States of Play: Ohio

After Trump maxed out the Buckeye State’s rural areas and small town areas, can Biden max out the suburbs?

KEY POINTS FROM THIS ARTICLE

— Ohio insiders believe that the state is closer than last time, and that Donald Trump is struggling mightily in suburban areas.

— Still, Ohio should vote considerably to the right of the nation, thanks to its high percentage of white voters who don’t have a four-year college degree — a strong group for Trump — and its smaller-than-average nonwhite population, a group that is very Democratic.

— Suburban areas in general, and the Cincinnati and Dayton areas in particular, would likely be a key part of a Biden path to victory. But Trump is still better-positioned to win the state.

Ohio may be back in play in 2020

With Joe Biden generally leading nationally in the range of 6-8 points, Donald Trump is playing defense in a number of states that he won relatively comfortably in 2016.

One of those is Ohio, the traditional bellwether state that took a sharp right turn four years ago.

No Republican has ever won the White House without Ohio, and the state has typically voted to the right of its brethren in the industrial North. For instance, Ohio voted for George W. Bush twice, in 2000 and 2004, even as Bush was twice losing Michigan, Pennsylvania, and Wisconsin. Barack Obama carried Ohio twice, but his margin was narrower than it was in many other places in the region. And while Donald Trump carried Ohio by eight points, he captured that trio of “Blue Wall” states each by less than a point apiece.

There’s not a ton of recent public polling data in Ohio, and the data we do have do not tell a consistent story. Rasmussen Reports, a pollster that has generally produced national numbers that are more favorable for Trump than other pollsters, had Biden leading by four points in Ohio right after the Republican National Convention. Meanwhile, Morning Consult and the Democratic pollster Civiqs had Biden down five and three points, respectively, in the Buckeye State earlier this month. Yesterday, Lauren Copeland of Baldwin Wallace University, which helps produce the Great Lakes Poll of several states in the region, released polling showing Ohio about tied.

Certainly one would expect Ohio to vote significantly to the right of the nation again in 2020. But the fact that Ohio is apparently not totally locked down for Trump is an indication of the immense amount of work the president needs to do across the Midwest in the final 40 days of the campaign.

Trump was in Ohio on Monday, visiting the Toledo and Dayton areas in the western half of the state. Meanwhile, Joe Biden will step foot in Ohio for next week’s presidential debate in Cleveland, but it’s unclear whether he will campaign in the state beyond that.

Let’s look back at what happened in Ohio in 2016, and then ahead to what we might see in November.

The best case scenario for Trump is that he wins the state comfortably again, with that margin indicative of enduring strength in other parts of the region.

The best case scenario for Biden is that his improved margins with white voters helps him cut Trump’s towering margins in rural and small town areas, and that the realignment toward Democrats in suburban areas and with white voters with a four-year degree that effectively skipped Ohio in 2016 hits the state in force this November.

The 2016 Ohio story

There are a few eye-popping data points that illustrate how the 2016 presidential race knocked Ohio from its decades-long perch as the state most consistently reflective of national presidential voting patterns and pushed it more firmly into the hands of Republicans.

Based on data compiled by Daily Kos Elections, four of the 10 congressional districts nationwide that shifted most strongly toward the Republicans from 2012 to 2016 were in Ohio. That included the district that moved further right than any of the nation’s other 434 districts: Appalachian OH-6, which runs along the state’s eastern border and shifted from giving Mitt Romney a 13-point victory in 2012 to giving Donald Trump a 43-point margin in 2016.

The others in the top 10 were OH-13, a usually very Democratic Akron-to-Youngstown district held by Rep. Tim Ryan (D), a fringe 2020 presidential candidate; OH-7, another northeast Ohio district that covers some rural areas, parts of exurban Cleveland, and part of Stark County (Canton), historically a swing county that cut hard to Trump in 2016 (and then stayed Republican in 2018); and OH-4, a sprawling district stretching from western to northern Ohio held by polarizing Rep. Jim Jordan (R).

These shifts at the congressional district level are borne out by isolating the state’s three big urban counties: Cuyahoga (Cleveland), Franklin (Columbus), and Hamilton (Cincinnati). Together, these counties cast a little less than 30% of the statewide vote, and Hillary Clinton effectively matched Barack Obama’s 2012 showing in these counties: Her collective margin in them was only about 5,600 votes worse than Obama’s. Meanwhile, in the state’s 85 other counties, Clinton’s margin lagged roughly 608,000 votes behind Obama’s. (Ohio has cast a little more than 5.5 million presidential votes in each of the last four elections.)

Trump won the state by eight points while losing nationally by two, meaning Ohio voted 10 points to the right of the nation. Ohio had not voted so differently than the nation since 1932, Franklin Roosevelt’s first election.

The American electorate, demographically, can be split into three groups: white without a four-year degree, white with a four-year degree, and nonwhite. According to the Center for American Progress’ post-2016 demographic analysis by Rob Griffin, Ruy Teixeira, and John Halpin — the Center for American Progress is left of center, but the numbers produced by this report are widely respected — the 2016 national electorate was roughly 45% non-college white, 29% college white, and 26% nonwhite.

