One more thing before moving on from this topic...
I've been talking about Walter Shapiro's claim that if Sarah Palin could win 35% of GOP primary votes, winner-take-all rules would make her the nominee. I argued that Shapiro is wrong.
However, I've been focused on Shapiro's 35% assumption. It's worth mentioning that as Palin approached 50%, the odds start to change.
We're talking about a candidate who could win a certain, minority, percentage of the primary vote but be unacceptable to the rest of the party. At 35%, I think it's clear that the field would winnow, and the 35% candidate would lose some of the pre-winnowed contests and all of the post-winnowing contests. I think that still holds at 40%.
As the candidate approaches 50%, however, her chances improve. Now, she'll win most or all of the contests before the field winnows, racking up a delegate lead. Meanwhile, if the field winnows imperfectly, and as she gets closer to 50%, her chances of winning are pretty good. If the party is solidly against her 52-48, then all it takes is a small number of voters to cast their ballots for candidates who have already withdrawn, or perhaps for a Kucinich or Ron Paul type spoiler, for her to beat the broadly acceptable candidate. For example, in the 2008 Kentucky primary, held after the nomination was locked up, Romney and Giuliani combined for 6% of the vote (and Ron Paul took another 7%). So it's possible to imagine a scenario in which a 45% Palin edges out a Romney or Pawlenty, who draws only 44%, leaving the remaining 11% to be scattered among the also-rans. Again, that's more likely the closer Palin gets to 50.
The kicker here, however, is that the same thing could happen in a PR system. A candidate on the Democratic side who gets 48% of the vote will probably get a bit over 50% of the (pledged) delegates, because candidates below a threshhold (typically 10% or 15%) do not receive any delegates.
Is there some sweet spot in which a minority candidate could win the nomination under winner-take-all but not proportional representation? Sure. If the candidate's votes were distributed just the right way across states, a minority candidate could, in theory, win a system that was winner-take-all by state. It's even theoretically easier in a system that combines the two -- if the minority candidate can manage to concentrate most of her votes in the winner-take-all states, still picking up some delegates in the proportional states, that would help. Three years out (only two plus years before Iowa!), it's fun for political junkies, myself included, to speculate about such things. But realistically, a candidate who is solidly opposed by 50% or more of a political party isn't going to win a presidential nomination. If Sarah Palin wants to win in 2012, she'll have to do better than that.
Excellent series of posts, Jonathan.
ReplyDeleteThe thing that struck me about Shapiro's argument is that it was missing two valuable pieces: whether the winner-take-all primaries are opened/closed and when they all take place on the calendar. Well, I finally did what I should have done when I left my comment on your Palin's Future post: I put all that in a table (all 20 winner-take-all states plus DC by opened/closed and timing.) You and your readers may be interested. You can find it here (Oh, the shameful self-plug.).
Thannks!
ReplyDeleteI'm interpreting what he says pretty strictly, with the 35% as a hard cap on Palin's possibilities. If that's true, it really doesn't matter where in the process the winner-take-alls fall, because she won't win them, at least not after the first couple of states.
That said, I agree that this is one case where open/closed could matter -- and I do recommend Josh's nice post on where, when, and how the winner-take-alls are (he didn't mention it above, but the answer is that they aren't really very friendly to Palin at all).
Anyway, follow his link to Josh's site for much more detail.