As we know from the past couple of elections exit polling is hardly an exact science, but I believe that this year it may be completely worthless for several reasons that I will explain below.
First, this is not a "Bradley Effect" diary or anything resembling a concern for Obama's victory diary, as I am pretty certain he is going to win (and big), just not as big as I think the exit polls will have us believe (the regular polls going into the election are likely to be more accurate). Here are the reasons:
- The enthusiasm of Obama's supporters means that they may be more likely to talk with the exit pollsters than people have in the past, which could heavily skew polling for Obama, which leads us to
- A lack of enthusiasm by McCain supporters, who may not wish to admit that they voted for McCain/Palin and thus likely avoid the pollsters or (much worse, lie)
- A real problem with estimating turnout in light of both early voting, but especially as a compounding factor of the enthusiasm issue. Pollsters may very well overstate turnout (on the actual election day) by not accurately factoring in the hordes who voted early (this could obviously go the other way as well).
- A jump on the winner bandwagon effect that may lead many McCain supporters to potentially lie to pollsters if the election looks like it was going to go big to Obama in the days before the election (they likely still vote for McCain, but again don't want to admit it (this is different from #2 because 2 is factoring out people who were embarrassed by voting for McCain, while #4 is people who just want to "be right").
- The last one, and one I hope is wrong, but think is probably right is underestimating the number of people that simply vote Republican no matter what in the end. These are the people that brought us Bush in 2004 and will likely come home to roost on election day no matter who the Democrats run out, but aren't counted easily because they fall into category #2 above.
In the end, Obama is likely to win this big (350+ electoral votes) and with well over 50% of the vote, but I have a feeling that exit polling is going to lead us to believe that the win will be much bigger earlier in the day and thus should be taken with a grain of salt.