Americans prefer Democratic House candidates in 2018, for now

Throughout the Trump presidency, Democrats have had one glimmer of optimism looking ahead to 2018. Polls continue to show that the party is well ahead of Republicans on the "generic ballot" — the term for when pollsters ask voters which party they would like to win the House of Representatives in the next election, or which party's House candidate they would likely vote for.

Last week, an ABC News/Washington Post poll showed Democrats with a 14-point lead on the generic ballot. That's one of several polls since May that have shown Democrats with a double-digit lead in the generic ballot. Altogether, Democrats lead Republicans by around 9 points in the generic ballot, according to the latest average from RealClearPolitics.

That's likely encouraging for many Democrats after suffering stinging losses in 2016. But how big of a deal is that?

One simple (and, admittedly, simplistic) gauge is to look at past performance. In both 2006 and 2008, Democrats went into Election Day with a lead of 9-plus points...and gained 30 and 23 House seats, respectively, according to data compiled by Kyle Kondik, managing editor for Sabato's Crystal Ball at the University of Virginia. To win the House, Democrats would have to win 24 sets next year.

This, of course, comes with the usual caveats of any well-ahead-of-time midterm election story: we're still 15 months from Election Day 2018, lots could change and so on. But the point here is that nine points has in the past translated to a wave of House seat gains for Democrats.

On top of all that, there are a couple of other encouraging signs for Democrats: for example, in midterm elections, there is almost always a backlash against the party in power in the White House.

Since 1934, the president's party has only ever gained House seats in a midterm three times, according to data from the University of California Santa Barbara's American Presidency Project. And those gains are always small; they are never as big as the sometimes-massive losses that presidents' parties suffer in midterms.

Trump's low approval rating — right now, at around 37 percent, according to Gallup — could be an additional source of hope for Democrats as well. But as FiveThirtyEight's Harry Enten noted earlier this year, the correlation between approval rating and House seat gains is "rough."

(And as with the generic ballot numbers, Trump's approval rating could turn around in the next year-and-a-quarter.)

One potential — but surmountable — hurdle for Democrats is that, altogether, House seats skew more Republican than the nation as a whole.

"[Democrats] are gonna have to win some seats that are more Republican than the national average," Kondik told NPR. "Which I think they can do; it happens in midterm elections all the time," he continued.

Kondik provided data illustrating this. For recent election years, he determined the median House district in terms of presidential performance — that is, if you rank districts from the most Democratic to the most Republican in terms of presidential votes, the median is the 218th out of 435, or the exact middle district.

What he found is that in 2004, 2008, 2012 and 2016 the median House district was more Republican than the total national performance was. In 2016, for example, the median district was Iowa's First District. Trump won that district by 3.8 points, while the rest of the nation voted for Clinton by 2.2 points — a six-point differential.

That might seem like a good sign for the GOP for 2018, but not necessarily — in 2004, for example, the median district was 5 points more Republican (by this measure) than the nation was. But in 2006, Democrats nevertheless picked up 23 seats.

And yet: there are a few reasons why Americans' seeming preference for a Democratic House shouldn't get Democrats too excited just yet.

One is the simple benefits of incumbency. In any given race, the incumbent often has natural advantages in terms of fundraising potential and name recognition. Considering that Republicans hold more districts than Democrats, that's one simple hurdle to retaking the House for Democrats.

It's also true that votes don't necessarily equal seats. In every election since 2010, Democrats have won a smaller share of congressional seats than votes. This is a relatively new trend — as the Brookings Institution found in November, Democrats used to regularly win a "seats bonus."

In other words, even assuming Democrats lead decisively in the generic ballot in November 2018, and assuming that lead translates to a greater vote share, it won't necessarily mean a similarly sizable gain in House seats.

And, of course, it is more than fifteen months until Election Day 2018. At this point in 2013, Democrats had a lead of around 3.5 points in the generic ballot. They would go on to slip behind Republicans by around 2.4 points just before Election Day 2014 — and lose 13 seats.

One final point: Democrats generally have a tougher time turning out voters in midterms than Republicans do.

It's true that recent special elections in Georgia, Kansas and Montana showed that Democrats could mobilize — if not win — in some heavily Republican districts.

But at least for now, Democrats nationwide don't look particularly energized for 2018. While Democrats lead by 14 points in the generic ballot, Trump supporters, in particular, are far more likely than Democrats to say they are "absolutely certain" to vote in midterms, as the Washington Post's Mike DeBonis and Emily Guskin wrote. Seventy-two percent of Trump supporters said this, compared to 65 percent of all Republicans and Republican-leaners, and just 57 percent of Democrats and Democrat-leaners.

So once again, even if Americans prefer Democratic House candidates in November 2018, it won't mean a thing if those people don't get out to their polling places.