Republicans win big margins among the white non-college group, Democrats win even bigger margins with nonwhites, and the white college group is competitive but trending Democratic: Hillary Clinton won white college voters by seven points nationally, according to the study, after Barack Obama had won them by less than a point.

Trump dominated in Ohio for three key reasons:

1. Ohio’s electorate was 57% non-college white — a dozen points larger than the nation as a whole — and that group shifted strongly toward Trump in Ohio in 2016.

2. Just about 16% of Ohio’s voters are nonwhite, a significantly smaller share than the nation.

3. White voters with a four-year degree make up a very similar proportion of Ohio’s electorate — 28% instead of 29% nationally — but the group did not become more Democratic from 2012 to 2016 (the group basically split down the middle in both elections).

That Clinton did no better than Obama in Ohio with white voters with a four-year degree may come as a surprise to observers (it did to this one), because Clinton did run better in some highly-educated suburban areas in the state despite her big statewide deficit. For instance, in highly-educated Upper Arlington in suburban Columbus — the electorate there likely has substantially more than double the share of college graduates as the state electorate as a whole — an eight-point Romney victory in 2012 turned into a 15-point Clinton win in 2016, according to numbers compiled by the Columbus Dispatch. But not all suburbs are created equal, and the Dispatch’s accounting of suburban areas around Akron, Cleveland, Columbus, Cincinnati, Dayton, Toledo, and Youngstown found that the aggregate presidential vote in these suburbs was split in both 2012 and 2016 (Trump won them by a point after Obama and Romney effectively tied).

With the suburbs as a wash, Trump’s dominance in small town, non-urban/suburban Ohio — as exemplified in the strong shifts in some of the congressional districts noted above — helped him win the state comfortably.

What’s going on in 2020?

Informed Ohio observers generally still see Trump as having an edge in Ohio, but most would be surprised if he matches his 2016 margin. Part of this is a widely-held belief in both parties, backed up by their district-level polling, that the numbers in the suburbs have moved this time after collectively staying stagnant from 2012 to 2016, and Trump has lost a significant amount of ground. State Sen. Peggy Lehner (R) recently told the Associated Press that Trump’s support has “crashed” in her suburban Dayton district.

The tantalizing possibility for Biden is that Ohio got the Republican end of the realignment in 2016, and now it’s getting the Democratic end in 2020.

How much ground Trump has lost in the suburbs is unclear, but there is some polling that helps shed a light.

Democrats have released several polls of Ohio’s 1st Congressional District, where Rep. Steve Chabot (R) is trying to hang on against Kate Schroder (D), a health care executive. Biden has consistently led in the Cincinnati-based district, albeit by small margins (the most recent Democratic survey had him up just one), after Trump carried the district by seven points. We have not seen any publicly-released data from Republicans in this district, although Congressional Leadership Fund — a major outside group connected to House GOP leadership — has booked money in the district, an acknowledgement that Chabot needs air cover.

Trump’s margin, at the very least, is likely to be smaller in OH-12, a suburban/exurban/rural district north and east of Columbus. This district, held by Rep. Troy Balderson (R), hosted a very competitive special election in 2018, and Biden could actually carry it if he has a good night, although the likeliest outcome appears to be a Trump win, albeit reduced from 2016’s 11-point margin.

So there are places where one can imagine Biden doing markedly better than Clinton. But where else? Elections from 2018 may provide something of a roadmap.

Overall, the Democratic showing in Ohio two Novembers ago was a disappointment. While Sen. Sherrod Brown (D-OH) won a seven-point reelection victory over an overmatched Republican opponent, then-Rep. Jim Renacci (R, OH-16), Republicans held the state’s five elected executive positions — governor and a number of other row offices like attorney general and secretary of state — by close but clear margins: roughly 3.5-6.5 points apiece.

Map 1 shows four recent top races in Ohio by congressional district: Barack Obama’s (D) three-point win in 2012; Donald Trump’s (R) eight-point win in 2016; Sherrod Brown’s (D) seven-point win in 2018; and now-Gov. Mike DeWine’s (R) four-point win the same year. Thanks to Crystal Ball Associate Editor J. Miles Coleman for producing the map.

Map 1: Recent Ohio elections by congressional district

Somewhat conveniently for illustrative purposes, the gap between Obama’s 2012 and Trump’s 2016 Ohio victories and that of Brown and DeWine’s respective 2018 wins is about 11 points apiece.

First off, the 2012 and 2016 maps illustrate the differences laid out above: Both Obama and Clinton only won four of the state’s 16 congressional districts — the state has a GOP gerrymander that has led to Republicans consistently winning the other 12 districts — but Clinton’s margin was a couple of points worse than Obama in OH-11 (a Cleveland-to-Akron district drawn to be majority Black) and OH-3 (a Columbus-based district designed to take in as many Democratic voters as possible); 15 points worse in the Toledo-to-Cleveland OH-9; and 21 points worse in the Akron-to-Youngstown OH-13. Those latter two districts cover a lot of white working-class territory where Democratic strength appears to be waning.

As noted above, a number of the less densely-populated and already-red Republican districts became way redder from 2012 to 2016.

