>86,564</td><td>11.3\\%</td><td>60.7\\%</td><td>74.6\\%</td></tr><tr><td></td><td>Large US</td><td>73,837</td><td>9.9\\%</td><td>56.5\\%</td><td>71.9\\%</td></tr><tr><td></td><td>Small Foreign</td><td>56,233</td><td>13.5\\%</td><td>51.1\\%</td><td>57.7\\%</td></tr><tr><td></td><td>Small US</td><td>131,459</td><td>9.5\\%</td><td>49.3\\%</td><td>57.4\\%</td></tr></table>"
+        },
+        {
+          "bbox": [
+            78,
+            536,
+            614,
+            582
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Note: Table shows the number of progenitor applications filed in each of the six NBER patent technology fields by\neach applicant type (between 1996 and 2005), and the percentage of each type's applications that produced the\ndifferent types of continuations."
+        },
+        {
+          "bbox": [
+            337,
+            826,
+            354,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 3,
+          "text": "21"
+        }
+      ]
+    },
+    {
+      "page_number": 22,
+      "elements": [
+        {
+          "bbox": [
+            78,
+            61,
+            524,
+            80
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Figure A1: Trends in allowance rates with adjustments for censoring, 1991-2010"
+        },
+        {
+          "bbox": [
+            116,
+            93,
+            566,
+            372
+          ],
+          "label": "fig",
+          "reading_order": 1,
+          "text": "![Figure](figures/figure_001.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_001.png"
+        },
+        {
+          "bbox": [
+            79,
+            386,
+            615,
+            537
+          ],
+          "label": "anno",
+          "reading_order": 2,
+          "text": "Notes: Figure shows trends in the three types of allowance rates for the 4.2 million applications filed at the PTO for\nthe first time between 1991 and 2010. A significant number of applications filed after 2005 were pending as of June\n30, 2013 and the dotted lines (for the first-action allowance rate and progenitor allowance rate) represent the\ncorresponding rates if all the pending applications are, in fact, allowed. Thus, they represent the theoretical upper-\nbound for the allowance rates (a vast majority of the applications filed for the first time in 2010 were past the first\naction, but still pending at the office: if all of these pending applications were to issue, then the progenitor allowance\nrate for 2008 applications would be around 68 % ). For progenitor applications that produced continuation\napplications that are still pending, we calculate the maximum possible family allowance rate by assuming that every\npending continuation application produced by the progenitors will eventually be allowed. This maximum possible\nfamily allowance rate is represented by the corresponding dashed line."
+        },
+        {
+          "bbox": [
+            337,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 3,
+          "text": "22"
+        }
+      ]
+    },
+    {
+      "page_number": 23,
+      "elements": [
+        {
+          "bbox": [
+            78,
+            60,
+            540,
+            80
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Figure A2: Trends in First Action Allowance Rate by Technology Field, 1996-2005"
+        },
+        {
+          "bbox": [
+            126,
+            93,
+            561,
+            402
+          ],
+          "label": "fig",
+          "reading_order": 1,
+          "text": "![Figure](figures/figure_001.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_001.png"
+        },
+        {
+          "bbox": [
+            78,
+            418,
+            532,
+            438
+          ],
+          "label": "cap",
+          "reading_order": 2,
+          "text": "Figure A3: Trends in Progenitor Allowance Rate by Technology Field, 1996-2005"
+        },
+        {
+          "bbox": [
+            83,
+            452,
+            536,
+            761
+          ],
+          "label": "fig",
+          "reading_order": 3,
+          "text": "![Figure](figures/figure_003.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_003.png"
+        },
+        {
+          "bbox": [
+            337,
+            826,
+            354,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 4,
+          "text": "23"
+        }
+      ]
+    },
+    {
+      "page_number": 24,
+      "elements": [
+        {
+          "bbox": [
+            78,
+            60,
+            513,
+            80
+          ],
+          "label": "cap",
+          "reading_order": 0,
+          "text": "Figure A4: Trends in Family Allowance Rate by Technology Field, 1996-2005"
+        },
+        {
+          "bbox": [
+            120,
+            94,
+            566,
+            402
+          ],
+          "label": "fig",
+          "reading_order": 1,
+          "text": "![Figure](figures/figure_001.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_001.png"
+        },
+        {
+          "bbox": [
+            337,
+            826,
+            355,
+            841
+          ],
+          "label": "foot",
+          "reading_order": 2,
+          "text": "24"
+        }
+      ]
+    }
+  ],
+  "metadata": {}
+}

processed/00071_W2085332066/document.md

@@ -0,0 +1,656 @@
+GETTING PATENTS AND ECONOMIC DATA TO SPEAK TO EACH OTHER:
+
+AN `ALGORITHMIC LINKS WITH PROBABILITIES' APPROACH FOR JOINT ANALYSES OF PATENTING AND ECONOMIC ACTIVITY
+
+by
+
+Travis J. Lybbert, * University of California, Davis and
+
+Nikolas J. Zolas, * U.S. Bureau of the Census
+
+CES 12-16 September, 2012
+
+The research program of the Center for Economic Studies (CES) produces a wide range of economic analyses to improve the statistical programs of the U.S. Census Bureau. Many of these analyses take the form of CES research papers. The papers have not undergone the review accorded Census Bureau publications and no endorsement should be inferred. Any opinions and conclusions expressed herein are those of the author(s) and do not necessarily represent the views of the U.S. Census Bureau. All results have been reviewed to ensure that no confidential information is disclosed. Republican in whole or part must be cleared with the authors.
+
+To obtain information about the series, see www.census.gov/ces or contact C.J. Krizan, Editor, Discussion Papers, U.S. Census Bureau, Center for Economic Studies 2K130F, 4600 Silver Hill Road, Washington, DC 20233, CES.Papers.List@census.gov.
+
+Electronic copy available at: http://ssrn.com/abstract=2161982
+
+---
+
+## Abstract
+
+International technological diffusion is a key determinant of cross-country differences in economic performance. While patents can be a useful proxy for innovation and technological change and diffusion, fully exploiting patent data for such economic analyses requires patents to be tied to measures of economic activity. In this paper, we describe and explore a new algorithmic approach to constructing concordances between the International Patent Classification (IPC) system that organizes patents by technical features and industry classification systems that organize economic data, such as the Standard International Trade Classification (SITC), the International Standard Industrial Classification (ISIC) and the Harmonized System (HS). This `Algorithmic Links with Probabilities' (ALP) approach incorporates text analysis software and keyword extraction programs and applies them to a comprehensive patent dataset. We compare the results of several ALP concordances to existing technology concordances. Based on these comparisons, we select a preferred ALP approach and discuss advantages of this approach relative to conventional approaches. We conclude with a discussion on some of the possible applications of the concordance and provide a sample analysis that uses our preferred ALP concordance to analyze international patent flows based on trade patterns.
+
+NOTE: The ALP concordances described and used in this paper will soon be available for download on the WIPO website at http://www.wipo.int/ipstats/en/statistics/patents/
+
+*The authors would like to thank Prantik Bhattachayya for superb research assistance. We are grateful for the assistance and guidance provided by researchers and programmers at the World Intellectual Property Organization (WIPO), including Carsten Fink, Hao Zhou, Christophe Mazenc, Sacha Wunsch-Vincent, and others. We thank seminar participants at WIPO and Morrison Foerster. We also acknowledge the financial support the project has received from the National Science Foundation. The research in this paper was undertaken while Nikolas Zolas was at the University of California, Davis. Any opinions and conclusions expressed herein are those of the authors and do not necessarily represent the views of the U.S. Census Bureau or the NSF. The research in this paper does not use any confidential Census Bureau information.
+
+Electronic copy available at: http://ssrn.com/abstract=2161982
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+## 1 Introduction
+
+International technological diffusion is an important driver of technological change which is a key determinant of cross-country differences in income and economic growth (Romer 1990; Aghion and Howitt 1992; Grossman and Helpman 1991; Keller 2004) . International trade and foreign direct investment are often considered to be key catalysts of technology transfer (Coe and Helpman 1995; Eaton and Kortum 2002; Branstetter et al. 2006; Archaya and Keller 2009) , but directly studying this process is often hampered by the fact that measuring transferred technology empirically is challenging. Thus far, data and statistics on patents have served as one of the more useful proxies for measuring technological change (Griliches 1990; Basberg 1987) and diffusion (Jaffe et al. 1993) . However, in order to fully exploit patent data in economic analyses, researchers must be able to link patents to economic activity at a level of disaggregation that allows for different technological, industrial and spatial patterns. Such a detailed link between technological and economic activity would further improve our assessment of policies that aim to promote innovation, as well as assess the relationship between technological change and economic development.
+
+Patent statistics have frequently been used as both technological and economic indicators due to the widespread availability of patent data and the assumption that patents reflect direct inventive activity and innovation. In his survey reviewing the different ways patents are used as technology indicators, Basberg (1987) describes how patents have been incorporated in innovation models to measure to technology diffusion and to evaluate the output of research activity. In a similar survey, Griliches (1990) documents the numerous instances patents have been used as economic indicators and finds that patents have held diverse roles from serving as proxy for R & D output to predicting stock-market activity and total factor productivity. Within this literature, however, the validity of patents as technological or economic indicators remains a somewhat of an open question. Important concerns include the commercial use and value of patents, heterogeneity across countries and industries in patent institutions, legislation and enforcement, and pronounced changes over time in patenting and patent institutions. We believe that more disaggregate analyses of patent statistics – particularly when matched with equally disaggregate economic data – will help to address these concerns and enable new empirical research related to patents.
+
+In general, there are three levels at which patents can be linked to economic activity. At the coarsest macro-level, aggregate patent data taken from a specific country in a specific year can be associated with aggregate economic data, respectively. Linking patent and economic data at this aggregate level is based simply on the country-year unit of analysis and has enabled research on questions such as measuring the rate of innovation (Porter and Stern 2000) , a country's innovative capacity (Furman et al. 2002) and the effects of patent harmonization (McCalman 2001) . Analyses of foreign patent flows and economic activity (Eaton and Kortum 1996; Xu and Chiang 2005; Falvey and Foster 2006; Harhoff et al. 2007) is similarly based on an aggregate association of patents to economic data through a shared space-time unit of analysis.
+
+At the finest level, patents and economic activity can be linked at the firm-level. While this micro-linkage between patent and economic data enables rigorous and insightful research on patenting as part of firm-level strategies (Brouwer and Kleinknecht 1999; Austin 1993)
+
+2
+
+Electronic copy available at: http://ssrn.com/abstract=2161982
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+constructing and maintaining such a firm-level database requires substantial effort, is only feasible for a fraction of the firms represented in patent databases, and may miss broader considerations regarding relevant products, competitors and industrial dynamics. Among other things, these limitations constrain our ability to use a firm-level linkage between patents and economic activity to learn much about patenting in important emerging economies where firmlevel data is sparse.
+
+Between these macro- and micro-level linkages is a meso- or industry-level linkage that associates patents and economic data based on the domain of goods and services they represent. At this level, patents on biomedical and semiconductor inventions, for example, are linked to industry or product classes that use biomedical and semiconductor inventions, respectively. We argue that a robust industry-level linkage – perhaps in conjunction with macro- and micro-level analyses – will enable researchers to better analyze the relationship between patenting and economic activity over time, across space and technology classes. Most industry-level linkages are based on concordances. For example, the Yale Technology Concordance (YTC) (Kortum and Putnam 1997) links the International Patenting Classification code (IPC) to the Canadian Standardized Industrial Classification system. Thus, with the YTC a researcher can link patent data organized by IPC, country and year to the value of production organized by Canada SIC, country and year. Unfortunately, conventional concordance approaches like the YTC suffer from a host of flaws that limit their usefulness in empirical research. After describing these limitations, we propose new methods for constructing concordances and, thereby, industry-level linkages between patent and economic data. These methods use text analysis, data mining and probabilistic matching to build these links in ways that can be applied broadly or narrowly across time and space, can be easily updated, and can create direct linkages between patent data and a variety of industry and trade classification schemes in a way that that does not require layers of concordances.
+
+We refer to the general approach we develop in this paper as an Algorithmic Links with Probabilities (ALP) approach to constructing concordances. We propose and test two different versions of this approach. First, a data mining approach (ALP-DM) identifies patents that contain manually-assigned keywords where each word pertains to a specific industry. The patents are aggregated and reveal a frequency matches between the keywords and IPC subclasses. This frequency then provides the basis for weighting each of the matches. Second, a keyword extraction and probabilistic matching approach (ALP-PM) extracts keywords from the patents themselves and then matches these keywords probabilistically to industry or trade classifications. By implementing these two approaches using the full PATSTAT database provided by the European Patent Office (EPO), we generate probability distributions of the technologies used within each industry and, conversely, distributions of the industries using certain types of technology. Since these distributions create linkages in both directions – from patents to economic data and vice versa – researchers can use these direct concordances for industry-level analyses of the relationships between patents and an array of economic activity organized by different classification schemes such as SITC, NACE, ISIC and HS. Given that these methods require minimal manual or subjective intervention, the concordances they generate are easy to update and refine.
+
+3
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+After providing a brief background of related patent concordance research, we discuss the prevailing IPC concordances in some detail and describe a fundamental limitation of these conventional concordances when applied to economic data. We then describe our ALP approaches to constructing more useful concordances and generate IPC concordances for both trade (SITC) and industry (ISIC) classification schemes. To provide a test our approach, we use our ALP-DM and ALP-PM approaches to generate concordances that can be directly compared with two prevailing concordances, including the YTC. Before concluding, we demonstrate the use of ALP concordances with a specific analysis that compares patent and trade flows.
+
+## 2 Background
+
+Patents are a potentially powerful data source for technology and innovation analyses because the patents themselves contain a wealth of information, including the names of the inventee, date, prior art, technologies used, as well as a full description of the embedded technology with numerous figures and references. Recently, there has been a large push initiated by the private sector to develop novel ways of analyzing, organizing and making this patent information accessible to firms interested in exploiting or diversifying their patent portfolios and formulating R & D strategies (Moehrle et al. 2010) . This form of patent analysis – called “patinformatics” – aims to reveal relationships between individual patents and broader technological fields in order to inform commercial, legal and policy decisions and includes grouping similar concepts and technologies, creating patent landscape maps, tracking the evolution of these maps over time, and analyzing and interpreting citation networks. These approaches typically use the latest developments in text analysis and text clustering software, and then uses the findings from these programs to create different visualization and mapping schemes. The methods we develop are conceptually similar to these tools and could ultimately provide a valuable economic layer to patent landscapes, networks and other patinformatic analyses.
+
+The ALP concordances we construct are designed to enable more rigorous econometric analysis at the industry-level. By doing this, we continue to build on other efforts to link patent and economic data through technology-industry associations. While these industry-level linkages are facilitated by the fact that the IPC and economic classification systems share a detailed hierarchical structure, they are complicated by the fact that these classification systems are motivated by different objectives. Whereas economic classification systems are intended to disaggregate goods and services into meaningful and related sub-groups, the IPC system is intended to facilitate the patent examination process by enabling patent examiners to precisely identify the novel technical features of the disclosed invention and to define the prior art against which they can assess novelty. Since goods or services in very different economic classifications can use the same technical feature (e.g., an electronic motion control device may be used in washing machines and satellites), this difference in intended usage implies that linking patents to economic data through a concordance of their respective classification systems is never straightforward. Whereas one could manually construct a one-to-one concordance between two industrial classification schemes that share the same unit of analysis (i.e., industry), constructing a concordance between the IPC and an economic classification at any useful level of resolution is effectively a many-to-many mapping that may not amenable to a manual approach.
+
+4
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+The first attempt to link patent data with industry data was conducted by Schmookler in 1966 (Comanor and Scherer 1969) who assigned “industries-of-use” to patents organized by the US patent class (USPC). The classification scheme used in this concordance assigned patent classes to industries where at least 2/3 of patents in that class were used for that particular industry. A later concordance developed by a branch of the US Patent and Trademark Office (USPTO) used a similar methodology and assigned equal weighting to patent classes which related to multiple industries. The first comprehensive concordance, the YTC, emerged in the early 1990s (Evenson and Putnam 1994; Kortum and Putnam 1997) . The YTC was constructed by leveraging a useful feature of the roughly 250,000 patents issued in Canada between 1978 and 1993. For each of these patents, the Canadian Patent office examiners were required to assign a technology field from the IPC system (standard practice worldwide) and to indicate the Industry of Manufacture (IOM) and Sector of Use (SOU) of the invention according to the Canadian Standard Industrial Classification (1980 cSIC-E Version). The patents examined in this window implicitly concord IPC to cSIC since examiners were assigning patents to both systems concurrently. The YTC tabulated these assignments to make this an explicit IPC-cSIC concordance.
+
+Because it is based on assignments made by patent examiners – presumably, experts in the field – the YTC benefits from hundreds of thousands of hours of expertise and consideration. Furthermore, this structure implies that the YTC comprehensively covers all technologies and industries included in the 250,000 patents that were cross-classified. An additional benefit is that the YTC uses probabilistic rather than subjective weights, which allows for the same technical feature to be used in multiple sectors. On the other hand, the YTC suffers from several serious limitations. First, it is only possible to possible to directly link to one classification system, the cSIC, which is not commonly used in industry-level studies. Bridging to any other economic classification system introduces noise and can hopelessly atrophy the resulting composite concordance (as discussed below). Second, it is frozen in time and space, as it were, because it will always be based on Canadian patents examined between 1978 and 1993. This introduces potential technological, temporal and spatial biases (Schmoch et al. 2003) .
+
+## 3 The IPC & Prevailing IPC-Industry Concordances
+
+In this section, we describe in more detail the structure of the prevailing concordances that attempt to link the IPC to industry classification systems. First, we describe briefly the structure of the IPC system and contrast it with existing economic classification systems. We then differentiate between the prevailing concordances that that build on the YTC and those that chart a different path entirely and discuss them in reverse order.
+
+The IPC was established in 1971 by the Strasbourg Agreement to provide a harmonized, language independent, hierarchical system for classifying technology embedded in patents and utility models 2 . Given its role in defining the scope of prior art considered in patent examination,
+
+2 For a complete guide to the IPC, including useful training resources, see http://www.wipo.int/export/sites/www/classifications/ipc/en/guide/guide_ipc_2009.pdf
+
+To explore the IPC interactively with complete notes see http://www.wipo.int/ipcpub
+
+5
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+the IPC is a central feature to the global network of national patent systems. The current version of the IPC divides technology into eight sections, which are further divided into a total of nearly 70,000 “ subgroups ” . To illustrate the structure of the IPC, consider the example of IPC “ subgroup ” B64C 11/18, which covers “ Aerodynamic features of propellers used in aircraft. ” This group number is composed of section B ( “ Performing operations; Transporting ” ), class B64 ( “ Aircraft; Aviation; Cosmonoautics ” ), subclass B64C ( “ Aeroplanes; Helicopters ” ), main group B64C 11/00 ( “ Propellers ” ), and subgroup B64C 11/18. We construct our concordance at the four-digit subclass level (e.g., B64C, A21B, etc.), of which a total of 639 exist (in the most recent version). In terms of how the IPC is used in practice, patent examiners around the world classify the inventions claimed by the patents they examine. Where multiple inventive features are evident in an invention, examiners often cross-list the patent in multiple IPCs. $^3$
+
+With this brief description of the IPC in mind, consider the structure of existing IPC-industry concordances. Two of these concordances, the “ DG Concordance ” (Schmoch et al. 2003) and the MERIT Concordance (Verspagen et al. 1994) , chart a different path than the YTC. Both of these concordances attempt to match IPC subclasses to ISIC industry classifications using the official descriptions of these respective categories. In order to do this manually, both efforts are based on one-to-one matches, which is only feasible at a relatively coarse resolution. Specifically, the DG concordance assigns 625 IPC subclasses to one of 44 different manufacturing sectors, of which one or more ISICs are associated. The MERIT Concordance matches IPC subclasses to 22 industrial classes based on a mix of two- and three-digit ISIC codes. Both approaches are notable for their attempt to manually and directly (i.e., one-to-one) translate the IPC to the ISIC industry classification system. While the mapping to the ISIC that emerges from these efforts is undeniably coarse, it can nevertheless enable some useful empirical and policy analysis.
+
+For more rigorous analysis, higher resolution economic data can be particularly useful – but leveraging these higher resolution data requires a higher resolution concordance. To construct a higher resolution concordance, researchers have had little choice but to trod the YTC path and rely on the same narrow base of Canadian patents. Two other prevailing concordances take this approach and seek to build on the YTC. Specifically, the OECD Concordance (Johnson 2002) and PATDAT Concordance used by Silverman 4 simply layer an additional concordance to translate the IPC to more commonly used industry classification systems such as ISIC (used in OECD) and the US Standard Industrial Classification (SIC) (used in PATDAT). This conventional composite concordance approach introduces additional complications, such as causing the strength of the technology-industry linkage to atrophy. To illustrate this problem, Table 1 takes a random IPC subclass, B64D “ Aircraft; Aviation; Cosmonautics Equipment for Fitting In or To Aircraft ” , and shows what happens during the layering process. Whereas the initial concordance is sensible, the composite concordance has clearly atrophied – even when the additional concordance layer (cSIC-ISIC in this case) is itself quite robust. Obviously, the severity of this problem intensifies with additional concordance layers.
+
+3 In some jurisdictions, examiners must designate a primary IPC and list the remaining IPCs as secondary. The PATSTAT database compiles patent data from many jurisdictions, only some of which follow this convention, so a primary IPC designation is not always available when multiple IPCs are listed on a patent
+
+4See http://www.rotman.utoronto.ca/~silverman/ipcsic/documentation_ipc-sic_concordance.htm for documentation and procedure
+
+6
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+In summary, any effort to analyze the relationship between patents and economic activity at the industry-level faces a serious concordance dilemma. While there is rich, high resolution data for both patents and economic activity, and these data would seem to enable a host of insightful empirical analyses, jointly harnessing the high resolution on both sides requires a robust, accurate and high resolution concordance. Manual, one-to-one concordances are too crude for many research questions, but up-to-date more sophisticated concordances have little choice but to build on a relatively narrow set of Canadian patents that are effectively frozen in time, space and technology. Furthermore, since very few (if any) datasets are described with the cSIC classification system, additional concordance layers are required to construct more broadly useful concordances from this narrow patent base, which quickly atrophies the integrity of the concordance.
+
+## 4 Guiding Principles and Methodology
+
+To escape the dilemma described in the previous section, an ideal concordance would replicate the human process of reviewing each patent and assigning industry codes based on the information contained within the patent, while also including a much broader set of patents from around the world, allowing for direct translation into multiple economic classification schemes, and facilitating updates to reflect technological and classification system changes. In this section, we formalize a set of guiding principles based on this ideal and then describe the methods we develop to approximate an ideal concordance according to these principles.
+
+Three principles have guided our effort to approximate an ideal concordance to link patents to economic activity:
+
+- 1. Use the descriptive content of individual patents as the basis for the concordance .
+Since technical features classified in the IPC can pertain to several different classes of
+economic activity, it is important to consider each patent individually. An ideal
+concordance would be based on an effective evaluation of the content of each patent,
+including how and where the underlying invention may be used. The patent applicant is
+best suited to assess the potential uses of the invention and, in most jurisdictions, has an
+incentive to discuss this industrial usefulness in the application.
+2. Eliminate the need for concordance layering by constructing direct concordances .
+To avoid the composite concordance problem, we aim to devise methods that can be
+directly applied to the most common economic classification schemes, including SITC
+(Rev. 2 and 3), ISIC (Rev. 2, 3, 3.1 and 4), NAICS, HS and SIC. As new versions of
+these concordances or the IPC are released, new direct concordances are preferable to
+indirect ones that update the older to the newer version via a concordance.
+3. Automate the construction process as much as possible . Technology changes rapidly,
+and the concordance should reflect these changes. A proper concordance will therefore
+need continuous updating to reflect new technologies as they emerge. Automating the
+process implies that it should:
+a. Involve minimal manual work in order to rapidly process millions of patents
+at a time . The process should not require, for example, manually sifting through
+patents or classification schemes.
+7
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+- b. Be relatively easy to implement and flexible enough to capture changing
+technologies and industries . Through the process, generating a new version of
+the concordance should be relatively cheap and easy to do. The process should
+also be flexible enough to allow for adjustments in the technological focus or
+years considered to tailor the concordance as needed.
+c. Rely more on objective algorithms than subjective judgments . This helps to
+reduce the manual workload of constructing the concordance, but can also
+provides a critical objective basis on which to construct the weights in a many-to-
+many concordance.
+The ALP methodology we describe below is guided by these principles and leverages recent advances in computing power and search techniques. Programs that perform tasks such as keyword extraction and text mining allow for specific bits of information to be extracted from individual patents, making it possible to approximate a manual assignment of industry classifications. As with any algorithmic search technique, our methods cannot perfectly replicate careful manual inspection and assignment, but because they can sift through millions of patents they may be able to converge on accurate implied linkages. Because our ALP approach statistically relies on the “Law of Large Numbers,” we expect the resulting concordances to improve as the number of patents processed increases.
+
+Patents are a natural candidate for mining and clustering techniques because of their wealth of information. We use the PATSTAT database available from the European Patent Office (EPO) as the source of these patent data. The PATSTAT database contains patent data for 86 countries since 1990 and contains details for more than 100 million patent applications, some of which relate to the same invention in different jurisdictions. Included in this database are almost 20 million unique patent abstracts and titles. In contrast, there is no comparable information-rich source of qualitative data on economic activity by industry classification. Instead, these economic classification systems typically have only one source of qualitative information: the brief descriptions used to characterize a particular category of goods or services. Standard keyword extraction from these concise industry descriptions is challenging and often produces too narrow a set of keywords. To expand these keywords, we exploited the Cross-Lingual Expansion tool embedded in WIPO's PATENTSCOPE. 5 This tool is ideal for our purposes because it generates synonyms based on the full text of patents in different languages and therefore expands our keyword lists based on terms that appear frequently in patent documents.
+
+To showcase these mining and matching methods, we focus on directly mapping four-digit IPC subclasses to four-digit SITC trade classifications (SITC Rev. 2) and vice versa. This same process can be replicated for industrial classification schemes such as ISIC and SIC. The next two sections describe our methodology in detail.
+
+### 4.1 Data Mining Approach (ALP-DM)
+
+The data mining approach (ALP-DM), as the name implies, relies on data mining the patent abstracts and titles included in the PATSTAT database using keywords from the industry
+
+$^{5}$ This tool is available here: http://www.wipo.int/PATENTSCOPE/search/elir/elir.jsp?interfaceLanguage=en
+
+8
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+classification descriptions. Specifically, the approach uses search terms generated by hand for the industry descriptions and identifies all of the patents that contain these specific keywords in either the title or abstract. Based on the number of patents that match the search terms, we obtain a frequency of IPC subclasses, which are then reweighted according to how frequent the IPC subclass is used overall.
+
+The key process in the ALP-DM approach is producing the search terms that represent each industry description and reweighting the number of matches. For each industry code, we generate each of the search terms by hand from the text contained in the industry description. In some cases, the search terms have been augmented by additional keywords and synonyms generated by PATENTSCOPE. The search terms are designed to be as industry-specific as possible to reduce the noise coming from patent matches. In many ways, the search terms are similar to coming up with a Google search term for a specific industry. This becomes tricky for a number of reasons. On the one hand, we would like to include as many patent matches as possible to ensure proper coverage of the industry. However, increasing the scope of possible matches tends to introduce more noise and reduced accuracy. Therefore, the process requires careful treatment and we remove all terms that have multiple meanings or are considered too general. We also incorporate the use of “ not ” terms, since many industry descriptions include “ not elsewhere specified ” or refer to a particular sub-group within an industry. The final result from the assignment of search terms is that each industry is typically assigned anywhere from one to several dozen search terms, with additional “ not ” terms. Table 2 provides an example of the search terms generated for a grouping of SITC industry codes.
+
+Once the search terms are generated, it is then a straightforward process to query the PATSTAT database using these terms. Specifically, we identify patents that contain the exact phrases of each search term either in their title or in the abstract. We do not limit the patents by year or country, since we want the pool of patents to be as large and as varied as possible. After identifying the patents, we obtain a frequency of all the IPC subclasses that are contained within those patents. For patents containing multiple IPCs, each IPC is equally reweighted by the total number of IPCs contained within each patent. The (unweighted) frequency share for IPC subclass $j$ is computed as:
+
+$$Unweighted Frequency(SITC_{i},IPC_{j})=\frac{m_{ij}}{M_{i}}\quad (1)$$
+
+where $m_{ij}$ indicates the number of patents that list IPC subclass $j$ among those retrieved by the keywords for SITC $i$ and $M_i$ is the total number of patents retrieved by the keywords for SITC $i$ .
+
+In the next step, we reweight the frequency shares in (1) by how frequently their corresponding IPC subclasses appear in the PATSTAT database. Subclasses that appear very frequently in the PATSTAT database are more likely to generate spurious matches with the search terms, so it is important to reduce the potential for noise by reweighting the matches. We explore two separate weighting schemes. The first weighting scheme ( “ Specificity Weights ” ) reweights the matches by the total number of IPC subclasses found in the database as follows:
+
+$$Specificity Weighted Frequency\left(SITC_{i}, IPC_{j}\right)=\frac{s_{i j}}{M_{i}}=\frac{m_{i j} / N_{j}}{M_{i}}\quad (2)$$
+
+9
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+where $N_j$ indicates the total number of patents in PATSTAT that list IPC subclass $j$ and $s_{ij}$ represents the share of patents listing IPC subclass $j$ that link to SITC $i$ . These specificity weighted frequencies are then normalized to sum to one. The goal of this weighting scheme is to effectively adjust IPC subclasses so they have equal probability of matching the SITC search terms. This reduces much of the spurious matches caused by dominant IPC subclasses. There is a downside to this weighting scheme, however: it tends to disproportionally overweight sparsely used IPC subclasses relative to frequently used IPC subclasses in ways that may not accurately reflect the way these technologies are used in practice. We therefore formulate a second, alternative weighing scheme as a middle ground. This hybrid of unweighted and specificity weighted frequency is defined as:
+
+$$Hybrid Weighted Frequency(SITC_{i},IPC_{j})=\frac{s_{ij}m_{ij}}{\sum_{j}s_{ij}m_{ij}}\quad (3)$$
+
+where we weight each of the IPC subclasses that match to SITC $i$ by share $s_{ij}$ in both the numerator and denominator. This hybrid weighting approach is less extreme than pure specificity weights and may provide a better reflection of the nature of the technologies used in the industries.
+
+Table 3 illustrates the differences between the two weighting schemes using an example. In the example, the search terms for an SITC industry yields two IPC subclasses: A and B . In the initial raw frequency, IPC subclass A has a weight of 10 % , while B has a weight of 90 % . However, due to the fact that IPC subclass B shows up in the dataset 10,000 times, while A shows up in the dataset only 100 times, the “ true ” technological nature of this industry should weigh more heavily towards A since we can be reasonably sure that all of those matches are not spurious. Applying the specificity weighting approach reverses these weights, with A having a weight of 91.74 % and B having a weight of 8.26 % . This may be too extreme, since A only appears in 10 % of the raw estimates, while B appears in 90 % of the raw estimates. It may be the case that B is a widely applicable technology, while A is a narrowly defined technology that is rarely used. Applying the hybrid weighting approach moderates the results, assigning a 55.25 % weight for A and 45.75 % weight for B .
+
+With the differences in our two candidate weighting schemes in mind, note that since the nature of these technology-industry linkages is likely to vary across sectors, we would not expect one weighting scheme to dominate the other universally. In the subsequent section, we test how well these ALP concordances based on these different weighting schemes match existing concordances, which is the best test we could formulate for comparing these weighting options.
+
+Once all of the IPCs have been reweighed, the final step in the process purges the low-frequency IPCs and renormalizes the results. We set an arbitrary cutoff of 2 % so that all IPCs whose frequencies are less than 2 % are excluded and the remaining results are reweighted. This significantly reduces the amount of noise in each SITC. It may be worthwhile experimenting with different cutoff conditions to ascertain the optimal cutoff value. Based on our own explorations, we believe 2 % represents a reasonable cutoff.
+
+To better illustrate the full outcome of the process, we provide the results for SITC code 8484, which is described as "Headgear and fitting thereof". We first queried the PATSTAT database
+
+10
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+using the search terms found in Table 1 . This initial query yielded 11,660 unique patents and 379 unique IPCs. Table 3 shows the results once the matches are reweighted, expunged of the lowfrequency IPCs and renormalized. The final result conforms nicely with our own expectations of the types of technology that would be embedded in this industry.
+
+We repeat these steps for every SITC description and generate weights that match all 4-digit SITCs to 4-digit IPCs. We also apply the same methodology to the product descriptions from other common industry classification systems, such as ISIC, NAICS, HS and US SIC. The benefit of such a concordance is that no layering of concordances is required and the results are based on actual concurrent data, with minimal subjectivity. Researchers will have the flexibility to use a variety of different industry classification systems and get customized technology reports for each industry. Going forward, we will continue to explore different search queries and weighting schemes that may more accurately represent the true results.
+
+### 4.2 Indexing and Probabilistic Matching Approach (ALP-PM)
+
+Our second approach uses a similar methodology as the ALP-DM approach, but incorporates a separate matching process. In this case, we first extract keywords from the patents and then match them to the industry descriptions using probability weights. While the data mining approach would typically be used to translate industries into technologies, this approach might better be used in the opposite direction and match technologies to industries. This approach may also ultimately enable patent-specific matching to economic classifications, although this would require further refining.
+
+In the initial step of this approach, we order the patents by IPC cluster. We then run each of the patents through a keyword extraction program. For our initial approach, we utilize an opensource Python-based keyword extraction program called “Topia Term Extract 1.10.” 6 This extraction program is a generalized text extraction program that identifies the important terms within written content. The benefit of this program is that it also uses language patterns and statistical analysis to determine the strength of each keyword, so that it is possible to rank the keywords by order of importance. There are many other keyword extraction programs in existence, each with their own niche and specialty. While the results from each program will differ slightly, the programs generate very similar results on the whole.
+
+Because of the large quantity of words contained in both the patent abstracts and titles (especially when compared to the quantity of words found in the industry descriptions), it makes sense to weigh the keywords extracted from each patent according to relative importance. In this case, we weigh the keywords from the title to be twice the weight of the extracted keywords from the abstract. This is due to our belief that a single word from the title will provide a better clue as to the real nature of the invention rather than a single word from the abstract. We also limit the number of keywords extracted from each patent to be 10 total words from both the title and abstract. Patent titles and abstracts vary greatly in length, so in order for all patents to receive equal weighting, it is important to limit the matching process to the ten strongest keywords so that certain patents are not more influential.
+
+$^{6}$ The program package and description can be found at http://pypi.python.org/pypi/topia.termextract/
+
+11
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Another more nuanced step in the keyword extraction process is the use of a “ blacklist. ” Early in our analysis, we found that certain words kept appearing on the keyword extractions that were too general to be used in the matching process, such as “ system ” , “ device ” , “ model ” , “ invention ” and more. To construct this blacklist of keywords, we ran the keyword extraction program over 500,000 random patents and tabulated the keywords. We looked at the top 100 keywords and ran the PATENTSCOPE cross-lingual expander on certain keywords, which left us with a blacklist of roughly 250-300 keywords. We remove all of the blacklisted words from the extraction results.
+
+Once all of the keywords have been extracted and tabulated for the IPC cluster, we are left with a list of keywords and weights, which were obtained by summing the number of times each keyword appeared in all of the analyzed patents. Each of the keywords and weights are then matched against the industry classification descriptions generated in the ALP-DM approach with additional augmentations. For our initial runs, we used “ exact string ” matching, although it is possible to do “ like ” matching and set the tolerance level. For the “ exact string ” matching portion, we used an expanded word list based on the ALP-DM search terms, full industry descriptions, PATENTSCOPE synonyms and additional plurals, root words and alternative spellings. The reason for this augmentation of the industry terms is that the pool of possible industry matches is much smaller than the pool of patent matches (a couple hundred versus almost 20 million), so we wanted to maximize the quantity of matches and utilize a filtering system and reweighting process to reduce the false positives and thereby improve quality.
+
+For each match, we weighed the importance of the match by the weight of each keyword. The industries that matched with the keywords that have the highest weight after the extraction process were weighed the most. Once the industry and weights have been tabulated, we are left with our raw results.
+
+Next, to reduce the number of spurious matches, we employed a filtering process to the raw results. The first filtering process involved assigning allowable IPC-SITC correspondences. To implement this filter, we assigned lower level IPC's (3-digit) with lower-level SITCs (2-digit). If the correspondence did not make sense, i.e. agricultural production with steel technology, then we disregarded the weights for that specific match. We did this for all 3-digit IPC's and 2-digit SITCs. The next filter involved the 2 % cutoff condition, which was similarly employed in the ALP-DM approach. All weights that represented less than 2 % of the total weights between IPC and SITC were disregarded and the remaining weights were retabulated and normalized. We then implemented the same “ Specificity ” and “ Hybrid ” weighting schemes to these results
+
+To better illustrate the results, we run the full approach for IPC subclass A42B which is described as being “ Headwear/Hats; Head Coverings ” . These results can be found in Table 5 below. Overall, there are 20,988 patents that contain this particular IPC subclass. After running the keyword extraction program through these patents, we find that the 5 most common keywords are “ utility model ” , “ cap ” , “ hat ” and “ helmet ” and “ head ” . We then used “ exact string ” matching to get the corresponding SITCs. Once again, the end result matches closely with our own preconceptions of the industries that use headwear technology or whose industries
+
+12
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+might be used to make headwear technology. The next section compares how our software-based methodology stacks up against the existing concordances.
+
+## 5 Comparison with Existing Concordances
+
+In this section, we use the ALP-DM and ALP-PM approaches with hybrid and specificity weights to generate concordances that are structurally comparable to two existing concordances and the devise tests for how well the ALP concordances fit the familiar concordances. Given that the two comparison concordances – the YTC and the DG concordance – are structurally very different, we view these tests as complementary. Specifically, we consider the YTC test to be the best high resolution test of how well the ALP approach can match careful human classification since it is based on patent examiners' classification of patent applications into high resolution industrial categories and provides probabilistic weights that are directly comparable to ALP concordances. The DG concordance provides a test of how well the ALP concordance can match more aggregate, one-to-one matches.
+
+### 5.1 YTC Comparison
+
+Unfortunately, we cannot identify the exact patents used in the YTC, but we can limit our ALP methodology to only the Canadian patents issued in the same time period between 1978 and 1993. This provides us coverage of more than 350,000 Canadian patents and abstracts (30 % more than was used in the YTC). We then convert the IPC's from those patents into the Canadian SICs using both the ALP-DM and ALP-PM algorithm. Note that our algorithm is more heavily weighted towards tradeable goods, since the specific purpose behind our approach is to convert technology data into specific product-types. The Canadian SICs are comprised of both tradeable and non-tradeable goods (e.g. services), so we expect there to be some inherent differences between the two approaches.
+
+We compare the YTC concordance against the concordances based on each of the three weighting options (unweighted, specificity weight and hybrid weight) for both the ALP-DM and ALP-PM. We do this first for the 4-digit cSIC-E for both the Sector of Use (SOU) and Industry of Manufacture (IOM). Since the YTC is constructed as a mix of 3- and 4-digit cSIC concordances with 4-digit IPC, we aggregate both the YTC and ALP concordances to the 3-digit cSIC. Therefore, in all of the comparisons that follow, our ALP results and the YTC results all concord 3-digit cSIC to 4-digit IPC.
+
+The first ALP-YTC comparison we conduct is provided in Table 6 , a simple cross-tabulation of zero and positive values of the respective results where the off-diagonal elements provide a crude measure of errors. The ALP-DM approach generates matching zero values roughly 75 % of the time and matching positive values 2.4-4.1 % of the time. Conditional on YTC=0, the probability that ALP-DM correctly generates a zero weight is 78-79 % . When YTC>0, the probability that this approach correctly generates a positive weight is 62-65 % . In the case of the ALP-PM results, 90-94 % of the results are matching zero values. The conditional probabilities of a matched zero and matched positive weight, respectively, are 96-98 % and 28-38 % . Table 6 captures a key tradeoff between the type I and type II errors associated with the ALP-DM and
+
+13
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+PM approaches: relative to the ALP-PM approach, the ALP-DM approach produces fewer false positives at the expense of more false negatives.
+
+Next, we compute the difference between the YTC and our ALP results. Given that the majority of these differences are zero due to matching zero values (see Table 6 ), we compare these differences across all possible combinations of 3-digit cSIC and 4-digit IPC excluding matching zero values. This provides a strong test of our results with the YTC. Figure 1 shows the distribution of these differences in standard deviation (of the YTC) units. Several things are noteworthy in this figure. First, these differences are extremely small relative to the standard deviation of the YTC. Even after excluding matching zero values, the vast majority of these differences are less than 10 % of the standard deviation of the YTC. Second, the ALP approach and weighting noticeably affects the fit of the ALP results to the YTC results. The ALP-DM approach produces the smallest errors, which seems consistent with Table 6 results since this figure excludes zero values. In both approaches, the hybrid weights generate the best fit to the YTC results. In the case of the ALP-DM approach with hybrid weights, the bulk of the results are within 2 % of the standard deviation of the YTC. Finally, although it is not clear the differences are significant, the weighted ALP-DM approach appears to better fit IOM than SOU results.
+
+As a final comparison of our ALP results and the YTC, we assess how the fit between the two changes with the number of patents available to process, which is determined by the number of Canadian patents in each IPC subclass (4-digit) from 1978 to 1993. Since the ALP approach is a statistical approach that relies on the law of large numbers, we hypothesize that it will more closely approximate the human classification-based YTC as the number of patents processed increases. For future use of ALP approaches, it is important to demonstrate this pattern and to characterize how the number of patents processed affects the quality of the results. The YTC comparison offers a convenient test of this hypothesis since the number of patents in different IPC subclasses varies widely in these Canadian patents (see x-axis in Figure 2 ). To exploit this variation, we non-parametrically regress the absolute deviation of the YTC with our ALP results – normalized again by the standard deviation of the YTC – on the number of patents processed. This regression (Figure 2 ) confirms that the fit improves as the number of patents processed increases. When the number of patents processed is less 2000, the rate of improvement is very apparent. Beyond this threshold, doubling or tripling the number of patents analyzed does nothing to improve the fit. This result provides a useful benchmark for future applications of the ALP approach, which, incidentally, will almost always include many more patents than are contained in this subset of Canadian patents.
+
+Overall, the comparison of the ALP concordances with the YTC shows some systematic biases that are mainly attributed to the methodological construction of the concordance. Our concordance matches to tradeable classes better than non-tradeable classes. $^7$ While these differences can be seen occasionally at high resolution (e.g., 4-digit), the differences quickly fade
+
+7 As we pushed further into the comparison with the YTC, we ran some basic fixed-effect regressions on the 4-digit weights to identify any specific differences between certain class levels. We found that our algorithmic approach tends to under-weigh most of the non-tradeable cSIC-E (these are cSIC1 greater than 5). This is unsurprising since our algorithm relies most frequently on identifying specific products and goods, and it is much more difficult to match specific services.
+
+14
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+with aggregation. There may be more that could be done to refine the matches on the margin, but we expect these improvements to be modest at best and will instead focus our attention on applying the ALM methodology to other trade and industry classifications.
+
+### 5.2 DG Concordance Comparison
+
+As a second check, we compare the results of our concordance with the DG Concordance constructed by Schmoch et al. (2003) . The DG Concordance linked IPCs to both the NACE and ISIC (Rev. 3) classification system using a one-to-one mapping of 4-digit IPC groups into 44 different manufacturing fields, which are then assigned to one or more ISICs. The assignment of IPCs to manufacturing fields was based on the industry of operation of firms filing patent applications. The DG Concordance used more than 3,000 applicant firms that accounted for more than 150,000 patents from 1997 to 1999. Once they identified the industry of the firm, they summed up the IPC counts of the patents filed by the firm and assigned the largest IPC weight a one-to-one match with the manufacturing field.
+
+We generated ALP concordances for IPC-ISIC (Rev. 3), then aggregated the 4-digit ISICs to match the 44 industry fields used in the DG approach. The overall correlation between our ALP weights and the DG weights – which are binary indicators for whether a given IPC subclass is included in an industrial field (1) or not (0) – ranges from 0.36 to 0.53 depending on the methodology used. Given the structural differences between these approaches (i.e., one-to-one matching versus probabilities), these correlations seem quite encouraging. Beyond this overall correlation, we find that the ALP approach matches the DG concordance better in some fields than in others. The ALP concordances matched better with well-defined industrial fields such as "Tobacco", "Wood Products" and "Accumulators", but matched less well with more broadly defined industry types such as "Non-specific Machinery", "Agricultural Machinery" and "Electrical Components".
+
+As a more quantitative comparison of these differences across industrial fields, we take the average weight for each field across all of the IPCs within that field. For instance, Field 1 has 19 different IPCs associated with that specific field. The DG Concordance assigns a 100 % weight for each those 19 IPCs into ISIC (Rev. 3). We compute a similar average for the ALP concordances by taking the average weight of these same 19 IPCs (and similarly for each field). Table 7 summarizes these average weights for each field using the different weighting schemes. As another comparison, we compare the mean ALP weight for all IPC subclass-field pairs that are not matched by the DG concordance (i.e., DG=0) with the mean ALP weight for those that are matched by the concordance (i.e., DG=1). As shown in Table 8 , the differences between these mean weights are statistically significant for all ALP approaches and particularly stark for the ALP-DM approach with hybrid weights. For a final comparison, we rank order the weights within each IPC subclass across fields and compare these ranked weights to the binary DG weights for these subclasses. The final two columns of Table 8 show how frequently the three largest ALP weights for a given IPC subclass include the IPC-field linkage implied by the DG (i.e., DG=1). Roughly 50 % of the time the ALP-DM with hybrid weights captures the DG match in the top three.
+
+15
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Taken together, these comparisons seem to indicate that ALP concordances – especially from the ALP-DM approach – provide a reasonably strong match to the DG concordance. With this comparison in mind, it is worth noting that there are added benefits to the ALP approach relative to the DG approach. If one is interested primarily in the 44 fields contained in the DG concordance, the ALP approach generates a probability structure that in many contexts is preferable to the one-to-one binary matches of the DG concordance. Potentially even more important, the ALP approach provides much more disaggregated linkages that enable economic data to speak to patent data at a much higher industrial resolution if necessary.
+
+## 6 Sample Analysis of Disaggregated Patent & Trade Flows
+
+ALP concordances offer a promising way to jointly analyze patents and economic data at an unprecedented level of resolution. To demonstrate one potential use of this concordance, we analyze how bilateral patent flows are related to bilateral trade flows. We expect patent flows between two countries to be highly correlated with trade because the fixed costs and benefits of both activities are similar and the two are closely linked with technology transfer (Coe and Helpman 1995; Eaton and Kortum 1996) . Previous analyses of international patent flows have relied heavily on a “gravity” model of trade, where bilateral patent flows are determined by the economic size of the countries (i.e. GDP), distance and other country-specific factors. These studies have all looked at the aggregate flows (Harhoff et al. 2007; Bosworth 1983; Eaton et al. 2004; Slama 1981) with no breakdown of industry-level or sectoral differences. A more detailed analysis of the same topic at the industry level can yield additional insights into international patenting strategies across the different industries and technologies. $^8$
+
+To make this comparison, we use bilateral trade flows from the UN-Comtrade database organized by 4-digit SITC (Rev. 2) and bilateral patent flows from the PATSTAT database organized by 4-digit IPC. To use these disaggregated patent and trade data jointly, we concord the patents to the 4-digit SITC (Rev. 2) using the ALP-DM approach with hybrid weights. Thus, the bilateral patent flows associated with a given 4-digit SITC are computed as weighted bilateral patent flows of the 4-digit IPCs that concord to the SITC in the ALM-DM concordance, which provides the weights on each of these IPCs. In addition to trade flows, our gravity model specification includes country-specific variables such as the origin and destination country GDP (obtained from World Bank Indicators) as well as some industry-specific measures. 9 As an extension of our basic specification, we bifurcate our sample using the Broad Economic Classification (BEC) system 10 to see how patent flows differ across different industry types. After applying the concordance and organizing all the variables, we are left with 634 different 4digit SITC industries that filed for patents in at least one of 68 possible destination countries
+
+8 In a survey of the literature regarding patents as measures of technological change, Basberg (1987) notes that patents applied for abroad are most likely to be highest quality patents due to time and costs involved with the application process. Similar statements acknowledging the value of foreign patents were made in Putnam (1996).
+
+9 We use the elasticity of substitution measures obtained from Broda and Weinstein (2006) which are organized by 4-digit SITC
+
+10 Provided by the United Nations (2002 Version). Note that several SITC's qualify under multiple BEC classifications, in which case we still counted that SITC among each bifurcated group. Hence, the total number of observations from the bifurcation will exceed the total number for all industries.
+
+16
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+between 2001 and 2005 (14,442,520 possible observations). Table 9 provides a summary of the variables used.
+
+At the disaggregated level, there is a much higher proportion of zero patents than zero trade. In the cases where both trade and patent flows were zero between the two countries, these observations were dropped. Otherwise, we retained the observation and treat the zeros as informative. Since we are regressing count data (patents), we run a pseudo (Poisson) maximum likelihood (PPML) regression as recommended in Santos Silva and Tenreyro (2006) . $^11$
+
+We first estimate a gravity equation of aggregate country-level bilateral patent flows based on GDP and country-level trade barriers, such as distance and border effects. We provide this as comparison with the previous studies looking at international patenting flows. The GDP measures are intended to capture market supply, demand and absorption capabilities for new technologies of the origin and destination countries, while the gravity terms capture the transaction costs of doing business abroad. We then estimate the same equation and include bilateral trade. We expect countries which are more economically integrated to more readily file patent applications with each other (after controlling for trade barriers and market sizes) since firms' incentives to protect innovations in foreign markets are increasing in export revenues earned in those markets.
+
+The aggregate results in Table 10 , which are based on total bilateral patent and trade flows (i.e., not disaggregated using the ALP-DM concordance), provide a benchmark for the (shaded) disaggregated specifications that use the concordance. Column (1) provides the PPML estimates for the simple gravity model of patent flows. The overall fit of the gravity model is quite high, with market size playing the largest role in determining patent flows. The regression also shows that besides distance, none of the other gravity variables are significant and distance is only significant at the 5 % level. These findings are similar to the Eaton et al. (2004) study that also found a low elasticity of patent flows with respect to distance. Columns (2) and (3) include bilateral trade and the same country-specific variables. We can see that trade is positively related to patent flows with an estimated trade elasticity of 0.41-0.56 at the aggregate level. When similar regressions are estimated at the disaggregate 4-digit SITC level (Columns 3 through 7), we see some noticeable differences in the values of the coefficients. Across all industries, the trade elasticity decreases to 0.24-0.28 and 0.15-0.17 without and with industry fixed effects, respectively. The decline in the elasticity at the disaggregate level seems to imply that while trade flows continue to shape patent application decisions, other industry-specific factors enter importantly into these patenting decisions once we can model the relationship at higher resolution. Specifically, GDP and other gravity variables (e.g., distance, common language, colonial relationship) play a larger and statistically clearer role at the disaggregate level.
+
+As an additional exercise, we further leveraged the ALP-DM concordance to estimate the full disaggregated model in column (7) by subsamples as defined by selected Broad Economic
+
+11 In addition to the Poisson regression, we also experimented with OLS using ln ( Patents + a ) where $a$ is a relatively small constant. We also ran similar regressions using the Zero-Inflated Poisson (ZIP) regressions. The results from these estimations are qualitatively similar to our current estimates and are available upon request.
+
+17
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Categories. $^12$ This allows us to investigate whether there are any noticeable differences in patenting behavior across broad industries. Specifically, we break our sample into three BEC classes across which we expect there to be differences in how firms use patents: Industrial Supplies (BEC-2), Capital Goods (BEC-4), and Consumer Goods (BEC-6). Since Industrial Supplies encompass primarily intermediate goods and Consumer Goods encompass primarily final goods, we expect bilateral exports to shape bilateral patent flows more in the former than the latter. This pattern is evident in our results (Table 11 ). The trade elasticities for capital and consumer goods are nearly twice as large as the elasticity for Industrial Supplies. Once we can match patent flows to SITC and subsequently to BEC, we also see that the destination country's market size plays a larger role in patent flows for capital and consumer goods and that capital goods are more sensitive to geographic barriers such as distance and border effects.
+
+There is obviously much more that could be done to push this analysis further, which is the focus of ongoing research. Our objective here is simply to illustrate how an ALP concordance might be used to better understand determinants of international patenting strategies. In addition to enabling joint analysis of disaggregated patent and trade flows, such a concordance opens other modeling possibilities because many other data sources are structured using economic classifications such as SITC and ISIC. Finally, note that while this sample analysis involves model estimation, there are many descriptive analyses that are enabled by ALP concordances that are potentially just as insightful and policy relevant. For example, these concordances make it possible to add layers of economic and industrial activity to standard patent landscapes, making it easier to detect key innovation trends and patterns in specific fields.
+
+## 7 Conclusion
+
+There is a long and important literature that uses patents to understand the innovation and diffusion of technology. While economists have made important contributions to this field of inquiry, economic analyses of patents have often been constrained by the mismatch between patent and economic data. The ALP methods we develop in this paper enable patents and economic data to speak to each other at an unprecedented level of disaggregation.
+
+There are many policy-relevant questions that could be addressed by joint, high resolution analyses of patent and economic data, including both descriptive exercises (e.g., enhanced patent landscapes) and more rigorous model estimation (e.g., dynamics models of the economic impacts associated with innovation, international technology transfer and patenting strategies, etc.). By making the ALP concordances we have constructed widely available to the research community and continuing to refine these methods as yet more powerful algorithmic tools are developed, we hope to enable these kinds of industry-level analyses in order to complement the insightful but scarce firm-level patent data and analyses that exist.
+
+In this paper, we have developed and tested two ALP approaches to constructing concordances along with various weighting options. Based on testing these approaches against existing
+
+12 These are provided by the UN and are constructed based on SITC categories. For more details, see http://unstats.un.org/unsd/pubs/gesgrid.asp?id=331 .
+
+18
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+concordances, the data mining approach (ALP-DM) with hybrid weights outperforms the probability matching approach (ALP-PM). For near term research, concordances based on the ALP-DM approach with hybrid weights will provide the most reliable means of linking patents to economic data. With continued advances in text and semantic analysis tools and richer databases, however, new possibilities will emerge for building these linkages at yet greater levels of disaggregation. For example, an enhanced ALP-PM approach may soon be able to match individual patents to economic classifications – or even to actual products or processes that use the invention. Although effectively leveraging high resolution linkages like this will demand real research creativity, we believe the potential gains associated with a flurry of creative work on this frontier are extraordinary.
+
+19
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+## REFERENCES
+
+Aghion, Phillipe and Peter Howitt, "A Model of Growth Through Creative Destruction," Econometrica, Vol. 60, No. 2, March 1992.
+
+Andrews, Nicholas and Edward Fox, "Recent Developments in Document Clustering," Working Paper , October 2007.
+
+Acharya, Ram and Wolfgang Keller, "Technology Transfer Through Imports," Canadian Journal of Economics, Vol. 42, No. 4, November 2009.
+
+Barro, Robert and Jong-Wha Lee, "A New Data Set of Educational Attainment in the World: 1950-2010," NBER Working Paper No. 15902, April 2010.
+
+Basberg, Bjorn, "Patents and the Measurement of Technological Change: A Survey of the Literature," Research Policy , Vol. 16, No. 2-4, August 1987.
+
+Bosworth, Derek, "Foreign Patent Flows To and From the United Kingdom," Research Policy, Vol. 13, No. 2, 1984.
+
+Branstetter, Lee and Raymond Fisman and C. Fritz Foley, “Do Stronger Intellectual Property Rights Increase International Technology Transfer? Evidence from U.S. Firm-Level Panel Data,” Quarterly Journal of Economics, Vol. 121, No. 1, February 2006.
+
+Broda, Christian and David Weinstein, "Globalization and the Gains from Variety," Quarterly Journal of Economics, Vol. 1212, No. 2, 2006.
+
+Coe, David and Elhanan Helpman, "International R&D Spillovers," European Economic Review, Vol. 35, No. 5, May 1995.
+
+Comanor, William and F. M. Scherer, "Patent Statistics as a Measure of Technical Change," Journal of Political Economy, Vol. 77, No. 3, May-June 1969.
+
+Eaton, Jonathan and Samuel Kortum , “Trade in Ideas: Patenting and Productivity in the OECD,” Journal of International Economics , Vol. 40, No. 3-4, May 1996.
+
+Eaton, Jonathan and Samuel Kortum, "International Patenting and Technology Diffusion: Theory and Measurement," International Economic Review, Vol. 40, No. 3, August 1999.
+
+Eaton, Jonathan and Samuel Kortum , “Technology, Geography and Trade,” Econometrica , Vol. 70, No. 5, September 2002.
+
+Eaton, Jonathan and Samuel Kortum, Josh Lerner, “International Patenting and the European Patent Office: A Quantitative Assessment,” Patents, Innovation and Economic Performance: OECD Conference Proceedings , 2004.
+
+20
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Evenson, Robert and Jonathan Putnam, "Inter-Sectoral Technology Flows: Estimates from a Patent Concordance with an Application to Italy," Yale University Mimeo, 1994.
+
+Falvey, Rod and Neil Foster, "The Role of Intellectual Property Rights in Technology Transfer and Economic Growth: Theory and Evidence," UNIDO Working Paper, 2006.
+
+Griliches, Zvi, "Patent Statistics as Economic Indicators: A Suvery," Journal of Economic Literature, Vol. 28, December 1990.
+
+Grossman, Gene and Elhanan Helpman, "Innovation and Growth in the Global Economy," MIT Press, 1991.
+
+Harhoff, Dietmar and Karin Hoisl, Bettina Reichl, Bruno Van Pottelsberghe De La Potterie, “Patent Validation at the Country Level: The Role of Fees and Translation Costs,” Research Policy, Vol. 38, No. 9, 2009.
+
+Jaffe, Adam and Manuel Trajtenberg and Rebecca Henderson, "Geographic Localization of Knowledge Spillovers as Evidenced by Patent Citations," Quarterly Journal of Economics, Vol. 108, No. 3, August 1993.
+
+Johnson, Daniel, "The OECD Technology Concordance (OTC): Patents by Industry of Manufacture and Sector of Use," OECD Science, Technology and Industry Working Papers, March 2002.
+
+Keller, Wolfgang, "International Technology Diffusion," NBER Working Paper #8573. October 2001.
+
+Kortum, Samuel and Jonathan Putnam, "Assigning Patents to Industries: Tests of the Yale Technology Concordance," Economic Systems Research, Vol. 9, No. 2, 1997.
+
+Moehrle, Martin and Lother Walter, Isumo Bergmann, Sebastian Bobe and Svenja Skrzipale , “Patinformatics as a Business Process: A Guideline Through Patent Research Tasks and Tools,” World Patent Information , Vol. 32, 2010.
+
+Park, Walter, "International Patent Protection: 1960-2005," Research Policy, Vol. 37, No.4, May 2008.
+
+Putnam, Jonathan, "The Value of International Patent Protection", Yale University Dissertation, 1996.
+
+Romer, Paul, "Endogenous Technological Change", Journal of Political Economy, Vol. 98, No. 5, October 1990.
+
+Santos Silva, J.M.C and Silvana Tenreyro, "The Log of Gravity," The Review of Economics and Statistics, Vol. 88, No. 4, November 2006.
+
+21
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Schmoch, Ulrich, Francoise LaVille, Pari Patel, and Rainer Frietsch , “Linking Technology Areas to Industrial Sectors: Final Reports to the European Commission,” DG Research , November 2003.
+
+Slama, Jiri, "Analysis by Means of a Gravitation Model of International Flows of Patent Applications in the Period 1967 - 1978," World patent Information, Vol. 3, No. 1, 1981.
+
+Verspagen, Bart, Ton van Moergastel, and Maureen Slabbers, "MERIT Concordance Tables: IPC-ISIC (Rev. 2)," MERIT Research Memorandum, February 1994.
+
+Xu, Bin and Eric Chiang, "Trade, Patents and International Technology Diffusion," Journal of International Trade and Economic Development, Vol. 14, No. 1, 2005.
+
+22
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+![Figure](figures/figure_002.png)
+
+Figure 1 Kernel densities of differences between the YTC and ALP results in standard deviation (YTC) units excluding matching zero values.
+
+23
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+![Figure](figures/figure_002.png)
+
+Figure 2 Non-parametric LOWESS regression of the normalized absolute deviation of ALP results from the YTC (IOM) as a function of the number of patents analyzed (i.e., the number of Canadian patents in the 1978-93 window by IPC subclass (4-digit)). Tick marks along x-axis depict the distribution of the number of patents analyzed.
+
+24
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 1 Concordance for IPC subclass B64D which is “ Aircraft; Aviation; Cosmonautics / Equipment for Fitting in or to Aircraft ”
+
+<table><tr><td colspan="2">Initial Concordance: IPC-cSIC</td><td colspan="2">Composite Concordance: IPC-cSIC- ISIC</td></tr><tr><td>Description</td><td>Weight</td><td>Description</td><td>Weight</td></tr><tr><td>Aircraft and Aircraft Parts Industry</td><td>43.2%</td><td>Manufacture of other fabricated metal products; metal working service activities</td><td>10.8%</td></tr><tr><td>Other Communication and Electronic Equipment Industries</td><td>9.4%</td><td>Manufacture of motor vehicles</td><td>10.8%</td></tr><tr><td>Other Machinery and Equipment Industries</td><td>6.3%</td><td>Manufacture of bodies (coachwork) for motor vehicles; manufacture of trailers and semi-trailers</td><td>10.8%</td></tr><tr><td>Indicating, Recording and Controlling Instruments Industry</td><td>5.8%</td><td>Steam and air conditioning supply</td><td>10.8%</td></tr><tr><td>Other Textile Products Industries</td><td>5.0%</td><td>Freshwater fishing</td><td>1.4%</td></tr><tr><td>Electrical Switchgear and Protective Equipment Industry</td><td>2.9%</td><td>Marine aquaculture</td><td>1.4%</td></tr></table>
+
+
+25
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 2 Example search terms used for SITC Industry Descriptions
+
+<table><tr><td>SITC Code</td><td>SITC Full Description</td><td>Search Terms</td><td>“Not” Search Terms</td></tr><tr><td>8484</td><td>Headgear and fitting thereof, nes</td><td>“Headgear”, “Head Gear”, "Helmet"</td><td></td></tr><tr><td>8510</td><td>Footwear</td><td>“Footwear”</td><td></td></tr><tr><td>8710</td><td>Optical instruments and apparatus</td><td>“Optical Instruments”, “Eyeglasses”</td><td></td></tr><tr><td>8720</td><td>Medical instruments and appliances, nes</td><td>“Medical Instrument”, “Medical Appliance”</td><td></td></tr><tr><td>8731</td><td>Gas, liquid and electricity supply or production meters; etc</td><td>“Gas Meter”, “Liquid Meter”, “Electric Meter”</td><td>“Part”</td></tr><tr><td>8732</td><td>Counting devices non-electrical; stroboscopes</td><td>“Counting Device”, “Stroboscope”</td><td>“Part”, “Electric”</td></tr><tr><td>8741</td><td>Surveying, navigational, compasses, etc, instruments, nonelectrical</td><td>“Surveying Equipment”, “Surveying Instrument”</td><td></td></tr></table>
+
+
+Table 3 Illustration of Weighting Schemes
+
+<table><tr><td>SITC</td><td>IPC</td><td>Match</td><td>IPC Total</td><td>\shortstack{Absolute</td></tr><tr><td>Frequency}</td><td>Specificity</td><td>\shortstack{Hybrid</td></tr><tr><td>Weights}</td><td>\shortstack{Hybrid</td></tr><tr><td>Weights}</td></tr><tr><td>1</td><td>A</td><td>100</td><td>100</td><td>10\%</td><td>91.74\%</td><td>55.25\%</td></tr><tr><td>1</td><td>B</td><td>900</td><td>10,000</td><td>90\%</td><td>8.26\%</td><td>44.75\%</td></tr></table>
+
+
+26
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 4 IPC Frequency for Industry Group, "Headgear and Fitting Thereof" (SITC 8484)
+
+<table><tr><td>\textbf{IPC}</td><td>\textbf{Raw}</td><td>\textbf{Specificity}</td><td>\textbf{Hybrid}</td><td>\textbf{IPC Description}</td></tr><tr><td>A42B</td><td>43.1\A42C</td><td>1.5\A62B</td><td>5.2\B68B</td><td>0.1\F41H</td><td>1.7\B63C</td><td>1.6\</td></tr></table>
+
+
+27
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 5 ALP-PM Approach Example for IPC Class A42B
+
+<table><tr><td>IPC Number</td><td>A42B</td><td></td><td></td></tr><tr><td>IPC Description</td><td>Headwear -- Hats; Head Coverings</td><td></td><td></td></tr><tr><td>Top Keywords</td><td>“Helmet”, “Utility Model”, “Cap”, “Hat”, “Head”</td><td></td><td></td></tr><tr><td>\# of Patents Analyzed</td><td>20,988</td><td></td><td></td></tr><tr><td></td><td>Raw Weight</td><td>Specificity Weight</td><td>Hybrid Weight</td></tr><tr><td>8484 - Headgear and fitting thereof, nes</td><td>65.0\%</td><td>13.8\%</td><td>72.3\%</td></tr><tr><td>6576 - Hat shapes, hat-forms, hat-bodies and hoods</td><td>19.0\%</td><td>7.4\%</td><td>11.4\%</td></tr><tr><td>6571 - Articles of felt, nes</td><td>8.1\%</td><td>-</td><td>-</td></tr><tr><td>8421 - Overcoats</td><td>7.9\%</td><td>-</td><td>-</td></tr><tr><td>6579 - Special products of textile material</td><td>-</td><td>20.6\%</td><td>-</td></tr><tr><td>6517 - Yarn of regenerated fibres</td><td>-</td><td>20.1\%</td><td>6.7\%</td></tr><tr><td>6577 - Wadding, wicks</td><td>-</td><td>14.5\%</td><td>4.6\%</td></tr><tr><td>6543 - Woven fabric of wool or fine hair, nes</td><td>-</td><td>12.8\%</td><td>5.0\%</td></tr><tr><td>6581 - Textile material used for packing of goods</td><td>-</td><td>7.9\%</td><td>-</td></tr><tr><td>6121 - Articles of leather used in mechanical appliances</td><td>-</td><td>3.0\%</td><td>-</td></tr></table>
+
+
+28
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 6 Cross-tabs of zero and positive values of YTC and ALP results (both DM and PM results are based on hybrid weights)
+
+
+
+
+29
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 7 Comparison of the DG and ALP Concordances across industrial field
+
+<table><tr><td></td><td></td><td colspan="4">ALP-DM Approach</td><td colspan="2">ALP-PM Approach</td></tr><tr><td></td><td></td><td>DG</td><td>Raw</td><td>Specificity</td><td>Hybrid</td><td>Raw</td><td>Specificity</td><td>Hybrid</td></tr><tr><td>Field</td><td>Description</td><td>(1)</td><td>(2)</td><td>(3)</td><td>(4)</td><td>(5)</td><td>(6)</td><td>(7)</td></tr><tr><td>1</td><td>Food</td><td>100\2</td><td>Tobacco</td><td>100\3</td><td>Textiles</td><td>100\4</td><td>Wearing</td><td>100\5</td><td>Leather</td><td>100\6</td><td>Wood Products</td><td>100\7</td><td>Paper</td><td>100\9</td><td>Petroleum</td><td>100\10</td><td>Basic Chemicals</td><td>100\11</td><td>Pesticides</td><td>100\12</td><td>Paint</td><td>100\13</td><td>Pharmaceuticals</td><td>100\14</td><td>Soaps</td><td>100\15</td><td>Other Chemicals</td><td>100\16</td><td>Man-made Fibres</td><td>100\17</td><td>Plastic Products</td><td>100\18</td><td>Mineral Products</td><td>100\19</td><td>Basic Metals</td><td>100\20</td><td>Metal Products</td><td>100\21</td><td>Energy Machinery</td><td>100\22</td><td>Non-specific Machinery</td><td>100\23</td><td>Agricultural Machinery</td><td>100\24</td><td>Machine Tools</td><td>100\25</td><td>Special Machinery</td><td>100\26</td><td>Weapons</td><td>100\27</td><td>Domestic Appliances</td><td>100\28</td><td>Computers</td><td>100\29</td><td>Electric Motors</td><td>100\30</td><td>Electrical Distribution</td><td>100\31</td><td>Accumulators</td><td>100\32</td><td>Lightening</td><td>100\33</td><td>Other Electrical Components</td><td>100\35</td><td>Telecommunications</td><td>100\36</td><td>Television</td><td>100\37</td><td>Medical Equipment</td><td>100\38</td><td>Measuring Instruments</td><td>100\39</td><td>Industrial Control</td><td>100\40</td><td>Optics</td><td>100\41</td><td>Watches</td><td>100\42</td><td>Motor Vehicles</td><td>100\43</td><td>Other Transport</td><td>100\44</td><td>Consumer Goods</td><td>100\</td></tr></table>
+
+
+30
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 8 Summary of comparison of ALP concordance with the DG concordances (Schmoch et al. 2003) where DG=1 indicates that the DG assigns a particular IPC subclass to an industrial field
+
+<table><tr><td></td><td colspan="2">Mean Weight</td><td colspan="3"> $\%$  ranked in top 3 by IPC subclass</td></tr><tr><td></td><td>DG=0</td><td>DG=1</td><td>t-statistic</td><td>DG=0</td><td>DG=1</td></tr><tr><td colspan="6">Data Mining (ALP-DM)</td></tr><tr><td>Raw</td><td>0.02</td><td>0.07</td><td>24.2</td><td>5.9%</td><td>42.5%</td></tr><tr><td>w/ 2% cutoff</td><td>0.02</td><td>0.13</td><td>37.6</td><td>5.8%</td><td>44.5%</td></tr><tr><td>Specificity</td><td>0.02</td><td>0.07</td><td>14.2</td><td>5.9%</td><td>40.3%</td></tr><tr><td>w/ 2% cutoff</td><td>0.02</td><td>0.11</td><td>19.9</td><td>5.0%</td><td>38.5%</td></tr><tr><td>Hybrid</td><td>0.02</td><td>0.17</td><td>41.9</td><td>5.7%</td><td>52.6%</td></tr><tr><td>w/ 2% cutoff</td><td>0.02</td><td>0.20</td><td>45.0</td><td>5.1%</td><td>47.5%</td></tr><tr><td colspan="6">Probability Matching (ALP-PM)</td></tr><tr><td>Raw</td><td>0.02</td><td>0.06</td><td>14.5</td><td>6.1%</td><td>20.5%</td></tr><tr><td>w/ 2% cutoff</td><td>0.02</td><td>0.06</td><td>14.9</td><td>6.0%</td><td>20.1%</td></tr><tr><td>Specificity</td><td>0.02</td><td>0.04</td><td>8.5</td><td>6.0%</td><td>12.2%</td></tr><tr><td>w/ 2% cutoff</td><td>0.02</td><td>0.04</td><td>8.0</td><td>6.1%</td><td>12.2%</td></tr><tr><td>Hybrid</td><td>0.02</td><td>0.07</td><td>13.7</td><td>6.0%</td><td>20.2%</td></tr><tr><td>w/ 2% cutoff</td><td>0.02</td><td>0.07</td><td>13.3</td><td>5.6%</td><td>19.7%</td></tr></table>
+
+
+31
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 9 Summary statistics for variables used in sample analysis of bilateral trade and patent application flows.
+
+<table><tr><td>Variable</td><td>Source</td><td>Mean</td><td>Min.</td><td>Max.</td><td>Percent Zero</td><td>\# of Obs.</td></tr><tr><td>Bilateral Patent Flows</td><td>PATSTAT</td><td>2.798 (26.520)</td><td>0</td><td>4572</td><td>95.8</td><td>14,442,520</td></tr><tr><td>Bilateral Trade Flows</td><td>UN-Comtrade</td><td>7010.57 (121965)</td><td>0</td><td>3.65e7</td><td>70.5</td><td>14,442,520</td></tr><tr><td>Origin \</td><td>Destination GDP</td><td>WB Indicator</td><td>5.35e11 (1.50e12)</td><td>8.15e8</td><td>1.26e13</td><td>-</td><td>14,442,520</td></tr><tr><td>Elasticity of Substitution</td><td>Broda \</td><td>Reinstein (2006)</td><td>5.85 (13.642)</td><td>1.1</td><td>131.5</td><td>-</td><td>9,197,184</td></tr><tr><td>Distance</td><td>CEPII</td><td>7080.58 (4952.76)</td><td>60</td><td>19,772</td><td>-</td><td>14,442,520</td></tr><tr><td>Border</td><td>CEPII</td><td>0.035 (0.185)</td><td>0</td><td>1</td><td>-</td><td>14,442,520</td></tr><tr><td>Common Language</td><td>CEPII</td><td>0.114 (0.318)</td><td>0</td><td>1</td><td>-</td><td>14,442,520</td></tr><tr><td>Colony</td><td>CEPII</td><td>0.027 (0.163)</td><td>0</td><td>1</td><td>-</td><td>14,442,520</td></tr></table>
+
+
+Note: Standard Deviations are in parenthesis. The Means and Standard Deviations for bilateral patent flows and trade flows are for the nonzero observations.
+
+32
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 10 PPML regression results for extended gravity model of bilateral patent flows, 20012005. Shaded variables and results use the ALP-DM concordance with hybrid weights to match patents (IPCs) to SITC categories.
+
+<table><tr><td></td><td colspan="3">Aggregate</td><td colspan="4">Disaggregated by 4-digit SITC</td></tr><tr><td>Dependent Variable</td><td>Bilateral Patent Flows (1)</td><td>Bilateral Patent Flows (2)</td><td>Bilateral Patent Flows (3)</td><td>Bilateral Patent Flows (4)</td><td>Bilateral Patent Flows (5)</td><td>Bilateral Patent Flows (6)</td><td>Bilateral Patent Flows (7)</td></tr><tr><td>In Trade</td><td></td><td>0.407*** (0.0335)</td><td>0.563*** (0.162)</td><td>0.276*** (0.116)</td><td>0.170*** (0.009)</td><td>0.237*** (0.0136)</td><td>0.148*** (0.0116)</td></tr><tr><td>In Destination GDP</td><td>1.288*** (0.104)</td><td></td><td>0.761*** (0.127)</td><td></td><td></td><td>1.036*** (0.0318)</td><td>1.116*** (0.0293)</td></tr><tr><td>In Origin GDP</td><td>1.190*** (0.101)</td><td></td><td>0.718*** (0.146)</td><td></td><td></td><td>0.921*** (0.0312)</td><td>0.984*** (0.0305)</td></tr><tr><td>In Elasticity of Substitution</td><td></td><td></td><td></td><td></td><td></td><td>-0.187*** (0.0378)</td><td>-0.0236 (0.0373)</td></tr><tr><td>In Distance</td><td>-0.296* (0.137)</td><td></td><td>0.00167 (0.139)</td><td></td><td></td><td>-0.0908* (0.0388)</td><td>-0.163*** (0.0372)</td></tr><tr><td>Border Dummy</td><td>-0.0662 (0.417)</td><td></td><td>-0.449 (0.422)</td><td></td><td></td><td>-0.258* (0.125)</td><td>-0.174 (0.119)</td></tr><tr><td>Same Language Dummy</td><td>0.392 (0.242)</td><td></td><td>0.222 (0.235)</td><td></td><td></td><td>0.305*** (0.0900)</td><td>0.339*** (0.0788)</td></tr><tr><td>Colonial Dummy</td><td>-0.696 (0.373)</td><td></td><td>-0.304 (0.321)</td><td></td><td></td><td>-0.575*** (0.120)</td><td>-0.630*** (0.104)</td></tr><tr><td>Year Fixed Effects</td><td>Yes</td><td>Yes</td><td>Yes</td><td>Yes</td><td>Yes</td><td>Yes</td><td>Yes</td></tr><tr><td>Country Fixed Effects</td><td>No</td><td>Yes</td><td>No</td><td>Yes</td><td>Yes</td><td>No</td><td>No</td></tr><tr><td>Industry Fixed Effects</td><td></td><td></td><td></td><td>No</td><td>Yes</td><td>No</td><td>Yes</td></tr><tr><td>Constant</td><td>-60.27*** (0.728)</td><td>-9.94*** (0.767)</td><td>-44.20*** (5.038)</td><td>-12.93*** (0.671)</td><td>-7.342*** (0.361)</td><td>-55.47*** (1.580)</td><td>-57.57*** (1.555)</td></tr><tr><td>Observations</td><td>22,570</td><td>21,801</td><td>21,801</td><td>4,253,941</td><td>4,253,941</td><td>2,894,659</td><td>2,894,659</td></tr><tr><td>Psuedo  $R^{2}$ </td><td>0.780</td><td>0.966</td><td>0.797</td><td>0.654</td><td>0.747</td><td>0.507</td><td>0.582</td></tr></table>
+
+
+Note: Robust standard errors are in parentheses. Standard errors are clustered by origin-destination pairs in aggregated and origin-destination-industry match in the disaggregated. Industries are denominated by 4-digit SITC (Rev. 2). Patents are matched to SITC using weights generated by the ALP-DM approach with hybrid weights and a 2 % cutoff. Country fixed effects include origin and destination country fixed effects. Industry fixed effects are at the 2-digit SITC level. Significance denoted by: * p < 0:05, ** p < 0:01, *** p < 0:001
+
+33
+
+---
+
+Lybbert and Zolas
+
+August 2012
+
+Table 11 PPML regression results for extended gravity model of bilateral patent flows, 20012005, by selected Broad Economic Categories (BEC). Shaded variables and results use the ALPDM concordance with hybrid weights to match patents (IPCs) to SITC categories.
+
+<table><tr><td>Dependent Variable</td><td>All Industries Bilateral Patent Flows (1)</td><td>Industrial Supplies Bilateral Patent Flows (2)</td><td>Capital Goods Bilateral Patent Flows (3)</td><td>Consumer Goods Bilateral Patent Flows (4)</td></tr><tr><td>In Trade</td><td>0.148***</td><td>0.133***</td><td>0.200***</td><td>0.209***</td></tr><tr><td></td><td>(0.0116)</td><td>(0.0141)</td><td>(0.0271)</td><td>(0.0241)</td></tr><tr><td>In Destination</td><td>1.116***</td><td>1.031***</td><td>1.239***</td><td>1.133***</td></tr><tr><td>GDP</td><td>(0.0293)</td><td>(0.0329)</td><td>(0.0549)</td><td>(0.0383)</td></tr><tr><td>In Origin</td><td>0.984***</td><td>0.968***</td><td>0.986***</td><td>0.985***</td></tr><tr><td>GDP</td><td>(0.0305)</td><td>(0.0372)</td><td>(0.0575)</td><td>(0.0437)</td></tr><tr><td>In Elasticity</td><td>-0.0236</td><td>0.0360</td><td>-0.0492</td><td>-0.0413</td></tr><tr><td>of Substitution</td><td>(0.0373)</td><td>(0.0355)</td><td>(0.0752)</td><td>(0.0930)</td></tr><tr><td>In Distance</td><td>-0.163***</td><td>-0.123**</td><td>-0.232***</td><td>-0.140*</td></tr><tr><td></td><td>(0.0372)</td><td>(0.0447)</td><td>(0.0588)</td><td>(0.0557)</td></tr><tr><td>Border Dummy</td><td>-0.174</td><td>0.00741</td><td>-0.515*</td><td>-0.360*</td></tr><tr><td></td><td>(0.119)</td><td>(0.135)</td><td>(0.205)</td><td>(0.167)</td></tr><tr><td>Same Language</td><td>0.339***</td><td>0.326***</td><td>0.334*</td><td>0.411***</td></tr><tr><td>Dummy</td><td>(0.0788)</td><td>(0.0899)</td><td>(0.140)</td><td>(0.0922)</td></tr><tr><td>Colonial</td><td>-0.630***</td><td>-0.546***</td><td>-0.777***</td><td>-0.660***</td></tr><tr><td>Dummy</td><td>(0.104)</td><td>(0.134)</td><td>(0.172)</td><td>(0.131)</td></tr><tr><td>Year Fixed Effects</td><td>Yes</td><td>Yes</td><td>Yes</td><td>Yes</td></tr><tr><td>Country Fixed Effects</td><td>No</td><td>No</td><td>No</td><td>No</td></tr><tr><td>Industry Fixed Effects</td><td>Yes</td><td>Yes</td><td>Yes</td><td>Yes</td></tr><tr><td>Constant</td><td>-57.57***</td><td>-62.69***</td><td>-62.28***</td><td>-58.05***</td></tr><tr><td></td><td>(1.555)</td><td>(1.826)</td><td>(2.822)</td><td>(2.080)</td></tr><tr><td>Observations</td><td>2,894,659</td><td>1,488,977</td><td>774,233</td><td>753,960</td></tr><tr><td>Psuedo  $R^{2}$ </td><td>0.582</td><td>0.575</td><td>0.639</td><td>0.614</td></tr></table>
+
+
+Note: Robust standard errors are in parentheses. Standard errors are clustered by origin-destination pairs in aggregated and origin-destination-industry match in the disaggregated. Industries are denominated by 4-digit SITC (Rev. 2). Patents are matched to SITC using weights generated by the ALP-DM approach with hybrid weights and a 2 % cutoff. Country fixed effects include origin and destination country fixed effects. Industry fixed effects are at the 2-digit SITC level. Significance denoted by: * p < 0:05, ** p < 0:01, *** p < 0:001
+
+34
+

