Friday, January 18, 2008
Compelling assessment of John Edwards from Russ Feingold:
The one that is the most problematic is (John) Edwards, who voted for the Patriot Act, campaigns against it. Voted for No Child Left Behind, campaigns against it. Voted for the China trade deal, campaigns against it. Voted for the Iraq war ...
MyDD is temporarily stunned! ... 2:33 P.M.
Thursday, January 17, 2008
'Bradley Effect' Update: Obama is ahead by 9 points over Hillary in the most recent Mason-Dixon poll of South Carolina voters. But can we trust voters to have told pollsters the truth--or are racial concerns (including the desire not to offend) leading them to give inaccurate answers?
a) Black 'Bradley' Voters? Noam Scheiber weighs in again on the possiblity of such a "Bradley Effect" for black voters. It all depends on the race of the interviewer, he argues--suggesting that when the interviewer is black, some black voters may opt to (falsely) show racial solidarity, but that
when African-Americans are in the presence of whites, the greater social fear is being considered a "race man" ...
Debra Dickerson isn't buying that, and neither am I--though it's an empirical question that presumably could be resolved one way or another.
b)White 'Bradley' Voters? Meanwhile, Emailer Z, who knows his or her polls, argues the Mason-Dixon poll might not have such good news for Obama after all--given the more-often discussed tendency of white voters to occasionally mislead pollsters:
Here's how the Bradley Effect works: A stranger calls you to ask how you intend to vote. You do NOT intend to vote for the African American, but you don't want to get a lot of guff from this stranger about how you must be a racist if you won't vote for the African American. So you answer, "Not sure." In all the classic Bradley Effect elections (and NH fit the pattern), the polls got the vote for the African American about right, but OVERREPORTED not sure and UNDERREPORTED the other candidate's vote.
So when the brand new MSNBC-McClatchy-Mason Dixon poll in SC says there are twice as many undecided in the Dem race (15%) than in the GOP race (8%), you might suspect Bradley-ism in that poll. So what looks like a 9-point Obama lead with a fat undecided might in fact portend a very close race, no? [E.A.]