Early voting could create a Rudy surprise
The polls are open in Florida, where Mitt Romney and John McCain are in a knock-down, drag-out fight for Republican front-runner status. Even Rudy Giuliani says the winner of the Sunshine State's primary is likely to win the nomination -- a statement that has many folks speculating that Rudy will drop out if he loses, which polls suggest is extremely likely; he's in a battle for a distant third place with Mike Huckabee.
A word of caution about those polls, though. The polls are snapshots; the Florida primary is not. Floridians have actually been voting for weeks already. Absentee voting began in late December, and "early voting" started on January 14. As of last Friday, a whopping 400,000 Republicans had already voted. [UPDATE: Make that 474,000 through Sunday.] To give you an idea of how significant a number that is, a total of 699,500 voters cast ballots in Florida's 2000 GOP primary. Now, turnout will probably be much higher in 2008, since this year's primary is much more significant and hotly contested. (The 2000 primary was held a week after Super Tuesday; McCain had already conceded.) But even if the raw turnout total doubles, we're still talking about something on the order of one-third of the electorate having voted before election day. (Florida has 3.8 million registered Republicans. If 1.4 million of them vote, that'd be a 37% turnout -- which would be quite high for a primary.)
Giuliani's campaign has specifically tried to get their candidate's supporters in Florida to vote early (though hopefully not often), in hopes of "locking them in" before the inevitable decline in Rudy's momentum and poll numbers as the early-state results took their toll. As far as I know, the other candidates -- who, unlike Giuliani, actually focused their resources on competing in those early states -- have not focused on early and absentee voting nearly as much. So if Rudy does much better tonight than the current polls indicate, the reason is probably early voting.
[UPDATE: In comments, Derek suggests that I'm overstating the impact of early voting, as many of the pollsters have taken it into account. I didn't realize that.]
P.S. Judging from the RCP trend lines, absentee voting may also help Huckabee, who was running a close second to Giuliani throughout late December and early January, while early voting may help McCain, who consistently led Romney in the polls (albeit by a narrowish margin) during the first week the polls were open. The big loser from early/absentee voting would seem to be Romney, who has . He'll need strong election-day support to make up for his presumed deficit among those who voted before election day.
On the other hand, it also occurs to me that the standard polling models for determining who is a "likely voter" are probably thrown for a loop by early voting -- in other words, "snapshot" polls that captured the overall mood of the electorate at various points in the past month cannot necessarily be trusted to have captured the mood of those who were voting early during those same points in time. So, who knows? The main point is this: the ridiculously high level of early voting in Florida is an enormous wild card in today's race, and one that seems reasonably likely to redound to Giuliani's benefit (the big question is how much).
P.P.S. The prevalence of early voting also means that exit polls will be far less useful than usual, both in determining tonight's result in advance and in divining demographical and statistical information about who voted for whom and why. Exit pollsters will be missing ~33% of the electorate!