Wednesday, 11 February 2004

Intelligence failures

A lot of people have egg on their face over this one. Lots of bloggers staked their reputations, in some way, on it, and we utterly failed. The latest, and perhaps the most high-profile, blogger to acknowledge it, is Dan Drezner.

No, I’m not talking about WMD in Iraq. I’m talking about John Kerry’s cakewalk to the Democratic nomination. A month ago, everyone thought that Dean was going to win this thing, even though (as we now know) his poll numbers started eroding about that time.

This is, no doubt, the point where our good buddies the Perestroikans will come out and say this proves, once and for all, that attempts at a science of politics are futile (one suspects they might also think a science of chemistry is futile, but that is neither here nor there). And, if I were someone who believed that the ultimate test of science is prediction rather than explanation, I might agree with them. But in all good science, when our observations don’t conform with our theories it’s a good time to revisit our hypotheses. The hypothetical reasons why Dean should win were sound:

  1. Dean will win the nomination because he’s got the support of the base: This was the clearest argument for Dean’s candidacy. He had a massive fundraising advantage. Dean had captured the energy of the Internet and young people. He captured the anger of many Democratic rank-and-file voters against both Bush and the Iraq war.
  2. Dean has organization: Dean had spent all of 2003 setting up a formidable organization in Iowa; the only candidate who was close was near-native-son Dick Gephardt.
  3. The primary schedule favors Dean: Coming off an anticipated win in Iowa, most observers expected him to surge into New Hampshire and create an air of inevitability around his campaign.
  4. Dean was a governor: Three of the last five Democratic nominees were state governors. In a field where Dean was the only governor, the prediction that Dean would win would be reasonable.
  5. Dean has elite support: Throughout the fall, establishment Democrats jumped on the Dean bandwagon. Elites don’t generally hand out endorsements, particularly in primaries, unless they’re pretty sure they won’t be shown to be incorrect.

Yet a funny thing happened on the way to Boston. As we now know, none of these things came to pass. Why?

  1. We overestimated the base: Dean’s base was deep, but not very broad—perhaps 20% of the Democratic electorate. Dean’s gambit to widen his appeal—being the hardcore “anti-war” candidate not named Kucinich—backfired when Saddam Hussein got dragged out of a foxhole in mid-December. Dean correctly recognized that Democrats were anti-war, but not why they were anti-war: much of the Dem base doesn’t really oppose the war on principle; instead, they oppose it mostly because George W. Bush called the shots. I’d call this the “process” critique of the war, and one that John Kerry was able to capitalize on by being downright shifty with his votes and rhetoric.
  2. Organization didn’t matter in Iowa: Somehow a lot of Iowa Democrats got excited about the campaign without getting very excited about the candidates themselves, and when it came time to vote the undecided voters chose Kerry and Edwards—two candidates whose operations in Iowa were weak.
  3. The primary schedule favored a New Englander: Kerry essentially borrowed the Dean playbook—convert a win in Iowa to unstoppable momentum coming out of New Hampshire as a “favorite son.” Kerry also benefitted from the unexpected nature of the Iowa win—a Dean victory in Iowa would have been expected, and the momentum bump would have been far smaller.
  4. Endorsements don’t matter: The big kahuna, Bill Clinton, stayed out, so the endorsements Dean picked up from has-beens like Al Gore, Bill Bradley, and (maybe) Jimmy Carter were devalued.
  5. Dean pissed off the Iowans: Not just in making stuff up when asked about his connection with military veterans, but in old videotapes where Dean had an all-too-candid moment in revealing that the Iowa caucuses are mainly an excuse for presidential candiates to go and kiss farmers’ asses for months on end.
  6. Finally, Democrats recognized that Republicans were teeing up for Dean: Anyone with half a brain knew that Dean was Karl Rove’s dream candidate: an inexperienced opponent, prone to displays of short temper and highly vulnerable on social issues and defense policy. Democratic voters, who are desperate to be rid of Bush, siezed on electability as the one and only issue that matters—and decided a patrician senator from Massachusetts, like party idol/martyr JFK, was preferable to the thumb-shaped Vermont wonder in that department.

I think the bottom-line lesson here is that peoples’ political behavior is both idiosyncratic and hard to predict. Perhaps more importantly, I won’t call for an investigation of my fellow political scientists for blowing the search for the Democratic nominee if you won’t.

This is today’s entry in the OTB Beltway Traffic Jam.