processed/00100_W1008094143/document.md

@@ -0,0 +1,394 @@
+Dossier
+
+# The Right to Science: Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Lea Shaver
+
+## Abstract
+
+The right to enjoy the benefits of scientific progress has long been neglected, both in theory and in practice. Even scholars, advocates, and jurists deeply involved in the human rights field are likely to express uncertainty as to what the right to science concretely requires... if they are even aware of its existence. This article seeks to remedy that obscurity, providing a highly accessible account of the right to science that is both philosophically grounded and concrete. In short, the right to science calls for treating scientific research, scientific knowledge, and technology as global public goods, to be cultivated for the benefit of humanity and made accessible to all, just as with other socioeconomic rights such as education and healthcare. This article then elaborates what that broad vision means for minimum core content. Particular emphasis is given to reconciling the potential tension between the right to science and intellectual property regimes.
+
+## Résumé
+
+Le droit de bénéficier du progrès scientilfique et de ses applications a pendant longtemps été négligé, tant en théorie qu'en pratique. Même les chercheurs, avocats ou juristes profondément impliqués dans le domaine des droits de l'homme expriment une incertitude quant à ce que le droit à la science requiert concrètement ... si tant est qu'ils aient connaissance de son existence. Cette contribution a pour but de remédier à cette obscurité en apportant des précisions, tant philosophiques que pratiques, relatives au droit à la science. En résumé, le droit à la science appelle à appréhender la science et la technologie en tant que bien public global, à développer au bénéfice de l'humanité et à rendre accessible à tous, au même titre que d'autres droits économiques, sociaux et culturels tels le droit à la santé ou le droit à l'éducation. Cette contribution élaborera par la suite ce que cette vision large du droit à la science aura comme effet sur le contenu obligatoire minimum de ce droit. L'accent sera ainsi mis sur la réconciliation d'une tension potentielle entre le droit à la science et les régimes de propriété intellectuelle.
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights 411
+
+This is the author's manuscript of the article published in final edited form as: Shaver, L. (2015). The Right to Science: Ensuring that Everyone Benefits from Scientific and Technological Progress. European Journal of Human Rights 2015/4, 411-430. http://dx.doi.org/10.2139/ssrn.2564222
+
+---
+
+Dossier
+
+Lea Shaver
+
+## I. Introduction and Summary
+
+Thinking about the right to science is at once very old and very new. It is old, in the sense that the textual basis for this right, and the debates about its inclusion in the Universal Declaration of Human Rights (UDHR), date back to the 1940s.1 During the 1960s and 1970s this debate was rekindled in the context of the Covenant on Economic, Social and Cultural Rights. But in another sense the debate is also still new. The right to science remains today at a very early stage of conceptualization compared to the right to education or the right to health – and even more so compared to freedom of expression or the right to privacy. Even people deeply involved in the human rights field are frequently unaware of the existence of a right to science, much less of its meaning. Thus the right to science is a human right whose conceptual content needs to be both recovered and further developed.
+
+Recognizing this need, the Office of the High Commissioner for Human Rights (OHCHR) organized a Seminar on the Right to Enjoy the Benefits of Scientific Progress and its Applications in October of 2013.2 This Seminar implemented one of the recommendations made by Farida Shaheed, the UN Special Rapporteur in the field of Cultural Rights, in her Report to the Human Rights Council the prior year.3 Shaheed had suggested a participatory process to improve the conceptual clarity of “the right to science and related obligations,” as an area of human rights law that had long been neglected as a matter of both theory and practice. The Seminar proved to be lively and generative, bringing together a global group of experts to share perspectives over two days. This essay reflects some important new insights that came out of that seminar, as well as a later academic workshop hosted by the University of Fribourg, at which the papers of this symposium were presented and developed.
+
+This essay seeks to do two things to contribute to greater conceptual clarity regarding the right to science.
+
+Part II establishes a theoretical foundation for the right through a discussion of its fundamental principles. Toward this end, the discussion emphasizes the animating spirit of science as a public good, with both instrumental and intrinsic value, to be directed toward the service of humanity, guided by values of participation and inclusion. Science and technology have significant power as a means to the end of improving the human situation, but the scientific endeavor also has inherent value as a way in which individuals and communities give expression
+
+1 UN General Assembly, Resolution 217 A (III), 10 December 1948, (A/RES/3/217 A). Article 27(1) state that "Everyone has the right...to share in scientific advancement and its benefits. "
+
+2 Report of the United Nations High Commissioner for Human Rights the seminar on the right to enjoy the benefits of scientific progress and its applications, presented at the twenty-sixth session of the Human Rights Council (1 April 2014) (A/HRC/26/19).
+
+3 Report of the Special Rapporteur in the field of cultural rights Ms. Farida Shaheed on the right to enjoy the benefits of scientific progress and its applications, presented at the twentieth session of the Human Rights Council (14 May 2012) (A/HRC/20/26).
+
+412 Journal européen des droits de I'homme European Journal of Human Rights 2015/4
+
+---
+
+Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Dossier
+
+to a unique aspect of the human personality, much like the arts and other forms of culture. To bring life to these values, however, not just any science will do. Realizing the human rights potential of the science and technology requires a philosophical and practical commitment to science and technology in service of humanity, rather than in service of state power or private profit. In particular, the human rights approach requires a conception of science and technology as public goods, which must be supported and cultivated and made accessible to all people – as with education and health care. To achieve this goal, scientific norms and innovation policy must consciously prioritize broad public participation in the scientific and technological process, and ensure widespread access to new technologies, particularly for the poor and other vulnerable populations.
+
+Yet it is one thing to talk about the general principles and spirit of the right to science, and quite another to define the specific legal obligations and concrete standards it entails. Particularly in the context of socioeconomic rights, it is often difficult to make this translation from abstract principles to concrete legal obligations. To take a contrasting example, freedom of expression also began its legal life as an abstract principle or aspiration. Over the course of centuries, however, this right has benefitted from extensive advocacy, debate, and clarification. The resulting clarity gives us greater confidence today that we understand what “freedom of expression” actually means: that this human right is more than just a rhetorical claim, but is capable of judicial review and carries specific content... even though reasonable people may disagree about some aspects of that content. This much-needed process of advocacy, debate, and clarification, however, is still at an early stage when it comes to the right to science.
+
+To advance this goal of concretization, Part III then proceeds to speak more specifically about what States must do to honor the right to science, by exploring what “minimum core content” might be attributed to this right. Four conceptually distinct approaches to elaborating the minimum core content of human rights are deployed, seeking to translate the high-level principles elaborated in the first part of the essay into more concrete legal obligations. The discussion begins by highlighting the problematic nature of a “core consensus” approach to defining the content of the right to science, at a time when the right still struggles for recognition. Next, the “normative essence” approach is identified as a more promising method, suggesting a concept of “essential technologies” to which all people should enjoy affordable access. Third, the “minimum obligations” approach focuses more explicitly on the duties of States in respect of the right to science, highlighting universal access to clean water, sanitation, electricity, the Internet, and other essential technological services; academic and Internet freedom; protection against the use of technology to abuse privacy or other human rights; public access to publicly funded research; and intellectual property rules that are adopted through a publicly transparent process enabling an appropriate balancing of interests in protection and in access. Finally, the essay proposes a fourth, “pragmatic approach” to minimum core content, which responds to particular issues
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights | 413
+
+---
+
+Dossier
+
+Lea Shaver
+
+and challenges of our time. This section discusses the relevance of the right to science to current battles over access to medicines for addressing the HIV/AIDs crisis, other key conflicts between the right to science and the current expansionist trend in regulation of intellectual property, and debates over Internet governance and freedom.
+
+Throughout this work, I will use the phrase “the right to science” rather than the more formal phrase “the Right to Enjoy the Benefits of Scientific Progress and its Applications” or its common abbreviation “REBSP”. The task before us is one of promoting dialogue and discussion about this right and advancing conceptual clarity; these goals are best facilitated when we have a simple and straightforward way to name what we are talking about. Not so long ago, the human rights community dutifully spoke of “the right of everyone to the enjoyment of the highest attainable standard of physical and mental health.” Fortunately we have by now exchanged that awkward phrasing for the shorter and simpler phrase “the right to health,” without losing sight of the rich complexity of meaning behind this convenient shorthand. The time has come to similarly speak of “the right to science.”
+
+This of course still begs the question of what we mean by “science.” As with any powerful concept – such as “rights,” “equality,” or “law” – the word “science” is subject to many different usages. Indeed, this essay intentionally draws upon multiple meanings of the term. Perhaps the best definition I can offer of the term “science” as I use it in this essay would be: the systematic application of the human powers of inquiry, observation, and reason to better understand the world; often, but not necessarily, with the aim of finding ways to improve it. In the broadest sense, the term “science” is a placeholder for the scientific endeavor, the body of knowledge produced by science, and its technological applications. “Science” in this conception includes anthropology and philosophy as much as medicine and engineering. It is broad enough to encompass traditional knowledge systems and other epistemologies foreign to the university.4 It also holds room for the efforts of amateurs as well as professional scientists who are highly trained in specific traditions. This approach to the concept of science specifically rejects as too narrow the common usage of the term to refer only to specific branches, disciplines, or methods of modern academic enquiry that are empirical, quantitative, or positivist; for example defining “the sciences” in contrast to “the humanities” or “the arts.” Science is a form of human culture, a complex collaborative endeavor of meaning-making and creativity. It inevitably relies upon subjective interpretation and even metaphor, as much as some might like to pretend it can be purely objective, mathematical, or centered in laboratories.5 Whether you come to this essay as a philosopher, a lawyer, or a student, or even
+
+4 B. DE SOUSA SANTOS (ed.), Cognitive justice in a global world: Prudent knowledges for a decent life, Lantham, Maryland, Rowman & Littlefield, 2007.
+
+5 D. McCluskey, "The Rhetoric of Economics", Journal of Economic Literature, vol. 31, n' 2, 1983, p. 481.
+
+414 Journal européen des droits de l'homme European Journal of Human Rights 2015/4
+
+---
+
+Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Dossier
+
+simply an activist who wishes to be thoughtful and reflective about your work, you are in my estimation, a scientist.
+
+## II. Foundational Principles
+
+This section develops four ideas about the foundational principles underlying the right to science. First, it emphasizes the instrumental and intrinsic value of science – both as a means to a technological end, and as a process or activity in which human beings individually and collectively give expression to an important aspect of our humanity. Second, the discussion juxtaposes three conceptions about the aims of science and argues that the human rights vision requires a prioritization of science in service of humanity. Third, the essay emphasizes the importance of recognizing science and technology as global public goods, to be cultivated and encouraged by States, civil society, and the international community for the benefit of all. Fourth and finally, the discussion highlights attention to the touchstone values of inclusivity and participation, both for conducting the scientific process and for ensuring access to its technological fruits.
+
+### A. The Value of Science: Both Instrumental and Intrinsic
+
+Why should science and technology find a place in the international bill of rights? Occasionally, access to science and technology may be fundamental to human survival. This is the case, for example, when we are talking about vital health research, essential medicines, or the technology that supports sanitation services and clean water. These most essential aspects of science and technology, however, are already referenced by other human rights, including the right to health, the right to education, and the right to food. The separate recognition of the right to science implies a further purpose for science well beyond providing for these basic human needs.
+
+The key to uncovering that further purpose lies in looking at the context in which the right to science was enshrined in the international human rights texts. The right to science always appears right beside the right to culture, within the very same article.6 These two concepts are deeply intertwined, so much so that I generally prefer to speak of “the right to science and culture” in a unified sense, because there is so much overlap between the scientific and cultural aspects.7 In the human rights treaties, the right to science and culture always follows immediately after the right to education. This placement is also significant. Unlike the rights to health, housing, or food, access to science and technology is not usually
+
+6 Article 27(1) Universal Declaration of Human Rights; article 15(1-4) International Covenant on Economic, Social, and Cultural Rights.
+
+L. SHAVER, "The Right to Science and Culture", Wisconsin Law Review, vol. 2010, n° 1, 2010, p. 121.
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights | 415
+
+---
+
+Dossier
+
+Lea Shaver
+
+a matter of life and death. It does, however, go to the heart of what kind of life we live. Like education and culture, science and technology hold particular power to improve human life, raise standards of living and promote other human rights. The rights to education, culture, and science have in common a vision of dignified human life and community engagement that goes well beyond mere survival and security needs. Through education, culture, and science, human beings collaborate to realize values of beauty, creativity, the search for truth, and realization of a better tomorrow.
+
+The value of science then, is not purely instrumental. Yes, science and technology also have significant utilitarian value. They can be deployed to solve social problems and improve our material situation. But there is also a value inherent in the process itself, as with the educational process. Engaging in cultural manifestations such as art, literature, music, and theatre helps us to realize and express parts of our shared humanity, which has value from the perspective of individual development and the shared life of the community. Engaging in scientific discovery and technological innovation does as well. Human beings are naturally curious about our world. We seek to understand it. We seek to solve the problems we perceive in it. This is a beautiful and precious part of the human personality. The right to science envisages the scientific and technological endeavor as a process that every person is entitled to participate in – a collective and collaborative process that can help to unite a frequently fragmented world.
+
+## B. SCIENCE IN SERVICE OF HUMANITY
+
+Although science will ideally reflect and serve these humane values, it is important to acknowledge that scientific inquiry and technology are not inherently good. They are rather vehicles that will serve whatever values they are guided by, for good or for evil. The international bill of rights is not neutral as to these values. When the international community first came together to recognize and enshrine a right to science in the post-WWII moment, historical circumstances made them keenly aware of the immense harm that can come from the misuse of science. Science in service of authoritarianism had advanced the ends of violence, torture, murder, and genocide. Nazi scientists had declared the biological inferiority of non-Aryan races and provided the ideological support for “social cleansing” campaigns that would target Jews, homosexuals, and the mentally and physically handicapped, among other minority populations. American scientists had perfected the means to annihilate cities through nuclear attack. Fire bombing, chemical gassing, the atom bomb, and many other technologies for mass murder... these were among the fruits borne through the vision of science in service to the State.
+
+Bearing in mind these bitter lessons, the Universal Declaration of Human Rights articulated a decidedly different vision: that of science in service of humanity. A science that is deployed to alleviate human suffering and to improve the human
+
+416 Journal européen des droits de l'homme European Journal of Human Rights 2015/4
+
+---
+
+Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Dossier
+
+condition. A science that is committed to high ethical standards, conducted always in ways that are respectful of other human rights. By enshrining this right in the Declaration and by establishing UNESCO, the international community articulated an alternative vision for scientific and technological development, one which recognizes and honors our common humanity by advancing norms of dignity, equality, and freedom. When these humane values are placed at the heart of the scientific process, it becomes more likely that the resulting applications will enhance, rather than threaten, the enjoyment of the full range of human rights.8
+
+## C. SCIENCE AS A PUBLIC GOOD
+
+In our own time, however, this vision of science in service to humanity is threatened by a new competing vision: that of science in service of profit. In much of our contemporary public discourse, financial profit and economic growth have come to be seen as both the purpose of science and technological innovation, as well as its primary incentive. This shift in philosophical emphasis corresponds with a decline in public investment in science and increasing support for the privatization of research and the commodification of innovation. To be sure, there are many things that markets and for-profit businesses do more efficiently than governments or the social sector, and private entrepreneurship is an essential pillar of economic welfare and human freedom. Yet the philosophy of science in service of profit seems to me to have the ordering of means and ends backward. Harnessing private enterprise to advance scientific research and technological development is all for the good. But to view scientific research and technological development as the servant to enterprise is to put the cart before the horse, and to unwisely divorce science from its much-needed ethical grounding. Science in service of profit is likely to deliver on its promise of delivering a return on private investment, but it is likely to fail in realizing the larger potential of service to humanity.
+
+What is needed is a renewed political and ethical commitment to the pursuit of science as a public good. To call science a human right is precisely to insist that the supply of scientific knowledge and the development of technology must not be left entirely – or even primarily – to market forces. This is true, in the first instance, because science and technology are dependent upon state support to realize their fullest potential. 9 Just as important is the need for distributive justice. As with health care and education, there is a moral as well as economic value in making science and technology accessible to all, regardless of any particular individual's ability to pay. To claim the right to science is to insist that both the process and the products of science must be understood as public goods intended for the benefit of all, not merely the already privileged, who are best positioned to purchase access in a marketplace. This implies that scientific and
+
+8 R. CLAUDE, Science in the service of human rights, Philadelphia, Pennsylvania, University of Pennsylvania Press, 2002.
+
+9 M. Mazzucato, The entrepreneurial state: Debunking public vs. private sector myths, New York, Anthem Press, 2014.
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights 417
+
+---
+
+Dossier
+
+Lea Shaver
+
+technological research should be the target of public funding, and that innovation policy should prioritize socially valuable ends and the widespread diffusion of technological benefits, especially to benefit vulnerable populations. Science in pursuit of profit will not accomplish this end; science must be ethically grounded in a vision of service to humanity.
+
+## D. Inclusivity and Participation
+
+Ensuring that everyone benefits from scientific and technological progress, however, cannot be a top-down effort. Achieving this goal depends instead upon broad participation in the process of science. Because the forces shaping scientific advancement are complex, technological progress is often mystified. To outsiders, it may seem that scientific progress is a natural process that simply “happens.” To the casual observer, it may seem that a new technology simply “appears,” and a short while later, everyone seems to have one. Scholars in the field of Science and Technology Studies offer an important corrective perspective. This discipline investigates science and technology as products of social processes engaged in by real people. Science and technology, like politics and culture generally, do not proceed inevitably in a predetermined direction. Rather, the path they follow in any particular social and historical context is the product of both the individual choices of scientists and larger social forces. Public policy and legal regulation shape which technologies are pursued and set the conditions under which their spread may be accelerated or delayed. These individual and collective choices can and must be guided by ethical judgments, including a commitment to widespread public benefit.
+
+When these normative choices are obscured or neglected, the scientific process can easily drift from its mission of service to humanity. Technological development may end up catering to only a narrow elite, failing to serve those most in need. The challenge of ensuring that everyone benefits from scientific and technological progress requires broader participation and ethical accountability. The ethical emphasis on participation extends not only to ensuring universal access to the ultimate technological fruits of the scientific endeavor; public participation must also inform the values that guide the scientific process itself. Scientific disciplines and technological fields must ensure that they are truly open to equal participation by women and minority populations. Scientists and technologists should also take up the responsibility to ensure that their work is responsive to social needs, informed by outside perspectives and knowledge, and translated to reach beyond “the ivory tower.” As the scientific process is guided by such values, it becomes more and more likely that the technological results will in fact be useful and accessible to all, promoting lives with dignity, especially for the most vulnerable.
+
+418 Journal européen des droits de l'homme European Journal of Human Rights 2015/4
+
+---
+
+Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Dossier
+
+## III. Minimum core content
+
+So far this essay has offered a view of the right to science as shaped by four foundational principles: recognition of the intrinsic as well as instrumental value of science, an ethical insistence upon science in service of humanity, a political commitment to science as a public good, and an emphasis on the importance of broad participation. The second half of this essay explores what these general principles underlying the right to science imply, in terms of specific legal obligations and policy priorities.
+
+One important tool that scholars and jurists have used to concretize the legal obligations corresponding to various human rights is the concept of “minimum core content.” Socioeconomic rights are subject to the logic of “progressive realization” in the context of resource constraints. Yet the Committee on Economic, Social, and Cultural Rights has repeatedly emphasized that it is also possible to identify a “minimum core obligation to ensure the satisfaction of, at the very least, minimum essential levels of each of the rights.”10 “Minimum core” approaches to human rights interpretation identify specific standards around which there is widespread agreement, which apply even in contexts of very limited resources, or which a nation's failure to honor will be subject to legal censure. The explicit understanding is that these minimum standards are not meant to limit broader understandings of the right. They serve as a baseline or floor, from which upward movement should be continuously pursued. This “minimum core content” approach has been deployed by juridical bodies such as the Committee on Economic, Social, and Cultural Rights (CESCR) and national courts, as well as by human rights scholars, as a method of translating human rights principles into concrete obligations.
+
+One such scholar Katherine Young, has sought to clarify the concept of minimum core content by delineating multiple possible approaches to defining the minimum core content of a right, each of which has precedent in human rights law and Committee practice.11 One approach seeks to locate the “core consensus” content of the right, distinguishing this from marginal aspects of the right, upon which disagreement is to be honored. A second approach seeks to identify the “normative essence” of the right, defining a minimum level of the right that is necessary to honor fundamental principles of dignity, equality, and freedom. The third approach attempts to define “minimum obligations” that States must implement as a matter of priority, or be judged to have violated the right through omission. Each of these approaches has a unique emphasis and offers a unique perspective. Efforts to clarify the right to science should draw on all three of these approaches, ideally with an explicit awareness of their complementarities and limitations.
+
+10 Committee on Economic, Social and Cultural rights, Report on the Fifth Session, 26 November-14 December 1990, (E/1991/23).
+
+11 K. YOUNG, "The Minimum Core of Economic and Social Rights: A Concept in Search of Content", Yale Journal of International Law, vol. 33, n° 1, 2008, p. 113.
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights | 419
+
+---
+
+Dossier
+
+Lea Shaver
+
+The sections that follow apply each of these three approaches to help identify minimum core content for the right to science. This discussion will first highlight some problems with utilizing the “core consensus” approach for the right to science. Next I will recommend the “normative essence” approach as a more promising starting point for this particular human right. Finally, the discussion will examine how to translate the right to science into “minimum obligations.” The article then concludes with a proposed fourth “pragmatic approach” to defining the minimum core content of the right to science.
+
+## A. THE "Core Consensus" APPROACH
+
+The essence of the “core consensus” approach is to locate a minimum core content of a right upon which there is widespread agreement, whereas debate may still exist at the margins of a right. Young describes the “core consensus” approach as being more positivist in nature, since it looks to State practice to identify areas of agreement.12 Because it builds upon consensus, this approach has political advantages for institutions that must carefully tend to their legitimacy. A drawback of this approach is that it may be overly conservative, tending toward the “lowest common denominator,” and thereby failing to adequately defend the interests of vulnerable individuals. Normative consensus can shift to become either more or less accommodating of human rights claims; it can also reflect political considerations at odds with human rights norms.
+
+For example, there was long a stable political and juridical consensus in Brazil that all citizens were entitled to receive prescribed medications free of charge. 13 This consensus went hand-in-hand with an administrative and regulatory structure that emphasized public-sector pharmaceutical research and development and forbade the granting of patents on products important to human health. During international trade negotiations in the 1990s, pharmaceutical industry groups successfully pushed for new international patent rules. As a result, Brazil and many other countries were required to revise their domestic laws to extend patent protection to pharmaceuticals. The prices of medicines have risen significantly as a result, and Brazil's health budgets are now under significant strain. Reflecting this new financial pressure, the political and judicial consensus in favor of free provision of medicines as a basic human right now shows signs of unraveling. 14
+
+This story of access to medicines illustrates several problematic results of the “core consensus” approach as applied to the right to science. First, the consensus on how to balance patent protection and access to medicines has shifted and
+
+12 K. YOUNG, op. cit., pp. 142-144.
+
+15 S. Monica, R. Guise, D. Wang, T. de Campos, "Access to Medicines: Pharmaceutical Patents and the Right to Health", in L. Shaver (ed.), Access to Knowledge in Brazil: New Research on Intellectual Property, Innovation and Development, New Haven, Connecticut, Information Society Project, 2010, p. 103.
+
+14 Ibidem, pp. 103-132.
+
+420 Journal européen des droits de I'homme European Journal of Human Rights 2015/4
+
+---
+
+Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Dossier
+
+continues to shift over time...a nd not always in the direction of expanded sensitivity to human rights. Were this discussion of the right to science taking place thirty years ago, we would naturally have pointed out that there was no international consensus in favor of patent protection for pharmaceutical technologies. Yet today there is as a matter of positive law a strong international consensus that patents must be granted in all fields, including pharmaceuticals. This legal consensus has emerged because multinational companies successfully leveraged international trade negotiations to advance their own financial interests... often at the expense of public interests. 15 It would be a mistake, however, to bless a consensus of State practice produced in this manner with the human rights stamp of approval.
+
+The Brazilian example also highlights a second dynamic: the troubling tendency of an emphasis on consensus to empower restrictive interpretations of human rights. The Venice Statement emphasizes that the right to science is often in tension with intellectual property protections, “which should be managed in accordance with a common responsibility to prevent the unacceptable prioritization of profit for some over benefit for all”.16 The Special Rapporteur in the field of cultural rights has similarly recommended States to “guard against promoting the privatization of knowledge to an extent that deprives individuals of opportunities to take part in cultural life and enjoy the fruits of scientific progress, and consequently to reconsider the current maximalist intellectual property approach....”17 Given the economic value of patent rights to politically powerful actors, however, it is highly unlikely that we will ever observe a consensus in favor of restricting them, no matter how strong the public policy arguments for doing so might be. Emphasizing a consensus approach to defining the minimum core content of the right to science, therefore, could empower powerful groups to successfully oppose recognition of the human rights of the vulnerable.
+
+On the other hand, some aspects of the right to science do already have a stronger consensus behind them. For example, calls to respect academic and scientific freedom, and to enforce safeguards for human research subjects, are ones that admit little disagreement, at least in principle. There is also widespread support in scientific fields for the desirability of open access publishing. Thus in certain areas it may be possible to point to some minimum core content on the basis of a consensus principle. Care must be taken, however, to ensure that the emphasis on consensus does not become a tool for limiting rights, particularly where intellectual property regimes are concerned. The “minimum core” of the right to science should not be confined only to respect for academic freedom and ethical safeguards on research – both of which are already justifiable on other
+
+15 S. SELL, Private Power, Public Law: The Globalization of Intellectual Property Rights, Cambridge, Cambridge University Press, 2003.
+
+16 United Nations Economic, Social and Cultural Organization, Venice Statement on the Right to Enjoy the Benefits of Scientific Progress and its Applications, 16-17 July 2009.
+
+17 Report of the Special Rapporteur on the right to enjoy the benefits of scientific progress and its applications, op. cit.
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights 421
+
+---
+
+Dossier
+
+Lea Shaver
+
+human rights grounds, such as freedom of speech and the right to health. The potentially unique contributions of the right to science, for instance in underscoring the need to cultivate scientific knowledge and research as a public good, and addressing the problems of inequitable access to technology, require looking beyond easy consensus.
+
+In sum, the core consensus approach is problematic when it comes to the right to science, because the recently dominant approach to technology policy so relentlessly emphasizes market orientation, privatization, and exclusivity of access-values antithetical to the grounding principles of the right. The right to science is an area of human rights law where the gap between right and reality looms particularly wide. If we seek to locate the right within an existing political consensus, we may miss it entirely. The project must be understood as one of building consensus around the right to science, rather than recognizing and formalizing a consensus already present.
+
+## B. THE "NORMATIVE ESSENCE" APPROACH
+
+The “normative essence” approach reasons from foundational normative values of dignity, equality, and freedom, to specify a minimum core of each human right that is essential to upholding these values. This may take the form of a “basic needs” emphasis, seeking to define the degree of enjoyment of the right that is necessary and generally sufficient to preserve human life. Or the approach may be more expansive, seeking to guarantee not only survival but also to provide conditions for a broader conception of human flourishing-protecting not just life, but life with dignity. In this second vein, the “human capabilities” approach seeks to define a set of basic entitlements well beyond mere survival, to which every individual has a strong moral claim. Either way, both approaches have in common the desire to specify a degree of enjoyment of the right to which no individual should be denied, which is defined with regard to underlying universal values.
+
+In the area of the right to science, the "normative essence" approach can be applied in several ways.
+
+First, we might define certain “essential technologies” as fundamental to a dignified life, and require States to ensure that these technologies are accessible to all. It would, for example, be easy to place water purification technology, sanitation, and essential medicines on this list. These technologies are understood as important to basic survival. Moving beyond mere survival to include criteria of dignity, equality, and freedom would expand the list of essential technologies further. Electricity, telephone service, and Internet access probably qualify for this more inclusive list of technologies essential for realizing human capabilities. This approach to a minimum core already finds support in the Special Rapporteur’s Report on the right to enjoy the benefits of scientific progress and its applications, which has emphasized that: “A core principle is that innovations
+
+422 Journal européen des droits de I'homme European Journal of Human Rights 2015/4
+
+---
+
+Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Dossier
+
+essential for a life with dignity should be accessible to everyone, in particular marginalized populations.”18 The Report recommends that this goal be achieved through consultation to identify the priority needs of marginalized populations for public subsidies and targeted research,19 as well as through public utilities to ensure universal access to electricity, telephone, and Internet services.20
+
+Second, beyond the emphasis on access to specific technologies – the concrete benefits of scientific progress – the right to science also emphasizes sharing in the process of scientific progress itself. Here, the essential minimum approach points to minimum core content such as access for all to basic scientific education, access to the tools for continually studying the world around them (such as literacy, books, and the Internet), protection of their safety and dignity when they participate as research subjects or are otherwise subjected to new technologies in a context of vulnerability, and consideration of their needs and priorities in shaping the direction of scientific research and technological development. These aspects of minimum core content, too, already find recognition in the Report of the Special Rapporteur.
+
+One virtue of the essential minimum approach to defining a minimum core content is that it works well to focus attention on the basic needs of vulnerable populations. Emphasizing universal access to water and electricity will deliver the greatest benefit to the poorest groups within each society. Yet poverty is not the only dimension of social vulnerability that can be addressed by this approach. From a gender perspective, access to certain technologies can also greatly relieve the disparate psychological, physical, and health burdens placed upon women. Technologies fundamental to gender equality include family planning methods to reduce the health burdens of high-multiples pregnancy, an easily accessible water supply to relieve girls and women of the burden of water-carrying, and modern systems of fuel delivery for cooking food to avoid unhealthy daily exposure to smoke. Women's rights advocates have also pointed out the importance of access to a simple yet often socially taboo technology: the sanitary pad. Women and girls who lack the resources to purchase this modern technology tend to endure shame, miss school, and be socially isolated. Disability advocates could likewise identify certain adaptive technologies as essential to ensuring lives of dignity and equality for persons with special needs.
+
+A unique challenge in applying the “essential minimum” approach to the right to science lies in the special nature of technology as the object of this right. We must take care to guard against two common errors. The first is related to our conception of what counts as technology. The second is related to which technologies
+
+18 Report of the Special Rapporteur on the right to enjoy the benefits of scientific progress and its applications, op. cit.
+
+1s Ibidem.
+
+$^{30}$ Ibidem.
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights 423
+
+---
+
+Dossier
+
+Lea Shaver
+
+qualify as essential. Both of these pitfalls can tempt us to take too limited a vision of the scope of essential technologies.
+
+First, we must not be too narrow in our concept of what qualifies as technology. There is an incredible gap in the level of access to technology enjoyed by the most privileged sector of humanity and the least privileged sector. This might cause us to conceive of technology too narrowly as only the latest and most “cuttingedge” innovations, such as sophisticated smartphones and gene therapy. The technologies of greatest relevance to vulnerable groups, however, are likely to be much more basic and may even be decades old. I have in mind examples such as indoor electricity for lighting, systems for delivering running water to homes and containing waste, and oral rehydration salts to treat acute gastrointestinal illnesses. These are technologies in the sense that they are tools developed by human ingenuity to solve particular problems. Individually and collectively they have improved and saved many millions of lives. To leverage the right to science in a way that is actually useful for marginalized groups, we will need to be broad and inclusive in our conception of technology, both old and new.
+
+Second, we must resist the temptation to be too stingy in our concept of which technologies are essential. Technology advances, and the list of technologies deemed essential to a life of dignity and freedom must expand accordingly. 21 It may feel particularly awkward to recognize newer technologies – such as Internet access – as essential ones. After all, until very recently, everyone made do without that particular technology, and we would not say that the lives we led then were lacking in dignity or freedom. But if “the right to enjoy the benefits of scientific progress and its applications” means anything, it is precisely that these new innovations and technologies are to be enjoyed by all. Fifty years ago, we could not have said that there was a universal human right to antiretroviral therapy for HIV infection; neither antiretrovirals nor HIV was known at the time. Yet today it is not difficult to recognize such medicines as an essential innovation from a human rights perspective. 22
+
+These two cautionary principles operate in a complementary way. The first reminds us not to overlook technologies that might seem too old. The second reminds us not to rule out technologies that might seem too new. An essential technology, from the view of human rights, may be very old or very new. The limiting principle is not the age of the technology, but its importance for promoting human freedom, dignity, and equality.
+
+This last caveat points to one final challenge in applying the "essential minimum" approach to the right to science or any human right: it often remains difficult
+
+21 M. LAND, "Toward an International Law of the Internet", Harvard International Law Journal, vol. 54, n° 2, 2013, p. 393.
+
+22 Tribunal Supremo de Justicia (Supreme Tribunal of Justice – Venezuela), 15 July 1999, Cruz del Valle Bermúdez y otros vs. MSAS s/amparo. Expediente N°. 15.789. Sentencia N°. 196. This case recognized access to state of the art HIV medications as required by the right to life, the right to health, and the right to science.
+
+424 Journal européen des droits de l'homme European Journal of Human Rights 2015/4
+
+---
+
+Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Dossier
+
+to get very specific in defining the content of the right. To translate the right to science as universal access to essential technologies is to replace a very abstract principle with an only somewhat less abstract one. It still remains to be defined which technologies are essential to dignity, freedom, and equality. It also remains to be defined what exactly States must do to ensure that access, ranging from subsidies for research and commercialization to direct procurement or provision. This remaining ambiguity may be a fault or a virtue. On the negative side, we may be left with less clarity and specificity than had been sought. On the positive side, it may be appropriate to leave this clarification and concretization to domestic processes of advocacy, policy-making, and adjudication, in light of particular national priorities and needs.
+
+## C. THE "MINIMUM Obligations" APPROACH
+
+In contrast to the “normative essence” and “consensus core” approaches, the “minimum obligations” approach has been more explicitly focused on defining not the right itself, but the corresponding duties of States. This emphasis is intended to make human rights particularly useful for guiding public policy, to facilitate more effective international supervision, and to enable domestic and regional rights adjudication. The “minimum obligations” approach goes handin-hand with the “violations approach” to human rights, which seeks to define human rights and their corresponding State duties specifically enough to enable their justiciability in particular cases. It may also serve as a framework for priority setting in national policymaking and international cooperation. The minimum obligations approach often builds on the normative essence or consensus core approaches, translating the rights identified there into duties, and identifying which corresponding State duties are most appropriate to insist upon.23
+
+The “minimum obligations” approach can also be related to the effort to distinguish between positive obligations requiring States to act in certain ways that promote the enjoyment of human rights versus negative obligations requiring States to refrain from activities that would prejudice human rights. A more elaborate three-part approach, conceiving of government duties to respect, protect, and fulfill human rights, is often used to elaborate different ways in which government actions or inactions relate to the right. For example, governments have a duty to respect the right to science by refraining from activities that would interfere with academic freedom.24 Mere inaction in respect of the right to science, however, will not go very far to ensuring its enjoyment; active steps are also required. Governments can protect the right to science by ensuring that intellectual property rules are well designed to promote creativity and innovation without unduly sacrificing
+
+$^{23}$ K. YOUNG, op. cit.
+
+34 Y. DONDERS, "The Right to Enjoy the Benefits of Scientific Progress: in Search of State Obligations in relation to Health", Medicine, Health Care and Philosophy, vol. 14, n° 4, 2011, p. 371.
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights 425
+
+---
+
+Dossier
+
+Lea Shaver
+
+participation and access, 25 and by using effective regulatory procedures to protect the safety and dignity of human research subjects. 26 Governments can fulfill the right to science by funding research and development, establishing mechanisms that enhance popular participation in science and providing science education through public schooling and publicly supported media.
+
+Another way to think of the minimum obligations approach is in tandem with the principle of progressive realization. It is well understood that the realization of socioeconomic rights is often significantly constrained by limits on government resources, which differ greatly from country to country. Yet the principle is also well established that certain priority aspects of human rights require immediate implementation by all countries. This may be so because implementation of that priority aspect does not require great resources. Alternatively, even if the resource investment may be substantial, the cost-to-benefit calculus is nevertheless compelling. For example, in the area of the right to housing, minimum obligations include the government duty to respect the right to housing by not conducting illegal evictions. A government can hardly claim that it is too poor to grant due process and consideration for human rights before evicting people from their land or homes.27 Minimum obligations with respect to the right to education have similarly been defined to include providing free and universal primary education.28 No doubt, significant financial resources must be mobilized to comply with this duty. Yet the normative and utilitarian justifications for universal primary education are so overwhelming that a State's failure to do so simply cannot be reconciled as reasonable priority setting.
+
+Applying this approach, the minimum core content of the right to science would include efforts to expand access to technology and opportunities for scientific participation that are highly cost-effective. Access to clean water, sanitation services, electricity and other essential technologies should be universalized. Academic freedom and Internet freedom should be respected. Technology should not be used in ways that abuse privacy or other human rights. Governments should ensure that intellectual property rules are adopted through a publicly transparent process that allows the concerns of authors and the public to be addressed. 29 Scientific publications subsidized by government funding should be made available to the public at large, rather than only through private services
+
+25 Committee on Economic, Social and Cultural Rights, Statement by the Committee on Economic, Social and Cultural Rights (14 December 2001) (E/C.12/2001/15).
+
+24 Y. DONDERS, op. cit
+
+27 Committee on Economic, Social and Cultural Rights, Report on the Sixteenth and Seventeenth Sessions, 28 April-16 May 1997, 17 November-5 December 1997, (E/1991/22).
+
+28 Article 13(2)(a) International Covenant on Economic, Social and Cultural Rights.
+
+20 Report of the Special Rapporteur in the field of cultural rights Ms. Farida Shaheed on copyright policy and the right to science and culture, presented at the twenty-eighth session of the Human Rights Council (24 May 2014) (A/HRC/28/57); United Nations Office of the High Commissioner for Human Rights, Independent expert calls for an end to secret negotiations of free trade and investment agreements until public consultation and participation is ensured and independent human rights impact assessments are conducted (30 March 2015).
+
+426 Journal européen des droits de I'homme European Journal of Human Rights 2015/4
+
+---
+
+Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Dossier
+
+that restrict public access. $^30$ These are just a few examples of highly cost-effective ways of respecting, protecting, and fulfilling the right to science.
+
+## IV. A Pragmatic approach
+
+Katharine Young's work delineating the three major approaches to defining the minimum core content of human rights points out that there are multiple methods and purposes to defining a minimum core. As Young herself suggests, a self-conscious examination of those purposes may help to guide the process. Some familiar motivations for defining a minimum core are to focus public pressure on the most urgent issues, to prioritize the needs of vulnerable populations, to deflate excuses of “limited resources” and “progressive realization,” or to advance human rights along the lines most compatible with preserving institutional legitimacy. Bearing these or similar goals in mind, we may work backward to guide our definition of minimum core content in a way that most effectively addresses these needs. Young does not offer a label for this alternative approach to defining a minimum core, but I propose we think of it as a “pragmatic approach” to minimum core content.
+
+In line with this recognition, I suggest that it is natural and appropriate for efforts to elaborate the minimum core content of a right to be responsive to the particular challenges and issues of the time. An emphasis on particular content of urgency today need not limit efforts to recognize and emphasize other aspects of the right in the future, as new needs and challenges are encountered. For instance, the current emphasis on access to essential medicines is an appropriate and necessary response to a particular human rights crisis of our own time: the deaths of millions of people in the prime of their lives from diseases for which effective treatments exist, but which are being denied in the name of intellectual property. A focus today on assuring access to essential medicines today need not mean that the right to science is inherently tied to pharmaceuticals more so than other forms of technology. It is simply the emphasis of a particularly important and timely aspect of the right. The right to science perspective helps to emphasize that the human rights issue is not only one of ensuring universal access to the drugs that exist today, but also reorienting pharmaceutical policy to better meet the needs of vulnerable populations through future research and development.
+
+The arena of copyright law also reveals urgent conflicts between the privatization of knowledge and the right to science. Digital technology today offers the ability to reproduce and share written works at extremely low cost, unimpeded by traditional geographic barriers or the weak state of book publishing and retail in developing countries. We finally have the tools to end the “book famine” that has
+
+30 Report of the Special Rapporteur in the field of cultural rights Ms. Farida Shaheed on copyright policy and the right to science and culture, op. cit.
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights | 427
+
+---
+
+Dossier
+
+Lea Shaver
+
+traditionally plagued higher education and scholarly research in many countries of the world. Yet there is a conflict between scholars and students who wish to access these works easily and affordably, and the companies that hold the copyrights in them, who wish to obtain as much revenue as possible. One study calculated that to legally purchase the required readings for the first year of university studies in Brazil would cost six to ten months' income at the minimum wage. 31 Such a large gap between price and the ability to pay is not sustainable. Solutions must be found that better balance the needs of authors and readers to promote broader access to scientific knowledge.
+
+Also related is the call to require that scientific research be published on an Open Access model, ensuring its ability to be legally distributed and shared. This call responds to the increasing financial pressures faced by academic libraries even at institutions as wealthy as Harvard University. But it is particularly important to the ability of scholars in developing countries to participate in the scientific process. Because academic works in particular are produced according to incentive structures based on university employment, public subsidy, academic reputation, and the individual desire to contribute to shared knowledge, this is an area in which it makes particular sense to emphasize openness and intellectual freedom over treatment as private property. Similarly, Open Access initiatives for primary and secondary textbooks can help address the textbook shortage that critically undermines education in many developing countries, particularly for children from poorer families. These calls have recently found support in the report of the Special Rapporteur on copyright policy and the right to science and culture. 32
+
+There are also positive developments that need encouragement to be carried forward. The World Intellectual Property Organization (WIPO) recently concluded a treaty designed to expand access to copyrighted works : The Marrakesh Treaty to Facilitate Access to Published Works for Persons Who Are Blind, Visually Impaired, or Otherwise Print Disabled. Previously international treaty-making had focused only on expanding protection for intellectual property. This is the first international instrument designed to ensure that copyright law does not act as a barrier to access and participation. There is great potential in continuing this approach to further advance the right to science. Already debate is underway on international instruments to facilitate additional exceptions and limitations to copyright to assist the work of educational and research institutions and libraries.
+
+The several examples presented above all reflect a common theme. One of the great challenges of our time, to which the right to science must respond, is the privatization of the scientific enterprise and the neglect of public welfare in the name of intellectual property. The emphasis has come to be placed too strongly
+
+21 P. MIZUKAMI et al., "Exceptions and Limitations to Copyright in Brazil: A Call for Reform" in L. SHAVER (ed.), Access to Knowledge in Brazil : New Research on Intellectual Property, Innovation and Development, New Haven Connecticut, Information Society Project, 2010, p. 103.
+
+32 Report of the Special Rapporteur in the field of cultural rights Ms. Farida Shaheed on copyright policy and the right to science and culture, op. cit.
+
+428 Journal européen des droits de I'homme European Journal of Human Rights 2015/4
+
+---
+
+Ensuring that Everyone Benefits from Scientific and Technological Progress
+
+Dossier
+
+on the prioritization of profit and the logic of the market, too often to the neglect of the moral imperative for science to serve human needs. The tendency to regulate intellectual property in the sphere of international trade has aggravated this imbalance. A human rights perspective is urgently needed in this debate. The call to reinvigorate the orientation of the scientific enterprise as a public good in service of humanity and the call to insist upon universal access to the benefits of technology will come from human rights institutions or it will not come at all. The right to science offers a particularly apt and timely framework for reasserting this ethical perspective in the international sphere. Human rights institutions may feel that they have lesser expertise or lesser competency to speak about intellectual property law. That competency must be acquired, just as it was when human rights institutions began to participate in conversations about global public health.
+
+Of course, the challenges of today are not only about intellectual property. The age-old struggle between freedom and despotism continues to play out, in the arena of science and more broadly. Many States are tempted to restrain the academic enterprise or to control their citizens’ Internet use as a means to repress political criticism. This too, must be condemned from the perspective of the right to science. We must ensure that the Internet remains a force for promoting freedom, and that controls are not imposed in the name of national security or intellectual property, which will later be abused to restrict the free exchange of ideas. We also continue to face the challenge of expanding access to education, improving its quality at all levels, and protecting academic independence. The right to science can offer a normative framework for guiding respect for intellectual freedom both online and offline.
+
+## V. Conclusion
+
+By now we are well accustomed to viewing education and health care as public goods, to be publicly supported and made available for the benefit of all. The right to science encourages us to approach science and technology in a similar way. Technology has a great capacity to save and improve lives, when it is directed to those ends. Beyond the utilitarian value of technology, participation in the collective process of scientific and technological development has an intrinsic value – as an opportunity to give expression to our human nature, cultivate the human personality, and build international understanding. For both sets of reasons, it is vital that active efforts be taken to ensure that all people enjoy opportunities to participate in the scientific process and benefit from essential technologies, both old and new.
+
+Translating this broad vision of science in service of humanity into minimum core content is both fruitful and challenging. In some areas, such as academic freedom and protection of research subjects, substantial consensus exists on specific
+
+2015/4 Journal européen des droits de I'homme European Journal of Human Rights | 429
+
+---
+
+Dossier
+
+Lea Shaver
+
+norms. In other areas, particularly with respect to access to technology, the challenge remains one of building consensus. Indeed, the modern direction of international rulemaking around intellectual property has tended to be one that marginalizes and undermines the right to science, rather than respecting and fulfilling it. Here the problem of pharmaceutical patents and access to essential medicines is merely a particularly high-stakes example of the broader tension between the right to science and intellectual property regimes. This tension presents both a challenge to enjoyment of the right to science and an opportunity for human rights institutions to make a difference. The essence of the right to science is to insist that scientific learning and essential technologies be made available to all. Patent and copyright rules must be designed to strike an appropriate balance between incentivizing innovation and creativity and ensuring broad access to scientific knowledge and new technologies. Public funding must fill the gap to ensure that the needs of marginalized groups are being addressed, despite the necessarily lower profit potential. Leveraging the human rights perspective can help these goals to become a reality.
+
+Lea Shaver is Associate Professor of Law at the Indiana University Robert H. McKinney School of Law. She can be reached at lbshaver@iu.edu.
+
+430 Journal européen des droits de I'homme European Journal of Human Rights 2015/4
+

