On The Bradley Effect

Bouncing around the Free Republic posts in their 2008 Polls folder, I've found that their forum commenters react to Obama's increasing lead in the polls with three different talking points.

1. All polls are skewed to favor Obama. McCain is really leading but pollsters are "in the tank" for Obama. They are either tweaking their data or only polling people and areas where they know Obama would lead.

2. McCain is getting what he deserves for not being a true conservative and for not getting "down in the gutter" with Obama. This group usually uses "Hussein" to refer to Obama and is more likely to repeat the "Marxist, Muslim, closeted homo" trio of descriptors.

3. The polls are real but we have nothing to worry about thanks to the Bradley Effect.

That last one is the one is the one that actually worries me. In case you're unaware, the Bradley Effect is used to describe the phenomenon of voters telling pollsters that they will support the black candidate but then vote for the white opponent. It's named after California gubernatorial candidate Tom Bradley, who in 1982 led polls before the election and even led in the exit polls, prompting the SF Chronicle to publish its Deweyesque headline Bradley Win Projected. White candidate George Deukmejian was the actual winner.

As Obama's lead has stretched in recent weeks, virtually all op-ed writers in the country have discussed the Bradley Effect. Google shows hundreds of Bradley Effect articles in the last month alone. Some say that polling in 1982 was not the exact science it is today. Some dismiss the phenomenon as a relic of racist America that has greatly dissipated since the Bradley election. But others say that in our increasingly PC times, even more voters are inclined to tell a pollster they'll support a black candidate when they actually have no such intent. I fear the last to be true.