“Democrats were projected to gain 15 House seats and instead — and [House Republican Leader] Kevin McCarthy gives us the credit, good man, very good man — my coattails, we swept our Republican House candidates to victory,” Trump said at a rally in Georgia on Jan. 4. (The two GOP Senate candidates Trump was there to support both lost their runoff elections the next day.)
Trump’s relationship with McCarthy soured a bit two days later, but you get the gist. Trump was hyping how well his party did and pointing to his coattails as the cause.
It is true that Republicans overperformed in House races, and it may be true that this was a function of Trump’s presence on the ticket. But a more accurate way of describing the results in the House may be to say that they were more tightly partisan than in any other election in the past several decades.
You’re no doubt familiar with the idea of split-ticket voting — that voters will sometimes vote for a Democrat for one office and a Republican for another. In 2020, that happened far less frequently than normal. In fact, data released by DailyKos on Friday shows that the presidential results in House districts hewed very closely to the presidential results almost across the board.
It’s immediately apparent when graphed. Each dot is a House district, and the closer to the white dotted line a dot falls, the more closely the House result matched the presidential result.
The districts that split their results — electing a Democrat to the House while voting for Trump, for example — appear at upper left and bottom right. As you can see, there aren’t many of them.
A few districts had unusually large disparities between their House and presidential results. We identified a few. Two were in Minnesota, which isn’t really a surprise: Excluding House districts in which the results were lopsided (generally because of a candidate running uncontested), the average gap between the House and presidential margins in Minnesota was 9.8 points, the largest of any state. In Utah, the average gap was 9.5 points and in Hawaii, 8.7 points.
You may look at the graph above and think that it looks the way you’d expect. Why wouldn’t voters who preferred a Democrat for the House similarly want to support the Democratic presidential nominee? But such a tight correlation between the two votes simply doesn’t match past elections. If we compare the House and presidential results for each election since 1992 (using the preceding presidential results in a district for midterm years), you can see that it’s only in the past few cycles that the dots have clustered around that white dotted line.
To put a fine point on it, the DailyKos data suggests that only 16 House districts saw split results in 2020 — 10 fewer than split their vote in 2012, the year with the next-fewest such districts.
Since 1992, there have been an average of 70 split districts in each cycle, though the average since 2012 has been only 29. 2020 stands out even in recent history.
Despite the infrequency of split tickets in 2020, there were some notable shifts since 2016. Most House districts — 317 of them — voted more heavily Democratic in the presidential contest than they had four years previously. This isn’t a surprise; Joe Biden was elected, after all.
What’s interesting about those figures, though, are the outliers. Those two districts way at the right are Florida’s 25th and 26th districts, places where Biden fared unexpectedly poorly (perhaps because of large populations of Cuban Americans). A lot of those big blue circles (indicating larger Biden win margins) fall to the right of the centerline. In heavily Democratic areas, Trump often did better than he had in 2016, thanks to improvement with non-White voters relative to his first election.
Overall, districts that voted for Biden and a Democratic House candidate saw the House candidate do slightly better than Biden. (Here again we’re excluding the results in districts where House candidates ran uncontested.) The same held in Trump/Republican districts: There, the House Republican’s margin was about four points better than Trump’s. In those 16 split districts, though, the margin was more favorable to the presidential candidate by much larger margins.
That, too, makes sense: The presidential candidates won, while their party’s House candidates didn’t.
Because there were only 16 split-ticket districts, eight for each party, the number of House districts won by Biden and the number won by Trump were the same as the number of seats won by the Democrats and the Republicans. Trump may have had coattails — but there’s no reason to think they were any longer than his opponent’s.