How RepRadar calculates everything

When you vote Yea or Nay on a bill, RepRadar compares your position to how your saved representatives actually voted on the same roll call (from GovTrack, which mirrors official House Clerk and Senate records). Match % = matching votes ÷ bills you've both weighed in on. Your votes are stored only in your browser (localStorage) — we never see them, and there is no account.

A match score based on 2 votes means little; one based on 30 means a lot. The vote count is always shown next to the percentage so you can judge.

All federal campaign finance data comes from the Federal Election Commission (candidate totals and itemized Schedule A receipts, refreshed at most daily). "Small donors" are unitemized individual contributions (under $200). Top-donor lists exclude refunds and transfers (e.g. U.S. Treasury entries), negative amounts, committee-to-committee transfers, and conduit committees (ActBlue, WinRed) — conduits are payment processors passing through individual donations, not PAC influence.

This is currently RepRadar's only issue-specific funding tracker, so it deserves an explanation. It exists because pro-Israel PACs were among the largest issue-focused outside spenders of the 2024 cycle, the underlying FEC data is unusually clean (a well-defined set of committees), and reader demand for it is high. It tracks direct contributions (Schedule A/B) from 20+ registered pro-Israel PACs and independent expenditures in support (Schedule E) from 4 Super PACs. Every number links to FEC filings.

A low or zero figure is reported as-is ("none on record"), and totals are career-to-date across cycles. We plan to generalize this into a set of issue-funding trackers (pharma, oil & gas, guns, crypto, unions) that you can choose from; this one shipped first, not exclusively.

Attendance is votes present ÷ eligible roll calls, from GovTrack. Bills Enacted counts sponsored bills that became law this Congress — note that 1–2% is typical for members; passing bills is rare by design. Votes w/ Majority is how often the member voted with the chamber majority over recent roll calls — it measures contrarianism, not quality. Ideology & Leadership are GovTrack's published statistical analyses of cosponsorship patterns — not our editorial judgment.

Stock trades come from STOCK Act disclosures (House Clerk and Senate eFD, via Financial Modeling Prep), accumulated daily. Members report ranges, not exact amounts, and have 45 days to file — recent trades may not appear yet.

Scorecards (Heritage Action, LCV, NRA, and others) are republished from each organization's own public scorecard, labeled with the issuing group and year. These are the groups' opinions, not ours — a 0% from one group and a 100% from its opposite tell you the same thing from two directions.

"Explain" buttons and daily briefings are generated by an AI model (Groq-hosted Llama) under instructions to summarize only what the underlying official text says — no opinion, no speculation, source citations required. AI summaries can still contain errors, and they are always labeled as AI-generated. The official text (CRS summary, roll call, Federal Register document) is linked next to every summary; when it matters, read the source.

Verified verdict positions require email confirmation and a ZIP-matched state. Full methodology, including what counts as "verified" and how tallies close: How verdicts work →

Representative data: Cicero, GovTrack, OpenStates. Votes: House Clerk, Senate.gov, GovTrack. Bills: Congress.gov (Library of Congress). Finance: FEC, FollowTheMoney. News: Federal Register, Congress.gov, GovInfo, AP, Reuters, C-SPAN. Vote data refreshes every 15 minutes during session hours; finance data daily; scorecards daily. Every module shows its source, and pages show "last updated" stamps.

Found something wrong? That's a bug, not a feature — tell us and we'll fix it and say so.