The 2018 Election Heads to Kasich-land

The policy stakes here aren’t high—the winner will finish up Tiberi’s term, which is only a few more months—but it’s a potentially important political data point. House special election results (when compared to an appropriate baseline and aggregated) give us a read of the national political climate. This race will also give us more information on which districts will be vulnerable in November and what sort of candidates can win in the Trump era.

What you need to know about the district

Ohio’s 12^th District is north of Columbus. It contains a few slivers of Franklin County (home to Columbus), some suburban areas (like Delaware County) as well as some comparatively rural, further northern areas. This map (created by DDHQ’s excellent cartographer, J. Miles Coleman) provides a good sense of where the district sits and how it voted.

In 2016, the district voted for President Trump by 11.3 points, and in 2012 it voted for Mitt Romney by 10.5 points. It has some underlying Republican DNA (Delaware County, which cast about a quarter of the district’s votes in the 2016 presidential election, has voted for every Republican presidential candidate since Woodrow Wilson left the ballot), but there are also some populous, relatively Democratic areas (those blue slivers of Franklin County made up about a third of the districtwide presidential vote in 2016).

The district is whiter, more educated, and more affluent than the median congressional district. But it’s not that different from other districts that have held a special during the Trump era.

These scatter plots compare the districts where special elections have been held (the red points—every special except CA-39 where no Republican made the runoff) to those where they haven’t (shown in blue). Data is from American Community Survey 5-year estimates, and the lines for 115^th Congress are used.

The first plot shows that Ohio-12 is less Republican than other districts that have held specials (like Pennsylvania’s 18^th or Arizona’s 8^th), but the rate of college education among whites is comparable. And the second plot shows that it’s whiter and richer than most other districts (measured in terms of household income) that have held specials, but not by an enormous amount.

Ohio’s 12^th District, like most of the other districts that held specials, didn’t move much between 2012 and 2016—Trump improved on Romney’s win margin by less than a percentage point. But there is some evidence that this district contains potential swing voters

This table compares Romney’s 2012 performance to Trump’s 2016 performance using data from Daily Kos Elections.

Clinton improved on Obama’s margin in Delaware County as well as the bits of Franklin County that are in the district. That makes sense—you’d expect to find some well-educated, suburban, Romney-to-Clinton type voters there. But Trump improved on Romney’s showing basically everywhere else, signaling that there may be some Obama-to-Trump voters in the more rural, less Columbus-ish parts of the district.

It’s hard to use geographic data like this to firmly conclude much (beware the ecological fallacy). But it’s not crazy to look at this data and think that there are persuadable voters in this district. There may be some Trump voters who, given a more normal Republican candidate and a more throwback Democrat, would go blue. And there seem to be some historically Republican, Clinton-voting suburbanites who might come back into the fold for the right candidate.

This is a real race

The demographic and political data suggests that this race could be competitive. If Democrats managed to re-create some of their best performances, they would win this district by a solid margin (e.g. Democrat Conor Lamb won Pennsylvania’s 18^th District after Trump won it by 20 points; Trump won Ohio’s 12^th only by 11 points). And if Republicans re-create some of their better special election performances, they’ll underperform the district’s overall partisan lean by a couple points and hold the seat.

At this point, the race leans slightly toward the more Republican scenario. But our data isn’t great, and it’s easy to imagine how things might change.

There are only two reliable public polls of the race and both are a bit stale. Monmouth University, a historically accurate pollster, put Balderson (the Republican) ahead of O’Connor by 9 points in mid-June. JMC Analytics basically agreed, showing O’Conner behind by 11 points around the same time. So Balderson was likely ahead about a month ago, but we don’t know if the race has changed since then.

Major handicappers also think that the race leans right. RealClearPolitics and Sabato’s Crystal Ball say the race “Leans Republican” while Inside Elections says it only “Tilts Republican” and Cook Political Report lists it as a toss-up.

In some ways, these ratings make sense. Neither candidate seems truly great or extremely bad – Balderson appears to be a pretty standard Republican, and O’Connor is trying to signal independence by saying he’d vote against Nancy Pelosi as House Speaker. And unlike most other House special elections, this race is competing for attention with both primaries and an upcoming general election. So Republicans might be able to cruise through this race while enthusiastic Democrats focus their attention elsewhere.

But this race is nowhere close to over. Ratings like “Leans” or “Tilts” (or in Cook’s case, Toss-up) are far from certain. More importantly, Democrats have outperformed their baseline partisan strength in every House special since Trump took office. The data is noisy, but if Democrats managed to post an average-level overperformance here (meaning they don’t have to recreate Pennsylvania 18), they’d win the election.