My previous post on election-forecasting models—they’re showing an easy Bush victory in November because of the strengthening economy—generated a lot of reaction like this: "These models are crazy. Just look at the polls right now. It shows Bush and Kerry in a dead heat. Can eight or nine more months of good economic growth really make that kind of difference?" Sounds like a reasonable criticism, so I E-mailed Yale Prof. Ray Fair and asked him to defend his forecasting model. His response:

"1) Polls are not very reliable this far in advance. I think in July of 1988 they had Dukakis beating Bush by 17 points!

"2) The statistical results strongly suggest that the growth of the economy in the year of the election is important

"3) It may be that the equation is no longer any good for future predictions. This is discussed in my book Predicting Presidential Elections and Other Things.

# posted by James M. Pethokoukis at 4:00 PM EST
Return to Next News
Go to the archive