Methodology

The Influenced Index was built to answer one question: when a member of Congress takes money, does it change how they vote?

This page explains how we calculate each member's score. If you just want the quick version: we look at twelve things — who gives them money, how much, whether they vote the way their donors want, whether contributions show up right before key votes, and whether the money comes from groups trying to hide who's behind it. We weigh the evidence of influence more heavily than the money itself. A member who takes modest donations but always votes with their funders will score higher than one who raises millions but votes independently.

Every score comes from public records — FEC (Federal Election Commission) filings, lobbying disclosures, and recorded votes.

What's scored

Every member's score is built from twelve categories. Seven behavioral and network categories — vote alignment, contribution timing, stock trades, revolving-door lobbying access, outbound money distribution, the breadth of funding networks a member draws from, and the power of the committees they sit on — carry the bulk of the weight. Five money categories — direct contributions, outside spending, lobbying inside the policy areas the member regulates, dark money, and Israel-policy PAC money — measure the financial channel. Every category counts in the score.

The twelve-category score

Every sitting member of Congress gets an exposure score built from twelve numbers, each tracked separately and summed, then placed on a 0–100 scale against a fixed ceiling — the same ceiling for every Congress. Evidence of capture — how members vote, when money arrives, whether they trade in the sectors they regulate — carries the heaviest weight. Money alone carries the least. The remaining categories measure who has access, how widely a member's support spreads across funding networks, where the member sends money in turn, how much institutional power their committee seats carry, and whether the money serves a foreign government's policy agenda. The number shown on a member's page is that 0–100 score, capped at 100.

Contributions

Direct contributions to the member's main campaign committee and leadership PAC. Contributions from donors whose policy area matches the industries and policy areas their committees oversee carry additional weight.

Independent expenditures

Outside spending reported on FEC Schedule E filings — money spent for the member, or against their general-election opponent. Policy-aligned outside spending (sector- or issue-specific, like AIPAC's UDP, Fairshake on crypto, or KochPAC on energy) carries more weight than generic partisan outside spending (NRCC, DCCC, partisan SuperPACs) — partisan outside spending tells you about party caucus, not about a specific funding network buying a member.

Lobbying

Senate lobbying disclosures (LD-1 and LD-2 filings, reported quarterly). Each filing is mapped to the industries and policy areas implicated by its issue codes. Lobbying that touches the policy areas this member regulates counts toward this category.

Revolving door

Former members and staff who left Congress to become registered lobbyists, now lobbying the committees they once served on. Measured as a count of matched revolving-door lobbyists per member from Senate lobbying disclosures (covered_position fields cross-referenced against the congressional roster).

Vote alignment

For each funding network, direction is inferred from its outside-spending pattern — which side the network backs in close races. Then every member is scored by how often their recorded votes agree with the direction the network's supported candidates voted. Higher alignment means the member votes with the network's bloc.

Bundle timing

Contributions that arrive in close proximity to a related floor vote, measured as a share of the member's cycle receipts.

Dark money

The share of a member's total independent expenditures that traces to groups that don't disclose their donors — 501(c)(4) tax-exempt intermediaries, Senate Majority PAC / Leadership Fund style vehicles where the underlying donor base is hidden. The modifier does not double-count outside spending — it reweights existing independent-expenditures points for opacity.

Trade direction

Member stock trades from House and Senate Periodic Transaction Reports, filtered to purchases (not sales) of equities whose sector maps to a policy area the member legislates. The score reflects the dollar share of in-jurisdiction buys against total trading. Party leadership — Speaker, former Speakers, Majority and Minority Leaders — get jurisdiction over all policy areas (Pelosi has no committee seat as Speaker Emerita but legislates across every area she chooses).

Outbound money distribution

Money a member's own committees send back out — transfers from their campaign and leadership committees to party committees and to other members (FEC 24K/24Z transactions). A member who funds the party apparatus and bankrolls colleagues is exercising influence, not only receiving it. Measured on a log scale against the largest distributor in Congress.

Cluster network breadth

How many distinct coordinated funding networks a member draws money from. A member tied to a single network looks different from one woven into many; breadth counts the sponsor networks with material support for the member, normalized against the range across Congress. (This replaces the earlier donor-concentration measure, which is no longer scored.)

Committee jurisdiction power

Measures the weight of a member's committee assignments. Tier 1 committees (Appropriations, Ways & Means, Finance, Armed Services, Energy & Commerce, Intelligence) carry the highest weight. Committee chairs receive a multiplier. Members with no standing committee assignments score zero.

Foreign interest

Israel-policy PAC money and FARA-registered institutional lobbying. Direct PAC money carries full weight. Institutional lobbying is allocated by committee jurisdiction at a reduced weight (0.002), reflecting that a dollar given to a member’s campaign is more direct than a dollar spent lobbying their committee.