processed/00111_W2894120170/document.md

@@ -0,0 +1,496 @@
+ARTICLE IN PRESS
+
+Information Processing and Management xxx (xxxx) xxx-xxx
+
+![Figure](figures/figure_002.png)
+
+Contents lists available at ScienceDirect
+
+Information Processing and Management
+
+journal homepage: www.elsevier.com/locate/infoproman
+
+![Figure](figures/figure_006.png)
+
+# Capturing information on technology convergence, international collaboration, and knowledge flow from patent documents: A case of information and communication technology
+
+Changjun Lee$^{a}$, Dieter Franz Kogler$^{a}$, Daeho Lee$^{b,*}$
+
+a Spatial Dynamics Lab, School of Architecture, Planning & Environmental Policy, University College Dublin, Ireland
+
+b Interaction Science, Sungkyunkwan University, Seoul, Republic of Korea
+
+## ARTICLE INFO
+
+Keywords: Technology evolution Patent document Knowledge flow Evolutionary trajectory Information and communication technology
+
+## ABSTRACT
+
+In addressing persistent gaps in existing theories, recent advances in data-driven research approaches offer novel perspectives and exciting insights across a spectrum of scientific fields concerned with technological change and the socio-economic impact thereof. The present investigation suggests a novel approach to identify and analyze the evolution of technology sectors, in this case, information and communications technology (ICT), considering international collaboration patterns and knowledge flows and spillovers via information inputs derived from patent documents.
+
+The objective is to utilize and explore information regarding inventors' geo-location, technology sector classifications, and patent citation records to construct various types of networks. This, in turn, will open up avenues to discover the nature of evolutionary pathways in ICT trajectories and will also provide evidence of how the overall ICT knowledge space, as well as directional knowledge flows within the ICT space, have evolved differently. It is expected that this data-driven inquiry will deliver intuitive results for decision makers seeking evidence for future resource allocation and who are interested in identifying well-suited collaborators for the development of potential next-generation technologies. Further, it will equip researchers in technology management, economic geography, or similar fields with a systematic approach to analyze evolutionary pathways of technological advancements and further enable exploitation and development of new theories regarding technological change and its socio-economic consequences.
+
+## 1. Introduction
+
+Patent documents contain a wealth of unstructured data, including multidimensional information, which, once extracted and processed, enables scholars and experts to carry out a variety of exploratory analysis tasks depending on their purposes ( Liu, Liao, Pi, & Hu , 2011 ) . The importance of knowledge as a key asset in an innovation-driven economy is well documented ( Jaffe & Trajtenberg , 2002 ; Teece , 1998 ) . A key challenge for decision makers in charge of managing knowledge assets is the effective exploration and exploitation of technological opportunities via the information stored in patent data ( Abbas, Zhang, & Khan , 2014 ; Bonino, Ciaramella, & Corno , 2010 ; Codina-Filbà et al. , 2017 ) .
+
+The increasing dominance of teams (Wuchty, Jones, & Uzzi, 2007) and levels of complexity in the creation of novel products and
+
+" Corresponding author.
+
+E-mail address: daeho.lee@skku.edu (D. Lee).
+
+https://doi.org/10.1016/j.ipm.2018.09.007
+
+
+Received 30 June 2018; Received in revised form 22 August 2018; Accepted 18 September 2018
+
+
+0306-4573/ c 2018 Elsevier Ltd. All rights reserved.
+
+Please cite this article as: Lee, C., Information Processing and Management, https://doi.org/10.1016/j.ipm.2018.09.007
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+processes of economic value ( Kodama , 1992 ; Patel & Pavitt , 1997 ) have sparked interest in patterns of international collaboration as one effective way to create or recombine new and useful knowledge ( Guan & Chen , 2012 ; Hird & Pfotenhauer , 2017 ; Kim & Park , 2009 ; Rycroft & Kash , 2004 ; Wagner & Leydesdorff , 2005 ; Zhang , 2017 ) . Consequently, understanding trends in patterns of international collaboration and knowledge flows among countries in specific technology sectors, alongside an understanding of how knowledge spillovers facilitate technology convergence processes, has become a focal point in the study of the evolutionary patterns of technological change ( Dosi & Nelson , 1994 ) . Developing a multi-dimensional understanding and analysis framework of a specific technology sector could be even more critical when it comes to rapidly evolving sectors, such as information and communication technology (hereafter, ICT).
+
+Researchers have utilized patent documents to analyze technology trends ( Khasseh, Soheili, Moghaddam, & Chelak , 2017 ; Kim, Suh, & Park , 2008 ; Yoon & Park , 2004 ) , perform tasks of technology forecasting ( Chen, Zhang, Zhu, & Lu , 2017 ; Daim, Rueda, Martin, & Gersri , 2006 ; G. Kim & Bae , 2017 ; Kyebambe, Cheng, Huang, He, & Zhang , 2017 ; Yoon & Park , 2005 ) , investigate relationships among scientists ( Jiang, Shi, An, Yu, & Wang , 2017 ) , and undertake strategic technology planning ( Joung & Kim , 2017 ; Lee, Kim, & Shin , 2017 ; Yu & Zhang , 2017 ) . Still lacking, however, is a holistic understanding of technological change that considers four distinct but interrelated dimensions and evolutionary paths that lead to advances in any specific technology sector. These four dimensions / paths include (1) technology convergence, (2) collaboration networks and knowledge flow among (3) technologies and (4) countries. To the best of our knowledge, no quantitative investigation concurrently employing each of these four dimensions has been conducted so far concerning the evolutionary trajectories of a discrete technology sector. Considering today's landscape, the investigation of trends in collaboration and knowledge flows among countries is of particular relevance when aiming to understand a country's capacity in a specific technology sector.
+
+The present investigation suggests a novel approach for using patent documents to identify and analyze technology convergence, international collaboration patterns, and knowledge flows taking place within a specific technology sector (the specific technology sector herein is ICT) and among inventors residing in different countries. Based on this approach, the objective is to discover the nature of evolutionary paths in ICT, and further, to study the potential different paths of inventor collaboration and knowledge flow patterns over time. To convert this vision to practical application, data from the United States Patent and Trademark Office (USPTO) are utilized with information on inventors' location (i.e., to detect the exact origin of invention), technology classifications (i.e., to delineate the building blocks of an invention), and citations to prior art (i.e., to identify directional knowledge flows). It is expected that the data-driven approach herein will provide intuitive results for decision makers, such as research and development (R&D managers and others, who are concerned with future resource allocation tasks or who are interested in finding potential well-suited collaborators in their quest for the development of next-generation technology. Further, it is also expected that the suggested approach will provide a toolset for researchers to analyze complex matters of technological change more systematically, while also allowing the experts to test existing theories and to exploit new theories along these lines.
+
+The remainder of this study is organized as follows. Section 2 outlines the data and the various methodological approaches employed in the present investigation. Section 3 subsequently highlights the findings of the study, while Section 4 discusses the results and offers concluding remarks.
+
+## 2. Data and methods
+
+### 2.1. Patent documents and data sample
+
+In order to conduct an international patent analysis, patent data from the USPTO were obtained. Patent data from the USPTO is publicly accessible, and due to the size and competitiveness of the US market, these data are quite suitable for international technology and innovation studies (Jaffe, 1986; Jaffe & Trajtenberg, 2002; Lee, 2013). While the limitations of patent data are well understood (Griliches, 1998), patent data also provide a wealth of information. In the context of the present investigation of information on the geo-location and technology field classifications of inventors, patent citations are relevant in order to delineate exactly where an invention originated from, which specific technology fields an invention is related to, and what technology classes provided essential knowledge inputs for the development of a novel product or process of economic value.
+
+Herein the focus is on patent documents related to information and communication technology (ICT) fields via patents applied for at the USPTO from 1980 to 2014. To filter ICT-related patents from the larger dataset, the International Patent Classification (IPC)ICT concordance table developed by Inaba and Squicciarini (2017) was utilized. Originally established by the World International Property Organization (WIPO) and the European Patent Office (EPO), IPC is a patent classification tree based on fundamental knowledge and technology categories. The total number of USPTO patents applied for in the timeframe of interest, regardless of inventors' country of residence at the time of invention, is about 5.5 million. Following application of the IPC-ICT concordance table, the sample results in about 1.8 million patent documents related to ICT technologies over the time period 1980–2014.
+
+### 2.2. Capturing knowledge and collaboration space of countries in ICT
+
+Fig. 1 illustrates how to extract and reorganize the information of interest from patent documents. The example highlights three individual patents. Each of patent #1 and patent #3 were developed by three inventors, respectively, while patent #2 features two co-inventors. Patent #1 was co-invented by two individuals residing in the United States at the time of invention together with one inventor from Germany. On the other hand, patent #2 was a collaboration between one inventor from the US and another inventor from Japan. In addition to co-inventor collaboration patterns, patent documents also provide insights into the technological fields
+
+2
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+![Figure](figures/figure_003.png)
+
+Fig 1. Example in knowledge space and collaboration network from patent document.
+
+wherein the patents are allocated. All three patents in the example displayed in Fig. 1 are classified in three distinct technology classes—that is, technology combinations—of (A, B, C), (B, C, E), and (B, E, F), respectively.
+
+To understand the relationship between ICT technology sub-classes and their recombination patterns, a methodology based on knowledge space ( Kogler, Rigby, & Tucker , 2013 ) is applied by constructing a technology class (or country) co-occurrence matrix. Following the work of Hidalgo, Klinger, Barabási, and Hausmann ( 2007 ) , a variety of scholars have used this methodological approach to outline the relatedness between industries ( Neffke, Henning, & Boschma , 2011 ) , products ( Hidalgo et al. , 2007 ) , and knowledge ( Kogler, Essletzbichler, & Rigby , 2017 ) . In order to construct an ICT knowledge space, a technology class co-occurrence matrix in ICT-related patents needs to first be developed. For instance, technology class B and technology class C occur together in patent # 1 and patent # 2, but class A and class E never co-occur in any patent in our example. Fig. 1 -B and C show the co-occurrence matrix and the network visualization thereof based on the information derived from the three patent documents in this example.
+
+The co-occurrence matrix approach is also utilized to construct parameters for the international collaboration space. Similar to the knowledge space methodology, a country co-occurrence matrix that counts the resident countries of each pair of inventors found in every patent document of interest is constructed. For instance, the US and Germany co-occur in patent #1 (in other words, a collaboration between US and German inventors took place while developing this invention). Similarly, in patent #2 there was a collaboration between an inventor residing in the US and an inventor who lived in Japan at the time of invention. Patent #3 features three inventors from three different countries and thus we find three instances of international collaboration (i.e., Germany and Switzerland, Switzerland and France, and Germany and France). Following this approach, the international collaboration matrix is shown in Fig. 1 -D. This co-occurrence matrix can also be translated into a network visualization, which is displayed in Fig. 1 -E.
+
+### 2.3. Capturing knowledge flow among technologies and countries from patent data
+
+While there is no direction between pairs of technologies or countries in the co-occurrence matrix, there exists both a source and a target in knowledge flow. In other words, knowledge flow shows the origins of the knowledge and the destinations toward which the knowledge flows. We use information on technology elements and patent citations to capture knowledge flow in the two dimensions of technology and country. Fig. 2 illustrates an example of the process in capturing knowledge flow among technologies and countries from a patent. Assuming the knowledge (or the technology) in patent # 1 is influenced by patent # 2 and patent # 3, and assuming patent # 2 is created by applying the knowledge in patent # 3, the knowledge flow among countries is shown to have been started in Germany, Switzerland, and France, to have flowed via the US and Japan, and to have arrived in the US and Germany (see Figs. 2 -D and E). Likewise we can determine the direction of the knowledge flow among technologies, flowing, in our example, from tech (B, E, F) through (B, C, E) to (A, B, C). For instance, knowledge flow from tech C to tech B happens only once in patent # 2 to patent # 1, while flow from tech B to tech C happens three times in patent # 3 to patent # 2, from patent # 3 to patent # 1, and from patent # 2 to patent # 1 (see Fig. 2 -B). This pattern of technology flow is visualized in Fig. 2 -C.
+
+3
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+![Figure](figures/figure_003.png)
+
+Fig 2. Example in knowledge flow among technologies and countries from patent document.
+
+### 2.4. Evolutionary trajectories of each network
+
+To understand the evolutionary trajectories of the four spaces of (1) the technology co-occurrence matrix, (2) the country cooccurrence matrix, (3) the technology flow matrix, and (4) the country knowledge flow matrix, we exploit those four matrices over five-year periods of 1980–1984, 1985–1989, 1990–1994, 1995–1999, 2000–2004, 2005��2009, and 2010–2014. By using all of these 28 (4 × 7) adjacency matrices, we visualize and observe how the four spaces have evolved over the time periods. This helps us easily discern some initial ideas, however, we still need support from the numbers with regard to overall network properties for further understanding. Thus we utilize basic indicators in network analysis studies such as the number of nodes, edges, network density, average path length, and average cluster coefficients. In the data sample herein, the large numbers of nodes and edges mean that many technologies (or countries) are participating during the designated time periods and that their combinational links are diverse, respectively. Network density is calculated as the sum of the weight of edges divided by every possible combination of edges, representing the degree of connectivity. Average path length is calculated by summing the shortest path between all pairs of nodes and dividing by the total number of pairs. Average path length represents how long it takes for a node to get to another node on average. The average clustering coefficient is the ratio of existing edges connecting a node's neighbors to each other to the maximum possible combination.
+
+### 2.5. Trend analysis with network indices
+
+We apply several network indices with the attributes of nodes and edges to capture trends of technology convergence, collaboration, and knowledge flow among technologies and countries. Edge weight, $A_{ij}$ , which is the value of the cell in the co-occurrence between $i$ and $j$ or knowledge flow from $i$ to $j$ , enables us to capture a trend of distinguished convergent combination between technologies.
+
+Centralities are the most representative indicator in network analysis. We use weighted degree and betweenness centrality to track the evolutionary trajectories in top-tiered technologies in ICT as the pertinent knowledge spaces have evolved. Weighted degree centrality, $C_{WD}(v)$ , is the sum of edge weights to a node, measured by each node. A co-occurrence matrix does not consider the in-orout-degree centrality because there is no meaning in the direction of the links. In a knowledge flow matrix, however, the direction of links by nodes has different implications whether they are in-coming or out-going links. Thus, in-degree $C_{WD-in}(v)$ and out-degree $C_{WD-out}(v)$ centralities are measured separately. We can also capture the tendency of a technology (or country) to be a source technology (or country) in knowledge flow by using the directional information (based herein on patent citation). Here, we use outdegree ratio to the overall degree, ODR(v).
+
+4
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+$$C_{WD}(v)=C_{WD-in}(v)+C_{WD-out}(v)$$
+
+$$where, \quad C_{WD-in}(v)=\sum_{i}A_{iv}, C_{WD-out}(v)=\sum_{j}A_{ij}\quad (1)$$
+
+$$ODR(v)=C_{WD-out}(v) / C_{WD}(v)\quad (2)$$
+
+We also use eigenvector centrality, $C_E(v)$ , to measure the overall influence of the nodes across the networks ( Newman , 2008 ) . Compared to the way of calculation for degree centrality where each neighbor contributes equally to centrality, a node's eigenvector centrality is calculated by considering the relative scores assigned to all other nodes based on a concept wherein connections to highscoring nodes indicate higher influence than an equal number of connections to low-scoring nodes. Eigenvector centrality of a node is calculated by proportional value to the sum of scores of its neighbors (see Eq. (3)). Therefore, a high eigenvector score means that a technology (or a country) is connected to many technologies (or countries) with high eigenvector scores.
+
+$$C_{E}(v)=\frac{1}{\lambda} \sum_{\imath \in M(v)} C_{E}(t)=\frac{1}{\lambda} \sum_{\imath \in G} A_{\imath \imath} \times C_{E}(t)\quad (3)$$
+
+where $M(v)$ is a set of the neighbors of $v$, and $\lambda$ is a constant
+
+Betweenness centrality, C B (v), is measured by the number of the shortest paths that pass through each node (see Eq. (4) ). Betweenness centrality indicates how well a node is bridging other nodes or communities through the node, measured by each node. The node with the highest betweenness centrality is the most connective technology in the knowledge space, the biggest platform country in the collaboration space, and the most effective transferring knowledge and country in the knowledge flow space.
+
+$$C_{B}(v)=\sum_{i \neq v \neq j \in V} \frac{\sigma_{ij}(v)}{\sigma_{ij}}\quad (4)$$
+
+where $\sigma_{ij}$ is the number of shortest paths from $i$ to $j$ and $\sigma_{ij}(v)$ is the number of shortest paths from $i$ to $j$ passing through node $v$.
+
+### 2.6. Trend analysis based on the positioning of each node using centralities
+
+We apply a matrix based on eigenvector and betweenness centrality to identify the positions of the nodes in the collaboration space and knowledge flow among countries (see Fig. 3 ). Unlike the work of Lee and Kim (2018) , which adopts weighted degree centrality for building a positioning matrix, we use eigenvector centrality partnered with betweenness centrality to more effectively reflect the overall influence of nodes in the networks. This is because eigenvector centrality, in comparison to weighted degree centrality, weights more for each node's overall influence in a network than the node itself. Thus if a technology (or a country) has both high eigenvector and betweenness centralities, we classify the node into a global hub of the entire network system because it has a highly influential power as well as a high possibility of overlapping with other technologies (or countries). If a node has a high eigenvector but low betweenness centrality, however, the node is likely to be a local hub of a community because it has a strong influence, but only in a specific local community. In addition, if a node has low eigenvector centrality but high betweenness centrality, then the node is likely to be a bridge node linking other local communities, meaning that the technology (or the country) is not influential, but could be a bottleneck or a gatekeeper when important information is transferred from one community to another.
+
+![Figure](figures/figure_014.png)
+
+Fig 3. Network positioning matrix.
+
+5
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+Table 1
+
+Descriptive statistics of the networks (Knowledge space and flow in ICT).
+
+<table><tr><td>Period</td><td>80–84</td><td>85–89</td><td>90–94</td><td>95–99</td><td>00–04</td><td>05–09</td><td>10–14</td></tr><tr><td colspan="8">Knowledge space in ICT</td></tr><tr><td>No. nodes</td><td>12</td><td>12</td><td>12</td><td>12</td><td>12</td><td>12</td><td>12</td></tr><tr><td>No. edges</td><td>124</td><td>132</td><td>132</td><td>131</td><td>132</td><td>132</td><td>132</td></tr><tr><td>Net. Density</td><td>54</td><td>88</td><td>205</td><td>520</td><td>919</td><td>1,328</td><td>2,069</td></tr><tr><td colspan="8">Knowledge flow in ICT</td></tr><tr><td>No. nodes</td><td>12</td><td>12</td><td>12</td><td>12</td><td>12</td><td>12</td><td>12</td></tr><tr><td>No. edges</td><td>132</td><td>132</td><td>132</td><td>132</td><td>132</td><td>132</td><td>132</td></tr><tr><td>Net. Density</td><td>1,587</td><td>4,595</td><td>21,971</td><td>290,898</td><td>232,707</td><td>793,250</td><td>2,938,818</td></tr></table>
+
+
+## 3. Findings
+
+### 3.1. Evolutional paths of knowledge space and flow in ICT
+
+Table 1 shows descriptive statistics of knowledge space and flow in ICT by periods of time. All networks have twelve nodes, representing that twelve technology areas of ICT (high speed networks, mobile communication, security, sensor and device networks, high speed computing, large-capacity and high speed storage, large-capacity information analysis, cognition and meaning understanding, humaninterface, imaging and sound technologies, information communication devices, electronic measurement, and other ) are connected to each other at least once. Thus there are no isolated nodes in both networks. Network densities in both networks have been increased from 54 to 2,069 and from 1,587 to 2,938,818 as the weights of the edges have grown in the fixed number of nodes. An increasing trend of network density in knowledge space means that ICT technology has become more complicated with various inventions (rigorous recombination). On the other hand, an increasing trend of network density in knowledge flow space means that the amount of inflowing and out-flowing knowledge among sub-technological fields in ICT has grown rapidly.
+
+Fig. 4 shows the evolutional paths of ICT knowledge space and knowledge flow space over alternating periods of five years. In knowledge space, the distance between technologies indicates technical relatedness between them measured by the degree to which they co-occur frequently. Thus a short distance means high technical relatedness, so that a pair of technologies is more likely to be combined than other pairs. During 1980 – 1984, four technologies of information communication devices, high speed networks, largecapacity and high speed storage, and imaging and sound technologies are located in the center of knowledge space. Since 1990, however, technologies related to security and large-capacity information analysis are shown to have grown rapidly and combined with the previous big four technologies. During 2000 – 2004, sensor technologies such as human-interface or computer input-output and other technologies began to grow, and the recent period of 2010 – 2014 shows big growth in mobile communication .
+
+While the knowledge space has changed dramatically over time, the evolutional path of knowledge flow space seems not to have changed across the sample periods. This might be due to the fact that foundation or source knowledge is not typically replaced. Once a technology is used as a base knowledge, then the technology is cited repeatedly whilst knowledge complexity increases. High speed networks have been used as a source of knowledge over the sample periods. In the first period of 1980 – 1984, we identified specific technology pairs wherein both paired technologies were shown to have highly affected each other. For example, high speed networks and information communication devices are influenced by each other, and large-capacity and high speed storage and imaging and sound technologies cite each other as well. However, this trend is not shown to have lasted longer than the period of 1990 – 1994. We hardly find strong evidence of which technologies are more greatly based on others, because most technologies end up being connected and citing each other. One unchanging fact is that the four technologies of high speed networks, information communication devices, largecapacity and high speed storage, and imaging and sound technologies are knowledge sources for other technologies including electronic measurement and sensor and device networks.
+
+Aggregated edges' weight by period hints at the relationship between technologies and their evolution. Table 2 shows trends in the top five convergent ICT technologies and their weights in the knowledge space network (on the left side) together with the top five directions of ICT technology flow in the knowledge flow network (on the right side). The most frequent convergent combination during the initial period from 1980–1984—the combination between high speed networks and information communication device technologies—continued to make top five convergent lists until the period of 2000–2004, disappearing thereafter. In contrast, the combination between high speed networks and mobile communication did not make the list until 1994, when it suddenly emerged at the second spot, remaining in the first tier up to the period of 2010–2014.
+
+In knowledge space, the two technologies of high speed networks and information communication devices are shown to have cooccurred frequently in the 1980s, but there is no information on the direction of knowledge flow. The right side of Table 2 shows the top five directions of knowledge flow in ICT. Starting with the period of 1980–1989, it turns out that knowledge flow is mostly bidirectional, as opposed to unilateral, until a change is seen in the period from 1990–1994. For example, during the period of 1980–1984, the first ranking trend was the flow from imaging and sound technologies to large-capacity and high speed storage , and the second ranking trend was in the opposite direction between the same two sub-classes. The period of 1985–1989 shows the same relationship between the two. This trend changed as of 1990. During the period of 1995–1999, most trends in knowledge flow went from high speed network technology to various technologies. In other words, high speed network technology became a well-known knowledge source for other technologies.
+
+6
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+![Figure](figures/figure_003.png)
+
+Fig 4. Evolutional path of knowledge space and flow in ICT sector
+
+Note: A, C, E, and G indicate an evolutionary path of knowledge space, and B, D, F, and H indicate an evolutionary path of Knowledge flow.
+
+7
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+Table 2
+
+Trends in Top 5 convergent technologies and technology flow in ICT sector.
+
+<table><tr><td>Period</td><td>Knowledge space in ICT Weight</td><td>ICT</td><td>Knowledge flow in ICT Weight</td><td>ICT Technology Knowledge Flow (A = 8)</td><td>Large-capacity and high speed storage</td></tr><tr><td>80–84</td><td>8323 High speed network</td><td>Information communications device</td><td>8.134</td><td>Imaging and sound technology</td><td>Large-capacity and high speed storage</td></tr><tr><td>855</td><td>1655 Imaging and sound technology</td><td>Large-capacity and high speed storage</td><td>8.108</td><td>Large-capacity and high speed storage</td><td>Imaging and sound technology</td></tr><tr><td>8554</td><td>548 Information communication device</td><td>Large-capacity and high speed storage</td><td>7.487</td><td>Information communications device</td><td>Large-capacity and high speed storage</td></tr><tr><td>85–89</td><td>473 High speed network</td><td>Large-capacity and high speed storage</td><td>7.226</td><td>High speed network</td><td>Information communication device</td></tr><tr><td>85–89</td><td>1.14 High speed network</td><td>Imaging and sound technology</td><td>8.272</td><td>Imaging and communication device</td><td>High speed network</td></tr><tr><td>975</td><td>1755 Information communication device</td><td>Information communication device</td><td>25.381</td><td>Large-capacity and high speed storage</td><td>Imaging and sound technology</td></tr><tr><td>9740</td><td>Imaging and sound technology</td><td>Large-capacity and high speed storage</td><td>18.987</td><td>High speed network</td><td>Large-capacity and high speed storage</td></tr><tr><td>9740</td><td>Imaging and sound technology</td><td>Information communications device</td><td>18.893</td><td>High speed network</td><td>Information communication device</td></tr><tr><td>9740</td><td>2.910</td><td>Imaging and sound technology</td><td>Large-capacity and high speed storage</td><td>16.477</td><td>High speed network</td></tr><tr><td>2.010</td><td>High speed network</td><td>Information communications device</td><td>14.776</td><td>High speed network</td><td>Imaging and sound technology</td></tr><tr><td>2.016</td><td>High speed network</td><td>Imaging and sound technology</td><td>128.565</td><td>Imaging and sound technology</td><td>Large-capacity and high speed storage</td></tr><tr><td>1.221</td><td>High speed network</td><td>Large-capacity and high speed storage</td><td>11.114</td><td>High speed network</td><td>Imaging and sound technology</td></tr><tr><td>95–99</td><td>6.426</td><td>Imaging and sound technology</td><td>Large-capacity and high speed storage</td><td>11.114</td><td>High speed network</td></tr><tr><td>4.744</td><td>4.744</td><td>High speed network</td><td>2.0025,540</td><td>2.0025,540</td><td>High speed computing</td></tr><tr><td>4.744</td><td>4.744</td><td>High speed network</td><td>Multiple communication technologies</td><td>2.532,435</td><td>High speed network</td></tr><tr><td>5.969</td><td>5.969</td><td>High speed network</td><td>Information communications device</td><td>1.694,055</td><td>Imaging and sound technology</td></tr><tr><td>5.979</td><td>3.479</td><td>Imaging and sound technology</td><td>2.147</td><td>High speed network</td><td>Computer input-output and Others</td></tr><tr><td>6.479</td><td>3.479</td><td>Imaging and sound technology</td><td>2.147</td><td>High speed network</td><td>Computer input-output and Others</td></tr><tr><td>6.127</td><td>1.1704</td><td>High speed network</td><td>2.147</td><td>High speed network</td><td>High speed computing</td></tr><tr><td>1.1704</td><td>High speed network</td><td>Mobile communication</td><td>1.728,288</td><td>High speed network</td><td>Security</td></tr><tr><td>7.286</td><td>Large-capacity information analysis</td><td>Security</td><td>2.147</td><td>High speed network</td><td>High speed network</td></tr><tr><td>6.938</td><td>1.638</td><td>Imaging and sound technology</td><td>2.147</td><td>High speed network</td><td>Large-capacity information analysis</td></tr><tr><td>6.59</td><td>2.834</td><td>Imaging and network</td><td>Information communications device</td><td>933,365</td><td>High speed network</td></tr><tr><td>6.59</td><td>2.834</td><td>Imaging and sound technology</td><td>2.147</td><td>High speed network</td><td>High speed network</td></tr><tr><td></td><td>1.1277</td><td>Imaging and sound technology</td><td>2.147</td><td>High speed network</td><td>High speed network</td></tr><tr><td></td><td>1.0684</td><td>Large-capacity information analysis</td><td>2.147</td><td>High speed network</td><td>Security</td></tr><tr><td>9.936</td><td>9.936</td><td>High speed network</td><td>Security</td><td>2.147,500,988</td><td>Security information analysis</td></tr><tr><td>9.978</td><td>9.878</td><td>Imaging and sound technology</td><td>2.147,512</td><td>High speed network</td><td>Large-capacity information analysis</td></tr><tr><td>10–14</td><td>15.608</td><td>High speed network</td><td>Security</td><td>2.146,633</td><td>High speed network</td></tr><tr><td></td><td>1.2360</td><td>High speed computing</td><td>High speed network</td><td>2.003,179,989</td><td>High speed network</td></tr><tr><td>1.1749</td><td>Imaging and sound technology</td><td>Large-capacity and high speed storage</td><td>2.005,1749</td><td>High speed network</td><td>Computer input-output and Others</td></tr><tr><td>11.532</td><td>Imaging and sound technology</td><td>Large-capacity information analysis</td><td>1.476,313</td><td>Mobile communication</td><td>Large-capacity information analysis</td></tr></table>
+
+
+8
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+![Figure](figures/figure_003.png)
+
+Fig 5. Trends in weighted degree by technologies in ICT sector Note: Fig. 5 represents trends in $C_{WD}(v)$ by ICT technologies in knowledge space (A) and knowledge flow space (B).
+
+Node attributes (the basis of several centralities herein) give useful information, such as the recombination and positioning of subtechnologies in ICT. Fig. 5 shows how each technology's weighted degree centrality has changed over time. Fig. 5 -A and B represent trends in the weighted degree in knowledge space and knowledge flow, respectively. Most technologies show increasing trends in both graphs, with similar orders of rank. We found two ICT technologies mismatched in the two networks: imaging and sound and computer input-output technologies. Imaging and sound technology has maintained the second spot in the rankings, showing a high growth rate in knowledge space, but lagging with a slow growth rate in knowledge flow. This can be understood to mean that imaging and sound is a well-known dimension of convergent technology, but has not greatly contributed to ICT knowledge flow in terms of being source knowledge or a target of knowledge. On the other hand, computer input-output technology has skyrocketed in the knowledge flow network, but has shown only slow growth in the knowledge space network. This means that computer input-output technology has greatly contributed to being source knowledge or a target of knowledge creation instead of being a dimension of convergent technology.
+
+By showing trends in absolute weighted degree values in Fig. 5 , we can discern absolute differences between technologies, but hardly capture the relative dynamics among technologies. In this sense, Fig. 6 stands to better elucidate technology dynamics and forecasting by showing trends in the rank of weighted degree and eigenvector centralities in the ICT sector. Fig. 6 -A and B refer to the ranking dynamics of weighted degree centrality in knowledge space and knowledge flow, respectively. Fig. 6 -C and D refer to the ranking dynamics of eigenvector centrality in both spaces. We highlight the two technologies of mobile communication and largecapacity information analysis as the fastest growing technologies in all centralities and spaces. In comparing the two spaces, what is interesting is that the increasing rate in the ranking of mobile communication is faster than that of large-capacity information analysis in knowledge space, while the opposite holds true in knowledge flow space. This means that the main power of growth in mobile communication technology is driven by the fusion of mobile communication with other technologies rather than by the adoption or affecting of other technologies on the part of mobile communication. Information and communication device technology shows decreasing trends in both spaces. Trends in eigenvector centralities have no distinct differences in comparison to weighted degree centralities.
+
+### 3.2. Evolutional paths of collaboration space and knowledge flow in ICT
+
+Table 3 shows descriptive statistics of the collaboration space and knowledge flow among countries in ICT by periods of time. In both networks, the amount of nodes has been growing over time. Only 87 countries engaged in international collaborations during the period of 1980 – 1984, but in today's environment, more than approximately 180 countries are collaborating or contributing to knowledge flow in the ICT sector. In both networks of collaboration space and knowledge flow, network density has increased from 0.11 to 1.48 and from 3.36 to 24.83, respectively, representing that collaboration has become more multinational and that knowledge flow has become more dense among countries as ICT technology has grown more complicated. Average path length in both networks has decreased from 2.51 to 2.11 in the collaboration space and from 2.02 to 1.93 in the knowledge flow space. Decreasing trends in average path length suggest how the two spaces have co-evolved. The transferring of knowledge (or information) has become faster in knowledge flow networks as many countries are increasingly connected and thereby better able to find and reach appropriate
+
+9
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+![Figure](figures/figure_003.png)
+
+Fig 6. Ranking dynamics of centralities in knowledge space and flow among ICT technologies Note: Fig. 6 represents ranking dynamics of $C_{WD}(v)$ in knowledge space (A) and knowledge flow space (B), and $C_E(v)$ in knowledge space (C) and knowledge flow space (D).
+
+Table 3
+
+Descriptive statistics of the networks (Collaboration space and knowledge flow among countries).
+
+<table><tr><td>Period</td><td>80–84</td><td>85–89</td><td>90–94</td><td>95–99</td><td>00–04</td><td>05–09</td><td>10–14</td></tr><tr><td colspan="8">Collaboration space in ICT</td></tr><tr><td>No. nodes</td><td>87</td><td>97</td><td>119</td><td>162</td><td>183</td><td>183</td><td>184</td></tr><tr><td>No. edges</td><td>268</td><td>362</td><td>563</td><td>1,091</td><td>1,659</td><td>2,019</td><td>2,189</td></tr><tr><td>Net. Density</td><td>0.11</td><td>0.19</td><td>0.27</td><td>0.44</td><td>0.76</td><td>1.15</td><td>1.48</td></tr><tr><td>Ave. Pathlength</td><td>2.51</td><td>2.42</td><td>2.31</td><td>2.30</td><td>2.22</td><td>2.10</td><td>2.11</td></tr><tr><td>Ave. CC</td><td>0.20</td><td>0.24</td><td>0.25</td><td>0.31</td><td>0.34</td><td>0.36</td><td>0.39</td></tr><tr><td colspan="8">Knowledge flow among countries in ICT</td></tr><tr><td>No. nodes</td><td>108</td><td>119</td><td>134</td><td>175</td><td>199</td><td>188</td><td>189</td></tr><tr><td>No. edges</td><td>612</td><td>849</td><td>1,215</td><td>2,297</td><td>3,101</td><td>3,143</td><td>3,407</td></tr><tr><td>Net. Density</td><td>3.36</td><td>6.22</td><td>9.56</td><td>17.64</td><td>18.47</td><td>18.75</td><td>24.83</td></tr><tr><td>Ave. Pathlength</td><td>2.02</td><td>2.00</td><td>1.97</td><td>1.95</td><td>1.96</td><td>1.95</td><td>1.93</td></tr><tr><td>Ave. CC</td><td>0.23</td><td>0.27</td><td>0.31</td><td>0.33</td><td>0.36</td><td>0.39</td><td>0.42</td></tr></table>
+
+
+collaborative partners. Evidence of steady increases in average clustering coefficients in both networks leads us to conclude that local communities have been forming slowly within collaboration and knowledge flow networks.
+
+Fig. 7 shows the evolutional paths of collaboration space and knowledge flow in the ICT sector over alternating periods of five years. During the period of 1980 – 1984, only a few countries (in comparison to the current multinational nature of both networks) joined in collaboration and knowledge flow, and that four or five countries, including the United States (US), Japan, Germany, France, and the United Kingdom (UK), led the collaboration or knowledge-creating scene. What is clearly observed in this period is that the majority of ICT knowledge flowed between the US and Japan. However, there was a gradual rise in the number of countries participating in collaboration during the period of 1985 – 2004. At that point, the US and Japan were no longer the only countries collaborating and prioritizing the flow of knowledge among themselves. Although the distribution of collaboration and knowledge flow had become more even, uneven distribution remained worldwide, with about ten to twelve countries being the main participants in collaboration and knowledge flow in the fairly recent past. For example, countries initially positioned at the periphery of the knowledge flow network, such as South Korea, have emerged as recently as 2000. Currently, more countries are crowded in the
+
+10
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+![Figure](figures/figure_003.png)
+
+Fig 7. Evolutional paths of collaboration space and knowledge flow in ICT sector Note: A, C, E, and G indicate an evolutionary path of collaboration space, and B, D, F, and H indicate an evolutionary path of Knowledge flow space among countries Fig 8 . Ranking dynamics of centralities in ICT knowledge space and flow among countries.
+
+11
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+Table 4
+
+Trends in Top 5 international collaboration partners in ICT sector.
+
+<table><tr><td rowspan="2">Period</td><td colspan="2">Collaboration space in ICT</td><td colspan="3">Knowledge flow in ICT</td></tr><tr><td>Weight</td><td>Collaboration partner (A-B)</td><td>Weight</td><td>Technology Knowledge Flow (A→B)</td><td></td></tr><tr><td rowspan="5">80–84</td><td>61</td><td>United States</td><td>Japan</td><td>8,649</td><td>United States</td><td>Japan</td></tr><tr><td>51</td><td>United States</td><td>United Kingdom</td><td>7,132</td><td>Japan</td><td>United States</td></tr><tr><td>49</td><td>United States</td><td>Germany</td><td>1,742</td><td>United States</td><td>Germany</td></tr><tr><td>37</td><td>United States</td><td>Canada</td><td>1,673</td><td>Germany</td><td>United States</td></tr><tr><td>32</td><td>United States</td><td>Netherland</td><td>1,457</td><td>France</td><td>United States</td></tr><tr><td rowspan="5">85–89</td><td>184</td><td>United States</td><td>Japan</td><td>20,712</td><td>United States</td><td>Japan</td></tr><tr><td>114</td><td>United States</td><td>Canada</td><td>19,708</td><td>Japan</td><td>United States</td></tr><tr><td>83</td><td>United States</td><td>United Kingdom</td><td>3,136</td><td>Germany</td><td>United States</td></tr><tr><td>74</td><td>United States</td><td>Germany</td><td>2,680</td><td>United States</td><td>Germany</td></tr><tr><td>65</td><td>Japan</td><td>Germany</td><td>2,569</td><td>United Kingdom</td><td>United States</td></tr><tr><td rowspan="5">90–94</td><td>517</td><td>United States</td><td>Japan</td><td>43,930</td><td>Japan</td><td>United States</td></tr><tr><td>224</td><td>United States</td><td>United Kingdom</td><td>33,915</td><td>United States</td><td>Japan</td></tr><tr><td>172</td><td>United States</td><td>Canada</td><td>5,029</td><td>United Kingdom</td><td>United States</td></tr><tr><td>168</td><td>United States</td><td>Israel</td><td>4,292</td><td>Germany</td><td>United States</td></tr><tr><td>158</td><td>United States</td><td>Germany</td><td>3,846</td><td>France</td><td>United States</td></tr><tr><td rowspan="5">95–99</td><td>1,179</td><td>United States</td><td>Japan</td><td>107,446</td><td>Japan</td><td>United States</td></tr><tr><td>694</td><td>United States</td><td>Canada</td><td>81,122</td><td>United States</td><td>Japan</td></tr><tr><td>690</td><td>United States</td><td>United Kingdom</td><td>16,548</td><td>Canada</td><td>United States</td></tr><tr><td>552</td><td>United States</td><td>Germany</td><td>15,991</td><td>United States</td><td>United Kingdom</td></tr><tr><td>530</td><td>Taiwan</td><td>China</td><td>15,410</td><td>United States</td><td>South Korea</td></tr><tr><td rowspan="5">00–04</td><td>1,622</td><td>United States</td><td>United Kingdom</td><td>99,614</td><td>Japan</td><td>United States</td></tr><tr><td>1,560</td><td>United States</td><td>Germany</td><td>88,555</td><td>United States</td><td>Japan</td></tr><tr><td>1,547</td><td>United States</td><td>Japan</td><td>22,764</td><td>Canada</td><td>United States</td></tr><tr><td>1,441</td><td>United States</td><td>Canada</td><td>20,516</td><td>United Kingdom</td><td>United States</td></tr><tr><td>1,127</td><td>Taiwan</td><td>China</td><td>20,055</td><td>United States</td><td>United Kingdom</td></tr><tr><td rowspan="5">05–09</td><td>2,526</td><td>United States</td><td>Canada</td><td>63,318</td><td>Japan</td><td>United States</td></tr><tr><td>2,254</td><td>United States</td><td>United Kingdom</td><td>50,171</td><td>United States</td><td>Japan</td></tr><tr><td>1,965</td><td>United States</td><td>India</td><td>24,136</td><td>United States</td><td>South Korea</td></tr><tr><td>1,892</td><td>United States</td><td>Germany</td><td>22,289</td><td>Canada</td><td>United States</td></tr><tr><td>1,738</td><td>United States</td><td>Japan</td><td>21,823</td><td>United States</td><td>Canada</td></tr><tr><td rowspan="5">10–14</td><td>3,779</td><td>United States</td><td>India</td><td>67,307</td><td>Japan</td><td>United States</td></tr><tr><td>3,670</td><td>United States</td><td>Canada</td><td>45,108</td><td>Canada</td><td>United States</td></tr><tr><td>2,672</td><td>United States</td><td>China</td><td>36,505</td><td>United States</td><td>Japan</td></tr><tr><td>2,614</td><td>United States</td><td>United Kingdom</td><td>35,639</td><td>United States</td><td>Canada</td></tr><tr><td>2,190</td><td>United States</td><td>Germany</td><td>30,755</td><td>South Korea</td><td>United States</td></tr></table>
+
+
+center of the collaboration space, and knowledge flow sources and channels are becoming more diverse.
+
+Table 4 shows trends in the top five international collaboration partners in the ICT sector. As shown in Fig. 7 , since 1980, the United States has been a global hub in both collaboration space and the knowledge transfer network. Collaborations between the US and Japan had been among the most frequent collaborations until 2009, but this relationship is not on the list of top five collaborative partnerships in the period of 2010 – 2014. Other collaborative partnerships such as US-Germany, US-Canada, and US – UK have stayed strong over time. Collaborations that stand out in the more recent scene are collaborations between China and Taiwan and the US and India. China – Taiwan collaboration, which had not appeared in the top ranks of the initial time periods, suddenly ranked fifth during the period from 1995 – 2004. US-India collaboration emerged in 2005 – 2009 as the third highest-ranking collaborative partnership, and the relationship ranked as the most popular collaboration in the period of 2010 – 2014.
+
+There are also interesting findings in the direction of knowledge flow in ICT over time. Until 1989, the US was the main source of ICT knowledge, with knowledge flowing mostly toward Japan. Since 1990, however, the flow of knowledge has changed to the opposite direction, and Japan remains the main knowledge source in ICT sectors worldwide. What is interesting here is the emer- gence of South Korea. While South Korea did not exist on the collaboration scene during whole initial periods, ICT knowledge began suddenly to flow to them in the period from 2004–2009. Today South Korea is one of the world's main sources of ICT knowledge.
+
+Fig. 8 shows trends in the ranking of weighted degree, eigenvector, and betweenness centralities in the ICT sector. We find a prominent increasing trend in the weighted degree centrality of Canada, India, and China (see Fig. 8 -A) as well as in the eigenvector centrality of these countries (see Fig. 8 -C) in collaboration space. In knowledge flow, however, a prominent increasing trend is found in weighted degree and eigenvector centralities of Canada and South Korea (see Fig. 8 -B), while a prominent decreasing trend is found in these factors in France and Switzerland (see Fig. 8 -D). When it comes to betweenness centrality, Japan remains top-ranked in both collaboration and knowledge flow networks (see Fig. 8 -E and F). In contrast, Canada shows a decreasing trend, despite remaining top-ranked in weighted degree centrality and eigenvector centrality (see Fig. 8 -E). In knowledge flow, China is shown to have made rapid growth in betweenness, while South Korea made significant growth until 2009, but has currently stopped growing (see Fig. 8 -F). Even though South Korea has grown to become an important knowledge source, its abrupt fall may lead the country to be classified as a local hub, which has influential power only within specific communities.
+
+12
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+![Figure](figures/figure_003.png)
+
+Fig. 8. represents ranking dynamics of $C_{WD}(v)$ in knowledge space (A) and knowledge flow space (B), $C_E(v)$ in knowledge space (C) and knowledge flow space (D), and $C_H(v)$ in knowledge space (E) and knowledge flow space (F).
+
+### 3.3. Capturing the relationships among countries by using a network position matrix
+
+To capture the relationships among countries in terms of collaboration and knowledge flow, we apply a network position matrix. Fig. 9 shows how each country has shifted position in collaboration and knowledge flow spaces over time. We find that, first, the US is the only country to keep the global hub position in both spaces at the same time. Japan is a global hub in knowledge flow, but has moved from a global hub to a bridge position in collaboration space. Canada was also one of the global hubs in both spaces at one point, but has fallen into a local hub in both spaces. Switzerland is losing their influence (eigenvector centrality) and their gatekeeping role (betweenness centrality) in both spaces. While South Korea is not a famous player in collaboration space, in knowledge flow space South Korea has grown from the periphery, through the local hub, to the global hub. Currently, however, South Korea has stopped growing toward the global hub and has landed instead as a local hub. China was peripheral in both spaces, but the country has grown to become a local hub in collaboration space and to have a bridge role in knowledge flow.
+
+### 3.4. Capturing information on knowledge sources among technologies and countries
+
+Finally, we use the weighted out-degree ratio to determine which technologies (or countries) have emerged as sources or
+
+13
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+![Figure](figures/figure_003.png)
+
+Fig 9. Trends of changing position in collaboration space and knowledge flow among countries Note: A and B refer to positioning change in collaboration space and in knowledge flow space respectively.
+
+foundations, and which are more applied technologies (or, in the case of countries, which are better in applying knowledge rather than creating new knowledge) in terms of knowledge flow. Fig. 10 shows trends in the weighted out-degree ratio, ODR(v), in the ICT sector. Overall, it seems difficult to detect a macro trend, but there are some patterns we can capture. In Fig. 10 -A, an increasing trend in electronic measurement is observed, indicating that electronic measuring technology has changed from an applied to a source technology over time. In other words, electronic measurement is no longer a trendy technology, but is poised to be a cash cow technology in the ICT sector going forward. A technology showing a decreasing trend of ODR is large-capacity information analysis, indicating that this dimension has changed from source to applied technology. This means that big-data analysis may be a future technology largely created by other source technologies. The risk, however, is that this kind of targeted technology stands to easily disappear.
+
+Fig. 10 -B shows the ranking dynamics of ODR in knowledge flow space among countries. In terms of source technologies, Japan was a first rising star, and has managed to maintain this top position. In relatively recent periods, South Korea has emerged as a second rising star in terms of owning source technologies. However, because South Korea is only actively participating in the knowledge flow scene without any effort to join the collaboration space, their growth may be limited in comparison to Japan's strong participation in both collaboration and knowledge flow spaces. While the US has always been the leader in both collaboration and knowledge spaces, the ODR of the US has dropped. Thus, at least in the ICT sector, the US is no longer a strong technology sourcing country worldwide.
+
+![Figure](figures/figure_007.png)
+
+Fig 10. Ranking dynamics of weighted out-degree ratio in knowledge space and flow among ICT technologies and countries Note: A and B refer to ranking dynamics of ODR(v) among technologies and countries.
+
+14
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+## 4. Discussion and conclusion
+
+Insights into the multiple dimensions that shape the evolutionary pathways of ICT technology advancements highlight the importance of knowledge structure and network properties in an environment of rapid change. In parallel, these insights also raise awareness of the dynamic capabilities and significant factors that should be considered by organizations and entities in the development of strategies and policies intended to maximize future growth. The present investigation identifies four specific dimensions that are relevant in this context. Technology convergence measures based on the knowledge space methodology developed by Kogler et al. ( 2013 ) provide an understanding of proximity among technological sub-classes, which in turn can be interpreted as an indicator of re-combination potential (i.e., technologies that are near to each other in the knowledge space have a much higher potential for recombination than technologies that are not). This has important implications for development strategies, where a dense knowledge space might lend itself to approaches that are geared toward increased patterns of specialization, while a more evenly distributed space might offer opportunities for diversification ( Boschma , 2017 ; Kogler , 2017 ) . Globalization, advancements in transportation, and communication are all processes that reinforce the notion of a globalized knowledge economy—and the growing numbers of international inventor collaborations certainly confirm this. Thus, patterns of international collaboration in knowledge production, here measured by the development of novel products and processes of economic value (i.e., inventions) are yet another important dimension to consider in the present context. Knowledge and collaboration spaces are measures based on co-occurrence, which indicate a symmetrical relationship between the respective units of interest, thereby providing important insights on the overall structural properties of pertinent aspects of technology. While knowledge exchange is not necessarily unidirectional, some organization units or spatial configurations might be in more favorable positions than others to source knowledge. To account for this fact, two further dimensions are considered in the present investigation—namely, knowledge flow among technology sub-classes and among similar countries within the global ICT sector as indicated by patent citation knowledge flow patterns. The competitive position of an organization is directly linked to its level of dynamic capabilities, and thus, understanding the evolutionary trajectories concerning technology convergence, international collaborative relationships, and knowledge flows among technologies and countries is an important building block in this regard. In summary, all four dimensions should provide a balanced picture of technology evolution in an international context, something that is not only relevant for practitioners and policy makers, but also for a number of scientific disciplines, including strategy and management, innovation studies, and evolutionary economic geography.
+
+Since the advent of smartphones, ICT has become a very dynamic market, driven by fierce “ ecosystem competition ” ( Lee, Lee, & Hwang , 2015 ; Lee, Park, & Lee , 2016 ; Lee, Park, & Lee , 2018 ) based on the capability to recombine knowledge embedded in technologies rather than on the domination of a single technology ( Basole, Park, & Barnett , 2015 ; Fransman , 2010 ; Lee, Kim, & Lee , 2017 ) . Therefore, units that are competent in very few technology sub-classes should look toward greater diversification of their technology base (albeit sticking at first to sub-classes that are related to their current expertise as well as locally and internationally embedded technologies) ( Kim & Lee , 2017 ) . Needless to say, this is a difficult task insofar as practitioners and policy makers must anticipate future technology trajectories in order to develop the most favorable strategies going forward. The proposed approach based on technology class and citation flow information from USPTO patent documents, combined with the suggested methodological tools that produce a number of basic indicators in network analysis and network positioning matrix measures, stand to provide essential insights for such planning purposes.
+
+Results show that mobile communication and large-capacity information analysis are the fastest growing technologies in colla- boration and knowledge flow spaces in terms of both weighted degree centrality and eigenvector centrality. On the other hand, information and communication devices indicate a tendency to decrease in both spaces. Therefore, when determining strategies or policies that aim for expansion in the ICT sector, companies or countries would be wise to consider these trends.
+
+However, even if a particular company or country tries to expand the scope of technologies in its portfolio, it is certainly not possible to be competitive across the whole spectrum due to resource restrictions. To circumvent such restrictions, the utilization of collaboration networks with other companies or other countries in a framework of open innovation is recommended ( Kim, Lee, & Kim , 2016 ) . In this context, it is of course more favorable for a company or country to engage with an advanced and influential partner (rather than a laggard), as this stands to provide an opportunity to take advantage of significant knowledge inputs. In the present investigation, evolutional paths in the collaboration space are analyzed by considering the location of inventors at the time of invention. The results indicate that the US is the only country to have maintained a global hub position in both spaces of investigation. Further, in the case of South Korea, no significant influence on the collaboration space was found, even though the country has exhibited rapid growth in the knowledge flow space. In contrast, it is observed that Switzerland has lost its influence in both spaces over time.
+
+Although the present study examines technology-specific as well as country-specific trends, one limitation that remains is the omission of technology-specific trends by country or country-specific trends by technology. This leaves important questions that should be investigated in follow-up studies (e.g., which technologies have played a major role in maintaining the global hub position or which technological drivers have enabled certain countries to grow in terms of knowledge flow position?) Also, an enrichment of the present analysis with further data sources, such as financial data on ICT-related revenue and ICT import/export data, would further enhance our understanding of evolutionary trajectories in technology advancement and how these trends are linked to economic performance, growth, and change.
+
+## Acknowledgments
+
+Dieter F. Kogler and Changjun Lee would like to acknowledge funding from the European Research Council under the European
+
+15
+
+---
+
+ARTICLE IN PRESS
+
+C. Lee et al.
+
+Information Processing and Management xxx (xxxxx) xxx-xxx
+
+Union’s Horizon 2020 research and innovation programme (grant agreement No 715631, TechEvo). This research was also supported by the MIST (Ministry of Science and ICT), Korea, under the National Program for Excellence in SW supervised by the IITP (Institute for Information & communications Technology Promotion) (2015-0-00914), and by the Ministry of Education of the Republic of Korea, the National Research Foundation of Korea (NRF-2017R1C1B5017518).
+
+## References
+
+Abbas, A., Zhang, L., & Khan, S. U. (2014). A literature review on the state-of-the-art in patent analysis. World Patent Information,, 37, 3-13.
+
+Basole, R. C., Park, H., & Barnett, B. C. (2015). Coopetition and convergence in the ICT ecosystem. Telecommunications Policy, 39(7), 537-552.
+
+Boschma, R. (2017). Relatedness as driver of regional diversification: A research agenda. Regional Studies, 51(3), 351-364.
+
+Bonino, D., Ciaramella, A., & Corno, F. (2010). Review of the state-of-the-art in patent information and forthcoming evolutions in intelligent patent informatics. World Patent Information, 32(1), 30–38.
+
+Chen, H., Zhang, G., Zhu, D., & Lu, J. (2017). Topic-based technological forecasting based on patent data: A case study of Australian patents from 2000 to 2014. Technological Forecasting and Social Change, 119, 39-52.
+
+Codina-Filbà, J., Bouayad-Agha, N., Burga, A., Casamayor, G., Mille, S., Müller, A., et al. (2017). Using genre-specific features for patent summaries. Information Processing & Management, 53(1), 151-174.
+
+Daim, T. U., Rueda, G., Martin, H., & Gerdstri, P. (2006). Forecasting emerging technologies: Use of bibliometrics and patent analysis. Technological Forecasting and Social Change, 73(8), 981-1012.
+
+Dosi, G., & Nelson, R. R. (1994). An introduction to evolutionary theories in economics. Journal of Evolutionary Economics, 4(3), 153-172.
+
+Fransman, M. (2010). The new ICT ecosystem: Implications for policy and regulation. Cambridge University Press.
+
+Griliches, Z. (1998). Patent statistics as economic indicators: A survey. R&D and productivity: The econometric evidence (pp. 287-343). University of Chicago Press.
+
+Guan, J., & Chen, Z. (2012). Patent collaboration and international knowledge flow. Information Processing & Management, 48(1), 170-181.
+
+Hidalgo, C. A., Klinger, B., Barabási, A.-L., & Hausmann, R. (2007). The product space conditions the development of nations. Science, 317(5837), 482-487.
+
+Hird, M. D., & Pfotenhauer, S. M. (2017). How complex international partnerships shape domestic research clusters: Difference-in-difference network formation and research re-orientation in the MIT Portugal Program. Research Policy, 46(3), 557-572.
+
+Inaba, T., & Squicciarini, M. (2017). ICT: A new taxonomy based on the international patent classification. OECD Science, Technology and Industry Working Papers.
+
+Jaffe, A. B. (1986). Technological opportunity and spillovers of R & D: Evidence from firms' patents, profits, and market value. The American Economic Review, 984-1001.
+
+Jaffe, A. B., & Trajtenberg, M. (2002). Patents, citations, and innovations: A window on the knowledge economy. MIT Press.
+
+Jiang, J., Shi, P., An, B., Yu, J., & Wang, C. (2017). Measuring the social influences of scientist groups based on multiple types of collaboration relations. Information Processing & Management, 53(1), 1–20.
+
+Joung, J., & Kim, K. (2017). Monitoring emerging technologies for technology planning using technical keyword based analysis from patent data. Technological Forecasting and Social Change, 114, 281–292.
+
+Khasseh, A. A., Soheili, F., Moghaddam, H. S., & Chelak, A. M. (2017). Intellectual structure of knowledge in iMetrics: A co-word analysis. Information Processing & Management, 53(3), 705-720.
+
+Kim, E., Lee, D., & Kim, J. (2016). How collaboration networks affect innovation in Korea's ICT industry in the era of Internet of Things. Asian Journal of Technology Innovation, 24(2), 202–221.
+
+Kim, G., & Bae, J. (2017). A novel approach to forecast promising technology through patent analysis. Technological Forecasting and Social Change, 117, 228-237.
+
+Kim, H., & Park, Y. (2009). Structural effects of R&D collaboration network on knowledge diffusion performance. Expert systems with applications, 36(5), 8986-8992.
+
+Kim, J., & Lee, S. (2017). Forecasting and identifying multi-technology convergence based on patent data: The case of IT and BT industries in 2020. Scientometrics, 111(1), 47–65.
+
+Kim, Y. G., Suh, J. H., & Park, S. C. (2008). Visualization of patent analysis for emerging technology. Expert systems with applications, 34(3), 1804-1812.
+
+Kodama, F. (1992). Technology fusion and the New R&D. Harvard Business Review, (July-August), 70-78.
+
+Kogler, D. F. (2017). Relatedness as driver of regional diversification: A research agenda-a commentary. Regional Studies, 51(3), 365-369.
+
+Kogler, D. F., Essletzbichler, J., & Rigby, D. L. (2017). The evolution of specialization in the EU15 knowledge space. Journal of Economic Geography, 17(2), 345-373.
+
+Kogler, D. F., Rigby, D. L., & Tucker, I. (2013). Mapping knowledge space and technological relatedness in US cities. European Planning Studies, 21(9), 1374-1391.
+
+Kyebambe, M. N., Cheng, G., Huang, Y., He, C., & Zhang, Z. (2017). Forecasting emerging technologies: A supervised learning approach through patent analysis. Technological Forecasting and Social Change, 125, 236-244.
+
+Lee, C., Lee, D., & Hwang, J. (2015). Platform openness and the productivity of content providers: A meta-frontier analysis. Telecommunications Policy, 39(7), 553-562.
+
+Lee, C., & Kim, H. (2018). The evolutionary trajectory of an ICT ecosystem: A network analysis based on media users' data. Information & Management.
+
+Lee, C., Kim, J. H., & Lee, D. (2017a). Intra-industry innovation, spillovers, and industry evolution: Evidence from the Korean ICT industry. Telematics and Informatics, 34(8), 1503–1513.
+
+Lee, J., Kim, C., & Shin, J. (2017b). Technology opportunity discovery to R&D planning: Key technological performance analysis. Technological Forecasting and Social Change, 119, 53-63.
+
+Lee, K. (2013). Schumpeterian analysis of economic catch-up: Knowledge, path-creation, and the middle-income trap. Cambridge University Press.
+
+Lee, K., Park, Y., & Lee, D. (2016). Effect of the ICT ecosystem structure on the sustainable growth of ICT firms: A metafrontier analysis on China, South Korea, the United States, and Japan. Sustainability, 8(5), 469.
+
+Lee, K., Park, Y., & Lee, D. (2018). Measuring efficiency and ICT ecosystem impact: Hardware vs. software industry. Telecommunications Policy, 42(2), 107-115.
+
+Liu, S. -H., Liao, H. -L., Pi, S. -M., & Hu, J. -W. (2011). Development of a patent retrieval and analysis platform--A hybrid approach. Expert systems with applications, 38(6), 7864–7868.
+
+Neffke, F., Henning, M., & Boschma, R. (2011). How do regions diversify over time? Industry relatedness and the development of new growth paths in regions. Economic Geography, 87(3), 237-265.
+
+Newman, M. E. (2008). The mathematics of networks. The New Palgrave Encyclopedia of Economics, 2(2008), 1-12.
+
+Patel, P., & Pavitt, K. (1997). The technological competencies of the world's largest firms: Complex and path-dependent, but not much variety. Research Policy, 26(2), 141–156.
+
+Rycroft, R. W., & Kash, D. E. (2004). Self-organizing innovation networks: Implications for globalization. Technovation, 24(3), 187-197.
+
+Teece, D. J. (1998). Capturing value from knowledge assets: The new economy, markets for know-how, and intangible assets. California management review, 40(3), 55-79.
+
+Wagner, C. S., & Leydesdorff, L. (2005). Network structure, self-organization, and the growth of international collaboration in science. Research Policy, 34(10), 1608-1618.
+
+Wuchty, S., Jones, B. F., & Uzzi, B. (2007). The increasing dominance of teams in production of knowledge. Science, 316(5827), 1036-1039.
+
+Yoon, B., & Park, Y. (2004). A text-mining-based patent network: Analytical tool for high-technology trend. The Journal of High Technology Management Research, 15(1), 37–50.
+
+Yoon, B., & Park, Y. (2005). A systematic approach for identifying technology opportunities: Keyword-based morphology analysis. Technological Forecasting and Social Change, 72(2), 145–160.
+
+Yu, X., & Zhang, B. (2017). Obtaining advantages from technology revolution: A patent roadmap for competition analysis and strategy planning. Technological Forecasting and Social Change.
+
+Zhang, J. (2017). Uncovering mechanisms of co-authorship evolution by multiplications-based link prediction. Information Processing & Management, 53(1), 42-51.
+
+16
+

