special elections: forecast or fluke?

by Joseph Bandera-Duplantier

The recent special elections have been nothing if not sensational. From the GOP's Star Wars-themed attack ads in Georgia to its real attacks on journalists in Montana, these races have provided an anomalous level of drama. But are their results equally anomalous — or are the double-digit swings toward Democrats a bellwether for 2017 and 2018?

Here we summarize five recent articles that analyze the current electoral climate and offer our own perspective.

Sifting through a bizarre race for 2018 clues. Kyle Kondik of Sabato’s Crystal Ball writes that last Thursday’s results in Montana provide both parties with reasons for optimism. Overall, though, the national environment bodes better for Democrats. On average, Democrats in special elections this year have outperformed Clinton’s margin in 2016 by 11 points.

How to tell whether Democrats are “on track” for a House majority. Using a shorthand method to evaluate this year’s special elections, David Wasserman of the Cook Political Report predicts that Democrats may currently be on track to retake the House in 2018. But there are two important caveats to this approach. First, special elections are, by definition, open elections, without any built-in incumbency advantage (incumbents typically have upwards of an 80% chance of winning re-election). Second, much can change between now and November 2018.

Trump’s base is shrinking. Nate Silver writes for FiveThirtyEight that an enthusiasm gap between Democratic and Republican voters may be emerging: “Trump has already lost almost a third of his strong support,” likely caused by the GOP’s first attempt to pass the unpopular AHCA. If this trend continues, Republican prospects in 2018 will likely take a beating.

Who’s turning out for these special elections? Writing for The Upshot last month after the primary in Georgia’s 6th congressional district, Nate Cohn points out that Democrats outperformed midterm expectations by about 10 percent. This was in part due to Jon Ossoff winning over a significant number of highly educated, previously Republican voters. Patrick Ruffini of FiveThirtyEight writes that, on the other hand, Democratic turnout among black voters is also on a noticeable decline. Democrats haven’t been able to rely on the Obama coalition, and it’s clear Democrats need to do more to reach out to black voters beyond fear of the Trump administration.


our take

Last week’s results leave us even more optimistic about state races. At flippable, we’re more convinced than ever that our approach — using data to identify the most flippable state districts, observing emerging trend lines among the electorate, and using effective storytelling to amplify our message to voters — will yield dividends in 2017 and beyond.

First, results up and down the ballot show us the potential for significant gains at the state legislative level. Democrats won five more counties in the recent special election than we won in the 2016 presidential race. In a winner-take-all federal or statewide election, that result doesn’t translate into any seats. At the state legislative level, however, that result might yield five to ten seats in Montana’s House. When we write about how margins matter, it’s most applicable to picking up seats in state legislative elections.

Second, it’s clear that money matters. Consider the fact that Democrats lost the Montana race by 6 points — roughly the same margin of loss in KS-4. Ultimately, Republicans can blunt the impact of a poor electoral environment — with both Trump and generic Congressional approval ratings steadily declining — by pouring money into these elections. If the GOP is continually forced to spend on safe districts, Democrats will be able to target and pick off key districts.

There is actually empirical evidence for this Republican dilemma of resource allocation in last week’s state legislative elections. While the GOP was distracted by a close national race in Montana, spending millions of dollars on a formerly safe Republican district, Democrats quietly picked up their first two flips — and major upsets — of the year in New York and New Hampshire. This is especially interesting because Trump significantly outperformed Romney’s 2012 results in both states by about five points. The Northeast, in contrast to the sun belt and Southwest, seemed more open to Trump’s messaging and appeared to be trending red, but that momentum may have slowed.

Finally, it’s important to monitor the shifting electoral trend lines, rather than single proxies. It is not enough to rely on Clinton’s margins in 2016 or Obama’s in 2012 to identify vulnerable districts. We’re interested in determining how the resistance translates into actual votes on the ground — and that means observing how demographics, education, and population density, among other factors, influence these early elections, and forecasting their impact on future races. We’ll continue to keep you updated on our findings and invite you to join us as we apply what we learn to our targeting in Virginia and Washington in 2017.