Hugging Face's logo

Datasets:
chuxuan
/
REPRO-Bench

Tasks:
Image-Text-to-Text
Languages:
English
ArXiv:
Tags:
reproducibility
social-science
agents
Dataset card Files
xet
Community
1
REPRO-Bench / 30 /replication_package /Log /AppendixC_simulations_May2021.log
anonymous-submission-acl2025's picture
anonymous-submission-acl2025
add 30
57fe0a1
raw
history blame contribute delete
68.3 kB
 ---------------------------------------------------------------------------------------------------------------------------------
name: <unnamed>
log: C:/Users/paserman/Dropbox/Research/GenderCooperativeness/EJ/3 replication package/Log/AppendixC_simulations_May2021.
> log
log type: text
opened on: 10 May 2021, 18:23:16
.
.
. cap prog drop rd_sim
. prog def rd_sim, rclass
1. version 15.1
2. syntax [, nobs(integer 10000) beta_a(real 1.0) beta_b(real 1.0) rho_x(real 0.7) /*
> */ zbar_R(real 0.5) zbar_D(real -0.5) /*
> */ alpha_R(real 0.5) alpha_D(real 0.5) /*
> */ kappa_ksi_R(real 0.0) beta_ksi_R(real 10.0) kappa_ksi_D(real 0.0) beta_ksi_D(real 10.0) /*
> */ phi0_R(real -1.0) phi1_R(real -1.0) phi0_D(real -1.0) phi1_D(real -1.0) /*
> */ kappa_u(real 0.0) beta_u(real 1.0) /*
> */ gamma0(real 0.0) gamma1(real -5.0) gamma2(real 0.0) gamma3(real 0.0)/*
> */ tau0(real 0.3) tau1(real -1.0) tau2(real 0.0)]
3. drop _all
4. set obs `nobs'
5.
. * Overall district ideology
. gen z = 2*(rbeta(`beta_a',`beta_b')-0.5)
6. gen x = z + (sqrt((1-`rho_x'^2)/`rho_x'^2))*(2*(rbeta(`beta_a',`beta_b')-0.5))
7.
. * Ideology of R and candidates: weighted average of national party and local ideology, plus noise
. gen z_R = `alpha_R'*`zbar_R' + (1-`alpha_R')*z + `kappa_ksi_R'*(rbeta(`beta_ksi_R',`beta_ksi_R')-0.5)
8. gen z_D = `alpha_D'*`zbar_D' + (1-`alpha_D')*z + `kappa_ksi_D'*(rbeta(`beta_ksi_D',`beta_ksi_D')-0.5)
9.
. * Gender is correlated with candidate ideology
. gen byte female_D = rnormal(`phi0_D' + `phi1_D'*z_D)>0
10. gen byte female_R = rnormal(`phi0_R' + `phi1_R'*z_R)>0
11.
. * Voteshare depends on ideology of the candidates plus noise
. gen u = `kappa_u'*(rbeta(`beta_u', `beta_u')-0.5)
12. gen voteshare_D = (exp(`gamma0' + `gamma1'*(z - (z_D+z_R)/2) + `gamma2'*female_D - `gamma3'*female_R + u)/ /*
> */ (1+ exp(`gamma0' + `gamma1'*(z - (z_D+z_R)/2) + `gamma2'*female_D - `gamma3'*female_R + u )))
13.
. gen voteshare_female = voteshare_D if female_D==1 & female_R==0
14. replace voteshare_female = (1-voteshare_D) if female_D==0 & female_R==1
15.
. * Outcome: depends on who is elected
. gen y = `tau0' + `tau1'*abs(z_D) + `tau2'*female_D + rnormal() if voteshare_D>=0.5
16. replace y = `tau0' + `tau1'*abs(z_R) +`tau2'*female_R + rnormal() if voteshare_D<0.5
17.
. * Now four types of RD analyses
. * (1) Density test
. rddensity voteshare_female, c(0.5)
18. local denstest_pval_all = e(pv_q)
19.
. rddensity voteshare_female if voteshare_D>=0.5, c(0.5)
20. local denstest_pval_D = e(pv_q)
21.
. rddensity voteshare_female if voteshare_D<0.5, c(0.5)
22. local denstest_pval_R = e(pv_q)
23.
. * (2) is ideology continuous at the threshold
. rdrobust z voteshare_female, c(0.5) kernel(uniform)
24. mat b = e(b)
25. mat V = e(V)
26. local b_ideology_all = b[1,1]
27. local se_ideology_all = sqrt(V[1,1])
28.
. rdrobust z voteshare_female if voteshare_D>=0.5, c(0.5) kernel(uniform)
29. mat b = e(b)
30. mat V = e(V)
31. local b_ideology_D = b[1,1]
32. local se_ideology_D = sqrt(V[1,1])
33.
. rdrobust z voteshare_female if voteshare_D<0.5, c(0.5) kernel(uniform)
34. mat b = e(b)
35. mat V = e(V)
36. local b_ideology_R = b[1,1]
37. local se_ideology_R = sqrt(V[1,1])
38.
. * (3) Estimate treatment effect with simple RD
. rdrobust y voteshare_female, c(0.5) kernel(uniform)
39. mat b = e(b)
40. mat V = e(V)
41. local b_rd_all = b[1,1]
42. local se_rd_all = sqrt(V[1,1])
43. local band_all = e(h_l)
44.
. rdrobust y voteshare_female if voteshare_D>=0.5, c(0.5) kernel(uniform)
45. mat b = e(b)
46. mat V = e(V)
47. local b_rd_D = b[1,1]
48. local se_rd_D = sqrt(V[1,1])
49. local band_D = e(h_l)
50.
. rdrobust y voteshare_female if voteshare_D<0.5, c(0.5) kernel(uniform)
51. mat b = e(b)
52. mat V = e(V)
53. local b_rd_R = b[1,1]
54. local se_rd_R = sqrt(V[1,1])
55. local band_R = e(h_l)
56.
. * (4-5) Estimate the treatment effect with weighted RD
. gen byte female = female_D if voteshare_D>=0.5
57. replace female = female_R if voteshare_D<0.5
58. gen voteshare_female_adj = voteshare_female-0.5
59.
. * (4) using x
. probit female x if abs(voteshare_female_adj)<=`band_all'
60. predict pscore if e(sample)==1
61. gen wt =1/pscore if female==1
62. replace wt = 1/(1-pscore) if female==0
63. reg y female voteshare_female_adj i.female#c.voteshare_female_adj [aw=wt] if abs(voteshare_female_adj)<=`band_all'
64. local b_rdwt_all = _b[female]
65. local se_rdwt_all = _se[female]
66. drop pscore wt
67.