processed/00138_W2613989322/document.md

@@ -0,0 +1,1765 @@
+## SURVEY RESULTS
+
+FROM
+
+THE 2003 INTELLECTUAL PROPERTY OWNERS ASSOCATION SURVEY ON STRATEGIC MANAGEMENT OF INTELLECTUAL PROPERTY
+
+Iain M. Cockburn
+
+Boston University and NBER
+
+and
+
+Rebecca Henderson MIT and NBER
+
+October 2003
+
+Contact information: Prof. Iain M. Cockburn, School of Management, Boston University, 595 Commonwealth Ave, Boston MA 02215. cockburn@bu.edu, 617-353-3775. Acknowledgements: Ronald A. Bleeker, IPO Corporate IP Management Committee Vice Chair, Robert R. Schroeder, IPO Corporate IP Management Chair, and Rob Williamson, PetrashWilliamson, provided extensive advice and assistance with survey design and administration. Tony Briggs, doctoral candidate at MIT, also provided valuable input, and we thank Rohit Dogra and Raghu Lalgudi for research assistance. This project would not have been possible without the input of the individuals and companies who devoted scarce time to responding to the survey questionnaire. Support from the Intellectual Property Owners Association, and the sponsorship of Delphion Inc. (now Thomson Delphion), is gratefully acknowledged.
+
+---
+
+## A. Sample Frame and Respondents
+
+The survey questionnaire was administered by mail in the late summer of 2002 to contacts at IPO member companies, as well as to a list of senior IP managers maintained by Delphion Inc. After an initial round of questionnaires were returned, follow-up phone calls were made to the IPO membership list, as well as to a small number of companies identified by Delphion as “high priority targets.” In all, 66 usable questionnaires were returned, representing a response rate of slightly over 30% for the IPO member companies, and under 5% from the Delphion mailing list.
+
+Respondents were largely senior legal staff of the corporation with responsibility for IP or technology: 44% identified themselves as “Chief Patent Counsel” or the equivalent, 21% as “Assistant General Counsel”. Of the remainder, 34% identified themselves as General Counsel of the corporation or senior executives (VicePresident or equivalent).
+
+Respondents overwhelmingly had some university-level technical education. 77% had at least a Bachelor's degree in engineering, physics, or life sciences, and 12% of respondents had a graduate degree in a technical subject.
+
+88% of respondents had a law degree, and they averaged just over 20 years since passing the bar exam. 75% had been admitted to the Patent Bar. Only 10% of respondents had no legal experience outside IP, with more than 2/3 reporting "extensive" experience with patent prosecution, and 1/4 reporting "extensive" experience with litigation.
+
+Respondent's companies spanned a wide range of industries, with the majority drawn from the chemical (22%), IT and communications (44%), life sciences (15%) and mechanical (16%) sectors.
+
+76% of respondents thought that it made sense to treat their company as a single entity in terms of strategic decision making or corporate policy, and more than 86% felt able to answer survey questions in terms of their company as whole.
+
+A.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Legislation and Policy
+
+## B. Legislation, Policy etc.
+
+1. How have recent patent reforms and court decisions in the US affected your company?
+
+(i) Publication of applications
+
+Very few respondents reported a negative impact.
+
+(ii) Patents on methods of doing business
+
+A large majority of respondents were neutral. Of the remainder, slightly fewer reported a positive rather than negative impact.
+
+(iii) Festo
+
+Opinion on the impact of the Festo decision was split.
+
+Most respondents reported a neutral impact, among those who did not, slightly more reported that the decision in this case had damaged their company.
+
+(iv) Recent trends in CAFC decisions
+
+A wide range of assessments were obtained. 1/3 of respondents reported a negative impact, and 1/5 a reported a positive impact.
+
+![Figure](figures/figure_012.png)
+
+![Figure](figures/figure_013.png)
+
+![Figure](figures/figure_014.png)
+
+![Figure](figures/figure_015.png)
+
+B.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Legislation and Policy
+
+2. How do you think the following changes to the US patent system would affect your company?
+
+### (i) European-style post-grant opposition
+
+On balance, respondents predicted a positive impact of introducing a post grant opposition process.
+
+### (ii) Extensive prior user rights
+
+Although half of respondents predict a beneficial impact almost 1/4 took the opposite view.
+
+### (iii) Registration without automatic examination
+
+Assessments of the impact of introducing a registration system were strongly negative, with less than 5% of respondents predicting a beneficial effect.
+
+### (iv) Increased prior art search requirements
+
+The majority of respondents anticipate a beneficial effect of increased search requirements, though 1/5 predict a negative impact.
+
+![Figure](figures/figure_011.png)
+
+![Figure](figures/figure_012.png)
+
+![Figure](figures/figure_013.png)
+
+![Figure](figures/figure_014.png)
+
+B.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Legislation and Policy
+
+## (v) "Raising the bar" for patentability
+
+Only 5% of respondents anticipate being damaged by reforms to the patent process that would make patentability standards more difficult to reach.
+
+![Figure](figures/figure_004.png)
+
+## (vi)Adopting a first-to-file basis for priority
+
+On balance, opinion on the impact of introducing a first-to-file system was positive, with half of respondents predicting a beneficial effect on their company. Just under 1/5 expect to be damaged by moving away from the first-to-invent principle.
+
+![Figure](figures/figure_007.png)
+
+B.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+## C. The Role of IP in Company Strategy
+
+### 1. IP strategy as an aspect of my company's day-to-day business decisions
+
+IP issues are an integral part of doing business for 2/3 of respondents' companies
+
+![Figure](figures/figure_005.png)
+
+### 2. Impact of my company's involvement in IP disputes in the past 5 years
+
+A majority of respondents report that their companies are successful in IP disputes
+
+![Figure](figures/figure_008.png)
+
+### 3. My company's competitive advantage:
+
+### (i) is built on proprietary technology
+
+Only a handful of responding companies do not compete on the basis of proprietary technology
+
+![Figure](figures/figure_012.png)
+
+### (ii) is driven by rapid new technology development
+
+Speed in developing new technology is a source of competitive advantage for 3/4 of respondents
+
+![Figure](figures/figure_015.png)
+
+C.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+### (iii)is driven by marketing
+
+Just under 1/3 of respondents derive competitive advantage from capabilities other than marketing
+
+![Figure](figures/figure_004.png)
+
+### (iv)is driven by manufacturing
+
+Just under 1/3 of respondents derive competitive advantage from capabilities other than manufacturing
+
+![Figure](figures/figure_007.png)
+
+## 3. My company's competitive advantage would quickly erode:
+
+### (i) without patent protection
+
+Competitive advantage is sustained by patents for 2/3 of respondents
+
+![Figure](figures/figure_011.png)
+
+### (ii) without trade secret protection
+
+Loss of trade secret protection would damage 80% of respondents' ability to sustain their competitive advantage
+
+![Figure](figures/figure_014.png)
+
+C.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+## (iii) without copyright protection
+
+Copyright is an important means of sustaining competitive advantage for more than 1/3 of respondents
+
+![Figure](figures/figure_004.png)
+
+## (iv)without trademark protection
+
+More than 2/3 of respondents rely on trademarks to sustain competitive advantage
+
+![Figure](figures/figure_007.png)
+
+## (v) without other IP protection (mask rights, breeders rights etc.)
+
+Sui generis forms of IP protection protect competitive advantage of less and 1/5 of respondents
+
+![Figure](figures/figure_010.png)
+
+## (vi)without our internally developed know-how
+
+3/4 of respondents build competitive advantage on internal know-how
+
+![Figure](figures/figure_013.png)
+
+C.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+4. My company would spend significantly less on R&D and technology development:
+
+### (i) without patents
+
+A wide range of responses was obtained. Losing protection afforded by the patent system would strongly affect R&D spending of 1/3 of responding companies, but over 40% would not lower their spending.
+
+![Figure](figures/figure_005.png)
+
+### (ii) without trade secrets
+
+1/3 of respondents would maintain R&D spending after losing trade secret protection
+
+![Figure](figures/figure_008.png)
+
+### (iii) without trademarks
+
+2/3 of respondents would maintain R&D spending after losing trademark protection for their products
+
+![Figure](figures/figure_011.png)
+
+### (iv)without copyrights
+
+4/5 of respondents would maintain R&D spending in the absence of copyright
+
+![Figure](figures/figure_014.png)
+
+C.4
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+5. The most profitable companies in our industry:
+
+### (i) have built up significant IP assets
+
+A large majority of respondents report a strong connection in their industry between profitability possessing IP
+
+![Figure](figures/figure_005.png)
+
+### (ii) aggressively assert their IP rights
+
+2/3 of respondents report a connection between profitability and a strong IP “offence”
+
+![Figure](figures/figure_008.png)
+
+### (iii) react aggressively to IP activity by competitors
+
+Strong "defense" is associated with profitability in more than 2/3 of respondents' industries
+
+![Figure](figures/figure_011.png)
+
+### (iv)invest in IP mostly for defensive reasons
+
+Slightly less than 1/2 of respondents report an association between profitability and a defensive posture on IP strategy
+
+![Figure](figures/figure_014.png)
+
+C.5
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+6. Some firms have dominated our industry by:
+
+### (i) controlling key patents
+
+1/5 of respondents report that control of key patents leads to a dominant position in their industry, though overall respondents were evenly split
+
+![Figure](figures/figure_005.png)
+
+### (ii) holding important technology as trade secrets
+
+Overall, respondents were evenly split on the competitive value of keeping technology secret
+
+![Figure](figures/figure_008.png)
+
+### (iii) owning key trademarks
+
+2/3 of respondents reject the idea that trademarks can confer a dominant position in the market.
+
+![Figure](figures/figure_011.png)
+
+C.6
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+7. Competitors' patent portfolios seriously constrain my company's freedom to operate by:
+
+(i) foreclosing technology development in important areas
+
+Competitor IP is a significant constraint on the scope of technology development for 1/4 of respondents
+
+![Figure](figures/figure_005.png)
+
+(ii) slowing the pace of technology development
+
+Less than 1/5 of respondents report that competitor
+
+
+IP slows their technology development
+
+
+1
+
+![Figure](figures/figure_008.png)
+
+(iii) blocking access to important markets
+
+Competitor IP blocks market access for about 1/4 of respondents
+
+![Figure](figures/figure_011.png)
+
+C.7
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+8. In our industry:
+
+(i) product lifecycles are typically shorter than the time it takes to get a patent issued
+
+The pendency period is a problem for 1/4 of respondents
+
+![Figure](figures/figure_005.png)
+
+(ii) patents are a major obstacle to establishing technology standards
+
+Very few respondents report that patents negatively affect standard-setting in their industry
+
+![Figure](figures/figure_008.png)
+
+(iii) IP is primarily important as a bargaining chip in negotiating access to technology
+
+Respondents were evenly split
+
+![Figure](figures/figure_011.png)
+
+(iv)it is very difficult to keep new technology secret for long
+
+Information about new technology diffuses rapidly in the industries of 4/5 of respondents
+
+![Figure](figures/figure_014.png)
+
+C.8
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+(v) Non-Compete and Non-Disclosure agreements are an effective way to control tech transfer
+
+Just under 1/3 of respondents report that NCAs and NDAs are ineffective in their industry
+
+![Figure](figures/figure_004.png)
+
+(vi)to retain control of technology companies have to be able to retain key individuals
+
+3/4 of respondents agreed
+
+![Figure](figures/figure_007.png)
+
+C.9
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+9. In my company, thought leaders generally believe that:
+
+(i) patent documents tend to disclose too much valuable information to our competitors.
+
+3/4 of respondents disagree
+
+![Figure](figures/figure_005.png)
+
+(ii) many of our most important ideas cannot be effectively protected with patents.
+
+A significant minority of respondents agree, reporting serious limits to patent protection.
+
+![Figure](figures/figure_008.png)
+
+(iii) the really important intellectual assets are the skills and knowledge of our people
+
+Human capital is highly valued in the great majority of responding companies
+
+![Figure](figures/figure_011.png)
+
+(iv)with enough money and the right people most patents can be invented around
+
+In 3/4 of responding companies, patents are not thought to be insuperable obstacles
+
+![Figure](figures/figure_014.png)
+
+C.10
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+10. In my company:
+
+(i) top leadership is only rarely involved in IP issues
+
+Senior management pay little attention to IP in about of 1/3 of responding companies
+
+![Figure](figures/figure_005.png)
+
+(ii) each business unit and product team has its own individual IP policies/plans/objectives
+
+More than 1/2 of responding companies allow considerable autonomy in IP strategy
+
+![Figure](figures/figure_008.png)
+
+(iii) business units which obtain IP assets can compete more effectively for internal resources
+
+![Figure](figures/figure_010.png)
+
+2/3 disagree
+
+C.11
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+11. We think of our patent strategy as a success if we:
+
+(i) avoid being sued for patent infringement.
+
+Almost 90% agree
+
+![Figure](figures/figure_005.png)
+
+(ii) settle IP disputes on favorable terms.
+
+Only 5% disagree
+
+![Figure](figures/figure_008.png)
+
+(iii)generate licensing revenue
+
+1/3 of respondents do not measure success in terms of licensing revenue
+
+![Figure](figures/figure_011.png)
+
+(iv)prevent competitors from copying our products.
+
+For more than 10% of respondents, prevention of copying is not an important goal of patent strategy
+
+![Figure](figures/figure_014.png)
+
+C.12
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+(v) make it more expensive for competitors to be in business against us.
+
+90% of respondents measure success in terms of their ability to raise rivals' costs
+
+(vi)have a unique product position.
+
+90% of respondents measure success in these terms
+
+(viii) develop new partner relationships
+
+1/3 of respondents do not measure success in these terms
+
+(vii)influence adoption of technology by our industry partners
+
+80% of respondents agree
+
+![Figure](figures/figure_010.png)
+
+![Figure](figures/figure_011.png)
+
+![Figure](figures/figure_012.png)
+
+![Figure](figures/figure_013.png)
+
+C.13
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+(ix)are able to measure and manage legal risk to our businesses.
+
+85% of respondents measure success of patent strategy in terms of ability to manage risk
+
+(x) maintain freedom to operate in core technologies and businesses.
+
+98% of respondents measure success of their patent strategy in these terms, with almost 3/4 “strongly agreeing”
+
+![Figure](figures/figure_006.png)
+
+![Figure](figures/figure_007.png)
+
+C.14
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+12. The rank of my company's IP assets in order of their dollar value is: (1=most valuable, 2=next most valuable etc.
+
+(i) Patents
+
+Over half of respondents identified patents as their company's most valuable IP asset
+
+![Figure](figures/figure_005.png)
+
+(ii) Trademarks
+
+Almost 1/5 of respondents rated trademarks as their company's most valuable IP asset
+
+![Figure](figures/figure_008.png)
+
+(iii) Copyrights
+
+Almost all respondents rated the value of
+
+copyrights very low compared to other IP assets
+
+![Figure](figures/figure_011.png)
+
+C.15
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Role of IP in Strategy
+
+### (iv) Trade secrets
+
+Though only occasionally rated the most valuable among IP assets, trade secrets were the most frequently chosen no. 2 or 3
+
+![Figure](figures/figure_004.png)
+
+### (v) Know-How
+
+1/5 of respondents rated know how as the most valuable of their companies' IP assets, know how was also often rated as 2 nd or 3 rd most valuable IP asset class
+
+![Figure](figures/figure_007.png)
+
+### 12.a How much of the total monetary value of your company's IP assets do the following represent:
+
+<table><tr><td colspan="5">Fraction of total monetary value of IP assets</td></tr><tr><td></td><td>N</td><td>Average</td><td>Minimum</td><td>Maximum</td></tr><tr><td>Patents</td><td>23</td><td>44.1%</td><td>0</td><td>80%</td></tr><tr><td>Trademarks</td><td>21</td><td>19.0%</td><td>0</td><td>80%</td></tr><tr><td>Copyrights</td><td>16</td><td>6.8%</td><td>0</td><td>45%</td></tr><tr><td>Trade secrets</td><td>21</td><td>17.4%</td><td>0</td><td>40%</td></tr><tr><td>Know-how</td><td>21</td><td>13.8%</td><td>0</td><td>40%</td></tr></table>
+
+
+C.16
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Technology Development
+
+## D. Technology Development
+
+1. For your company, please rank the three most important sources of new scientific and technical ideas at each stage of the technology development cycle: (1=most important, 2=next most important etc.)
+
+<table><tr><td></td><td>Early stage discovery, basic research</td><td>New product development, applied research</td><td>Product enhancement, incremental innovation</td></tr><tr><td>Internal R&amp;D</td><td>1.2</td><td>1.2</td><td>1.4</td></tr><tr><td>Vendors</td><td>2.5</td><td>2.5</td><td>2.3</td></tr><tr><td>Customers</td><td>2.3</td><td>2.1</td><td>1.7</td></tr><tr><td>Competitors</td><td>2.3</td><td>2.4</td><td>2.6</td></tr><tr><td>Outside consultants</td><td>2.1</td><td>2.4</td><td>2.8</td></tr><tr><td>Partners e.g. alliances, JVs, etc.</td><td>2.2</td><td>2.3</td><td>2.4</td></tr><tr><td>“Arm’s length” licensed-in technology</td><td>2.3</td><td>2.5</td><td>2.6</td></tr><tr><td>R&amp;D performed in unrelated industries</td><td>3.0</td><td>N/A</td><td>N/A</td></tr><tr><td>Government/University relationships</td><td>2.5</td><td>2.6</td><td>3</td></tr><tr><td>Professional or academic publications</td><td>2.3</td><td>2.5</td><td>2.3</td></tr><tr><td>Patent disclosures</td><td>2.3</td><td>2.0</td><td>2.4</td></tr></table>
+
+
+(table entries are average ranking, lower number means more important)
+
+Internal R&D is the most important source of new ideas at all stages of the technology development cycle. Customers are an important source during late stage, incremental innovation and product enhancement. Patent documents are rated more important than competitors, in-licensing, professional publications or government and university partnerships, and roughly equivalent to partnerships and joint ventures. Government/university relationships received a surprisingly low average ranking.
+
+D.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Technology Development
+
+2. Competitors' patents play an important role in shaping the pace and direction of my company's technology development and R&D through:
+
+(i) decisions not to pursue otherwise promising technologies.
+
+1/2 of respondents agree
+
+![Figure](figures/figure_005.png)
+
+(ii) decisions to abandon later-stage development of otherwise promising technologies.
+
+Over 2/3 disagree
+
+![Figure](figures/figure_008.png)
+
+(iii) slowing down the technology development cycle.
+
+3/4 disagree
+
+![Figure](figures/figure_011.png)
+
+D.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Technology Development
+
+3. At my company, scientific and technical personnel are:
+
+(i) recognized for invention disclosures
+
+Typical reward: plaque, dinner, t-shirt
+
+![Figure](figures/figure_005.png)
+
+(ii) recognized for initial patent filing
+
+Typical reward: $500-$1000. Maximum was $3000
+
+![Figure](figures/figure_008.png)
+
+(iii) recognized for patent grants
+
+Typical reward $1000 plus dinner and plaque.
+
+Maximum was \$10million!
+
+![Figure](figures/figure_012.png)
+
+D.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Technology Development
+
+### (iv) recognized for generating other IP
+
+If reward is given, it is typically for “ trade secret ” and is a token amount of money. “ Part of the job ” was a common comment from respondents.
+
+### (v) promoted, at least in part, on their record of generating patents.
+
+The majority of companies do taken inventors record of generating patents into account in promotion decisions, though very few registered strong agreement and 40% of companies do not appear to use this practice.
+
+### (vi) given significant monetary rewards for generating patents.
+
+“Significant” appears to have been an important modifier for this question. The majority of companies do reward inventors, but rarely with substantial amounts of money.
+
+![Figure](figures/figure_008.png)
+
+![Figure](figures/figure_009.png)
+
+![Figure](figures/figure_010.png)
+
+D.4
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Technology Development
+
+4. We routinely monitor our competitors' IP activity:
+
+(i) for competitive awareness
+
+Almost 90% of companies do this.
+
+![Figure](figures/figure_005.png)
+
+(ii) for oppositions or interferences
+
+More than 1/3 do not monitor competitors for this
+
+
+purpose. This may reflect differences in international patenting practices.
+
+![Figure](figures/figure_008.png)
+
+(iii) for technology
+
+1/3 of companies do not do this.
+
+![Figure](figures/figure_011.png)
+
+D.5
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Operations
+
+## E. IP Organization and Operations
+
+1. What is the exact title of the senior person responsible for intellectual property issues in your company?
+
+### General Counsel, Chief Patent Counsel
+
+2. Does that person sit of the Board of Directors / Main Board of your company
+
+### 11% YES
+
+3. Is that person a member of your company's senior internal management committee?
+
+### 29% YES
+
+4. Do you have a high-level policy group dedicated to IP issues?
+
+### 41% YES
+
+If so,
+
+(i) How often does this group meet?
+
+For those companies with such a group, 1/2 meet at least once per month.
+
+![Figure](figures/figure_014.png)
+
+(ii) How many people are involved?
+
+Typical group size was 5 to 8 people.
+
+![Figure](figures/figure_017.png)
+
+E.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Operations
+
+5. Do you have a formal Patent Review Board, or similar committee or process that makes decisions about whether to patent inventions brought forward?
+
+## 51% YES
+
+If so,
+
+(i) How often does this group meet?
+
+![Figure](figures/figure_006.png)
+
+(ii) How many people are involved?
+
+![Figure](figures/figure_008.png)
+
+(iii)Other than IP counsel, which business functions are normally represented in this group?
+
+![Figure](figures/figure_010.png)
+
+E.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Operations
+
+6. What fraction of your company's total IP headcount report directly to operating units, as opposed to the legal department?
+
+A substantial share of IP staff report to operating units in a very small minority of responding companies. In 3/4 of companies, all of the IP staff report to the legal function.
+
+![Figure](figures/figure_004.png)
+
+7. Who has significant input to pay and promotion decisions for IP attorneys and staff?
+
+<table><tr><td>74%</td><td>Senior IP Counsel / Head of IP function.</td></tr><tr><td>88%</td><td>Chief Legal Officer / Head of legal function.</td></tr><tr><td>17%</td><td>Business unit managers.</td></tr><tr><td>27%</td><td>Senior non-legal executives.</td></tr><tr><td>18%</td><td>Corporate HR</td></tr></table>
+
+
+8. Who has authority over pay and promotion decisions?
+
+General Counsel
+
+9. IP attorneys or specialist staff are frequently / occasionally / never assigned to product development teams, or similar line activities in operating business units.
+
+27% NEVER, 73% N/A
+
+10.Do you routinely outsource IP related tasks? 97% YES
+
+If yes, what fraction of the total activity in each of the following tasks do you outsource?
+
+<table><tr><td>44%</td><td>Prior Art searches</td></tr><tr><td>62%</td><td>Drafting patent applications</td></tr><tr><td>56%</td><td>US patent prosecution</td></tr><tr><td>68%</td><td>Foreign filing</td></tr><tr><td>15%</td><td>Scanning competitive IP</td></tr><tr><td>66%</td><td>Opinions on validity/infringement</td></tr><tr><td>38%</td><td>Maintaining the patent portfolio (renewals etc.)</td></tr><tr><td>38%</td><td>Enforcement</td></tr><tr><td>88%</td><td>Litigation</td></tr><tr><td>12%</td><td>Contracts \&amp; Licensing</td></tr></table>
+
+
+11.Do you use in-house patent agents?
+
+42% YES
+
+E.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Operations
+
+12.Do you use a packaged docketing system to manage your IP portfolio? 46% YES
+
+13.Do you use any other specialized software applications to automate or support IP management decisions?
+
+
+26% YES
+
+
+1
+
+14. For what fraction of the patents in your portfolio have you calculated a monetary value
+
+Very few companies appear to be able to calculate a monetary value for more than a handful of patents in their portfolio. More than 1/3 of respondents who answered this question indicated that their company has not valued ANY of its patents.
+
+(Category shares are reported as percent of nonmissing responses.)
+
+![Figure](figures/figure_007.png)
+
+15. Do you use analytical tools to evaluate your patent portfolio? 29% YES If so, which?
+
+<table><tr><td>3%</td><td>Real options</td></tr><tr><td>18%</td><td>Visualization tools</td></tr><tr><td>20%</td><td>Citation mapping</td></tr><tr><td>3%</td><td>Monte Carlo simulation</td></tr><tr><td>9%</td><td>Other: typically "standard accounting tools"</td></tr></table>
+
+
+16. How do you measure the effectiveness of your IP group: (check all that apply)
+
+<table><tr><td>47%</td><td>Number of invention disclosures reviewed</td></tr><tr><td>38%</td><td>Time from initial disclosure to patenting decision</td></tr><tr><td>80%</td><td>Number of patent applications filed</td></tr><tr><td>74%</td><td>Number of patents granted</td></tr><tr><td>14%</td><td>Time from application to issuance</td></tr><tr><td>39%</td><td>Royalties received</td></tr><tr><td>18%</td><td>Royalties paid</td></tr><tr><td>77%</td><td>Satisfaction of operating company managers</td></tr><tr><td>27%</td><td>Measures of patent quality (if so describe): typically "claim coverage"</td></tr><tr><td>17%</td><td>Number of times named as a defendant in a patent dispute</td></tr><tr><td>20%</td><td>Number of or ratio of favorable/unfavorable dispute resolutions</td></tr><tr><td>9%</td><td>Other: typically "profitability" or "cost savings"</td></tr></table>
+
+
+E.4
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Operations
+
+17. My company's biggest challenges in managing IP are: (rank all that apply in order of importance: 1=most important etc.)
+
+### (i) Budget constraints
+
+70% of respondents who ranked this option rated it the top priority
+
+![Figure](figures/figure_005.png)
+
+### (ii) Lack of internal expertise.
+
+Almost 2/3 did not rank this option, of those who did, less than 1/2 rated it 2nd or higher.
+
+![Figure](figures/figure_008.png)
+
+### (iii) A lack of understanding of the importance of IP
+
+More than 1/2 of respondents did not rank this option, of those who did more than 1/2 rated it the top priority
+
+![Figure](figures/figure_011.png)
+
+E.5
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Operations
+
+### (iv) Lack of management support.
+
+Almost 2/3 did not rank this option, of those who did, 1/3 ranked it 4 th or below. The level of nonresponse is difficult to interpret. Respondents may have had difficulty understanding the question, or may have been concerned about the consequences of registering a response even in an anonymous survey.
+
+![Figure](figures/figure_004.png)
+
+### (vi) No perceived need.
+
+![Figure](figures/figure_006.png)
+
+E.6
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Management of IP
+
+## F. IP Management
+
+1. Our IP strategy has been developed:
+
+![Figure](figures/figure_004.png)
+
+2. Our written Intellectual Asset Plan explicitly incorporates business strategy goals.
+
+30% of responding companies appear not to have a written intellectual asset plan.
+
+![Figure](figures/figure_007.png)
+
+3. A concrete decision as to whether to patent, publish or hold as trade secret is reached for every significant invention.
+
+2/3 of companies do not do this.
+
+![Figure](figures/figure_010.png)
+
+4. We use specific hard-and-fast quantitative criteria to guide our patenting decisions.
+
+The high non-response rate suggests that, in total, over 1/3 of companies do not do this.
+
+![Figure](figures/figure_013.png)
+
+F.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Management of IP
+
+5. We have developed a clear set of guidelines as to what should be patented.
+
+1/3 of companies have not done this
+
+6. We patent nearly every non-trivial idea that is brought forward.
+
+Patenting is not "automatic" except in a small minority of companies.
+
+7. Patenting decisions are often controversial.
+
+8. Budget issues tend to dominate our filing decisions
+
+![Figure](figures/figure_008.png)
+
+![Figure](figures/figure_009.png)
+
+![Figure](figures/figure_010.png)
+
+![Figure](figures/figure_011.png)
+
+F.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Management of IP
+
+9. We audit our IP portfolio at the following intervals:
+
+![Figure](figures/figure_003.png)
+
+10. In my company, patentability has been compromised by:
+
+(i) publication before establishing priority
+
+![Figure](figures/figure_006.png)
+
+(ii) violating the on-sale bar
+
+![Figure](figures/figure_008.png)
+
+(iii) non-written disclosure
+
+![Figure](figures/figure_010.png)
+
+F.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Management of IP
+
+11.Normally, when decisions are made about whether a patent should be filed:
+
+<table><tr><td></td><td>Is consulted</td><td>Plays a significant role</td><td>Has a veto</td><td>Normally controls the decision</td><td>Is not Involved</td></tr><tr><td>R&amp;D</td><td>13%</td><td>56%</td><td>3%</td><td>28%</td><td>0</td></tr><tr><td>IP counsel</td><td>8%</td><td>37%</td><td>19%</td><td>37%</td><td>0</td></tr><tr><td>Non-IP legal</td><td>5%</td><td>4%</td><td>2%</td><td>2%</td><td>88%</td></tr><tr><td>Marketing</td><td>30%</td><td>30%</td><td>3%</td><td>5%</td><td>33%</td></tr><tr><td>Manufacturing</td><td>63%</td><td>17%</td><td>15%</td><td>3%</td><td>2%</td></tr><tr><td>Product managers</td><td>27%</td><td>40%</td><td>7%</td><td>2%</td><td>25%</td></tr><tr><td>Finance</td><td>2 %</td><td>2%</td><td>0</td><td>0</td><td>97%</td></tr><tr><td>Senior corporate management</td><td>18 %</td><td>4%</td><td>4%</td><td>5%</td><td>70%</td></tr><tr><td>Corporate Business Development</td><td>7%</td><td>7%</td><td>2%</td><td>0</td><td>84%</td></tr><tr><td>Cross-functional IP group</td><td>4%</td><td>25%</td><td>2%</td><td>32%</td><td>38%</td></tr></table>
+
+
+12. Normally, when strategic decisions must be made during the patent application process
+
+<table><tr><td></td><td>Is consulted</td><td>Plays a significant role</td><td>Has a veto</td><td>Normally controls the decision</td><td>Is not Involved</td></tr><tr><td>R&amp;D</td><td>39%</td><td>39%</td><td>5%</td><td>11%</td><td>7%</td></tr><tr><td>IP counsel</td><td>5%</td><td>21%</td><td>8%</td><td>67%</td><td>0</td></tr><tr><td>Non-IP legal</td><td>4%</td><td>4%</td><td>2%</td><td>2%</td><td>89%</td></tr><tr><td>Marketing</td><td>19%</td><td>7%</td><td>3%</td><td>0</td><td>71%</td></tr><tr><td>Manufacturing</td><td>16%</td><td>2%</td><td>2%</td><td>2%</td><td>77%</td></tr><tr><td>Product managers</td><td>36%</td><td>10%</td><td>2%</td><td>2%</td><td>50%</td></tr><tr><td>Finance</td><td>0</td><td>0</td><td>2%</td><td>0</td><td>98%</td></tr><tr><td>Senior corporate management</td><td>9%</td><td>5%</td><td>4%</td><td>2%</td><td>81%</td></tr><tr><td>Corporate Business Development</td><td>5%</td><td>7%</td><td>0 %</td><td>2%</td><td>89%</td></tr><tr><td>Cross-functional IP group</td><td>16%</td><td>7%</td><td>0 %</td><td>10%</td><td>67%</td></tr></table>
+
+
+F.4
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Management of IP
+
+## 13. Normally, when decisions must be made about paying maintenance fees
+
+<table><tr><td></td><td>Is consulted</td><td>Plays a significant role</td><td>Has a veto</td><td>Normally controls the decision</td><td>Is not Involved</td></tr><tr><td>R&amp;D</td><td>33%</td><td>23%</td><td>2%</td><td>25%</td><td>17%</td></tr><tr><td>IP counsel</td><td>9%</td><td>41%</td><td>13%</td><td>33%</td><td>6%</td></tr><tr><td>Non-IP legal</td><td>6%</td><td>2%</td><td>2%</td><td>2%</td><td>88%</td></tr><tr><td>Marketing</td><td>29%</td><td>20%</td><td>4%</td><td>5%</td><td>43%</td></tr><tr><td>Manufacturing</td><td>16%</td><td>8%</td><td>2%</td><td>2%</td><td>43%</td></tr><tr><td>Product managers</td><td>16%</td><td>8%</td><td>2%</td><td>2%</td><td>73%</td></tr><tr><td>Finance</td><td>23%</td><td>23%</td><td>2%</td><td>9%</td><td>43%</td></tr><tr><td>Senior corporate management</td><td>2%</td><td>0</td><td>0</td><td>0</td><td>98%</td></tr><tr><td>Corporate Business Development</td><td>11%</td><td>4%</td><td>2%</td><td>7%</td><td>76%</td></tr><tr><td>Cross-functional IP group</td><td>15%</td><td>4%</td><td>0</td><td>2%</td><td>80%</td></tr></table>
+
+
+## 14. When we want to understand a competitor's IP position we use:
+
+(i) In-house IP specialists
+
+![Figure](figures/figure_006.png)
+
+(ii) Reports from in-house R&D
+
+![Figure](figures/figure_008.png)
+
+F.5
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Management of IP
+
+(iii) Out-sourced search services
+
+(iv) Outside IP counsel
+
+![Figure](figures/figure_004.png)
+
+![Figure](figures/figure_005.png)
+
+## Patent Search
+
+15. We always do a patent search before initiating any R&D or product development effort.
+
+![Figure](figures/figure_008.png)
+
+F.6
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Management of IP
+
+16. Now that patent applications are being published in the US, we are becoming aware of competitors' IP activity much sooner than in the past.
+
+![Figure](figures/figure_003.png)
+
+17.Our company has adequate internal search capabilities for:
+
+(i) supporting R&D activities.
+
+![Figure](figures/figure_006.png)
+
+(ii) competitor intelligence
+
+![Figure](figures/figure_008.png)
+
+(iii) patent searching and analysis
+
+![Figure](figures/figure_010.png)
+
+F.7
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Management of IP
+
+(iv) trademark search and analyses
+
+![Figure](figures/figure_003.png)
+
+18. Patent searches are normally performed by people from: (Check all that apply)
+
+<table><tr><td>61%</td><td>R&amp;D</td></tr><tr><td>83%</td><td>In-house legal/IP</td></tr><tr><td>8%</td><td>Business function</td></tr><tr><td>26%</td><td>Out-sourced specialist search services</td></tr><tr><td>50%</td><td>Outside IP counsel</td></tr></table>
+
+
+19.Average number of hours of search per disclosure or per patent
+
+(i) to support Patent Review Board
+
+![Figure](figures/figure_008.png)
+
+(ii) to prepare an application for filing
+
+![Figure](figures/figure_010.png)
+
+F.8
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Management of IP
+
+(iii) for product clearance
+
+![Figure](figures/figure_003.png)
+
+(iv) to prepare for litigation
+
+![Figure](figures/figure_005.png)
+
+F.9
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Trade Secrets
+
+## G. Trade Secrets
+
+1. Please rank the importance to your company of the following means of controlling the use of technology:
+
+(i) Formal IP: Patents, Copyrights, Trademarks etc.
+
+Ranked most important by more than 2/3 of respondents.
+
+![Figure](figures/figure_006.png)
+
+(ii) Secrecy
+
+Ranked most important by more than 1/4 of respondents
+
+![Figure](figures/figure_009.png)
+
+(iii) Contract law: NDAs, licensing agreements, partnership agreements etc.
+
+Ranked second most important by many respondents.
+
+![Figure](figures/figure_012.png)
+
+G.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Trade Secrets
+
+2. Theft/misappropriation of trade secrets is a serious problem for our company
+
+2/3 of respondents disagreed. Very few registered strong agreement.
+
+![Figure](figures/figure_004.png)
+
+3. Trade secret law is an effective way for us to retain control of important technology
+
+2/3 of respondents agreed.
+
+![Figure](figures/figure_007.png)
+
+4. In our industry, it is usually easy to reverse-engineer competing products.
+
+The majority agreed.
+
+![Figure](figures/figure_010.png)
+
+5. In the past 5 years, about how many times has your company initiated legal action to protect trade secrets?
+
+Only a small minority report having initiated legal action on trade secrets more than once per year.
+
+![Figure](figures/figure_013.png)
+
+G.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Trade Secrets
+
+6. When trade secrets leak out, how important the following channels? Approximately what fraction would you say was through each channel?
+
+(i) inadvertent disclosure
+
+![Figure](figures/figure_004.png)
+
+(ii) employee turnover
+
+![Figure](figures/figure_006.png)
+
+(iii) employees in contact with competitors
+
+![Figure](figures/figure_008.png)
+
+Fraction
+
+Mean=13% Min=0 Max=50%
+
+Computed from non-missing responses
+
+Fraction
+
+Mean=44% Min=5% Max=100%
+
+Computed from non-missing responses
+
+Fraction
+
+Mean=26% Min=0 Max=100%
+
+Computed from non-missing responses
+
+G.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Trade Secrets
+
+### (iv) vendors
+
+![Figure](figures/figure_003.png)
+
+### (v) customers
+
+![Figure](figures/figure_005.png)
+
+### (vi) industrial espionage
+
+![Figure](figures/figure_007.png)
+
+## Fraction
+
+Mean=19% Min=5% Max=45%
+
+Computed from non-missing responses
+
+## Fraction
+
+Mean=19% Min=2% Max=75%
+
+Computed from non-missing responses
+
+## Fraction
+
+Mean=9% Min=0 Max=25%
+
+Computed from non-missing responses
+
+G.4
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Resources
+
+## H. Resources
+
+1. The IP function in my company is adequately funded.
+
+Resource constraints are a problem for 1/3 of respondents
+
+![Figure](figures/figure_005.png)
+
+2. Operating managers frequently say that we spend too much time/money on IP issues.
+
+If line management are dissatisfied with the level of resources devoted to IP, they are not communicating this to respondents.
+
+![Figure](figures/figure_008.png)
+
+3. Relative to the benefit derived, the amount of time/resources spent on IP issues by the following people is
+
+<table><tr><td></td><td>Too much</td><td>Appropriate</td><td>Too little</td></tr><tr><td>R&amp;D</td><td>38%</td><td>62%</td><td></td></tr><tr><td>Legal</td><td>26%</td><td>74%</td><td></td></tr><tr><td>Line management</td><td>41%</td><td>59%</td><td></td></tr><tr><td>Top leadership</td><td>37%</td><td>63%</td><td></td></tr></table>
+
+
+Again, no evidence for under-spending on IP.
+
+H.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Resources
+
+4. If more resources were devoted to IP in my company, the top three uses in order of priority should be:
+
+### (i) exploring licensing opportunities
+
+Overall, the second choice for top priority, though 1/2 of respondents did not choose this option.
+
+### (ii) getting involved in the new product design process earlier
+
+More respondents rated this the top priority for use of additional resources than any other option, though more than 1/2 of respondents did not choose this option.
+
+### (iii) obtaining more patents
+
+1/3 of respondents who chose this option rating this 1st or 2nd priority for use of additional IP resources,
+
+### (iv)enforcing our patent portfolio more aggressively
+
+Very few respondents rated this the top priority for for use of additional IP resources, even among the minority to chose this option.
+
+![Figure](figures/figure_011.png)
+
+![Figure](figures/figure_012.png)
+
+![Figure](figures/figure_013.png)
+
+![Figure](figures/figure_014.png)
+
+H.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Resources
+
+### (v) bringing prosecution in-house
+
+Most of the respondents who saw an advantage to in-house patent prosecution had evidently already done so.
+
+![Figure](figures/figure_004.png)
+
+### (vi)competitive intelligence
+
+Very few respondents rated this a priority use of additional IP resources.
+
+![Figure](figures/figure_007.png)
+
+### (vii) training
+
+Again, very few respondents rated this a priority use of additional IP resources.
+
+![Figure](figures/figure_010.png)
+
+H.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Resources
+
+## (viii) administration
+
+Very few respondents rated this a priority use of additional IP resources.
+
+![Figure](figures/figure_004.png)
+
+## (ix)adding IP staff
+
+Understaffing appears to be a source of concern for a small number of respondents, though 2/3 did not identify adding staff as a priority.
+
+![Figure](figures/figure_007.png)
+
+## (x) other: typically "more foreign filing"
+
+Very few respondents rated this a priority use of additional IP resources.
+
+![Figure](figures/figure_010.png)
+
+H.4
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Strategy
+
+## I. IP Strategy
+
+1. In our industry, companies actively publish technical material with the aim of limiting competitors' ability to obtain patents
+
+Almost 2/3 report little tendency in their industry to proactively use the public domain to “fence in” competitors
+
+![Figure](figures/figure_005.png)
+
+2. Defensive publication is an important strategic tool for my company
+
+2/3 of respondents discount the value of defensive publication as a strategic tool.
+
+![Figure](figures/figure_008.png)
+
+3. When we proactively place information in the public domain we use:
+
+<table><tr><td>17%</td><td>Publication in house-sponsored journals</td></tr><tr><td>42%</td><td>Placement of articles in the trade press</td></tr><tr><td>52%</td><td>Publication in academic or professional journals</td></tr><tr><td>8%</td><td>Statutory Invention Registration</td></tr><tr><td>27%</td><td>Marketing materials</td></tr><tr><td>20%</td><td>On-line publication services</td></tr><tr><td>14%</td><td>Company web site</td></tr><tr><td>5%</td><td>Other</td></tr></table>
+
+
+Technical journals and the trade press are the dominant channels for putting information in the public domain
+
+4. The value of any particular patent in our portfolio is determined by interrelationships with other patents that we hold: for most of our portfolio, “the whole is greater than the sum of the parts
+
+Almost 3/4 of respondents agree that inter-relationships among patents in their company's portfolio are important.
+
+![Figure](figures/figure_014.png)
+
+1.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Strategy
+
+5. For a typical new product developed by my company:
+
+(i) a single patent with broad claims would provide ample protection
+
+Respondents are evenly split
+
+![Figure](figures/figure_005.png)
+
+(ii) imitation would only be effectively prevented by a "thicket" of related patents
+
+3/4 of respondents who answered this question see “thicketing” as an effective means of preventing imitation
+
+![Figure](figures/figure_008.png)
+
+6. Approximately what fraction of your patent portfolio is
+
+<table><tr><td>47%</td><td>used to protect current product line from imitation</td></tr><tr><td>35%</td><td>used to establish freedom to operate</td></tr><tr><td>10%</td><td>generating license revenue</td></tr><tr><td>21%</td><td>held only for potential future own business use</td></tr><tr><td>9%</td><td>held only for potential licensing</td></tr><tr><td>32%</td><td>likely to be allowed to expire before full term</td></tr></table>
+
+
+Patents have multiple uses: prevention of imitation is the dominant motivation for holding patents. On average, 1/5 of the portfolio is being held for option value, and 1/3 of the portfolio is likely to be allowed to expire before term.
+
+1.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Strategy
+
+## 7. Filing lots of patents in a new area is likely to trigger an "arms race" among competitors
+
+Only 1/3 of respondents anticipate an aggressive response by competitors to this type of development
+
+![Figure](figures/figure_004.png)
+
+## 8. We always evaluate competitors' reactions before filing patents
+
+Only a tiny minority of respondents report that their company does this.
+
+![Figure](figures/figure_007.png)
+
+## 9. We always file patents as quickly as possible to avoid competitors getting priority
+
+“Brinksmanship” is not a popular game. Only 1/3 of respondents’ companies do not file promptly, whether for strategic or other reasons.
+
+![Figure](figures/figure_010.png)
+
+1.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+IP Strategy
+
+10.For the following types of innovations my company would normally respond by
+
+<table><tr><td>For an innovation with:</td><td>Patent intensively to obtain maximum coverage</td><td>File a single patent</td><td>Maintain as a trade secret</td><td>Proactively place it in the public domain</td><td>"Wait and see"</td></tr><tr><td>Minor technological significance</td><td>0</td><td>53%</td><td>30%</td><td>20%</td><td>14%</td></tr><tr><td>Major technological significance</td><td>86%</td><td>12%</td><td>3%</td><td>0</td><td>0</td></tr><tr><td>Minor market Impact</td><td>3%</td><td>52%</td><td>32%</td><td>12%</td><td>21%</td></tr><tr><td>Major market impact</td><td>92%</td><td>8%</td><td>3%</td><td>2%</td><td>0</td></tr></table>
+
+
+Major innovations result in intensive patenting, to the exclusion of other strategies, whereas minor inventions result in a variety of responses.
+
+11. If it came to our attention that a competitor firm was filing patents on the following types of innovations, my company would normally respond by:
+
+<table><tr><td>For an innovation with:</td><td>Filing or in-licensing large numbers of patents to “fence” in the competitor</td><td>Immediately obtaining legal opinion on scope of competitor's claims</td><td>Filing or in-licensing patents with a view to securing rights to practice our technology</td><td>Placing current technology in the public domain as a “spoiling” tactic to limit the competitor's activity</td><td>Adopting a “Wait and see” strategy</td><td>Gathering prior art for a possible validity challenge</td></tr><tr><td>Minor technological significance</td><td>0</td><td>8%</td><td>14%</td><td>5%</td><td>67%</td><td>21%</td></tr><tr><td>Major technological significance</td><td>15%</td><td>44%</td><td>44%</td><td>5%</td><td>11%</td><td>50%</td></tr><tr><td>Minor market Impact</td><td>0</td><td>6%</td><td>14%</td><td>5%</td><td>64%</td><td>17%</td></tr><tr><td>Major market Impact</td><td>17%</td><td>44%</td><td>48%</td><td>5%</td><td>12%</td><td>52%</td></tr></table>
+
+
+Competitor patents generate a broad range of responses for both major and minor innovations. Responses to minor innovations are generally “relaxed”, whereas competitor patents on major innovations generate prompt action on a variety of fronts.
+
+12. Does your company routinely make use of provisional applications? 62% YES
+
+1.4
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Licensing
+
+## J. Licensing
+
+1. In the past five years, net licensing revenue received by us has been:
+
+Respondents' companies were significant net beneficiaries from licensing
+
+![Figure](figures/figure_005.png)
+
+2. In negotiating licenses, maximizing licensing revenue is our number one priority
+
+Licensing is a complex phenomenon: almost 2/3 of respondents report that other priorities dominate revenue generation
+
+![Figure](figures/figure_008.png)
+
+3. Additional licensing revenue could only be realized from our portfolio at the cost of significantly impairing our competitive advantage
+
+Apparently there is "money on the table". Only 1/4 of respondents appear to have reached the limit of their ability to profitably increase licensing revenue.
+
+![Figure](figures/figure_011.png)
+
+J.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Licensing
+
+4. In decisions over whether a patent should be licensed:
+
+<table><tr><td></td><td>Is consulted</td><td>Plays a significant role</td><td>Has a veto</td><td>Normally controls the decision</td><td>Is not Involved</td></tr><tr><td>R&amp;D</td><td>55%</td><td>47%</td><td>20%</td><td>9%</td><td>6%</td></tr><tr><td>IP Counsel</td><td>35%^</td><td>76%</td><td>12%</td><td>12%</td><td>0</td></tr><tr><td>Non-IP legal</td><td>17%</td><td>14%</td><td>3%</td><td>2%</td><td>44%</td></tr><tr><td>Marketing</td><td>50%</td><td>29%</td><td>12%</td><td>6%</td><td>12%</td></tr><tr><td>Manufacturing</td><td>26%</td><td>6%</td><td>5%</td><td>0</td><td>42%</td></tr><tr><td>Product managers</td><td>49%</td><td>26%</td><td>11%</td><td>5%</td><td>14%</td></tr><tr><td>Finance</td><td>15%</td><td>6%</td><td>2%</td><td>0</td><td>53%</td></tr><tr><td>Senior corporate management</td><td>42%</td><td>23%</td><td>21%</td><td>20%</td><td>8%</td></tr><tr><td>Corporate Business Development</td><td>26%</td><td>20%</td><td>6%</td><td>6%</td><td>36%</td></tr><tr><td>Cross-functional IP group</td><td>27%</td><td>20%</td><td>5%</td><td>8%</td><td>32%</td></tr><tr><td>Other: (typically “contract group”)</td><td>2%</td><td>3%</td><td>3%</td><td>3%</td><td>6%</td></tr></table>
+
+
+A relatively broad range of functions are involved in licensing decisions, though manufacturing, finance, and non-IP legal are often excluded. R&D and IP counsel have the dominant role.
+
+5. What fraction of your patent portfolio is
+
+<table><tr><td>out-licensed?</td><td>17.6%</td></tr><tr><td>in-licensed?</td><td>8.4%</td></tr></table>
+
+
+6. What fraction of your trademark portfolio is
+
+<table><tr><td>out-licensed?</td><td>12.2%</td></tr><tr><td>in-licensed?</td><td>3%</td></tr></table>
+
+
+7. What fraction of your licenses are non-exclusive? 75.5%
+
+8. What fraction of your technology licensing activity does not involve any patents? 25.4%
+
+9. Does your company ever initiate technology development in the expectation that returns will be realized solely through licensing revenue, rather than through product sales?
+
+12% “YES”
+
+10. What share of your patent licensing activity involves cross-licensing, alliance agreements, etc, in which there is no direct financial consideration? 26.8%
+
+J.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Licensing
+
+11. What constrains your licensing revenue?
+
+<table><tr><td>65%</td><td>Difficulty in finding appropriate licensees</td></tr><tr><td>59%</td><td>Insufficient resources to pursue opportunities</td></tr><tr><td>15%</td><td>Difficulty in negotiating the terms of agreements</td></tr><tr><td>38%</td><td>Operating managers are reluctant to let us license</td></tr><tr><td>18%</td><td>Other: typically "lack of interest/focus/priority"</td></tr></table>
+
+
+12. Who in the organization receives revenue credit from a license?
+
+<table><tr><td>47%</td><td>the business unit which originated the technology</td></tr><tr><td>17%</td><td>the business unit which initiated the license negotiation</td></tr><tr><td>26%</td><td>Corporate</td></tr><tr><td>11%</td><td>IP/legal</td></tr><tr><td>21%</td><td>Other: typically "holding company"</td></tr></table>
+
+
+J.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Litigation and Enforcement
+
+## K. Litigation and Enforcement
+
+1. It would be straightforward to identify infringement of:
+
+(i) most of our product patents
+
+3/4 of respondents agreed
+
+![Figure](figures/figure_006.png)
+
+(ii) most of our process patents
+
+3/4 of respondents disagreed
+
+![Figure](figures/figure_009.png)
+
+(iii) most of our trademarks
+
+Only 2% of respondents disagreed
+
+![Figure](figures/figure_012.png)
+
+(iv) most of our copyrights
+
+Respondents were evenly split
+
+![Figure](figures/figure_015.png)
+
+K.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Litigation and Enforcement
+
+2. Compared to other kinds of civil litigation, patent suits tend to:
+
+(i) be too costly relative to the benefits that we receive.
+
+A slender majority of respondents who answered this question agreed.
+
+![Figure](figures/figure_005.png)
+
+(ii) have much less certain outcomes
+
+2/3 of respondents who answered this question agreed.
+
+![Figure](figures/figure_008.png)
+
+(iii) be harder to bring to resolution fast enough to properly protect our interests
+
+2/3 of respondents who answered this question agreed.
+
+![Figure](figures/figure_011.png)
+
+K.2
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Litigation and Enforcement
+
+3. In our industry, courts and arbitrators involved in IP litigation generally reach the "right" decisions.
+
+A clear majority of respondents appear to be satisfied with the accuracy of dispute resolution.
+
+4. The threat of a patent suit is usually enough to make us revise business decisions
+
+A slim majority of respondents disagree.
+
+5. In the past 5 years, about how many times did you file suit to enforce any of your patents?
+
+Just under 1/3 of respondents have not filed suit recently. A tiny minority are very active litigators, while the majority initiate, on average, one to two suits per year.
+
+6. In your industry, when damages are awarded or a financial settlement is reached in IP disputes, net of the cost of litigation, how do these amounts typically compare to the economic loss experienced by plaintiff
+
+Respondents' experience is highly varied: a significant fraction did not answer this question, and among those who did, there is no clear consensus.
+
+![Figure](figures/figure_010.png)
+
+![Figure](figures/figure_011.png)
+
+![Figure](figures/figure_012.png)
+
+![Figure](figures/figure_013.png)
+
+K.3
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Litigation and Enforcement
+
+7. Did any firm take action to enforce their IP rights against your firm? 89% YES
+
+How would you rate this action?
+
+For almost half of those who answered this question, defending IP suits consumed “major amounts of managerial time and attention”
+
+![Figure](figures/figure_005.png)
+
+8. Dealing with "nuisance" IP litigation from non-competitors consumes significant amounts of my company's time and resources
+
+Litigation by non-competitors was a significant resource drain for 1/3 of respondents.
+
+![Figure](figures/figure_008.png)
+
+K.4
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Quantitative Data
+
+## L. Quantitative Data
+
+<table><tr><td>In 2001:</td><td>N</td><td>Average</td><td>Std. Dev.</td><td>Minimum</td><td>Maximum</td></tr><tr><td>Number of disclosures from employees</td><td>50</td><td>442.5</td><td>614</td><td>0</td><td>3000</td></tr><tr><td>Fraction of internally generated disclosures being pursued</td><td>52</td><td>55.2%</td><td>24.6</td><td>0</td><td>100%</td></tr><tr><td>Unsolicited disclosures from non-employees</td><td>46</td><td>79.3</td><td>239</td><td>0</td><td>1500</td></tr><tr><td>Fraction of unsolicited disclosures being pursued</td><td>43</td><td>1.2%</td><td>2.7%</td><td>0</td><td>13%</td></tr><tr><td>Number of US patents applied for</td><td>54</td><td>264.1</td><td>379</td><td>0</td><td>2000</td></tr><tr><td colspan="6">Fraction of these for which also filed in Europe and Japan</td></tr><tr><td>Other OECD</td><td>54</td><td>58.7%</td><td>36.3%</td><td>0</td><td>100%</td></tr><tr><td>Emerging Markets</td><td>49</td><td>44.9%</td><td>34.4%</td><td>0</td><td>100%</td></tr><tr><td>Number of articles published in open literature</td><td>28</td><td>172.9</td><td>316.9</td><td>0</td><td>1500</td></tr><tr><td>Fraction of articles published in trade press vs. academic journals</td><td>18</td><td>51.8%</td><td>37.9</td><td>2%</td><td>100%</td></tr><tr><td>Approximate licensing revenue $\$ MM</td><td>27</td><td>91.5</td><td>294.2</td><td>0</td><td>1000</td></tr><tr><td>Approximate licensing fees paid $\$ MM</td><td>23</td><td>25.5</td><td>108.6</td><td>0</td><td>500</td></tr><tr><td colspan="6">Full time &amp; equivalent employees in IP Group</td></tr><tr><td>IP attorneys</td><td>58</td><td>13.9</td><td>24.3</td><td>0</td><td>100</td></tr><tr><td>Patent Agents</td><td>44</td><td>6.5</td><td>25.6</td><td>0</td><td>150</td></tr><tr><td>Support staff</td><td>53</td><td>18.2</td><td>29.3</td><td>0</td><td>125</td></tr><tr><td colspan="6">Full time &amp; equivalent IP employees in business units</td></tr><tr><td>IP attorneys</td><td>31</td><td>1.2</td><td>4.7</td><td>0</td><td>25</td></tr><tr><td>Patent Agents</td><td>30</td><td>2.0</td><td>8.2</td><td>0</td><td>40</td></tr><tr><td>Support staff</td><td>33</td><td>2.6</td><td>5.5</td><td>0</td><td>25</td></tr><tr><td colspan="6">Productivity measures for the IP function:</td></tr><tr><td>US applications per disclosure</td><td>49</td><td>0.59</td><td>0.312</td><td>0.2</td><td>2.01</td></tr><tr><td>US applications per IP attorney</td><td>47</td><td>43.9</td><td>103.8</td><td>1</td><td>700</td></tr><tr><td>Net licensing revenue per IP attorney $\$ MM</td><td>23</td><td>0.83</td><td>7.7</td><td>-19</td><td>28.2</td></tr></table>
+
+
+L.1
+
+---
+
+Results From the IPO Survey on Strategic Management of Intellectual Property
+
+Quantitative Data
+
+<table><tr><td>2001 data for 59 publicly traded companies in the sample:</td><td>Average</td><td>Std. Dev.</td><td>Minimum</td><td>Maximum</td></tr><tr><td>Sales ($\$bn)</td><td>19.9</td><td>35.7</td><td>&lt;0.2</td><td>&gt;150</td></tr><tr><td>Market cap ($\$bn)</td><td>43.8</td><td>74.7</td><td>&lt;0.2</td><td>&gt;250</td></tr><tr><td>Employees (1000s)</td><td>50.4</td><td>60.4</td><td>&lt;0.5</td><td>&gt;150</td></tr></table>
+
+
+## Frequency distributions of selected variables, 2001
+
+Number of patents applied for:
+
+![Figure](figures/figure_005.png)
+
+Licensing revenues ($\MM):
+
+![Figure](figures/figure_007.png)
+
+Sales (\$bn)
+
+![Figure](figures/figure_009.png)
+
+Employees (1000s)
+
+![Figure](figures/figure_011.png)
+
+L.2
+

