Washington a surprising choice for analysts forecasting elections

Political polls, betting markets and statistical forecasts have become ubiquitous as November approaches. But there’s one predictive tool you may have missed: primary elections in Washington state, which took place last month.

It’s an indicator that election nerds pay particular attention to, for good reason: The general election results in the state a few months later tend to move in the same direction – whether becoming more Republican or more Democratic – as the rest of the country.

This year, results from Washington state are cause for some optimism among Democrats. They suggest a national environment somewhat similar to 2020, when Joe Biden outgained Donald Trump by 4.5 percentage points in the national popular vote and Democrats retained the House.

Here’s what to know about one of the lesser-known bellwethers of U.S. elections:

Why Washington?

Unlike the primaries of other states, Washington’s are strong predictors of its general elections. This is true for a few reasons.

• Washington has a top-two system in which all candidates, regardless of party, compete on the same ballot, with the top two advancing to the general election. This incentivizes candidates to campaign harder, to ensure they’re on the final ballot in November. And it motivates voters to participate rather than risk having their preferred candidate – or even their party – knocked off the ballot if they don’t vote.

• Voting is conducted almost completely by mail, and voters have almost three weeks to vote, which lowers the bar for participation. (Primaries that rely mostly or exclusively on in-person Election Day voting are more likely to attract only the most engaged sliver of the electorate.)

• The primary date is usually in August or September, much later than most other primaries, leaving less time for moods to change before the November election.

There are states with similar primary structures, and there are states with later primaries. But no state has Washington’s combination of primary structure, large voter numbers and late calendar date. Put them all together, and you get a high-turnout election with a broad, relatively diverse electorate just a few months before the rest of the country votes for president.

What does Washington’s primary tell us this year?

If you used only Washington’s primaries to forecast voting trends in the U.S. House, you’d do relatively well: The primary’s subtle shifts left or right have tracked with the country’s in all but two election cycles since 2000.

But Washington state and the country don’t just move in the same direction. They also tend to move one way or another by similar degrees. If after every Washington primary election between 2000 and 2022 you predicted that the country as a whole would shift left or right by the same level as the state’s primaries, you would have been off by less than 3 percentage points, on average. Pretty good!

Last month, Washington’s House Democrats received 58% of the major party vote share across the state’s 10 House districts – 3 points better for Democrats than their result in 2022.

If that shift translated perfectly to the national vote, that would mean a move 3 points to the left of the 2022 House vote (when Republicans narrowly captured the House), lifting Democrats to 51% of the popular vote from 48%. That total would be about where Democrats finished in 2020 and 2 points better than Democrats’ result in 2016.

An analysis by SplitTicket, an independent election analysis site, went further, examining election results only in the nonurban parts of Washington state. They found that if you remove Seattle, Washington looks demographically similar to the key swing states of Wisconsin, Michigan and Pennsylvania. Since 2012, the primaries in those nonurban districts have consistently tracked the national House vote in November, though they are typically 3 to 5 points further to the right.

As with the rest of the state, the districts outside of Seattle have shifted modestly to the left since 2022, and are slightly to the left of where they were in 2020, perhaps another piece of good news for Democrats.

Can it tell us anything about the presidency?

Only to a point. The Washington primaries are helpful in thinking about the general mood of the country as it translates to the national popular vote, but it’s less useful as a predictor of presidential elections, where specific candidates, issues or economic conditions make a bigger difference. And, more important, there’s the Electoral College. As we know from the 2000 and 2016 presidential elections, the popular vote in a close race is only a partial indicator for how the most important battleground states will go.

As it happens, the 2016 primary results in nonurban Washington suggested Hillary Clinton might underperform in the Northern battlegrounds. She did.

Why Washington might be less predictive in 2024

There are other reasons to be cautious about the predictive power of Washington’s primary this year.

• We don’t know to what extent the August primary results were influenced by a highly turbulent July. President Joe Biden announced his decision to exit the race July 21, and the immediate Democratic excitement over Kamala Harris’ ascent to the nomination may have increased turnout in Washington’s primary, reflecting an early enthusiasm that may not be matched by Election Day.

• A lot has already happened since that primary and still might. Since Washington’s primary, the nation has seen the Democratic National Convention, the first debate between Harris and Trump, a Taylor Swift endorsement and a second assassination attempt against Trump.

• There’s also no guarantee that Washington’s primaries will continue to move in parallel with the country at large. In an earlier era, between 1978 and 1998, Washington and the country as a whole trended in opposite directions more than they trended together.

If recent trends hold, however, the Washington primary suggests the country will shift 3 points to the left from 2022 – plus or minus 3 points.

That’s roughly in line with today’s polling averages, whereas in 2020 the primary results correctly suggested that Biden’s big polling lead was overstated.

Still, that’s a big margin of error for the most hotly contested House races, and a presidential race that is essentially a tossup.

This article originally appeared in The New York Times.