Data sources
Roll-call votes — Voteview
The complete member-by-member vote matrices for the 117th, 118th and 119th Congresses were downloaded from Voteview (Lewis, Poole, Rosenthal, Boche, Rudkin, Sonnet — UCLA), the standard academic source for congressional roll-call data. Files used: H117/H118/H119_votes.csv, _rollcalls.csv, _members.csv.
Coverage: 3 January 2021 through 23 July 2026 — 996 + 1,235 + 643 = 2,874 House roll calls. Any vote after 23 July 2026 is not reflected here.
Sponsorship — Congress.gov API
Sponsored and cosponsored legislation was retrieved from the official Library of Congress API endpoint for member M001213 on 15 August 2026. Counts are the API's own: 97 sponsored, 622 cosponsored.
The two percentages, defined precisely
Measure A — all votes with a Republican position
For every House roll call, count Republicans voting Yea and Nay. If one side has more, that is "the Republican position." If Moore voted the other way, that is a break. Votes where Moore was absent, voted Present, or where Republicans split exactly evenly are excluded.
Result: 188 breaks in 2,842 qualifying votes — 6.6%. He took the same side as most Republicans 93.4% of the time.
Measure B — contested party-unity votes (the standard measure)
This is the conventional political-science definition, used by CQ Roll Call and others. It counts only votes where a majority of Republicans opposed a majority of Democrats — i.e. votes that actually divided the parties. Lopsided and unanimous votes are dropped, because they inflate any loyalty score.
Result: 173 defections in 1,886 party-unity votes — a party-unity score of 90.8%.
Measure B is the fairer number to quote and it is the less flattering one for him, which is a good sign that it is not being cherry-picked. Both are on this site so nobody can accuse you of picking whichever looked worse.
Per-Congress results
| Congress | Party-unity votes | Defections | Unity score | Median R defection rate | Rank (most independent) | Nokken-Poole |
|---|
Limitations you should know about
- Ranking is restricted to comparable members. Republicans who cast fewer than 60% as many qualifying votes as Moore (mid-term arrivals, departures, long absences) are excluded from the rank denominators so the comparison is like-for-like.
- A "break" is not the same as moderation. In the 118th Congress especially, many of Moore's defections were votes alongside Republican leadership against his party's right flank. The raw defection count overstates ideological independence, which is why every break is listed individually rather than only summarised.
- Most breaks are amendment votes. 144 of the 188 breaks (77%) are amendment votes rather than final passage or other final action. Use the vote-type filter on the breaks page before drawing conclusions.
- DW-NOMINATE is an estimate, not a fact about beliefs. The 0.349 → 0.381 → 0.425 progression is a model output describing his voting position relative to his colleagues, not a measurement of his sincerity.
- Sponsorship counts are not merit. Two standalone public laws out of 97 is genuinely low, but provisions routinely become law inside larger bills — as Moore's child-investment-account proposal did inside H.R. 1. Do not claim he "passed nothing."
- Third-party estimates are labelled as such. The 77,000-Utahns and $870-million figures come from KFF and the Utah Hospital Association respectively, are projections rather than observations, and are linked to their sources on the overview page.
- The Mayorkas first vote. Moore is recorded voting no on 6 February 2024 on a resolution that failed, then yes on 13 February when it passed. Members of leadership sometimes switch to no in order to preserve a motion to reconsider. This site reports both votes and does not assert his motive.
Using this honestly
This site was built to argue against Blake Moore's re-election. That is a point of view, and it is stated openly rather than hidden. What it does not do is invent, omit or shade anything: the votes to his credit — certifying the 2020 election, the January 6 commission, the Respect for Marriage Act, Ukraine aid, expelling Santos, the Trump veto override — are given the same prominence as the votes against him. An argument that survives the counter-evidence is the only kind worth making at a town hall, because someone in that room will know the counter-evidence.