processed/00179_W1970856530/document.md

@@ -0,0 +1,498 @@
+Int Environ Agreements (2016) 16:579-596 DOI 10.1007/s10784-014-9271-4
+
+![Figure](figures/figure_001.png)
+
+ORIGINAL PAPER
+
+# Realizing access and benefit sharing from use of genetic resources between diverging international regimes: the scope for leadership
+
+Kristin Rosendal · Steinar Andresen
+
+Accepted: 3 December 2014/Published online: 9 December 2014 Springer Science+Business Media Dordrecht 2014
+
+Abstract This article examines how access and benefit sharing (ABS) in international transactions with genetic resources can be achieved and how Norway contributes to their realization. Regarding the first question, progress on the ground has been slow, but important principles have been agreed within the convention on biological diversity (CBD) and its Nagoya Protocol (NP). Although domestic legislation is adopted, key user countries remain reluctant. They argue that the ABS regime needs to be supplemented with sector approaches within forums such as the Food and Agriculture Organization. In principle, this may sound logical, but sector approaches may risk undermining the ABS regime of the CBD/NP. The principle of access is more user-oriented and benefit sharing is weaker in the relevant FAO negotiations. Against this background, the future practical significance of the ABS regime remains uncertain. Norway has played an important leadership role in ABS within the CBD/NP framework. This stems in part from ‘fortunate circumstances’, as Norway has relatively few stakes in this issue area, but also includes strong normative elements: Norway’s inclination to support weaker part, the South. The Norwegian position has also been solidified by good coordination and strong institutional capacity among the actors involved. However, there are indications of a growing split in the Norwegian position along sector lines. We do not yet have sufficient empirical evidence that this is the case — but if it is, achieving an effective ABS regime may be even more difficult.
+
+Keywords Genetic resources · Benefit sharing · Global governance · Institutional complexity · Leadership · Norway
+
+K. Rosendal ( ) · S. Andresen Fridtjof Nansen Institute (FNI), PO Box 326, 1326 Lysaker, Norway e-mail: kristin.rosendal@fni.no
+
+![Figure](figures/figure_009.png)
+
+---
+
+580
+
+K. Rosendal, S. Andresen
+
+## 1 Introduction
+
+A main challenge of international biodiversity negotiations is to achieve an effective regime for access and benefit sharing (ABS) in the face of North/South conflicts. Norway has supported the Southern states since the start of negotiations in the late 1980s and has been an active supporter for the threefold objective of the convention on biological diversity (CBD): the conservation and sustainable use of biodiversity and equitable sharing of benefits derived from utilization of genetic resources (Rosendal 2000 ) . This article examines whether Norway can continue this role as the current process of negotiations over access and benefits to genetic resources now take place in a situation of increased complexity. As a rich and sparsely populated country, with relatively low pressure on ecosystems and with little vested interests in biotechnology at the time, this was a low-cost option for Norway to enhance a positive bridge-builder image internationally (Rosendal 2005 ) . However, this is also in line with Norway's explicit emphasis on the rights and interests of the South more generally. As the CBD goals proceed through the more demanding and complex implementation phase and Norway increases their biotechnology ambitions, the leadership role may be challenged.
+
+The CBD principles of ABS have since been sought strengthened through the 2010 Nagoya Protocol (NP) of the CBD. The early phases of the negotiations started out with the CBD as a wildlife conservation treaty, but its scope was soon expanded to genetic resources, including the value of domesticated material. This met the demands from developing countries, where the bulk of terrestrial biodiversity is found. That move strengthened their negotiating clout and led to the inclusion of the ABS principles in international transactions involving genetic resources (Rosendal 2000) . As the inclusion of domesticated genetic material is where the CBD most clearly interacts with the UN Food and Agricultural Organization (FAO), this relationship is a central focus of the present study.
+
+Against this backdrop, our research question concerns how Norway may contribute to the realization the ABS objectives. In addition to considering domestic explanatory factors, we discuss how new and more complex international regime linkages and actor constellations may affect Norway's opportunities for continuing this leadership role.
+
+The CBD and its ABS regime have been ratified by 194 countries and the NP by 51 (as of 4 August 2014). Acknowledging countries' national sovereignty over genetic resources, the regime stipulates that access to genetic resources is to be based on prior informed consent (PIC) and mutually agreed terms (MAT). The regime enjoys considerable legitimacy among actors engaged in international transactions with genetic resources. Realizing the need for trust in such transactions with provider countries, many public and private bioprospectors have incorporated PIC and MAT in their statutes as regards accessing genetic resources (Laird 2000 ) . Still, legal compliance in user countries and financial results from benefit sharing remain imperfect and uncertain. The ABS principles have been a source of alienation between providers and users of genetic resources and the regime is poorly implemented in user countries. Implementation in provider countries was delayed during the negotiations of the NP. The ABS principles also go a long way towards explaining why the CBD has not been ratified by the USA (Rosendal 1995 ) . Although the NP helped to strengthen and concretize the ABS regime, conflicts remain over access standards, user measures to comply with ABS, traditional knowledge, functional and temporal scope, and the relationship to other international forums.
+
+In September 2013, Norway became the first industrialized country to ratify the NP. Norway also has an explicit mandate not to support legally binding ABS agreements under
+
+![Figure](figures/figure_008.png)
+
+---
+
+Realizing access and benefit sharing from use of genetic resources
+
+581
+
+the CGRFA/FAO, as this is seen as not supporting the NP. $^1$ Norway therefore seems to be pursuing a different policy from most other developed countries. Norway has been among the few user countries to strive for coordinated CBD/ABS implementation domestically and has pursued a leadership role throughout the international negotiations.
+
+In the following, we develop the analytical approach for the study of leadership in a situation of complexity. We then go on to consider hurdles and possibilities for the ABS regime in a complex governance situation, before applying our analytical approach to the case.
+
+## 2 Approach and methodology
+
+Let us consider the means through which leadership can be conducted and then the motivation to choose a leadership role. Leadership can be defined as `an asymmetrical relationship of influence, where one actor guides or directs the behaviour of others towards a certain goal over a certain period of time' (Underdal 1992 : 3). Scholars have noted four avenues through which leadership in international negotiations can be pursued (Young 1991 ; Underdal 1992 ). There is `structural leadership', based on unilateral power and out of the league of small states (Young 1991 ). Remaining channels, accessible also for small countries, are instrumental, intellectual and directional leadership, associated with setting a good example (Underdal 1992 ). `Intellectual leadership' may affect the systems of thought that shapes the perspectives of those participating in institutional bargaining (Young 1991 ). `Instrumental leadership' is characterized by the ability to come up with creative solutions and then move negotiations in the preferred direction. This type has similarities with `directional leadership', as a proposed solution is likely to gain greater international credibility if the proposer can show that it has already been successful at the domestic level.
+
+What, then, about the motivations for leadership? Putnam ( 1988 : 434) notes that, while domestic groups pursue their interests by pressuring governments to adopt advantageous international policies, there may also be powerful incentives for consistency in policies between the national and international levels. In areas of typically low-salience politics, maintaining a positive international profile can be valuable for smaller states with few other tools for demarcating itself on the international arena. Ideational and cognitive perspectives within regime theory point to how reiterated learning and development of common norms may affect the behaviour of negotiating parties, independently of structural power (Haas 1992 ; Young 1982 ) . Despite scant implementation, the ABS system constitutes the central international regime for the governance of genetic resources (Oberthür and Rosendal 2014 ) . Small states can be assumed to be even more likely to comply with such norms and principles, especially when they have helped to bring them into being.
+
+Among the motivating factors for assuming the bridge-builder role is a normative and material interest in carving out a positive international profile — coupled with the means or capacity to maintain this role. Capacity comes largely in the form of institutional ability to allocate staff and resources to pursuing certain policies. In addition to these domestic factors, such a role has been shown to be easier pursued in collaboration with like-minded
+
+1 Norway's mandate to the CGRFA-14:"Norway should recommend ratification of the Nagoya Protocol. It is not an option to support proposals for developing legally binding agreements on ABS under the CGRFA". The MoE added that the last entry is due to the perception that this would not support the Nagoya Protocol". Gaute Voigt-Hanssen, Senior Advisor, Ministry of the Environment, e-mail to author, 12 April 2013.
+
+![Figure](figures/figure_009.png)
+
+---
+
+582
+
+K. Rosendal, S. Andresen
+
+countries (Rosendal 2007 ) . While increased international complexity may enlarge the scope for package deals, it could also make for more unruly coalitions. A central question becomes whether there are changes in these domestic (normative and/or material interest, institutional capacity) and international (complexity and coalitions) factors that may affect the motivation and means for maintaining a leadership role.
+
+Against this backdrop, Norway's leadership position in ABS might be hampered or strengthened by changes in the following:
+
+- 1. Domestic normative persuasion as basis for leadership One explanation for assuming
+an intellectual leadership role could be a normative persuasion regarding the ABS
+principles (equitable sharing). Our examination of the equity dimension aims to shed
+light on Norway's implementation of ABS and domestic coordination of policies
+relating to the CBD and FAO. An examination of norms and values must acknowledge
+the potential for inconsistency between them. Norway's position might rest on a
+genuine belief that the equity inherent in the ABS is worth promoting. Adding to this
+explanatory perspective is how Norway has ratified the CBD with its ABS, which
+could imbue the pacta sunt servanta view (Franck 1990 ) .
+2. Domestic material interests as basis for leadership Norway's persistent interest in
+pursuing a leadership role could be accounted for by stable or evolving/diverging
+relevant material interests. Here, Norway could be expected to have an interest
+structure that differs from most other developed countries. We would assume that
+domestic interests have not been diverging much, historically or recently. The
+empirical task is to check for factors that might change the cost – benefit analysis for
+Norway. We also compare Norwegian domestic ABS legislation with that of other
+relevant OECD countries, providing a background for discussing why and how
+Norway deviates from others.
+3. Domestic institutional capacity for coordinated implementation The degree of
+national-level continuity and coordination between ministries could be explained
+through institutional factors, in terms of their capacity to coordinate positions between
+related international negotiation processes. In this category, we find leadership by an
+example: the ability to coordinate and agree on domestic legislation that can be used as
+a lever to sway international negotiations in the desired direction.
+4. International regime complexity and interest coalitions Increased regime complexity
+tends to produce inter-institutional competition and even open conflict and turf battles,
+as it opens for forum shopping (Gehring and Faude 2013 ) . Change or stability in
+leadership could hinge on how the Norwegian position links up to increased regime
+complexity and shifting coalitions. Norway aims at bridge-building through intellec-
+tual and instrumental leadership — and has a history of maintaining this position across
+a range of forums. Such tightly knit linkages across issue areas can enhance trust
+through the effects of iterative negotiation games and cooperation (Keohane 1984 ) . As
+a corollary, motivation might decrease in value if the original coalition partners change
+their positions. The main question here concerns the collaborative robustness of
+Norway's position and leadership role in a setting of greater complexity and altered
+coalitions.
+These four explanatory perspectives are additive rather than mutually exclusive. They encompass different aspects of the motivation and ability to maintain a robust policy position, both internationally and at home. If they all pull in the same direction, this could indicate persistency in Norwegian leadership on ABS. If some factors are becoming less important, then that could spell a change in Norwegian policy. In addition, the examination
+
+![Figure](figures/figure_006.png)
+
+---
+
+Realizing access and benefit sharing from use of genetic resources
+
+583
+
+may shed light on how Norway differs from most other user countries, as well as indicating how the ABS objectives may be achieved.
+
+### 2.1 Methodology
+
+Based on extensive earlier studies, we trace the Norwegian ABS position back to the early phases of the CBD negotiations and through the negotiations of the NP. This is compared to the Norwegian mandate in the FAO–CGRFA meetings. The next section traces how Norway has followed up (or not) the CBD obligations relating to ABS. This sheds light on eventual changes in Norway's approach to ABS policies. We base these examinations largely on academic studies of the role of Norway in the CBD and FAO. The investigation also includes interviews with key actors in Norwegian ministries and from the CBD negotiations. The authors' direct participation in a FAO commission meeting has helped to bring out nuances. We have also discussed with stakeholders from domestic NGOs and academics. From the ministry sector, actors from agriculture, environment, coastal and fisheries are represented.
+
+We asked key Norwegian actors in the negotiations to identify specific items where Norway had special interests and impact on the treaty text. Further, we enquired about specific barriers to ABS policy and legislation, and whether the actors have seen changes in the mandates relating to ABS — due to shifting norms, changes in interests, or changes emanating from complexity. Our interviewees have been very close to the negotiation processes for more than a decade, holding central positions. Still, these actors may admittedly inflate the actual Norwegian impact. Hence, their views have also been checked with a non-Norwegian key actor, who has also been closely linked to the ABS negotiations. A total of six key actors have been interviewed. In sum, this has allowed for an in-depth study of Norway's role and a comparison of the Norwegian arguments in the CBD/NP negotiations with those at the FAO Commission talks. Still, ideally, we should have had more non-Norwegians in our interview sample, but time and resources have not allowed for that.
+
+## 3 Negotiations and state of the ABS regime
+
+### 3.1 Early phase (1989-1992): CBD and ABS
+
+The bulk of terrestrial species diversity is found in the tropical South (Heywood 1995 : 749) , while developed countries (their private companies in particular) are largely in a position to gain the biotechnological revenues from utilization of genetic resources. This North/South controversy still colours the ABS conflict, although the negotiating coalitions have been changing along with the pattern of providers and users.
+
+Developed and developing countries came to the CBD negotiations with widely differing agendas and largely incompatible interests. Most developing countries wanted genetic material from both wild and cultivated/agricultural species included in the Convention, to ensure that they would receive a fair share of the proceeds from their use. They also wanted compensation and incentives to preserve their biological diversity, and thus avoid having to shoulder the greatest burdens. By contrast, most developed countries wanted a straightforward conservation treaty. Implicitly, they preferred the current arrangement to remain as it was, allowing them to enjoy free access to genetic resources in the South. Speaking for the interests of the biotechnology sector, they did not want to link
+
+![Figure](figures/figure_010.png)
+
+---
+
+584
+
+K. Rosendal, S. Andresen
+
+conservation with economic obligations in developed countries, or to link the use of genetic resources to benefit sharing (Schei 1997; Rosendal 2000, 2011).
+
+How to parcel out the economic responsibility for conserving biological diversity has remained among the toughest topics in negotiations. From the outset, the interaction between the CBD and the Agreement on Trade-related Aspects of Intellectual Property Rights (TRIPS) of the World Trade Organization (WTO/TRIPS) was contested due to the relationship between ABS and intellectual property rights (IPR). There was also political and organizational strife concerning the demarcation between the CBD and FAO, as not only wild but also domesticated genetic resources (seeds) became subject to the CBD (Rosendal 1991 ) . This turf battle coincided with the FAO seed wars in the late 1980s and hinged on property rights to seeds, as the FAO Undertaking on PGRFA was reinterpreted in 1989 to accommodate IPR, waiving the `common heritage of mankind' principle for systematically bred seeds (Raustiala and Victor 2004 ; Pistorius and van Wijk 1999 ; Rosendal 1991 ) . Common heritage and patents/plant breeders' rights were at the centre of the FAO seed dispute; the CBD's response was to balance increased Northern seed patents with (re)acknowledging natural resources as subject to national sovereignty. For seeds in international gene banks collected prior to the CBD entering into force, the strife resulted in the Nairobi Final Act (1992) by the CBD Parties, which referred to the role of FAO in dealing with this material (Andersen 2008 ; Rosendal 2000 ) . In 2001, the FAO Parties concluded the International Treaty on Plant Genetic Resources for Food and Agriculture (ITPGRFA) on this basis.
+
+### 3.2 Strengthening ABS (2002-2010, to present): Nagoya Protocol
+
+At its sixth meeting (2002), the COP6 adopted the Bonn Guidelines, 2 and considered the role of IPRs in ABS and the relationship with the WTO/TRIPS (ENB 2014 ) . The 2002 World Summit on Sustainable Development in Johannesburg called for a more detailed ABS agreement, which, like the CBD and unlike the Bonn Guidelines, should be legally binding. COP7 decided to start negotiations.
+
+These negotiations took place in a setting of rising international regime complexity, as concrete efforts to link access to benefit sharing made the practical interaction with other international regimes apparent. The G77 coalition, which had remained stable throughout the ABS negotiations (Najam 2005 ) , became fragmented; it included the African group, the like-minded Asia–Pacific countries, the like-minded megadiverse countries and group of Latin American and Caribbean countries (GRULAC). More technologically advanced countries like India and Latin American countries stressed the need for compliance mechanisms and inclusion of derivatives $^3$ from genetic resources. The African group was concerned with genetic material in ex situ (gene bank) collections accessed prior to and since the CBD. Still, the North/South schism persisted as the dominating conflict line (Wallbott et al. 2014 ) . While the G77 split may have reduced the overall strength of the providers, the loosening up of the deep conflict lines may nevertheless have made it possible to reach agreement in Nagoya. This agreement is characterized as impressive, given the high level of technicality characterizing the legal and biological challenges that
+
+2 Bonn Guidelines on Access to Genetic Resources and Fair and Equitable Sharing of the Benefits Arising Out of Their Utilization, adopted by Decision VI/24, Doc. CBD UNEP/CBD/COP/6/20 (2002) [Bonn Guidelines].
+
+3 The debate concerns the distance (derivation) from the functional units of heredity (genetic resources) at which the obligation to share benefits remains in force (Tvedt 2014).
+
+![Figure](figures/figure_009.png)
+
+---
+
+Realizing access and benefit sharing from use of genetic resources
+
+585
+
+negotiators had to deal with. $^4$ Another element that may have furthered agreement on the NP was the perceived failure of the Climate Change Summit in Copenhagen: advocates of multilateral UN collaboration sorely needed a success story.
+
+Five core elements remain contested in the NP (see Wallbott et al. 2014 and Oberthür and Rosendal 2014 ) . Firstly, (1) elaboration of international access standards (as demanded by user countries) and (2) ` user measures ' (as demanded by provider countries). The Protocol became highly detailed concerning access standards and obliges provider countries to grant legal certainty and transparency in their ABS legislation. Repeating original obligations from the CBD, user countries are required to take appropriate measures to ensure that the genetic resources are accessed in accordance with PIC and MAT in provider-country legislation. However, the long-standing demand of providers for a mandatory disclosure requirement regarding genetic resources in patent application failed again in the negotiations. This contested area of ABS governance includes balancing access to genetic resources with intellectual property rights (IPR) (Swanson 1995 ; Esquinas-Alcazar 2005 ; Rosendal 2011 ; Pavoni 2013 ) . ABS seeks to balance IPR as both systems aim to set economic conditions on legal use of the material. Concerning (3) traditional knowledge , the Protocol establishes similar requirements regarding PIC and MAT, more detailed than what was formulated in the CBD. As regards the (4) scope of resources covered, providers succeeded in that the Protocol covers genetic resources along with their derivatives, which are currently more often the source of the benefits. However, derivatives are still subject to debate, as they are mentioned in the NP only in the definitions (Koester 2012 ) . The exact temporal and functional scopes of the NP remain disputed due to ambiguous language (Tvedt 2014 ) . Finally, the (5) relationship with other international institutions forms the fifth core contested element of the Protocol (Art 4.3, 8.b). Crucial here is that sector-specific ABS regimes can be created (as demanded by users), if they are supportive of and do not run counter to the objectives of the NP and the CBD (as demanded by providers). It is with regard to this last issue that the complexity and a potential turf war with the FAO become apparent.
+
+### 3.3 The sector approach to ABS: general challenges and interaction with the FAO
+
+The norms and principles of the ABS regime have come to penetrate an increasing number of institutions and arenas, providing the regime with normative (albeit not very strong practical) momentum (Oberthür and Rosendal 2014 ) . The success of ABS depends on whether user countries will create incentives for benefit sharing (Tvedt and Young 2007 ) . With regard to user measures, there is an untapped potential for synergy with the WTO (Kamau and Winter 2009 ; Pavoni 2013 ) . The lack of political willingness to make the patent system a useful tool for workable benefit sharing reduces its potential to contribute to ABS implementation. This is seen in the reluctance of countries in the World Intellectual Property Organization (WIPO) and WTO to take on board an obligation to provide information about PIC and MAT in patent applications (disclosure). The imbalance between ABS and IPR is strengthened by the restrictions on access that emanate from IPRs. $^5$
+
+4 As pointed out by head of Norwegian delegation, Birthe Ivars, Ministry of the Environment, 27 January 2014.
+
+5 IPR tends to impede research and development on the specific needs of poor populations in developing countries, especially in agriculture and health (Oberthür et al. 2011 ) .
+
+![Figure](figures/figure_008.png)
+
+---
+
+586
+
+K. Rosendal, S. Andresen
+
+Several arenas, including the FAO and the World Health Organization (WHO) 6 have opened discussions on sectoral approaches. Such approaches could represent a fine-tuning of governance within specific areas to enhance access. The arguments for a sectoral approach to plant genetic resources in food and agriculture stem from the incremental improvement and multiple sources characterizing seeds and plant breeding, indicating high interdependence among providers and users in plant breeding (Morgera et al. 2012 ) . Incremental improvement, multiple sources and interdependence suggest that there is not one end-product tied to the accession, that it is hard to identify one source country and that a provider can become a user and vice versa (Schloen et al. 2011 ; Chiarolla 2011 ) . These characteristics provide the rationale for decoupling benefit sharing from both the plant genetic resource and the provider in the legal instruments deliberated within the FAO. However, whereas these characteristics can to some extent be documented for plant breeding, the other sectors currently contemplated by the FAO — farm animals, forest trees, aquatics, microorganisms and invertebrates — follow different patterns (Medaglia et al. 2013 ) . These differ from plant seeds in that there is less or no incremental improvement involved and generally far less dependency on multiple sources or interdependence among users/providers.
+
+There are also difficulties of principle with sectoral approaches. Tvedt ( 2014 ) argues that opening for sector-wise ABS rules for pathogens, academic use and other groupings of domesticated genetic resources might entail fragmenting what is covered by the general ABS rules of the CBD/NP. ABS establishes a mechanism to counterbalance private exclusive rights (patents) to innovation and discovery based on genetic material. When certain uses or resources are removed from the scope of ABS under the Protocol, there is no guarantee that the balance acquired in the CBD is maintained. Hence, the sectoral approach could also be interpreted as forum shopping by powerful actors seeking to circumvent benefit sharing and as an illustration of persistent turf wars between the CBD and FAO. Competing sector approaches could challenge implementation of the CBD/NP regime for ABS (Tvedt 2014 ; Medaglia et al. 2013 ) . More tailored and streamlined solutions could be beneficial for some actors and such efforts must be considered on a caseby-case basis. Nevertheless, poor provider countries would generally be financially burdened by entering into prolonged negotiations in yet more forums.
+
+### 3.4 The sectoral approach in FAO and the CGRFA: in practice
+
+Within months of concluding the NP, the parties to the FAO Commission on Genetic Resources for Food and Agriculture (CGRFA) started preparing for negotiations on sectoral approaches for access to farm animal, forest tree, aquatic, microorganisms and invertebrate genetic resources, as well as for plants outside the multilateral system of the ITPGRFA. 7 This harks back to the early turf struggles during the establishment of the CBD. Within the CBD, ABS is principally regarded as a prerequisite for conservation and sustainable use of biodiversity through increased equity , whereas access to seeds has top priority with the FAO.
+
+6 In May 2011, a Pandemic Influenza Preparedness (PIP) Framework (Agreement) for the Sharing of Influenza Viruses and Access to Vaccines and Other Benefits was adopted by the WHO Assembly.
+
+7 On ratifying the FAO Plant Treaty, countries agree to make their genetic diversity and related information about the crops stored in their gene banks available to all through the Multilateral System (MLS). http:// www.planttreaty.org/content/what-multilateral-system Accessed 21 November 2014.
+
+![Figure](figures/figure_008.png)
+
+---
+
+Realizing access and benefit sharing from use of genetic resources
+
+587
+
+Within the CBD multilateral system of ABS, benefit sharing is to be negotiated bilaterally between a user and a provider. Monetary benefits may include fees per sample, milestone payments, royalties on net sales and/or licensing agreements; non-monetary benefit sharing might cover training, capacity building, research exchanges, supply of equipment and technology transfer. Within the FAO, benefit sharing is decoupled from the provider and is defined in non-monetary terms — as access to improved breeding material from the FAO multilateral system (MLS). This is not in line with the ABS system under the CBD/NP, and it does not balance the IPR of multinational seed corporations. Users favour this definition and this constitutes the FAO position of most OECD countries — with the notable exception of Norway. 8 The multinational seeds sector has economic interests in flexible, open-access regimes and argues that ABS may restrict this. However, the increased private appropriation of genetic resources through intellectual property rights represents a much more serious barrier to free and flexible access — and the FAO debate, unlike that in the CBD, does not acknowledge this.
+
+In discussions in the FAO – CGRFA 14 (April 2014), the sector approach was hotly debated and patent issues were excluded from the agenda. At the first session of the Ad Hoc Technical Working Group on ABS for GRFA (CGRFA-14/13/6), the document on the need for and modalities of ABS arrangements for GRFA (CGRFA-14/13/7) were debated. $^9$ The European Regional Group (ERG), the USA and Japan all emphasized the need to ensure that ABS measures accommodate FAO and do not impede food security. They stressed the users' arguments in favour of a sectoral approach to genetic resources (see Morgera et al. 2013). This evoked the core conflict of the NP negotiations and was rejected by the African group and the Southeast Asia Regional Initiatives for Community Empowerment (SEARICE), an NGO which has been central to and followed the ABS issue closely since the start of the CBD negotiations (Rosendal 2000). $^10$ However, other G77 countries remained almost silent on the issue. The former two groups emphasized sovereign rights over genetic resources and urged all FAO parties `not to splinter the ABS into a landscape that small actors cannot navigate in, as this would also not be helpful for food security'. $^11$ They also made calls for a study on the impact of IPRs on ABS for GRFA, but that debate was not kept on the agenda nor reported in the ENB. $^12$ The USA demanded that mention of the relationship with the WIPO be struck from the debate. The report of the CGRFA debate ended up without a single reference to IPR issues.
+
+Removing all domesticated genetic material (GRFA) from the NP would seem to run counter to the interests of the developing world. For developing provider countries to open negotiations here could mean going back to square one in negotiating ABS regulations. Benefit sharing is mandatory under the CBD and practically voluntary within the system in place with the FAO, the ITPGRFA. Also the volume of access to samples distributed under the ITPGRFA is much greater than what is done legally under the CBD. It would certainly be in the interest of users to move the bulk of genetic resources to a forum without mandatory benefit sharing and where the relationship between ABS and patents has a weaker position. By contrast, providers are likely to remain interested in mandatory and
+
+8 Grethe Evjen, LMD, 10 September 2013, Seminar at FNI, Norway.
+
+9 Both authors were present as observers in this plenary debate; here we present the elements of the ensuing debate that relates to ABS. See also ENB 2013 , vol. 9, no. 597.
+
+10 SEARICE promotes and implements community-based conservation, development and sustainable use of plant genetic resources. http://searice.org.ph/about-searice/what-we-do/ , accessed 20 February 2014.
+
+11 Authors' observations from the CGRFA meeting.
+
+12 ENB, vol. 9/no. 597.
+
+![Figure](figures/figure_010.png)
+
+---
+
+588
+
+K. Rosendal, S. Andresen
+
+monetary benefit sharing, irrespective of which forum deals with the genetic resources. The FAO efforts to expand could mean that providers must devote resources and efforts to ensure that new ABS mechanisms are supportive of and do not run counter to the CBD/NP, as required by Article 4. Neither the FAO multilateral system nor the CBD–ABS system has yet provided much in terms of financial allocations, but the systems rest on very different principles.
+
+## 4 Role of Norway in ABS
+
+### 4.1 Norwegian leadership in the international ABS negotiations
+
+Throughout the CBD negotiations, Norway has assumed the role of bridge-builder. Initially along with the Nordic countries, Denmark and Sweden in particular, Norway insisted that all countries have a common responsibility for sharing the costs of biodiversity conservation. The Nordics argued for a system that would allow a returning flow of benefits in compensation for the use of genetic resources from the South (Koester 1997 ; Rosendal 2000 ) . Further central Nordic goals were the principles that the CBD should include all biological diversity (Schei 1997 ; Svensson 1993 ) . These goals were successfully achieved in liaison with the G77. One of the most controversial articles, which caused greatest discomfort to the USA, is Article 16.5 (Rosendal 1995 ) . It reads: `The Contracting Parties, recognizing that patents and other intellectual property rights may have an influence on the implementation of this Convention, shall cooperate in this regard subject to national legislation and international law in order to ensure that such rights are supportive of and do not run counter to its objectives'. According to central negotiators, this text was crafted by the Norwegians. $^13$ That the working group drafting this text was led by the chief Brazilian negotiator also helped to promote the essence of the Article, as the whole of G77 was aligned with Norway on these issues and Brazil was a strong advocate. $^14$
+
+During the NP negotiations, Norway played a central role in the architecture of the core compliance provisions (Articles 15, 16, 17 and 18)—especially the principle that users must comply with ABS regulations in provider countries. Achieving strong user-country measures was of utmost importance to Norway in these negotiations. $^15$
+
+For Norway, these objectives go back to the period preceding the CBD negotiations (Rosendal 2005) . The objectives were already set out in White Paper no. 46 (Ministry of the Environment 1989) ; and the accompanying Blue Book (Ministry of Foreign Affairs 1988) and were put into effect during the CBD negotiation phase. White Paper 46, 1989 (p. 113) has the following international biodiversity goals, `To support international activities that give developing countries a more equitable share of the benefits accruing from the use of genetic resources'. The same goals are set out in the Blue Book of the Min of foreign affairs (1988, p 23) (further elaborated in Rosendal 2005, 2007) . The principles were also central in the report of the World Commission on Environment and Development (1987:
+
+13 Interview, Veit Koester, Danish delegation leader to the CBD negotiations, 28 April 2014. Koester presided over the negotiation of the portion of the Convention on Biological Diversity resulting inter alia in the Access and Benefit Sharing (ABS) system.
+
+14 Personal communication (7 April 2014) from Peter Johan Schei, who headed the Norwegian delegation to the CBD negotiations and COP meetings from 1989 to 2004. Personal communication, Veit Koester, Danish delegation leader to the CBD negotiations, 28 April 2014.
+
+15 Personal communication from Birthe Ivars, Ministry of the Environment and member of the Norwegian delegation to the CBD and the Nagoya Protocol, and chair for the Bonn Guidelines process. 8 April 2014.
+
+![Figure](figures/figure_011.png)
+
+---
+
+Realizing access and benefit sharing from use of genetic resources
+
+589
+
+the `Brundtland Report'), which urged: `Industrialised nations seeking to reap some of the economic benefits of genetic resources should support the efforts of the Third World nations to conserve species' and `Developing countries must be ensured an equitable share of the economic profit from the use of genes for commercial purposes' (WCED 1987 ) .
+
+Also in the more recent phases of ABS negotiations Norway has had an acknowledged role as a significant player, with a level of engagement mirrored by its own national ABS legislation (Burton 2012 ) . Compared with many other countries, Norway has high capacity to coordinate the positions of negotiation delegations across international arenas. Moreover, a typical characteristic is continuity, as the same civil servants — mostly from the Ministries of Foreign Affairs, Agriculture, Fisheries and the Environment, with the latter in the lead — have attended these UN forums for several years (Rosendal 2007 ) . Here, Peter Johan Schei and Birthe Ivars have played central roles in leading the ABS negotiations through crucial phases. In the negotiations leading up to Nagoya, the delegation was strengthened with a member from the Ministry of Legal Affairs in order to include expertise on negotiations underway in the WTO and WIPO. 16 These elements have provided Norway with expertise, commitment and continuity in negotiations. 17 Because of this expertise and commitment, Norway was chosen by the Japanese hosts along with Brazil, the EU and Namibia to form a group to assist in the final rounds of the Nagoya negotiations. 18
+
+This bolsters the explanation that Norway has a high degree of institutional capacity and continuity for coordination of domestic policies regarding ABS, and also that the normative basis for ABS remains part of the Norwegian politics. Both pacta sunt servanta and iterative negotiation games with stable coalition partners appear to strengthen the likelihood that Norway will be able to continue its ABS position. Moreover, due to the broad representation from various ministries, Norway has a long history of supporting the CBD principles across international forums. The growing complexity in ABS governance could, however, affect the normative basis, as this could be seen as a case of competing norms: FAO parties cite the normative principle linked to food security and access to GRFA and worry that the ABS could obstruct access to genetic material. As argued in Sect. 3.4 , this normative controversy could also be interpreted as part of the interests of structurally powerful parties and organizational turf wars; as the FAO, more so than the CBD, is dominated by users and an increasingly strong multinational seeds industry. This illustrates the difficulty of distinguishing between normative and interest-based motivations. In order to examine this more closely, the next sections take a closer look at the material basis for Norway's ABS position. According to our analytical presumptions, how ABS objectives are reflected in domestic legislation and how this corresponds with domestic material interests may affect the leadership role.
+
+### 4.2 Norway and domestic ABS implementation: directional leadership
+
+We have shown how Norway has been active in advocating four core principles of the ABS regime: equitable sharing, the inclusion of domesticated genetic resources (seeds), balancing IPR and ABS systems (Article 16.5) and strengthening user measures in order to comply with the ABS regime. Let us now see how these items are reflected in domestic legislation.
+
+16 Interview with Birthe Ivars, Norwegian Ministry of the Environment, 27 January 2014.
+
+17 Substantiated in interview with Birthe Ivars, Norwegian Ministry of the Environment, 27 January 2014.
+
+18 Interview with Birthe Ivars, Norwegian Ministry of the Environment, 27 January 2014.
+
+![Figure](figures/figure_010.png)
+
+---
+
+590
+
+K. Rosendal, S. Andresen
+
+Sections 57–60 of Norway's Nature Diversity Act (2009) directly address user-measure obligations in the ABS. The import of genetic material into Norway from a provider state that requires prior informed consent may take place only in accordance with such consent. When such material is utilized for research or commercial purposes in Norway, such use shall be accompanied by information about the country of origin according to the amended Patent Act, section 8b, of 2003. Moreover, any person receiving genetic material from a public collection shall refrain, in Norway or abroad, from claiming intellectual property rights to the material, unless the material has been modified in a way that results in a substantial change. Contrary behaviour is subject to sanctions through legal action (see also Medaglia et al. 2013 ; Tvedt 2010 ) . By tying sanctions to non-compliance , Norway goes further than required by the ABS obligations of the NP.
+
+Among user countries, the most significant ABS measures are the legal amendments in patent legislation regarding disclosure of origin of genetic resources in patent applications . Norway and eight other European countries, including Sweden and Denmark, have provided such far-reaching legislation. Other significant legal measures relate to legal language demanding respect for PIC and MAT , found in legislation in Norway and a few other OECD countries. The measures vary considerably, however: in Norway, the legislation on PIC and MAT is aimed at Norwegian activities abroad, whereas Australian PIC and MAT legislation is directed towards external bioprospectors within their own country (Prip et al. 2014 ) . A third significant legal activity is the introduction of penalties/sanctions for noncompliance , which has been enacted in Norway, Denmark and Switzerland.
+
+Compared with most other user/OECD countries, Norway has held a leading position in establishing compatible domestic regulations with a view to implementing ABS. Norway was the first among the user countries thus far to ratify the NP. Among developed countries, Australia, Norway and Switzerland are the earliest implementers of the ABS regime. They have contributed to the NP through domestic legislation (directional leadership) which allows the NP to build on key elements that have already been tested and found to work (Burton 2012 ) .
+
+This part of the picture thus seems to confirm that Norway has invested institutional capacity in achieving a coordinated policy and a normatively consolidated approach to the ABS principles. During the legislative process, the translation of these principles into domestic legislation was not subject to diverging sub-national interests. 19 The domestic enactment of compatible legislation may have bolstered Norwegian instrumental leadership during negotiations, indicating leadership by example. We now turn to domestic ABS legislation in Norway and its relation to national material interests.
+
+### 4.3 Norwegian international efforts in light of domestic material interests
+
+Do Norwegian biotechnological and live-sector interests deviate significantly from those of the environmental authorities over ABS? Since the entry into force of the CBD in 1993, seven regions 20 and 57 countries (40 developing countries and 17 developed countries) have formulated domestic legal ABS measures. 21 The emerging economies of Brazil, China, India and South Africa have all enacted fully fledged ABS legislation. Compared
+
+19 Personal communication, Ole Kristian Fauchald (FNI); central in developing the legal framework on Norway's Nature Diversity Act (Naturmanafoldloven).
+
+20 The African Regional Intellectual Property Organization, African Union, Andean Pact, Central American countries, Commission des Forêts d'Afrique Centrale, European Union, and the Nordic region.
+
+21 http://www.cbd.int/abs/measures/groups.shtml Accessed 30.06.2014.
+
+![Figure](figures/figure_011.png)
+
+---
+
+Realizing access and benefit sharing from use of genetic resources
+
+591
+
+with those countries, Norway was a latecomer here with the amended Patent Act of 2003 and Nature Diversity Act of 2009. Compared with other users, which is more relevant, Norway is a pioneer in enacting user measures.
+
+This type of legislation would seem to have come at relatively low domestic costs for Norway. Through most of the negotiations, its biotechnology sector has been comparatively small (Ernst and Young 2001 ) . The USA, the EU and Japan together accounted for about 80 % of all applications for biotechnology-related patents under the system administered by the WIPO (OECD 2009 ) . In Norway, there have been fairly few domestic economic interests linked to intellectual property rights to genetic resources.
+
+However, Norway may expect greater attention from multinational corporations within the pharmaceutical sector and in the rapidly growing aquaculture sector. There is considerable interest in prospecting for marine genetic material, which has displayed promising medicinal traits. Still a relatively small user country, Norway might also have a particular interest in ABS legislation as a provider, due to its high levels of interesting marine genetic material. Also Australia deviates from most OECD countries by being not only a typical user country: it also ranks among the 17 megadiverse countries—a typical provider (Burton 2012 ) . This goes a long way towards explaining Australia's elaborate ABS legislation as a provider country. By contrast, Switzerland is hardly megadiverse and is home to a very strong biotechnology sector, but has nevertheless gone far in developing legal user measures such as disclosure and has also pursued the role of bridge-builder in the negotiations (Hufty et al. 2014 ) . Denmark and Switzerland still rank higher than Norway in patents granted, although also in Norway the biotechnology sector has increased during the last decade (Ernst and Young, 2005 , 2008, 2012; OECD 2011 ) . $^22$
+
+The study of domestic interests also warrants some scrutiny of evolving trends in agriculture, relevant for the interests Norway may pursue vis-à-vis the FAO. All countries, including Norway, depend on exchange of seeds (Kloppenburg 2004 ) . Still, the specific climatic conditions for Norwegian agriculture may set it apart from most other developed countries. Does this make Norway more or less dependent on the large seed corporations? Do multinational corporations have a less dominant role in Norway, accompanied by much lower demand for IPRs in seeds? The argument here is that multinational corporations have strong economic interests in pursuing broader patent protection; at the same time, they can evade the governmental control and regulations intended to enforce ABS behaviour (Louwaars et al. 2009 ) .
+
+Although the latter issue remains uncertain, the upshot is in line with the expectation that Norway does not seem to have strong domestic interests that deviate from the ABS principles. Norway differs from most OECD countries in having a relatively small biotechnology sector, being well endowed with genetic resources and with climatic conditions that place Norwegian agriculture on the outskirts of the radar of multinational seeds corporations. This would seem to furnish Norway with a somewhat different interest-base compared with most OECD countries. There might be altered policy deliberations ahead as Norwegian biotechnology picks up, although the Swiss example indicates that ABS compliance is possible also in countries with a strong biotechnology sector. Norway and Switzerland both seem inclined to advocate equitable sharing even though their material interests might dictate otherwise. This could be because both countries think that it does not necessarily mean great losses to support ABS and that the benign international profile and coming across as a legitimate bioprospector would make-up for possible economic losses in their biotechnology sector.
+
+22 http://www.oecd.org/sti/inno/keybiotechnologyindicators.htm Accessed 14 October 2013.
+
+![Figure](figures/figure_008.png)
+
+---
+
+592
+
+K. Rosendal, S. Andresen
+
+4.4 Increased complexity and altered coalitions: affecting Norwegian leadership role?
+
+As noted, Norway has long had a leadership role in this issue area. Its independent role from the EU and the JUSCANZ 23 group provided Norway with wider scope in the negotiations to develop its own position. The increased complexity did not appear to have negative consequences for exerting leadership within the NP negotiations — indeed, this may have broadened the scope for manoeuvre and leadership. 24 This may be different within the framework of the FAO, where Norway is a member of the European group, indicating less scope for independent leadership. The spilt in the FAO – CGRFA meeting, with only the African group arguing against sectoral approaches, may also affect Norway's role and position. It is important to note that the Norwegian delegation participated in CGRFA with a mandate not to support legally binding sectoral ABS agreements under the CGRFA. 25 Nevertheless, Norway did not express vocal support to Africa at the 2013 meeting. How can that be explained?
+
+The Norwegian position cannot provide an explanation. Norwegian administrators are not comfortable with the FAO definition of ABS in terms of non-monetary benefits. The Ministry of Agriculture and Food (MAF) and the Ministry of the Environment (MoE) concur that Norway does not follow the OECD interpretation and position in this definition. 26 Across ministries, there is recognition of the need to be aware of the different power and interest structures of the various international forums, as well as the need for nationallevel coordination when negotiating interacting regimes. As regards the FAO multilateral system, the MAF is aware that it is problematic that the multinational seed corporations enjoy full access without providing benefit sharing. These corporations generate large sums of money that are not returned to farmers. 27 Norway is also the third largest contributor to the ITPGRFA fund, 28 acknowledging that the seed industry is not likely to provide the necessary funding. Moreover, as fiscal support is unlikely to be forthcoming through the ITPGRFA, that concern is meant to be implemented through the CBD-ABS. 29
+
+However, the composition of the Norwegian delegation might explain why they did not stand up in defence of the African group/Namibia/SEARICE's proposal in the plenum debate. Norway's delegation did not include representation from the MoE, even though that ministry had had a noteworthy role in forging the coordinated mandate. The MoE representative fell out of the delegation at the last minute before the CGRFA meeting. If MoE had been present, might Norway have made an intervention in favour of the African view? Perhaps, the ministries represented did not have enough stakes in the mandate to make a public statement in its defence. To the sector-based ministries, the FAO and its CGRFA are more important venues than the CBD and its ABS regime, so these sector-
+
+23 JUSCANZ is for all practical purposes a group for coordination and exchange of information, not a negotiating group.
+
+24 Interview with Birthe Ivars, Norwegian Ministry of the Environment, 27 January 2014.
+
+$^{25}$ Gaute Voigt-Hanssen, Senior Advisor, Ministry of the Environment, e-mail 12 April 2013.
+
+26 Interview, Birthe Ivars, Ministry of the Environment, 27 January 2014. Grethe Evjen, Ministry of Agriculture and Food, FNI seminar on farmers' rights, Lysaker, 10 September 2013.
+
+27 Elisabeth Koren, Ministry of Agriculture and Food, FNI seminar on farmers' rights, Lysaker, 10 September 2013.
+
+28 IT/ACFS-7/12/3. “Resource Mobilisation: Implementation of the Strategic Plan for the Implementation of the Benefit-sharing Fund”. http://www.planttreaty.org/content/seventh-ad-hoc-advisory-committeefunding-strategy .
+
+29 Grethe Evjen, Ministry of Agriculture and Food, FNI seminar on farmers' rights, Lysaker, 10 September 2013.
+
+![Figure](figures/figure_013.png)
+
+---
+
+Realizing access and benefit sharing from use of genetic resources
+
+593
+
+based ministries might cater to a different set of interests than the MoE. On the other hand, the fact that the MoE representative was not replaced may also signify that the FAO – CGRFA is not considered to be very important for the MoE. In that case, this may point towards a division of labour between the sector ministries (focusing on FAO) and the more general MoE (focusing on CBD/NP). If this is the case, it points towards a less coordinated position in the future and maybe therefore also less scope for leadership when the ABS issue is discussed in the FAO. However, there is as yet not enough empirical material to substantiate such a conclusion.
+
+## 5 Conclusions
+
+With this article we have sought to shed light on how Norway, through leadership, has attempted to contribute to the realization of the ABS objectives. Regarding the implementation of ABS in general, progress on the ground has been rather slow, but important principles have been agreed within the framework of the CBD/NP. Although the necessary legislation is gradually being adopted in an increasing number of countries, key user countries have been reluctant to comply with the ABS regime. Most OECD countries argue that the general framework of the ABS within the CBD/NP needs to be supplemented with sector approaches within forums like the FAO and argue that this is necessary for finetuning specific application in different issue areas. In principle, this may sound logical, but we believe that some sector approaches may risk undermining the ABS regime as it is fleshed out in the CBD/NP. In our case in point, the FAO and the CGRFA, we find that the principles related to ABS are framed in a more user-oriented manner. Against this situation of increased complexity, the future practical significance of the ABS regime remains uncertain.
+
+As to the role of Norway in realizing the ABS regime, we have shown how Norway has played an important role as instrumental and directional leader. Both our literature reviews and interviews with a range of key actors indicate that Norway has long been central in getting acceptance for the gradual strengthening of the ABS regime within the CBD/NP framework. The basis for this leadership role stems in part from `fortunate circumstances', as Norway has relatively few stakes in this issue area compared with many other OECD countries. However, we have also argued that there are fairly strong normative elements in Norway's position, with an inclination to support the weaker part, the South, reflecting a long-standing tradition in many important global environmental processes. The Norwegian position has also been solidified by good coordination and strong institutional capacity among the actors involved. However, judging from the FAO meeting on the issue, there are indications of a growing split in the Norwegian position along sector lines. Managing institutional complexity requires skill and capacity and hence tends to increase the burden of domestic cross-sectoral co-ordination. In this case, increased complexity may have led to weaken the G77 coalition, with which Norway has traditionally supported ABS. Both of these factors may eventually weaken Norwegian leadership in the ABS issue area. We do not yet have sufficient empirical evidence that this is the case — but if it is, Norway's leadership role may become less pronounced and achieving an effective ABS regime may be even more difficult.
+
+Acknowledgments The authors would like to thank Morten Walløe Tvedt and two anonymous reviewers for their helpful comments on previous versions of this paper. Funding for this research was provided by the
+
+![Figure](figures/figure_007.png)
+
+---
+
+594
+
+K. Rosendal, S. Andresen
+
+Norwegian Research Council through the project 220630 "Biotechnology in agriculture and aquaculture— effects of intellectual property rights in the food production chain".
+
+## References
+
+Andersen, R. (2008). Governing agrobiodiversity: Plant genetics and developing countries. Aldershot: Ashgate.
+
+Burton, G. (2012). The JUSCANZ Group in Nagoya negotiations; the unlikely lot. In E. Morgera, M. Buck, & E. Tsioumani (Eds.), The 2010 Nagoya protocol on access and benefit sharing in perspective: Implications for international law and implementation challenges (pp. 295–330). Leiden: Martinus Nijhoff.
+
+Chiarolla, C. (2011). Intellectual property, agriculture and global food security: The privatization of crop diversity. Cheltenham: Edward Elgar.
+
+ENB. (2013). 14th meeting of the CGRFA. Highlights, Rome, Volume 09, Number 597, 16 April, 2013.
+
+ENB. (2014). Third meeting of the intergovernmental committee for the Nagoya protocol on access and benefit sharing. volume 09 Number 612, Monday, 24 February 2014.
+
+Ernst and Young (2001). Focus on fundamentals: The Biotechnology Report. 15th Annual Review.
+
+Ernst and Young (2005; 2008; 2012). Beyond Borders: The Global Biotechnology Report. Annual Review.
+
+Esquinas-Alcazar, J. (2005). Protecting crop genetic diversity for food security: Political, ethical, and technological challenges. Nature, 6(12), 946-953.
+
+Franck, T. M. (1990). The power of legitimacy among nations. New York: Oxford University Press.
+
+Gehring, T., & Faude, B. (2013). The dynamics of regime complexes. Global Governance, 19(1), 119-130.
+
+Haas, P. M. (ed.) (1992). Knowledge, power, and international policy coordination, Special issue of International Organization, 46(1).
+
+Heywood, V. H. (Ed.). (1995). Global biodiversity assessment, UNEP. Cambridge: Cambridge University Press.
+
+Hufty, M., Schulz, T., & Tschopp, M. (2014). The role of Switzerland in the Nagoya protocol negotiations. In S. Oberthür & G. K. Rosendal (Eds.), Global governance of genetic resources: Access to and benefit sharing after the Nagoya protocol (pp. 96–113). London: Routledge.
+
+Kamau, E. C., & Winter, G. (Eds.). (2009). Genetic resources, traditional knowledge and the law. London: Earthscan.
+
+Keohane, R. O. (1984). After hegemony: Cooperation and discord in the world political economy. Princeton: Princeton University Press.
+
+Kloppenburg, J. R. (2004). First the seed (2nd ed.). Cambridge: Cambridge University Press.
+
+Koester, V. (1997). The Biodiversity Convention negotiation process and some comments on the outcome. Environmental Policy and Law, 27(3), 175-192.
+
+Koester, V. (2012). The Nagoya protocol on ABS: Ratification by the EU and its member states and implementation challenges. Studies N 03/12, Paris: IDDRI.
+
+Laird, S. A. (2000). Benefit-sharing ‘best practice’ in the pharmaceutical and botanical medicine industries. In H. Svarstad & S. S. Dhillion (Eds.), Bioprospecting: From biodiversity in the South to medicines in the North (pp. 89–99). Oslo: Spartacus.
+
+Louwaars, N., Dons, H., von Overwalle, G., Raven, H., Arundel, A., Eaton, D., & Nelis, A. (2009). Breeding business: The future of plant breeding in the light of developments in patent rights and plant breeders' rights. Wageningen: Center for Genetic Resources.
+
+Medaglia, J. C., Tvedt, M. W., Perron-Welch, F., Jørem, A. & Phillips, F. -K. (2013). The interface between the Nagoya protocol on ABS and the ITPGRFA at the international level: Potential issues for consideration in supporting mutually supportive implementation at the national level . Lysaker: FNI Report 1/2013.
+
+Ministry of Foreign Affairs. (1988). Norwegian contributions to international activities for sustainable development (the 'Blue Book'). Oslo: Ministry of Foreign Affairs.
+
+Ministry of the Environment. (1989). Environment and development: Programme for Norway’s follow-up of the Report of the World Commission on Environment and Development, Report No. 46 to the Storting (1988–1989), Ministry of the Environment, Oslo.
+
+Morgera, E., Buck, M., & Tsioumani, E. (Eds.). (2012). The 2010 Nagoya protocol on access and benefit sharing in perspective: Implications for international law and implementation challenges. Leiden: Martinus Nijhoff.
+
+Najam, A. (2005). Developing countries and global environmental governance: From contestation to participation to engagement'. International Environmental Agreements, 5(3), 303-321.
+
+![Figure](figures/figure_029.png)
+
+---
+
+Realizing access and benefit sharing from use of genetic resources
+
+595
+
+Oberthür, S., & Rosendal, G. K. (2014). Global governance of genetic resources: Access and benefit-sharing after the Nagoya protocol. Abingdon: Routledge.
+
+Oberthür, S., Gerstetter, C., Lucha, C., McGlade, K., Pozarowska, J., Rabitz, F. and Tedsen, E. (2011) Intellectual Property Rights on Genetic Resources and the Fight against Poverty , December 2011, Brussels: European Parliament, Directorate-General for External Policies of the Union, Policy Department.
+
+OECD. (2009). Biotechnology statistics, 2009. www.oecd.int.
+
+OECD. (2011). Key biotech indicators, share of countries in biotechnology patents filed under PCT, 2007–2009. Available at: www.oecd.org/dataoecd/41/46/48719943.xls.
+
+Pavoni, R. (2013). The Nagoya protocol and WTO law. In E. Morgera, M. Buck, & E. Tsioumani (Eds.), The 2010 Nagoya protocol on access and benefit sharing in perspective (pp. 185–213). Leiden: Martinus Nijhoff.
+
+Pistorius, R., & van Wijk, J. (1999). The exploitation of plant genetic information. Amsterdam: Amsterdam University Press.
+
+Prip, C., Rosendal, G. K., Andresen, S. & Tvedt, M. W. (2014). The Australian ABS framework: A model case for bioprospecting? FNI Report 1/2014, Lysaker.
+
+Putnam, R. D. (1988). Diplomacy and domestic politics: The logic of two-level games. International Organization, 42(3), 427-460.
+
+Raustiala, K., & Victor, D. G. (2004). The regime complex for plant genetic resources. International Organization, 58(2), 277-309.
+
+Rosendal, G. K. (1991). International conservation of biological diversity: The quest for effective solutions, R:012-1991. Lysaker: FNI.
+
+Rosendal, G. K. (1995). The Biodiversity Convention: an analysis of the background for and implications of the US 'no' in Rio. In J. Calestous & V. Sánchez (Eds.), Biodiplomacy (pp. 87-103). Nairobi: ACTS Press.
+
+Rosendal, G. K. (2000). The convention on biological diversity and developing countries. Dordrecht: Kluwer Academic.
+
+Rosendal, G. K. (2005). Biodiversity: International bungee jump—Domestic Bungle. In J. B. Skjerseth (Ed.), International regimes and Norway's environmental policy: Crossfire and coherence (pp. 161–194). Hampshire: Ashgate.
+
+Rosendal, G. K. (2007). ‘Norway in UN environmental policies: Ambitions and influence’. Journal of International Environmental Agreements: Politics, Law and Economics. Special Issue.
+
+Rosendal, G. K. (2011). Biodiversity protection in international negotiations: Cooperation and conflict. In Shlomi Dinar (Ed.), Beyond resource wars: Scarcity, environmental degradation, and international cooperation (pp. 59–86). Cambridge: MIT Press.
+
+Schei, P. J. (1997). 'Konvensjonen om biologisk mangfold'. In W. M. Lafferty, O. S. Langhelle, P. Mugaas, & M. Holmboe Ruge (Eds.), Rio+ 5: Norges oppfølging av FN-konferansen om miljø og utvikling. Oslo: Tano Aschehoug.
+
+Schloen, M., Louafi, S. & Dedeurwaerdere, T. (2011). Access and benefit sharing for genetic resources for food and agriculture. Current use and exchange practices. Commonalities, differences and user community needs. Report from a multi-stakeholder expert dialogue. Rome: CGRFA Background study paper no. 59.
+
+Svensson, U. (1993). The Convention on Biodiversity—A new approach. In S. Sjøstedt & H. Aniansson (Eds.), International environmental negotiations: Process, issues and context. Stockholm: FRN Utrikespolitiska Institutet.
+
+Swanson, Timothy (Ed.). (1995). Intellectual property rights and biodiversity conservation. Cambridge: Cambridge University Press.
+
+Tvedt, M. W. (2010). Norsk genressursrett [Norwegian Gene Resource Law]. Oslo: Cappelen Akademiske Forlag.
+
+Tvedt, M. W. 2014. Beyond Nagoya: Towards a legally functional system of access and benefit-sharing? In S. Oberthür. & G.K. Rosendal (eds.) (pp. 158-177).
+
+Tvedt, M. W., & Young, T. (2007). Beyond access: Exploring implementation of the fair and equitable sharing commitment in the CBD. Gland: IUCN.
+
+Underdal, A. (1992). Leadership in international environmental negotiations: Designing feasible solutions. Working Paper 1992:08. Oslo: CICERO.
+
+Wallbott, L., Wolff, F., Pozarowska, J. (2014). The negotiations of the Nagoya protocol: Issues, coalitions and process. In Oberthür, S. & G.K. Rosendal (Eds.) (pp. 33-59).
+
+WCED, World Commission for Environment and Development. (1987). Our common future. Oxford: Oxford University Press.
+
+![Figure](figures/figure_027.png)
+
+---
+
+596
+
+K. Rosendal, S. Andresen
+
+Young, O. R. (1982). Resource regimes: Natural resources and social institutions. Berkeley: University of California Press.
+
+Young, O. R. (1991). Political leadership and regime formation: On the development of institutions in international society. International Organization, 45(3), 281-308.
+
+![Figure](figures/figure_004.png)
+

