I have some tips for keeping track of who is most likely to win a presidential election. You all know about the prediction markets, here are a few other mental categories which I find useful. There is real evidence for them, but I can’t pretend they all command a consensus in the political science literature. I do, however, think they are true, in part because they are consistent with my underlying views of human behavior:
1. People tend to overreact to the news of the moment in predicting a winner, don’t make this mistake. Ultimately election outcomes are determined by the fundamentals of the comparison. For instance if you wish to argue that Hillary Clinton will still be the Democratic nominee, just ponder all those Latinos and blue collar workers out there. They’re not responding to most of the cues analyzed by the net roots bloggers. For any forecast you make, imagine yourself telling it to the guy sitting next to you at the West Virginia K-Mart, and see if it passes his laugh test.
2. Party disunity predicts an electoral loss; if you are a Democrat you should worry about this. It remains to be seen how deep the Republican squabbles will run. Read the work on Martin P. Wattenberg on this question, of course party disunity can either be a cause of loss or a symptom of other problems. The state of the party (just like market prices) also aggregates information.
3. The swing voters in the American citizenry don’t really trust the Democrats with foreign policy and won’t anytime soon, whether this is rational or not. Signs that the election will center around the economy help the Democrats. Signs that the economy will focus on foreign policy help the Republicans.
4. When a woman or an African-American or a former first Lady is running for President, that is a huge issue in the minds of voters, whether anyone admits it or not.
There is reasonable though not decisive academic evidence for points #2 and #3. #4 is of course a wildcard, and #1 I have never seen tested; a study might find mean-reversion in the betting markets, I do not know. Based on this list, I am still thinking that most people are underestimating the chances of a Republican President (the ascendancy of John McCain is starting to reverse this tendency), noting that #2 and #4 are working for my view, but #3 is working against it, at least at the moment.