Saturday, 26 January 2008

Back of the envelope

As noted by me at OTB, one question going forward for Democrats is whether or not Barack Obama’s voter breakdown by race carries into the next primary states. Here’s some quick-and-dirty math for Florida’s zero-delegates-except-if-Hillary-says-so primary on Tuesday.

Assuming Obama gets 80% of the black vote and 25% of the non-black vote, and 24% of Florida Democratic primary voters are black (assuming no differential turnout, based on Florida’s registration statistics from December 31st), Obama should get around 38% of the Florida vote. That’s well ahead of how Obama has been polling in Florida, so I’m not at all convinced that the extrapolation works well even though Obama’s average has been tracking upwards slightly in the state and one would expect that Florida whites would be less racially conservative than South Carolina whites. I think the safe money is that Clinton will still win the state and its 0 delegates comfortably, but I wouldn’t be overly surprised with a result like 40–35 or so (with both candidates receiving about equal numbers of those 0 delegates at stake, given the Democrats’ high threshold-PR rules).

Old wine in new bottles

Josh Patashnik of The New Republic discovers that Republicans and Democrats have divergent beliefs about the state of the national economy (þ: JustOneMinute). Clearly he doesn’t have a subscription to the American Journal of Political Science, where my dissertation chair and two co-authors showed this to be the case seven years ago based on 1990s data, well before George W. Bush set up camp in the Oval Office (see also Duch and Palmer 2001, which demonstrates the same effect among Hungarian voters).

The moral of the story: those who do not read the political science literature are condemned to reinvent it.