. probit female x if voteshare_D>=0.5 & abs(voteshare_female_adj)<=`band_D'
68. predict pscore if e(sample)==1
69. gen wt =1/pscore if female==1
70. replace wt = 1/(1-pscore) if female==0
71. reg y female voteshare_female_adj i.female#c.voteshare_female_adj [aw=wt] if voteshare_D>=0.5 & abs(voteshare_female_adj
> )<=`band_D'
72. local b_rdwt_D = _b[female]
73. local se_rdwt_D = _se[female]
74. drop pscore wt
75.
. probit female x if voteshare_D<0.5 & abs(voteshare_female_adj)<=`band_R'
76. predict pscore if e(sample)==1
77. gen wt =1/pscore if female==1
78. replace wt = 1/(1-pscore) if female==0
79. reg y female voteshare_female_adj i.female#c.voteshare_female_adj [aw=wt] if voteshare_D<0.5 & abs(voteshare_female_adj)
> <=`band_R'
80. local b_rdwt_R = _b[female]
81. local se_rdwt_R = _se[female]
82. drop pscore wt
83.
.
.
. * (5a) using ideology of the district
. probit female z if abs(voteshare_female_adj)<=`band_all'
84. predict pscore if e(sample)==1
85. gen wt =1/pscore if female==1
86. replace wt = 1/(1-pscore) if female==0
87. reg y female voteshare_female_adj i.female#c.voteshare_female_adj [aw=wt] if abs(voteshare_female_adj)<=`band_all'
88. local b_rdwtideodistrict_all = _b[female]
89. local se_rdwtideodistrict_all = _se[female]
90. drop pscore wt
91.
. probit female z if voteshare_D>=0.5 & abs(voteshare_female_adj)<=`band_D'
92. predict pscore if e(sample)==1
93. gen wt =1/pscore if female==1
94. replace wt = 1/(1-pscore) if female==0
95. reg y female voteshare_female_adj i.female#c.voteshare_female_adj [aw=wt] if voteshare_D>=0.5 & abs(voteshare_female_adj
> )<=`band_D'
96. local b_rdwtideodistrict_D = _b[female]
97. local se_rdwtideodistrict_D = _se[female]
98. drop pscore wt
99.
. probit female z if voteshare_D<0.5 & abs(voteshare_female_adj)<=`band_R'
100. predict pscore if e(sample)==1
101. gen wt =1/pscore if female==1
102. replace wt = 1/(1-pscore) if female==0
103. reg y female voteshare_female_adj i.female#c.voteshare_female_adj [aw=wt] if voteshare_D<0.5 & abs(voteshare_female_adj)
> <=`band_R'
104. local b_rdwtideodistrict_R = _b[female]
105. local se_rdwtideodistrict_R = _se[female]
106. drop pscore wt
107.
. * (5b) using ideology of the elected representative
. gen z_elected = z_D if voteshare_D>=0.5
108. replace z_elected = z_R if voteshare_D<0.5
109.
. probit female z_elected if abs(voteshare_female_adj)<=`band_all'
110. predict pscore if e(sample)==1
111. gen wt =1/pscore if female==1
112. replace wt = 1/(1-pscore) if female==0
113. reg y female voteshare_female_adj i.female#c.voteshare_female_adj [aw=wt] if abs(voteshare_female_adj)<=`band_all'
114. local b_rdwtideoelected_all = _b[female]
115. local se_rdwtideoelected_all = _se[female]
116. drop pscore wt
117.
. probit female z_elected if voteshare_D>=0.5 & abs(voteshare_female_adj)<=`band_D'
118. predict pscore if e(sample)==1
119. gen wt =1/pscore if female==1
120. replace wt = 1/(1-pscore) if female==0
121. reg y female voteshare_female_adj i.female#c.voteshare_female_adj [aw=wt] if voteshare_D>=0.5 & abs(voteshare_female_adj
> )<=`band_D'
122. local b_rdwtideoelected_D = _b[female]
123. local se_rdwtideoelected_D = _se[female]
124. drop pscore wt
125.
. probit female z_elected if voteshare_D<0.5 & abs(voteshare_female_adj)<=`band_R'
126. predict pscore if e(sample)==1
127. gen wt =1/pscore if female==1
128. replace wt = 1/(1-pscore) if female==0
129. reg y female voteshare_female_adj i.female#c.voteshare_female_adj [aw=wt] if voteshare_D<0.5 & abs(voteshare_female_adj)
> <=`band_R'
130. local b_rdwtideoelected_R = _b[female]
131. local se_rdwtideoelected_R = _se[female]
132. drop pscore wt
133.
. * (6-7) Propensity score methods
. gen absMV = abs(voteshare_D-0.5)
134.
. * (6a) pscore - x
. cap teffects ipw (y) (female x absMV, probit), pstolerance(1e-6) osample(osample)
135. teffects ipw (y) (female x absMV, probit) if osample==0, pstolerance(1e-6)
136. drop osample
137. mat b = e(b)
138. mat V = e(V)
139. local b_pscorex_all = b[1,1]
140. local se_pscorex_all = sqrt(V[1,1])
141.
. cap teffects ipw (y) (female x absMV, probit) if voteshare_D>=0.5, pstolerance(1e-6) osample(osample)
142. teffects ipw (y) (female x absMV, probit) if voteshare_D>=0.5 & osample==0, pstolerance(1e-6)
143. drop osample
144. mat b = e(b)
145. mat V = e(V)
146. local b_pscorex_D = b[1,1]
147. local se_pscorex_D = sqrt(V[1,1])
148.
. cap teffects ipw (y) (female x absMV, probit) if voteshare_D<0.5, pstolerance(1e-6) osample(osample)
149. teffects ipw (y) (female x absMV, probit) if voteshare_D<0.5 & osample==0, pstolerance(1e-6)
150. drop osample
151. mat b = e(b)
152. mat V = e(V)
153. local b_pscorex_R = b[1,1]
154. local se_pscorex_R = sqrt(V[1,1])
155.
.
. * (7a) pscore - district ideology
. cap teffects ipw (y) (female z absMV, probit), pstolerance(1e-6) osample(osample)
156. teffects ipw (y) (female z absMV, probit) if osample==0, pstolerance(1e-6)
157. drop osample
158. mat b = e(b)
159. mat V = e(V)
160. local b_pscoreideodistrict_all = b[1,1]
161. local se_pscoreideodistrict_all = sqrt(V[1,1])
162.
. cap teffects ipw (y) (female z absMV, probit) if voteshare_D>=0.5, pstolerance(1e-6) osample(osample)
163. teffects ipw (y) (female z absMV, probit) if voteshare_D>=0.5 & osample==0, pstolerance(1e-6)
164. drop osample
165. mat b = e(b)
166. mat V = e(V)
167. local b_pscoreideodistrict_D = b[1,1]
168. local se_pscoreideodistrict_D = sqrt(V[1,1])
169.
