« The insipid Iran threat debate | Main | How bad was MS-01 for the GOP? »
Updating my series on the state-level predictors of Obama support, the graphs below (which include 95% confidence intervals around predicted linear fits using data from before yesterday's primaries) show that white support for Obama in Oregon and Kentucky fits the pattern we expect by education but not by black population:
TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d83451d25c69e200e55252be4f8833
Listed below are links to weblogs that reference Quick KY and OR results analysis:
The comments to this entry are closed.
I am an assistant professor in the Department of Government at Dartmouth College. I received my Ph.D. from the Department of Political Science at Duke University in 2009 and served as a RWJ Scholar in Health Policy Research at the University of Michigan from 2009-2011. I also tweet at @BrendanNyhan and serve as a media critic for Columbia Journalism Review. Previously, I co-edited Spinsanity, a non-partisan watchdog of political spin, and co-authored All the President's Spin. For more, see my bio or academic website.
Content feeds:
-Blog RSS
-Twitter feed
-Twitter RSS
Email:
bnyhan@yahoo.com
Oregon and Kentucky don't match your racial trend, but they seem to fit perfectly into the regional one.
Looking at maps on the county-by-county level (at the link below), Hillary has won by 30% and more throughout the Appalachians while Obama has done extremely well in the North West and the South Eastern coastal plains.
This has been true throughout the election.
http://www.openleft.com/showDiary.do;jsessionid=060D3B4D8BD64FE039A5CCD6CAFA7DF4?diaryId=4503
Posted by: Jinchi | May 21, 2008 at 12:11 PM
It looks like you should consider modeling support vs education via an asymptotic exponential rather than a simmple linear relationship.
Posted by: | May 24, 2008 at 06:44 PM