« The Kenyan Electoral Train Wreck Continues | Main | Caitlin Flanagan Sounds Like Anthony Kennedy, or Desperately Seeking Social Science »

Hoisted from Comments: Jim Gimpel

In response to my earlier post, Jim Gimpel offers this take on the consequences of frontloading in 2008:

There are very good reasons for why a highly compressed primary schedule may not winnow the candidates as quickly as everyone expected it to.

It’s because winnowing requires momentum, and building momentum requires time, days and weeks to drag out, between contests. People need time to process information so as to vote in sophisticated, (e.g., voting for a second or third favorite candidate); vs. sincere ( e.g., voting for their first preference) fashion.

What this new schedule has done is allowed each of the top 3-4 candidates to now focus on their own distinct pocket of electoral support: SC, FL, MI, NV, etc.

So, for example, on the GOP side, Huckabee goes for SC, Giuliani hopes to take FL, Romney wants a win in Michigan, and so forth. Romney and Giuliani, in particular, are only a token presence in the other states.

This development, oddly enough, could delay winnowing, not speed up the winnowing process like so many expected it to. That’s because without time for momentum to build, and sophisticated voting to kick-in, voters cast ballots according to their sincere (first) preferences, and each of the top 3 or 4 candidates each take a state, or two, and claim to be on their way to the nomination.

So suppose Giuliani takes FL, Huckabee takes SC, Romney takes Mich, perhaps. McCain is still in. Still no frontrunner.

Now tell me that ISN’T what is happening this year in the GOP? If so, it could be awhile before we see the nominee. Ironically, it might have been a faster route to the nomination under the former, non-compressed, system.

Yes?

Now let’s see how long it takes for political scientists to figure this out and actually start writing about it. No doubt it will take years, decades perhaps!

Let’s hope not! Thanks to Jim for these comments. Any thoughts in response?

[“Hoisted from Comments” brazenly stolen from Brad DeLong (example).]

Comments

For me, the less winnowing the better. Keep more candidates in the race and give then entire country, not just a few randomly selected states, a chance to chose the nominees. Best possible outcome would be conventions that actually serve as more than a party love-fest for someone named as nominee months earlier. Additionally, keeping them fighting longer might show something beyond the scripted talking points and stump speeches.

Might.

Well, first, isn't winnowing happening just as rapidly as expected? For
the Dems, we're down to three candidates, only two viable, after only two
events, with three plausible candidates that made it to Iowa entirely out, and a major candidate (Edwards) barely hanging on. On the GOP side, we're
down to three still-plausible candidates, although Thompson and Guiliani are still around; granted, they've had three (sort of) events, but it seems that winnowing is proceeding more or less normally. Just four years ago, Kerry, Edwards, Dean, and Clark all survived NH, and IIRC each won at least one later event -- that's not going to happen on the Dem side this time, and I think it's unlikely that four different GOPers will win an event this time.

More to the point, though: the main actors here aren't voters (by switching to strategic voting); the main actors are parties, who winnow by means of endorsements, donations, etc. See e.g. Cohen et al in the current _Forum_ and their forthcoming book, and more broadly any of the party network literature. I've speculated (Rise & Fall of Howard Dean, in the Forum a while back) that party elites are now using voting the way that pre-1972 parties used primaries, as useful information; thus party leaders on the Dem side chose against Richardson, Dodd, and Biden before Iowa, but used Iowa and NH information to further narrow to Obama and Clinton, and in particular to conclude that Obama was
electorally viable. On the GOP side, the problem isn't that voters aren't winnowing; it's that GOP party leaders can't figure out what to do given the lack of a broadly acceptable candidate(again, see Cohen et al.).

Is winnowing really happening as rapidly as before? Sure, the also-rans have generally hopped off the bandwagon in the Democratic party but we knew they were also rans a year ago. The true dynamic of the race -- a contest between Hill, Obama, and Edwards -- doesn't seem like it has changed to me. Moreover, both of them seem to have enough organization and fundraising to get through this next stage. Given strategic considerations that will probably mean that the candidates don't compete in all of the same states, it looks possible that February 5th won't do much but get rid of Edwards (and then only because he lacks funds).

The Republican side is just a mess, partly because the guys who can campaign don't have money and the guys who have money don't really inspire anyone as candidates.

Interesting side note -- there is a real, and generally overlooked, race for superdelegates in the Democratic primary. As long as someone like Hillary can persist in keeping those folks on board, she can probably take a few more hits in the primaries. (Sidenote: good article in the LA Times on this today, so I can't claim credit for thinking about it all on my own.)