Spaces:
Running on CPU Upgrade
Running on CPU Upgrade
Political Influence Analysis - Complete Data Integration
The Complete Picture
This integration combines 4 major data sources to reveal political connections in civic engagement and grant funding:
┌─────────────────────────────────────────────────────────────────────┐
│ DATA ECOSYSTEM │
├─────────────────────────────────────────────────────────────────────┤
│ │
│ 1. IRS FORM 990 2. FEC CONTRIBUTIONS │
│ ─────────────── ─────────────────── │
│ • Nonprofit orgs • Political donations │
│ • Officer names • Contributor employer ─────┐ │
│ • Compensation • Donation amounts │ │
│ • Revenue sources • Recipients (candidates) │ │
│ │ │
│ 3. GRANTS.GOV 4. VOTER DATA │ │
│ ─────────────── ──────────── │ │
│ • Available grants • Party affiliation │ │
│ • Deadlines • Elected officials │ │
│ • Award amounts • Turnout patterns │ │
│ • Eligibility • Demographics ────┴───┐ │
│ │ │
│ ┌───────────────────────┐ │ │
│ │ ANALYSIS QUESTIONS │ <─────────────┘ │
│ └───────────────────────┘ │
│ │
│ • Political influence on grant awards? │
│ • Partisan patterns in oral health funding? │
│ • Advocacy networks and donor circles? │
│ • Timeline: Donation → Policy → Grant? │
│ │
└─────────────────────────────────────────────────────────────────────┘
Data Source Comparison
What Each Source Provides
| Data Source | What You Get | Key Value |
|---|---|---|
| IRS 990 (existing) | Nonprofit finances, officers, grants received | WHO got funded, HOW MUCH |
| Grants.gov (new) | Available grant opportunities, deadlines | WHAT'S available NOW |
| FEC (new) | Political donations by individuals | WHO donated to WHOM |
| Voter Data (new) | Jurisdiction politics, elected officials | POLITICAL CONTEXT |
Combined Insights
When you integrate all four:
EXAMPLE: Massachusetts Dental Health Clinic
IRS 990:
├─ Organization: MA Dental Health Clinic (EIN: 12-3456789)
├─ Executive Director: Dr. Jane Smith
├─ Compensation: $150,000/year
├─ Revenue: $2M total
│ ├─ Government grants: $800,000 (40%)
│ ├─ Program service: $1M (50%)
│ └─ Donations: $200,000 (10%)
FEC Contributions:
├─ Dr. Jane Smith
│ ├─ $2,500 → Senator Elizabeth Warren (D-MA) - 2024-03-15
│ ├─ $1,000 → Rep. Joe Kennedy (D-MA-04) - 2024-06-01
│ └─ Total: $3,500 in 2024 cycle
│
└─ Board Member: Dr. John Doe
├─ $5,000 → HCFP PAC (Health Care for People) - 2024-02-10
└─ Total: $5,000
Grants.gov:
├─ HRSA Oral Health Workforce Grant
│ ├─ Opportunity: HRSA-24-123
│ ├─ Posted: 2024-07-15
│ ├─ Deadline: 2024-12-31
│ ├─ Award: $500k-$2M
│ └─ Eligible: Community health centers (MATCH!)
Voter Data:
├─ Location: Boston, MA (Congressional District 4)
├─ District: D+15 (strong Democratic)
├─ Representative: Joe Kennedy (D) - recipient of Dr. Smith's donation
├─ Senator: Elizabeth Warren (D) - recipient of Dr. Smith's donation
└─ State Legislature: Democratic majority
TIMELINE ANALYSIS:
├─ 2024-03-15: Dr. Smith donates to Sen. Warren
├─ 2024-06-01: Dr. Smith donates to Rep. Kennedy
├─ 2024-07-15: HRSA grant posted (eligible for clinic)
├─ 2024-09-01: Clinic receives $500k HRSA grant (hypothetical)
└─ 2025-03-01: IRS 990 reports $800k government grants
QUESTIONS:
• Correlation or causation?
• Is this normal civic engagement?
• Does political activity predict grant success?
• Are there partisan patterns in health funding?
Analysis Examples
1. Political Influence on Grant Awards
Research Question: Do nonprofits with politically active leadership receive more federal grants?
