One of the most difficult things to do is to predict Ohio politics. The state does not register by party until a person votes in an election. If you did not vote in the previous election, guess what? You get dumped into the “U” bin (“unaffiliated”) until you vote again.
Anyway, the numbers produced by the Ohio Secretary of State office are exceedingly inaccurate, at least according to my source, “The Accountant.” In 2016, the Accountant and my other Ohio source “OhioWan” were dead on, saying Trump would win big.
Actually, hold that thought: they weren’t dead on. They were low. Real low. In late September 2016 they told me Trump would win Ohio by “four or five.” I transmitted that info to Team Trump and urged the Trump team to get out of Ohio and spend time in Pennsylvania and Michigan. They listened. Meanwhile, as you know, Trump won Ohio by nine.
Why was the Accountant, in particular, so clued in when everyone else thought Ohio was, at least, competitive? It gets back to the old “Unaffiliated” voters in Ohio. One might think these are “independents,” and thus (as we thought wrongly in 2012) if a candidate like Mitt Romney is leading with independents in Ohio, it will be hard to lose.
Wrong. In 2012 those “unaffiliated” voters were in large part Obama Democrats who didn’t vote in the primary, but who showed up on election day, handing Romney a loss.
The Accountant vowed to address that issue in the next election, and he did so through painstaking historical research into how each household in a single county—the swing county of Montgomery—voted going back to 2000. He established a ranking system where he could thereby much more accurately identify an “unaffiliated” as a “Lazy D,” a “Lazy R,” or truly an independent.
The result? The Accountant was able to allocate the large majority of “U” voters to either the Democrats or the Republicans, and therefore when the early ballots came in, could with extremely high accuracy say “this is a Trump Republican” or “this is a Hillary Democrat.” There were, of course, people in each party who crossed to vote for the other candidate, and there were, of course, some remaining true “Us” whose vote couldn’t be predicted. But . . .
Using this tool, the Accountant constantly “re-allocated” all the early votes based on voting history, and by early October was convinced Trump was doing exceptionally well in the county. Neither he nor OhioWan wanted to predict Trump would win the county (which he did), but even being close in Montgomery historically meant a big GOP victory.
Fast forward to 2018. Recall I said the Secretary of State registration numbers were not accurate?
Here is the “official” status as us August 2018:
In most states, your reaction would be “Wow, the Republicans have a huge lead (603,000). You would likely be tempted to say, “Well, Trump in polling is getting 51% of the indie/unaffiliated vote so Rs will win by more than 603,000.” And you’d be wrong.
How about this? You know I have been tracking voter registration changes in battleground states and that with exceptions of CO, OR, and NJ, virtually all of the battleground states have seen GOP net gains, some of them very large.
Here are the net Ohio registrations since November 2016:
Based on this, you’d think the Ohio Democrats have registered 38,000 more voters since 2016 than Ohio Republicans. And you’d be wrong. Ohio registers people to the same party each time they change counties!!
We don’t know how much, but a large percentage of those are not new registrations at all, but are simply people who changed counties. Especially in Cuyahoga, Democrat voters who live in apartments can change residence fairly frequently. Therefore, unlike, say, Arizona or Florida, voter registration changes aren’t as meaningful in Ohio.
Here is another tidbit to file away before I hit you with the “BOOM”: looking at the history of unaffiliated voters actually voting, one finds that in fact they don’t vote very reliably . . . at all. Democrats in Ohio actually voted at a slightly higher rate than Republicans (76% to 73%) but “unaffiliateds” only voted at a rate of 42%. But this could mean that more Democrats vote in primaries, but not the general. Since 2000, 41 million Republicans have cast a ballot in an Ohio election, vs. only 32 million Democrats and only 5 million “unaffiliated” voters.
At any rate, here is the good stuff: the Accountant used his historical tables to more accurately assess the Secretary of State’s numbers as he did with Montgomery County in 2016. Here is are the more accurate estimates:
You’ll notice that now the Republican advantage is down by half, that the GOP lead is only about 268,000 now. Gee, combining that with the “new registrations,” doesn’t that mean Ohio Republicans are in trouble?
No. Just the opposite. The key lays in the reduction by half of the “unaffiliated” voters. By assigning them where they really should go, you have a much more predictable and likely outcome of a Republican victory of over 250,000 (allowing for an even split of the remaining “Us”), vs. a perceived GOP edge of 600,000, which is in fact eaten up by a large number of “Unaffiliated” voters showing up to vote Democrat.
The Accountant used a model of 2018 primary and 2016 general election turnout to produce two estimates for 2018, a “low turnout” estimate and a “high turnout” estimate, in which he factored in the very high turnout for the 2010 midterm.
His estimates for 2018 Ohio are as follows:
Low Turnout Model
It is obvious that when you account for all the “Lazy Rs” or “Lazy Ds,” the Republicans emerge with a 427,000 advantage.
High Turnout Model
In this model, the Republicans have a larger advantage (463,000) and would still win by 231,000 votes if every single unaffiliated voter turned out to be a Democrat.
In short, despite the surface perception that a more accurate assessment of the registration numbers favors the Democrats, the fact is by more accurately accounting for unaffiliated voters, this refined look at Ohio voter registration suggests that barring a miracle, Republicans should beat the Democrats by anywhere from 420,000 to over 463,000 in 2018.
Hard to win elections, even in stacked districts, with those numbers.
Larry Schweikart is the co-author of How Trump Won with Joel Pollak and A Patriot’s History of the United States with Michael Allen and author of Reagan: the American President, due out next summer.
Join the conversation!
We have no tolerance for comments containing violence, racism, profanity, vulgarity, doxing, or discourteous behavior. Thank you for partnering with us to maintain fruitful conversation.