How your members of Congress vote on this agenda

How this is scored

Where the data comes from: every mark is computed only from official Congress.gov records - cosponsorship lists and the House Clerk and U.S. Senate roll-call votes - refreshed weekly. The records are facts. The scoring is editorial: which votes count as key votes, what the platform position is, and how votes translate into a stance label are judgments this project makes and publishes in full below, so you can disagree with the rules while trusting the records.

• Votes drive the stance, and the denominator is small. The key recorded votes are One Big Beautiful Bill Act (both chambers) and the SAVE Act (House). The platform's position is to vote no on both. That means a senator's stance currently rests on one vote and a representative's on at most two; more roll-calls are added as platform bills reach the floor. Until then, the bills-backed meter is the better breadth measure, which is why both are always shown together.
• Cosponsorship only adds support. Not cosponsoring a bill is recorded as "no record," never a penalty - so the support meter counts the platform bills a member has actively backed in their own chamber.
• Every mark is ask-level, and that is the point. A member who cosponsors one platform bill is backing that ask, not endorsing this platform; a member who votes against another is opposing that ask, not everything here. Most members will agree with parts of this agenda and disagree with others, and the scorecard treats that as normal, not as inconsistency to be punished. That's why each member page lists every ask separately, so you can see exactly which elements they back, which they oppose, and where there's no record, instead of collapsing a mixed record into one number.
• Every mark is sourced. Each member page links each vote to its roll-call and each cosponsorship to the official bill record.
• No causal claims. Records are shown next to outcomes, never as proof one caused the other. As of 2026-07-19.
• The data is downloadable. The complete scored dataset (scorecard.json) and the ask catalog with the full scoring rules are published here, so anyone can audit the grades or apply different rules.