Monday, January 29, 2007

The Iowa Caucuses

Despite an outpouring of support and enthusiasm from my six readers, I regret to announce that I have decide against forming an exploratory committee to run for President. As a member of the minority of Americans not running for the nomination, I have therefore turned my attention to the majority, and who will win my invaluable support. (My backing in 2004 was instrumental to bouying the Kucinich, and later, Dean campaigns during their slow failures.) I begin by looking at the eye of the needle, the Iowa Caucuses, though which the eventual winner must thread him/herself a little less than a year from now.

Here's what we know: the Iowa Caucuses are a terrible predictor of who will receive the nomination, but they're really good at determining who won't. Except in very rare cases, the candidate who receives the nomination has finished in the top two in Iowa. And those rare cases have very specific factors.

In1992, when Clinton finished third (fourth, if you consider the Iowans who voted "uncommitted"). But that was a weird year, because Iowa Senator Tom Harkin sucked all the wind out of the state by winning 76% of the vote--more than any candidate in history (including incumbents) and 28 points higher than the next closest non-incumbent received (Mondale in '84). A second exception came in '88, when eventual nominee Mike Dukakis was edged out by Midwesterners Paul Simon (Illinois) and Dick Gephardt (Missouri).

To win the Presidency, then, you must be in the top two unless you're running against a candidate from Iowa or a neighboring state. It's not an exact science, but it puts into context the candidates in '08, three of whom are heavy hitters and one of whom could become one if he enters the race (Gore). It's possible that Gore won't enter the race and that the big three (Clinton, Edwards, and Obama) will stay alive by virtue of a one, two, three finish--after all, Obama's from Illinois.

But winning Iowa isn't always the greatest thing. Of the last seven non-incumbent caucuses, the eventual nominee won only four. On the Republican side, where there were fewer non-incumbent elections eventual nominees went two of four.

However, Iowa's a better predictor of the mood of the country (if not its politics). The eventual Presidential winner won the popular vote in Iowa eight of nine times (including 2000, when it went for Gore). The only outlier is '88, when Iowans gave Dukakis a resounding ten-point win. That was also the only election in which Iowans were more than just a percent or two off the national vote (except in '84, when neighboring-stater Mondale got 6% more in Iowa, in a losing effort, than he got nationally).

So for the Democratic candidate looking to change the curtains in 1600 Pennsylvania Avenue, these are some key things to remember: Iowans reflect the mood of the country in broad terms; being out of step with Iowans is a bad thing. Winning Iowa isn't any prescription for winning the country, but losing it big means not being around to win the nomination in the subsequent primaries (this year in Nevada, New Hampshire, and South Carolina).

This post has been updated. One of the interns caught a typo; I've beaten him for missing it before we went live.


Anonymous said...

You need to beat him again...

Zak J. said...

I think Iowa is more important for an up and coming like Obama or Richardson than for an established name like Clinton or even Edwards. Winning among the party faithful who show up for caucuses is going to hard for DLC support like Clinton anyway. But winning in Iowa will get you national attention, and party faithful are probably (let's hope) paying more attention to substance and positions than media hype, so even the votes of a demographically non-representative state like Iowa are still important for the rest of the nation to pay attention to.

If she loses Iowa, Clinton will shrug it off; but if Obama loses Iowa, he'll have a very long row to hoe.

Zak J. said...

Meant to say "going to BE hard for A DLC supportER like Clinton." Your proofreading virus must be catching.