. cap teffects ipw (y) (female z absMV, probit) if voteshare_D<0.5, pstolerance(1e-6) osample(osample)
170. teffects ipw (y) (female z absMV, probit) if voteshare_D<0.5 & osample==0, pstolerance(1e-6)
171. drop osample
172. mat b = e(b)
173. mat V = e(V)
174. local b_pscoreideodistrict_R = b[1,1]
175. local se_pscoreideodistrict_R = sqrt(V[1,1])
176.
.
. * (7b) pscore - elected representative ideology
. cap teffects ipw (y) (female z_elected absMV, probit), pstolerance(1e-6) osample(osample)
177. teffects ipw (y) (female z_elected absMV, probit) if osample==0, pstolerance(1e-6)
178. drop osample
179. mat b = e(b)
180. mat V = e(V)
181. local b_pscoreideoelected_all = b[1,1]
182. local se_pscoreideoelected_all = sqrt(V[1,1])
183.
. cap teffects ipw (y) (female z_elected absMV, probit) if voteshare_D>=0.5, pstolerance(1e-6) osample(osample)
184. teffects ipw (y) (female z_elected absMV, probit) if voteshare_D>=0.5 & osample==0, pstolerance(1e-6)
185. drop osample
186. mat b = e(b)
187. mat V = e(V)
188. local b_pscoreideoelected_D = b[1,1]
189. local se_pscoreideoelected_D = sqrt(V[1,1])
190.
. cap teffects ipw (y) (female z_elected absMV, probit) if voteshare_D<0.5, pstolerance(1e-6) osample(osample)
191. teffects ipw (y) (female z_elected absMV, probit) if voteshare_D<0.5 & osample==0, pstolerance(1e-6)
192. drop osample
193. mat b = e(b)
194. mat V = e(V)
195. local b_pscoreideoelected_R = b[1,1]
196. local se_pscoreideoelected_R = sqrt(V[1,1])
197.
.
. * (8) OLS
. reg y female
198. local b_ols_all =_b[female]
199. local se_ols_all = _se[female]
200.
. reg y female if voteshare_D>=0.5
201. local b_ols_D = _b[female]
202. local se_ols_D = _se[female]
203.
. reg y female if voteshare_D<0.5
204. local b_ols_R = _b[female]
205. local se_ols_R = _se[female]
206.
. * (9) Return
. return scalar denstest_pval_all = `denstest_pval_all'
207. return scalar denstest_pval_D = `denstest_pval_D'
208. return scalar denstest_pval_R = `denstest_pval_R'
209.
. return scalar b_ideology_all = `b_ideology_all'
210. return scalar se_ideology_all = `se_ideology_all'
211. return scalar b_ideology_D = `b_ideology_D'
212. return scalar se_ideology_D = `se_ideology_D'
213. return scalar b_ideology_R = `b_ideology_R'
214. return scalar se_ideology_R = `se_ideology_R'
215.
. return scalar b_rd_all = `b_rd_all'
216. return scalar se_rd_all = `se_rd_all'
217. return scalar b_rd_D = `b_rd_D'
218. return scalar se_rd_D = `se_rd_D'
219. return scalar b_rd_R = `b_rd_R'
220. return scalar se_rd_R = `se_rd_R'
221.
. return scalar b_rdwt_all = `b_rdwt_all'
222. return scalar se_rdwt_all = `se_rdwt_all'
223. return scalar b_rdwt_D = `b_rdwt_D'
224. return scalar se_rdwt_D = `se_rdwt_D'
225. return scalar b_rdwt_R = `b_rdwt_R'
226. return scalar se_rdwt_R = `se_rdwt_R'
227.
. return scalar b_rdwtideodistrict_all = `b_rdwtideodistrict_all'
228. return scalar se_rdwtideodistrict_all = `se_rdwtideodistrict_all'
229. return scalar b_rdwtideodistrict_D = `b_rdwtideodistrict_D'
230. return scalar se_rdwtideodistrict_D = `se_rdwtideodistrict_D'
231. return scalar b_rdwtideodistrict_R = `b_rdwtideodistrict_R'
232. return scalar se_rdwtideodistrict_R = `se_rdwtideodistrict_R'
233.
. return scalar b_rdwtideoelected_all = `b_rdwtideoelected_all'
234. return scalar se_rdwtideoelected_all = `se_rdwtideoelected_all'
235. return scalar b_rdwtideoelected_D = `b_rdwtideoelected_D'
236. return scalar se_rdwtideoelected_D = `se_rdwtideoelected_D'
237. return scalar b_rdwtideoelected_R = `b_rdwtideoelected_R'
238. return scalar se_rdwtideoelected_R = `se_rdwtideoelected_R'
239.
. return scalar b_pscorex_all = `b_pscorex_all'
240. return scalar se_pscorex_all = `se_pscorex_all'
241. return scalar b_pscorex_D = `b_pscorex_D'
242. return scalar se_pscorex_D = `se_pscorex_D'
243. return scalar b_pscorex_R = `b_pscorex_R'
244. return scalar se_pscorex_R = `se_pscorex_R'
245.
. return scalar b_pscoreideodistrict_all = `b_pscoreideodistrict_all'
246. return scalar se_pscoreideodistrict_all = `se_pscoreideodistrict_all'
247. return scalar b_pscoreideodistrict_D = `b_pscoreideodistrict_D'
248. return scalar se_pscoreideodistrict_D = `se_pscoreideodistrict_D'
249. return scalar b_pscoreideodistrict_R = `b_pscoreideodistrict_R'
250. return scalar se_pscoreideodistrict_R = `se_pscoreideodistrict_R'
251.
. return scalar b_pscoreideoelected_all = `b_pscoreideoelected_all'
252. return scalar se_pscoreideoelected_all = `se_pscoreideoelected_all'
253. return scalar b_pscoreideoelected_D = `b_pscoreideoelected_D'
254. return scalar se_pscoreideoelected_D = `se_pscoreideoelected_D'
255. return scalar b_pscoreideoelected_R = `b_pscoreideoelected_R'
256. return scalar se_pscoreideoelected_R = `se_pscoreideoelected_R'
257.
. return scalar b_ols_all = `b_ols_all'
258. return scalar se_ols_all = `se_ols_all'
259. return scalar b_ols_D = `b_ols_D'
260. return scalar se_ols_D = `se_ols_D'
261. return scalar b_ols_R = `b_ols_R'
262. return scalar se_ols_R = `se_ols_R'
263.
. end
.
.
. ********************************************************************************
. ********************************************************************************
. ********************************************************************************
.
. * Now actually run the simulations
.
. set seed 1234567
.
