Its very early still in the battle for the GOP presidential nomination, but to judge by the numbers, the real race for the top spot on the Republican ticket has already tightened considerably. As of the morning on Sept. 9, Rick Perry is the most probable Republican nominee for president, scoring a 38.8 percent likelihood on our predictions model. Mitt Romney (35.3 percent) is a close second. These likelihoods are derived mainly as an average of the prices in the major prediction markets; I've explained the process in more detail here.
Once the focus moves off the two main front-runners, the probability numbers drop off steeply--to the tune of 30 points. Jon Huntsman (at 5.8 percent) places third, and Sarah Palin is fourth, at 4.7 percent--even though Palin is not yet declared as a candidate.
Despite the big gap separating Romney and Perry from their nearest competitors, it's still early enough in the race--without any candidate yet tested by an actual primary vote--so that the third and fourth candidates still have a meaningful possibility of winning the nomination. One immediate example of how volatile the contest still is actually comes from the two top-tier contenders: Shortly after he announced for the presidency, Perry bolted out to a huge lead over Romney--but after Perry's first debate performance this week, the two candidates became much more closely matched.
Chart of Republican Nomination
Events such as Wednesday's debate fuel a great deal of speculation over the perceived "electability" of GOP contenders--that is to say, how likely a given candidate would be to prevail over President Barack Obama in a general-election vote.
Prediction markets such as Betfair and Intrade provide data to handicap the likely winner of both the GOP nomination and the presidency. If we can effectively gauge the likelihood that a given GOP candidate will win the general election in comparison to his or her odds of winning the Republican nomination next summer, we can start to get a reliable measure of the otherwise elusive quality of electability. So to compute, say, Romney's electability, we need to divide Romney's likelihood of winning the presidential election (which stands currently at 16.3 percent) by his present likelihood of earning the Republican nomination (35.3 percent). The result--46.2 percent--is Romney's likelihood of winning in November 2012 if he is nominated. In other words, this ratio reflects the markets' opinion of Romney's electability in the general election.
The chart below shows the electability of Romney, and his chief rival Perry:
Chart of Republican Electability
In the balmier days of early August, all Republican candidates appeared more electable, as Obama continued to be battered by bad economic news. Yet now that the field is separating out a bit, we can see several important differences between the top two major Republican candidates.
Perry steadily gained an aura of electability as he moved toward the announcement of his presidential bid. His standing in the prediction markets has gone up and down since his announcement--a volatile measure for any candidate on a day to day basis--but the overall trajectory of his prospects since his official announcement has been downward. This is especially true after Wednesday's debate, when his electability numbers plunged.
Meanwhile, Romney is in the middle of the GOP pack when it comes to electability. Until Wednesday, he had been comparatively quiet on the national stage, even as he continued to campaign vigorously in early-primary states. Not surprisingly, this trend has coincided with a dip in his relative electability compared with the other major Republican candidates. After Romney sparred with with Perry at several key points during this past Wednesday's debate, he's seen a sharp increase in his perceived electability.
The chart above focuses on Perry and Romney in order to avoid clutter--but the last week of the campaign has shown some significant movement among candidates beneath the top tier.
Palin, for example, is in the middle of the Republican pack. Using our measure of electability going back to January, Palin was way down. Over the past few months, however, Palin has been relatively silent--while picking up some ground in the electability battle. This suggests a somewhat paradoxical trend for her possible candidacy: She appears to have gained a measure of credibility among the prediction markets by virtue of staying out of the national spotlight--a pointed contrast to the experience of Romney, and indeed to that of any other candidate in the field.
And Huntsman--the third-place candidate here--has consistently performed best in battling Obama; he is actually been above 50 percent for most of the last three weeks. This quite clearly reflects his own deliberate strategy; as he recently told ABC News, "I think when you find yourself at an extreme end of the Republican Party you make yourself unelectable." In short, Huntsman is staking his campaign on being the "electable" Republican, with a far more moderate profile than that of the other candidates. And that message appears to be getting through, with a growing segment of the prediction markets regarding him as more competitive with Obama than any other Republican. He has seen little forward movement in his likelihood of winning the nomination--but it seems plausible, based on the way the numbers have been trending, that Huntsman could emerge as the clear third choice in the next few days or weeks.
Finally, the collective wisdom of the markets now suggests that Michele Bachmann--who was in front of the GOP field after her victory in last month's Ames straw poll--is much less electable than any of the other major candidates.
All of this forecasting still comes with a big and obvious disclaimer, however. The calculation of a candidate's electability gauges the likelihood he or she wins the general election after emerging victorious from a grueling and uncertain nomination process; it does not rate their chances of winning if they were somehow simply handed the nomination. In other words, numbers don't account for the many ways in which a campaign can transform a candidate; there is a reason, after all, that Bill Clinton--one of the most accomplished presidential campaigners in recent memory--was known as the Comeback Kid.
David Rothschild is an economist at Yahoo! Research. He has a PhD in applied economics from the Wharton School of Business at the University of Pennsylvania. His dissertation discussed the creation of aggregated forecasts from individual-level information. He can be found on Twitter at @DavMicRot.
