Ray C. Fair, a yale professor, has revised his election prediciton model to take into account first quarter 04 economic numbers. His methodology, which, when applied to past elections, accurately predicts the two-party vote share within no more than about 3%. Of course, he's tweaked the model to make it work for elections going back to 1916, and provides this caveat:
it may in fact be a poor approximation and only look good because I have tried thousands of versions to come up with one that explains well. It may also be that there will be a sea change in voting behavior beginning with the 2004 election. In this case even though the equation may have been good in explaining voting behavior through 2000, it may no longer be any good. It is never easy predicting human behavior. These caveats aside, the bottom line here is that if there is anything to the vote equation, President Bush will be hard to beat in 2004.As of now, he is predicting that Bush will win 58.74% of the two party vote. If I slap that number into the Composite Poll matrix, my computer belches smoke before predicting a Bush electoral landslide 376-162 (and that doesn't even take into account the discouraged west coast democrats who want show up when Bush sweeps the midwest). If that happens, I'll be getting drunk quickly as they tell me Bush wins Michigan and Pennsylvania, and that Illinois is "too close to call". I'll be passed out when a gloomy Dan Rather (and enthused Brit Hume) say California is too close to call.
Heck, even if he's off by 3%, that gives Bush a 360-178 win. I don't see that happening unless Kerry self-destructs, which is possible, though unlikely. I bet he's going to have to change the model after this election. And he lists his potential mistakes in his Note to the Media.
This guy's a social scientist who's done a lot of hard work on this model. I don't mean to disparage his numbers (and I sure hope his analysis is not "misspecified"), but a bunch of amateurs' analyses like the Composite Poll, Election Projection and Dales' Electoral College Breakdown seem more reliable at this point.