. * Run one simulation as a test
. rd_sim, beta_a(5) beta_b(5) kappa_ksi_R(0.4) kappa_ksi_D(0.4) kappa_u(1) tau1(-5) rho_x(0.6) nobs(100000) /*
> */ gamma2(0.0) gamma3(0.6) phi1_D(-1) phi1_R(-1)
number of observations (_N) was 0, now 100,000
(79,715 missing values generated)
(8,263 real changes made)
(53,038 missing values generated)
(53,038 real changes made)
Computing data-driven bandwidth selectors.
Point estimates and standard errors have been adjusted for repeated observations.
(Use option nomasspoints to suppress this adjustment.)
RD Manipulation test using local polynomial density estimation.
c = 0.500 | Left of c Right of c Number of obs = 28548
-------------------+---------------------- Model = unrestricted
Number of obs | 11425 17123 BW method = comb
Eff. Number of obs | 4542 4450 Kernel = triangular
Order est. (p) | 2 2 VCE method = jackknife
Order bias (q) | 3 3
BW est. (h) | 0.096 0.083
Running variable: voteshare_female.
------------------------------------------
Method | T P>|T|
-------------------+----------------------
Robust | -0.2327 0.8160
------------------------------------------
P-values of binomial tests. (H0: prob = .5)
-----------------------------------------------------
Window Length / 2 | <c >=c | P>|T|
-------------------+----------------------+----------
0.000 | 5 15 | 0.0414
0.000 | 19 29 | 0.1934
0.001 | 31 40 | 0.3425
0.001 | 42 54 | 0.2615
0.001 | 51 69 | 0.1203
0.001 | 64 83 | 0.1374
0.002 | 69 97 | 0.0358
0.002 | 83 112 | 0.0447
0.002 | 100 127 | 0.0842
0.002 | 115 132 | 0.3086
-----------------------------------------------------
Computing data-driven bandwidth selectors.
Point estimates and standard errors have been adjusted for repeated observations.
(Use option nomasspoints to suppress this adjustment.)
RD Manipulation test using local polynomial density estimation.
c = 0.500 | Left of c Right of c Number of obs = 13774
-------------------+---------------------- Model = unrestricted
Number of obs | 2457 11317 BW method = comb
Eff. Number of obs | 1615 6442 Kernel = triangular
Order est. (p) | 2 2 VCE method = jackknife
Order bias (q) | 3 3
BW est. (h) | 0.141 0.173
Running variable: voteshare_female.
------------------------------------------
Method | T P>|T|
-------------------+----------------------
Robust | 9.3618 0.0000
------------------------------------------
P-values of binomial tests. (H0: prob = .5)
-----------------------------------------------------
Window Length / 2 | <c >=c | P>|T|
-------------------+----------------------+----------
0.000 | 5 15 | 0.0414
0.000 | 19 29 | 0.1934
0.001 | 31 40 | 0.3425
0.001 | 42 54 | 0.2615
0.001 | 51 69 | 0.1203
0.001 | 64 83 | 0.1374
0.002 | 69 97 | 0.0358
0.002 | 83 112 | 0.0447
0.002 | 100 127 | 0.0842
0.002 | 115 132 | 0.3086
-----------------------------------------------------
Computing data-driven bandwidth selectors.
Point estimates and standard errors have been adjusted for repeated observations.
(Use option nomasspoints to suppress this adjustment.)
RD Manipulation test using local polynomial density estimation.
c = 0.500 | Left of c Right of c Number of obs = 14774
-------------------+---------------------- Model = unrestricted
Number of obs | 8968 5806 BW method = comb
Eff. Number of obs | 6647 2312 Kernel = triangular
Order est. (p) | 2 2 VCE method = jackknife
Order bias (q) | 3 3
BW est. (h) | 0.211 0.148
Running variable: voteshare_female.
------------------------------------------
Method | T P>|T|
-------------------+----------------------
Robust | -12.2100 0.0000
------------------------------------------
P-values of binomial tests. (H0: prob = .5)
-----------------------------------------------------
Window Length / 2 | <c >=c | P>|T|
-------------------+----------------------+----------
0.000 | 5 15 | 0.0414
0.000 | 19 29 | 0.1934
0.001 | 31 40 | 0.3425
0.001 | 42 54 | 0.2615
0.001 | 51 69 | 0.1203
0.001 | 64 83 | 0.1374
0.002 | 69 97 | 0.0358
0.002 | 83 112 | 0.0447
0.002 | 100 127 | 0.0842
0.002 | 115 132 | 0.3086
-----------------------------------------------------
Sharp RD estimates using local polynomial regression.
Cutoff c = .5 | Left of c Right of c Number of obs = 28548
-------------------+---------------------- BW type = mserd
Number of obs | 11425 17123 Kernel = Uniform
Eff. Number of obs | 2282 2410 VCE method = NN
Order est. (p) | 1 1
Order bias (q) | 2 2
BW est. (h) | 0.046 0.046
BW bias (b) | 0.092 0.092
rho (h/b) | 0.497 0.497
Outcome: z. Running variable: voteshare_female.
--------------------------------------------------------------------------------
Method | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------------+------------------------------------------------------------
Conventional | .01855 .00924 2.0073 0.045 .000437 .036665
Robust | - - 2.0807 0.037 .001262 .042242
--------------------------------------------------------------------------------
Sharp RD estimates using local polynomial regression.
Cutoff c = .5 | Left of c Right of c Number of obs = 13774
-------------------+---------------------- BW type = mserd
Number of obs | 2457 11317 Kernel = Uniform
Eff. Number of obs | 882 2485 VCE method = NN
Order est. (p) | 1 1
Order bias (q) | 2 2
BW est. (h) | 0.065 0.065
BW bias (b) | 0.118 0.118
rho (h/b) | 0.552 0.552
Outcome: z. Running variable: voteshare_female.
--------------------------------------------------------------------------------
Method | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------------+------------------------------------------------------------
Conventional | .2107 .00961 21.9190 0.000 .191862 .229544
Robust | - - 18.9995 0.000 .191055 .235007
--------------------------------------------------------------------------------
Sharp RD estimates using local polynomial regression.
Cutoff c = .5 | Left of c Right of c Number of obs = 14774
-------------------+---------------------- BW type = mserd
Number of obs | 8968 5806 Kernel = Uniform
Eff. Number of obs | 2933 1286 VCE method = NN
Order est. (p) | 1 1
Order bias (q) | 2 2
BW est. (h) | 0.083 0.083
BW bias (b) | 0.151 0.151
rho (h/b) | 0.551 0.551
Outcome: z. Running variable: voteshare_female.
--------------------------------------------------------------------------------
Method | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------------+------------------------------------------------------------
Conventional | -.19744 .0082 -24.0813 0.000 -.213511 -.181372
Robust | - - -20.4594 0.000 -.214516 -.177009
--------------------------------------------------------------------------------
Sharp RD estimates using local polynomial regression.