Total contributions and outside spending from the canonical Israel-policy PAC network in the current cycle (2024 and 2026). These are American PACs that organize around Israel policy — both pro-Israel groups and pro-Israel-critic groups such as J Street, which opposes the current Israeli government. Includes both direct PAC-to-candidate contributions and outside spending supporting the candidate, summed from 16 hand-curated Israel-policy PACs: AIPAC PAC, United Democracy Project, Pro-Israel America PAC, NORPAC, American Pro-Israel PAC, NATPAC 1947, DMFI PAC, Pro-Israel America Action Fund, JStreetPAC, J Street Action Fund, RJC-PAC, RJC Victory Fund, Democratic Majority for Israel, Democrats for Israel, COPAC, CITYPAC. The category measures current-cycle Israel-policy network exposure, not historical donor accumulation — a member raising small amounts whose career donor file shows tenure-accumulated Israel-policy money will not score high here unless the money is flowing now. Outside spending against the member (OPPOSE) is not counted in this category (that flow is captured by the general outside-spending category).

How the tiers work

A raw score means something different in a Congress that has run its full two years than in one still in session, and it means something different in the Senate than in the House. So the Index does not cut tiers off a fixed number. Instead, each member is ranked against the others scored in the same Congress, and placed in one of four tiers by where they fall in that ranking:

• Most exposed — top 5% of the Congress (crimson)
• Highly exposed — next 15% (gold)
• Moderately exposed — the middle 50% (warm gray)
• Least exposed — the lowest 30% (dark green)
• Not scored — members with no usable voting record to score (see below)

The labels are comparative. They describe where a member stands relative to colleagues in the same Congress — not an absolute verdict, and not a legal finding. Least exposed means lowest-ranked among those scored, not free of influence. Because the tiers are relative, the score that marks the top 5% differs from one Congress to the next. That is intended: each Congress is judged as its own field, so the same raw score can place a member in different tiers in different congresses, because the colleagues they are ranked against change.

Why ranked, not absolute

The displayed score sits on a fixed 0–100 scale — the same ceiling for every Congress — so the numbers are comparable across time. The tier labels, though, are ranked within each Congress: a member is judged against the colleagues seated alongside them, not against members from a different era.

The current Congress is in progress

The 119th Congress runs through January 2027. Its votes and filings are still being recorded, so scores and tier placements will continue to move as the term plays out. Treat any current-Congress placement as provisional.

Members scored, not the full chamber

For past congresses the Index covers the members for whom complete public records were assembled — not every seat that existed. A member's history may have gaps for terms in which they were not scored. Within the set of members scored for each Congress, the counts and rankings are exact.

"Not scored"

A member is marked Not scored — rather than placed in a tier — when there is no usable voting record to score against. Two cases: a sitting member whose vote and contribution records have not yet been matched to them in the data (often because the records are filed under a different FEC candidate identifier), and a non-voting delegate (American Samoa, D.C., Guam, Northern Mariana Islands, Puerto Rico, U.S. Virgin Islands), who casts no floor votes for the model to read. A Not scored label is the honest absence of a score, not a finding of low exposure.

What "Most exposed" means

The top tier reflects a combination of signals across the twelve categories, measured relative to the rest of the Congress. It describes measured financial exposure — it does not describe intent or motive, and it does not mean corrupt or unlawful. The signals it draws together include:

• High direct contributions from organized interest networks.
• Substantial independent-expenditure support from sponsor-aligned outside groups.
• Significant lobbying exposure inside the policy areas the legislator regulates.
• Former-staff or former-member lobbyists from the same network now lobbying the member's committees.
• Strong roll-call alignment with the financial backers.
• Contributions clustered around related floor votes.
• Outside spending concentrated in dark-money vehicles rather than disclosed sources.
• Personal trades aimed at sectors the member legislates.
• Money routed back out to party committees and colleagues, and support spread across many coordinated funding networks.
• Seats on the most powerful committees, with extra weight for chairs.

Any one of these in isolation is normal. Several together is the pattern the score is built to surface.

Sponsor network detection

The system detects coordinated funding networks by scoring structural signals between FEC committees — shared donors, shared treasurers, shared addresses, direct transfers, joint candidate support, and joint outside-spending patterns. Dense sub-networks are extracted algorithmically.

Data sources

• Federal Election Commission bulk data — committees, candidates, contributions, disbursements, c2c transfers, c2c-to-cand transfers, independent expenditures. Cycles 2022, 2024, 2026.
• Voteview / DW-NOMINATE — roll-call votes for the U.S. House and Senate, ICPSR identifiers, member metadata.
• Senate Office of Public Records — Lobbying Disclosure — quarterly LD-1 / LD-2 filings.
• House Financial Disclosures — periodic transaction reports for member stock trades.
• congress-legislators — Bioguide IDs and the legislator-FEC crosswalk maintained by the civic-tech project.
• Bioguide portrait archive via unitedstates/images — public-domain US government works.

