Which Washington legislators break with their party

Everyone in Olympia has a sense of who the independents are. Almost nobody has the number.

We keep a copy of every recorded floor vote in the Washington State Legislature, so we can check rather than guess. What follows is the 2023-24 biennium — 1,108 bills with recorded floor votes, every member of both chambers, counted from the roll calls themselves.

The short version: the Legislature is far more unanimous than its reputation, and the two obvious ways to measure independence produce almost opposite lists.

First, most floor votes aren't close

Party-line splits are the exception. In the House, about 200 of the roughly 980 floor votes a member was present for had the two caucuses on opposite sides. In the Senate it was about 189 of 990.

So roughly four out of five floor votes in Olympia are not a party fight at all. Most bills that reach the floor have already been negotiated, amended, or quietly killed upstream — by the time there's a vote, there's usually a consensus. The fights people remember are real, but they are a fifth of the record.

Within that, both caucuses hold together tightly:

The median legislator broke with their own caucus on 1.6% of votes in the House and 1.7% in the Senate. That is the baseline against which everything below should be read: near-total cohesion, with a handful of genuine exceptions.

The distinction that changes the answer

There are two different things people mean by "breaks with their party," and they are not the same measurement:

• Breaking — voting against the majority of your own caucus, on any vote.
• Crossing over — voting with the other party on a vote where the two caucuses split.

You can do a great deal of the first while almost never doing the second. Two House Republicans make the point better than any explanation:

Walsh breaks with his caucus nearly twice as often as Cheney and crosses to the Democrats essentially never. Cheney does the opposite. They are the two most independent Republicans in the House by different measures, and they are not doing the same thing at all. One is voting no when his caucus votes yes; the other is voting yes with the other side.

Any list that reports a single "independence score" is averaging those two behaviours together and telling you about neither.

Who breaks with their own caucus most

Counted as: votes against the majority position of the member's own caucus.

House

Senate

Bob Hasegawa is the only Democrat in either top five. That is not an accident of personality — it's what majority and minority status look like in the record, and it's the next section.

Who crosses over to the other party

Counted as: on votes where the caucuses split, the member voted with the other caucus. This is the number that says who is genuinely reachable across the aisle.

House

Senate

The pattern nobody mentions: crossing over is mostly a minority-party behaviour

Look at those two tables again. Every one of the ten highest crossover rates in both chambers belongs to a Republican — and the Republican caucuses are the less cohesive ones in both (95.0% and 95.4%, against 99.0% and 98.7%).

That isn't a claim about who is more open-minded. It's structural. When you are in the majority, your caucus's position usually is the outcome — holding together is how bills pass, and a defection costs something. When you are in the minority, a party-line split is one you're going to lose anyway, and voting with the majority on a bill you can live with costs you very little.

So a high crossover rate says something real about a member, but it says it in the context of which side of the chamber they sit on. Comparing a majority-party member's 5% to a minority-party member's 25% is not comparing the same decision.

How this was counted, and what it doesn't tell you

• Source: the Legislature's own roll calls for the 2023-24 biennium, as published by the Washington State Legislature's web services. 1,108 bills with recorded floor votes.
• "Own caucus position" is the majority position of the member's own party on that vote. Members the record lists with no party are excluded from these tables.
• Absences and excused votes are not counted as anything. A member's percentages are over the votes they were actually recorded on, which is why the denominators differ slightly between members.
• Chambers are counted separately. A House member and a Senate member never voted on the same roll call, so comparing them directly is comparing different votes.

What this is not: a rating. There is no good or bad score here, and none of these numbers say whether a member voted the right way on anything. A legislator with zero breaks may represent a district that agrees with their caucus; one with a hundred may be doing exactly what they were elected to do. This is the record of what happened, and it is the beginning of a question about a member rather than the answer.

It is also only floor votes. Most of what determines a bill's fate happens in committee, in negotiation, and in the decision about what reaches the floor at all — none of which appears in a roll call.

Every number here is computed from the public roll-call record and is on OlympiaTrack's Insights page under Coalitions, for any biennium we hold — including each individual bill behind each member's count, so you can check any figure above against the vote it came from.

Last updated August 5, 2026.