processed/00207_W2994601296/document.md

@@ -0,0 +1,596 @@
+Maurer School of Law: Indiana University Digital Repository @ Maurer Law
+
+Indiana Journal of Global Legal Studies
+
+Volume 3 | Issue 2
+
+Article 2
+
+Spring 1996
+
+# The Externalization of Domestic Regulation: Intellectual Property Rights Reform in a Global Era
+
+Paul N. Doremus U.S. Department of Commerce
+
+Follow this and additional works at: https://www.repository.law.indiana.edu/ijgls
+
+![Figure](figures/figure_008.png)
+
+Law Commons
+
+## Recommended Citation
+
+Doremus, Paul N. (1996) "The Externalization of Domestic Regulation: Intellectual Property Rights Reform in a Global Era, " Indiana Journal of Global Legal Studies: Vol. 3: Iss. 2, Article 2. Available at: https://www.repository.law.indiana.edu/ijgls/vol3/iss2/2
+
+This Article is brought to you for free and open access by the Maurer Law Journals at Digital Repository @ Maurer Law. It has been accepted for inclusion in Indiana Journal of Global Legal Studies by an authorized editor of Digital Repository @ Maurer Law. For more information, please contact kdcogswe@indiana.edu.
+
+![Figure](figures/figure_014.png)
+
+JEROME HALL LAW LIBRARY
+
+INDIANA ENIVERSITY Manure School of Law Bloomington
+
+---
+
+# The Externalization of Domestic Regulation: Intellectual Property Rights Reform in a Global Era
+
+PAUL N. DOREMUS'
+
+Intellectual property rights (IPR) issues in the software, biotechnology, and semiconductor industries exemplify the pressure that new technologies and international competition are placing on domestic and international regulatory systems. Traditional patent and copyright rules cannot easily accommodate any of these technologies. At the same time, the high costs of research and development, relative ease of replication, and global markets characteristic of these technologies heighten the importance of both domestic and foreign IPR protection. In the context of rapidly changing technological conditions, borderless markets, and inflexible international regimes, national policymakers face a political dilemma: how to accommodate new technologies at home, encourage similar accommodation abroad, and do both without undermining either long-standing domestic IPR arrangements or the international patent and copyright regimes. This article reviews the different strategies of externalization associated with IPR reform in the software, biotechnology, and semiconductor industries. Variations across these cases indicate that fundamentally different technological, market, and political conditions can lead to different strategies for equilibrating incompatible and highly contested domestic and international regulatory rules.
+
+* Senior Analyst at the U.S. Department of Commerce, Technology Administration, responsible for analyses of regulatory, tax, and legal policies that affect corporate innovation and competitiveness. Mr. Doremus received a M.A. and Ph. D. from Cornell University. Mr. Doremus has researched and written on various topics related to corporate technology development, national innovation systems, intellectual property rights, multinational corporations, and international trade, investment, and competition. The research and views expressed in this article are the sole responsibility of the author and do not in any way reflect the work or the policies of the U.S. Department of Commerce.
+
+341
+
+---
+
+342
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+## I. INTRODUCTION
+
+In 1984, Congress created the first new intellectual property right (IPR) in nearly a century. The legislation, which became the Semiconductor Chip Protection Act of 1984 (SCPA), established a sui generis form of IPR protection for semiconductor chip products, a set of rights designed solely for semiconductor technology and statutorily separate from either copyright or patent law. $^1$ The SCPA is notable not only for its unique contribution to domestic IPR law, but also for its unique approach to coordinating domestic and foreign IPR rules. Instead of abiding by the national treatment provisions of the international patent and copyright regimes, the SCPA links domestic and foreign IPR rules through bilateral reciprocity based on “substantial similarity” to U.S. rules. $^2$ The rule of national treatment would have required the United States to provide the same IPR terms to foreign as to domestic intellectual property owners in the U.S. market, while accepting whatever IPR terms foreign countries established for semiconductor products in their own markets ––which, at the time, were literally nothing. Instead, the SCPA borrowed a more coercive technique from contemporary trade policy: the United States will grant IPR protection for foreign semiconductor products only if the property owner's home country establishes IPR rules similar to those in the SCPA. By most measures, this fairly coercive tactic has been rather successful. Japan negotiated interim reciprocity almost immediately and was followed quickly by several European nations. Numerous countries subsequently negotiated reciprocal protection with the United States.
+
+The primary congressional sponsor of the legislation, Robert W. Kastenmeier, hailed the SCPA as “pav[ing] the way” for innovative policy
+
+1. Semiconductor Chip Protection Act of 1984, Pub. L. No. 98-620, tit. III, 98 Stat. 3347 (codified at 17 U.S.C. §§ 901-14 (Supp. II 1984)). Until the 1984 legislation, the U.S. Congress had not created a new intellectual property right since 1881, when it recognized trademarks as protectable property. As a general term, “intellectual property” encompasses four categories of rights. The two primary categories, patents and copyrights, date to the U.S. Patent and Copyright Acts of 1790 and are rooted in Art. 1 Sec. 8 of the Constitution; the third, trademark protection, dates to 1881; and the fourth, semiconductor mask works, dates to 1984. There are also two somewhat lesser categories of intellectual property which are regulated at the state level: trade secrets and misappropriation of other proprietary information. These are much weaker forms of protection, being based less on property rights per se than on variable notions of fair business practices.
+
+2. The SCPA also allows for reciprocity via multilateral treaty, but the provision is virtually meaningless. No multilateral treaty existed in 1984, and none was expected in the near future–despite the efforts of the World Intellectual Property Organization (WIPO) to create one. To a large extent, the bilateral mechanism established by the SCPA obviated the need for a multilateral treaty.
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+343
+
+responses to new information-based technologies that fall outside the range of patent and copyright law, including such important technologies as biotechnology, software, and computer data bases. 3 But instead of taking the newly paved road, Congress and the courts have hesitated and chosen different routes with each of these new technologies. Semiconductor technology has prompted the most radical departure from traditional IPR rules: the United States legislated sui generis domestic protection and sought international coordination through bilateral reciprocity. Biotechnology has prompted an entirely different response: U.S. courts have accommodated biotechnology innovations by expanding the scope of patent law, and the United States (along with the European Union and Japan) has sought international coordination through the patent regime. IPR developments for software have shifted substantially over time. In 1980, Congress formally extended copyright protection to software, and did so in a manner consistent with the international copyright regime (even though neither of the international copyright conventions formally recognizes software copyrights). Since the mid-1980s, however, the politics of IPR reform for software have become considerably more complicated. At home, the courts have widened the scope of copyright protection for software and have even accorded patent rights to certain forms of software, 4 developments that have injected considerable uncertainty into U.S. IPR rules for software and further fueled already serious discord within the software industry over the proper scope and form of IPR protection. Abroad, the United States has pursued a series of Section 301 investigations centering on software copyright violations, representing a relatively coercive strategy for coordinating domestic and foreign copyright rules. 5
+
+3. Robert W. Kastenmeier & Michael J. Remington, The Semiconductor Chip Protection Act of 1984: A Swamp or Firm Ground?, 70 Minn. L. Rev. 417, 468-69 (1985).
+
+4. Software per se is not patentable. The types of software patents that have been granted are for patentable inventions that include software components cr software-controlled processes. Diamond v. Diehr, 450 U.S. 175 (1981); In re Alappat, 31 U.S.P.Q. 2d 1545 (Fed. Cir. 1994); In re Lowry, 32 U.S.P.Q. 2d 1031 (Fed. Cir. 1994). See also U.S. Dep't of Com., Patent and Trademark Office, Proposed ROPOSED Examination XAMINATION Guidelines UIDELINES for Computer-Implemented OMPUTER -IMPLEMENTED Inventions NVENTIONS ( 1995). Patentability criteria for software-related inventions remain uncertain in many respects. As technological developments blur the distinctions between hardware and software, and between tangible machines and data structures or programs, the scope and applicability of software patents will likely become even more contentious than it is today.
+
+5. Section 301 of the Trade Act of 1974, as amended, essentially provides the U.S. Trade Representative with the power to investigate unfair trade practices in foreign countries, and, under certain circumstances, to impose retaliatory sanctions. Trade Act of 1974, Pub. L. 93-618, 88 Stat. 1978 (codified at 19 U.S.C. § 2101 et seq.) (as amended by The Trade and Tariff Act of 1984, Pub. L. 98-573, 98 Stat. 2948, and the Omnibus Trade and Competitiveness Act of 1988, Pub. L. No. 100-418, 102 Stat. 1107).
+
+---
+
+344
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+Reforming domestic and foreign IPR rules to accommodate new, information-intensive technologies entails multiple political and economic factors operating at different levels and in different arenas. The technologies are complex, rapidly changing, and not easily accommodated by traditional classes of IPR. The markets involved are global, highly competitive, and integrally linked to other economic sectors. Consequently, the policy dilemma is simultaneously domestic and international, and the policy arena typically spans several sectors and competing political interests. In essence, IPR policy is no longer a matter of domestic regulatory choices alone. Changes in U.S. patent and copyright laws automatically have international implications due to U.S. participation in the international patent and copyright regimes as well as the recently negotiated Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPs) in the Uruguay Round of the General Agreement on Tariffs and Trade (GATT). Furthermore, many new information-intensive technologies are created in and used by highly global industries, which not only heightens the importance of having cognate IPR protection in foreign markets but also exposes the domestic reform process to international economic considerations.
+
+This article develops two general analytical propositions for interpreting and explaining how U.S. IPR policy has accommodated technological change, on both domestic and international fronts. First, international economic factors affect domestic IPR reform choices in ways that vary with the market structure of the affected industries. Specifically, different types of markets convey different degrees of trade leverage: the greater the trade leverage, the more likely externalization strategies will be coercive and bilateral; the lesser the trade leverage, the more likely externalization strategies will be cooperative and multilateral.
+
+Second, the style of accommodation and externalization associated with IPR reform must be viewed relative to the domestic political settlement between affected producers and users, whether they are competing producers, dependent users, or even economically unrelated industries (as can occur when existing holders of related property rights may perceive a threat to their interests). Like all forms of regulation, IPR rules embody adversarial political relationships. IPR reform pits contrary societal interests against each other in the courts and in Congress, where public officials seek to balance the interests of intellectual property owners against those of users--not just individual users but also the public at large, which benefits over the long run from the development and dissemination of new knowledge. Consequently, as the
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+345
+
+political and economic consequences of IPR reform become more acute for specific domestic actors (which often happens as technologies mature), coalitional conflicts can rise to the degree that IPR reform becomes much more problematic and prone to stalemate. Under conditions of domestic legal uncertainty and coalitional conflict, U.S. policymakers cannot easily pursue coercive externalization strategies.
+
+Although both market structure and domestic politics are highly influential, neither is fully determinative. The politics of IPR reform invariably involve a complicated relationship between international economic pressures and the domestic political dynamic that emerges from the politics of adversarial regulation. The full story of IPR reform is rarely compact, nor is the policy problem straightforward and easily negotiated. In this kind of territory, theoretical propositions are necessarily contingent.
+
+In the broadest sense, then, this article investigates the inexorable pressures of technological change and economic globalization on domestic regulatory policy. In essence, the globalization of commerce and capital has redefined the concept and practice of trade to include international transactions in services as well as global exchanges of knowledge and technology. Many of these new forms of global economic exchange have internationalized the scope and impact of domestic policy. International flows of goods and services as well as technology and capital are increasingly entwined with a range of fundamental domestic regulatory policies, including not only IPR but also investment, taxation, antitrust, environmental, and numerous other forms of regulation. The very nature of these relationships demands an interdisciplinary analytical perspective because they inherently involve law, public policy, foreign economic policy, and international political economy. In essence, one needs new analytical tools and theoretical frameworks to better understand the relationship between technological change, international competition, and domestic regulatory policy during a period that some now call “ the global era. ” 6
+
+The following section begins this analytical agenda by introducing the cases and describing the two fundamental dilemmas associated with IPR in advanced information-based industries. First, for innovators, obtaining adequate IPR protection is increasingly both important and uncertain. Second, for policymakers, the coordination of domestic and foreign IPR rules is simultaneously necessary yet difficult to achieve through traditional routes.
+
+6. ALFRED C. AMAN, JR., ADMINISTRATIVE LAW IN A GLOBAL ERA (1992).
+
+---
+
+346
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+In each of the three cases considered in this article--software, biotechnology, and semiconductors--the United States has sought to externalize domestic regulatory reform in substantially different ways, for reasons that are explored in Part III.
+
+## II. New EW Technologies, ECHNOLOGIES , New EW IPR Rules, ULES , and Different IFFERENT Styles TYLES of Externalization XTERNALIZATION
+
+This article examines U.S. IPR reform by comparing three case studies involving the software, biotechnology, and semiconductor industries. Each of these industries produces goods whereby the value of the information content far exceeds the value of the physical product on which the information is stored; indeed, information itself is the primary product. The cases are further similar in that they each involve leading-edge, information-intensive industries that are highly international, intensely competitive, and rapidly changing. Moreover, each involves technologies that confound the two main classes of intellectual property--patents and copyrights. $^7$
+
+Despite these fundamental similarities, each case represents a different process of domestic accommodation and a different style of externalizing domestic regulatory reform. Software has been granted copyright protection and limited patent protection, while the externalization strategy has varied over time; biotechnology has been accommodated through expanded patent rights that have been cooperatively externalized through multilateral channels; and semiconductor technology has been granted a sui generis IPR accompanied by coercive externalization conditions (see Table 1). In short, the analytical problem can be approached with the classic method of comparing similar cases with divergent outcomes.
+
+7. Intellectual property law for biotechnology is unique in that the extension of property rights to life forms invokes ethical and moral issues that often conflict with market development and competitiveness concerns. See U.S. CONGRESS, OFFICE OF TECHNOLOGY ASSESSMENT, PATENTING LIFE (1989).
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+347
+
+Table 1: Varying Styles of Externalization
+
+<table><tr><td></td><td>Cooperative Externalization</td><td>Coercive Externalization</td></tr><tr><td>Software, pre-1985</td><td>Through multilateral channels</td><td></td></tr><tr><td>Software, post-1985</td><td></td><td>Through bilateral and multilateral channels</td></tr><tr><td>Biotechnology</td><td>Through multilateral channels</td><td></td></tr><tr><td>Semiconductors</td><td></td><td>Through bilateral channels</td></tr></table>
+
+
+In the United States, the traditional justification for providing IPR protection in any industry is based on two simple but powerful assumptions: first, that innovators are motivated by the prospect of reward; and second, that society benefits from innovation by virtue of the economic and cultural growth that ensues. Given these assumptions, the state grants innovators limited monopoly rights in exchange for making the innovation public. Innovators benefit from the economic power of the monopoly, and society benefits from the generation and dissemination of new ideas. $^8$ By this logic, inadequate IPR protection eviscerates the incentive to innovate, which over the long run dampens economic growth and cultural advancement.
+
+New information-intensive technologies are particularly vulnerable to inadequate IPR protection because most can be disseminated very quickly and reproduced relatively easily (and at a fraction of the cost of development). Semiconductor chips can be reverse-engineered to decipher their design, software products can be readily copied and decompiled to reveal their
+
+8. Most property rights analysts accept that the rewards established by IPR are necessary to achieve socially optimal amounts of innovation, although there is a dissenting tradition that argues that neither IPR nor any other governmental support for innovative activity is necessary to achieve socially optimal innovation rates. The classic justification for maintaining an IPR system is Kenneth J. Arrow, Economic Welfare and the Allocation of Resources for Invention, in The Rate and Direction of Inventive Activity: Economic and Social Factors 609 (National Bureau of Economic Research ed., 1962). The dissenting tradition dates to Arnold Plant's work in the mid-30s. Arnold Plant, The Economic Aspects of Copyright in Books, 1 Economica 167 (1934); Arnold Plant, The Economic Theory Concerning Patents for Inventions, 1 Economica 30 (1934). More recent dissents include Stephen Breyer, The Uneasy Case for Copyright: A Study of Copyright in Books, Photocopies, and Computer Programs , 84 Harv. L. Rev. 281 (1970); Justin Hughes, The Philosophy of Intellectual Property , 77 Geo. L.J. 287 (1988).
+
+---
+
+348
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+structure, and biotechnology products can often be reproduced or readily imitated. Moreover, the type of intellectual property characteristic of these technologies cannot be protected through alternative channels, such as trade secrets, because the sale of information products usually reveals the commercially relevant proprietary information. In addition, the new information industries can only rarely rely on lead time advantages to recoup research and development (R&D) costs. R&D costs for these technologies have skyrocketed, product cycles have shortened, and competitors have improved their ability to rapidly replicate and market new technologies at a cost equal to, if not substantially lower than, the costs of innovation. When faced with imitative competitors who did not bear the considerable costs of R&D, innovative firms lose the ability to price their products competitively. Under these conditions, inadequate IPR protection can stifle innovation by increasing the uncertainty and risk of R&D investment.
+
+However great the need, though, it is extremely difficult to establish and enforce IPR protection for these types of technologies. The root problem is that information-intensive technologies tend to confound the criteria which differentiate the main classes of intellectual property. Semiconductor and software technology have characteristics that cross the borders of copyright and patent law; biotechnology products by no means look like the industrial inventions for which patent law was designed. In addition, new and powerful reproduction and transmission technologies--such as advanced photocopiers, audio and video recording devices, computers, electronic storage and retrieval systems, satellites, and cable--have created unparalleled domestic and international enforcement problems.
+
+Herein lies the dilemma for innovators in the new information technologies: their products are extraordinarily expensive to create yet comparatively cheap to duplicate and disseminate, which makes adequate IPR protection simultaneously more valuable and more problematic. The dramatic costs of inadequate IPR protection have created political pressure for IPR reform at home and abroad. In addition to the costs of lost markets and legal uncertainties at home, U.S. intellectual property owners have been absorbing increasingly large financial losses from inadequate intellectual property protection abroad. Estimates reach $ 61 billion a year, which alone captures
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+349
+
+political attention. 9 The economic and political significance of IPR protection is magnified by the wide-ranging economic relevance of the new information technologies, as their relative economic health and innovative capacity reverberate throughout such critical manufacturing and service industries as telecommunications, microelectronics, pharmaceuticals, and finance.
+
+The natural difficulty of adjusting IPR rules to accommodate the new information technologies is exacerbated by the inherently international character of the industries involved. Unlike traditional trade in goods, the commercial development and success of intellectual property products frequently depends on foreign laws and enforcement actions. For all practical purposes, the U.S. government simply cannot protect the IPR of its citizens through domestic policy alone. To strengthen the effective protection available to American intellectual property owners, the U.S. government must also pursue stronger foreign and/or international rules and enforcement practices.
+
+For over a century, the primary channels for coordinating domestic and foreign IPR rules have been the international patent and copyright regimes. $^10$ Each regime is a formal system of rules created by treaty and administered by an international organization. Neither regime establishes rights independently; rather, they adjust in the wake of domestic change and gradually harmonize the divergent patent and copyright rules of member states. IPR rules for new technologies generally develop within the countries in which the technology emerged, usually through an expansion or modification of either patent or copyright laws. The international patent and copyright regimes then harmonize domestic-level adjustments as they develop.
+
+Most of the new information-based technologies emerged in the United States, and the U.S. IPR system has often been the first to adapt to these
+
+9. Pinpointing monetary losses resulting from inadequate IPR protection is notoriously difficult, for it requires estimates of reduced profit margins, domestic and foreign sales displaced by infringing goods, damage to corporate reputations, and business foregone for lack of adequate property rights. The first comprehensive study of IPR-related business losses was conducted by the International Trade Commission (ITC) in 1988; although problematic in many respects, their estimate that $ 43 to $ 61 billion is lost annually remains a widely cited figure. U.S. Int'l Trade Comm'n, Pub. No. 2065, Foreign Protection Protection of Intellectual NTELLECTUAL Property ROPERTY Rights IGHTS and the Effect FFECT on US Industry NDUSTRY and Trade RADE (1988).
+
+10. The international patent regime centers on the Paris Convention for the Protection of Industrial Property (1883), which is administered through the World Intellectual Property Organization (WIPO). Paris Convention for the Protection of Industrial Property, Mar. 20, 1883, 13 U.S.T. I, 828 U.N.T.S. 107. The international copyright regime centers on the Berne Convention on Copyrights, Sept. 9, 1886, 168 Consol. T.S. 185 and the Universal Copyright Convention, Sept. 6, 1952, 6 U.S.T. 2731, 216 U.N.T.S. 132, the former being administered through WIPO and the latter through the United Nations Educational, Scientific and Cultural Organization (UNESCO).
+
+---
+
+350
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+technologies. Some IPR regimes have responded to particular technologies more quickly than the United States, which appears to be the case with copyright protection for computer databases in the European Union. Other nations have avoided IPR reform altogether, whether for databases or any other new information technology. Among those national IPR regimes that have responded, there are considerable differences and inconsistencies in approach and outcome.
+
+The uncoordinated and divergent development of domestic IPR rules magnifies the natural difficulty of coordinating IPR reform through the international patent and copyright regimes. Despite years of effort, the WIPO has gotten little support for its proposed multilateral treaty to protect semiconductor mask works; neither of the copyright conventions explicitly recognizes software copyrights, although many nations are treating software copyrights in a manner consistent with the regime; and the WIPO’s consideration of biotechnology patents lags considerably behind domestic developments in the United States, the European Union, and Japan.
+
+Herein lies the dilemma for U.S. policymakers facing domestic pressure for IPR reform at home and abroad: at the very time that international coordination is most needed, it is simultaneously most unlikely. In the context of rapidly changing technologies, international markets, and inadequate international regimes, national policymakers must decide how to adjust domestic policy in a manner that appropriately accommodates the new technologies at home and encourages similar accommodation abroad, preferably without undermining the otherwise useful international IPR regimes.
+
+Different strategies of externalization carry different political implications. As is the case in trade policy generally, a coercive approach to international IPR reform could undermine foreign support for the particular policy goals and perhaps damage the prospects for international cooperation in other areas. The choice of forum--multilateral or bilateral--has equally wide-ranging implications. The international patent and copyright regimes may move slowly, but they do provide a sure and predictable method of coordinating domestic and foreign IPR rules, which facilitates investment in technologies that are financially sustainable only in global markets, as is generally the case
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+351
+
+in information-intensive technologies. 11 In addition, the international agreements help to disseminate new technology globally by reducing the transaction costs of obtaining and enforcing exclusive rights in different countries. Bilateral strategies may be quick and effective in the short term. However, in the long run, they can introduce policy inconsistencies, create interpretive difficulties for domestic businesses, increase administrative complexity, and otherwise raise the transaction costs of obtaining IPR abroad (as firms are compelled to seek intellectual property protection for the same product in several countries with different standards and procedures). If it circumvents the regimes, the United States could lose its ability to influence how other nations and the international regimes accommodate the new technologies and protect U.S. intellectual property products in general.
+
+The broad political stakes and serious economic consequences of IPR reform have exacerbated major cleavages in the world economy, cleavages which not only set rich countries against poor but also set rich countries against each other. Many countries do not even share the basic normative underpinnings of western IPR concepts. For example, copying in traditional Korean society is an expression of honor, not an infringement of an inherent ownership right. 12 Other countries object to U.S.-style IPR protection on straightforward economic grounds. Brazil, for example, has long denied IPR protection for U.S. pharmaceutical products on the dual grounds that doing so would lock Brazil into technological dependency and would also create an enormous public health problem by making much needed pharmaceutical products prohibitively expensive. 13 Given these conditions, it is of little surprise that the United States has pursued high-profile bilateral trade disputes designed to force individual countries to establish IPR rules consonant with the IPR regimes (at least) and U.S. IPR rules (at best). The Reagan Administration initiated this tactic and secured major de jure IPR revisions in Korea, Taiwan, Singapore, Indonesia, Malaysia, Mexico, and Thailand. The Bush and Clinton Administrations have carried the agenda forward, pressing bilateral IPR cases
+
+11. Because most markets for information-intensive products are global, the availability of foreign
+
+
+IPR protection can significantly affect the R&D investment choices and market development strategies of multinational firms.
+
+12. See R. Michael Gadbaw, Republic of Korea, in INTELLECTUAL PROPERTY RIGHTS: GLOBAL CONSENSUS, GLOBAL CONFLICT? (1988).
+
+13. See Claudio R. Frischtak, The Protection of Intellectual Property Rights and Industrial Technology Development in Brazil, in INTELLECTUAL PROPERTY RIGHTS IN SCIENCE, TECHNOLOGY, AND ECONOMIC PERFORMANCE: INTERNATIONAL COMPARISONS 61, 61-98 (Francis W. Rushing & Carole Ganz Brown eds., 1990).
+
+---
+
+352
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+with China, Brazil, and other countries, and raising the political sights of the United States to de facto as well as de jure IPR protection. $^14$
+
+Although most of the advanced industrial states have IPR regimes with similar normative underpinnings, there are nonetheless substantial conflicts between them over the proper scope and terms of IPR protection. For instance, the whole issue of designing IPR protection for semiconductor mask works developed largely in response to powerful U.S. semiconductor manufacturers' claims that Japanese copying of U.S. innovations led to the rapid and devastating loss of the 64K DRAM market in the early 1980s. 15 In a different vein, large U.S.-based software producers have bitterly battled several European countries, accusing them--all net importers of software--of exploiting loopholes in the Berne Convention to avoid paying U.S. copyright royalties. U.S. software producers, led by the Business Software Alliance, also have turned to fairly dramatic legal efforts to combat corporate and residential copying within the member countries of the Organisation for Economic Cooperation and Development (OECD), which they estimate deprives them of several billion dollars annually in sales revenue. 16 IPR rules in many OECD countries remain unclear or unsettled on software patents and decompilation rights as well as on whether certain categories of biotechnology are patentable.
+
+At a time when most OECD economies are shifting toward services and information-based products, the problem of inadequate IPR protection has taken on profound significance. Information and information-based products and services are not just intrinsically valuable commodities; they also affect productivity and competitiveness across numerous sectors. The character of domestic and international IPR rules for new information-based technologies
+
+14. United States' progress on international IPR issues and other trade disputes is documented on an annual basis in two reports published by the U.S. Trade Representative. U.S. Trade RADE Rep., NE . Trade RADE Policy OLICY Agenda AGENDA and Annual ANNUAL Report EPORT of the President RESIDENT of the United NITED States on the Trade RADE Agreements GREEMENTS Program ROGRAM (1986-1996); U.S. Trade RADE Rep., EP., National ATIONAL Trade RADE Estimate STIMATE on Foreign OREIGN Trade RADE Barriers ARRIERS (1986-1996). Annex II of the 1996 Trade RADE Policy OLICY Agenda and 1995 Annual ANNUAL Report PPORT provides a comprehensive listing of formal trade agreements entered into by the United States since 1984. The annual National TRADE Estimates STIMATES provides the USTR's assessment of progress on IPR reform and other trade matters on a country by country basis.
+
+15. The accuracy of this claim is a matter of dispute, but there is no denying that U.S. semiconductor producers made the claim frequently and were rarely questioned about it in policy circles.
+
+16. The most recent survey conducted by the Business Software Alliance (BSA) estimates that foreign piracy in the software publishing and distribution industries cost U.S. firms $ 15.2 billion in lost revenue for 1994. BSA attempts to counter software piracy in foreign markets by conducting enforcement, education, and public policy campaigns in over 60 countries. Since 1988, BSA has filed approximately 600 lawsuits worldwide against suspected infringers of software copyrights. Business OSFTNFARE Alliance, The Impact of Software Piracy on the International Marketplace, in Annual REPORT (1995).
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+353
+
+can profoundly alter transborder data flows, trade in information products and services, and trade in a wide range of information-intensive industries. Given the central position of information-based technologies in the economic food chain, the nature of IPR reform could strongly affect overall levels of innovation and growth in the post-industrial service economies: inadequate IPR protection raises the uncertainties and risks of conducting R & D, which slows the pace of innovation and consequently could eviscerate the long-term competitiveness of the economy's most critical growth sectors.
+
+Serious political attention to the linkages between IPR protection, innovation, and competitiveness emerged in the early 1980s, congruent with a sense of impending crisis over the intractable trade and budget deficits as well as the comparative decline in U.S. competitiveness in general. The profound consequences of inadequate IPR protection became an important topic not just in trade circles but also in rather high political circles, as reflected in the 1985 report on the “new reality” of global competition conducted by the President's Commission on Industrial Competitiveness:
+
+Technological innovation is a mainstay of the American economy. It is the foundation of our economic prosperity, our national security, and our competitiveness in world markets. Yet, despite its importance, technological innovation has not always been optimally nurtured in America. One weakness in U.S. policy has been the lag between the advances in technology and the adaptations of U.S. intellectual property law to protect them. . . . If ownership rights are not improved domestically and internationally, the United States will limit its innovative capability and, consequently, its national economic and social development for generations to follow. At a period when other nations are focusing on ways to enhance their technological base, the United States could well be left behind. $^17$
+
+This concern with the implications of IPR rules for national competitiveness has not been purely rhetorical. U.S. courts have been
+
+17. President's Comm'n on Industrial Competitiveness, 2 Global Competition: The New Reality 306 (1985).
+
+---
+
+354
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+according stronger proprietary rights to intellectual property owners, 18 Congress has explicitly linked trade sanctions to foreign IPR conditions in the 1984 and 1988 Trade Acts, 19 and the Executive Branch has pursued foreign IPR reform through Section 301 and related bilateral tactics. 20
+
+Such political attention to the link between IPR and international competitiveness illustrates just one way in which the United States has been grappling with profound changes in the international economy, changes that have imposed new and distinct pressures upon domestic policy establishments. The globalization of commerce and capital has redefined the concept and practice of trade to include international transactions in services as well as global exchanges of knowledge and technology. Many of these new forms of global economic exchange have internationalized the scope and impact of domestic policy. International flows of goods and services as well as technology and capital are increasingly entwined with a range of domestic regulatory policies, including not only IPR but also investment, tax, antitrust, environmental, and other forms of regulation (see Table 2 ).
+
+18. The evidence that U.S. IPR rules have shifted in favor of proprietary rights is compelling and widely accepted. For one, the scope of intellectual property protection has been extended to previously unprotected objects (such as life forms and mathematical algorithms). In addition, U.S. courts have been ruling increasingly in favor of intellectual property owners; most often noted is the clearly pro-patent decision record of the U.S. Court of Appeals for the Federal Circuit (CAFC), which was established in 1982 to consolidate and systematize patent law. On the CAFC, see Rochelle Cooper Dreyfuss, The Federal Circuit: A Case Study in Specialized Courts 64 N.Y.U. L. Rev. 1, 26-30 (1989). According to one critic, the CAFC has decided in favor of patent holders in over 60% of the cases heard; overall, “the CAFC has greatly strengthened the presumption of patent validity and upheld royalties ranging from 5 to 33 percent.” Brian Kahin, The Software Patent Crisis , 93 Tech. Rev. 57 (1990).
+
+19. During the 1980s, Congress passed several legislative provisions that allowed trade retaliation for inadequate IPR protection in foreign markets. See Caribbean Basin Economic Recovery Act of 1983, 26 U.S.C.A. § 6015, repealed (linking foreign aid to IPR protection); Trade and Tariff Act of 1984, 19 U.S.C.A. § 1654 (explicitly linking GSP preferences to the protection of U.S. IPR and also making foreign IPR protection subject to direct trade retaliation through the Section 301 provisions of the Trade Act of 1974, supra note 5 (which is the principal statutory basis for addressing unfair or discriminatory foreign government practices that burden or restrict U.S. commerce—such as inadequate IPR protection.)); Omnibus Trade and Competitiveness Act of 1988, supra note 5 (Special 301 provisions require the U.S. Trade Representative to identify “priority” foreign countries that deny adequate and effective IPR protection and/or fair and equitable market access for firms that rely on IPR protection and countries placed on this list subsequently become the focus of increased bilateral attention).
+
+20. As noted above, the USTR has successfully achieved IPR reform through bilateral trade disputes with Korea, Taiwan, Singapore, Indonesia, Malaysia, Mexico, China, Thailand, and others. See sources cited supra notes 12, 13, & 14.
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+355
+
+Table 2: The Intersection of Domestic Regulation and International Competition: Relevant Policy Arenas
+
+<table><tr><td>Policy Arena</td><td>Policy Instruments</td><td>Affected Industries</td></tr><tr><td>Intellectual property rights</td><td>Patent, copyright, and trademark rules and enforcement</td><td>Semiconductors, software, biotechnology, recording and publishing, pharmaceuticals, and others</td></tr><tr><td>Foreign investment</td><td>Ownership limits; local content rules; licensing and other technology transfer rules</td><td>Global manufacturing and service industries</td></tr><tr><td>Taxation</td><td>Investment credits; R&amp;D credits; depreciation rules; transfer pricing</td><td>Global manufacturing and service industries</td></tr><tr><td>Antitrust</td><td>Merger and acquisition rules and enforcement; R&amp;D consortia; traditional antitrust (pricing; market concentration)</td><td>Oligopolistic industries, such as aerospace, semiconductors, automotive, telecommunications, and—perhaps—software</td></tr><tr><td>Environmental regulation</td><td>Emissions and other output rules; "green" standards; take-back rules; design conventions</td><td>Refrigerants, automotive, electronics, forestry, utilities</td></tr><tr><td>Standards</td><td>Product quality, process and interoperability controls, health and safety</td><td>Electronics, data processing, computers, software, biotechnology</td></tr><tr><td>Technology policy</td><td>Direct and indirect forms of government R&amp;D support; sector targeting; public/private alliances; procurement</td><td>Semiconductors and electronics, aerospace, instruments, optics, advanced materials, information technology</td></tr></table>
+
+
+---
+
+356
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+For the most part, the trade implications of domestic regulatory policies are not suitably addressed by contemporary international trade institutions. The dominant postwar international institution for governing trade, the GATT, applied only to trade in goods and consequently left an increasingly large part of international economic relations uncovered. Largely as a result of U.S. pressure, the Uruguay Round sought to expand the GATT's scope by negotiating three agreements covering the “new trade issues”–the General Agreement for Trade in Services (GATS), the Agreement on Trade-Related Investment Measures (TRIMs), and the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPs).
+
+At root, the Uruguay Round strategy of expanding and harmonizing international IPR rules through the GATT/WTO represented an attempt to thoroughly revise the long-standing rules of engagement structured by the international patent and copyright regimes. The TRIPs agreement may strengthen IPR protection worldwide and could change the rules of international coordination through the enforcement power of trade retaliation. However, the TRIPs agreement builds upon minimal rules in the existing IPR regimes and does not address the manifold problems associated with accommodating new information-intensive technologies. 21 In addition to having a narrow scope, the TRIPs mechanism has the additional liability of being excruciatingly slow; it is far too cumbersome to accommodate the rapid rate of technological change and economic development in new informationbased industries.
+
+U.S. industry and the U.S. government consequently have been left to seek alternative strategies for promoting foreign IPR reform. The primary strategic choice is whether to remain within the confines of the international IPR regimes, which in essence is a choice between multilateralism and bilateralism. The United States has opted for a bilateral strategy with regard to semiconductor mask works, but has retained an essentially multilateral strategy with biotechnology and has used both mechanisms in software. 22 Whatever the choice of forum, domestic reformers can pursue foreign accommodation either
+
+21. As stated by the OECD, “A more fundamental approach appears to be needed in order to devise internationally consistent rules that are aligned more closely with the characteristics of contemporary knowledge generation, invention and diffusion.” OECD, Science and Technology Policy: Review and Outlook 1991 , 93 (1992).
+
+22. The style of externalization in software may be confusing at first glance. Although the United States has used coercive tactics such as Section 301 investigations, the goal has been to win foreign adherence to the international copyright regime; accordingly, the strategy is essentially multilateral, despite the aggressive attitude.
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+357
+
+coercively or cooperatively. The most cooperative strategy is to facilitate the regime's harmonization process through international conferences, crossnational policy reform discussions, and so forth—as has primarily been the case with regard to biotechnology patent rules. Coercive tactics are somewhat broader in range—the United States can encourage foreign adherence to regime rules through “carrot and stick” tactics such as the reciprocity provisions of the SCPA, or it can use even more forceful tactics such as the threat of trade retaliation (as has been the case in software).
+
+In short, the United States has used diverse styles of externalization to accommodate IPR reform in new information-intensive technologies. The following section begins the process of interpreting these differences, developing lines of inquiry based on the particular market structure of the affected industries as well as the political character of IPR regulation.
+
+## III. Explaining XPLAINING Styles TYLES of Externalization: XTERNALIZATION : Public UBLIC Policy OLICY in the Context ONTEXT of International NTERNATIONAL Competition OMPETITION
+
+Until relatively recently, IPR policy and trade policy were entirely separate arenas. Each had its own set of domestic laws and international agreements, administered by entirely distinct agencies and organizations whose efforts rarely required coordination. In their operations abroad, U.S. corporations generally dealt with IPR matters on local terms, negotiating agreements with individual governments on a case-by-case basis. However, in the late-1970s and early-1980s, a number of changes in the international economic environment rendered particular domestic industries increasingly vulnerable to inadequate IPR protection abroad. That vulnerability led to private-sector activism that ultimately dissolved the political and institutional barriers between international trade and domestic IPR policy.
+
+The first U.S. firms to target inadequate foreign IPR protection were industries sensitive to trademark protection. In fact, the very idea of linking IPR and trade dates to 1978, when the Levi Strauss Corporation began a concerted effort to combat foreign counterfeiting of their trademark blue jeans. Levi Strauss' first step was to strengthen border sanctions; it lobbied Congress to add seizure and forfeiture provisions to Section 1526 of the Tariff Act of 1930, and Congress promptly obliged in the Customs Procedural Reform and
+
+---
+
+358
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+Simplification Act of 1978. 23 In the spring of 1978, Levi Strauss went one considerable step further. Encouraged by the adoption of “codes of agreement” during the Tokyo Round of GATT negotiations, the company banded together with a group of trademark-sensitive firms and pressed for an anti-counterfeiting code. The Office of the U.S. Trade Representative (USTR) took up the cause and, despite introducing the initiative very late in the GATT negotiations, nearly succeeded. 24 Although unsuccessful in the Tokyo Round, Levi Strauss and the newly created International Anti-Counterfeiting Coalition (IACC) had set a powerful precedent: they had swung U.S. trade policy behind the idea of linking IPR conditions to trade, and had firmly established the desirability of a new international mechanism for coordinating national IPR rules through the GATT.
+
+By the early-1980s, some of the largest manufacturers of industrial and leading-edge high technology products joined the trademark industries in demanding improved foreign IPR protection. The pressure for foreign IPR reform became more widespread and more urgent as the U.S. economy and international trade in general became increasingly oriented around products and services with a high intellectual property content. In addition, inadequate IPR protection abroad became more costly as new product generations commanded higher and higher R & D outlays at the same time that relatively inexpensive reproduction technologies emerged.
+
+These economic and technological trends brought diverse industries-ranging from automotive products and agricultural chemicals to pharmaceuticals and electronics--into the IPR debate. For the most part, these industries have been interested in establishing foreign IPR rules equivalent to those in the United States, or at least to those mandated by the international patent and copyright regimes. Those ends alone have proven to be challenging.
+
+For new information-based technologies, the problem of achieving adequate IPR protection has been doubly vexing, for these technologies do not easily fit within existing categories of IPR. For example, domestic IPR protection for the semiconductor, software, and biotechnology industries did not take shape until relatively recently. In 1980, Congress extended the Copyright Act of 1976 to software products, although the scope and form of
+
+23. Tariff Act of 1930, 19 U.S.C. §1526. For an account of the anti-counterfeiting code in the Tokyo Round, see William N. Walker, Private Initiative to Thwart the Trade in Counterfeit Goods, 4 WORLD ECON. 29 (1981).
+
+24. Walker, supra note 23, at 29.
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+359
+
+that protection has remained highly contentious. The Supreme Court set the precedent allowing patents for biotechnology products in 1980, with the landmark Diamond v. Chakrabarty decision, 25 and Congress created a sui generis form of industrial property right for semiconductor mask works in 1984.
+
+Because these technologies are so complex and unusual, and because the industries that deploy them are largely international in scope, the politics of adjusting domestic IPR rules invariably coexists with political pressure for foreign IPR reform. The dual character of IPR reform in these technologies suggests two lines of inquiry regarding externalization strategies. First, because the entire problem of accommodating the new information technologies emerged during a period of renewed political attention to innovation and competitiveness, strategies of externalization are likely to be shaped by the pressures of international competition. Second, because these technologies have required fundamental adjustments to domestic IPR rules, strategies of accommodation and externalization are also likely to be shaped by the domestic political bargaining involved in creating and regulating new forms of IPR. The interpretation and explaination of variances in externalization strategies, in other words, requires a multi-level analysis: the first stage involves outlining fundamental variances in market structure and associated levels of trade leverage, while the second stage situates domestic political conditions within that broader market context.
+
+In terms of market structure, the semiconductor, software, and biotechnology industries are each relatively new and are based on rapidly changing, information-intensive technologies. In addition, the markets for each are global in the sense that products are developed, produced, and sold on an international scale. However, neither condition is uniform. Technology varies in terms of its level of development (developing or mature), and markets vary roughly in terms of the product cycle (market creation, market expansion, and market saturation). Accordingly, competition involving emerging technologies in new markets tends to center on R & D and market creation -- that is, competition is a positive sum game. Competition involving mature technologies in highly developed markets tends to be a more contentious battle for market share; as technology disperses and markets become saturated, competition invariably becomes a zero-sum game. Intermediate markets --
+
+25. Diamond v. Chakrabarty, 447 U.S. 303 (1980).
+
+---
+
+360
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+highly commercialized but not saturated--tend to be characterized by competition for market expansion.
+
+By extension, the political and economic implications of IPR reform should vary across different technological and market circumstances. For instance, in very developed markets involving mature technologies (like semiconductors), IPR reform will tend to have immediate implications for competition over market share. In new markets involving developing technologies (like biotechnology), IPR reform will have more long-term implications for economic competitiveness. 26
+
+The three industries assessed in this article nicely illustrate the relationship between market structure and the political salience of IPR reform. At the time of the SCPA, the commercial implications of IPR reform in semiconductors were sobering. 27 Large U.S.-based semiconductor firms had been steadily losing market share to Japanese competitors, and attributed at least part of that loss to inadequate IPR protection. 28 The implications of IPR reform in software have been significant but, arguably, less dramatic. U.S. software firms clearly dominate world markets, such that inadequate IPR protection threatens not their very existence (as semiconductor producers characterized the stakes for themselves) but instead their ability to compete “fairly” in domestic and foreign markets. The commercial implications of IPR reform in biotechnology may also be significant but are less compelling than in semiconductors or software. Since the technology remains largely precommercial, IPR reform primarily affects the potential for developing new products and markets. Consequently, IPR reform does not provoke immediatè trade conflicts.
+
+In short, the different structural features of each market convey a different degree of potential trade leverage for IPR reform. The semiconductor industry has had a high degree; the software industry, a moderate yet increasing degree; and the biotechnology industries, a relatively low degree. These variances
+
+26. The character of IPR rules tends to vary with the technology's level of development. The scope of IPR applied to developing technologies often is rather broad, and the terms generous; as knowledge expands and the technology matures, the scope narrows and the terms become more specific and restrictive. This pattern is generally more true of patents than copyrights, although the extension of copyrights to software (and perhaps other utilitarian articles) tends to follow this pattern.
+
+27. U.S. DEP'T OF COM. INT'L TRADE ADMIN., THE SEMICONDUCTOR INDUSTRY (1983).
+
+28. See The Semiconductor Chip Protection Act of 1983: Hearings on S. 1201 Before the Subcomm. on Patents, Copyrights, and Trademarks of the Senate Comm. on the Judiciary, 98th Cong. 1st Sess. 66, 7582 (1983) (statements of Thomas Dunlap Jr., Intel Corporation, and Christopher K. Layton, Intersil Corporation).
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+361
+
+generate the hypothesis that trade leverage can shape IPR reform in particular ways: the higher the trade leverage, the more likely that externalization strategies will be coercive and bilateral; the lower the trade leverage, the more likely that externalization strategies will be cooperative and multilateral.
+
+However, although trade leverage may be highly influential, it is rarely determining. IPR rules embody a fundamental domestic political tension. On the positive side, IPR provide financial incentives to individual innovators, incentives needed to promote creative activity and, over the long run, maximize societal welfare through technological and economic growth. On the negative side, IPR grant economic and legal powers to innovators that, if too encompassing, can be used to extract excessive profits and restrict competition—which, over the long run, would damage societal welfare. IPR, in short, are a form of adversarial regulation. IPR rules distribute costs and capabilities among competing groups that are in a zero-sum relationship (as opposed to policies that regulate individuals or groups for their own individual or collective benefit). $^29$
+
+In essence, IPR rules represent a state-sanctioned distribution of power, a form of social contract designed by the government to balance the rights of innovators and users and to further society's interest in economic and cultural innovation. IPR reform potentially affects not only the interests of prospective rights holders and related users, but also the interests of long-standing rights holders, whether or not they are in economically related arenas. Accordingly, one would expect the politics of IPR reform to be shaped considerably by the classic organizational and associational characteristics of affected groups--their structure (concentrated or dispersed), their cohesiveness (cooperative or conflictual), and the general nature of the policy cleavage (within an industry or between industries).
+
+29. This policy characterization is based on Lowi's policy typology, which was first outlined in Theodore J. Lowi, American Business, Public Policy, Case-Studies, and Political Theory , 16 World Politics 677 (1984). On the definitions of and distinctions among distributive, regulatory, redistributive, and constituent policies, see Theodore J. Lowi, Four Systems of Policy, Politics, and Choice , in Public Administration Review 298 (1972). As Lowi noted, adversarial regulation can involve either behavior that is intrinsically undesirable (e.g. crime) or behavior that is undesirable only in its consequences (e.g. monopolies); the critical condition is that one set of actors is being regulated to protect or enhance the interests of an opposing set. IPR policy, along with similar forms of economic regulation (such as antitrust regulation) and social regulation (such as environmental and civil rights regulation), clearly fits in the category of behavior that matters because of its consequences. Wilson, to the contrary, treats social regulation and regulation of competitive practices as different classes; his differentiation is based more on the issues at stake than the fundamental relationships characteristic of adversarial regulation. See The Politics of Regulation (James Q. Wilson, ed. 1980).
+
+---
+
+362
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+Of course, the problem of adapting IPR rules to the new information technologies is not simply a domestic matter. Choosing how to regulate private gain and public welfare in the new information-based technologies automatically assumes an international dimension, because private gain involves foreign IPR rules and public welfare involves international technologies in international markets. The global character of the new information technologies exposes the politics of adversarial regulation to the considerable pressures of trade and international competition.
+
+In sum, the politics of IPR reform invariably involves a complicated relationship between international economic pressures and the domestic political dynamic that emerges from the politics of adversarial regulation. The complex role and significance of variances in market structure and domestic policy dynamics assumes full force when the three cases are viewed as a set. The next section provides a brief and highly stylized explanatory overview of the cases, illustrating the significance of a multilevel analysis of market structures and political dynamics. Building theory around this type of research question is, however, a highly contingent venture: the technologies and industries are complex and rapidly changing, the pressures on domestic and international rule systems are multitudinous and unrelenting, and the legal and political responses to these pressures at home and abroad are multifaceted and usually uncoordinated. These circumstances are certainly less conducive to theory-building than the comparatively static international environment of past decades. Then again, the research question approaches one of the most politically and theoretically salient needs of our times--how to grapple with the profound impact of technological change and the globalization of capitalism on domestic policy choices and the process of international policy coordination.
+
+## IV. IPR Reform EFORM in Semiconductors, EMICONDUCTORS, Software, OFTWARE, and Biotechnology IOTECHNOLOGY
+
+The software, biotechnology, and semiconductor industries are fundamentally similar in that they each involve leading-edge, informationintensive industries that are highly international, intensely competitive, and rapidly changing. In addition, each involves technologies that confound the two main classes of intellectual property – patents and copyrights. Despite these fundamental similarities, each industry has displayed a different process of domestic accommodation and a different style of externalizing domestic
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+363
+
+regulatory reform. These variances conform to fundamental differences in the interaction between structural market conditions and the political dynamics associated with adversarial regulation (see table 3 ).
+
+---
+
+364
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+Table 3: Explaining Different Styles of Externalization: Market Structure and Domestic Politics
+
+<table><tr><td></td><td>Software, pre-1985</td><td>Software, post-1985</td><td>Biotechnology</td><td>Semiconductor</td></tr><tr><td>MARKET STRUCTURE</td><td></td><td></td><td></td><td></td></tr><tr><td>Technological development</td><td>Developing, low commercialization</td><td>Intermediate, highly commercialized</td><td>Developing, low commercialization</td><td>Mature, highly commercialized</td></tr><tr><td>Competition</td><td>Positive sum, over market creation</td><td>Mixed sum, over market expansion</td><td>Positive sum, over market creation</td><td>Zero-sum, over market share</td></tr><tr><td>Trade leverage</td><td>Low (negligible trade volume)</td><td>Mid to high (increasing export sensitivity, overseas expansion threatened.)</td><td>Low (negligible trade volume)</td><td>High (both domestic and export market shares contested)</td></tr><tr><td>DOMESTIC POLITICS</td><td></td><td></td><td></td><td></td></tr><tr><td>Industry organization</td><td>Dispersed, cooperative</td><td>Dispersed, increasingly conflictual</td><td>Dispersed, cooperative</td><td>Concentrated, cooperative</td></tr><tr><td>Legal-regulatory conflict</td><td>Low</td><td>Mid to high</td><td>Low</td><td>High</td></tr><tr><td>Adversarial coalition</td><td>Extra-industry</td><td>Intra-industry</td><td>Extra-industry</td><td>Inter-industry</td></tr><tr><td>SYCLE OF EXTERNALIZATION</td><td>Cooperative, multilateral</td><td>Coercive, billmutilateral</td><td>Cooperative, multilateral</td><td>Coercive, bilateral</td></tr></table>
+
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+365
+
+In the case of the software industry, the shift from cooperative multilateralism to a more coercive style of externalization with both bilateral and multilateral elements conforms to fundamental changes over time in the industry’s market structure, the increasingly complex character of the technology, and the gradual evolution of coalitional factionalization.
+
+In the late 1970s, software technology was relatively new, retail markets were only developing, and competition centered primarily on creating new products. The very extension of copyright protection to computer software in the Copyright Act of 1980 was intended to encourage the growth of independent software producers and reinforce the open and fluid competitiveness characteristic of the software industry in the late 1970s. By design, the 1980 Copyright Act amendments set in motion a gradual harmonization process that encouraged similar adjustments in member nations. 30 Between 1980 and 1985, other nations moved toward copyright protection for computer programs to a strong enough degree that WIPO abandoned its effort to draft an international treaty for IPR in software. During this period, cooperative multilateralism was a viable and productive externalization strategy. Other nations were moving in the same direction, and cross-national differences in copyright protection had not yet begun to create significant economic problems.
+
+In the mid-1980s, however, international trade and competition in the software industry changed dramatically, which in turn encouraged an entirely different approach toward equilibrating incompatible national IPR laws. As the software industry internationalized and shifted toward market-expansion competition, U.S. firms became increasingly exposed and sensitive to foreign counterfeiting and other difficulties caused by cross-national differences in IPR protection for software. Under these conditions, the software industry needed quick, widespread adaptation of copyright laws to accommodate software as a literary work. The best route for gaining widespread change was through the long-standing international copyright regime. However, the regime's harmonization processes are slow, its minimum rules are quite weak, and it entirely lacks enforcement power. Consequently, the U.S. software industry created a powerful alliance with other copyright-sensitive industries and pressed for wholesale reform of the multilateral copyright regime.
+
+30. Marla R. Bloch, The Expansion of the Berne Convention and the Universal Copyright Convention to Protect Computer Software and Future Intellectual Property, 11 Brooklyn J. INT'L L. 283 (1985).
+
+---
+
+366
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+Encouraged by the near passage of an anti-counterfeiting code in the Tokyo Round, the coalition pressed the U.S. government for what eventually became the TRIPs proposal in the Uruguay Round, which essentially strengthened the minimal copyright terms provided by the Berne Convention and added the enforcement power of trade retaliation. 31 In the face of strong early opposition to this strategy, the copyright coalition encouraged the use of Section 301 tactics to develop both the substantive basis and the political support for the TRIPs program, as well as to provide a fail-safe in case the Uruguay Round might collapse. 32 From the outset, the USTR treated the bilateral Section 301 cases and the multilateral TRIPs agreement as symbiotic exercises; the bilateral cases were designed to convey U.S. intentions and gradually develop a web of agreements that would substantively support the multilateral policy goals of the TRIPs agreement.
+
+In essence, the externalization strategy of coercive multilateralism grew out of the high and increasing degree of internationalization in the software industry and the unusual distribution of market share. The size and dominance of the U.S. market has provided a platform from which U.S. software firms have been able to compete in foreign markets. As U.S. software firms internationalized, they became increasingly sensitive to export market access; in this context, inadequate IPR emerged as a serious market barrier in numerous foreign countries. In effect, the burgeoning export trade sensitivity of U.S. producers created an imperative for rapid and widespread international IPR reform. The international copyright regime held out the best prospect for achieving widespread reform, although the pace of adjustment was excruciatingly slow. Section 301 offered a channel for accelerating the process, and the software industry had sufficient internal coherence, external alliances, and political clout to see its preferred strategy implemented. The result was coercive externalization, through both bilateral and multilateral channels.
+
+In merely five years, the United States took a relatively arcane regulatory issue, legitimized its treatment as an important trade issue, and catalyzed what quickly became an international consensus on substantive copyright protection for computer software. Granted, several countries may have adopted a similar copyright solution either for internal reasons or to remain consistent with the
+
+31. Interviews with representatives of the International Property Rights Alliance, the International Anti-Counterfeiting Coalition, and the Office of the U.S. Trade Representative, in Washington D.C.,( JuneJuly 1992).
+
+32.1d.
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+367
+
+international copyright regime. Nevertheless, numerous other countries (especially in the developing world) would not have moved in that direction were it not for the coercive element of the U.S. strategy. Indeed, the United States continues to exert pressure of this sort in its quest to improve enforcement conditions in several countries.
+
+Although impressive, the international acceptance of software as a copyrightable literary work has not closed the book on copyright reform for computer software. In the United States and abroad, the increasing complexity of software technology and the gradual onset of market-share competition have unraveled the former consensus on basic copyright reform. Since the late 1980s, developments in software technology and copyright law have exacerbated the legal uncertainties inherent in the 1980 copyright settlement and have fostered increasingly bitter disputes within the U.S. software industry. 33 Current legal and political trends indicate that the early public concern with encouraging innovation (as voiced in 1980) has shifted to concern with preserving competition. The complex and contentious legal battles over interface protection and decompilation rights have fostered interindustry and intra-industry hostilities that foretell a major conflict over IPR reform in the near future. In the context of domestic legal flux, little can be done about discrepancies between U.S. copyright rules and those of other nations, such as those between the United States and the European Union regarding software decompilation rights.
+
+Changes in the market structure of the software industry, in short, have produced different IPR dynamics and have encouraged different strategies of externalization. The widespread lack of adequate IPR protection abroad and the increasing export sensitivity of U.S. software producers established strong incentives for a multilateral externalization strategy; the shift from marketcreation competition to market-expansion competition established incentives to shift from a cooperative to a coercive externalization attitude. The software industry's current shift into a mode of market-share competition corresponds with divisive IPR disputes that portend the end of an era of consensus in the international harmonization of IPR rules for computer software.
+
+Although changes in the market structure of the software industry have directly shaped the strategy of externalization used to equilibrate U.S. and
+
+33. Pamela Samuelson, A Case Study on Computer Programs, in Global Dimensions of Intellectual Property Rights in Science and Technology 284 (Mitchel B. Wallerstein et al., eds., 1993).
+
+---
+
+368
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+foreign copyright rules, the analytical significance of market conditions cannot be separated from the domestic politics of IPR reform. As the software case also illustrates, the politics characteristic of adversarial regulation change as the political and economic consequences of regulatory policies become more acute for specific domestic actors; as policy choices move toward a zero-sum game, coalitional conflicts can rise to the degree that IPR reform becomes much more problematic and, consequently, much more prone to stalemate.
+
+In 1980, the U.S. copyright system was able to develop basic IPR protection for software technology without creating discord within the industry and without upsetting existing IPR bargains in other industries. The 1980 copyright amendments aroused little opposition for the very reason that they did not factionalize the software industry or adversely affect other copyright industries. The potential for blocking the proposal certainly existed: the panel that recommended copyright protection in 1980 was composed partly of representatives from other copyright industries, and throughout its deliberations it consulted a wide range of representatives from commercial interests other than the software and computer industry. 34
+
+However, the political and economic consequences of IPR policy typically become more acute as technologies develop and become more mature. In the early stages of a technology, most developments appear novel and original, and there are few reference points from which to judge developments as either incremental or purely innovative. Consequently, early IPR grants tend to be broad, fairly unrefined, and politically inconsequential. As the technology develops, its dimensions become more fully understood, and the truly novel characteristics of the technology become more well defined. Consequently, IPR rules tend to become more narrow and precise. Ultimately, as the technology becomes mature, what were once innovative products or processes become standardized, and technological change becomes much more incremental. In this context, IPR rules tend to define important boundaries that can have distinctly different consequences for affected industries and firms. In the context of market-share competition, strengthening IPR rules is not a positive sum game; stronger IPR protection may increase the competitiveness of individual firms, but usually at the expense of competing firms and possible the industry as a whole. Consequently, at the very time that market competition increases the natural level of tension between firms in the same industry, the narrowing of IPR disputes increases the potential for adversarial
+
+34. NAT'L COMM. ON NEW TECHOLOGICAL USES OF COPYRIGHTED WORKS, FINAL REPORT (1979).
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+369
+
+relations between competing firms. In the context of market-share competition in a developed technology, IPR rules begin to take on a zero-sum character: firms use IPR claims to protect their competitive advantage and either maintain or increase market share, which typically comes at the expense of immediate competitors. In the context of market-share competition, policy debates over IPR reform shift from considerations of innovation to considerations of competition. The evolution of IPR debates over software technologies follows this pattern closely: recent IPR disputes over computer interfaces, decompilation rights, and the applicability of patents indicate that policy judgments have been shifting away from preserving innovation incentives (as was the concern in 1980) and toward preserving competition by restraining monopoly power.
+
+Despite the considerable ambiguities and uncertainties in current IPR law, neither the software industry nor U.S. legislators have pressed in any concentrated fashion for a coherent form of sui generis protection. Instead, the software industry has debated the terms of copyright protection, while Congress has stood back and watched the courts struggle to secure copyright guidelines for software technology. This situation will likely prevail for some time, although not indefinitely. Software has been accumulating enough unique provisions within copyright law that in many respects it already looks like a sui generis form of protection. 35 Many analysts and activists note that software is getting complex enough that statutory reform may be inevitable. 36 In a sense, the reform of domestic and foreign IPR laws to accommodate software technology has only just begun.
+
+By comparison, IPR reform in biotechnology conforms to the early period of software reform. The cooperative, multilateral style of externalization in the biotechnology case is consistent with the industry's relatively low level of
+
+35. Software has special provisions for fair use, for depositing (developers can mask out up to 40% of deposited programs), for rental rights, and perhaps also for reverse engineering. Moreover, it is the only form of intellectual property that can be covered by both patents and copyrights. Cf. Copyright Act, 17 U.S.C. §§ 101, 117 (1976).
+
+36. A number of members of the IPR legal community agree with this characterization, as do members of the USTR. Officials at the U.S. Copyright Office, on the other hand, argue that copyright law is fully capable of handling software technology. However confident the U.S. Copyright Office, though, unsettled legal conflicts continue to deepen the already profound uncertainty over the scope of copyright protection. For instance, the U.S. Supreme Court recently deadlocked on a crucial and widely followed case pitting Lotus Development Corporation against Borland International Inc.; by issuing no opinion and setting no precedent for future cases, the Court failed to resolve longstanding legal disputes over the copyright status of user interfaces. For a brief account of the case, see Linda Greenhouse, Supreme Court Deadlocks in Key Case on Software, N.Y. TIMES, Jan. 17, 1996, at C2.
+
+---
+
+370
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+commercialization and generally cooperative intra-industry relations. Although biotechnology is highly international in terms of R&D and marketing requirements, it is still a developing technology and consequently confers a relatively low degree of trade leverage. Because international markets are underdeveloped and expanding, IPR disputes have not provoked trade tensions. Moreover, biotechnology is characterized by a symbiotic and generally cooperative relationship between small, highly innovative R&D firms and large multinational enterprises, both of which stand to gain from stronger patent protection for biotechnology products. Their ability to pursue IPR reform has been aided by the lack of a contrary industry coalition, either from the biotechnology field or from unrelated holders of industrial patent rights. Rather, the primary resistance to increased IPR protection for biotechnology products has come from consumer advocates and some public sector agencies who are wary of both the ethical implications of patenting life and the uncertain health and safety implications of new biotechnology products. The tenuous socio-legal response to biotechnology, in the United States as well as abroad, has counterbalanced the industry's relatively consensual preference for expanded patent protection. Moreover, the industry's relatively low level of trade sensitivity has not conveyed any countervailing form of leverage. Collectively, these circumstances have been conducive to a slow, cooperative externalization strategy that relies on the traditional method of multilateral harmonization established by the Paris Convention and administered by the WIPO.
+
+IPR reform in the semiconductor industry represents the other end of the spectrum from biotechnology, in terms of both market structure and domestic politics. Since the mid-1980s, competition in the semiconductor industry has existed within a fully developed market that is dominated by American and Japanese firms, each highly trade sensitive and both focused primarily on the U.S. market. This basic market structure has conveyed an enormous amount of real and rhetorical trade leverage upon the U.S. semiconductor industry. Rhetorically, the extreme nature of U.S.-Japanese competition for market share produced a high degree of trade leverage; indeed, the very need for IPR reform was repeatedly defined in terms of competition with Japanese semiconductor producers. Substantively, the dependence of foreign producers (particularly the Japanese) on the U.S. market gave advocates of sui generis reform reason to expect that coercive bilateralism would work. Since Japanese producers were dependent on selling in the U.S. market, they could be expected to press the Japanese government to pass cognate IPR reforms and consequently obtain
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+371
+
+the reciprocal protection in the U.S. market provided by the SCPA. Overall, in light of the serious commercial implications of IPR protection in the semiconductor industry, Congress eventually considered the coercive bilateral strategy worth the risk of undermining the international IPR regimes and foregoing the certainty of regime-sanctioned harmonization.
+
+A solely market-based interpretation of the semiconductor case is, however, entirely misleading. The semiconductor industry reached a domestic consensus on a sui generis IPR solution only after its preferred option, copyright protection, was repeatedly stalemated by the vociferous and unrelenting opposition of the American Association of Publishers (AAP). The AAP .represented a large and diverse group of industries that uniformly opposed copyright protection for semiconductor mask works, and viewed the proposed terms as a serious breach of fundamental copyright principles that would ultimately undermine their own copyright protection. 37 The coalitional stalemate between the AAP and the semiconductor industry played out over several years and was essentially unresolvable. Eventually, the semiconductor industry abandoned the copyright initiative and pursued an alternative, sui generis solution. The unusual character of the ultimate solution is in no small part a testament to the semiconductor industry's oligopolistic structure and its cohesive representation through the Semiconductor Industry Association, which formed an intra-industry consensus on a sui generis IPR solution and pursued a coherent and politically powerful reform strategy. The serious commercial implications of IPR reform in semiconductors gave the industry sufficient political weight to create an entirely new category of IPR with coercive externalization provisions.
+
+Of course, these are but sketches of very complex political patterns, sufficient to illustrate only the general empirical pattern of the cases. Viewed in their entirety, the cases demonstrate how variances in international competition and trade leverage can interact with the dynamic politics of adversarial regulation to shape domestically and internationally contentious IPR policy choices.
+
+37. The AAP maintained this position throughout the legislative debates that preceded the SCPA of 1984. See Copyright Protection for Semiconductor Chips: Hearings on H.R. 1028 Before the Subcomm. on Courts, Civil Liberties, and the Administration of Justice of the H..R. on the Judiciary, 98th Cong., 1st Sess. (1983); The Semiconductor Chip Protection Act of 1983, supra note 28.
+
+---
+
+372
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+## V. CONCLUSIONS
+
+In analytical terms, the cases surveyed in this article support two general propositions for explaining variances in the strategies the United States has pursued to equilibrate national IPR rules. First, the choice of externalization strategy conforms in part to the market structure of the affected industries. Global industries characterized by zero-sum competition for market share convey a high degree of trade leverage, a necessary (but not sufficient) condition for supporting coercive and bilateral externalization strategies. Industries characterized by positive-sum competition for market creation or market expansion convey considerably less trade leverage, a condition that tends to support more cooperative and multilateral externalization strategies. Second, as technologies mature, domestic coalitional conflicts can rise to the degree that IPR reform becomes much more problematic and prone to stalemate. Under conditions of domestic legal uncertainty and coalitional conflict, U.S. policymakers cannot easily pursue coercive externalization strategies.
+
+These two explanatory propositions are necessarily contingent, as they derive from three complicated cases within one very complicated regulatory arena, and altogether from the perspective of the United States alone. At a minimum, the cases suggest the need for interdisciplinary analytical approaches to studying the complex intersection of domestic regulation and economic globalization. The particular explanatory propositions can be developed, refined, and tested through additional cases of IPR reform, as well as cognate cases in other regulatory areas, such as competition policy, environmental regulation, standards, and the regulation of foreign direct investment.
+
+The cases and general trends in IPR reform discussed in this article also suggest broader observations regarding both the impact of technological change on intellectual property systems and the adequacy of existing institutional mechanisms and political strategies for equilibrating incompatible national IPR systems.
+
+First, technological change clearly has been testing the boundaries of the U.S. and other national IPR regimes. Like most complex legal-regulatory institutions, intellectual property regimes are inherently more static than dynamic; when faced with rapid rates of technological change, coalitional conflict and associated legal and policy uncertainties are likely to emerge. In the United States and elsewhere, IPR systems are being tested most severely
+
+---
+
+1996]
+
+INTELLECTUAL PROPERTY RIGHTS REFORM
+
+373
+
+by rapidly growing industries that create and/or use technologies that do not clearly fit within existing classes of intellectual property. The software industry alone illustrates how new information technologies challenge fundamental intellectual property concepts, such as those of authorship, expression, novelty, and nonobviousness. The continued controversy over the patentability of certain forms of software, along with the failure of the legal system to resolve such vexing copyright problems as user interface protection and decompilation rights, suggests that statutory reform may be inevitable.
+
+In the interim, ongoing legal uncertainty may affect innovation and competition in the software industry or other information-intensive industries, although to an unknown degree. In some industries, the lack of certain intellectual property protection may not significantly affect innovation and competition. For instance, few would argue that the fantastic technological advancements and economic growth in the semiconductor industry derive from the IPR certainty provided by the Semiconductor Chip Protection Act of 1984. In the semiconductor industry, the salience of IPR protection is considerably lower than in other industries due to industry-specific factors such as rapid product cycles (based largely on incremental technological advancements), enormously steep price curves, and the significance of manufacturing capability and know-how to competitive success. The scope and terms of intellectual property protection appear to have a stronger bearing on the software industry; in biotechnology, patent protection is undoubtedly critical. Eventually, deeper legal uncertainty may have a deleterious effect on innovation and/or competition in these and other industries.
+
+Second, much like domestic intellectual property regimes, the institutional structure of the international IPR regimes reduces their capacity to harmonize divergent national IPR responses to rapidly changing technologies. The international IPR regimes may become increasingly irrelevant to important realms of IPR law, as the widening gaps between national IPR regimes outstrip the gradual harmonization processes fostered by the practice of national treatment. Integrating international IPR issues with trade concepts and institutions, such as the TRIPs agreement in the GATT/WTO, is unlikely to speed international harmonization processes substantially: adjustment processes are still very slow, dispute resolution mechanisms remain weak, and the system overall is far less able to accommodate rapid rates of technological change in new information industries than are most domestic IPR regimes. In this context, bilateral and plurilateral trade-based strategies provide the only
+
+---
+
+374
+
+GLOBAL LEGAL STUDIES JOURNAL
+
+[Vol. 3:341
+
+alternative routes toward resolving particular discrepancies between national IPR regimes.
+
+Ultimately, bilateral trade-based strategies may not be effective. For instance, although bilateral U.S. pressure has led to significant de jure IPR reform in many countries, effective and enforceable de facto IPR reform often remains elusive. Such limits to policy coordination and behavioral reform do not augur well for international policy harmonization in other, even more contentious regulatory arenas, such as labor and health standards, environmental regulations, and competition policy. Among the major industrialized countries alone, policymakers are divided and uncertain about how to harmonize national regulatory approaches to competition policy and foreign direct investment. Policy discussions of these issues at the OECD level remain largely formative. 38 The fact that national regulatory regimes are far more similar among the OECD member nations than between the OECD countries and the rest of the world suggests that economic globalization is unlikely to foster an equivalent globalization of law and regulation.
+
+38. See, e.g., OECD, NEW DIMENSIONS OF MARKET ACCESS IN A GLOBALISING WORLD ECONOMY (1995).
+

