Unless you’ve been living in a cave for the last six weeks, you know that the Democrats won a pretty smashing victory on Election Day, picking up a net 40 seats in the House of Representatives and overperforming most prognosticators’ expectations. You’re probably also aware that the Republican casualties weren’t randomly distributed among their entire caucus, or that Republican losses weren’t tilted toward their loudest-mouthed doofuses and their most incompetent political practitioners. (Granted, Steve King nearly managed to lose, quite an accomplishment given how dark-red his district is. But, for instance, Louie Gohmert’s asparagus managed to escape without any aspersions being cast upon it.)
Instead, the losses were heavily concentrated in the nation’s swing districts, the ones closest to the national political midpoint. Accordingly, the Republicans who lost were disproportionately among the most moderate members of the GOP caucus—not because moderation is more off-putting to voters in swing districts, though.
Elected officials in swing districts tend to be more moderate because they believe it’ll help them get re-elected (and, if multiple political science studies are correct, that's true). Any moderation advantage, however, is small, and as political polarization continues to increase, it becomes even more likely that if you’re in the wrong party in a closely-divided district in a wave year, you’re probably dead regardless of how much you try to distance yourself from your party. (Just as many moderate Democrats found out to their chagrin in, for example, 2010.)
The orderly precision with which a wave strikes swing district representatives down, though, regardless of their voting record and regardless of their skill as a campaigner, is pretty abstract unless you see it depicted visually. So, I’ve tried to create a graphic that shows just that. Try thinking of the House as a coastal town, with some streets lined with oceanfront property and some set further back from the water. With a big enough wave, just about everything in the first few blocks facing the water gets wiped out. Compared with previous decades when there was more ticket-splitting, it just doesn’t matter much anymore whether you’ve tried to build your own durable brand on a solid foundation of moderation or good constituent service or aw-shucks demeanor; the wave simply doesn’t care. The dark blue blocks, mostly in the middle of the chart, are where all the wave damage is: in other words, the seats that the Democrats picked up in 2018.
On my graphic, each successive “street” is rows of 10 congressional districts each, ordered from left to right and then from top to bottom according to presidential results in 2016 (using Daily Kos Elections’ calculations). The upper-leftmost block is New York’s 15th congressional district, where Hillary Clinton got 93.8 percent of the vote; the lower-rightmost block is Texas’s 13th district, where she got 16.9 percent of the vote. In the very middle—the 218th seat, halfway between the two ends—is Nebraska’s 2nd district, where Clinton got 46 percent of the vote (oddly, it’s one of the few remaining red blocks left on the center line; I’ll have more on that shortly).
The light blue blocks on the chart are the seats that the Democrats already held (you can think of them as the places that were already underwater, even before the tide came in). You’ll also notice, to the right of the midpoint, several shades of red. The dark red blocks are the ones that were competitive races (according to our 2018 qualitative ratings) but where the Republican won. The light red blocks are the Republican-held seats that were, according to our qualitative ratings, uncompetitive (which doesn’t mean that they were uncontested; it means that there was no solid evidence, in terms of fundraising, polls, and so on, that the Democratic candidate had a shot at winning—and, in fact, no Democrat won a race that we listed as “safe Republican”). There are also two yellow boxes, which are the sad duo of seats that the Republicans picked up.
Looking at the 2018 results this way, you can really see the precision with which the wave moved through the first few ranks of the Republicans and then stopped, right around the point where Clinton was getting around 44 percent in 2016. There are several more “streets” set further back from the water where there were still a lot of competitive races, though where the Democrats didn't win many more seats, extending down to around Clinton’s 41 percent mark. Beyond that, though, almost all the remaining Republicans were left high and dry.
You’ll also notice this isn’t completely uniform; there are still a few Republican ruins sticking up out of the water, and a few Democratic puddles far inland. There are still five remaining GOP-held districts with presidential numbers that are bluer than the national median among house districts, and there are decent explanations in each of these cases. Texas’s 23rd district may be the most surprising survivor (though Democratic nominee Gina Ortiz Jones did manage to get much closer in the end than polls in the 23rd predicted); this is a heavily-Latino district in the San Antonio area, though where turnout simply wasn’t quite high enough to get her over the top, at least not compared with the bonkers turnout we saw in California that helped Democrats pick up similar districts like CA-10 and CA-21.
There are also three districts where there were hotly-contested progressive vs. moderate primaries where the progressive won but then had trouble pivoting after the primary to go against established, moderate GOP incumbents (Pennsylvania’s 1st district (Scott Wallace), New York’s 24th district (Dana Balter), and Nebraska’s 2nd (Kara Eastman). Finally, there’s Florida’s 25th district, which is a tougher nut to crack because its mostly Cuban-American population tends to be more supportive of Cuban Republicans running downballot than of national Republicans. (Even in the other two mostly-Cuban districts that Democrats picked up this year, Florida’s 26th and 27th, the Democratic nominees won by much narrower margins than the presidential numbers would presage; as you can see from the chart, they were the second- and third-bluest districts the Democrats picked up this year, behind only the newly un-gerrymandered Pennsylvania’s 5th district.)
As something of a mirror image, there were also five very red districts where Clinton got 40 percent or less in 2016, but which were Democratic pickups this year. (And these are probably also the five districts that are likeliest to flip back to the GOP in 2020, if the Democrats don’t maintain the same level of support on the generic House ballot.) Two of these seats were open seats, where there was also a particularly appealing Democratic candidate, New Mexico’s 2nd and South Carolina’s 1st districts. (SC-01 is an open seat to the extent that Mark Sanford lost in the GOP primary, and continued sniping between Sanford and nominee Katie Arrington probably helped Joe Cunningham sneak through in the general.)
Two more of these seats saw Republican incumbents who were either totally sleepwalking (Steve Russell in OK-05) or too odious even for her dark-red seat (Claudia Tenney in NY-22). And, finally, much further to the right of even those other four, is Utah’s 4th district, where Ben McAdams defeated Mia Love despite Clinton’s measly 32.4 percent here. McAdams, who’s the mayor of Salt Lake County, benefits from already representing most of the 4th’s residents at the municipal level. But the main problem for Love here was her failure to adequately separate herself from Donald Trump, deeply disliked in this heavily Mormon district. (In fact, the 32.4 percent may be a rather misleading number in this case, because Trump only got 39 percent here, with most of the balance going to Evan McMullin.)
Finally, there are the two districts that the Republicans picked up from the Democrats. Here, the chart shows just how difficult those would have been for the Democrats to hold; these districts are redder, at the presidential level, than anything the Democrats picked up other than UT-04. Compounding that, Minnesota’s 1st and 8th districts were open seats, vacated by Tim Walz for his successful gubernatorial run and Rick Nolan’s retirement. These rural and poorly-educated, but once-reliably Democratic, seats simply fell off too far in the 2016 election, and they didn’t snap back adequately enough this year.
(Some sources you might see would say there’s a third one that the Republicans flipped, Pennsylvania’s 14th district, left open after redistricting. There are two ways to think about this and the neighboring 17th in Pittsburgh’s suburbs: one could say that Democrat Conor Lamb picked up the 17th by defeating Republican incumbent Keith Rothfus, while Lamb vacated the 14th, the geographical successor from the bulk of his old 18th district, which GOPer Guy Reschenthaler gained. You could also, say if you prefer, though, that Lamb’s victory was a Democratic hold while Reschenthaler’s win was a Republican hold.)