Romney had the institutional support of the Christian right in this election; remember
Billy Graham scrubbing "Mormonism is a cult" from his website? I literally cannot find a significant Christian conservative voice headed into the election who was arguing that voters should stay home rather than vote for Romney (and obviously, none of these people were endorsing Obama). That's not to say that, if you dig deep enough in the bowels of right-wing-nuttery, you won't find the occasional oddball in the Deep South who refused to vote for Romney; I'm sure those people exist. But from a
systematic perspective, Romney's Mormonism wasn't a factor at all.
There's also no evidence that Romney's Mormonism was a drag on downballot races; the closest Senate race, for example, was Heidi Heitkamp in North Dakota, and she obviously ran well
ahead of Obama in her state. Ditto Jon Tester in Montana.
So the answer to the question is: absolutely nothing. That fits, by the way, with the evidence producing by polling aggregator sites such as the
Princeton Election Consortium and
538, which showed a remarkably stable race for all of 2012. In fact, back in August,
Michael Tomasky noticed that the 538 histogram was calling 332 EV -- the exact result of the election -- as the single most likely distribution of electoral votes. Here's what he said:
Most outcomes, in a range running from 150 EV’s up to 400, rate around a 2 percent chance of Obama receiving that number. The highest spike on the chart? It’s at around 330 EV’s, which [Nate] Silver [of fivethirtyeight.com] reckons Obama has a 14 percent chance of hitting. Now, most political journalists would chuckle derisively at the idea that Obama is going to carry home 330 EV’s. Deride away. And while you do, bear in mind that Silver called 50 out of 51 states last time (counting D.C.; he missed only Indiana) and every single Senate race.
...
An extremely close election that on election night itself stands a surprisingly good chance of being not that close at all.
That's the first week of
August, before either party's convention, before the bulk of the ads ran, before -- well, most of the campaign, really. And 538 had been running that histogram since the
primaries; Tomasky just
noticed in August.
The numbers suggest that -- notwithstanding the second-guessing and finger-pointing that inevitably surrounds a losing candidate -- Mitt Romney was a pretty good candidate who ran about as good a campaign as could have been run in 2012. He certainly produced the campaign's most dynamic moment, statistically speaking, after the first debate.
Had Romney drastically underperformed the statistical fundamentals; if, say, Obama had won not only North Carolina, but the Omaha congressional district as well, and had states like Arizona, Georgia, and Montana gone down to the wire --
then it might be worth asking 'what went wrong?' and wondering if something like anti-Mormon bias was in effect. But the data suggest that simply wasn't the case.