processed/00214_W2119474366/document.md

@@ -0,0 +1,264 @@
+Scholarship Repository
+
+University of Minnesota Law School
+
+Articles
+
+Faculty Scholarship
+
+2014
+
+# FTC v. Actavis, Inc.: When Is the Rule of Reason Not the Rule of Reason?
+
+Thomas F. Cotter University of Minnesota Law School, cotte034@umn.edu
+
+Follow this and additional works at: https://scholarship.law.umn.edu/faculty_articles
+
+![Figure](figures/figure_007.png)
+
+Part of the Law Commons
+
+## Recommended Citation
+
+Thomas F. Cotter, FTC v. Actavis, Inc.: When Is the Rule of Reason Not the Rule of Reason?, 15 Minn. J.L. Sci. & Tech. 41 (2014), available at https://scholarship.law.umn.edu/faculty_articles/148.
+
+This Article is brought to you for free and open access by the University of Minnesota Law School. It has been accepted for inclusion in the Faculty Scholarship collection by an authorized administrator of the Scholarship Repository.
+
+---
+
+## FTC v. Actavis, Inc.: When Is the Rule of Reason Not the Rule of Reason?
+
+Thomas F. Cotter*
+
+When is the lion neither inside nor outside the den?$^{1}$
+
+The U.S. Supreme Court's recent decision in FTC v. Actavis, Inc. 2 brings some resolution to the decade-long dispute over the level of antitrust scrutiny that is appropriate for evaluating the legality of “ reverse payment ” or “ pay-for-delay ” agreements settling pharmaceutical patent infringement litigation between brand-name and generic drug companies. I have written at length about this topic before and need not devote time and space to rehashing the facts or the arguments in favor of various proposed approaches. $^3$ Suffice to say that in the past I argued against the Eleventh Circuit's “ scope-of-the-
+
+c 2014 Thomas F. Cotter
+
+* Briggs and Morgan Professor of Law, University of Minnesota Law School.
+
+1. Whalen Lai, Koan, in Encyclopedia of Asian Philosophy 287, 288 (Oliver Leaman ed., 2001) (attributing the saying to Mazu Daeyi).
+
+2. 133 S. Ct. 2223 (2013).
+
+3. See Roger D. Blair & Thomas F. Cotter, Are Settlements of Patent Disputes Illegal Per Se?, 47 Antitrust Bull. 491 (2002); Thomas F. Cotter, Antitrust Implications of Patent Settlements Involving Reverse Payments: Defending a Rebuttable Presumption of Illegality in Light of Some Recent Scholarship, 71 Antitrust L.J. 1069 (2004); Thomas F. Cotter, Refining the “Presumptive Illegality” Approach to Settlements of Patent Disputes Involving Reverse Payments: A Commentary on Hovenkamp, Janis & Lemley , 87 Minn. L. Rev. 1789 (2003) [hereinafter Cotter, Presumptive Illegality]; see also Thomas F. Cotter, FTC v. Actavis: An Analysis, IntellectualIP.com (Feb. 25, 2013) http://intellectualip.com/2013/02/25/ftc-v-actavis-an-analysis/ [hereinafter Cotter, FTC v. Actavis: An Analysis]; Thomas F. Cotter, FTC v. Actavis, Case Law, IntellectualIP.com (Feb. 22, 2013), http://intellectualip.com/2013/02/22/ftc-v-actavis-case-law/ [hereinafter Cotter, FTC v. Actavis, Case Law]; Thomas F. Cotter, FTC v. Actavis: The HatchWaxman Framework, IntellectualIP.com (Feb. 20, 2013), http://intellectualip.com/2013/02/20/ftc-v-actavis-the-hatch-waxman-
+
+framework/; Thomas F. Cotter, FTC v. Actavis : Reverse Payments, IntellectualIP.com (Feb. 21, 2013), http://intellectualip.com/2013/02/21/ftcv-actavis-reverse-payments/.
+
+41
+
+---
+
+42
+
+MINN. J. L. SCI. & TECH.
+
+[Vol. 15:1
+
+patent" test, under which the agreement would be legal as long as the terms fell within the exclusionary potential of the patent, and the infringement action was not merely a sham or fraud. $^4$ Instead, I argued for a “ presumptive illegality ” approach, under which proof that a brand-name company paid a generic company to settle would shift the burden to the settling parties to rebut the presumption of illegality. $^5$ In this regard, I proposed that the most important factor in determining whether the settling parties have rebutted the presumption should be
+
+the amount of consideration flowing from the brand-name to the generic firm. Where that amount is less than the amount of the patent owner's expected litigation costs, this fact alone may be sufficient to rebut the presumption, and thus shift to the antitrust plaintiff the burden of proving that the anticompetitive harm outweighs the procompetitive benefit of the settlement. Under these circumstances, the payment may represent nothing more than a good-faith effort to avoid litigation costs . . . . Other relevant evidence may include the presence of other agreements between the settling parties (for example, authorizing the defendant to market an authorized generic drug, or licensing the defendant other intellectual property rights), which should be taken into account for the limited purpose of accurately estimating the value of the consideration flowing from plaintiff to defendant; whether the generic is “ cash-strapped, ” and therefore willing to accept a later entry date to remain in business; whether the patent owner sought, and succeeded in obtaining, a preliminary injunction against the generic manufacturer; whether the generic manufacturer agrees to waive its 180-day exclusivity, thus removing the risk of a bottleneck potentially blocking other ANDA applicants; and whether the patent in suit has withstood other validity challenges arising after the filing of the settled action. On the other hand, where the amount of consideration flowing from patent owner to generic manufacturer exceeds the generic firm's expected profit from the sale of the generic drug in question, the inference that the patent owner is simply paying a potential competitor to exit the market is much stronger, and the presumption of illegality should be very difficult to rebut. Moreover, although it probably would not be advisable to require the factfinder to estimate the ex ante probability that the patent would have been found valid and infringed had the infringement action not been settled — a matter that courts in some of the reverse payment cases understandably have been reluctant to undertake — all that the proposed approach requires is for courts to draw appropriate inferences from the amount of the settlement in
+
+4. See Cotter, FTC v. Actavis, Case Law, supra note 3.
+
+5. See, e.g., Blair & Cotter, supra note 3, at 534-35.
+
+---
+
+2014]
+
+NOT THE RULE OF REASON?
+
+43
+
+comparison to other expected costs and benefits, along with any other relevant facts and circumstances. $^6$
+
+Writing for a 5-3 majority in Actavis, Justice Breyer rejected both the scope-of-the-patent test and the presumptive illegality approach and held instead that courts should review reverse payment settlements under the rule of reason. $^7$ Or so the opinion states. In reality, the Court appears to have all but in name adopted the presumptive illegality approach it purported to reject. One might speculate about the political or prudential considerations that went into the majority's characterization of what it was actually doing, but as I read the opinion, reverse payment settlements of the type at issue in Actavis are now subject to a de facto regime of presumptive illegality. In my view, this is a welcome result.
+
+The reason I characterize the majority holding as adopting a de facto rule of presumptive illegality is that, as antitrust lawyers are well aware, in practice the rule of reason is hardly the sort of open-ended, totality-of-the-circumstances approach suggested by Justice Brandeis's classic definition of the rule of reason in the old Board of Trade case. $^8$ Rather, courts tend to apply a structured version of the rule of reason, $^9$ which Professor Hovenkamp nicely summarizes in his hornbook $^10$ and
+
+6. Cotter, FTC v. Actavis: An Analysis, supra note 3 (citing Michael A. Carrier, Innovation for the 21st Century: Harnessing the Power of Intellectual Property and Antitrust Law 378–82 (2009); Cotter, Presumptive Illegality , supra note 3, at 1812–15).
+
+7. FTC v. Actavis, Inc., 133 S. Ct. 2223, 2230-38 (2013).
+
+8. See Bd. of Trade of Chi. v. United States, 246 U.S. 231, 238 (1918). Justice Brandeis wrote for the Court:
+
+The true test of legality is whether the restraint imposed is such as merely regulates and perhaps thereby promotes competition or whether it is such as may suppress or even destroy competition. To determine that question the court must ordinarily consider the facts peculiar to the business to which the restraint is applied; its condition before and after the restraint was imposed; the nature of the restraint and its effect, actual or probable. The history of the restraint, the evil believed to exist, the reason for adopting the particular remedy, the purpose or end sought to be attained, are all relevant facts. This is not because a good intention will save an otherwise objectionable regulation or the reverse; but because knowledge of intent may help the court to interpret facts and to predict consequences.
+
+Id.
+
+9. See, e.g., Polygram Holding, Inc. v. FTC, 416 F.3d 29, 33-38 (D.C. Cir. 2005).
+
+10. Herbert Hovenkamp, Federal Antitrust Policy: The Law of Competition and Its Practice 279–80 (4th ed. 2011).
+
+---
+
+44
+
+MINN. J. L. SCI. & TECH.
+
+[Vol. 15:1
+
+which I paraphrase in the following manner in my own antitrust classes:
+
+1. Consider first whether there is a "contract, combination, or conspiracy" that restrains trade (in some sense).
+
+If yes (conscious parallelism coupled with plus factors?), go on.
+
+If not, § 1 doesn't apply (the "§ 1 gap").
+
+2. If necessary, consider next whether the restraint poses any possible risk to competition. I.e., is the restraint at issue one that poses a substantial risk of increasing price, lowering quantity, or causing some other anticompetitive harm?
+
+If yes, go on.
+
+If no, stop; judgment for defendant.
+
+3. If necessary, consider next whether the restraint is likely to generate any plausible, cognizable, procompetitive benefits.
+
+For example, does the restraint plausibly relate to the core activities of a lawful joint venture?
+
+Is it plausibly ancillary in the sense of being reasonably necessary to promote the legitimate activities of a joint undertaking? Reasonably necessary for the provision of some good or service that consumers demand but which might not be provided optimally if each competitor merely followed its own individual self-interest?
+
+If yes, go on.
+
+If no, stop; it is a naked restraint of trade, likely only to increase price or reduce output or quality, and is per se illegal.
+
+4. If necessary, consider next whether the defendant has market power (e.g., through substantial market share coupled with barriers to entry), or alternatively whether there is proof of actual anticompetitive effects, such as a reduction of output.
+
+If yes, go on.
+
+If no, stop; judgment for defendant.
+
+5. If necessary, consider next whether the restraint at issue provides actual (not just plausible) procompetitive benefits.
+
+If yes, go on.
+
+If no, stop; judgment for plaintiff.
+
+6. If necessary, consider next whether the restraint is the least restrictive means of attaining those benefits.
+
+If yes, go on.
+
+If no, stop; judgment for plaintiff.
+
+7. If necessary, balance the procompetitive benefits against the anticompetitive costs (good luck!). $^11$
+
+11. Thomas F. Cotter, PowerPoint Presentation, Antitrust Overview (unpublished document) (on file with author) [hereinafter Cotter PowerPoint];
+
+---
+
+2014]
+
+NOT THE RULE OF REASON?
+
+45
+
+Assuming that this analysis is correct, what exactly will courts be doing when they apply the rule of reason approach to pay-for-delay cases such as Actavis ? Will they be starting from step 1 above? No, because in any case in which a patentee agrees to pay money to an alleged infringer in return for the latter's agreement to settle the case and temporarily exit the market there is obviously a contract that potentially restrains trade; that much is indisputable. Equally obvious is the potential risk to competition (step 2). At the same time, there are potential procompetitive benefits (step 3), because (as a general matter) settlement conserves social resources that otherwise would be devoted to litigation and (in this specific context) in theory the settlement could speed up the entry of generic drugs to the market. $^12$
+
+Crucially, according to the majority, step 4 above is also likely to be present in the context of pay-for-delay settlements. As Justice Breyer wrote:
+
+First , the specific restraint at issue has the “ potential for genuine adverse effects on competition. ” The payment in effect amounts to a purchase by the patentee of the exclusive right to sell its product, a right it already claims but would lose if the patent litigation were to continue and the patent were held invalid or not infringed by the generic product. Suppose, for example, that the exclusive right to sell produces $ 50 million in supracompetitive profits per year for the patentee. And suppose further that the patent has 10 more years to run. Continued litigation, if it results in patent invalidation or a finding of noninfringement, could cost the patentee $ 500 million in lost revenues, a sum that then would flow in large part to consumers in the form of lower prices.
+
+....
+
+Second, these anticompetitive consequences will at least sometimes prove unjustified . . . .
+
+Third , where a reverse payment threatens to work unjustified anticompetitive harm, the patentee likely possesses the power to bring that harm about in practice. At least, the “size of the payment from a branded drug manufacturer to a prospective generic is itself a strong indicator of power”—namely, the power to charge prices higher than the competitive level. An important patent itself helps to assure such power. Neither is a firm without that power likely to pay “large sums” to induce “others to stay out of its market.” In any
+
+see also Thomas F. Cotter, Patent Holdup, Patent Remedies, and Antitrust Responses, 34 J. CORP. L. 1151, 1205-06 (2009).
+
+12. See Actavis, 133 S. Ct. at 2234-35.
+
+---
+
+46
+
+MINN. J. L. SCI. & TECH.
+
+[Vol. 15:1
+
+event, the Commission has referred to studies showing that reverse payment agreements are associated with the presence of higherthan-competitive profits — a strong indication of market power. $^13$
+
+If we have made it all the way through steps 1 through 4, what remains? Step 5 asks (in my formulation) “ whether the restraint at issue provides actual (not just plausible) procompetitive benefits. ” 14 Importantly, the burden of proof on step 5 normally rests with the defendant . 15 So if, under the Court's own analysis, steps 1 through 4 are satisfied in the typical pay-for-delay case and review really kicks in at step 5 — at which point the defendant has the burden of coming forward with exonerating evidence — it is a little hard to see how that framework differs in any functional manner from presumptive illegality. 16
+
+13. Id. at 2234-36 (citations omitted).
+
+14. See Cotter PowerPoint, supra note 11. Professor Hovenkamp states this step somewhat more forcefully, namely whether there is “ strong evidence that the challenged practice creates substantial efficiencies by reducing participants' costs or improving product or service quality. ” Hovenkamp , supra note 10, at 280.
+
+15. See, e.g., Polygram Holding, Inc. v. FTC, 416 F.3d 29, 36 (D.C. Cir. 2005) (“[T]he evidentiary burden shifts to the defendant to show the restraint in fact does not harm consumers or has ‘procompetitive virtues’ that outweigh its burden upon consumers.”); Schering-Plough Corp. v. FTC, 402 F.3d 1056, 1065 (11th Cir. 2005) (stating that “[o]nce the plaintiff meets the burden of producing sufficient evidence of market power, the burden then shifts to the defendant to show that the challenged conduct promotes a sufficiently procompetitive objective”); Clorox Co. v. Sterling Winthrop, Inc., 117 F.3d 50, 56 (2d Cir. 1997) (“[I]f the plaintiff succeeds [in showing an actual anticompetitive effect], the burden shifts to the defendant to establish the “pro-competitive ‘redeeming virtues” of the action. Should the defendant carry this burden, the plaintiff must then show that the same pro-competitive effect could be achieved through an alternative means that is less restrictive of competition.”).
+
+16. See Polygram Holding, 416 F.3d at 36. The Polygram court stated:
+
+For reasons we have already explained, we reject PolyGram's attempt to locate the appropriate analysis, and the concomitant burden of proof, by reference to the vestigial line separating per se analysis from the rule of reason. See Areeda & Hovenkamp, Antitrust Law, ¶ 1511a (“judges and litigants too often assume erroneously that the classification, per se or rule of reason, necessarily determines what must or may be alleged and proved, made the subject of detailed findings, or submitted to the jury”). At bottom, the Sherman Act requires the court to ascertain whether the challenged restraint hinders competition; the Commission's framework, at least as the Commission applied it in this case, does just that.
+
+We therefore accept the Commission's analytical framework. If, based upon economic learning and the experience of the market, it is obvious that a restraint of trade likely impairs competition, then the
+
+---
+
+2014]
+
+NOT THE RULE OF REASON?
+
+47
+
+This is particularly so given the Court's further statements that “it is normally not necessary to litigate patent validity to answer the antitrust question” and its discussion of the type of procompetitive justifications that might excuse a reverse payment. $^17$ As for the first issue, Justice Breyer wrote:
+
+An unexplained large reverse payment itself would normally suggest that the patentee has serious doubts about the patent's survival. And that fact, in turn, suggests that the payment's objective is to maintain supracompetitive prices to be shared among the patentee and the challenger rather than face what might have been a competitive market — the very anticompetitive consequence that underlies the claim of antitrust unlawfulness. The owner of a particularly valuable patent might contend, of course, that even a small risk of invalidity justifies a large payment. But, be that as it may, the payment (if otherwise unexplained) likely seeks to prevent the risk of competition. And, as we have said, that consequence constitutes the relevant anticompetitive harm. In a word, the size of the unexplained reverse payment can provide a workable surrogate for a patent's weakness, all without forcing a court to conduct a detailed exploration of the validity of the patent itself. $^18$
+
+In other words, the plaintiff is not going to have to prove, except inferentially by reference to the amount of the payment, that the probability of patent invalidity was high. As for the second, the Court noted that reverse payments
+
+may amount to no more than a rough approximation of the litigation expenses saved through the settlement. That payment may reflect compensation for other services that the generic has promised to perform — such as distributing the patented item or helping to develop a market for that item. There may be other justifications. Where a reverse payment reflects traditional settlement considerations, such as avoided litigation costs or fair value for services, there is not the same concern that a patentee is using its monopoly profits to avoid the risk of patent invalidation or a finding of noninfringement. In such cases, the parties may have provided for a reverse payment without having sought or brought about the anticompetitive consequences we mentioned above. $^19$
+
+All of this leads me to conclude that the reasons the Court chose not to characterize what it was doing as a presumptive illegality approach were either (1) political, e.g., to keep one or
+
+restraint is presumed unlawful and, in order to avoid liability, the defendant must either identify some reason the restraint is unlikely to harm consumers or identify some competitive benefit that plausibly offsets the apparent or anticipated harm.
+
+Id.
+
+17. See Actavis, 133 S. Ct. at 2236.
+
+18. Id. at 2236-37.
+
+19. Id. at 2236.
+
+---
+
+48
+
+MINN. J. L. SCI. & TECH.
+
+[Vol. 15:1
+
+more possibly gun-shy justices on board with the majority, on the theory that a rule of reason approach is not quite as proplaintiff as a presumptive illegality approach; or (2) based on concerns that courts might construe the adoption of a presumptive illegality approach in the present case as effectively holding that such an approach is appropriate in other cases, not arising in the byzantine shadow of HatchWaxman. The concluding section of the majority opinion seems to reflect this latter concern, 20 and thus may be viewed as a
+
+20. See id. at 2237-38. Justice Breyer concluded:
+
+[T]he likelihood of a reverse payment bringing about anticompetitive effects depends upon its size, its scale in relation to the payor's anticipated future litigation costs, its independence from other services for which it might represent payment, and the lack of any other convincing justification. The existence and degree of any anticompetitive consequence may also vary as among industries. These complexities lead us to conclude that the FTC must prove its case as in other rule-of-reason cases.
+
+To say this is not to require the courts to insist, contrary to what we have said, that the Commission need litigate the patent's validity, empirically demonstrate the virtues or vices of the patent system, present every possible supporting fact or refute every possible prodefense theory. As a leading antitrust scholar has pointed out, “ [t]here is always something of a sliding scale in appraising reasonableness, ” and as such “ the quality of proof required should vary with the circumstances. ”
+
+As in other areas of law, trial courts can structure antitrust litigation so as to avoid, on the one hand, the use of antitrust theories too abbreviated to permit proper analysis, and, on the other, consideration of every possible fact or theory irrespective of the minimal light it may shed on the basic question — that of the presence of significant unjustified anticompetitive consequences. We therefore leave to the lower courts the structuring of the present rule-of-reason antitrust litigation.
+
+Id. (citations omitted). I would expect that settlements of patent infringement litigation outside of the Hatch-Waxman context will rarely give rise to plausible antitrust claims under the rule of reason. In a typical case, settlement may increase output (e.g., by resulting in a nonexclusive license of a patent that has withstood a validity challenge), thus failing step 2 above; or the patentee lacks market power (step 4); or courts will conclude, as a general rule, that a settlement lacking any red flags such as the presence of a reverse payment in excess of the defendant's expected profit (or other suspicious conditions) necessarily has procompetitive benefits that outweigh any anticompetitive consequences that could be proven without unraveling the reduction of adjudicative costs that is the primary social good that settlement produces (and thus will not countenance attempts to question patent validity or infringement absent good reason). But I tend to agree with the majority that patent settlements should not be effectively immune from antitrust scrutiny absent conduct such as sham or fraud, which arguably was the implication of the scope-of-the-patent test.
+
+---
+
+2014]
+
+NOT THE RULE OF REASON?
+
+49
+
+rejoinder to the dissent's concern that the majority approach renders vulnerable even conventional patent settlements. $^21$
+
+In conclusion, it seems to me that the majority adopted a (de facto) presumptive illegality approach to pay-for-delay settlements entered into in the shadow of Hatch-Waxman — precisely the approach that many of us were hoping for $^22$ — even if, for political or prudential reasons, it suggests that it did not. As long as the lower courts correctly interpret the message, this seems an acceptable resolution to the pay-for-delay problem.
+
+21. See Actavis, 133 S. Ct. at 2244-45 (Roberts, C.J., dissenting).
+
+22. See Brief Amici Curiae of 118 Law, Economics, and Business Professors and the American Antitrust Institute in Support of Petitioners, FTC v. Actavis, Inc., 133 S. Ct. 2233 (2013) (No. 12-416), 2013 WL 391001.
+
+---
+
+**
+

processed/00214_W2119474366/metadata.json

@@ -0,0 +1,1394 @@
+{
+  "source_file": "/tmp/tmpdphfuibl/00214_W2119474366_FTC_v._Actavis,_Inc._When_Is_the_Rule_of_Reason_Not_the_Rule_of_Reason.pdf",
+  "total_pages": 11,
+  "pages": [
+    {
+      "page_number": 1,
+      "elements": [
+        {
+          "bbox": [
+            79,
+            126,
+            281,
+            163
+          ],
+          "label": "header",
+          "reading_order": 0,
+          "text": "Scholarship Repository\n\nUniversity of Minnesota Law School"
+        },
+        {
+          "bbox": [
+            79,
+            187,
+            126,
+            204
+          ],
+          "label": "header",
+          "reading_order": 1,
+          "text": "Articles"
+        },
+        {
+          "bbox": [
+            500,
+            187,
+            613,
+            205
+          ],
+          "label": "header",
+          "reading_order": 2,
+          "text": "Faculty Scholarship"
+        },
+        {
+          "bbox": [
+            79,
+            228,
+            112,
+            246
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "2014"
+        },
+        {
+          "bbox": [
+            78,
+            261,
+            591,
+            311
+          ],
+          "label": "sec_0",
+          "reading_order": 4,
+          "text": "FTC v. Actavis, Inc.: When Is the Rule of Reason Not the Rule of\nReason?"
+        },
+        {
+          "bbox": [
+            78,
+            325,
+            368,
+            358
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "Thomas F. Cotter\nUniversity of Minnesota Law School, cotte034@umn.edu"
+        },
+        {
+          "bbox": [
+            79,
+            633,
+            507,
+            649
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "Follow this and additional works at: https://scholarship.law.umn.edu/faculty_articles"
+        },
+        {
+          "bbox": [
+            79,
+            650,
+            98,
+            670
+          ],
+          "label": "fig",
+          "reading_order": 7,
+          "text": "![Figure](figures/figure_007.png)",
+          "figure_path": "/tmp/pdf_output/markdown/figures/figure_007.png"
+        },
+        {
+          "bbox": [
+            99,
+            657,
+            233,
+            671
+          ],
+          "label": "para",
+          "reading_order": 8,
+          "text": "Part of the Law Commons"
+        },
+        {
+          "bbox": [
+            79,
+            700,
+            214,
+            717
+          ],
+          "label": "sec_1",
+          "reading_order": 9,
+          "text": "Recommended Citation"
+        },
+        {
+          "bbox": [
+            79,
+            718,
+            604,
+            749
+          ],
+          "label": "para",
+          "reading_order": 10,
+          "text": "Thomas F. Cotter, FTC v. Actavis, Inc.: When Is the Rule of Reason Not the Rule of Reason?, 15 Minn. J.L.\nSci. & Tech. 41 (2014), available at https://scholarship.law.umn.edu/faculty_articles/148."
+        },
+        {
+          "bbox": [
+            79,
+            775,
+            578,
+            816
+          ],
+          "label": "para",
+          "reading_order": 11,
+          "text": "This Article is brought to you for free and open access by the University of Minnesota Law School. It has been\naccepted for inclusion in the Faculty Scholarship collection by an authorized administrator of the Scholarship\nRepository."
+        }
+      ]
+    },
+    {
+      "page_number": 2,
+      "elements": [
+        {
+          "bbox": [
+            99,
+            130,
+            487,
+            177
+          ],
+          "label": "sec_1",
+          "reading_order": 0,
+          "text": "FTC v. Actavis, Inc.: When Is the Rule of\nReason Not the Rule of Reason?"
+        },
+        {
+          "bbox": [
+            101,
+            194,
+            257,
+            215
+          ],
+          "label": "para",
+          "reading_order": 1,
+          "text": "Thomas F. Cotter*"
+        },
+        {
+          "bbox": [
+            130,
+            227,
+            462,
+            247
+          ],
+          "label": "para",
+          "reading_order": 2,
+          "text": "When is the lion neither inside nor outside the den?$^{1}$"
+        },
+        {
+          "bbox": [
+            101,
+            262,
+            513,
+            423
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "The U.S. Supreme Court's recent decision in FTC v.\nActavis, Inc. 2 brings some resolution to the decade-long dispute\nover the level of antitrust scrutiny that is appropriate for\nevaluating the legality of “ reverse payment ” or “ pay-for-delay ”\nagreements settling pharmaceutical patent infringement\nlitigation between brand-name and generic drug companies. I\nhave written at length about this topic before and need not\ndevote time and space to rehashing the facts or the arguments\nin favor of various proposed approaches. $^3$ Suffice to say that in\nthe past I argued against the Eleventh Circuit's “ scope-of-the-"
+        },
+        {
+          "bbox": [
+            101,
+            457,
+            236,
+            472
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "c 2014 Thomas F. Cotter"
+        },
+        {
+          "bbox": [
+            101,
+            472,
+            512,
+            498
+          ],
+          "label": "fnote",
+          "reading_order": 5,
+          "text": "* Briggs and Morgan Professor of Law, University of Minnesota Law\nSchool."
+        },
+        {
+          "bbox": [
+            101,
+            498,
+            511,
+            525
+          ],
+          "label": "fnote",
+          "reading_order": 6,
+          "text": "1. Whalen Lai, Koan, in Encyclopedia of Asian Philosophy 287, 288\n(Oliver Leaman ed., 2001) (attributing the saying to Mazu Daeyi)."
+        },
+        {
+          "bbox": [
+            125,
+            525,
+            264,
+            539
+          ],
+          "label": "fnote",
+          "reading_order": 7,
+          "text": "2. 133 S. Ct. 2223 (2013)."
+        },
+        {
+          "bbox": [
+            101,
+            539,
+            513,
+            739
+          ],
+          "label": "fnote",
+          "reading_order": 8,
+          "text": "3. See Roger D. Blair & Thomas F. Cotter, Are Settlements of Patent\nDisputes Illegal Per Se?, 47 Antitrust Bull. 491 (2002); Thomas F. Cotter,\nAntitrust Implications of Patent Settlements Involving Reverse Payments:\nDefending a Rebuttable Presumption of Illegality in Light of Some Recent\nScholarship, 71 Antitrust L.J. 1069 (2004); Thomas F. Cotter, Refining the\n“Presumptive Illegality” Approach to Settlements of Patent Disputes Involving\nReverse Payments: A Commentary on Hovenkamp, Janis & Lemley , 87 Minn.\nL. Rev. 1789 (2003) [hereinafter Cotter, Presumptive Illegality]; see also\nThomas F. Cotter, FTC v. Actavis: An Analysis, IntellectualIP.com (Feb.\n25, 2013) http://intellectualip.com/2013/02/25/ftc-v-actavis-an-analysis/\n[hereinafter Cotter, FTC v. Actavis: An Analysis]; Thomas F. Cotter, FTC v.\nActavis, Case Law, IntellectualIP.com (Feb. 22, 2013),\nhttp://intellectualip.com/2013/02/22/ftc-v-actavis-case-law/ [hereinafter Cotter,\nFTC v. Actavis, Case Law]; Thomas F. Cotter, FTC v. Actavis: The Hatch-\nWaxman Framework, IntellectualIP.com (Feb. 20, 2013),\nhttp://intellectualip.com/2013/02/20/ftc-v-actavis-the-hatch-waxman-"
+        },
+        {
+          "bbox": [
+            101,
+            739,
+            511,
+            777
+          ],
+          "label": "fnote",
+          "reading_order": 9,
+          "text": "framework/; Thomas F. Cotter, FTC v. Actavis : Reverse Payments,\nIntellectualIP.com (Feb. 21, 2013), http://intellectualip.com/2013/02/21/ftc-\nv-actavis-reverse-payments/."
+        },
+        {
+          "bbox": [
+            297,
+            786,
+            316,
+            802
+          ],
+          "label": "foot",
+          "reading_order": 10,
+          "text": "41"
+        }
+      ]
+    },
+    {
+      "page_number": 3,
+      "elements": [
+        {
+          "bbox": [
+            101,
+            68,
+            121,
+            84
+          ],
+          "label": "header",
+          "reading_order": 0,
+          "text": "42"
+        },
+        {
+          "bbox": [
+            219,
+            66,
+            388,
+            85
+          ],
+          "label": "header",
+          "reading_order": 1,
+          "text": "MINN. J. L. SCI. & TECH."
+        },
+        {
+          "bbox": [
+            451,
+            67,
+            512,
+            85
+          ],
+          "label": "header",
+          "reading_order": 2,
+          "text": "[Vol. 15:1"
+        },
+        {
+          "bbox": [
+            100,
+            99,
+            513,
+            258
+          ],
+          "label": "half_para",
+          "reading_order": 3,
+          "text": "patent\" test, under which the agreement would be legal as long\nas the terms fell within the exclusionary potential of the\npatent, and the infringement action was not merely a sham or\nfraud. $^4$ Instead, I argued for a “ presumptive illegality ”\napproach, under which proof that a brand-name company paid\na generic company to settle would shift the burden to the\nsettling parties to rebut the presumption of illegality. $^5$ In this\nregard, I proposed that the most important factor in\ndetermining whether the settling parties have rebutted the\npresumption should be"
+        },
+        {
+          "bbox": [
+            130,
+            258,
+            491,
+            707
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "the amount of consideration flowing from the brand-name to the\ngeneric firm. Where that amount is less than the amount of the\npatent owner's expected litigation costs, this fact alone may be\nsufficient to rebut the presumption, and thus shift to the antitrust\nplaintiff the burden of proving that the anticompetitive harm\noutweighs the procompetitive benefit of the settlement. Under these\ncircumstances, the payment may represent nothing more than a\ngood-faith effort to avoid litigation costs . . . . Other relevant\nevidence may include the presence of other agreements between the\nsettling parties (for example, authorizing the defendant to market\nan authorized generic drug, or licensing the defendant other\nintellectual property rights), which should be taken into account for\nthe limited purpose of accurately estimating the value of the\nconsideration flowing from plaintiff to defendant; whether the\ngeneric is “ cash-strapped, ” and therefore willing to accept a later\nentry date to remain in business; whether the patent owner sought,\nand succeeded in obtaining, a preliminary injunction against the\ngeneric manufacturer; whether the generic manufacturer agrees to\nwaive its 180-day exclusivity, thus removing the risk of a bottleneck\npotentially blocking other ANDA applicants; and whether the patent\nin suit has withstood other validity challenges arising after the\nfiling of the settled action. On the other hand, where the amount of\nconsideration flowing from patent owner to generic manufacturer\nexceeds the generic firm's expected profit from the sale of the\ngeneric drug in question, the inference that the patent owner is\nsimply paying a potential competitor to exit the market is much\nstronger, and the presumption of illegality should be very difficult to\nrebut. Moreover, although it probably would not be advisable to\nrequire the factfinder to estimate the ex ante probability that the\npatent would have been found valid and infringed had the\ninfringement action not been settled — a matter that courts in some\nof the reverse payment cases understandably have been reluctant to\nundertake — all that the proposed approach requires is for courts to\ndraw appropriate inferences from the amount of the settlement in"
+        },
+        {
+          "bbox": [
+            124,
+            753,
+            411,
+            768
+          ],
+          "label": "fnote",
+          "reading_order": 5,
+          "text": "4. See Cotter, FTC v. Actavis, Case Law, supra note 3."
+        },
+        {
+          "bbox": [
+            124,
+            768,
+            398,
+            781
+          ],
+          "label": "fnote",
+          "reading_order": 6,
+          "text": "5. See, e.g., Blair & Cotter, supra note 3, at 534-35."
+        }
+      ]
+    },
+    {
+      "page_number": 4,
+      "elements": [
+        {
+          "bbox": [
+            101,
+            67,
+            140,
+            85
+          ],
+          "label": "header",
+          "reading_order": 0,
+          "text": "2014]"
+        },
+        {
+          "bbox": [
+            204,
+            66,
+            405,
+            85
+          ],
+          "label": "header",
+          "reading_order": 1,
+          "text": "NOT THE RULE OF REASON?"
+        },
+        {
+          "bbox": [
+            493,
+            68,
+            513,
+            84
+          ],
+          "label": "header",
+          "reading_order": 2,
+          "text": "43"
+        },
+        {
+          "bbox": [
+            130,
+            98,
+            490,
+            126
+          ],
+          "label": "half_para",
+          "reading_order": 3,
+          "text": "comparison to other expected costs and benefits, along with any\nother relevant facts and circumstances. $^6$"
+        },
+        {
+          "bbox": [
+            101,
+            126,
+            513,
+            317
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "Writing for a 5-3 majority in Actavis, Justice Breyer\nrejected both the scope-of-the-patent test and the presumptive\nillegality approach and held instead that courts should review\nreverse payment settlements under the rule of reason. $^7$ Or so\nthe opinion states. In reality, the Court appears to have all but\nin name adopted the presumptive illegality approach it\npurported to reject. One might speculate about the political or\nprudential considerations that went into the majority's\ncharacterization of what it was actually doing, but as I read the\nopinion, reverse payment settlements of the type at issue in\nActavis are now subject to a de facto regime of presumptive\nillegality. In my view, this is a welcome result."
+        },
+        {
+          "bbox": [
+            101,
+            317,
+            513,
+            445
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "The reason I characterize the majority holding as adopting\na de facto rule of presumptive illegality is that, as antitrust\nlawyers are well aware, in practice the rule of reason is hardly\nthe sort of open-ended, totality-of-the-circumstances approach\nsuggested by Justice Brandeis's classic definition of the rule of\nreason in the old Board of Trade case. $^8$ Rather, courts tend to\napply a structured version of the rule of reason, $^9$ which\nProfessor Hovenkamp nicely summarizes in his hornbook $^10$ and"
+        },
+        {
+          "bbox": [
+            101,
+            480,
+            511,
+            532
+          ],
+          "label": "fnote",
+          "reading_order": 6,
+          "text": "6. Cotter, FTC v. Actavis: An Analysis, supra note 3 (citing Michael A.\nCarrier, Innovation for the 21st Century: Harnessing the Power of\nIntellectual Property and Antitrust Law 378–82 (2009); Cotter,\nPresumptive Illegality , supra note 3, at 1812–15)."
+        },
+        {
+          "bbox": [
+            124,
+            532,
+            421,
+            546
+          ],
+          "label": "fnote",
+          "reading_order": 7,
+          "text": "7. FTC v. Actavis, Inc., 133 S. Ct. 2223, 2230-38 (2013)."
+        },
+        {
+          "bbox": [
+            101,
+            546,
+            510,
+            572
+          ],
+          "label": "fnote",
+          "reading_order": 8,
+          "text": "8. See Bd. of Trade of Chi. v. United States, 246 U.S. 231, 238 (1918).\nJustice Brandeis wrote for the Court:"
+        },
+        {
+          "bbox": [
+            123,
+            572,
+            491,
+            715
+          ],
+          "label": "fnote",
+          "reading_order": 9,
+          "text": "The true test of legality is whether the restraint imposed is such as\nmerely regulates and perhaps thereby promotes competition or\nwhether it is such as may suppress or even destroy competition. To\ndetermine that question the court must ordinarily consider the facts\npeculiar to the business to which the restraint is applied; its condition\nbefore and after the restraint was imposed; the nature of the restraint\nand its effect, actual or probable. The history of the restraint, the evil\nbelieved to exist, the reason for adopting the particular remedy, the\npurpose or end sought to be attained, are all relevant facts. This is\nnot because a good intention will save an otherwise objectionable\nregulation or the reverse; but because knowledge of intent may help\nthe court to interpret facts and to predict consequences."
+        },
+        {
+          "bbox": [
+            101,
+            714,
+            118,
+            728
+          ],
+          "label": "fnote",
+          "reading_order": 10,
+          "text": "Id."
+        },
+        {
+          "bbox": [
+            101,
+            728,
+            510,
+            755
+          ],
+          "label": "fnote",
+          "reading_order": 11,
+          "text": "9. See, e.g., Polygram Holding, Inc. v. FTC, 416 F.3d 29, 33-38 (D.C. Cir.\n2005)."
+        },
+        {
+          "bbox": [
+            101,
+            755,
+            511,
+            782
+          ],
+          "label": "fnote",
+          "reading_order": 12,
+          "text": "10. Herbert Hovenkamp, Federal Antitrust Policy: The Law of\nCompetition and Its Practice 279–80 (4th ed. 2011)."
+        }
+      ]
+    },
+    {
+      "page_number": 5,
+      "elements": [
+        {
+          "bbox": [
+            101,
+            68,
+            121,
+            84
+          ],
+          "label": "header",
+          "reading_order": 0,
+          "text": "44"
+        },
+        {
+          "bbox": [
+            219,
+            67,
+            388,
+            85
+          ],
+          "label": "header",
+          "reading_order": 1,
+          "text": "MINN. J. L. SCI. & TECH."
+        },
+        {
+          "bbox": [
+            451,
+            67,
+            512,
+            85
+          ],
+          "label": "header",
+          "reading_order": 2,
+          "text": "[Vol. 15:1"
+        },
+        {
+          "bbox": [
+            100,
+            99,
+            512,
+            132
+          ],
+          "label": "half_para",
+          "reading_order": 3,
+          "text": "which I paraphrase in the following manner in my own\nantitrust classes:"
+        },
+        {
+          "bbox": [
+            131,
+            133,
+            490,
+            161
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "1. Consider first whether there is a \"contract, combination, or\nconspiracy\" that restrains trade (in some sense)."
+        },
+        {
+          "bbox": [
+            149,
+            161,
+            473,
+            178
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "If yes (conscious parallelism coupled with plus factors?), go on."
+        },
+        {
+          "bbox": [
+            149,
+            178,
+            349,
+            193
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "If not, § 1 doesn't apply (the \"§ 1 gap\")."
+        },
+        {
+          "bbox": [
+            131,
+            193,
+            490,
+            247
+          ],
+          "label": "para",
+          "reading_order": 7,
+          "text": "2. If necessary, consider next whether the restraint poses any\npossible risk to competition. I.e., is the restraint at issue one that\nposes a substantial risk of increasing price, lowering quantity, or\ncausing some other anticompetitive harm?"
+        },
+        {
+          "bbox": [
+            149,
+            248,
+            216,
+            263
+          ],
+          "label": "para",
+          "reading_order": 8,
+          "text": "If yes, go on."
+        },
+        {
+          "bbox": [
+            149,
+            263,
+            334,
+            279
+          ],
+          "label": "para",
+          "reading_order": 9,
+          "text": "If no, stop; judgment for defendant."
+        },
+        {
+          "bbox": [
+            131,
+            280,
+            489,
+            307
+          ],
+          "label": "para",
+          "reading_order": 10,
+          "text": "3. If necessary, consider next whether the restraint is likely to\ngenerate any plausible, cognizable, procompetitive benefits."
+        },
+        {
+          "bbox": [
+            149,
+            308,
+            489,
+            337
+          ],
+          "label": "para",
+          "reading_order": 11,
+          "text": "For example, does the restraint plausibly relate to the core\nactivities of a lawful joint venture?"
+        },
+        {
+          "bbox": [
+            149,
+            337,
+            490,
+            418
+          ],
+          "label": "para",
+          "reading_order": 12,
+          "text": "Is it plausibly ancillary in the sense of being reasonably\nnecessary to promote the legitimate activities of a joint\nundertaking? Reasonably necessary for the provision of some\ngood or service that consumers demand but which might not be\nprovided optimally if each competitor merely followed its own\nindividual self-interest?"
+        },
+        {
+          "bbox": [
+            149,
+            418,
+            216,
+            434
+          ],
+          "label": "para",
+          "reading_order": 13,
+          "text": "If yes, go on."
+        },
+        {
+          "bbox": [
+            149,
+            434,
+            489,
+            462
+          ],
+          "label": "para",
+          "reading_order": 14,
+          "text": "If no, stop; it is a naked restraint of trade, likely only to increase\nprice or reduce output or quality, and is per se illegal."
+        },
+        {
+          "bbox": [
+            131,
+            462,
+            490,
+            518
+          ],
+          "label": "para",
+          "reading_order": 15,
+          "text": "4. If necessary, consider next whether the defendant has market\npower (e.g., through substantial market share coupled with barriers\nto entry), or alternatively whether there is proof of actual\nanticompetitive effects, such as a reduction of output."
+        },
+        {
+          "bbox": [
+            149,
+            518,
+            216,
+            533
+          ],
+          "label": "para",
+          "reading_order": 16,
+          "text": "If yes, go on."
+        },
+        {
+          "bbox": [
+            149,
+            533,
+            334,
+            549
+          ],
+          "label": "para",
+          "reading_order": 17,
+          "text": "If no, stop; judgment for defendant."
+        },
+        {
+          "bbox": [
+            131,
+            549,
+            489,
+            578
+          ],
+          "label": "para",
+          "reading_order": 18,
+          "text": "5. If necessary, consider next whether the restraint at issue\nprovides actual (not just plausible) procompetitive benefits."
+        },
+        {
+          "bbox": [
+            149,
+            578,
+            216,
+            594
+          ],
+          "label": "para",
+          "reading_order": 19,
+          "text": "If yes, go on."
+        },
+        {
+          "bbox": [
+            149,
+            594,
+            324,
+            609
+          ],
+          "label": "para",
+          "reading_order": 20,
+          "text": "If no, stop; judgment for plaintiff."
+        },
+        {
+          "bbox": [
+            131,
+            609,
+            490,
+            637
+          ],
+          "label": "para",
+          "reading_order": 21,
+          "text": "6. If necessary, consider next whether the restraint is the least\nrestrictive means of attaining those benefits."
+        },
+        {
+          "bbox": [
+            149,
+            638,
+            216,
+            653
+          ],
+          "label": "para",
+          "reading_order": 22,
+          "text": "If yes, go on."
+        },
+        {
+          "bbox": [
+            149,
+            654,
+            324,
+            670
+          ],
+          "label": "para",
+          "reading_order": 23,
+          "text": "If no, stop; judgment for plaintiff."
+        },
+        {
+          "bbox": [
+            131,
+            670,
+            489,
+            699
+          ],
+          "label": "para",
+          "reading_order": 24,
+          "text": "7. If necessary, balance the procompetitive benefits against the\nanticompetitive costs (good luck!). $^11$"
+        },
+        {
+          "bbox": [
+            101,
+            753,
+            512,
+            782
+          ],
+          "label": "foot",
+          "reading_order": 25,
+          "text": "11. Thomas F. Cotter, PowerPoint Presentation, Antitrust Overview\n(unpublished document) (on file with author) [hereinafter Cotter PowerPoint];"
+        }
+      ]
+    },
+    {
+      "page_number": 6,
+      "elements": [
+        {
+          "bbox": [
+            101,
+            67,
+            140,
+            85
+          ],
+          "label": "header",
+          "reading_order": 0,
+          "text": "2014]"
+        },
+        {
+          "bbox": [
+            205,
+            66,
+            404,
+            85
+          ],
+          "label": "header",
+          "reading_order": 1,
+          "text": "NOT THE RULE OF REASON?"
+        },
+        {
+          "bbox": [
+            493,
+            68,
+            512,
+            83
+          ],
+          "label": "header",
+          "reading_order": 2,
+          "text": "45"
+        },
+        {
+          "bbox": [
+            101,
+            99,
+            513,
+            321
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "Assuming that this analysis is correct, what exactly will\ncourts be doing when they apply the rule of reason approach to\npay-for-delay cases such as Actavis ? Will they be starting from\nstep 1 above? No, because in any case in which a patentee\nagrees to pay money to an alleged infringer in return for the\nlatter's agreement to settle the case and temporarily exit the\nmarket there is obviously a contract that potentially restrains\ntrade; that much is indisputable. Equally obvious is the\npotential risk to competition (step 2). At the same time, there\nare potential procompetitive benefits (step 3), because (as a\ngeneral matter) settlement conserves social resources that\notherwise would be devoted to litigation and (in this specific\ncontext) in theory the settlement could speed up the entry of\ngeneric drugs to the market. $^12$"
+        },
+        {
+          "bbox": [
+            101,
+            321,
+            512,
+            369
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "Crucially, according to the majority, step 4 above is also\nlikely to be present in the context of pay-for-delay settlements.\nAs Justice Breyer wrote:"
+        },
+        {
+          "bbox": [
+            131,
+            370,
+            490,
+            530
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "First , the specific restraint at issue has the “ potential for\ngenuine adverse effects on competition. ” The payment in effect\namounts to a purchase by the patentee of the exclusive right to sell\nits product, a right it already claims but would lose if the patent\nlitigation were to continue and the patent were held invalid or not\ninfringed by the generic product. Suppose, for example, that the\nexclusive right to sell produces $ 50 million in supracompetitive\nprofits per year for the patentee. And suppose further that the\npatent has 10 more years to run. Continued litigation, if it results in\npatent invalidation or a finding of noninfringement, could cost the\npatentee $ 500 million in lost revenues, a sum that then would flow\nin large part to consumers in the form of lower prices."
+        },
+        {
+          "bbox": [
+            156,
+            537,
+            182,
+            560
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "...."
+        },
+        {
+          "bbox": [
+            131,
+            560,
+            489,
+            590
+          ],
+          "label": "para",
+          "reading_order": 7,
+          "text": "Second, these anticompetitive consequences will at least\nsometimes prove unjustified . . . ."
+        },
+        {
+          "bbox": [
+            131,
+            590,
+            491,
+            698
+          ],
+          "label": "para",
+          "reading_order": 8,
+          "text": "Third , where a reverse payment threatens to work unjustified\nanticompetitive harm, the patentee likely possesses the power to\nbring that harm about in practice. At least, the “size of the payment\nfrom a branded drug manufacturer to a prospective generic is itself\na strong indicator of power”—namely, the power to charge prices\nhigher than the competitive level. An important patent itself helps\nto assure such power. Neither is a firm without that power likely to\npay “large sums” to induce “others to stay out of its market.” In any"
+        },
+        {
+          "bbox": [
+            101,
+            726,
+            512,
+            754
+          ],
+          "label": "fnote",
+          "reading_order": 9,
+          "text": "see also Thomas F. Cotter, Patent Holdup, Patent Remedies, and Antitrust\nResponses, 34 J. CORP. L. 1151, 1205-06 (2009)."
+        },
+        {
+          "bbox": [
+            119,
+            754,
+            324,
+            767
+          ],
+          "label": "fnote",
+          "reading_order": 10,
+          "text": "12. See Actavis, 133 S. Ct. at 2234-35."
+        }
+      ]
+    },
+    {
+      "page_number": 7,
+      "elements": [
+        {
+          "bbox": [
+            101,
+            68,
+            121,
+            84
+          ],
+          "label": "header",
+          "reading_order": 0,
+          "text": "46"
+        },
+        {
+          "bbox": [
+            219,
+            67,
+            388,
+            85
+          ],
+          "label": "header",
+          "reading_order": 1,
+          "text": "MINN. J. L. SCI. & TECH."
+        },
+        {
+          "bbox": [
+            451,
+            67,
+            512,
+            85
+          ],
+          "label": "header",
+          "reading_order": 2,
+          "text": "[Vol. 15:1"
+        },
+        {
+          "bbox": [
+            130,
+            99,
+            490,
+            141
+          ],
+          "label": "half_para",
+          "reading_order": 3,
+          "text": "event, the Commission has referred to studies showing that reverse\npayment agreements are associated with the presence of higher-\nthan-competitive profits — a strong indication of market power. $^13$"
+        },
+        {
+          "bbox": [
+            101,
+            141,
+            513,
+            315
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "If we have made it all the way through steps 1 through 4,\nwhat remains? Step 5 asks (in my formulation) “ whether the\nrestraint at issue provides actual (not just plausible)\nprocompetitive benefits. ” 14 Importantly, the burden of proof on\nstep 5 normally rests with the defendant . 15 So if, under the\nCourt's own analysis, steps 1 through 4 are satisfied in the\ntypical pay-for-delay case and review really kicks in at step 5 —\nat which point the defendant has the burden of coming forward\nwith exonerating evidence — it is a little hard to see how that\nframework differs in any functional manner from presumptive\nillegality. 16"
+        },
+        {
+          "bbox": [
+            118,
+            346,
+            321,
+            361
+          ],
+          "label": "fnote",
+          "reading_order": 5,
+          "text": "13. Id. at 2234-36 (citations omitted)."
+        },
+        {
+          "bbox": [
+            101,
+            361,
+            512,
+            424
+          ],
+          "label": "fnote",
+          "reading_order": 6,
+          "text": "14. See Cotter PowerPoint, supra note 11. Professor Hovenkamp states\nthis step somewhat more forcefully, namely whether there is “ strong evidence\nthat the challenged practice creates substantial efficiencies by reducing\nparticipants' costs or improving product or service quality. ” Hovenkamp ,\nsupra note 10, at 280."
+        },
+        {
+          "bbox": [
+            101,
+            424,
+            513,
+            600
+          ],
+          "label": "fnote",
+          "reading_order": 7,
+          "text": "15. See, e.g., Polygram Holding, Inc. v. FTC, 416 F.3d 29, 36 (D.C. Cir.\n2005) (“[T]he evidentiary burden shifts to the defendant to show the restraint\nin fact does not harm consumers or has ‘procompetitive virtues’ that outweigh\nits burden upon consumers.”); Schering-Plough Corp. v. FTC, 402 F.3d 1056,\n1065 (11th Cir. 2005) (stating that “[o]nce the plaintiff meets the burden of\nproducing sufficient evidence of market power, the burden then shifts to the\ndefendant to show that the challenged conduct promotes a sufficiently pro-\ncompetitive objective”); Clorox Co. v. Sterling Winthrop, Inc., 117 F.3d 50, 56\n(2d Cir. 1997) (“[I]f the plaintiff succeeds [in showing an actual\nanticompetitive effect], the burden shifts to the defendant to establish the\n“pro-competitive ‘redeeming virtues” of the action. Should the defendant carry\nthis burden, the plaintiff must then show that the same pro-competitive effect\ncould be achieved through an alternative means that is less restrictive of\ncompetition.”)."
+        },
+        {
+          "bbox": [
+            118,
+            600,
+            489,
+            614
+          ],
+          "label": "fnote",
+          "reading_order": 8,
+          "text": "16. See Polygram Holding, 416 F.3d at 36. The Polygram court stated:"
+        },
+        {
+          "bbox": [
+            123,
+            614,
+            490,
+            745
+          ],
+          "label": "para",
+          "reading_order": 9,
+          "text": "For reasons we have already explained, we reject PolyGram's\nattempt to locate the appropriate analysis, and the concomitant\nburden of proof, by reference to the vestigial line separating per se\nanalysis from the rule of reason. See Areeda & Hovenkamp, Antitrust\nLaw, ¶ 1511a (“judges and litigants too often assume erroneously that\nthe classification, per se or rule of reason, necessarily determines\nwhat must or may be alleged and proved, made the subject of detailed\nfindings, or submitted to the jury”). At bottom, the Sherman Act\nrequires the court to ascertain whether the challenged restraint\nhinders competition; the Commission's framework, at least as the\nCommission applied it in this case, does just that."
+        },
+        {
+          "bbox": [
+            123,
+            745,
+            490,
+            783
+          ],
+          "label": "para",
+          "reading_order": 10,
+          "text": "We therefore accept the Commission's analytical framework. If,\nbased upon economic learning and the experience of the market, it is\nobvious that a restraint of trade likely impairs competition, then the"
+        }
+      ]
+    },
+    {
+      "page_number": 8,
+      "elements": [
+        {
+          "bbox": [
+            101,
+            67,
+            140,
+            85
+          ],
+          "label": "header",
+          "reading_order": 0,
+          "text": "2014]"
+        },
+        {
+          "bbox": [
+            204,
+            66,
+            405,
+            85
+          ],
+          "label": "header",
+          "reading_order": 1,
+          "text": "NOT THE RULE OF REASON?"
+        },
+        {
+          "bbox": [
+            493,
+            68,
+            513,
+            84
+          ],
+          "label": "header",
+          "reading_order": 2,
+          "text": "47"
+        },
+        {
+          "bbox": [
+            100,
+            98,
+            513,
+            180
+          ],
+          "label": "para",
+          "reading_order": 3,
+          "text": "This is particularly so given the Court's further statements\nthat “it is normally not necessary to litigate patent validity to\nanswer the antitrust question” and its discussion of the type of\nprocompetitive justifications that might excuse a reverse\npayment. $^17$ As for the first issue, Justice Breyer wrote:"
+        },
+        {
+          "bbox": [
+            130,
+            180,
+            491,
+            378
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "An unexplained large reverse payment itself would normally\nsuggest that the patentee has serious doubts about the patent's\nsurvival. And that fact, in turn, suggests that the payment's\nobjective is to maintain supracompetitive prices to be shared among\nthe patentee and the challenger rather than face what might have\nbeen a competitive market — the very anticompetitive consequence\nthat underlies the claim of antitrust unlawfulness. The owner of a\nparticularly valuable patent might contend, of course, that even a\nsmall risk of invalidity justifies a large payment. But, be that as it\nmay, the payment (if otherwise unexplained) likely seeks to prevent\nthe risk of competition. And, as we have said, that consequence\nconstitutes the relevant anticompetitive harm. In a word, the size of\nthe unexplained reverse payment can provide a workable surrogate\nfor a patent's weakness, all without forcing a court to conduct a\ndetailed exploration of the validity of the patent itself. $^18$"
+        },
+        {
+          "bbox": [
+            100,
+            378,
+            511,
+            444
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "In other words, the plaintiff is not going to have to prove,\nexcept inferentially by reference to the amount of the payment,\nthat the probability of patent invalidity was high. As for the\nsecond, the Court noted that reverse payments"
+        },
+        {
+          "bbox": [
+            130,
+            444,
+            491,
+            602
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "may amount to no more than a rough approximation of the litigation\nexpenses saved through the settlement. That payment may reflect\ncompensation for other services that the generic has promised to\nperform — such as distributing the patented item or helping to\ndevelop a market for that item. There may be other justifications.\nWhere a reverse payment reflects traditional settlement\nconsiderations, such as avoided litigation costs or fair value for\nservices, there is not the same concern that a patentee is using its\nmonopoly profits to avoid the risk of patent invalidation or a finding\nof noninfringement. In such cases, the parties may have provided for\na reverse payment without having sought or brought about the\nanticompetitive consequences we mentioned above. $^19$"
+        },
+        {
+          "bbox": [
+            101,
+            602,
+            512,
+            652
+          ],
+          "label": "para",
+          "reading_order": 7,
+          "text": "All of this leads me to conclude that the reasons the Court\nchose not to characterize what it was doing as a presumptive\nillegality approach were either (1) political, e.g., to keep one or"
+        },
+        {
+          "bbox": [
+            123,
+            676,
+            490,
+            727
+          ],
+          "label": "half_para",
+          "reading_order": 8,
+          "text": "restraint is presumed unlawful and, in order to avoid liability, the\ndefendant must either identify some reason the restraint is unlikely\nto harm consumers or identify some competitive benefit that\nplausibly offsets the apparent or anticipated harm."
+        },
+        {
+          "bbox": [
+            101,
+            726,
+            118,
+            739
+          ],
+          "label": "para",
+          "reading_order": 9,
+          "text": "Id."
+        },
+        {
+          "bbox": [
+            119,
+            740,
+            305,
+            753
+          ],
+          "label": "fnote",
+          "reading_order": 10,
+          "text": "17. See Actavis, 133 S. Ct. at 2236."
+        },
+        {
+          "bbox": [
+            119,
+            753,
+            224,
+            766
+          ],
+          "label": "fnote",
+          "reading_order": 11,
+          "text": "18. Id. at 2236-37."
+        },
+        {
+          "bbox": [
+            119,
+            767,
+            206,
+            781
+          ],
+          "label": "fnote",
+          "reading_order": 12,
+          "text": "19. Id. at 2236."
+        }
+      ]
+    },
+    {
+      "page_number": 9,
+      "elements": [
+        {
+          "bbox": [
+            101,
+            68,
+            122,
+            84
+          ],
+          "label": "header",
+          "reading_order": 0,
+          "text": "48"
+        },
+        {
+          "bbox": [
+            219,
+            67,
+            388,
+            85
+          ],
+          "label": "header",
+          "reading_order": 1,
+          "text": "MINN. J. L. SCI. & TECH."
+        },
+        {
+          "bbox": [
+            451,
+            67,
+            512,
+            85
+          ],
+          "label": "header",
+          "reading_order": 2,
+          "text": "[Vol. 15:1"
+        },
+        {
+          "bbox": [
+            100,
+            99,
+            513,
+            243
+          ],
+          "label": "half_para",
+          "reading_order": 3,
+          "text": "more possibly gun-shy justices on board with the majority, on\nthe theory that a rule of reason approach is not quite as pro-\nplaintiff as a presumptive illegality approach; or (2) based\non concerns that courts might construe the adoption of a\npresumptive illegality approach in the present case as\neffectively holding that such an approach is appropriate in\nother cases, not arising in the byzantine shadow of Hatch-\nWaxman. The concluding section of the majority opinion seems\nto reflect this latter concern, 20 and thus may be viewed as a"
+        },
+        {
+          "bbox": [
+            117,
+            279,
+            379,
+            294
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "20. See id. at 2237-38. Justice Breyer concluded:"
+        },
+        {
+          "bbox": [
+            123,
+            294,
+            490,
+            390
+          ],
+          "label": "para",
+          "reading_order": 5,
+          "text": "[T]he likelihood of a reverse payment bringing about anticompetitive\neffects depends upon its size, its scale in relation to the payor's\nanticipated future litigation costs, its independence from other\nservices for which it might represent payment, and the lack of any\nother convincing justification. The existence and degree of any\nanticompetitive consequence may also vary as among industries.\nThese complexities lead us to conclude that the FTC must prove its\ncase as in other rule-of-reason cases."
+        },
+        {
+          "bbox": [
+            123,
+            390,
+            490,
+            486
+          ],
+          "label": "para",
+          "reading_order": 6,
+          "text": "To say this is not to require the courts to insist, contrary to what\nwe have said, that the Commission need litigate the patent's validity,\nempirically demonstrate the virtues or vices of the patent system,\npresent every possible supporting fact or refute every possible pro-\ndefense theory. As a leading antitrust scholar has pointed out,\n“ [t]here is always something of a sliding scale in appraising\nreasonableness, ” and as such “ the quality of proof required should\nvary with the circumstances. ”"
+        },
+        {
+          "bbox": [
+            123,
+            486,
+            490,
+            581
+          ],
+          "label": "para",
+          "reading_order": 7,
+          "text": "As in other areas of law, trial courts can structure antitrust\nlitigation so as to avoid, on the one hand, the use of antitrust theories\ntoo abbreviated to permit proper analysis, and, on the other,\nconsideration of every possible fact or theory irrespective of the\nminimal light it may shed on the basic question — that of the presence\nof significant unjustified anticompetitive consequences. We therefore\nleave to the lower courts the structuring of the present rule-of-reason\nantitrust litigation."
+        },
+        {
+          "bbox": [
+            101,
+            581,
+            513,
+            782
+          ],
+          "label": "para",
+          "reading_order": 8,
+          "text": "Id. (citations omitted). I would expect that settlements of patent infringement\nlitigation outside of the Hatch-Waxman context will rarely give rise to\nplausible antitrust claims under the rule of reason. In a typical case,\nsettlement may increase output (e.g., by resulting in a nonexclusive license of\na patent that has withstood a validity challenge), thus failing step 2 above; or\nthe patentee lacks market power (step 4); or courts will conclude, as a general\nrule, that a settlement lacking any red flags such as the presence of a reverse\npayment in excess of the defendant's expected profit (or other suspicious\nconditions) necessarily has procompetitive benefits that outweigh any\nanticompetitive consequences that could be proven without unraveling the\nreduction of adjudicative costs that is the primary social good that settlement\nproduces (and thus will not countenance attempts to question patent validity\nor infringement absent good reason). But I tend to agree with the majority\nthat patent settlements should not be effectively immune from antitrust\nscrutiny absent conduct such as sham or fraud, which arguably was the\nimplication of the scope-of-the-patent test."
+        }
+      ]
+    },
+    {
+      "page_number": 10,
+      "elements": [
+        {
+          "bbox": [
+            100,
+            67,
+            140,
+            85
+          ],
+          "label": "header",
+          "reading_order": 0,
+          "text": "2014]"
+        },
+        {
+          "bbox": [
+            204,
+            66,
+            405,
+            85
+          ],
+          "label": "header",
+          "reading_order": 1,
+          "text": "NOT THE RULE OF REASON?"
+        },
+        {
+          "bbox": [
+            493,
+            67,
+            513,
+            84
+          ],
+          "label": "header",
+          "reading_order": 2,
+          "text": "49"
+        },
+        {
+          "bbox": [
+            100,
+            99,
+            512,
+            132
+          ],
+          "label": "half_para",
+          "reading_order": 3,
+          "text": "rejoinder to the dissent's concern that the majority approach\nrenders vulnerable even conventional patent settlements. $^21$"
+        },
+        {
+          "bbox": [
+            100,
+            132,
+            513,
+            245
+          ],
+          "label": "para",
+          "reading_order": 4,
+          "text": "In conclusion, it seems to me that the majority adopted a\n(de facto) presumptive illegality approach to pay-for-delay\nsettlements entered into in the shadow of Hatch-Waxman —\nprecisely the approach that many of us were hoping for $^22$ — even\nif, for political or prudential reasons, it suggests that it did not.\nAs long as the lower courts correctly interpret the message, this\nseems an acceptable resolution to the pay-for-delay problem."
+        },
+        {
+          "bbox": [
+            116,
+            728,
+            461,
+            743
+          ],
+          "label": "fnote",
+          "reading_order": 5,
+          "text": "21. See Actavis, 133 S. Ct. at 2244-45 (Roberts, C.J., dissenting)."
+        },
+        {
+          "bbox": [
+            100,
+            743,
+            511,
+            782
+          ],
+          "label": "fnote",
+          "reading_order": 6,
+          "text": "22. See Brief Amici Curiae of 118 Law, Economics, and Business\nProfessors and the American Antitrust Institute in Support of Petitioners,\nFTC v. Actavis, Inc., 133 S. Ct. 2233 (2013) (No. 12-416), 2013 WL 391001."
+        }
+      ]
+    },
+    {
+      "page_number": 11,
+      "elements": [
+        {
+          "bbox": [
+            294,
+            90,
+            318,
+            100
+          ],
+          "label": "para",
+          "reading_order": 0,
+          "text": "**"
+        }
+      ]
+    }
+  ],
+  "metadata": {}
+}

