Saturday night I was sitting with my wife drinking margaritas at the Mission Burrito on Richmond in Houston watching our kids play with about ten others when I heard the people at the table next to us talk about the Bradley Effect. I'm starting to get sick of it.
Anyway, the NYT featured a long discussion of it today, and made an interesting point--among several. Researchers noticed that the difference between a black candidates poll numbers and his or her electoral vote (the Bradley effect is the difference between the two, if the former exceeds the latter) declined after 1996. That was the year of welfare reform, so they speculate that this legislation lessened the negative images some whites had about black candidates. Since 1996, poll numbers have actually underestimated votes for black candidates -- the reverse Bradley Effect. In states where racism is still a problem, whites may actually be reluctant to express support for a black candidate.
The times they are a changin'.
Showing posts with label Bradley Effect. Show all posts
Showing posts with label Bradley Effect. Show all posts
Sunday, October 12, 2008
Saturday, October 11, 2008
More on the Bradley Effect
In response to inquiries from certain special students--or student ;)--here is a Gallup Poll story related to their attempt to determine whether race will impede or augment Barack Obama's electability:
While 6% of voters say they are less likely to vote for Barack Obama because of his race, 9% say they are more likely to vote for him, making the impact of his race a neutral to slightly positive factor when all voters' self-reported attitudes are taken into account.
At the same time, 6% of voters say John McCain's race will make them less likely to vote for him, with 7% saying it makes them more likely to vote for him, leading to the same basic conclusion: McCain's race, like Obama's, is on balance neither a plus nor a minus.
But the point of the Bradley effect is that poll respondents do not always tell pollsters (a human on the other end of a conversation) the truth. I'm not sure how much credence I'd give it.
Here's another take at what might be going on: a Reverse Bradley Effect. Whites may not want to admit to supporting a Black candidate.
While 6% of voters say they are less likely to vote for Barack Obama because of his race, 9% say they are more likely to vote for him, making the impact of his race a neutral to slightly positive factor when all voters' self-reported attitudes are taken into account.
At the same time, 6% of voters say John McCain's race will make them less likely to vote for him, with 7% saying it makes them more likely to vote for him, leading to the same basic conclusion: McCain's race, like Obama's, is on balance neither a plus nor a minus.
But the point of the Bradley effect is that poll respondents do not always tell pollsters (a human on the other end of a conversation) the truth. I'm not sure how much credence I'd give it.
Here's another take at what might be going on: a Reverse Bradley Effect. Whites may not want to admit to supporting a Black candidate.
Tuesday, October 7, 2008
Subscribe to:
Comments (Atom)