Meanwhile, the previously-mentioned OH-1 (Cincinnati and suburbs/exurbs) and OH-12 (Columbus suburbs/rural) hardly changed at all from 2012, while some other districts — OH-14 and OH-16 in northeast Ohio, OH-15 running from suburban Columbus to Ohio University in southeast Ohio, and OH-10 covering Dayton and some more rural turf to its east — shifted right by about 5-10 points apiece, less than the state’s overall 11-point presidential swing toward the GOP.

The 2018 maps illustrate how Brown won relatively easily while DeWine won a closer race. DeWine carried all of the 12 usually Republican districts, while Brown actually carried nine of the state’s 16 districts: the four usually Democratic ones, plus five other districts discussed in the previous paragraph (OH-1 and OH-10 in southwest Ohio, OH-12 and OH-15 in central Ohio, and OH-14 in northeast Ohio). These districts cover a lot of suburban turf.

DeWine did worse than Trump in 14 of the state’s 16 congressional districts, but the two exceptions were OH-1 (Cincinnati) and OH-10 (Dayton), where Trump and DeWine won basically the same margins. Both DeWine and his lieutenant governor, Jon Husted, have strong ties to the Dayton area. Additionally, DeWine is quite possibly a better fit for some suburban voters in these districts than Trump might be this year. Brown, meanwhile, carried both districts.

The two districts cover some key swing areas of the state, and in order to win Ohio, Biden likely needs to replicate or at least come close to Brown’s showing in both of them.

That’s because the flip side of the coin is that Biden is very unlikely to do as well as Brown did in eastern Ohio. In fact, he probably would be lucky to do as well as Richard Cordray, the 2018 Democratic gubernatorial nominee.

Cordray, like Clinton, did poorly in eastern Ohio relative to previous Democratic performance — but not as poorly. Cordray lost the extremely Trumpy OH-6 by 32 points — an 11-point improvement on Clinton’s 2016 showing. Cordray also restored, to some degree, Democratic margins in OH-13, the Akron/Youngstown district. He carried it by 16 points, nine points better than Clinton.

Brown only lost OH-6 by 18 points, or 25 points better than Clinton. Brown carried OH-13 by 25 points, nearly reaching Obama’s 2012 figure. Expecting Biden to match either of these Brown performances in eastern Ohio is extremely unrealistic given what appears to be Trump’s unusually strong appeal in the region.

In fact, one of the big question marks for this November is whether Trump can actually do better in some of the places where he overperformed so dramatically in 2016. Could he, for instance, win by 50 points in OH-6, as opposed to 43?

Ohio observers seem to believe Trump is holding up well in eastern Ohio but also that he doesn’t seem likely to do better and could very well do a bit worse. But Appalachia in general seems to love Trump: West Virginia, the only state that is entirely classified as part of the Appalachian region, gives Trump his highest approval rating of any state, according to 50-state polling calculations by both Ipsos — as part of Ipsos’ relaunched Political Atlas site, to which the Crystal Ball contributes race ratings — and Civiqs. This probably translates in some ways to the Appalachian portions of eastern Ohio and western Pennsylvania.

Biden may have an easier time making inroads with non-Appalachian white voters, such as those in northwest Ohio, a Republican but swingy area (OH-5 will be an important district to watch there).

One rough shorthand for measuring the presidential race in Ohio might be to look at the state as two halves, divided by Interstate 71, which cuts from northeast to southwest (see Map 2).

Map 2: Interstate 71 in Ohio

It may be that about the best Biden could hope for in the eastern half of Ohio — which contains all of its Appalachian counties — would be to replicate Cordray’s numbers from 2018. And even that may not be feasible.

Meanwhile, on the western side, Biden’s goal might be to look more like Sherrod Brown, particularly in places like Greater Cincinnati and Dayton, where Cordray did no better than Clinton (at least in OH-1 and OH-10) but where Brown actually carried these typically Republican-leaning congressional districts.

This is an oversimplified way of looking at the state — and, if you want to be pedantic, portions of both of those districts are technically east of 71 — but it acknowledges Trump’s strength in eastern Ohio and points to more potential for Biden west of 71 than east of it.

Ultimately, the Crystal Ball still rates Ohio as Leans Republican. It may be that the gubernatorial race from 2018 provides a decent proxy for what we might expect in Ohio, albeit with different voting patterns — Trump running better than DeWine in the eastern part of the state and approximating his incredibly impressive 2016 showing, Biden running better than Cordray in the western, and Trump still emerging victorious, albeit by a margin roughly half that of 2016.

If Biden were to win Ohio, though, Trump’s path in the Midwest — and to a second term — would be blocked.

That Trump spent precious time in Ohio earlier this week suggests that the battle for the state is not yet finished. In 2016, the Trump campaign felt good enough about Ohio that Trump skipped the state in his final tour of swing states the three days before the election. Whether Trump can do so again the final weekend of this campaign might tell us something about how well his campaign believes he’s doing.

Kyle Kondik is a Political Analyst at the Center for Politics at the University of Virginia and the Managing Editor of Sabato's Crystal Ball.