Cutoff c = .5 | Left of c Right of c Number of obs = 28548
-------------------+---------------------- BW type = mserd
Number of obs | 11425 17123 Kernel = Uniform
Eff. Number of obs | 4751 5448 VCE method = NN
Order est. (p) | 1 1
Order bias (q) | 2 2
BW est. (h) | 0.101 0.101
BW bias (b) | 0.190 0.190
rho (h/b) | 0.530 0.530
Outcome: y. Running variable: voteshare_female.
--------------------------------------------------------------------------------
Method | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------------+------------------------------------------------------------
Conventional | .17272 .04322 3.9964 0.000 .088013 .257427
Robust | - - 3.2304 0.001 .063486 .25937
--------------------------------------------------------------------------------
Sharp RD estimates using local polynomial regression.
Cutoff c = .5 | Left of c Right of c Number of obs = 13774
-------------------+---------------------- BW type = mserd
Number of obs | 2457 11317 Kernel = Uniform
Eff. Number of obs | 880 2484 VCE method = NN
Order est. (p) | 1 1
Order bias (q) | 2 2
BW est. (h) | 0.065 0.065
BW bias (b) | 0.117 0.117
rho (h/b) | 0.555 0.555
Outcome: y. Running variable: voteshare_female.
--------------------------------------------------------------------------------
Method | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------------+------------------------------------------------------------
Conventional | .31188 .07973 3.9117 0.000 .15561 .468147
Robust | - - 3.3740 0.001 .131626 .496528
--------------------------------------------------------------------------------
Sharp RD estimates using local polynomial regression.
Cutoff c = .5 | Left of c Right of c Number of obs = 14774
-------------------+---------------------- BW type = mserd
Number of obs | 8968 5806 Kernel = Uniform
Eff. Number of obs | 3622 1619 VCE method = NN
Order est. (p) | 1 1
Order bias (q) | 2 2
BW est. (h) | 0.104 0.104
BW bias (b) | 0.184 0.184
rho (h/b) | 0.566 0.566
Outcome: y. Running variable: voteshare_female.
--------------------------------------------------------------------------------
Method | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------------+------------------------------------------------------------
Conventional | .38413 .06763 5.6797 0.000 .251574 .516685
Robust | - - 4.6972 0.000 .218021 .530242
--------------------------------------------------------------------------------
(53,038 missing values generated)
(53,038 real changes made)
(71,452 missing values generated)
Iteration 0: log likelihood = -7045.573
Iteration 1: log likelihood = -7014.2341
Iteration 2: log likelihood = -7014.2334
Iteration 3: log likelihood = -7014.2334
Probit regression Number of obs = 10,199
LR chi2(1) = 62.68
Prob > chi2 = 0.0000
Log likelihood = -7014.2334 Pseudo R2 = 0.0044
------------------------------------------------------------------------------
female | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
x | -.2260739 .0286094 -7.90 0.000 -.2821473 -.1700006
_cons | .0707624 .0125909 5.62 0.000 .0460847 .09544
------------------------------------------------------------------------------
(option pr assumed; Pr(female))
(89,801 missing values generated)
(94,552 missing values generated)
(4,751 real changes made)
(sum of wgt is 20,398.6936738491)
Source | SS df MS Number of obs = 10,199
-------------+---------------------------------- F(3, 10195) = 51.96
Model | 193.062182 3 64.3540607 Prob > F = 0.0000
Residual | 12626.5639 10,195 1.23850554 R-squared = 0.0151
-------------+---------------------------------- Adj R-squared = 0.0148
Total | 12819.6261 10,198 1.25707258 Root MSE = 1.1129
-----------------------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------+----------------------------------------------------------------
female | .1739132 .0439874 3.95 0.000 .0876893 .2601372
voteshare_female_adj | 3.538998 .5419161 6.53 0.000 2.476736 4.60126
|
female#c.voteshare_female_adj |
1 | -6.258399 .7649045 -8.18 0.000 -7.757763 -4.759036
|
_cons | -1.052096 .0305361 -34.45 0.000 -1.111952 -.9922388
-----------------------------------------------------------------------------------------------
Iteration 0: log likelihood = -1933.3477
Iteration 1: log likelihood = -1878.7605
Iteration 2: log likelihood = -1878.6371
Iteration 3: log likelihood = -1878.6371
Probit regression Number of obs = 3,364
LR chi2(1) = 109.42
Prob > chi2 = 0.0000
Log likelihood = -1878.6371 Pseudo R2 = 0.0283
------------------------------------------------------------------------------
female | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
x | .5759299 .055828 10.32 0.000 .4665091 .6853507
_cons | .720536 .0252009 28.59 0.000 .6711433 .7699288
------------------------------------------------------------------------------
(option pr assumed; Pr(female))
(96,636 missing values generated)
(97,516 missing values generated)
(880 real changes made)
(sum of wgt is 6,733.80004513264)
Source | SS df MS Number of obs = 3,364
-------------+---------------------------------- F(3, 3360) = 40.81
Model | 139.293765 3 46.4312551 Prob > F = 0.0000
Residual | 3822.60101 3,360 1.13767887 R-squared = 0.0352
-------------+---------------------------------- Adj R-squared = 0.0343
Total | 3961.89478 3,363 1.17808349 Root MSE = 1.0666
-----------------------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------+----------------------------------------------------------------
female | .2406377 .0730094 3.30 0.001 .0974903 .3837851
voteshare_female_adj | 6.792886 1.412536 4.81 0.000 4.023369 9.562403
|
female#c.voteshare_female_adj |
1 | -9.439579 1.989195 -4.75 0.000 -13.33973 -5.539424
|
_cons | -1.285744 .0501982 -25.61 0.000 -1.384166 -1.187322
-----------------------------------------------------------------------------------------------
Iteration 0: log likelihood = -3240.1233
Iteration 1: log likelihood = -3133.1627
Iteration 2: log likelihood = -3132.9107
Iteration 3: log likelihood = -3132.9107
Probit regression Number of obs = 5,241
LR chi2(1) = 214.43
Prob > chi2 = 0.0000
Log likelihood = -3132.9107 Pseudo R2 = 0.0331
------------------------------------------------------------------------------
female | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
x | -.629661 .0437244 -14.40 0.000 -.7153593 -.5439628
_cons | -.5177479 .0184832 -28.01 0.000 -.5539742 -.4815216
------------------------------------------------------------------------------
(option pr assumed; Pr(female))
(94,759 missing values generated)
(98,381 missing values generated)
(3,622 real changes made)
(sum of wgt is 10,475.5064911842)
Source | SS df MS Number of obs = 5,241
-------------+---------------------------------- F(3, 5237) = 80.33
Model | 290.368848 3 96.7896161 Prob > F = 0.0000
