Many Questions About Size and Reach of Democrats’ Wave

There shouldn’t be much doubt that the Democrats are riding a wave heading into November’s midterm elections.

That was clear well before Conor Lamb’s apparent victory in Pennsylvania last week. There have been nearly 100 special elections over the last year, and at this point there’s so much evidence of Democratic strength that there’s not much more to glean from each additional result.

But the special elections haven’t taught us everything we need to know about the midterms. In particular, there are two big outstanding questions that the special elections can’t answer. Can Democrats sustain their performance in a higher-turnout general election? And will Republican incumbents run ahead of the national party by enough to survive?

The answers will determine whether the fight for control of the House is a Democratic-leaning tossup, as history and national polls imply, or whether Democrats are poised for a tsunami, as the special election results suggest.

So far this year, Democrats have often had a huge turnout advantage in special elections, based on an Upshot analysis of voter file data from L2, a nonpartisan voter file vendor.

The data is still fragmentary. It covers only about one-third of the special elections over the last year. But in states with party registration, the Democratic registration advantage in 2017 special elections was about five percentage points higher among special election voters than among all registered voters.

The party’s advantage was seven points if you exclude Florida, which makes up a big chunk of the available data. A majority of voters there have cast their ballot by mail absentee, and Republicans often have a big absentee ballot edge in the state — something that helps a lot in low-turnout special elections.

The Democratic turnout advantage is responsible for many of the party’s most impressive special election performances. In fact, if you want to predict special election results, you’re better off just knowing the Democratic turnout advantage than, say, how the district voted in the last few presidential elections.

The problem for Democrats is that it will be very hard to sustain this kind of advantage in a higher-turnout environment like the midterms. Even a five-point Democratic advantage would be without precedent in a modern general election. Any Democratic advantage at all would be rare.

Historical data of this kind is sparse, but in Iowa the Democrats have never enjoyed a turnout advantage over 30 years of comparable data. In the last 15 years of data in North Carolina general elections, it barely happened once.

The fact that something hasn’t happened before doesn’t mean it can’t happen in November. The growing affluence and educational attainment of white Democrats might give them a much better chance of earning a turnout advantage than in the past.

But so far, Democrats have not been able to sustain the same advantage in higher-turnout elections, such as in the general elections in New Jersey and Virginia, or in the special congressional election in Georgia’s 6th District.

In Virginia, Democrats still appeared to have a turnout advantage compared with all registered voters, but not a big one. The Republicans had an advantage in the Georgia 6th runoff election between Jon Ossoff and Karen Handel, the highest-profile race of all. The relatively unimpressive Democratic victories in Virginia and New Jersey are particularly noteworthy: The two elections are usually entirely omitted from analysis of recent races, even though more votes were cast in those two elections than in all of the special elections combined.

The Democrats could still fare extremely well even if their turnout advantage is diminished. Indeed, they did well in the special election in Pennsylvania’s 18th District and in the Alabama Senate race. Turnout was quite high for both sides in the two races.

There isn’t enough data to definitively resolve this question. But if, on average, Democrats are benefiting from unsustainable turnout, it seems reasonable to suppose that they might still fare a bit worse in the general election than in the broad set of special elections.

It has been common in the aftermath of Lamb’s victory for Democrats to note that there are nearly 120 congressional districts where President Donald Trump fared worse than he did in Pennsylvania’s 18th District. The implication seems to be that if Democrats can win there, they can win in the other 120 districts.

Maybe so. But buried in the result last week was a hint that it won’t be quite so easy: The Republican Rick Saccone ran six points better in his state legislative district than he did elsewhere, even after considering the area’s political and demographic characteristics.

In November, Democrats will generally face Republican incumbents who can claim a similar or perhaps even greater advantage, but across the entire district, not 9 percent of it.

At the very least, that’s what has happened historically. Even in wave elections.

Take the measure often used to discuss the special elections: the extent that Democrats are running ahead of their performance in recent presidential elections.

Historically, the party out of power often runs far ahead of recent presidential performance in open seats, just as Democrats have in 2017 and 2018’s special elections. But even in wave election years, like in 2006 or 2010, incumbents of the president’s party don’t run behind the president at all.

The hurdle for Democrats is straightforward: There just aren’t that many Republican-held congressional districts where the Democrats have fared well in recent presidential elections. If incumbency is worth what it has been, even a wave of 2006 or 2010 proportions might not be enough to give the Democrats a clear victory.

The fight for control of the House would be very close, although the Democrats would now probably be favored in such an election, because of all of the Republican retirements over the last few months. Of course, Democrats could still claim a more decisive victory in any number of ways. They could be riding a wave that’s even bigger than earlier versions.

Democrats could also claim a more decisive victory if the value of incumbency slips in today’s political environment. It has been slipping for decades, so it wouldn’t be much of a surprise if it did.

But we’re not going to find that out from any special election or any other kind of race until election night on Nov. 6.