running for office in the eighties. Many believe it occurs because white
respondents are more likely to tell a pollster they will vote Black than
they really will. Studies of Obama voters have shown a similar effect
among African Americans. This effect has minimized over the years.
Potential reasons are that modern pollsters adjust for it. Studies
suggest that you can minimize the effect by matching the surveyors race
up with the surveyee.
The gap between votes and polls is falling. In some states with larger
Black populations, a reverse Bradley Effect may even be occurring. This
suggests that it will be more pronounced in the general election than in
the primaries. I think the larger effect is the enthusiasm gap. Racially
influenced Whites are likely to tell a pollster they will vote Obama,
secretly plan to vote McCain because of race, but actually stay home
because they are disappointed in the Republicans. In short, more will
stay home than lie.
The impact of the effect on Obama cut both ways, helping him among Black
respondents and hurting him in some mostly White states. The impact in
the general election has many nervous. A wide margin in polls in key
states is his only defense.
Bradley Effect: http://en.wikipedia.org/wiki/Bradley_effect
Comments