Residual | 6310.12588 5,237 1.20491233 R-squared = 0.0440
-------------+---------------------------------- Adj R-squared = 0.0434
Total | 6600.49473 5,240 1.2596364 Root MSE = 1.0977
-----------------------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------+----------------------------------------------------------------
female | .3231791 .0614265 5.26 0.000 .2027576 .4436006
voteshare_female_adj | 4.05668 .722778 5.61 0.000 2.639734 5.473627
|
female#c.voteshare_female_adj |
1 | -5.84145 1.017686 -5.74 0.000 -7.836538 -3.846362
|
_cons | -.8815513 .0424921 -20.75 0.000 -.9648534 -.7982492
-----------------------------------------------------------------------------------------------
Iteration 0: log likelihood = -7045.573
Iteration 1: log likelihood = -6830.0923
Iteration 2: log likelihood = -6829.9941
Iteration 3: log likelihood = -6829.9941
Probit regression Number of obs = 10,199
LR chi2(1) = 431.16
Prob > chi2 = 0.0000
Log likelihood = -6829.9941 Pseudo R2 = 0.0306
------------------------------------------------------------------------------
female | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
z | -1.538549 .0747452 -20.58 0.000 -1.685046 -1.392051
_cons | -.0209974 .0136641 -1.54 0.124 -.0477786 .0057838
------------------------------------------------------------------------------
(option pr assumed; Pr(female))
(89,801 missing values generated)
(94,552 missing values generated)
(4,751 real changes made)
(sum of wgt is 20,472.7413574457)
Source | SS df MS Number of obs = 10,199
-------------+---------------------------------- F(3, 10195) = 70.48
Model | 262.88843 3 87.6294767 Prob > F = 0.0000
Residual | 12675.1194 10,195 1.24326821 R-squared = 0.0203
-------------+---------------------------------- Adj R-squared = 0.0200
Total | 12938.0078 10,198 1.2686809 Root MSE = 1.115
-----------------------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------+----------------------------------------------------------------
female | .1943662 .0436797 4.45 0.000 .1087454 .2799871
voteshare_female_adj | 3.946968 .5368838 7.35 0.000 2.894571 4.999366
|
female#c.voteshare_female_adj |
1 | -6.460541 .7641782 -8.45 0.000 -7.958481 -4.962601
|
_cons | -1.069798 .0301335 -35.50 0.000 -1.128866 -1.010731
-----------------------------------------------------------------------------------------------
Iteration 0: log likelihood = -1933.3477
Iteration 1: log likelihood = -1286.2133
Iteration 2: log likelihood = -1267.5265
Iteration 3: log likelihood = -1267.4815
Iteration 4: log likelihood = -1267.4815
Probit regression Number of obs = 3,364
LR chi2(1) = 1331.73
Prob > chi2 = 0.0000
Log likelihood = -1267.4815 Pseudo R2 = 0.3444
------------------------------------------------------------------------------
female | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
z | 7.444313 .2623512 28.38 0.000 6.930114 7.958512
_cons | 1.78103 .0537058 33.16 0.000 1.675768 1.886291
------------------------------------------------------------------------------
(option pr assumed; Pr(female))
(96,636 missing values generated)
(97,516 missing values generated)
(880 real changes made)
(sum of wgt is 6,608.81980001926)
Source | SS df MS Number of obs = 3,364
-------------+---------------------------------- F(3, 3360) = 112.97
Model | 469.699399 3 156.566466 Prob > F = 0.0000
Residual | 4656.54364 3,360 1.38587608 R-squared = 0.0916
-------------+---------------------------------- Adj R-squared = 0.0908
Total | 5126.24304 3,363 1.52430658 Root MSE = 1.1772
-----------------------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------+----------------------------------------------------------------
female | -.793492 .0774302 -10.25 0.000 -.9453071 -.6416768
voteshare_female_adj | 26.62825 1.637998 16.26 0.000 23.41667 29.83982
|
female#c.voteshare_female_adj |
1 | -31.34425 2.256755 -13.89 0.000 -35.769 -26.9195
|
_cons | -.2816657 .0479712 -5.87 0.000 -.3757214 -.1876099
-----------------------------------------------------------------------------------------------
Iteration 0: log likelihood = -3240.1233
Iteration 1: log likelihood = -2172.2326
Iteration 2: log likelihood = -2152.4686
Iteration 3: log likelihood = -2152.4172
Iteration 4: log likelihood = -2152.4172
Probit regression Number of obs = 5,241
LR chi2(1) = 2175.41
Prob > chi2 = 0.0000
Log likelihood = -2152.4172 Pseudo R2 = 0.3357
------------------------------------------------------------------------------
female | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
z | -7.049271 .1911968 -36.87 0.000 -7.42401 -6.674532
_cons | -.7631799 .0241643 -31.58 0.000 -.8105411 -.7158187
------------------------------------------------------------------------------
(option pr assumed; Pr(female))
(94,759 missing values generated)
(98,381 missing values generated)
(3,622 real changes made)
(sum of wgt is 9,903.25282597542)
Source | SS df MS Number of obs = 5,241
-------------+---------------------------------- F(3, 5237) = 22.54
Model | 81.2014589 3 27.067153 Prob > F = 0.0000
Residual | 6288.68375 5,237 1.20081798 R-squared = 0.0127
-------------+---------------------------------- Adj R-squared = 0.0122
Total | 6369.8852 5,240 1.21562695 Root MSE = 1.0958
-----------------------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------+----------------------------------------------------------------
female | -.1379332 .0656739 -2.10 0.036 -.2666815 -.0091849
voteshare_female_adj | 5.627017 .7120137 7.90 0.000 4.231173 7.022861
|
female#c.voteshare_female_adj |
1 | -7.189718 1.03427 -6.95 0.000 -9.217319 -5.162118
|
_cons | -.680452 .0396242 -17.17 0.000 -.758132 -.6027719
-----------------------------------------------------------------------------------------------
(53,038 missing values generated)
(53,038 real changes made)
Iteration 0: log likelihood = -7045.573
Iteration 1: log likelihood = -6040.1188
Iteration 2: log likelihood = -6037.2237
Iteration 3: log likelihood = -6037.2236
Probit regression Number of obs = 10,199
LR chi2(1) = 2016.70
Prob > chi2 = 0.0000
Log likelihood = -6037.2236 Pseudo R2 = 0.1431
------------------------------------------------------------------------------
female | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
z_elected | -1.958926 .0449516 -43.58 0.000 -2.04703 -1.870823
_cons | .0135668 .0133481 1.02 0.309 -.0125949 .0397285
------------------------------------------------------------------------------
(option pr assumed; Pr(female))
(89,801 missing values generated)
(94,552 missing values generated)
(4,751 real changes made)
(sum of wgt is 20,703.5746251345)
Source | SS df MS Number of obs = 10,199
-------------+---------------------------------- F(3, 10195) = 156.58
Model | 586.245291 3 195.415097 Prob > F = 0.0000
Residual | 12723.9437 10,195 1.24805725 R-squared = 0.0440
