A theory on a confusing Presidential poll, informed by the CD7 primary

Without research, one could be forgiven for assuming that John Edwards and Barack Obama are duking it out to become the liberal alternative to Hillary Clinton, whose pro-Iraq War, anti-free speech record has drawn the ire of activists. But the polls show something different. As noted at Eschaton:

Obama/Edwards supporters don't seem inclined to support the other one. Clinton is a popular 2nd choice as well as 1st one, and when Obama or Edwards are excluded from the poll their support largely shifts to Clinton.

Conventional wisdom, thrown under the bus by the truth.

That poll also shows support for Edwards as first choice nearly doubling from 9% to 19% since November 2006, while Obama's support has remained level at 19% - 20% since the day he declared. And there I go, evidencing Atrios' interpretation. I lean towards Edwards and I find myself bashing Obama.

Meanwhile, a very recent University of Iowa poll (found here) reveals this:

Results show that Edwards remains the leader among likely Democratic caucus goers, competing primarily with Clinton for caucus support. Edwards led by a substantial margin with 34.2 percent. Clinton followed with 28.5 percent, and Obama with 19.3 percent... Although Edwards led in support among Democratic caucus goers, this same group believed Clinton was the strongest candidate.

Why would the caucus-going supporters of two theoretically similar candidates default not to each other, but to a third, less-similar candidate? Consider the CD7 primary of 2006, where Peggy Lamm, a female centrist with a history of cooperating with Republicans, faced off against two ostensibly liberal male contenders. One was Ed Perlmutter, a Democratic insider with high name recognition and a long history of campaigning in the contested district. The other was Herb Rubenstein, an intelligent, affable newcomer who could claim early opposition to the Iraq War. Three months before the election, the polls put Perlmutter at 51%, Lamm at 31% and Rubenstein at 6%.

Now exit the realm of provable facts and consider a theory.

In a Democratic primary, voters are torn by competing forces. They want a candidate they can ethically support and they also want a candidate who can win a general election. An ideologically-driven candidate's mission, then, is not just to sell him or herself, but to sell the idea that an ideologically-driven candidate can win the general election. If the candidate fails, primary voters shrug their shoulders and vote for the centrist.

In CD7, Perlmutter succeeded because the voters of CD7 knew him as a person. Edwards is following a similar path in Iowa, where he has been campaigning forever.

Rubenstein did not have Perlmutter's name recognition or ground game, so he tried an alt-strategy, reaching out to anti-war voters, bloggers and activists. Similarly, Obama has not spent as much time in Iowa as Edwards, so he is trying his own alt-strategy, becoming a fundraising juggernaut and outright celebrity who the voters of Iowa may soon see as a potential winner.

If Edwards/Perlmutter or Obama/Rubenstein supporters conclude that their candidate is a loser, they are likely to default not to a similar candidate, but to someone they view as a general election winner.

There are doubtlessly imperfections in this comparison. But if you're thinking I'm forcing it just to give myself an excuse to start posting Christina Aguilera videos again, well, that's just crazy. Since you brought it up, though: