The latest polls
show Kerry tied among likely voters. That's extremely promising given the circumstances, as it indicates that his support has, if anything, solidified over the past week. In the 2000 election, Gore performed about 5% better in the popular vote than predicted by the "likely voter" polls. Even though there may be some last minute thought by voters to the wisdom of "switching horses midstream", voters at the polls generally trend Democrat anyway. Expect him to pick up a few points on Bush.
Ultimately, what matters is the electoral vote count, however... and right now, things are looking pretty good for Kerry.
The big two states this year are Florida and Ohio, both of which supported Bush in the last election. Current trends favor a Kerry win in these states. By all indications, the Kerry campaign has done best in voter registration in both states. The reason for this isn't for lack of effort by the Republicans, but simply a matter of the Democrats having more room for improvement in these states than the Republicans have, with larger numbers of previous non-voters to target who are predisposed to vote for Democrats.
The recent Survey USA poll gives Kerry a 1% lead in Florida among likely voters
, a 3% lead in Ohio
among likely voters in the ABC News Poll, and a 2% lead there according to The Ohio Poll
by University of Cincinnati. This is significant, considering that the majority of the people arriving to vote early in Florida aren't likely voters. Rather, they have been disproportionately blacks and the elderly, often driven to the polls or marching on them as a group. The fact that these voters are turning out to vote early makes likely voter methodologies in these states increasingly irrelevant.
It should be noted that the most pro-Bush "likely voter" polls in Florida, the Republican-backed Strategic Vision polls, has a "likely voter" methodology that is presumably skewed in a method similar to that of their methodology in Georgia
. Their methodology says that just 18% of votes will be cast by African-Americans, when the actual figure in the last election was 23%. Likewise, the Gallup polls used a likely voter methodology that was skewed by 12% towards the Republicans, even though the last election showed more Democratic voters. Gallup polls are similarly racially biased
By all indications, the African-American turnout will be in record numbers, both in Florida and in Ohio. Certainly, registrations in Florida and Ohio have broken records, with new registrations disproportionately from Democratic inner-city strongholds.
Assuming that Kerry wins either Ohio or Florida and that Colorado passes -- as expected -- the initiative to support proportional electoral votes, we're looking at between a 23 and 31 electoral vote shift for Kerry compared to the last presidential election. That's huge.
That would mean that Bush would have to win several states he didn't win in the previous election in order to win this time -- probably at least 3 of the following:
Iowa, Minnesota, Wisconsin, Pennsylvania, Maine, New Jersey.
Bush would also have to hold onto battleground states he won in 2000, such as Missouri, New Mexico, Nevada, Arkansas, North Carolina, and Virginia.
This just isn't very likely to happen. So, yes... look to Florida and Ohio. If Kerry wins both states, we'll all be able to call it an early night on Election Day, no matter who gets the most votes.