The answer is actually pretty simple: Our elections increasingly look more like parliamentary ones, and given that, the results make a ton of sense.
New data from the election-reform group FairVote sheds some light on how the battle for the House played out. The big takeaway: Our politics are increasingly less about people and incumbents and more about party. We’ve been talking about increased polarization for many years, but the 2020 election really drove it home. The results for Congress affirm the fact that Republicans writ large lost the election, even though it might have been closer than many expected.
FairVote has for years studied an issue called “incumbency bump” — i.e., how much an incumbent benefits relative to other members of their party thanks to already being in office. The conventional wisdom on incumbency is that it’s a big advantage — that people might not like a politician’s party or Congress as a whole, but if they know that politician well or have any doubts, they’ll revert to supporting the person in the seat.
The 2020 election, though, continued an increasing decline in that advantage.
FairVote’s formula for “incumbency bump” is somewhat complex; it combines how much voters favor incumbents overall in a given election and the incumbent’s performance relative to how an average candidate would have been expected to do based upon past results.
While incumbents gained nearly eight percentage points from this advantage in the 2000 election, that number has steadily dropped to 1.5 points in the 2018 midterm elections and 1.4 points in the 2020 election. In other words, there is increasingly little daylight between being any old candidate with an “R” or a “D” next to your name and someone with those labels who currently holds the seat.
Perhaps more telling: The overall 2020 results also suggest not just that incumbency matters increasingly less, but also that candidates matter increasingly less. FairVote’s data compares the results in each district to how that district would be expected to perform based on recent results. In races in which we had candidates of both major parties — i.e. “contested” ones — none of them deviated more than 11 points from what you’d expect the winner to get.
Fully 350 races had winners within five points of where you’d expect a winner to be. Another 26 races didn’t even have two major-party candidates. That means only about 60 out of 435 races deviated from a predictable result by more than five points.
Some of the ones that did were certainly interesting. For example, three of the six Democratic incumbents who most underperformed the fundamentals in their districts were two members of “The Squad” — Reps. Alexandria Ocasio-Cortez (D-N.Y.) and Ilhan Omar (D-Minn.) — whose GOP opponents raised gobs of money despite running in uncompetitive districts, along with another popular target of Republican donors, Rep. Maxine Waters (D-Calif.). A fourth of the six was House Judiciary Committee Chairman Jerrold Nadler (D-N.Y.), who drew criticism for his handling of impeachment-related issues.
A Republican who did remarkably well, by contrast, was Rep. Adam Kinzinger (R-Ill.), who has emerged is a strong critic of Trump’s more controversial impulses. Many Republicans before him have paid a price with the GOP base and often headed for retirement, but Kinzinger exceeded his expected performance by more than five points — the ninth-best overperformance by a winning Republican.
The vast majority of Congress, though, inhabited the mushy middle. Given that, it isn’t a huge surprise that Republicans gained ground despite losing the presidential race.
After all, Republicans were starting at a significant deficit thanks to their losses in the 2018 midterms. And after both races, the balance of power closely reflected the popular vote. Democrats in 2018 won 54.1 percent of the two-party vote and 54 percent of the seats. In 2020, they won 51.5 percent of the vote and 51.2 percent of the seats.
We tend to look at these things in terms of party shifts from the previous election rather than the final shares of a chamber, especially given the expected advantage of incumbency. Increasingly, though, it looks like the real measure is not how much things changed, but where they finished regardless of the starting point.