The 2004 Election Prediction Map Contest

[muon2]

I could spend some time debating which poll to go on, but the forum seems to agree that the polls went haywire this weekend. So let's go from the gut.

I posted version 2 on Mar 3 after Kerry was the presumptive nominee. On Aug 21, I posted version 3, changing OR and NM back to Kerry and ME CD2 to Bush. Polls might cause me to switch NM and ME2 back to the Mar 3 prediction, but I've already commented on our collective opinion of those polls. Also, I've missed NM in almost every election I've predicted going back to 1976, so why stop now.

So, I shall blindly ignore the Swift Boat vets, the RNC, the debates, and everyone's final polls. My final call is from Aug 21, 2004. For President: Bush 292, Kerry 246, if electors vote as selected. With an abstemption on each side it becomes Bush 291, Kerry 245. Popular vote: Bush 50%, Kerry 48%, others 2%.

[muon2]

I think it is not a question of NY and CA as much as what hapened in the South and Mountain West. Bush drew substantially better in the  "hard R" states, probably due to the total lack of resources spent there by Kerry. This generated more votes for Bush without moving any states into his column. A similar pattern existed in 1976, where for a while there was some thought that Ford might beat Carter in the EC while Carter's strength in the South gave him an easy PV victory.