The Utility of Early Polls

The Monkey Cage's John Sides has a great post on some new research coming out by Wlezien and Erikson. It goes nicely with Jordan's post yesterday so I thought it was a fitting addition to the now popular topic: Do today's events and polls matter for the 2012 Election? Answer: Not really. As Wlezien and Erikson illustrate, polls taken closer to the election are much more accurate in predicting the outcome. This doesn't mean early polls are completely worthless. Early poll results may factor in the outcome by narrowing down the field or influencing candidates' fundraising. As events unfold some candidates will lose momentum, gain momentum, and some drop out of the race entirely; and this could affect the election's outcome. But, this research does support what I think most political scientists have been saying for a long while now: horse-race politics is overblown, particularly this early in the game. While many pundits like to impose meaning on events far in advance of elections ((cough) NY-26), these posts illustrate that it's a bit too early to do so. Take a deep breath. There is still a long way to go.