Polls show a dead heat. So why so many predictions of an Obama win?
Among pundits and prognosticators – as well as the public at large – there's an expectation that President Obama will win reelection, despite the fact that the race is still a virtual tie, nationally.
Washington — Is it just us, or does it seem like some of the suspense has leaked out of this campaign? In the final hours before Election Day, the mindset among the chattering class seems to have shifted from: "This thing is too close to call, it’s right down to the wire, a real nailbiter," to something more along the lines of: "It's close, but looks like President Obama’s got this."
Or as media-watcher Howie Kurtz put it in The Daily Beast: “The pundits have spoken: It’s Obama.”
Sure, plenty of caveats are still being thrown around: Mitt Romney could win, the polls are tight, yadda yadda yadda. But everywhere you look, the predictions are piling up in Mr. Obama’s favor.
In The Washington Post’s “Crystal Ball” contest on Sunday, only two participants out of 13 predicted Mr. Romney would win on Tuesday. For the record, that was the exact same number that predicted John McCain would win in 2008 – an election that was clearly heading for a more lopsided outcome than this one.
Now, some of the “predictions” being offered out there are clearly colored by partisanship. As in every election, there are strident voices on both sides of the aisle forecasting big wins for their guy (Fox News’s Dick Morris, for example, is predicting Romney will win with 325 electoral votes).
But most of the nonpartisan prognosticators seem to be calling it for Obama.
The University of Virginia’s Larry Sabato, a longtime professor of political science who produces a “Crystal Ball” newsletter of his own, is predicting that Obama will win reelection with 290 electoral votes. Veteran political prognosticator Charlie Cook has been more cautious about making outright predictions – but in his most recent column for National Journal, titled “Advantage, Obama,” he went through the electoral math in the battleground states and concluded: “Romney would need to win 64 of the 79 remaining electoral votes to win. Is that possible? Sure. But is it likely? It looks pretty tough.”
And of course, The New York Times’s political statistician Nate Silver – who has become something of a lightning rod among conservatives this cycle for his consistent predictions of an Obama victory – is now giving Obama an 86.3 percent chance of winning.
Perhaps most telling, this expectation of an Obama win is not limited to pundits. A newly released UPI poll, which found Obama and Romney locked in a dead heat at 49-48, also found voters predicting an Obama victory by double digits. In the 11 battleground states most likely to determine the election, voters picked Obama as the likely winner by 50 to 39.
So what’s driving this widespread perception?
One obvious answer is that the polls have shifted ever so slightly in Obama’s favor in the final days. It’s hardly overwhelming, but since political coverage and electoral predictions revolve to a large extent around the latest polls, the “echo chamber” effect may make any kind of shift, no matter how small, seem even more significant. Recent events have also seemed to work to Obama’s advantage – including last Friday’s better-than-expected jobs report, and his response to hurricane Sandy.
In addition, the growing prevalence of early voting has provided analysts with a more concrete metric – allowing prognosticators to base their assumptions not only on what polls suggest will happen on Election Day, but also on what early voting patterns suggest has already happened. They’re still basically guessing, of course, since the votes won’t actually be counted until Tuesday. But there are hard totals, and in many cases partisan affiliations, to factor in. All that has led many observers to conclude with more confidence that the president has a real edge.
Of course, it’s entirely possible that this expectation of an Obama victory could be totally wrong. As we’ve written before, this election seems in many ways the mirror image of 2004 – when, in the final days before the election, most polls showed George W. Bush and John Kerry in a virtual tie, but with President Bush holding a very slim lead overall. In the end, President Bush wound up barely edging Senator Kerry, taking Ohio by a mere 136,000 votes (which gave him 286 electoral votes to Kerry’s 251).
In other words, 2004 was an election that basically could have gone either way (though, of course Bush did win). And notably, heading into it, prognosticators were far more likely to hedge their bets: The Post’s Crystal Ball was almost evenly split, for example, and Professor Sabato actually forecast a 269-269 Electoral College tie. This time around, the chattering class seems significantly more confident in predicting an Obama victory, despite polls that are in many ways strikingly similar to those of 2004.
We'll soon see if that prevailing view is correct – or if a lot of pundits wind up with egg on their face.