An Iowa Presidential Caucus Primer

If you think that Pittsburgh politics are played by an arcane set of rules, then you've never been part of the Iowa presidential caucus system. As most adult Americans are aware, the 2008 presidential campaign is well underway, and the Iowa caucus on January 3 is the first "official" event leading up to the nomination of candidates by the two major parties. The Post-Gazette's pre-caucus story today (link here) does such an awful and misleading job of what the caucus system is about that I feel obligated, as a former political organizer in Iowa, to correct the record.

The PG hammers on some conventional themes: Iowa is an unrepresentative state, where a handful of voters wield disproportionate impact over the identity of the next president. Much, however, depends on turnout.

That narrative makes Iowa sound simply like an early primary in a thinly-populated state; the problem is that some unknown and possibly random population turns out to vote.

But there is no voting in an Iowa caucus, and turnout isn't unpredicable -- not, at least to the campaigns.

Campaigns win or lose in Iowa (or beat expectations, which is far more important) based on their organization, and organization is built in an old-fashioned face-to-face way: one caucus-goer at a time. For more than a year, the major campaigns have been assembling lists. By caucus night, each candidate wants to know the name, address, phone number, and (where possible) email address of every soul in the state who plans to attend a caucus for the candidate. That night, the task of the campaign is to ensure that each of those souls shows up, at the right place at the right time, and stands in the candidate's group to be counted. If extra souls show up, that's a bonus. But no campaign manager goes to sleep on January 2 (if she goes to sleep) wondering what will happen on January 3.

Today's New York Times feature on the Iowa caucus captures some of this flavor. The Edwards campaign is built on a backbone of true believers from 2004; Clinton and Obama are trying to expand the system by recruiting new caucus-goers. The endgame is the same for all three -- build a list -- though the Times overstates the novelty of the Clinton and Obama strategy. I did exactly the same thing in 1983 and 1984, when I worked the Iowa caucus. We didn't have computer data to generate predictions, however. We identified new caucus goers the old-fashioned way. We built phone banks, and we made thousands of telephone calls.

Among other things, the emphasis on organization means that the representativeness/unrepresentativeness argument loses almost all of its force. Iowa became a significant campaign event in 1976, and remains occasionally a significant campaign event, not because Iowans have any special sense (though Iowans would contend that they do!), but because Iowa is a testbed for a candidate's ability to organize and maneuver. Four years ago, I knew in the early Fall that Howard Dean's campaign was doomed -- even though he registered well in the polls and had all kinds of neat-o Internet stuff going on. No meaningful organization on the ground.

Voting? What do I mean that there is no voting?

Caucus-goers register their preferences by standing in the corner of a room with other supporters of the same candidate. There is no ballot box or voting booth; by definition, your preference is public, not private. The caucus occurs at a specific time -- it begins at 7 p.m. -- and you have to attend in person if you want to participate. There is no absentee balloting; you can't register your preference earlier in the day.

And not only is your preference public on the night that you express it, but it is public for all time. The Iowa Democratic Party runs its caucuses, not the State of Iowa, and the IDP has a comprehensive and robust database of caucus attendance that goes back at least to 1980, and which includes what candidate each caucus-goer supported.

I won't venture a meaningful prediction this year. But here's a tidbit. Mostly lost in the media's affection for polling data and campaign rallies is the fact that Clinton's Iowa campaign is being run by an Iowa native and caucus veteran named Teresa Vilmain. That's a strong signal that Clinton will do well; Teresa Vilmain knows how to run a caucus, and she may know how to run a caucus better than anyone else in the Democratic Party. I knew Teresa briefly during the 1983/84 campaign, when she worked for Walter Mondale. (I worked for Alan Cranston, the senior senator from my home state.) The Mondale staff and the Cranston staff each relied heavily on veterans of John Culver campaigns (Vilmain got her start working for Culver), and the two teams often drank together late at night at a Des Moines bar called Carl's. (To the best of my knowledge, Carl's is still around, though I don't know whether it's still a party hangout.) She's a winner; Mondale took almost 49% of the weighted "vote" in Iowa in 1984.

The larger lesson, however, is that Iowa is a fickle guide to national results. (In fact, on the Democratic side, only Jimmy Carter in 1976 has ridden a surprise showing in Iowa to victory in November. In 1992, Bill Clinton even finished behind "Uncommmitted.") Mondale was the Hillary Clinton of 1984: a candidate so overwhelmingly favored to win in a field of "lesser" candidates that anything less than a crushing win would be viewed as an effective defeat. (I'm not the first to make this comparison; the dean of Iowa political writers, David Yepsen, has written the same thing.) Mondale did secure a crushing win. Gary Hart finished second, with 16.5% of the "vote." But the failure to top 50%, and Hart's surprise emergence, cast a pall on Mondale that lasted through the general election.