processed/00221_W2122457513/document.md

@@ -0,0 +1,574 @@
+J Technol Transf (2010) 35:401-423 DOI 10.1007/s10961-009-9126-2
+
+# Inventing and patenting activities of scientists: in the expectation of money or reputation?
+
+Devrim Göktepe-Hulten · Prashanth Mahagaonkar
+
+Published online: 10 June 2009 ❾ The Author(s) 2009. This article is published with open access at Springerlink.com
+
+Abstract We propose that scientists use patents/invention disclosures as signals to gain reputation than financial benefits. Based on a newly created dataset on the commercial activities among 2,500 scientists affiliated with 67 institutes of the German Max Planck Society, we explore the relation between the expectations of scientists concerning the outcomes of commercial activities and the likelihood of their patenting and disclosure behaviors. We find that expectation of gaining financial benefits are not related with the patenting activities of scientists without industrial cooperation. Instead, their expectation to gain/increase reputation through commercial activities is correlated with their patenting and disclosures activities. This may in turn also increase the possibility to gain academic promotion, financial benefits through industrial collaboration etc., rather than the immediate personal financial gains.
+
+Keywords Academic commercialization · Patents · Rewards · Reputation · Signaling
+
+JEL Classification B31 · 031 · 034
+
+## 1 Introduction
+
+Scientists carry out the tasks of education, research and commercial activities (the third task) at universities. Despite their importance, the roles, motivations and perceptions of university inventors have been relatively neglected topics of study. As Link and Siegel ( 2007 ) have argued, since the initiation of the Bayh–Dole Act, scholars who assess university technology transfer have examined institutions that have emerged to facilitate entrepreneurial
+
+D. Göktepe-Hulten ( ) · P. Mahagaonkar Max Planck Institute of Economics, Jena, Germany e-mail: goktepe@econ.mpg.de
+
+P. Mahagaonkar Schumpeter School of Business and Economics-University of Wuppertal, Wuppertal, Germany
+
+![Figure](figures/figure_011.png)
+
+---
+
+402
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+commercialization, such as university technology transfer offices (TTOs), industry-university cooperative research centers, research/science parks, and incubators and different patent legislations. Most studies on university-industry relations have focused on a few selected elite universities in the United States, in specific science-based sectors. In these studies, the focus of interest is primarily the importance of institutions (patent legislation, policy mechanisms) and organizations (TTOs, university administration) in the patenting and other entrepreneurial activities of scientists (see recent reviews by Siegel and Phan 2005 ; Phan and Siegel 2006 ; Siegel et al. 2007 ; Rothermael et al. 2007 ) . Some studies initiated the importance of individuals, but rather limited themselves only to entrepreneurial traits, experience, scientific background and demographic factors such as age in order to analyze commercialization motives of scientists. In particular there have been a few studies that paid attention to the roles of individual inventors in the university-industry technology transfer and explore why academic scientists patent (Gulbrandsen 2005 ; Meyer 2005 ; Azoulay et al. 2007 ; Allen et al. 2007 ; Baldini et al. 2007 ; Bercovitz and Feldman 2008 ; Goktepe 2008 ) .
+
+The primary missions of scientists employed at universities and PROs (public research organizations) have been the creation and dissemination of knowledge and education of students. After the 1990s, scientists are also expected to carry out a third mission, namely commercial activities like patenting and company formation. Patenting is a mechanism to `privatize' information by excluding others to the intellectual property to gain monopoly rights over the commercial use of the inventions, and it is essentially an economic phenomenon. According to Thursby and Thursby ( 2007 ) , the common rationale behind university patenting and licensing is that they provide financial incentives for universities, faculty, and firms to engage in the commercialization of university research findings. $^1$ It is almost believed that any invention would barely come out of a human's brain if that human did not have the possibility to earn all or part of the stream of economic rents that results from the industrial exploitation of his or her invention, a preliminary condition for that being that he or she ought to own a propriety right (usually a patent) over that invention (Schmookler 1966 ) .
+
+Although commercial activities are seen as potential revenue sources for universities, it is not clear if scientists will also pursue commercial activities as potential revenue sources or it is bolted on to their traditional roles and expectations of being engaged in science. Expected monetary benefits due to patenting and licensing activities are more relevant for firms and TTOs. However benefits that patents provide to firms might not be the same as for the individual scientists. Academic scientists may have different concerns and expectations when they are involved in patenting activities compared to firms and university-TTOs. For instance, based on their research on the patenting and licensing activities in the US, Thursby and Thursby ( 2007 , p. 633 ) found licensing income is the most important objective for the TTO and the central administration, and firms. For faculty funds for sponsored research are the most important objective. Thursby and Thursby ( 2007 , p. 625 ) suggested “ patents are not necessary to provide incentives for university scientists and engineers to invent and disclose; the norms of science and the reward structure for science provide incentives for invention and public disclosure. Likewise, Rosenberg ( 1974 ) had already argued that inventive activity — along with technological change and the production of scientific and technical knowledge — as something that was independent of economic needs and motivations.
+
+Accordingly a deeper understanding of scientists' expectations will provide better policy insights on the initiation of entrepreneurial activities at the universities and PROs. We therefore specifically focus on the relationship between the likelihood of scientists'
+
+$^{1}$ This has been a common argument for the initiation of the Bayh-Dole Act as well.
+
+![Figure](figures/figure_007.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+403
+
+patenting and inventing activities and the rewards they are expecting from commercial activities which are measured in terms of financial benefits and scientific reputation. In line with these arguments, it is of particular interest to understand what matters for scientists to disclose their inventions to authorities and patent. Is it correlated to the expectations for gaining reputation or financial rewards?
+
+In order to address this question we use a unique database developed on the commercialization activities of over 2,500 scientists at 67 different institutes of Max Planck Society for Advancement of Sciences (hereafter referred as MPG). Using discrete choice models on patenting and invention disclosure to the MPG, we analyzed if expectation of financial benefits and reputation influence the inventing and patenting activities of scientists. We distinguish between industry-collaborating and non-collaborating scientists to focus on specific motivations of non-collaborators' innovation activities. We control for different socio-demographic as well as institutional factors and scientific fields in our analysis. We find that for non-collaborating academic inventors, invention disclosure and patenting activities are not automatically related to their financial expectations from the commercial activities. Instead, the expectation of reputation that drives patenting and invention disclosure activities of scientists. This confirms the assertions made by Long ( 2002 ) that patenting is basically an information transfer mechanism and patentees use patents not always for the expected financial benefits by excluding others but for the nonmonetary benefits that accrue due to the information conveyed. Individuals may resort to actions that signal their knowledge, skills and resources by conveying the right information to the relevant group.
+
+It can be interpreted that scientists who are working under the conditions of generous research funds, like MPG, but with a strong desire as well as stipulation of doing cuttingedge research will still be motivated by traditional academic values, like increasing recognition and reputation by showing the novelty of their research. Yet, we can not conclude scientists who are facing more scarce/limited resources for research maybe forced to engage in commercial activities or patenting in order to gain more money. Patenting activities could, to a certain extent, be independent from private economic incentives. Therefore rather than a total transition to an entrepreneurial identity or mission, scientists are still keeping their traditional expectations from science as focus. There may be other reasons that academic scientists are not willing or in need to abandon the norms of science even when they are involved in commercialization activities.
+
+The structure of the paper is organized as follows, the following section deals with the question of why scientists patent and disclose their inventions and takes the view of patents as reputation signals that scientists use. In the third section, perceptions and motivations of scientists are shed light upon and propositions are put forward after which in the fourth section the new dataset is introduced along with the variables of interest and methodology. The fifth section puts forward the estimation results and analysis and sixth concludes.
+
+## 2 Theoretical background
+
+In order to develop the principal arguments of this paper, we revisit some studies on norms and rewards in science $^2$ and the basic arguments behind patents. From a legal perspective,
+
+2 The concept of rewards (benefits) from science should not be misinterpreted as the sole reason of scientists is to obtain social status and financial gains. There is considerable evidence that scientists have a desire for inventing. Scientists at universities are intrinsically motivated to do research. Much of the incentive to
+
+![Figure](figures/figure_009.png)
+
+---
+
+404
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+a patent confers legal rights concerning the exploitation of an invention which allows the owner the best opportunity to profit from the invention by preventing others from copying it. With a patent, an inventor has the right to exclude others from commercial use of the intellectual property rights conferred by the patent. This allows the inventor to appropriate economic returns from her inventive activity (Arrow 1962 ) . The prospect of gaining profits from this special form of protection serves to promote research activity and to give an incentive for new investment. An inventor does not need a patent in order to exploit an invention; but without a patent the inventor would not be able to prevent others from copying the invention. Inventors are often not in a position to produce or market their invention from their own resources. Patents, being a form of commercial property, provide a basis for owners to negotiate with potential investors or other business partners while preserving their intellectual property rights.
+
+Academic inventors are also supposed to disclose the details of the invention/innovation. While making the research results publicly available allows others to build on the invention, as such it also compels others to recognize and respect to the scientific results shown in the invention. Although there have been concerns that increase in university patenting has challenged the open nature of university research and shifted academic research towards more commercialization, a number of scholars have investigated the relationship between patenting and open dissemination of research results by scientists in the forms of publications (Agrawal and Henderson 2002 ; Jensen and Murray 2005 ; Van Looy et al. 2006 ) . These studies have also found that publication and patenting are complementary and not competing activities of university researchers. Most of these studies have found a positive relationship between scientific publication and patenting activities. Jensen and Murray ( 2005 ) stated that most university research generates dual outcomes which can be utilized as paper-patent pairs. Scientific publications will follow as in most cases a research project can generate outputs that simultaneously contribute to public knowledge and to commercialization. Academic inventor manages to show his research value both as a scientific discovery and a potential commercial product.
+
+University research is typically conducted in the context of the norms of science (Merton 1973 ) . As such, there exists a `natural' incentive within universities both to invent and to disclose rather than there being a tension between the two incentives (Eisenberg 1987 ) . $^3$ Scientific knowledge principles should be `assigned to the community' rather than the scientist, and the scientist's claim to intellectual property from their work should be limited to `recognition and esteem'. As implied by communalism, the reward for discovery and research should be recognition. Stephan ( 1996 ) argues that the priority reward system confers a form of property right that differs from the patent system in providing for incentives to invent and disclose. The priority reward is exclusionary in that the first to discover a principle receives the recognition and then captures the reputation for the discovery, and, as such, it requires no separate disclosure requirement. This winner-takesall property provides an incentive to invent and, at the same time, it promotes the incentive to disclose. A scientist will disclose a discovery as soon as there is sufficient evidence of
+
+Footnote 2 continued invent comes from the joy of solving research questions (Levin and Stephan 1991 ; Stephan 1996 ) . But their behavior is inevitably influenced by social rewards. In particular, the possibility that someone else gets credit and due to himself is as unacceptable to a scientist as to anyone else (Ben-David and Sullivan 1975 ) .
+
+3 In the context of industrial research and development, there is a natural tension between the incentives for patenting and invention disclosure. If firm could not exclude others from commercial use; it would have an incentive to keep secret the inventions.
+
+![Figure](figures/figure_007.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+405
+
+validity in order to gain recognition and reputation in a timely fashion among their peers. As firms and TTOs are expecting monetary benefits on account of patenting and licensing activities it is also interesting to investigate if financial rewards are relevant for the invention disclosure and patenting activities of individual academic scientists.
+
+In the light of the brief overview as a point of departure we tip our hand with two basic assumptions: (1) scientists have an interest in recognition and prestige; (2) scientists have an interest in achieving economic gains $^4$ (Stephan and Levin 1992 ) .
+
+### 2.1 Norms and rewards of science
+
+### 2.1.1 Recognition and prestige:
+
+Merton ( 1957 ) stated that the institution of science has developed “ a reward system designed to give recognition and esteem to those who has best fulfilled their roles, to those who have made genuinely original contributions to the common stock of knowledge ” . The reward system operates to encourage creative scientists to be highly productive and to produce a higher correlation between the quantity and quality of their output (Cole and Cole 1967 ) . Therefore scientists are motivated by rewards of recognition and prestige among peers, and they have a strong interest in winning the game. By the nature of their work scientists constantly ask research questions and aim to show their research results among their peers to achieve reputation and recognition (Merton 1957 ) .
+
+Merton ( 1957 ) also noted the apparent contradiction between the norms of communality which require scientists to publish their research results and consider them as the property of mankind, and their sensitivity and desire for superiority in discoveries. He argued the proper recognition of discovery is a necessary condition for the maintenance of communality, since without recognition scientists could not defend their intellectual property. These statements created a theoretically meaningful basis for further empirical research of the correlation between rewards and the patenting activities of scientists.
+
+Patenting can enhance the prestige and increase the scientific productivity of the scientists by reaffirming the novelty and usefulness of their research (Owen-Smith and Powell 2001 , 2003 ) . More recent empirical findings also show patenting as a matter of doing something professionally satisfying and rewarding (Gulbrandsen 2005 ; Baldini et al. 2007 ; Goktepe 2008 ) . As we discuss below, due to ongoing changes in the role of scientific institutes and scientists, some scientists may use patents to signal the quality and novelty of their research. Although there is no explicit evidence that patents are used as a criterion to evaluate the academic merits of the scientists (e.g. in academic promotion), some scientists may consider patenting in order to increase their visibility and reputation.
+
+On the other hand, the extent to which scientists and their research milieus are ready for rewarding commercial activities as an academic merit will influence the decision of scientists to patent or not. Scientists who are concerned or surrounded by cohorts with more traditional academic values like open (public) nature of science might be less motivated to patent. Similarly, in order not to risk their career development scientists, e.g. junior scientists, who are more anxious or unsure about how peers, leaders and potential future employees will assess/react to their commercial/patenting activities, will be less likely to patent.
+
+4 Financial rewards are measured in two ways; (1) gains from commercial activities and (2) getting access to external funds, industrial research grants.
+
+![Figure](figures/figure_011.png)
+
+---
+
+406
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+### 2.1.2 Source of personal income
+
+Etzkowitz ( 1998 ) and Slaughter and Leslie ( 1997 ) underlined financial rewards, monetary compensation and profit motive in their analyses of the new entrepreneurial scientist. Universities that provide greater rewards for scientists' involvement in patenting (e.g. in the forms of equity shares, royalty distribution) are found to motivate scientists to commercialize (patent) more. Greater rewards are measured by the amount of royalty income received by the inventor. Owen-Smith and Powell ( 2001 ) argued that scientists' decisions to disclose are shaped by their perceptions of the benefits of patenting, licensing and startup company formation. The incentives to be involved in technology transfer are magnified or minimized by the perceived costs and gains of interacting with industry and TTOs. Siegel et al. ( 2003 ) concluded that organizational factors, in particular scientists' reward systems and technology transfer office compensation, influence the productivity of the technology transfer activities and thus the motivations of scientists to disclose their inventions.
+
+Bercovitz and Feldman ( 2008 ) assumed that faculty members would be responsive to financial incentives and that there would be a direct relationship between licensing royalty distribution rates and the amount of technology transfer across universities. Thursby et al. ( 2001 ) and Lach and Schankerman ( 2003 ) provided empirical evidence that milestone payments and share of license revenues from their inventions are positively related to the motivations of inventors to patent. Markman et al. ( 2004 ) investigated the relationship between entrepreneurial activities and payments to scientists, departments and TTO staff. They argued that scientists and their departments will be unlikely to disclose or participate in technology transfer activities unless they are given proper incentives to do so. They expect licensing revenues from technology transfer activities can motivate scientists and their departments towards entrepreneurial activities given the scarcity of resources on research.
+
+However, there are counter arguments why financial rewards may not influence commercialization activities. In addition to Mertonian norms 5 (see Merton 1973 ) ; there is considerable evidence that scientists have a desire for doing research and inventing. The puzzle-solving nature of research is described by the historian of science, Robert Hull (1988 in Stephan and Levin 2005 ) . Puzzle-solving involves a fascination for the research process itself (Stephan and Levin 2005 ) . Scientists at universities are intrinsically motivated to do research. Much of the incentive to invent comes from the joy of solving research questions (Levin and Stephan 1991 ; Stephan 1996 ) . Thus they are intrinsically motivated to conduct research, quite apart from the ability to earn financial rents from their effort (Hellmann 2007 ) .
+
+“A scientist, by choice of vocation, would heretofore have been assumed to have put aside all thoughts of business-like activity to live a monk-like existence as a searcher for truths about nature” (Etzkowitz 1998 ) . Etzkowitz continues—“attired in a white lab coat to protect their street clothing from chemical spills, the uniform of the scientist also signified a certain purity of motives, an abstraction from material concerns and a bemused tendency toward absentmindedness in daily life”. Further, “they were believed to find rewards for their discoveries not in pecuniary advantage but in recognition from their scientific peers through citation in the literature, election to a national academy and the ultimate accolade of the Nobel Prize”.
+
+5 Merton suggested four norms of science: universalism, communism (or communalism), disinterestedness, and organized skepticism.
+
+![Figure](figures/figure_008.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+407
+
+In a recent study Jeon and Menicucci ( 2008 ) discussed the allocation of talent (brain drain) between the science and private sectors when agents value money and fame. They assumed not only monetary rewards matter in agents decisions but fame, which is defined as peer recognition, matters as well. Recent empirical studies have also confirmed that the innate curiosity of scientists make them research that can be publishable. Gulbrandsen ( 2005 ) , Goktepe ( 2008 ) investigated the motives of inventors to patent. They asked whether monetary rewards or non-monetary rewards were important motivations for patenting. Consistently these studies although limited in scope found that personal satisfaction and doing something professionally enjoyable were important reasons for scientists to be involved in commercialization. They found that social and personal rewards (i.e. the fact that the innovation might increase the performance of the organization where the inventor works), personal satisfaction to show that something is technically possible, and gaining prestige and reputation) were considered by the inventors to be more important than other types of compensation like monetary rewards and career advancement.
+
+### 2.2 Patents as quality signals
+
+Anton and Yao ( 2004 ) find that many of the patents do not actually reveal complete information on the invention process, therefore leading to `little patents-big secrets'. So, with this finding it seems plausible that monetary benefits to patents can be still assured, without a danger to the knowledge underlying the invention process. But do all individuals patent just because they want money by excluding others? While we know about the monetary gains from patents, an equally intriguing gain is reputation. Since we are interested in individuals, reputation seems to be another interest that would drive them to act on different things.
+
+Recognition is allotted to scientists to the extent they fulfilled their academic tasks (Blume and Sinclair 1973 ) . Reputation can be achieved and shown by scientific publications, honorific awards, positions at top-ranked institutes, and citations. Scientists (researchers) have to publish in order not only to show the findings of their research but also the quality, novelty and uniqueness of their findings. Publishing internationally peerreviewed scientific articles in top journals, being cited, participating in or even prestigious being invited in top international conferences, teaching skills and receiving grants are always considered as academic merits and improve the chances of academic promotion and reputation. Due to the intensification of university-industry relations economic development through technology transfer has become a “third academic mission” on a par with universities’ traditional missions of teaching and research. We therefore wonder in addition to the abovementioned tools, if scientists’ expectation of gaining reputation (visibility/ recognition) is correlated with their patenting and disclosure behaviors.
+
+Competition for reputation among scientists, universities, public research organizations creates both rising demand for and supply of researchers (Ben-David and Sullivan 1975 ) . Higher education departments of German states and university presidents in the US were particularly inclined to accept scientific reputation as criterion for appointing professors and evaluating institutions. Reputable scientists and lobbies (e.g. Gessellschaft deutscher Natur und Ärzte) 6 persuade governments to establish new chairs and recognize new fields. Reputation is believed to be a relatively objective, almost measurable yardstick (BenDavid 1972 in Ben-David and Sullivan 1975 ) .
+
+6 http://www.gdnae.de/be89d0ed8a4810173299eb1891120558/de/start/ueber_die_gdnae/index.html.
+
+![Figure](figures/figure_008.png)
+
+---
+
+408
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+Quantity and quality of scientific publications, citations, conference presentations etc. have been and are still guarantees of recognition from the scientific community and employers. Although patents are not peer reviewed, the patenting process is exhaustive and extensive concerning the novelty, usefulness, non-obviousness and technical utility. In particular, for the fulfillment of Third Mission activities, scientists are expected to interact with the surrounding society and be more active in commercial activities in addition to their usual tasks. Patenting activities have thus become considered as tools for rewards to gain reputation/recognition, and financial benefits. In order to be reputable, in the first place, information has to be conveyed about the person in context. In this view, a scientist can be thought of conveying `his type' (highly productive–low productive) to specifically two or more groups of people. One major group would be the compatriots in the research field concerned while another can be the employer. To the first group, scientists have three ways to convey information about their type—either publish, or patent, or do both. To the second group in addition to these two ways, one specific channel would be to report their findings officially—meaning—disclose their invention to the employer on an official basis. 7
+
+While the rewards for academic researchers are still largely based on publications, they also receive some (limited) commercial returns to patenting and other forms of commercial science (Edwards et al. 2006 ) . Even when financially unsuccessful, commercial science provides additional scientific resources. For example, Murray and Graham ( 2007 ) show that participation in commercial science brings with it distinctive forms of status and resources and that patents have become an integral part of faculty strategies for the dissemination of ideas and for signaling interest in commercial activities. Patents can be used as a tool to trade with industry for access to funding, equipment, materials and other opportunities from industry (Stephan and Levin 1992 ; Owen-Smith and Powell 2001 ) . Scientists can use patents as a signal to show the industrial relevance and applicability of their research results in order to attract more industrial support. In this case, the research results would be likely to be patented together with the industrial financier of the project.
+
+Based on the arguments posed until now, we frame the following hypotheses for empirical analysis. Since we do not make a case for only reputation or only money drives patenting, we test several possibilities in terms of methodology. To empirically test these specific hypotheses, we also account for several individual and external (institution specific) factors that may influence patenting and invention disclosure decision of scientists.
+
+- – H1: Expected reputation affects scientists' patenting and invention disclosure
+activities.
+– H2: Expected monetary benefits affect scientists' patenting and invention disclosure
+activities.
+## 3 Research context
+
+In this section we first give brief information about public research organization, i.e. Max Planck Society, as the background for our research context. We then present the rules and regulations concerning industrial co-operation and patenting activities of scientists who are
+
+7 Invention disclosure to the employer is a job requirement. Different from the former university patent legislation (university teachers' privilege (Sect. 42 ArbNErfG—Law on Employees' Inventions), [organizational] ownership of intellectual property rights (IPR) regime has been valid since the 1970s.
+
+![Figure](figures/figure_009.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+409
+
+affiliated with different institutes at Max Planck Society (MPG hereafter). 8 MPG was founded in the late 1940s in Germany. The Max Planck Institutes are engaged in numerous disciplines which are highly regarded as national and international centers of excellence, and their scientists publish more than 12,000 scientific articles, books, conference reports and other publications each year. The research results (discoveries, inventions, patents) of MPG scientists have also industrial applications. For example, the so-called FLASHtechnology in magnetic resonance imaging and the novel cancer treatment Sutent ® are both based on the research work of scientists from the Max Planck Society. 9
+
+Rules concerning the ownership of intellectual property (IP) at MPG have been always different from the former university patent legislation (a.k.a. university teachers' privilege — Law on Employees' Inventions). Quite similar to the US Bayh – Dole Act, organizational ownership of intellectual property rights (IPR) regime has been valid since the 1970s. In 2002, this regime has become a model of organizing IPR for university inventions as well. Since 1971 MPG has also a well-established tradition of technology transfer through a dedicated technology transfer office (Max Planck Innovation, MPG – TTO) to promote technology transfer, and provide guidance e.g. patenting, licensing and venture creation (Buenstorf 2006 ) . Its primary aim is the transfer of patented and non-patented technologies developed by Max Planck Institutes to industry and to negotiate and close license agreements. $^10$
+
+The survey was conducted in the last part of 2007 at 67 institutes (out of 80 institutes) specialized in different scientific disciplines and located different cities in Germany. The MPG is funded to large extent by both the federal and state governments. Although the aim is to conduct basic research in the interest of general public in natural sciences, life sciences, social sciences and the humanities; the institutes takes up new and innovative ideas that the German universities are not in a position to conduct adequately. By providing equipments, facilities the research at the MPG complements the work done at the universities. Currently the MPG has 4,300 scientists and substantial amount of graduate students, post-docs and guests scientists. 51 % come from abroad. In 2006, the budget was around 1,379.1 million euros. 82 % is from federal and state governments, while 13 % is from projects supported by government, federal states and the EU. Donations, evaluation royalties etc. amount to 5 % . The MPG has not only a strong scientific base, but a wellestablished tradition of technology transfer, as well as has been a seedbed for technological developments.
+
+### 3.1 Industrial co-operations
+
+Industry often is interested to collaborate on a particular research area with a Max Planck Institute or to further develop a licensed invention in collaboration with the Max Planck Institute. The respective cooperation agreements usually provide for grants that allow the institute to carry out the developmental work. However it should be ensured that the subject of the collaboration is sufficiently narrowed down so that it does not comprise the whole research area of the department involved; that the freedom of the institute to publish is guaranteed (taking into account the interest of the collaboration partner); that inventions made by scientists of the MPG within the scope of the collaboration are owned by the MPG, and not by the industry partner (usually, the industrial partner is granted an option
+
+8 http://www.max-planck-innovation.de/en/inventors_founders/inventors_faq/#08.
+
+9 http://www.max-planck-innovation.de/en/industry/services_industry/.
+
+10 Source: Max Planck Innovation Website. http://www.max-planck-innovation.de/en/.
+
+![Figure](figures/figure_010.png)
+
+---
+
+410
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+for a license with fair terms), and that license fees are not waived in exchange for contributions in kind or research grants. Although MPG provides support for the closing of a consultancy agreement, the formal responsibility lies with institute and the finance department of the MPG. In contrast to license agreements, which are closed by Max Planck Innovation (MPG–TTO), cooperation agreements are concluded between the industrial partner and the institute itself. Since 1979 Max Planck Innovation (MPG–TTO have closed more than 1,500 contracts with companies of all sizes and from all sectors, from start-ups to global corporations. About half of the profits originate in the US, the other half in Germany, Europe and Japan.
+
+### 3.2 Invention disclosures and patenting $^{11}$
+
+According to Max Planck Society (MPG) employment contracts and also to the Employees' Inventions Act, all occupational findings or ideas that may have inventive character must be reported to the institute's management. Inventions made by MPG staff members usually emerge within the scope of their research activities or are based on the institute's experience or work. These inventions are thus called “ employee inventions ” . In accordance with the German Employees' Inventions Act, the employer, i.e. the Max Planck Society, is entitled to such inventions — in so far as the Max Planck Society claims them under the stipulations of the act. The claim will be examined and lodged within 4 months after the filing of the invention disclosure form. Once the Max Planck Innovation (MPG – TTO), has filed a patent application, priority is ensured under patent law and there is usually no obstacle to scientific publication.
+
+Scientists of the MPG are obliged to publish the results of their research as soon as possible, but they should try to protect their invention by filing for patent applications prior to publication. Due to the fact that such publications such as conference posters, abstracts, proceedings, handouts or masters and PhD theses etc., are damaging to novelty. They endanger later IP protection in relation to the inventive step. Therefore, scientists should contact Max Planck Innovation (MPG–TTO), prior to the publication of any research results that seem — or whose further development seems — to be commercially viable.
+
+After receiving the invention disclosure form, MPG–TTO examine whether the invention is likely to lead to a successful patent application and MPG–TTO evaluate the commercial potential. To this end, they conduct patent searches and market research. If the evaluation is positive—and after clearance with the inventor(s) and the Max Planck Institute, who meet the costs of the application—MPG–TTO instruct an independent patent attorney experienced in the relevant field to draw up the patent application. They aim to burden the inventor(s) as little as possible with this process. In general, the rough draft of a planned publication is a sufficient basis for a patent application. Subsequently, inventor(s) will receive an outline of the application for review and will be asked to answer any unresolved questions sometimes with a patent attorney.
+
+11 As we expect complementarities between patenting and publishing activities, i.e. most patentable research is also publishable we do not go into the details on the publication activities of the scientists. As the MPG-survey was conducted anonymously, we don't have the names—other type of personal information about the identity of the scientists—which could have been used to identify their publication rate. But as the previous literature confirmed most patentable research is also publishable and MPG scientists are expected to publish we are expecting these scientists are also quite active in publishing.
+
+![Figure](figures/figure_008.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+411
+
+Furthermore, the inventor must, to the best of his ability, support the Max Planck Society in its efforts to apply for and commercialize his invention. As a rule, scientists' complementary know-how is necessary to enable a future licensee to realize the economic potential of a product based on the invention. According to the current MPG regulations inventors generally stand to receive up to 30 % of the gross license income that MPG – TTO receives from the commercialization of the IP or know-how the scientists created. This compensation exceeds the minimum rates of indemnification for employee inventions provided for by guidelines currently in force in private industry and in the public sector, and is intended to motivate scientists to participate actively in technology transfer.
+
+## 4 Data characteristics, variables of interest and methodology
+
+This paper is based on a large-scale survey of over 2,500 scientists in Germany aimed at obtaining information about the commercialization activities. 12 The scientists pooled for this research are from the independent German non-profit research organization — the Max Planck Society for the Advancement of Science (MPG hereafter). The survey was conducted by a professional consultancy company TNS-EMNID from October 2007 till December 2007. It was a telephone-based survey and names of the participants were kept confidential and are not to be revealed. Previous studies on technology transfer, academic entrepreneurship and available interview guides and questionnaires were consulted before constructing the survey. To check for possible interpretation errors and mistakes, pilot surveys were conducted with randomly contacted scientists from other public research organizations in Germany. The survey has four parts in which, the first part is about invention, patenting and research cooperation activities. Second part focuses on entrepreneurial activities. The third part is about the perceptions of scientists on commercial activities in general. The final part deals with individual and professional demographic information (age, gender, academic title and education, citizenship).
+
+We chose inventions disclosed to MPG-innovation (TTO) and patents applied for as our measure of a scientist's involvement in commercialization activity. Disclosure is a process by which scientists inform the TTO that they have developed an invention that they believe has the potential for commercial applications. It is then the responsibility of the TTO and the institute where the researcher is employed to either pursue IP protection or decline the disclosure. While employment contracts at MPG mandate disclosure to the TTO, in practice the process may turn into a voluntary activity where scientists take the initiative to inform the TTO of their inventions. Under these conditions, the micro-motives and social factors identified earlier become salient influences on the scientist's disclosure and patenting activities.
+
+### 4.1 Empirical strategy
+
+In order to construct the variables we first concentrate on the variable of interest — patenting and invention disclosure. We use three groups of scientists to measure the relationship between likelihood of scientists' patenting activities and their expectations concerning the outcomes of commercial activities. Scientists who have only applied for a patent; scientists who have only disclosed inventions to the MPG and do not have a patent
+
+12 Survey tool can be made available upon request.
+
+![Figure](figures/figure_009.png)
+
+---
+
+412
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+and scientists who have both disclosed inventions to the MPG and also have applied for a patent.
+
+We further investigate if there is any difference regarding reputation and financial rewards between the scientists who patented and who had only disclosed their inventions to their employees. Inventing and patenting are two separable phenomena. It is accepted that not every invention can be patentable, even if scientists may have the expectations to patent. Investigating and comparing what are the perceptions of scientists who patented and who made invention disclosures to their employees would shed some light on the current debate on the role and ownership of IP at universities and (PROs).
+
+As mentioned before, we concentrate on the expectations of scientists who do not have any form collaboration with industry (joint projects, direct consultancy etc.). It is important to make this distinction due to some reasons. The choice to collaborate with academia may be an initiative driven by firm's objectives. Blind et al. ( 2006 ) showed that German firms collaborate with academia for several strategic reasons. Mainly firm's expect the patents generated from collaboration to leverage their own knowledge as well as their positions in negotiations with partners, suppliers and the financial sector. Therefore, the scientific outcomes from a collaborative effort may result in patents mainly due to the firm's interests, rather than of the scientists. Moreover concentrating on non-collaborating scientists gives us a chance to isolate the sample to those who have never been involved actively in commercial activities and would be an ideal sample to test our hypotheses: what would drive those scientists to patent who were never before involved in commercialization activities?
+
+Since our interest was also to control the demographic nature of the respondents, we have used age, gender (female or not), foreign-born scientist variables. We further utilize data on their industry experience, MPG experience, the position (whether a director, a group leader, a post doctoral fellow), and which field of science do they belong. In order to clearly track the patenting and invention disclosure behavior one has to also account for the personal opinions of the scientists with respect to the nature and mode of commercialization. Scientists were therefore asked if they want their research to be open (free from exclusion) and if they think a technology transfer office (TTO) is indeed needed to take their research to industry or commercialize it in any other fashion. We utilize this information in order to account for the personal opinion of scientists about commercialization in general that may affect their actual commercialization behavior.
+
+The group of studies that focuses on individuals is inspired partly by psychology and behavioral sciences. These studies have focused on the socio-demographic characteristics of inventors. Macdonald 1984 , 1986 , Sirilli 1987 , Amesse et al. 1991 , Klofsten and JonesEvans 2000 , investigated the characteristics, background and socio-demographic features of inventors. Stephan and Levin ( 2005 ) investigated whether personal characteristics, age (life-cycle), citizenship status, gender and receipt of federal funding were related to patenting behaviors. They found little evidence of age effects, yet they found that tenured scientists are more likely to patent than non-tenured ones (Levin and Stephan 1991 ; Stephan 1996 ) . Women patent less than men, although the effect is smaller since the number of women employed in science and engineering fields relative to men is low. The sociodemographic findings of these different studies are fairly consistent (see also Azoulay et al. 2007 ) . Inventors were most often men; the average age being between 45 and 48. They were highly educated and had technical and commercial knowledge and had experience above the average.
+
+![Figure](figures/figure_007.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+413
+
+In addition to the individual (socio-demographic) factors, we also account for the perceptions of scientists on the nature of their research (and scientific field), on the use of knowledge (whether research should be open) and commercialization is not found to be proper. Thursby et al. ( 2001 ) argued that scientists who specialize in basic research may not disclose because they are unwilling to spend time on the applied R & D required to interest business in licensing invention. Thursby et al. ( 2001 ) also stated scientists may not disclose because they believe that commercial activity is not appropriate for an academic scientists. Having this kind of perception or believing in the Mertonian norms of `disinterestedness' — scientists would perceive that their research results should be freely accessible to any other scientists and businesses. Such scientists are also expected to be less interested in patenting or other commercial activities.
+
+Additionally the role of organizational factors, like the need for technology transfer office (TTO) to help scientists in their commercial activities may also influence the likelihood of their patenting activities. Owen-Smith and Powell ( 2001 , 2003 ) found that scientists’ incentives to be involved in technology transfer are magnified or minimized by the perceived costs of interacting with industry, TTOs, or dealing with patenting, licensing and company formation individually. Faculty decisions towards commercialization are shaped by the institutional and organizational environments which are supportive or oppositional for university-industry technology transfer. We therefore control for scientists' perception on the role of technology transfer offices.
+
+The set up for the econometric model therefore is of a multinomial discrete choice model; specifically we use the multinomial logit estimation method. Measuring perceptions is a tricky issue. Since our main propositions are on reputation and money there are many ways that we could measure it. The scientists were asked whether they expect commercialization (patenting, starting up a new venture, industrial collaboration, consulting services etc.) to increase their reputation basing on a 5 point scale. In the same vein, the question on whether they expect commercialization to make money was asked. Using these two measures we constructed variables — high money, high reputation if the respondents strongly agree with the prospects of getting money, or getting reputation.
+
+The underlying model can be formulated as follows:
+
+Patenting/disclosure Activity = f (expected rewards; age, gender, citizenship, career experience, research milieu)
+
+## 5 Estimation results and analysis
+
+This section puts forward some statistics indicating on the nature of data, the variables considered and the estimation results from the multinomial logit model. In order to test if the multinomial specification is suitable we conducted Wald tests for combining alternatives in multinomial specification. After conducting Wald tests, we performed the Small – Hsiao test for independence of irrelevant alternatives assumption. Tables 1 and 2 (in Appendix) report the test statistics. As can be observed, the Wald tests support the choice of multinomial logit method, while for two categories in the non-cooperator sample the IIA assumption is not satisfied. However, for the cooperators sample this problem does not seem to occur. This dissimilarity is mainly attributed to the method deficiencies in Small – Hsiao tests mentioned by Long and Freese ( 2006 ) who cite a Monte-Carlo study by Cheng and Long ( 2007 ) concluding that Small – Hsiao tests are not sufficient to test for IIA
+
+![Figure](figures/figure_009.png)
+
+---
+
+414
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+assumption especially when sample sizes are small. Therefore, we depend on the Wald test and also on the fact that our categories are mutually exclusively coded in order to formulate a multinomial logit specification to test our hypotheses.
+
+After the necessary data requirements for the paper (cooperators, non-cooperators etc.) we had almost 1,100 usable responses. Out of this sample, 110 scientists reported only patenting, 99 reported only disclosure and 187 reported both patenting and disclosure. Tables 3 and 4 (in Appendix) provides the descriptive statistics on the variables we consider. It can be clearly seen that most of the scientists take both paths of patenting and invention disclosure, but only few of them do it for money. It's also interesting to see that scientists who consider their research to be freely available for everyone also patent and disclose inventions to MPG. The mean ages for every mechanism is around 40 while less than a quarter of scientists patenting, disclosing or doing both, is female. Almost half of the foreign-born scientists patent and the number is almost the same for disclosure, but lesser for both.
+
+Directors show a very high patenting and disclosing behavior, if not for each of them individually. There is almost an equal share of scientists patenting in the broad fields of biology and medicine compared to chemistry, physics and other technical subjects. Postdocs and group leaders seem to show very high patenting and disclosure behavior. This may be because they need to show performance mainly after Ph.D. and therefore they might be more active in inventing and patenting. Given this scenario, we tested a multinomial logit model where all the three categories (only patent, only disclose, both patent and disclose) are considered. We provide the estimation results for both cooperators and non-cooperators sample. Table 5 provides the estimation results based on the non-cooperators sample and Table 6 (in Appendix) provides estimation results on the cooperators sample.
+
+### 5.1 Non-cooperators sample
+
+Based on our estimation results in Table 5 (in Appendix), we can observe that the scientists who expect high reputation from commercialization activities are more likely to perform both patenting as well as invention disclosure. This confirms our first hypothesis that scientists who expect high reputation are more likely to use both mechanisms. It can be interpreted as the scientists who would expect to have high reputation would signal it through patenting and disclosing their invention to reach the relevant audience who receive the signal. Secondly, we can see the effect is so strong that if scientists want reputation they do not necessarily take any one of the paths, but are very highly likely to take both.
+
+Is money driving the patenting and invention disclosure behavior then? The answer seems to be no. As can be seen in Table 5 (in Appendix) monetary expectations do not affect the patenting and invention disclosure activities of scientists. In the light of these results, our hypothesis that expectation of monetary gains affects patenting activity stands to be rejected. It is indeed reputation that drives these two and scientists may view achieving reputation more important than money. Academic interests might be of more value to the scientists than monetary interests and this might be driven by the inner philosophy of science and interest in basic research in order to solve the puzzle, answer the questions that are left unanswered and other motivations.
+
+This leads to the result on the perception of scientists on research as being `open'. Even though descriptive statistics show that there are a number of scientists who patent and disclose while having the view of open research, the estimation findings confirm
+
+![Figure](figures/figure_009.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+415
+
+their opinion. Scientists who consider research to be open are less likely to take any of the three paths to commercialization. Scientists who consider costs of commercialization to be high are less likely to disclose their inventions but are more likely to patent and disclose. If a scientist considers costs as high, she would not be willing to approach the MPG to disclose the invention in the first place whereas if the research has high potential (may be through reputation), it might be possible that the scientist is willing to both patent and disclose.
+
+Another interesting result is on the position variable. As a sequential process — the group leaders and directors are more likely to only disclose or take both paths. This might be possible due to the experience that each of these persons have by understanding the rules, regulations and institutional culture of the MPG (i.e. existence of organizational ownership of patents and an active TTO since 1970s). If the scientists respond that TTOs are indeed needed for commercialization then that positively affects the likelihood to only patent or take up both the paths. It is as well as due to the fact that the personal responsibilities towards disclosing inventions may grow over time. This is confirmed by the MPG experience variable, that scientists having higher number of years with the MPG are more likely to disclose their inventions to the MPG.
+
+On the demographic aspects it can observed that older scientists are more likely to patent and rather than only disclose their inventions or do both. Female scientists are less likely to choose both paths and gender does not have an affect on any one of these paths exclusively. The subject-area effects of scientists are taken into account too.
+
+### 5.2 Cooperators sample
+
+Table 6 (in Appendix) presents the results on the non-cooperators sample. As can be observed the non-cooperators are not affected by reputational expectations when patenting or disclosing their inventions. In fact, a striking result shows up on the monetary expectations variable. Scientists who expect monetary rewards to be high in commercialization are less likely to disclose or patent and disclose. A simple explanation can be found in the fact that they are cooperators with the firms. As mentioned before, firstly, the choice to collaborate with academia may be an initiative driven by firm's objectives. Blind et al. ( 2006 ) showed that German firms collaborate with academia for several strategic reasons. Mainly firm's expect the patents generated from collaboration to leverage their own positions in negotiations with partners, suppliers and the financial sector. Therefore, the scientific outcomes from a collaborative effort take shape of patents mainly because of the firm's interests, rather than of the scientists. Scientists in cooperation agreements would very well know this fact and therefore, if they are driven by monetary interests, they might not choose the patenting path and look for other paths such as start-up activities or new product development that can be commercialized directly from the labs.
+
+The results on open research seem to be also valid for cooperators, in that, the scientists who prefer their research to be openly available to others are less likely to patent or disclose. The need for TTOs affects the patenting activities positively. As with the cooperators, the group leaders are more likely to patent and disclose while the directors are more likely to only disclose and do both. Years of experience in Max Planck affects the likelihood of disclosures and choosing both paths positively. Age has a positive effect on patenting and female scientists are less likely to patent and disclose. Industry experience effects disclosure likelihood positively.
+
+![Figure](figures/figure_008.png)
+
+---
+
+416
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+## 6 Discussion and concluding remarks
+
+Understanding of scientists' patenting activities is still a recent phenomenon. Although the patenting activities of scientists (universities and public research organizations) have been seen as sources for innovation and economic development, concerns have also been raised that scientists are moving towards applied research and away from fundamental research in order to patent with the expectations of financial benefits. Many therefore argued that patenting may challenge the culture and norms of open science. However, despite the ongoing debates on the detrimental influences of patenting on scientific production and norms of science, why researchers patent has not until recently received the same amount of attention. Only recently some studies started to examine the incentives and motivations behind scientists' invention disclosure and patenting behaviors. This paper aims to open this discussion and interest further.
+
+In this paper we investigate to what extent financial benefits or reputation and recognition expected to result due to commercial activities influence the inventing and patenting activities of scientists. These assumptions have been debated concerning the context of industrial knowledge creation, protection, research and development (Schmookler 1966 ; Rosenberg 1974 ; Eisenberg 1989 ; Merges and Nelson 1990 , 1994 ; Long 2002 ; Cohen 2005 ; Thursby and Thursby 2007 ) . We discussed this tension (money or fame) specifically within the context of academic knowledge creation and from the perceptions of scientists and their decisions to make inventions disclosures and patenting. Instead of making a case for or against one factor, we investigated both aspects. By doing so, we move beyond the traditional argumentation of financial incentives matter for inventing activities while norms of science loses its ground due to increasing commercial activities. We found despite scientists' involvement in inventing and patenting activities, such activities are related to their traditional academic concerns i.e. gaining reputation and visibility than financial expectations. This paper thus also contributed to the debate on the role of IPR and commercial activities at the universities and public research organizations.
+
+We used a newly created survey data on 2,500 scientists from 67 different institutes of the Max Planck Society in Germany conducted in 2007 – 2008. To observe the effects of individual factors (expectations and commercial activities), our identification strategy involved studying two different samples of non-cooperating and cooperating scientists respectively, in relation to industry collaboration. This identification strategy gives a straightforward test for assessing the effects of motivations by isolating the sample that entirely concentrates on laboratory activities for academic purpose, and hence how they drive patenting activities.
+
+Empirically, we show that non-cooperating scientists who have more expectations to gain scientific reputation and visibility will more likely to patent. On the other hand scientists' commercialization activities do not necessarily respond to monetary expectations. Scientists' inventing activities are rather related to their expectations of recognition and reputation while financial benefits are less important. Specifically, the scientists who expect high reputation from commercialization activities are more likely to perform both patenting as well as invention disclosure. This confirms our first hypothesis that scientists who expect high reputation are more likely to use both mechanisms. It can be interpreted as the scientists who would expect to have high reputation would signal it through patenting and disclosing their invention to reach the relevant audience who receive the signal.
+
+![Figure](figures/figure_007.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+417
+
+The scientists involved in industrial cooperation however, seem to be not driven by reputational expectations and their patenting and disclosure activities might be more or less affected by the firm in context and its motives. The scientists may rather choose some other path of earning monetary gains than choose patenting because they expect patenting to benefit the firms.
+
+Invention disclosure and patenting activities could to a certain extent be independent from private economic incentives. It can be clearly seen that most of the scientists take both paths of patenting and invention disclosure, but only few of them do it in the expectation of gaining financial benefits. It's also interesting to see that scientists who consider their research to be freely available for everyone patent and disclose inventions to a lesser extent. Both of these can be viewed as information transfer mechanisms, not necessarily for monetary gains but for the non-monetary benefits — such as reputation — and prestige that the academic researchers foresee to be accrued. These findings are also important because it means that the despite patenting activities and traditional academic values seemed intact. Even when there are less or no financial expectations, commercial science provides additional scientific resources and participation in commercial science brings some kind of status and resources and that patents have become an integral part of faculty strategies for the dissemination of ideas and for signaling interest in commercial activities (see Murray and Graham 2007 ) . However this does not mean that the design of intellectual property rights, other forms of incentives (e.g. accepting patenting activities as an academic merit, qualification for promotion or providing research funds to patenting scientists), in academic organizations would not have effects on economic growth and productivity. Controlling for a variety of other determinants, including age, gender, citizenship, scientific discipline, industrial and academic experience, scientists with high reputation expectation from commercial activities will more likely to patent. We acknowledge that, these factors (reputation and financial rewards) are not mutually exclusive meaning that under certain conditions (in the long term) reputation and visibility of scientists may bring financial rewards maybe in the forms of research funds, if not personal gains.
+
+Even though scientists do no longer have a monk-like existence searching for truths about nature, scientist's involvement in entrepreneurial activity is not a transition from their academic roles to another. By doing so, we indicate that, for these individuals, the decision to participate in commercial activity is akin to managing multiple identities (Pratt and Foreman 2000 in George et al. 2005 ) in order to signal that they have multiple skills and knowledge (both academic and industrial); and they are able to better respond to a variety of situations. Scientists are establishing a unique set of experiences and values that are closely linked to their roles and academic career. We assume scientists can not or will not easily suspend these sets of values even if they consider being involved in commercialization activity. Therefore albeit scientists are encouraged to get involved in commercialization activities (e.g. Third Mission), the financial prospects of commercial activities is balanced against the cost of giving up norms and thus expected rewards associated with their identity as scientists. Moreover, despite rules and regulations scientists often possess high levels of discretion when contemplating their involvement with commercial activity. More generally, although a wider range of commercialization of academic research has becoming a component of scientists activities, expectations from such activities are bolted on to the traditional streams of research and what scientists generally value as a reward. Given this scenario, we suggest that these individuals are likely to embrace valued aspects of their existing role identity even as they enter the realm of commercialization. Especially in
+
+![Figure](figures/figure_005.png)
+
+---
+
+418
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+the context of Max Planck Society and in Germany, as scientists remain as academic and having an academic focus is typically more respected than scientists having a commercial focus.
+
+Acknowledgment The authors acknowledge the useful feedback from David B. Audretch, Werner Boente and Diemo Urbig.
+
+Open Access This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
+
+## Appendix
+
+See Tables 1, 2, 3, 4, 5 and 6.
+
+Table 1 Specification tests: Wald tests for combining alternatives in the multinomial specification
+
+<table><tr><td>H 0 : All coefficients except</td></tr><tr><td>intercepts associated with a given</td></tr><tr><td>pair of alternatives are 0 (i.e.,</td></tr><tr><td>alternatives can be combined)</td></tr><tr><td>Categories 0, 1, 2 and 3 denote</td></tr><tr><td>no-commercialization, only</td></tr><tr><td>patenting, only invention-</td></tr><tr><td>disclosure, patenting plus</td></tr><tr><td>disclosure respectively</td></tr></table>
+
+
+<table><tr><td>Alternatives tested</td><td> $x^{2}$ </td><td>df</td><td> $P&gt;\chi^{2}$ </td></tr><tr><td colspan="4">Non-cooperators</td></tr><tr><td>1-2</td><td>7.214</td><td>15</td><td>0.951</td></tr><tr><td>1-3</td><td>44.048</td><td>15</td><td>0.000</td></tr><tr><td>1-0</td><td>24.443</td><td>15</td><td>0.058</td></tr><tr><td>2-3</td><td>61.473</td><td>15</td><td>0.000</td></tr><tr><td>2-0</td><td>28.901</td><td>15</td><td>0.017</td></tr><tr><td>3-0</td><td>16,167.954</td><td>15</td><td>0.000</td></tr><tr><td colspan="4">Cooperators</td></tr><tr><td>1-2</td><td>61.611</td><td>15</td><td>0.000</td></tr><tr><td>1-3</td><td>95.007</td><td>15</td><td>0.000</td></tr><tr><td>1-0</td><td>38,438.235</td><td>15</td><td>0.000</td></tr><tr><td>2-3</td><td>24.894</td><td>15</td><td>0.051</td></tr><tr><td>2-0</td><td>40.584</td><td>15</td><td>0.000</td></tr><tr><td>3-0</td><td>109.824</td><td>15</td><td>0.000</td></tr></table>
+
+
+Table 2 Specification tests: small Hsiao-Tests for independence of irrelevant alternatives assumption (IIA)
+
+<table><tr><td>Omitted variable</td><td>ln L(full)</td><td>ln L(omit)</td><td> $x^{2}$ </td><td>df</td><td> $P&gt;\chi^{2}$ </td><td>Evidence</td></tr><tr><td colspan="7">Non-cooperators</td></tr><tr><td>1</td><td>−182.943</td><td>−133.349</td><td>99.188</td><td>32</td><td>0.000</td><td>Against H0</td></tr><tr><td>2</td><td>−159.450</td><td>−138.222</td><td>42.457</td><td>32</td><td>0.102</td><td>For Ho</td></tr><tr><td>3</td><td>−174.547</td><td>−134.674</td><td>79.745</td><td>32</td><td>0.000</td><td>Against H0</td></tr><tr><td colspan="7">Cooperators</td></tr><tr><td>1</td><td>−216.996</td><td>−199.398</td><td>35.197</td><td>32</td><td>0.319</td><td>For H0</td></tr><tr><td>2</td><td>−191.277</td><td>−174.655</td><td>33.244</td><td>32</td><td>0.406</td><td>For H0</td></tr><tr><td>3</td><td>−160.783</td><td>−142.865</td><td>35.836</td><td>32</td><td>0.293</td><td>For H0</td></tr></table>
+
+
+Ho: Odds(Outcome-J vs. Outcome-K) are independent of other alternatives Categories 0, 1, 2 and 3 denote no-commercialization, only patenting, only invention-disclosure, patenting plus disclosure respectively
+
+![Figure](figures/figure_013.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+419
+
+Table 3 Descriptive statistics on scientist patenting and invention disclosures
+
+<table><tr><td>Variable</td><td>Only patent (110)</td><td>Only invention disclosure (99)</td><td>Patent and disclosure (187)</td></tr><tr><td>High financial benefits (3-4 on a 5 point scale)</td><td>28</td><td>24</td><td>37</td></tr><tr><td>High reputation (3-4 on a 5 point scale)</td><td>52</td><td>45</td><td>85</td></tr><tr><td>Open research</td><td>66</td><td>69</td><td>106</td></tr><tr><td>Commercialization costs are high</td><td>78</td><td>69</td><td>153</td></tr><tr><td>TTOs are needed</td><td>94</td><td>75</td><td>148</td></tr><tr><td>Age (mean)</td><td>41</td><td>40</td><td>44</td></tr><tr><td>Female</td><td>27</td><td>28</td><td>22</td></tr><tr><td>Foreign-born</td><td>42</td><td>40</td><td>49</td></tr><tr><td>Post-Doc</td><td>38</td><td>22</td><td>32</td></tr><tr><td>Group leader</td><td>26</td><td>24</td><td>78</td></tr><tr><td>Director</td><td>5</td><td>8</td><td>26</td></tr><tr><td>MPG experience(mean years)</td><td>8.3</td><td>8.9</td><td>12.2</td></tr><tr><td>Industry experience(mean years)</td><td>1.1</td><td>1.2</td><td>0.7</td></tr><tr><td>Biology \&amp; Medicine</td><td>50</td><td>46</td><td>105</td></tr><tr><td>Chemistry/Physics and other technical subjects</td><td>58</td><td>49</td><td>80</td></tr></table>
+
+
+Table 4 Partial Correlations of MNL categories of patenting and invention activity with variables in the estimation
+
+<table><tr><td rowspan="2">Variable</td><td colspan="2">Overall sample</td><td colspan="2">Non-cooperators</td><td colspan="2">Cooperators</td></tr><tr><td>Corr.</td><td>Sig.</td><td>Corr.</td><td>Sig.</td><td>Corr.</td><td>Sig.</td></tr><tr><td>Reputation</td><td>0.0735</td><td>0.001</td><td>0.076</td><td>0.004</td><td>0.0516</td><td>0.180</td></tr><tr><td>Monetary incentives</td><td>-0.0946</td><td>0.000</td><td>-0.0211</td><td>0.428</td><td>-0.1773</td><td>0.000</td></tr><tr><td>Open research</td><td>-0.1293</td><td>0.000</td><td>-0.0828</td><td>0.002</td><td>-0.1225</td><td>0.001</td></tr><tr><td>High costs of commercialization</td><td>0.0271</td><td>0.215</td><td>0.0255</td><td>0.340</td><td>0.0215</td><td>0.577</td></tr><tr><td>Need for TTOs</td><td>0.0333</td><td>0.127</td><td>0.0617</td><td>0.021</td><td>0.0158</td><td>0.682</td></tr><tr><td>Post-doc</td><td>0.0398</td><td>0.069</td><td>0.0308</td><td>0.248</td><td>0.0525</td><td>0.173</td></tr><tr><td>Group leader</td><td>0.2168</td><td>0.000</td><td>0.1914</td><td>0.000</td><td>0.1911</td><td>0.000</td></tr><tr><td>Director</td><td>0.2019</td><td>0.000</td><td>0.1886</td><td>0.000</td><td>0.1816</td><td>0.000</td></tr><tr><td>MPG experience</td><td>0.1098</td><td>0.000</td><td>0.0434</td><td>0.104</td><td>0.1301</td><td>0.001</td></tr><tr><td>Foreigner</td><td>0.01</td><td>0.648</td><td>0.0354</td><td>0.185</td><td>0.0051</td><td>0.895</td></tr><tr><td>Age (log)</td><td>0.0152</td><td>0.488</td><td>-0.0001</td><td>0.997</td><td>0.0175</td><td>0.650</td></tr><tr><td>Female</td><td>-0.0626</td><td>0.004</td><td>-0.0306</td><td>0.251</td><td>-0.067</td><td>0.082</td></tr><tr><td>Industry experience</td><td>0.0596</td><td>0.006</td><td>0.0166</td><td>0.534</td><td>0.0856</td><td>0.026</td></tr></table>
+
+
+![Figure](figures/figure_006.png)
+
+---
+
+420
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+Table 5 Multinomial logit estimates of reputation and financial benefits on inventing and patenting Activities of scientists: sample 1: non-cooperators
+
+<table><tr><td>MNL categories</td><td>Only patenting</td><td>Only disclosure</td><td>Patenting and disclosure</td></tr><tr><td>Explanatory variables</td><td></td><td></td><td></td></tr><tr><td>Reputation</td><td>0.0934 (0.19)</td><td>0.0254 (0.18)</td><td>0.499*** (0.19)</td></tr><tr><td>Monetary expectations</td><td>-0.214 (0.18)</td><td>-0.134 (0.21)</td><td>-0.161 (0.21)</td></tr><tr><td>Open research</td><td>-0.153 (0.18)</td><td>-0.177 (0.21)</td><td>-0.536*** (0.18)</td></tr><tr><td>High costs of commercialization</td><td>-0.108 (0.18)</td><td>-0.245 (0.22)</td><td>0.545 (0.39)</td></tr><tr><td>Need for TTOs</td><td>0.423* (0.23)</td><td>0.0376 (0.24)</td><td>0.576** (0.29)</td></tr><tr><td>Post doctoral fellow</td><td>0.446 (0.35)</td><td>0.296 (0.53)</td><td>0.277 (0.53)</td></tr><tr><td>Group leader</td><td>0.793 (0.64)</td><td>1.944*** (0.65)</td><td>2.232*** (0.62)</td></tr><tr><td>Director</td><td>1.402 (0.85)</td><td>3.104*** (0.89)</td><td>3.246*** (1.11)</td></tr><tr><td>Years in Max Planck</td><td>-0.0478 (0.040)</td><td>-0.0224 (0.040)</td><td>0.0503 (0.035)</td></tr><tr><td>Foreign-born scientists</td><td>0.516 (0.37)</td><td>0.502 (0.42)</td><td>0.312 (0.37)</td></tr><tr><td>Age (log)</td><td>2.428*** (0.90)</td><td>-0.129 (1.37)</td><td>-0.00615 (1.63)</td></tr><tr><td>Female</td><td>0.0461 (0.40)</td><td>0.0972 (0.43)</td><td>-0.940* (0.50)</td></tr><tr><td>Years work in industry</td><td>0.102 (0.25)</td><td>0.365 (0.30)</td><td>0.183 (0.32)</td></tr><tr><td>Biology \&amp; Medicine</td><td>0.725 (0.82)</td><td>0.477 (0.91)</td><td>19.14*** (5.77)</td></tr><tr><td>Chemistry/Physics/Technical subjects</td><td>0.558 (0.83)</td><td>0.653 (0.86)</td><td>18.19*** (5.79)</td></tr><tr><td>Constant</td><td>-13.70*** (3.07)</td><td>-3.135 (4.89)</td><td>-25.88 (0)</td></tr><tr><td>Observations</td><td>1,418</td><td>1,418</td><td>1,418</td></tr><tr><td>Pseudo  $R^{2}$ </td><td>0.17</td><td></td><td></td></tr></table>
+
+
+Robust standard errors in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Reputation and Monetary incentive variables are mean centered. Odds-ratios reported
+
+Table 6 Multinomial logit estimates of reputation and financial benefits on inventing and patenting Activities of scientists: sample 2: cooperators
+
+<table><tr><td>MNL categories</td><td>Only patenting</td><td>Only disclosure</td><td>Patenting and disclosure</td></tr><tr><td>\midrule\textit{Explanatory variables}</td><td></td><td></td><td></td></tr><tr><td>Reputation expectation</td><td>0.0614 (0.14)</td><td>0.101 (0.15)</td><td>0.196 (0.12)</td></tr><tr><td>Monetary expectation</td><td>-0.244 (0.16)</td><td>-0.349** (0.17)</td><td>-0.535*** (0.12)</td></tr><tr><td>Open research</td><td>-0.418*** (0.14)</td><td>-0.169 (0.14)</td><td>-0.333*** (0.11)</td></tr><tr><td>High costs of commercialization</td><td>-0.166 (0.18)</td><td>-0.260 (0.22)</td><td>0.178 (0.19)</td></tr><tr><td>Need for TTOs</td><td>0.316* (0.19)</td><td>0.288 (0.18)</td><td>-0.0153 (0.13)</td></tr><tr><td>Post doctoral fellow</td><td>0.339 (0.38)</td><td>0.186 (0.40)</td><td>0.425 (0.32)</td></tr><tr><td>Group leader</td><td>0.188 (0.40)</td><td>0.207 (0.41)</td><td>1.440*** (0.31)</td></tr><tr><td>Director</td><td>0.361 (0.82)</td><td>1.224* (0.66)</td><td>2.252*** (0.56)</td></tr><tr><td>Years in Max Planck</td><td>-0.00893 (0.028)</td><td>0.0470* (0.026)</td><td>0.0531*** (0.020)</td></tr><tr><td>Foreign-born scientists</td><td>-0.123 (0.34)</td><td>-0.00732 (0.34)</td><td>0.0901 (0.27)</td></tr><tr><td>Age (log)</td><td>3.045*** (0.96)</td><td>1.044 (1.07)</td><td>0.413 (0.74)</td></tr><tr><td>Female</td><td>-0.0677 (0.38)</td><td>0.209 (0.36)</td><td>-0.761** (0.35)</td></tr><tr><td>Years work in industry</td><td>0.235 (0.21)</td><td>0.594*** (0.19)</td><td>0.252 (0.17)</td></tr><tr><td>Biology \&amp; Medicine</td><td>18.35*** (3.35)</td><td>0.324 (0.85)</td><td>1.363* (0.75)</td></tr></table>
+
+
+![Figure](figures/figure_007.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+421
+
+Table 6 continued
+
+<table><tr><td>MNL categories</td><td>Only patenting</td><td>Only disclosure</td><td>Patenting and disclosure</td></tr><tr><td>Chemistry/Physics/Technical subjects</td><td>18.62*** (3.32)</td><td>0.111 (0.85)</td><td>0.996 (0.76)</td></tr><tr><td>Constant</td><td>−30.88</td><td>−6.558</td><td>−4.615</td></tr><tr><td></td><td>(0)</td><td>(4.00)</td><td>(2.87)</td></tr><tr><td>Observations</td><td>689</td><td>689</td><td>689</td></tr><tr><td>Pseudo  $R^{2}$ </td><td>0.15</td><td></td><td></td></tr></table>
+
+
+Robust standard errors in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. Reputation and Monetary incentive variables are mean centered. Odds-ratios reported
+
+## References
+
+Agrawal, A., & Henderson, R. (2002). Putting patents in context: Exploring Knowledge transfer from MIT. Management Science, 48, 44-60. doi:10.1287/mnsc.48.1.44.14279.
+
+Allen, S. D., Link, A. N., & Rosenbaum, D. T. (2007). Entrepreneurship and human capital: Evidence of patenting activity from the academic sector. Entrepreneurship: Theory and Practice, 31, 937-951. doi: 10.1111/j.1540-6520.2007.00207.x.
+
+Amesse, F., Desranleau, C., Etemad, H., Fortier, Y., & Seguin-Dulude, L. (1991). The individual inventor and the role of entrepreneurship: A survey of the canadian evidence. Research Policy, 20, 13–27. doi: 10.1016/0048-7333(91)90081-Z.
+
+Anton, J. J., & Yao, D. Y. (2004). Little patents and big secrets: Managing intellectual property. RAND Journal of Economics. The RAND Corporation, 35, 1-22. doi:10.2307/1593727.
+
+Arrow, K. J. (1962). The economic implications of learning by doing. The Review of Economic Studies, 29, 155-173. doi:10.2307/2295952.
+
+Azoulay, P., Ding, W., & Stuart, T. (2007). The determinants of faculty patenting behavior: Demographics or opportunities? Journal of Economic Behavior & Organization, 63, 599-623. doi:10.1016/j.jebo.2006. 05.015.
+
+Baldini, N., Grimaldi, R., & Sobrero, M. (2007). To patent or not to patent? A survey of Italian inventors on motivations, incentives, and obstacles to university patenting. Scientometrics, 70, 333-354. doi: 10.1007/s11192-007-0206-5.
+
+Ben-David, J. (1972). American higher education (p. 137). New York: McGraw-Hill.
+
+Ben-David, J., & Sullivan, T. A. (1975). Sociology of science. Annual Review of Sociology, 1, 203-222. doi: 10.1146/annurev.so.01.080175.001223.
+
+Bercovitz, J., & Feldman, M. (2008). Academic entrepreneurs: Organizational change at the individual level. Organization Science, 19, 69-89. doi: 10.1287/orsc.1070.0295 .
+
+Blind, K., Edler, J., Frietsch, R., & Schmoch, U. (2006). Motives to patent: Empirical evidence from Germany. Research Policy, 35, 655-672. doi: 10.1016/j.respol.2006.03.002.
+
+Blume, S. S., & Sinclair, R. (1973). Chemists in British universities: A study of the reward system in science. American Sociological Review, 38, 126-138. doi:10.2307/2094336.
+
+Buenstorf, G. (2006). Is academic entrepreneurship good or bad for science? Empirical evidence from the Max Planck society. Discussion paper No.2006-17, Max Planck Institute of Economics.
+
+Cheng, S., & Long, J. S. (2007). Testing for IIA in the multinomial logit model. Sociological Methods & Research, 35, 583-600. doi: 10.1177/0049124106292361 .
+
+Cohen, W. (2005). Patents and appropriation: Concerns and evidence. The Journal of Technology Transfer, 30, 57-71. doi: 10.1007/s10961-004-4358-7.
+
+Cole, S., & Cole, J. R. (1967). Scientific output and recognition: A study in the operation of the reward system in science. American Sociological Review, 32, 377-390. doi:10.2307/2091085.
+
+Edwards, M. G., Murray, F., & Yu, R. (2006). Gold in the ivory tower: Equity rewards of out-licensing. Nature Biotechnology, 24(5), 509-515. doi:10.1038/nbt0506-509.
+
+Eisenberg, R. (1987). Proprietary rights and the norms of science in biotechnology research. The Yale Law Journal, 97, 177-231. doi:10.2307/796481.
+
+Eisenberg, R. (1989). Patents and the progress of science—Exclusive rights and experimental use. The University of Chicago Law Review. University of Chicago. Law School, 56, 1017–1086. doi: 10.2307/ 1599761 .
+
+![Figure](figures/figure_025.png)
+
+---
+
+422
+
+D. Göktepe-Hulten, P. Mahagaonkar
+
+Etzkowitz, H. (1998). The norms of entrepreneurial science: Cognitive effects of the new university-industry linkages. Research Policy, 27, 823-833. doi:10.1016/S0048-7333(98)00093-6.
+
+George, G., Jain, S., & Maltarich, M. A. (2005). Academics or entrepreneurs? Entrepreneurial identity and invention disclosure behavior of university scientists. SSRN eLibrary.
+
+Goktepe, D. (2008). Inside the ivory tower: Inventors and patents at Land university, PhD thesis.
+
+Gulbrandsen, M. (2005). But Peter's in it for the money-the liminality of entrepreneurial scientists. VEST Journal for Science and Technology Studies, 18(1-2), 49-75.
+
+Hellmann, T. (2007). The role of patents for bridging the science to market gap. Journal of Economic Behavior & Organization, 63, 624-647. doi: 10.1016/j.jebo.2006.05.013.
+
+Jensen, K., & Murray, F. (2005). The intellectual property landscape of the human genome. Science, 310(10), 239-240. doi: 10.1126/science.1120014.
+
+Jeon, D. S., & Menicucci, D. (2008). Money, fame and the allocation of talent: Brain drain and the institution of science. Journal of Economic Behavior & Organization, 66, 558-581. doi: 10.1016/ j.jebo.2006.06.012 .
+
+Klofsten, M., & Jones-Evans, D. (2000). Comparing academic entrepreneurship in Europe: The case of Sweden and Ireland. Small Business Economics, 14, 299-309. doi:10.1023/A:1008184601282.
+
+Lach, S, & Schankerman, M. (2003). Incentives and invention in universities. CEPR discussion papers 3916.
+
+Levin, S., & Stephan, P. (1991). Research productivity over the life cycle: Evidence for academic scientists. The American Economic Review, 81, 114-132.
+
+Link, A., & Siegel, D. S. (2007). Innovation, entrepreneurship, and technical change. Oxford: Oxford University Press.
+
+Long, C. (2002). Patent signals. The University of Chicago Law Review, 69, 625-679.
+
+Long, J. S., & Freese, J. (2006). Regression models for categorical dependent variables using STATA (2nd ed.). College Station: Stata Press.
+
+Macdonald, S. (1984). The patent system and the individual inventor. The inventor, 24, 25-29.
+
+Macdonald, S. (1986). The distinctive research of the individual inventor. Research Policy, 15, 199-210. doi: 10.1016/0048-7333(86)90015-6 .
+
+Markman, G. D., Gianiodis, P. T., Phan, P. H., & Balkin, D. B. (2004). Entrepreneurship from the ivory tower: Do incentive systems matter? The Journal of Technology Transfer, 29, 353-364. doi: 10.1023/B:JOTT.0000034127.01889.86.
+
+Merges, R., & Nelson, R. R. (1990). On the complex economics of patent scope. Columbia Law Review, 90, 839-916.
+
+Merges, R., & Nelson, R. R. (1994). On limiting or encouraging rivalry in technical progress: The effect of patent scope decisions. Journal of Economic Behavior and Organization, 25, 1-24.
+
+Merton, R. C. (1957). Priorities in scientific discovery: A chapter in the sociology of science. American Sociological Review, 22, 635-659. doi:10.2307/2089193.
+
+Merton, R. C. (1973). The normative structure of science'. In R. C. Merton (Ed.), The sociology of science. Chicago: University of Chicago Press.
+
+Meyer, M. (2005). Independent inventors and public support measures: Insights from 33 case studies in finland. World Patent Information, 27, 113-123.
+
+Murray, F., & Graham, L. (2007). Buying science & selling science: Gender stratification in commercial science. Industrial and Corporate Change, 16, 657-689. doi:10.1093/icc/dtm021.
+
+Owen-Smith, J., & Powell, W. (2001). Careers and contradictions: Faculty responses to the transformation of knowledge and its uses in the life sciences research. Sociology of Work, 10, 109–140. doi: 10.1016/ S0277-2833(01)80023-6 .
+
+Owen-Smith, J., & Powell, W. (2003). The expanding role of university patenting in the life sciences: Assessing the importance of experience and connectivity. Research Policy, 32, 1695–1711. doi: 10.1016/S0048-7333(03)00045-3.
+
+Phan, P., & Siegel, D. S. (2006). The effectiveness of university technology transfer: Lessons learned, managerial and policy implications, and the road forward. Foundations and Trends in Entrepreneurship, 2, 77–144. doi: 10.1561/0300000006 .
+
+Pratt, M. G., & Foreman, P. O. (2000). Classifying managerial responses to multiple organizational identities. Academy of Management Review, 25, 18-42. doi: 10.2307/259261.
+
+Rosenberg, N. (1974). Science, invention and economic growth. The Economic Journal, 84, 90-108. doi: 10.2307/2230485.
+
+Rothermael, F. T., Agung, S. D., & Jian, L. (2007). University entrepreneurship: A taxonomy of the literature. Industrial and Corporate Change, 16, 691-792. doi:10.1093/icc/dtm023.
+
+Schmookler, J. (1966). Invention and economic growth. Cambridge, MA; London: Harvard University Press.
+
+![Figure](figures/figure_031.png)
+
+---
+
+Inventing and patenting activities of scientists
+
+423
+
+Siegel, D. S., & Phan, P. (2005). Analyzing the effectiveness of university technology transfer: Implications for entrepreneurship education. In G. Libecap (Ed.), Advances in the study of entrepreneurship, innovation, and economic growth (Vol. 16, pp. 1–38). Amsterdam: Elsevier Science/JAI Press.
+
+Siegel, D. S., Waldman, D., & Link, A. (2003). Assessing the impact of organizational practices on the relative productivity of university technology transfer offices: An exploratory study. Research Policy, 32 , 27–48. doi: 10.1016/S0048-7333(01)00196-2 .
+
+Siegel, D. S., Veugelers, R., & Wright, M. (2007). Technology transfer offices and commercialization of university intellectual property: Performance and policy implications. Oxford Review of Economic Policy, 23 (4), 640–660. doi: 10.1093/oxrep/grm036 .
+
+Sirilli, G. (1987). Patents and inventors: An empirical study. Research Policy, 16, 157-174. doi:10.1016/ 0048-7333(87)90029-1.
+
+Slaughter, S., & Leslie, L. L. (1997). Academic capitalism: Politics, policies, and the entrepreneurial university. Baltimore: Johns Hopkins University Press.
+
+Stephan, P. (1996). The economics of science. Journal of Economic Literature, 34, 1199-1235.
+
+Stephan, P., & Levin, S. (1992). Striking the mother lode in science: The importance of age, place, and time. New York: Oxford University Press.
+
+Stephan, P. E., & Levin, S. G. (2005). Leaving careers in IT: Gender difference in retention. The Journal of Technology Transfer, 30, 383-396. doi:10.1007/s10961-005-2583-3.
+
+Thursby, J. G., & Thursby, M. C. (2007). University licensing. Oxford Review of Economic Policy, 23, 620639. doi:10.1093/oxrep/grm031.
+
+Thursby, J. G., Jensen, R., & Thursby, M. C. (2001). Objectives, Characteristics and Outcomes of University Licensing: A Survey of Major U.S. Universities. The Journal of Technology Transfer, 26, 59-72. doi: 10.1023/A:1007884111883.
+
+Van Looy, B., Callaert, J., & Debackere, K. (2006). Publication and patent behavior of academic researchers: Conflicting, reinforcing or merely co-existing? Research Policy, 35, 596-608. doi: 10.1016/j.respol. 2006.02.003 .
+
+![Figure](figures/figure_013.png)
+