I've been reading a lot of hype around the Democratic advantage or disadvantage in early voting. While I understand that Democratic voters seem to be out-polling their counterparts in many States, I'm not really sure what to make of it. In this short diary, I would like to float a possible explanation for why these early voting stats don't really tell us much in hopes that someone will respond to tell me why I'm wrong.
It seems to me that in order to infer that the Democratic advantage in early polling is a positive indicator of Obama's chances on Tuesday, we must assume that the propensity to vote early is spread equally between Democrats, Republicans, and others. Otherwise, any advantage that Democrats have now may simply be a function of their greater motivation and energy to vote this cycle (and of course, fear that voter suppression efforts or voting machine problems will nullify their vote). If Democrats are more motivated to vote early, and more of them do, then doesn't that mean that there will be fewer Democratic voters available to vote on election day? Why doesn't an advantage in Democratic early voting just imply less election day voting?
I can imagine at least two possible reasons why early voting matters. First, if early voting is happening during a period in which Obama is enjoying a higher margin of support than exists on election day, then his position will be elevated by the fact that votes were cast in a more favorable period. So if there was a gaffe in the last week in the next 5 days, it would be less damaging because 30% of the electorate had already voted. It's sort of like an insurance policy I guess. So, inferring advantage from early voting from this perspective sounds reasonable. But simply projecting out future numbers based on early voting results is different, and probably doesn't make much sense.
Second, perhaps early voting spurs some sort of social network effect, where people who vote early are freed up to broadcast this fact to their peers and families. Also, they may be freed up on election day to volunteer for the campaign or otherwise spread the word.
Beyond this, though, I'm not sure what we can learn from analyzing the early voting results that can tell us anything meaningful about election day outcomes - especially if propensity to vote early is not spread evenly across parties. Perhaps looking at demographic composition might be useful, but that too is subject the same problem of unequal likelihood to vote early across those groups.
I be really curious to hear what others think, and hear some specific things that we can learn from early voting. Any thoughts?
STATE BLOGS