What the Index now tracks

• Revolving door. Former members and staff who registered as lobbyists are now tracked via Senate LDA filings. The system extracts covered_position fields and cross-references against the congressional roster.
• Dark money donor inference. When a 501(c)(4) tax-exempt group clusters with disclosed PACs in the same funding family, the disclosed PACs' donor base provides a probabilistic window into who funds the hidden entity. Coverage ratio — the share of disclosed network peers a donor funds — serves as the confidence signal. A donor who funds every disclosed peer in a dark-money group's network is almost certainly funding the hidden entity too. This is inference, not proof. The underlying donor identities remain legally undisclosed.

What the Index cannot see

• Dark-money donor identities. The upstream donor base of 501(c)(4) tax-exempt groups that hide their donors remains legally undisclosed. The Index infers probable donors from disclosed network peers, but this is inference, not proof.
• Network resolution. Funding-network detection is conservative — some related committees may be split, and the meta-network layer corrects some but not all of those splits.

Weight selection

The weight allocation prioritizes evidence of behavioral capture over raw money volume. Sensitivity testing against alternative weight schemes produces stable rankings — the same members anchor the top of every list regardless of how the weights are distributed. The ranking is driven by the data, not by the choice of weights.

What "Most exposed" is not

Tier labels are shorthand for a member's rank within their Congress. They are not legal findings, and a high rank against one funding network does not mean a member is captured by every network. The Index reports measured financial exposure from public records; it does not allege a crime, a quid pro quo, or any specific corrupt act.

Scoring 2026 candidates

The 2026 race pages score every FEC-registered candidate, not only sitting members. Coverage by candidate type:

• Incumbents carry the full twelve-category score and their exposure tier, identical to the dossier page (or Not scored, where the records have not been matched). Sitting members rolling into 2026 link straight through to their full record.
• Challengers and open-seat candidates are scored on three categories only: direct contributions (2026 cycle, principal committee), outside spending from sponsor-aligned groups, and dark-money exposure. The other seven categories require committee assignments, a recorded-vote record, or funding-network financial history, none of which exists for a non-sitting candidate. Because they have no voting record, the incumbent "exposed" tiers do not apply to them. Instead they are labelled by the organized money around the campaign — Heavily backed, Substantially backed, Modestly backed, or Lightly backed — a statement about funding, not about conduct.
• A challenger and an incumbent in the same race are not directly comparable on score alone — they are measured on different things.

Known limitations

• Chamber effect. In the current (119th) Congress, senators score about 6 points higher than House members on average, because more money flows through Senate races; the size of that gap varies from one Congress to the next. Tiers are ranked across the full Congress (both chambers together), so that skew carries into tier placement — senators are somewhat more likely to land in the higher tiers. Within-chamber comparisons are the most reliable.
• Network-level lobby sparseness. Not every funding network has direct lobbying attribution. Member-level lobby coverage, mapped through the policy areas each member regulates, is near-complete.
• Stock trades are buys-only and threshold-gated. The trade category only scores purchases inside the policy areas the member regulates. Members trading rarely or trading outside their jurisdiction score 0 on this category. Party leadership (Speaker, former Speakers, Majority and Minority Leaders) is treated as having jurisdiction over all policy areas so the filter doesn't drop Pelosi-style cases.
• Contribution-timing is window-based. The timing category captures contributions timed to related legislative action. Contributions outside that window are not scored.
• Challenger and open-seat candidates are partial. Lobbying, vote alignment, timing, and trade data require committee assignments and roll-call votes. A non-incumbent is measured only on the money around the campaign.

Validation

• Party confound. A model regressing the score on party alone has R² = 0.0006. The score is not a proxy for party affiliation.
• Weight sensitivity. Alternative weight schemes — heavier money categories, flat across all twelve categories, behavior-only — each produce stable rankings against the canonical ordering. The ordering doesn't depend on the weights.
• Score reconstruction. 532 of 532 sitting-member scores reproduce exactly from the raw inputs and the published methodology. Maximum deviation under floating-point math: 0.0. No member is hand-tuned.
• Versioning. The current scoring (twelve categories) correlates strongly with earlier versions. Three categories were added — outbound money distribution, the breadth of a member's funding networks, and committee jurisdiction power — and the earlier donor-concentration measure was retired in their favor. The change shifts mid-tier ordering while keeping the top of the list stable.
• Committee verification. 15 of 15 spot-checked committee assignments match the official roster.
• Outside-spending classification. 251 funding networks carry outside-spending attribution totaling $3.99B. Each network's outside-spending direction is sanity-checked against its largest disbursements.
• Alignment validation. The cluster-direction inference was validated against five known sponsor-aligned PACs. Inferred direction matched the documented direction in all five cases.