We've been paying attention to early voting in this election cycle, because it is both an increasing trend, and also a form of voting that has significant impact on some our next-stage election technology efforts around polling-place -- or early voting place -- operations and technology. As we were told by MN SoS Mark Ritchie (reflecting on the high vote-by-mail turnout in 2008), one reason for election officials' interest in early voting is that it may increase participation overall, but decrease reliance on vote-by-mail (VBM). VBM is another way of voting early, but one that places significant burden on local election officials (remember the "Lizard People" ballot?) that some might prefer to shift to early voting.
That made sense to me personally, but there is more than meets the eye, as reported by the New York Times: "Voting Early, but Not So Often"
Apparently, the availability of early voting, by itself, appears to actually decrease total participation. However, when combined with same-day registration, the net effect is positive. But don't take it from me, read the article and see what Professors Burden and Mayer of U. Wisconsin found out.
The connection with same-day registration brings up a tricky policy question, though. Because until recently, there was not a lot of early voting, the phrase "same-day registration" has typically been understood in the context of election day: you go to what you think is your proper polling place, fill out a voter registration form, vote (or vote provisionally) and hope that everything gets sorted out correctly in the election-day flurry and that you actually voted. (I've certainly heard first hand cases where this didn't actually work out exactly right.) Same-day registration is a bit controversial because of some folks' concern for abuse and voter fraud; and also because of other folks concern about proper follow-through and enfranchisement or lack thereof.
But in the context of early voting, same day registration might be less concerning on both counts, and more amenable to public transparency than your typical volunteer-run polling place -- with some assistance by appropriate election technology. :-) That's one reason for our interest in furthering our work on voter-registration technology, and polling-place technology, and maybe even coming up with a chocolate/peanut butter-like combo that hits the spot.
There's certainly more to learn about what's required, but the good news is that in this election cycle, we have many sources for that learning.