-------------+---------------------------------- Adj R-squared = 0.0438
Total | 13310.189 10,198 1.3051764 Root MSE = 1.1172
-----------------------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------+----------------------------------------------------------------
female | .3725144 .044218 8.42 0.000 .2858384 .4591904
voteshare_female_adj | 3.852766 .5269349 7.31 0.000 2.81987 4.885662
|
female#c.voteshare_female_adj |
1 | -6.090878 .7661041 -7.95 0.000 -7.592593 -4.589164
|
_cons | -1.201598 .0296283 -40.56 0.000 -1.259675 -1.14352
-----------------------------------------------------------------------------------------------
Iteration 0: log likelihood = -1933.3477
Iteration 1: log likelihood = -1587.7928
Iteration 2: log likelihood = -1582.0926
Iteration 3: log likelihood = -1582.0776
Iteration 4: log likelihood = -1582.0776
Probit regression Number of obs = 3,364
LR chi2(1) = 702.54
Prob > chi2 = 0.0000
Log likelihood = -1582.0776 Pseudo R2 = 0.1817
------------------------------------------------------------------------------
female | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
z_elected | 7.410054 .3114873 23.79 0.000 6.79955 8.020558
_cons | 3.042828 .1071286 28.40 0.000 2.83286 3.252796
------------------------------------------------------------------------------
(option pr assumed; Pr(female))
(96,636 missing values generated)
(97,516 missing values generated)
(880 real changes made)
(sum of wgt is 6,813.62821555138)
Source | SS df MS Number of obs = 3,364
-------------+---------------------------------- F(3, 3360) = 44.01
Model | 168.894906 3 56.2983019 Prob > F = 0.0000
Residual | 4298.15433 3,360 1.2792126 R-squared = 0.0378
-------------+---------------------------------- Adj R-squared = 0.0369
Total | 4467.04924 3,363 1.32829296 Root MSE = 1.131
-----------------------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------+----------------------------------------------------------------
female | -.4509505 .0758094 -5.95 0.000 -.5995877 -.3023132
voteshare_female_adj | 16.0325 1.518248 10.56 0.000 13.05572 19.00929
|
female#c.voteshare_female_adj |
1 | -19.80716 2.132745 -9.29 0.000 -23.98877 -15.62555
|
_cons | -.6497326 .0491586 -13.22 0.000 -.7461164 -.5533488
-----------------------------------------------------------------------------------------------
Iteration 0: log likelihood = -3240.1233
Iteration 1: log likelihood = -2594.5471
Iteration 2: log likelihood = -2583.7661
Iteration 3: log likelihood = -2583.7327
Iteration 4: log likelihood = -2583.7327
Probit regression Number of obs = 5,241
LR chi2(1) = 1312.78
Prob > chi2 = 0.0000
Log likelihood = -2583.7327 Pseudo R2 = 0.2026
------------------------------------------------------------------------------
female | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
z_elected | -7.805294 .2423385 -32.21 0.000 -8.280269 -7.33032
_cons | 1.29987 .0573838 22.65 0.000 1.1874 1.412341
------------------------------------------------------------------------------
(option pr assumed; Pr(female))
(94,759 missing values generated)
(98,381 missing values generated)
(3,622 real changes made)
(sum of wgt is 10,372.1911814213)
Source | SS df MS Number of obs = 5,241
-------------+---------------------------------- F(3, 5237) = 16.72
Model | 60.8606207 3 20.2868736 Prob > F = 0.0000
Residual | 6354.50433 5,237 1.21338635 R-squared = 0.0095
-------------+---------------------------------- Adj R-squared = 0.0089
Total | 6415.36495 5,240 1.22430629 Root MSE = 1.1015
-----------------------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------+----------------------------------------------------------------
female | -.1872573 .0632442 -2.96 0.003 -.3112422 -.0632723
voteshare_female_adj | 5.046656 .7215085 6.99 0.000 3.632198 6.461113
|
female#c.voteshare_female_adj |
1 | -5.702727 1.02625 -5.56 0.000 -7.714604 -3.690849
|
_cons | -.6934209 .0411359 -16.86 0.000 -.7740645 -.6127773
-----------------------------------------------------------------------------------------------
Iteration 0: EE criterion = 2.162e-28
Iteration 1: EE criterion = 4.110e-33
Treatment-effects estimation Number of obs = 100,000
Estimator : inverse-probability weights
Outcome model : weighted mean
Treatment model: probit
------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ATE |
female |
(1 vs 0) | .185551 .009143 20.29 0.000 .1676311 .2034709
-------------+----------------------------------------------------------------
POmean |
female |
0 | -1.51444 .0040466 -374.25 0.000 -1.522371 -1.506509
------------------------------------------------------------------------------
Iteration 0: EE criterion = 5.545e-18
Iteration 1: EE criterion = 2.033e-33
Treatment-effects estimation Number of obs = 46,962
Estimator : inverse-probability weights
Outcome model : weighted mean
Treatment model: probit
------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ATE |
female |
(1 vs 0) | -.0620271 .0113967 -5.44 0.000 -.0843642 -.03969
-------------+----------------------------------------------------------------
POmean |
female |
0 | -1.499856 .0061075 -245.58 0.000 -1.511827 -1.487886
------------------------------------------------------------------------------
Iteration 0: EE criterion = 2.364e-23
Iteration 1: EE criterion = 1.407e-32
Treatment-effects estimation Number of obs = 53,038
Estimator : inverse-probability weights
Outcome model : weighted mean
Treatment model: probit
------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ATE |
female |
(1 vs 0) | .4934254 .015274 32.30 0.000 .4634889 .5233619
-------------+----------------------------------------------------------------
POmean |
female |
0 | -1.519821 .0053757 -282.72 0.000 -1.530357 -1.509284
------------------------------------------------------------------------------
Iteration 0: EE criterion = 7.198e-22
Iteration 1: EE criterion = 6.756e-33
Treatment-effects estimation Number of obs = 100,000
Estimator : inverse-probability weights
Outcome model : weighted mean
Treatment model: probit
------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ATE |
female |
(1 vs 0) | .1960373 .0094185 20.81 0.000 .1775775 .2144971
-------------+----------------------------------------------------------------
POmean |
female |
0 | -1.527443 .004071 -375.20 0.000 -1.535422 -1.519464
------------------------------------------------------------------------------
Iteration 0: EE criterion = 4.047e-18
Iteration 1: EE criterion = 2.312e-32
Treatment-effects estimation Number of obs = 46,962
Estimator : inverse-probability weights
Outcome model : weighted mean
Treatment model: probit
------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ATE |
female |
(1 vs 0) | -.0158669 .0108021 -1.47 0.142 -.0370386 .0053048
