← Back to home
Methodology

We scored 406 of 437 House members using eight public data sources — no polling and no editorial judgment calls about which votes are "right." Everything below is exactly how those scores are built, in plain language.

Where The Data Comes From
Eight public data sources, refreshed on a schedule, not a poll or a survey in the mix.
Congress.gov APIMember identity, photo, and bill/vote metadataUsed in: Member name, party, district, and photo at the top of every member page; bill titles and policy areas in Recent Votes
Voteview (DW-NOMINATE)A member's ideology, scored from their actual voting recordUsed in: The member's half of the Voting Alignment score
MIT Election Lab — House returnsReal House election results by districtUsed in: The district's half of Voting Alignment; the baseline for a House member's race forecast
MIT Election Lab — Senate returnsReal statewide Senate election resultsUsed in: The baseline for a Senate member's race forecast
Census ACSPoverty rate, median income, uninsured rate by districtUsed in: The District Need panel — shown for context, never part of the grade
USASpending.gov APIFederal contracts and grants flowing into each districtUsed in: The Delivery score (compared against Census need, within similar-density districts)
FEC.gov APICampaign receipts — small-dollar vs. large, individual vs. PACUsed in: The Donor Capture score, and the Donor Composition bars
VoteHub APIGeneric congressional ballot and presidential approval pollingUsed in: The national swing built into every race forecast (not shown as its own number)
How These Scores Are Calculated
Plain-language version — no jargon, no formulas.
Delivery

We compare how much federal money (contracts and grants) actually flows into this district against how much the district needs, based on poverty, income, and insurance data. Then we compare that gap only against other districts with similar population density — rural vs. rural, city vs. city — so a rural district never gets unfairly judged against a big city, or vice versa.

Voting Alignment

This is not a poll. Both sides of this come from real, recorded actions. The district's lean comes from how people there actually voted in the last two House elections. The member's ideology comes from their actual voting record in Congress, using a standard method that looks at which bills they tend to vote for together over their whole career. We score both on the same 0 to 100 scale and compare them. The closer they match, the higher the score. It does not matter if that match is liberal or conservative, both score the same either way. This is not a "good votes" or "bad votes" score. It only asks one thing: does this member vote the way their district actually votes, or more extreme than that?

Donor Capture

We look at what share of a member's campaign money comes from small individual donors (giving $200 or less) and from individuals generally, versus PACs, corporations, and large-dollar donors. A higher score means more of their money comes from regular people rather than organized interests.

How the three combine into one grade

We rank every member nationally on each of the three scores above, then average those three rankings together for one composite score. Members are then graded on a curve: the top 10% get an A, the next 25% get a B, the middle 30% get a C, the next 25% get a D, and the bottom 10% get an F.

What's not in the grade

District Need (poverty rate, median income, uninsured rate) is shown for context only — it describes the district, not the member, so it never pushes the grade up or down.

Nonpartisan by construction — no component ever conditions on party. Grades and forecasts are mechanical outputs of the data above, not editorial opinions.