# Load all data
nonprofits = pd.read_parquet("data/gold/states/MA/nonprofits_organizations.parquet")
officers = pd.read_parquet("data/gold/states/MA/nonprofits_officers.parquet")
grants = pd.read_parquet("data/gold/states/MA/grants_revenue_sources.parquet")
contributions = pd.read_parquet("data/gold/fec/political_contributions.parquet")
# Merge everything
full_data = (
nonprofits
.merge(officers, on='EIN', how='left')
.merge(grants, on='EIN', how='left')
.merge(contributions, left_on='person_name', right_on='contributor_name', how='left')
)
# Compare groups
politically_active = full_data[full_data['contribution_amount'].notna()]
not_active = full_data[full_data['contribution_amount'].isna()]
print(f"Politically active orgs: Avg grant = ${politically_active['government_grants'].mean():,.0f}")
print(f"Not active orgs: Avg grant = ${not_active['government_grants'].mean():,.0f}")
# Statistical test
from scipy import stats
t_stat, p_value = stats.ttest_ind(
politically_active['government_grants'].dropna(),
not_active['government_grants'].dropna()
)
print(f"p-value: {p_value:.4f}")
2. Partisan Patterns in Oral Health Funding
Research Question: Do Democratic or Republican areas get more oral health grants?
# Enrich nonprofits with political context
from discovery.voter_data_integration import PoliticalContextEnricher
enricher = PoliticalContextEnricher()
enriched = enricher.add_political_context_to_nonprofits(
nonprofits_df=nonprofits,
contributions_df=contributions,
legislators_df=legislators
)
# Group by party control
by_party = enriched.groupby('state_legislature_control').agg({
'government_grants': 'mean',
'EIN': 'count'
}).rename(columns={'EIN': 'org_count'})
print("Average Government Grants by State Legislature Control:")
print(by_party)
# Visualization
import matplotlib.pyplot as plt
by_party['government_grants'].plot(kind='bar')
plt.title("Avg Federal Grants: Democratic vs Republican States")
plt.ylabel("Average Grant Amount ($)")
plt.show()
3. Donor Networks in Health Policy
Research Question: Which candidates receive the most from health sector?
# Health sector contributions
health_contribs = contributions[
contributions['contributor_employer'].str.contains(
'health|dental|clinic|hospital',
case=False,
na=False
)
]
# Top recipients
top_recipients = health_contribs.groupby('committee_name').agg({
'contribution_amount': 'sum',
'contributor_name': 'count'
}).sort_values('contribution_amount', ascending=False).head(10)
print("Top 10 Recipients of Health Sector Donations:")
print(top_recipients)
# Network analysis
import networkx as nx
# Build graph: Donors → Candidates
G = nx.Graph()
for _, row in health_contribs.iterrows():
donor = row['contributor_name']
recipient = row['committee_name']
amount = row['contribution_amount']
G.add_edge(donor, recipient, weight=amount)
# Find central candidates (most donors)
centrality = nx.degree_centrality(G)
print("\nMost Connected Candidates:")
for candidate, score in sorted(centrality.items(), key=lambda x: x[1], reverse=True)[:5]:
print(f" {candidate}: {score:.3f}")
4. Timeline Analysis: Donation → Grant
Research Question: How long between political donation and grant award?
# Merge contributions with grants by EIN
timeline_data = contributions.merge(
grants,
left_on='nonprofit_ein',
right_on='ein',
how='inner'
)
# Convert dates
timeline_data['contribution_date'] = pd.to_datetime(timeline_data['contribution_date'])
timeline_data['grant_date'] = pd.to_datetime(timeline_data['grant_award_date'])
# Calculate time difference
timeline_data['days_to_grant'] = (
timeline_data['grant_date'] - timeline_data['contribution_date']
).dt.days
# Filter to positive (donation before grant)
before_grant = timeline_data[timeline_data['days_to_grant'] > 0]
print(f"Donations made BEFORE grant award: {len(before_grant):,}")
print(f"Average time: {before_grant['days_to_grant'].mean():.0f} days")
print(f"Median time: {before_grant['days_to_grant'].median():.0f} days")
# Histogram
before_grant['days_to_grant'].hist(bins=50)
plt.title("Time from Political Donation to Grant Award")
plt.xlabel("Days")
plt.ylabel("Frequency")
plt.show()
Dashboard Features
Political Connections Widget
Add to nonprofit profile pages:
┌────────────────────────────────────────────┐
│ MA Dental Health Clinic │
│ EIN: 12-3456789 │
├────────────────────────────────────────────┤
│ 💰 Political Connections │
│ │
│ Leadership Political Activity: │
│ • 2 officers made political donations │
│ • Total: $8,500 (2024 cycle) │
│ │
│ Federal Grant Awards: │
│ • $800,000 government grants (2023) │
│ • HRSA Oral Health Workforce Grant │
│ │
│ Political Context: │
│ • District: MA-04 (D+15) │
│ • Representative: Joe Kennedy (D) │
│ • Senator: Elizabeth Warren (D) │
│ │
