Campaign Crawlers

Omens of Doom?

A statistical wizard gives 4-to-1 odds against a Romney win.

By 9.10.12

Send to Kindle

Fifty-seven days remain in the 2012 presidential campaign. Election Day is eight weeks from tomorrow, both party conventions are now in the rearview mirror, and Mitt Romney's uphill battle to unseat President Obama has reached its most crucial phase. Everything that happened before today was merely prelude to this, the heart of the fall campaign season, and no "expert" can confidently predict today what the final result will be on November 6.

These basic facts are important to establish at the outset of any discussion of the current state of the race, because there are many influential people who would like you to believe that the outcome of the election has somehow already been determined, and that they have clairvoyant insight on what that outcome will be. But why bring Nate Silver into this?

Silver is the poll-analyzing guru of the New York Times, whose reputation as a wizard was developed in crunching baseball statistics before being applied to political campaigns. On Saturday afternoon, Silver published an analysis which asserted that Obama now has a nearly 80 percent chance of winning the election, with 317 Electoral College votes and 52 percent of the popular vote. All of which is very interesting -- and very important, if true.

However, baseball isn't politics, and public-opinion polls are not batting averages or on-base percentages or any other such metric of past performance. Readers of Michael Lewis's bestseller Moneyball may appreciate this distinction, especially if they have any extensive experience in following polls and election campaigns. For myself, to cite just one example, I recall the Sunday in October 2010 when I arrived in New York's 25th Congressional District and was greeted by a Syracuse Post-Standard headline proclaiming that Democrat Rep. Dan Maffei had opened up a 12-point lead over Republican challenger Ann Marie Buerkle with barely two weeks remaining until Election Day. There was a mood of grim determination at the Buerkle campaign events I attended that Sunday and Monday, and I was far from certain that she could pull off an upset. On Election Night, the vote was "too close to call" and it was only after an extended recount that Buerkle was declared the winner -- two days before Thanksgiving -- by a margin of fewer than 600 votes. (See "The Republican Mandate," Nov. 24, 2010.)

Was Buerkle's victory a miraculous comeback? Had she actually erased a 12-point deficit in the span of 16 days? Of course not. The poll by Siena College was simply wrong, based on a faulty sample, and her chances of beating Maffei were as good on Sunday, Oct. 17, as they were on Tuesday, Nov. 2. Whatever caused the poll's misreading of voter sentiment, however, its publication as front-page news in the Syracuse paper posed a real danger to her campaign, one she addressed at a rally with her supporters. "I don't want anyone to let this discourage you," Buerkle said the day after the poll was published. "The voters will pick their congressman on Nov. 2.… It's up to the people to choose their congressman, not the Post-Standard and not Dan Maffei."

So it was in 2010 -- when despite all opposition and gloomy predictions of naysayers, Republicans won a victory of historic proportions -- and so it is now. It may well be the case, as Nate Silver would have us believe, that the odds are 4-to-1 against Romney winning. Defeating an incumbent president is a formidable task, one that has been accomplished only three times in the past 40 years: Jimmy Carter's narrow win over Gerald Ford in 1976; Ronald Reagan's victory over Carter in 1980; and Bill Clinton's defeat of George H.W. Bush in 1992. In that same span, incumbents have defeated challengers four times: Richard Nixon trounced George McGovern in 1972; Reagan stomped Walter Mondale in 1984; Clinton blitzed Bob Dole in 1996; and George W. Bush edged John Kerry in 2004.

The advantages of presidential incumbency -- the prestige of the "bully pulpit" and the power of executive authority -- are enormous, and the challenger always fights an uphill battle in attempting to unseat the Leader of the Free World. If we find the incumbent leading in polls two months before the election (and Obama is up by 5 points in the most recent Gallup tracking poll), this is not in the least surprising, coming as it does in the immediate aftermath of the Democratic National Convention. And it is worth noting that at no point in the past three months has Nate Silver's forecast shown Romney with more than a 41 percent chance of winning -- that was on June 2, when Obama, with a 59 percent chance according to Silver, held an 18-point advantage that has never been diminished. For most of the summer, the odds against Romney were at least 2-to-1 in Silver's calculus, so the shift toward Obama since Aug. 22 is merely the extension of a pre-existing trend.

What reason is there to believe that Silver's "Five Thirty Eight Forecast" has any predictive value? Or, better yet: What if its predictive value is entirely a function of its authority, as a self-fulfilling prophecy? If the poll expert at the most prestigious paper in the United States says that Romney's campaign is doomed beyond all hope of redemption, isn't it possible that this will have the effect of discouraging Republicans and influencing late-deciding "swing" voters to jump aboard the bandwagon of the Democrat candidate they believe to be a certain winner? And (here is where suspicions of liberal media bias arise) if the effect of such a prophecy is to assist Obama's campaign, is it not possible that the effect is the same as the intent?

This was certainly the suspicion on the part of Ann Marie Buerkle's supporters two years ago when the Syracuse paper announced in a front-page headline that she was trailing the Democrat incumbent by 12 points. There was never any apology from the newspaper for publishing the misleading poll, nor any explanation from the pollsters at Siena College as to how they had gotten it so absurdly wrong. When we look at Nate Silver's recent prediction that Obama will win precisely 316.9 Electoral College votes to Romney's 221.1, its explanatory authority must be regarded with profound skepticism, considering that the Real Clear Politics average of national polls Sunday evening showed Obama with less than a two-point lead, 47.8-46.0. Has there already been a decisive shift in the campaign, one that Silver's theoretical model has uniquely detected? There are reasons to doubt it, but no time now to explain those reasons beyond four words, "It's the economy, stupid."

To put it as bluntly as possible, the economy sucks, and the attempt by Democrats to exculpate Obama for this situation -- to place the blame on Republicans, or to say that the economy would suck even worse if Romney were elected -- is perhaps more difficult than Nate Silver's statistics suggest. If somebody were to offer you 4-to-1 odds on that proposition, how much would you bet? Mitt Romney's campaign reportedly raised $100 million last month, and the Obama campaign's embarrassed silence about its own August fundraising suggests that Democratic donors are less confident than the wizard of the New York Times.

Like this Article

Print this Article

Print Article
About the Author

Robert Stacy McCain is co-author (with Lynn Vincent) of Donkey Cons: Sex, Crime, and Corruption in the Democratic Party (Nelson Current). He blogs at The Other McCain.