Methodology · versioned formula
How we score influence severity (monkey-readable)
The domain name is editorial. Every dollar on this site links to a primary source. Severity labels are editorial summaries of public records — they do not prove quid pro quo, intent, or loyalty. We are monkeys with spreadsheets; the math is versioned.
PAC dependency ratio (headline number)
tagged_pac_contributions ÷ total_pac_contributions
The numerator is PAC dollars from committees in our published registry (AIPAC PAC, UDP, JStreetPAC, NORPAC, and others — see data/committee-registry.v1.json). The denominator is all PAC contributions to the member in the same cycle. We always show both dollar amounts and link each committee line to FEC filings.
Composite formula (influence_formula_v1)
funding_component = normalize(pac_dependency_pct) × 0.35
+ normalize(primary_pressure_usd) × 0.25
legislative_component = normalize(pinned_vote_alignment) × 0.25
+ normalize(committee_relevance) × 0.15
composite = funding_component × 0.55 + legislative_component × 0.45
Composite scores map to five levels: NONE (gray), LOW (green), MODERATE (yellow), HIGH (orange), SEVERE (red). SEVERE requires top-decile PAC dependency within chamber, high pinned vote alignment when vote data exists, and at least one primary-pressure signal. Phase 1 congress pages weight electoral capture heavily; vote and lobbying inputs expand in later phases.
Conduit contributions
Much "AIPAC PAC money" is individuals routing donations through a committee portal. FEC credits both donor and committee. Member pages show PAS2 committee→candidate contributions for the dependency ratio; Schedule Bitemizes every tagged PAC disbursement (vendors, transfers, refunds, candidate checks) from public bulk filings. Earmarked conduit breakdown remains a separate Phase 2 layer. We never merge foreign-agent FARA spend into domestic PAC totals.
Legal & editorial disclaimer
- We publish facts from FEC, Congress.gov, LDA, FARA, and other primary sources. Every dollar amount on member pages links to a public filing or summary.
- Severity scores are editorial composites, versioned and labeled. Correlation ≠ causation.
- We do not use quid pro quo language — no "bought," "owned," or "traitor." We avoid "agent" except the statutory Foreign Agents Registration Act label (e.g. FARA registrants).
- Contributor and payee names, employers, occupations, and addresses on Schedule B are public FEC fields — we publish them as filed on PAC spending pages. We do not sell or solicit using this data.
- Corrections are logged publicly at /corrections.
Influence Mixer
Argue with the formula like a drunk poli-sci major
Your priors move, the FEC filings don't.
Sliders re-weight the editorial composite for Example member. The filings are fixed; your priors are adjustable — like Congress, but honest about it. Top-level split stays 55% electoral / 45% legislative.
Live composite (monkey-adjusted)
72