Capitol Trail is a civic transparency platform that maps the financial relationships between political donors and U.S. Congress members. We analyze campaign finance data, voting records, and industry funding patterns to surface potential incentive alignments — without drawing causal conclusions.
Money flows through politics in ways that are difficult to track and harder to contextualize. Capitol Trail makes that data accessible — structured, searchable, and put in plain language — so citizens, journalists, and researchers can ask better questions.
We don't allege corruption. We present data. The goal is to make it easier to see which industries fund which legislators, how those legislators vote on related bills, and where financial relationships cluster.
All data originates from federal government sources and is used under public domain / open government data principles.
Industry alignment scores reflect the proportion of a politician's campaign donations that come from a given industry sector (e.g., Oil & Gas, Pharmaceuticals, Finance). A high alignment score means a significant share of documented contributions trace to that industry.
Scores are computed from FEC contribution records matched to OpenSecrets industry classifications. Aggregators like ActBlue and WinRed are excluded from industry scoring because they bundle donations from many sources. Scores are expressed as percentages of total tracked contributions.
Low confidence scores occur when donation data is sparse (e.g., some senators with limited individual contribution records). Confidence is noted where relevant.
The party-line vote percentage shown on politician profiles measures how often a member voted with their party's majority position on contested bills. For each bill where the politician cast a YEA or NAY vote, we compute the majority position of their party and check whether the politician's vote matched. Bills with no clear party majority (tie) are excluded.
A high party-line score indicates consistent alignment with party positions, not necessarily ideological rigidity — party majorities themselves vary by bill.
TrailAI is Capitol Trail's AI analysis layer. It generates politician profile summaries, bill analyses, and investigative insight narratives using structured donation and voting data as input. All TrailAI output is evidence-based, neutral in tone, and regenerated periodically as new data is ingested.
TrailAI is built on Anthropic — specifically the Claude model family. Anthropic is an AI safety company focused on building reliable, interpretable AI systems.
Anywhere you see the TrailAI label, the content was generated by AI and should be treated as an analytical starting point, not a definitive characterization. The underlying data is always sourced from federal government records.
Campaign finance data reflects disclosed contributions only. Dark money, in-kind contributions, and some bundled donations are not captured. FEC records have known data quality issues (misspellings, inconsistent employer fields).
Donation-to-vote correlations are observational. The presence of industry funding does not prove that funding influenced a vote. Many factors shape legislative decisions.
Coverage is limited to current Congress members (537 tracked). Former members, candidates, and non-congressional officials are not included.
Capitol Trail is an independent civic project. Data corrections, methodology questions, and partnership inquiries are welcome.