Monday, November 03, 2008

Whither the Bradley Effect?

So, the election is almost over and the nation is in a state of nail-biting suspense; the Republicans because they're pretty sure their guy is going to lose, and the Democrats because they're pretty sure their guy is going to lose. It would amuse me to no end if Ralph Nader won.

By all accounts, Barack Obama is ahead and will be the next president. Democrats are still freaking out, per usual, and refusing to count their chickens before they hatch. Not that this is a bad idea- everyone I ran into in upstate NY four years ago was sure that Kerry would beat Bush, so prudence is definitely called for.

What really isn't called for is endless talk about the "Bradley Effect", the idea that some voters will tell pollsters that they are likely to vote for a black candidate, and then change their mind once they get into the voting booth due to their unspoken racism. By listening to the pundits, you'd think the Bradley Effect is a proven scientific law and we just have to see how severe it will be this time around. Even more irritating are the Republicans who are gloating about how the Bradley Effect will sink Obama, praying for racism.

So, is the Bradley Effect a proven tenet of sociology, stretching into the mists of time and deserving of all this axiomatic talk? Well, no- the Bradley Effect only goes back to the 1982 election in which Tom Bradley lost his bid for Governor of California to George Deukmejian by a slim margin after the absentee ballots were counted, even though polls taken at the election offices showed him winning. Eagle-eyed analytical types might note the obvious problem here- I'll get to that in a minute. After Bradley, there are about ten elections that have been supposedy affected by the Bradley Effect. The last was in 1992. So we're talking about a ten year stretch and a handful of samples.

The Bradley Effect is really more of a theory than an "effect". Moreover, it has never really been tested for a somewhat obvious reason- it's almost impossible to find anyone who will be interviewed before and after voting testifying that they changed their mind in the voting booth due to racial factors. In fact, it's seemingly an unverifiable theory, since it is based on a belief that interview subjects are lying about an issue that we can't find out if they were actually lying about! There are still no hidden cameras in the voting booths. Even if we could interview people, get their names, and then check their ballots- which is probably illegal- no one has done this. So we're comparing the numbers gathered by pollsters in one context to the numbers gathered by ballot counters in a different context and with a different sample group.

So, it's still completely theoretical that people really change their minds at the last minute, although we might assume some do. What may have happened in the case of Bradley is that voters at the booths favored him, while absentee voters did not. In the absence of some way of verifying that voters who: A- stated they would vote for the black candidate, B- did not, we have no way of knowing if the Bradley Effect even exists. And it doesn't really seem to exist in the case of Tom Bradley.

Also, the Bradley Effect neglects every other possible explanation for why a voter might have voted the way they did, aside from racism. Elections change on the turn of a dime, and yet we're asked to believe that, when whites don't vote for blacks, their motivations can be easily nailed down to race. It's amazing to me that so many people think so little of so many Americans.

Lastly, why does the Bradley Effect die out in the early 90s? And, if it is such a factor in the choices of white voters, why did Obama win traditional white states in his bid to become the Democratic candidate? Should we now believe that the Bradley Effect only applies to Republicans and Independents? This idea might make Democrats feel better about themselves, but why in the world would Republicans hope that their voters are racists?


Look, I don't think that Obama has this election in the bag either. I'm a relatively pessimistic person by nature. But, what I do believe is that most people who vote McCain will do so because they agree with his vision of the future, like what he has to say, and approve of his character; and not because they're closet racists. Similarly, I think that most people who vote for Obama will do so because they agree with his vision of the future, like what he has to say, and approve of his character; and not because they're dupes, cultists, or just want to vote for the black guy.


And I do have to wonder what standard of evidence people who talk about the Bradley Effect have? It seems more to me like a fantasy than a verified theory.


(Note: I do think that The Bradley Effect would be a decent name for an indie rock band.)

2 comments:

Brian Dunbar said...

the idea that some voters will tell pollsters that they are likely to vote for a black candidate, and then change their mind once they get into the voting booth due to their unspoken racism.

With respect, the explanations I've heard for that do not imply racism this year but rather that the people who respond falsely don't wish to be _thought_ of as racist.

In it's original meaning, I understand there was racism involved.

I don't take it seriously for the reasons you note - speculating about it as useful as wondering how many angels can dance on the head of a pin.

rufus said...

Ah, well hopefully we can now retire it. Sure, there are some nutcases- of all stripes- but I think most Americans are pretty normal people.