Conservatives have spent much of the last few weeks expressing skepticism about polls that showed President Obama winning the election. Democrats claimed this was merely a case of premature sour grapes. But that disbelief, which I shared, was rooted in a reasonable argument. Most polls showing Obama ahead had samples that showed an electorate that seemed to match the 2008 turnout model, in which the Democrats had a large advantage in partisan identification. It seemed highly unlikely that the Democrats could maintain that lead after four dispiriting years of the Obama administration. Surely, they reasoned, the partisan split would be a lot more even in 2012, and polls with more balanced samples showed Romney ahead.
Yet the exit polls currently being discussed on the networks’ elections coverage are showing a turnout model remarkably similar to 2008. That makes the polls look smart and those that staked their reputations on them — like the New York Times’s Nate Silver — even smarter.
But the only problem with this is that exit polls are notoriously inaccurate. They also tend to favor Democrats. If you don’t believe that truism, just ask President John Kerry, who believed the exit polls that said he won the 2004 election before President George W. Bush was re-elected.
But the question is how inaccurate the exit polls are this year. Even if they are off by a bit, if they are somewhat close to the actual totals, that is very bad news for Mitt Romney. While the night is still young, the last ditch of Republican optimism is based on disbelief in the exit polls. Like the skepticism about the pre-election polls, this is not an unreasonable position. But the exits are going to have to be off by a lot in order for Romney to be the next president.