│ [View Full Disclosure] [Timeline Chart] │
└────────────────────────────────────────────┘
Transparency Report
Public disclosure page:
Political Finance Transparency Report
Massachusetts Oral Health Organizations
─────────────────────────────────────────
Summary:
• 145 health nonprofits in database
• 23 have politically active leadership (16%)
• $456,000 in political donations (2024)
• $45M in federal grants received (2023)
Top Donors (by organization):
1. MA Dental Association - $125,000
2. Community Health Network - $89,500
3. Boston Medical Center - $67,200
Top Recipients (candidates/committees):
1. Elizabeth Warren (D-MA) - $234,000
2. Ed Markey (D-MA) - $156,000
3. Various House candidates - $66,000
Partisan Breakdown:
• Democratic candidates: 78%
• Republican candidates: 18%
• Non-partisan: 4%
[Download Full Data] [Methodology]
Ethical Considerations
Important Disclaimers
Correlation ≠ Causation
- Political donations don't prove quid pro quo
- Many factors influence grant awards
- Present data, not conclusions
Personal vs. Institutional
- Officers donate as individuals, not on behalf of nonprofits
- Political views are personal
- Don't attribute individual actions to organizations
Context is Critical
- Civic engagement is protected speech
- Political donations are legal and transparent
- Focus on patterns, not individuals
Fairness and Balance
- Present all sides
- Avoid partisan framing
- Let users draw own conclusions
Best Practices
✅ DO:
- Present data transparently
- Provide full methodology
- Allow users to verify
- Show aggregate patterns
- Respect privacy laws
❌ DON'T:
- Imply wrongdoing without evidence
- Selectively present data
- Use partisan language
- Focus on individuals
- Draw unsupported conclusions
Implementation Checklist
- Get FEC API key (https://api.data.gov/signup/)
- Run FEC integration demo
- Match FEC data to your nonprofit officers
- Combine with Grants.gov data
- Add voter/political context
- Build analysis queries
- Create dashboard widgets
- Write transparency methodology
- Review for ethical concerns
- Launch with clear disclaimers
ROI Summary
Data Collection Effort
| Data Source | Setup Time | Cost | Maintenance |
|---|---|---|---|
| IRS 990 | ✅ Done | Free | Automatic |
| Grants.gov | 5 min | Free | Daily cron |
| FEC API | 5 min (signup) | Free | Weekly batch |
| Voter Data | Varies | $-$$$ | Monthly |
Total: 10-30 minutes + optional vendor costs
Value Delivered
For Transparency:
- Public disclosure of political-financial connections
- Trust through openness
- Accountability for nonprofits and officials
For Analysis:
- Understand influence patterns
- Track money in politics
- Identify advocacy networks
- Research civic engagement
For Advocacy:
- Strategic political engagement
- Optimize advocacy efforts
- Learn from successful orgs
- Navigate political landscape
Differentiation:
- Unique feature - no one else combines these
- Media interest - compelling stories
- Academic value - research partnerships
- Impact - better democracy
Example Queries
Save these for your analysis toolkit:
-- 1. Which nonprofits have politically active leaders?
SELECT
n.organization_name,
COUNT(DISTINCT c.contributor_name) as donor_count,
SUM(c.contribution_amount) as total_donated,
n.government_grants
FROM nonprofits n
JOIN officers o ON n.ein = o.ein
JOIN contributions c ON o.person_name = c.contributor_name
GROUP BY n.ein
ORDER BY total_donated DESC;
-- 2. Partisan breakdown by state
SELECT
state,
party_lean,
COUNT(*) as org_count,
AVG(government_grants) as avg_grants
FROM nonprofits
GROUP BY state, party_lean;
-- 3. Timeline: Donation to grant
SELECT
c.contribution_date,
c.contributor_name,
c.contribution_amount,
g.grant_award_date,
g.grant_amount,
DATEDIFF(g.grant_award_date, c.contribution_date) as days_between
FROM contributions c
JOIN grants g ON c.nonprofit_ein = g.ein
WHERE g.grant_award_date > c.contribution_date
ORDER BY days_between;
Next Steps
Get API Keys:
- FEC: https://api.data.gov/signup/ (instant, free)
- OpenStates: https://openstates.org/accounts/profile/ (optional)
- Google Civic: https://console.cloud.google.com (optional)
Run Demos:
python examples/demo_political_influence.py --api-key YOUR_FEC_KEYGenerate Full Dataset:
- Match all officers to FEC contributions
- Combine with Grants.gov opportunities
- Add voter/political context
Build Features:
- Political connections widget
- Transparency report page
- Timeline visualizations
- Network graphs
Launch with Context:
- Write clear methodology
- Add ethical disclaimers
- Provide full data downloads
- Allow verification
The bottom line: This creates the most comprehensive view of political influence in civic engagement available anywhere. Use it responsibly! 🎯