Andreas Graefe and Scott Armstrong developed a model for forecasting U.S. Presidential Elections. It provides fast advice on which issues candidates should focus on in their campaign by using information about how voters perceive the candidates’ ability to handle the single most important issue facing the country. It predicted the winner of the past ten elections with an accuracy of 97% (based on an examination of the forecasts on each of the last 100 days prior to each of the last ten U.S. Presidential elections). A working paper version of the paper, which was accepted for publication in the Journal of Behavioral Decision Making, is available here.

As of January 2010, Tim Pawlenty, Governor of Minnesota, ranks second at the intrade.com prediction market to be the Republican nominee for the 2012 U.S. Presidential Election. The market sees his chances to win the nomination at about 18%. 

According to the preliminary forecast of the biographical model, Pawlenty would not be a good choice for the Republicans. While Obama would achieve a score of 22, Pawlenty's score is 18. This translates to a predicted two-party share of the popular vote of 53.6% for Obama, or 46.4% for Pawlenty. 

As of January 2010, Mitt Romney is favored by the intrade.com prediction market to be the Republican nominee for the 2012 U.S. Presidential Election. The market sees his chances to win the nomination at about 25%, followed by Tim Pawlenty (20%), Sarah Palin (18%), and John Thune (13%).

According to the biographical model, the chances for Romney - for whom running in 2012 is "always a possibility" - to defeat Obama are not great. While Obama would achieve a score of 23, Romney's score is 21. This translates to a predicted two-party share of the popular vote of 51.7% for Obama, or 48.3% for Romney. Although Romney scored higher than Palin and Huckabee, he would need to achieve higher ratings than Obama on the four variables for which no data are yet available (i.e., facial competence, intelligence, attractiveness, and voice) in order to be predicted as the winner.

Here are the new scores after the model update for Barack Obama and Sarah Palin. While Obama achieves a score of 25, Palin’s score is 16. This translates to a predicted two-party vote share of 57.7% for Obama, with a 95% prediction interval of +/- 6.7%. If one removes predictor variables that might favor males in inter-gender races (such as the candidates’ height and weight), Obama would still be predicted to receive 56.3% of the vote. Even if Palin would be perceived as more competent, intelligent, and attractive than Obama – three cues for which no data are yet available – the model gives her little chance of defeating the incumbent president.