You’re going to hear a lot about something called the “top-two” primary in the lead up to California primary. Tuesday’s contest in the Golden State has been described as a potential “disaster” for Democrats in their march to pick up a net gain of 23 House seats and the majority in November.
And while that could eventually be the case, the truth is we have so little data as to be almost totally in the dark when looking at these races, and a look at the little data that is available suggests that such fears are probably overblown.
To understand why Democrats worry about the top two, you first have to realize what’s going on.
Simply put, all the Democratic candidates in a district and all the Republican candidates in a district run against each other in a primary, regardless of party affiliation. The two top vote getters advance to the general election in November, even if they are members of the same party. If there are more members of one party running than the other, it strengthens the chance that even if a particular party has fewer supporters its candidate can advance to the general election because so-said party is splitting its votes among fewer candidates.
It’s this vote splitting among many different candidates of the same party that is the big fear for Democrats in three Republican held districts: California 39 (where Ed Royce is retiring), California 48 (where Dana Rohrabacher is running for re-election), and California 49 (where Darrell Issa is retiring). These are all districts Hillary Clinton won, so they’re seen as big Democratic pickup opportunities.
Now obviously Democrats would not want two Republicans to advance in these districts, but it is important to put this all in perspective.
Winning or losing three seats is probably not going to make or break a House majority for either party. Right now, there are around 80 Republican held seats where Democrats are at least somewhat competitive (i.e. which CNN does not rate as solidly Republican). These three California seats are just a small chunk of that.
Put another way, if you were to look at the generic congressional ballot, losing three seats is roughly the equivalent of losing a point in national polling. Losing a point can make the difference in some elections (see the 2016 presidential election), though not in most.
Three seats is about how many seats Democrats are slated to gain because of the redistricting put into effect by the Pennsylvania Supreme Court earlier this year. When that decision came down, I noted that “Pennsylvania’s new congressional district lines are not a game-changer”.
And even if a Democrat and a Republican advance to the next round in these districts, it’s no guarantee that Democrats will win all of these districts. Yes, Clinton won these districts by margins ranging from 2 to 9 points, but Barack Obama lost all three of them in 2012. Obama lost both the 39th and 48th in 2008 and won the 49th by a point, even when he was winning nationally by 7 points. In other words, these are districts that traditionally lean Republican except for in 2016.
It’s not entirely clear whether the Democrats’ path to a majority necessarily goes through districts that have normally voted Republican but went for Clinton. In federal and state special elections, there has generally been a higher correlation with how a district or state voted in 2012 than in 2016. You might remember that Democrat Jon Ossoff came up short in the Georgia 6th congressional district special election, which took place in a district that was normally very Republican and Clinton barely lost in 2016.
That’s why it might actually be good news for Democrats that a top two primary is taking place in these districts. These are races where it’s not clear Democrats would necessarily win a one-on-one against the Republicans, and the very limited data suggests Republicans could be locked out of the general election in California 39 and 49.
Keep in mind no public polling has been released recently in these districts. We only have internal polling, which should always be looked at with a skeptical eye. Still internal polls in which pollsters were transparent about their data (as is the case with these polls) is better than internal anecdotes from the campaigns.
In the 39th, only a poll sponsored by Democrat Gil Cisneros was taken and released in May. It had Cisneros at 20%, then Republicans Bob Huff and Young Kim and Democrat Andy Thorburn all between 11% and 14%. I wouldn’t read too much into the Cisneros advantage here given he sponsored the poll. What I would read into it is that, given how un-predictive primary polling is, it’s conceivable that any of these candidates advance to a November general election. It could be two Republicans, though it could easily be two Democrats as well. Probably the most likely result is a Democrat and a Republican.
In the 49th, polls sponsored by Democrat Sara Jacobs and a group urging Democrats to vote for Democrat Doug Applegate show a similar mess. Both polls have Applegate, Republican Rocky Chávez, Republican Diane Harkey, Jacobs and Democrat Mike Levin all within a range of 10% and 16%. That means that like the 39th, the 49th could see two Democrats, two Republicans or one Democrat and one Republican advancing to the general election in November.
Indeed, the only the district of these three where Republicans looked assured of a spot is the 48th. Rohrabacher has been polling well ahead of all the Democrats in the field there, even in their own polling. Polls sponsored by Democrat Hans Keirstead and Democrat Harley Rouda and taken in May have Rohrabacher around 30% and Republican Scott Baugh, Keirstead and Rouda between 13% and 15%.
Keep in mind, though, that district is the one of the three where Democrats will have the hardest time winning in the fall. Clinton won it by only 2 points (compare to the 8 and 9 point margins she won in the 49th and 39th respectively), and incumbents generally run better than the district’s fundamentals would suggest.
Even so, it’s far from clear that Democrats are going to get locked out in the 48th.