California Will Have Clues for the Nation: How to Watch the Primary

The California primary elections Tuesday might give us a better idea of whether Democrats are headed for a so-called wave election — or whether their recent slide in the generic ballot and the bump in President Donald Trump’s approval rating mean they should worry.

The quirks that make the California primary risky for Democrats also make it a leading indicator of the general election.

In California’s nonpartisan, top-two system (also sometimes called a jungle or blanket primary), any voter can cast a ballot for any candidate, regardless of party, and the top two candidates advance to the general election, again regardless of party. Historically, these top-two primaries look a lot like the general election, when any voter can similarly pick any candidate.

Since 1990, the major party vote share in top-two congressional primaries in Washington (which also uses the top-two system) and California has differed from the general election result by an average of just 3 percent.

Of course, you can’t make an ironclad prediction about the general election based on these top-two primary results. An average error of 3 points on major party vote share (or 6 points of margin) means that most close results are still a tossup and that there are occasionally outlying results. The error is also bit larger in open seats (4 points) than in those with an incumbent (2.9 points).

But this is about as good as any data we’re going to get. The average House poll over the final three weeks of an election is off by an average margin of 6.2 points, according to FiveThirtyEight. So the primary results are a bit like getting a free round of more than 50 final House polls in early June.

And in contrast to late House polls, the direction of bias in the top-two primary is fairly predictable. There is generally a slight Republican bias on the order of a little more than a point, presumably because the primary electorate tends to be somewhat older and whiter than the general electorate. The exceptions to this pattern have come in years with one-sided presidential primaries, like 2016 when Democrats had a competitive presidential primary race after the Republican primary was over.

The results are good enough that you can put some stock in a surprise. In 2016, for instance, Darrell Issa’s seat was rated “Safely Republican” by the Cook Political Report heading into the primary. But he ended up claiming just 50.8 percent of the vote, to 45.5 percent for the Democrat, Doug Applegate, presaging the closest House election of the cycle. In the end, Issa won re-election with just 50.3 percent of the vote. If this is a wave environment like in 2006 or 2010, which would probably make the Democrats slight to modest favorites to retake the House, it shouldn’t be too hard to tell. Seat by seat, Democrats would probably be pretty close to half the vote in the seven California districts that Hillary Clinton won in Republican territory in 2016.

Perhaps Democrats would claim an outright lead in a district or two of the seven, like California’s 49th or 39th. Given the historical Republican bias in the primary results, that would make the Democratic candidate a 2-to-1 (or better) favorite to win in the general election. A district like California’s 10th or 48th might be close. California’s 21st and 45th might lean Republican, but it would be close enough — say, with Republicans claiming around 53 percent of the ballots — that the Democrats would be in range of a victory in a general election.

Conversely, the chance of a Democratic pickup would fall beneath 25 percent if their candidates fail to receive 55 percent of the vote. With all seven of those seats rated as “Lean Republican” or more competitive, a Democratic tally under 45 percent would merit re-evaluating the competitiveness of a race in the absence of meaningful contradictory evidence.

It’ll be interesting to see whether Democrats can pull within striking distance in any of their long-shot options, like California’s 50th or 4th. If Democrats have a shot against Devin Nunes (which appears unlikely) in the 22nd district, we should know.

In the aggregate, the easy way to judge the result is to look at the swing from the last election. Nationally, Democrats would be favored to win control of the House if they could gain an average of 4.5 points across the country, according to our estimates. A 3.5-point improvement between the 2016 general election and the California top-two primary results might put them on track to do so.

Historically, this approach would be pretty effective, even though California and Washington aren’t necessarily representative of the country. It would be as useful as looking at generic ballot polls, which ask voters whether they will vote for Democrats or Republicans for Congress. But given the fairly small number of elections in this analysis (just 11), it would probably be wise to be cautious about reading too much into the national picture. A weak turnout among Hispanic voters, for instance, could hurt Democrats a lot in California while doing little to hurt their chances elsewhere.

Here’s another cautionary note: It can take days and sometimes weeks — yes, weeks — for California to finish its count and late ballots often favor Democrats. So be careful about reaching any premature conclusions on election night, or even in the days after, particularly if those conclusions seem quite favorable to Republicans.