More surprises might be in store. One of the important functions that party identification serves is that of an educational tool. If I told you that you must choose between Mike Smith and Tim Jones for Congress, and that was all, you would not know whom to support. But what if I told you that Smith was a Republican and Jones was a Democrat? In that case, about 80 to 90 percent of voters would then have at least a direction to lean.
In nomination battles, party identification has absolutely no meaning. This implies it is hard for voters to develop a sense of which candidate is closest to their views, and in turn we can see some surprising swings throughout the season. We have already seen a lot of volatility this year. The RealClearPolitics average of the Iowa polls has had a whopping six candidates occupy the lead at one point or another.
And this is not the first time we have seen a lot of volatility. Barack Obama did not really begin his surge until November 2007, just a month or so before the contest. Ditto Mike Huckabee. And, what’s more, the final polls understated the size of Obama and Huckabee’s lead, suggesting that a good chunk of voters swung toward the frontrunner at the very last minute. The same was true in 2004: John Kerry was basically tied in the polls with Howard Dean, John Edwards, and Dick Gephardt until just a few days before the caucus; he ultimately won by a very solid 6 points.