-------------+----------------------------------------------------------------
POmean |
female |
0 | -1.51214 .0060289 -250.82 0.000 -1.523956 -1.500323
------------------------------------------------------------------------------
Iteration 0: EE criterion = 2.425e-27
Iteration 1: EE criterion = 1.682e-31
Treatment-effects estimation Number of obs = 52,919
Estimator : inverse-probability weights
Outcome model : weighted mean
Treatment model: probit
------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ATE |
female |
(1 vs 0) | -.1946729 .0624661 -3.12 0.002 -.3171042 -.0722417
-------------+----------------------------------------------------------------
POmean |
female |
0 | -1.45934 .0055571 -262.61 0.000 -1.470232 -1.448448
------------------------------------------------------------------------------
Iteration 0: EE criterion = 1.047e-24
Iteration 1: EE criterion = 3.917e-33
Treatment-effects estimation Number of obs = 100,000
Estimator : inverse-probability weights
Outcome model : weighted mean
Treatment model: probit
------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ATE |
female |
(1 vs 0) | .2416772 .0092623 26.09 0.000 .2235235 .2598309
-------------+----------------------------------------------------------------
POmean |
female |
0 | -1.519808 .0040542 -374.88 0.000 -1.527754 -1.511862
------------------------------------------------------------------------------
Iteration 0: EE criterion = 3.676e-18
Iteration 1: EE criterion = 4.790e-33
Treatment-effects estimation Number of obs = 46,962
Estimator : inverse-probability weights
Outcome model : weighted mean
Treatment model: probit
------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ATE |
female |
(1 vs 0) | -.000495 .0106006 -0.05 0.963 -.0212719 .0202818
-------------+----------------------------------------------------------------
POmean |
female |
0 | -1.515822 .0060067 -252.36 0.000 -1.527595 -1.504049
------------------------------------------------------------------------------
Iteration 0: EE criterion = 9.684e-16
Iteration 1: EE criterion = 8.458e-28
Treatment-effects estimation Number of obs = 53,038
Estimator : inverse-probability weights
Outcome model : weighted mean
Treatment model: probit
------------------------------------------------------------------------------
| Robust
y | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ATE |
female |
(1 vs 0) | -.1050439 .0401988 -2.61 0.009 -.183832 -.0262557
-------------+----------------------------------------------------------------
POmean |
female |
0 | -1.458219 .0054927 -265.48 0.000 -1.468984 -1.447453
------------------------------------------------------------------------------
Source | SS df MS Number of obs = 100,000
-------------+---------------------------------- F(1, 99998) = 115.72
Model | 155.270531 1 155.270531 Prob > F = 0.0000
Residual | 134179.437 99,998 1.3418212 R-squared = 0.0012
-------------+---------------------------------- Adj R-squared = 0.0011
Total | 134334.707 99,999 1.34336051 Root MSE = 1.1584
------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
female | .0988518 .0091894 10.76 0.000 .0808406 .1168629
_cons | -1.501196 .0040908 -366.97 0.000 -1.509214 -1.493178
------------------------------------------------------------------------------
Source | SS df MS Number of obs = 46,962
-------------+---------------------------------- F(1, 46960) = 81.37
Model | 107.755058 1 107.755058 Prob > F = 0.0000
Residual | 62185.463 46,960 1.32422196 R-squared = 0.0017
-------------+---------------------------------- Adj R-squared = 0.0017
Total | 62293.2181 46,961 1.32648832 Root MSE = 1.1507
------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
female | -.1088381 .0120654 -9.02 0.000 -.1324866 -.0851897
_cons | -1.487199 .0061843 -240.48 0.000 -1.499321 -1.475078
------------------------------------------------------------------------------
Source | SS df MS Number of obs = 53,038
-------------+---------------------------------- F(1, 53036) = 886.34
Model | 1182.47024 1 1182.47024 Prob > F = 0.0000
Residual | 70755.5305 53,036 1.33410383 R-squared = 0.0164
-------------+---------------------------------- Adj R-squared = 0.0164
Total | 71938.0008 53,037 1.35637387 Root MSE = 1.155
------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
female | .4290223 .0144105 29.77 0.000 .4007775 .457267
_cons | -1.511833 .0054114 -279.38 0.000 -1.522439 -1.501226
------------------------------------------------------------------------------
.
. fulijhkjhk
command fulijhkjhk is unrecognized
r(199);
end of do-file
r(199);
end of do-file
r(199);
. rdrobust z voteshare_female, c(0.5) kernel(uniform)
Sharp RD estimates using local polynomial regression.
Cutoff c = .5 | Left of c Right of c Number of obs = 28548
-------------------+---------------------- BW type = mserd
Number of obs | 11425 17123 Kernel = Uniform
Eff. Number of obs | 2282 2410 VCE method = NN
Order est. (p) | 1 1
Order bias (q) | 2 2
BW est. (h) | 0.046 0.046
BW bias (b) | 0.092 0.092
rho (h/b) | 0.497 0.497
Outcome: z. Running variable: voteshare_female.
--------------------------------------------------------------------------------
Method | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------------+------------------------------------------------------------
Conventional | .01855 .00924 2.0073 0.045 .000437 .036665
Robust | - - 2.0807 0.037 .001262 .042242
--------------------------------------------------------------------------------
. rdrobust z voteshare_female, c(0.5) kernel(uniform) masspoints(off) stdvars(on)
Sharp RD estimates using local polynomial regression.
Cutoff c = .5 | Left of c Right of c Number of obs = 28548
-------------------+---------------------- BW type = mserd
Number of obs | 11425 17123 Kernel = Uniform
Eff. Number of obs | 2279 2405 VCE method = NN
Order est. (p) | 1 1
Order bias (q) | 2 2
BW est. (h) | 0.046 0.046
BW bias (b) | 0.092 0.092
rho (h/b) | 0.495 0.495
Outcome: z. Running variable: voteshare_female.
--------------------------------------------------------------------------------
Method | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------------+------------------------------------------------------------
Conventional | .01803 .00924 1.9510 0.051 -.000083 .036141
Robust | - - 2.0378 0.042 .000813 .041726
--------------------------------------------------------------------------------
. adopath
[1] "C:/ado/plus/r/rd_2021"
[2] (BASE) "C:\Program Files (x86)\Stata15\ado\base/"
[3] (SITE) "C:\Program Files (x86)\Stata15\ado\site/"
[4] "."
[5] (PERSONAL) "c:\ado\personal/"
[6] (PLUS) "c:\ado\plus/"
[7] (OLDPLACE) "c:\ado/"
. doedit
. pwd
C:\Users\paserman\Dropbox
. doedit "C:\Users\paserman\Dropbox\Research\GenderCooperativeness\EJ\3 replication package\Dofiles\master.do"
. exit, clear