How Democrats Won The House

Midterms 2018: FiveThirtyEight reacts to election results

After two years of pent-up anticipation, Democrats have finally done what they have long been favorites to do: win control of the U.S. House of Representatives. As of this writing, our colleagues at ABC have projected 223 seats for Democrats and 201 seats for Republicans. Democrats have turned 29 Republican-held seats blue, while Republicans have flipped one Democrat-held seat, for a net Democratic gain of 28 (so far). Here are all the seats that have changed parties as of 3:43 a.m. on Wednesday:

Districts we saw as “toss-ups” split slightly in favor of Democrats, and many remain too close to call:

Districts that we rated as “lean” or “likely Democrat” (i.e., where the Democratic candidate had between a 60 percent and 95 percent chance) usually voted for the Democrat, but not always — which is exactly what should happen!

And districts that we rated as “lean” or “likely Republican” generally voted for the Republican — again with a few exceptions, which is what you’d expect.

Those three Democratic lightning strikes in “likely Republican” territory qualified as some of the night’s biggest surprises. Republican Rep. Dan Donovan was a 4 in 5 favorite in New York’s 11th District, but Democrat Max Rose won 53 percent to 47 percent. Republican Katie Arrington had an 11 in 12 chance of winning the South Carolina 1st, but Democrat Joe Cunningham eked out a 51 percent to 49 percent win. And in the Oklahoma 5th District, our forecast had given Republican Rep. Steve Russell a 14 in 15 chance to return to Washington, but Democrat Kendra Horn toppled him 51 percent to 49 percent. One thing all these districts have in common? They are all anchored by urban areas: New York City, Charleston and Oklahoma City, respectively.

Indeed, a theme of the evening was that suburban areas came up big for Democrats.¹ We’ve often used so-called “Romney-Clinton districts” as a stand-in for these areas — districts that voted for Republican presidential candidate Mitt Romney in 2012 but switched their allegiance to Democrat Hillary Clinton in 2016. Republicans had hoped that these places had voted for Clinton because of an aversion to President Trump and that they would remain loyal to their more traditional Republican representatives. That didn’t end up being the case. Not only did Democratic House candidates win most Romney-Clinton districts, but in at least six of the 13 races, they did so by margins that exceeded Clinton’s margin over Trump.

What’s more, there are still thousands of votes outstanding in California because so many voters there vote by mail and many ballots are probably still en route to elections offices. So those three Romney-Clinton districts in California where the Republican is currently leading could still flip Democratic as well.

Compare this to how Democrats performed in Obama-Trump districts, which are mainly working-class areas that disproportionately fall in the Midwest and Northeast. At first glance at the table below, it looks like Democrats did pretty well in these districts. But they were aided by the incumbency advantage in five of these districts.² In open or Republican-held Obama-Trump districts, Democrats have officially picked up eight seats.³ Republicans held on to four others and flipped at least one Democratic-held Obama-Trump seat red (the Minnesota 8th), so it was more of a mixed bag.

Finally, what about the races where no winner has been projected? Several of them are in California, which, as I mentioned, may take several days to count up all its results. Others, like Minnesota’s 1st and our old friend Georgia’s 6th, are simply too close to call. One uncalled race might even require special rules to resolve: In Maine’s 2nd District, it looks like no candidate will get a majority, which means that Maine’s ranked-choice voting rules will come into play. The second- and possibly third-place choices of the voters who supported the race’s two independent candidates will be redistributed between Republican Rep. Bruce Poliquin and Democrat Jared Golden. Currently, Golden leads 47 percent to 44 percent, but it could very well be Poliquin who ends up winning under these rules. Between that and California, it’ll be a bit before we know the exact margin that future Speaker Nancy Pelosi (or whoever!) will have to work with in 2019.