The million dollar question after every primary night of the 2018 midterm cycle so far: are Democrats on track to flip the House in November?
Data from last night’s primary elections in California say “yes — well, probably.”
Democrats will mostly likely avoid their disaster scenario in California, averting a “lockout” — a case in which no Democratic nominee makes it onto the ballot for the fall, a result of California’s atypical primary system — in three important contests: the 39th, 48th, and 49th congressional districts.
This would be a so-called “moral victory” for Democrats on its own, but where we find the real news is (you guessed it!) in the data.
Because of the strong relationship between the Democratic share of primary ballots in California and their eventual vote margins in the district, last night’s results (which have yet to be fully counted — more on this below) points to competitive elections in eight key California districts. Of these, Democrats were already favored to win three, with another five coming on the tail end of probable pickups in the November midterm elections.
Here, I’m showing you that there is a very strong relationship between Democrats’ June margin in the two-party share of the primary vote and their November margin in the two party vote in the general election. Statistically, this relationship is a 0.97 correlation (almost as high as you can go) over the past three election cycles, and it was even stronger in 2016 than in 2012 (because of an especially salient Democratic presidential primary between U.S. Senator Bernie Sanders and former Secretary of State Hillary Clinton). The folks at Sabato’s Crystal Ball and the University of Virginia’s Center for Politics were nice enough to share this data with me.
Notice that the points roughly cluster around the line running at a 45-degree angle upward — this is the line of perfect fit between the primary and general elections. In other words, if every point landed on that line, the Democratic margin in the total two-party primary vote would perfectly predict their two-party margin in the November general election vote.
However, the line we draw between the actual data is not perfect. Instead, it runs slightly above the line of perfect fit, intersecting the y-axis at 5.7 and thus indicating a typical 5.7% shift toward Democrats in margin of votes between the primary and general elections.
This means it is typical that Democrats in California slightly outperform their primary election performance in November.It also has a slightly less than 45 degree slope to it, indicating that districts where Democrats do better in the primary vote tend to give them worse numbers in the general election, with the reverse being true in districts where Republicans do well in the primary vote.
For 2018, that means Democrats can expect a slightly better performance in the general election than they saw last night. Below, I zoom in on the districts that are favored to flip from red to blue, based on last night’s primary results alone, and those that are within the margin of error for my predictions of vote margin in November (again, based only on this data).
|Dem Margin in Votes
|Seat||June Primary||Predicted Nov. General||Party|
It is evident that Democrats are positioned well to make gains in California in the November midterm elections, however, those gains are not certain, and there is a wide 15% margin of error around the projections of Democrats’ midterm votes. Anything can happen, we like to say, and it is unwise to follow any forecast blindly.
In addition, keep in mind this caveat: there are still many hundreds of thousands of ballots left to count in California. By the weekend after the 2016 primaries, some analysts estimated there were still 30% of votes remaining across the state. This year, that number is likely to be lower — turnout this cycle is much lower than in 2016 — but totals could still change by the time it is all said and done. However, large swings aren’t typical, and the current Democratic margin in the two-party primary vote is likely about 1-2% away from the final numbers. Additionally, outstanding ballots tend to be from large cities with more Democrats than Republicans, so if the movement goes in any direction it should ought to lean left.
Primary vote totals are just one of many indicators of how elections can play out. This year, the numbers suggest that Democrats are on track to flip 3-4 Republican-held seats in California, with an 8-seat gain very possible. The average swing in the 11 competitive seats detailed above was 5.5% — combined with the projected 5.7% drift left from now until election day, the Democratic Party could be poised for a double swing come election day on November 6. However, given the uncertainty in the data, if you were already thinking Democrats were going to do well (e.g. a 7-8 point win margin in the national popular in November) your prior beliefs about the race ought to remain roughly the same.
Per usual, we’ll just have to wait and see what happens.