Why did we lose? A Political Autopsy rebuilds a U.S. House or Senate race from the public record: the money, the expectations, the certified result, and the ads. Every figure is tied to a public source.
When a race is lost, the explanations start the same night. The Autopsy answers them from the filings, the market prices, and the certified count.
The report opens with numbered findings, then four chapters, each built from its own public data. It is delivered as a Word document, ready for your team to edit.
Four to six numbered findings, each drawn from the chapters that follow. A point the data doesn’t support is left out.
Itemized receipts week by week, gifts by size and by state, small-dollar totals, PAC money, and operating spending for both candidates. Outside spending is broken out by spender.
FEC filings
Daily prediction-market prices through the final stretch, with public polls where they exist. The chapter shows whether the loss was priced in early or arrived late.
Polymarket price history · public polls
The certified result, set against the district’s past general elections.
Certified results · MIT Election Data and Science Lab
The Google ads that ran in the race, advertiser by advertiser.
Google Ads Transparency Center
What each chapter covers and what it leaves out, followed by a source index listing every public record the report cites.
A post-mortem gets passed around, quoted, and argued over. Each figure in the Autopsy can be traced back to the public record it came from.
Every table is computed directly from the source records and captioned with its source. No figure in a table passes through an AI model.
Each public source is registered once for the report. The written findings can link only to that list, and any other link is removed before the document is built.
A figure with no public source is flagged. An analyst finds the source or cuts the figure before the report goes out.
Before delivery, an analyst checks a claim-by-claim citation map: each claim, its source, and how the figure was verified.
The candidate and senior staff deciding what to tell donors, and whether to run again.
Committees reviewing the seats they lost before recruiting for the next cycle.
Firms and outside groups writing the post-mortem for a client or a board.
Looking at the next race? A Viability Assessment rates a prospective candidacy from a finished research book.
A political autopsy is a post-election review of why a race was lost. Civly’s Political Autopsy rebuilds a U.S. House or Senate race from public data: fundraising and outside money, prediction-market expectations, the certified result, and Google ad spending. Every table cites its source, and the report ends with a full source index.
U.S. House and U.S. Senate races, in primaries and general elections. State legislative, statewide, and local races are not covered yet.
FEC filings for receipts, spending, and outside money; Polymarket price history and public polls where available; the certified result and the district’s election history; and Google’s Ads Transparency Center. TV, radio, and Meta ad spending are not included.
Tables are computed directly from the source records, and the written findings can link only to the sources registered for that report. Any figure without a public source is flagged, and an analyst sources it or removes it before delivery.
A Word document with numbered findings, the four chapters, a methodology and coverage section, and a source index. It is ready for your team to edit and share.
Yes, in early access. Tell us the race and we’ll confirm the data available for it.
Tell us the seat and the cycle, and we’ll confirm the data available for it.