Yale econometrician Ray Fair, whose model of presidential election outcomes I have discussed here, has issued his final prediction for the 2004 election. He believes that Bush will get 57.70 percent of the two-party popular vote. If that were to happen, Bush would walk off with 461 to 518 electoral votes (explanation here, see method 3).
I see a much closer election, with Bush getting about 51 percent of the two-party popular vote and somewhat more than 300 electoral votes. I'll issue a final prediction on election eve.