Monday, April 28, 2008


"Obama Youth are just part of the marketing plan"

Obama Girl
One of the terms that is being mentioned a lot in the presidential election this year is the "Bradley Effect." It comes up in context of the Democratic primaries choosing between Senators Obama and Clinton, and is sometimes not explained well.

The Bradley Effect originated in the 1982 gubernatorial campaign for California. Tom Bradley, black mayor of Los Angeles ran against George Deukmejian, white Attorney General for the state. Polling data before the election gave Bradley a consistent and handy lead, exit polls on election day proved he had the win. The media declared Bradley a winner based on these polls, but when it was all counted, Deukmejian had won. Analysis of the polls revealed that a lot of the people who claimed they were going to or had voted for Bradley voted for Deukmejian instead.

Deukmejian's campaign manager had suggested this would happen before the election, and while he resigned due to the suggestion that racism was the cause, in the end the analysis is generally that he was right: bigots hating African Americans tried to cover their shame by pretending they would vote for the hated brown person, or so we're told.

The reason the Bradley Effect is brought up in this election is because Senator Obama is seeing a similar pattern. Polling before the election and exit polling both show his number larger than they actually end up in the end. Racism is murmured about: America is packed with bigots! For example, CNN has this article on their website:
As both Sens. Barack Obama and Hillary Clinton race to make history, some political observers believe Obama might have a unique problem because he's African-American.

They claim voters often say they'll go one way, then act quite differently at the polls when it comes to a candidate's race.
Notice the "political observers" are not named, which suggests to me it means "guys at the news office" but the tip toeing around this issue is pretty common. Nobody wants to (yet) come out and say "if you don't vote for Senator Obama you're a bigot!" yet it is being implied in some places and whispered in others. CNN calls this effect "unspoken" then goes on to speak about it, but the implication is clear: this is out there, but not being mentioned much.

The thing is, I'm not sure that is accurate. Could there be another reason than racism why the polling data shows higher numbers for Bradley or Obama (or Wilder, in another election in Virginia) than they actually get?

Some suggest that it isn't so much racism as a desire to be perceived as not racist. When asked questions, pollsters have found in the past that at least some people will tend to give the answer that they think the pollster wants to hear or is looking for, so they will be "right" on this "test." Pollsters attempt to avoid this by how they word questions or order the queries, but in a direct "who will you vote for" analysis, that's not realistic.

I think there is some merit in that, and the thought that some people will give statements that they think they are supposed to in society, then act as they are otherwise inclined to when it comes to voting. Yet there's a post by Karl on Protein Wisdom that has a comment I think answers the question better than any of these options:
How do they know that people said one thing and did another? They don’t poll every potential voter, but they do count every actual vote.

What if the voters who say they’ll vote for the black guy are less likely to actually show up at the polls than those who don’t? Maybe they’re not in the habit of voting, or they don’t have a ride, or they’re just not that into it.

I’d call it the “Obama girl” effect, myself. It ain’t racism, folks: it’s inertia.
-by Dicentra
There are two offerings here, first the basic "polling is not as accurate or useful as the legacy media seems to think" which is my default position. You poll 800 people and claim to speak for 300 million? Get a real job. Polling is largely divination, like reading tea leaves: it can be accurate if done properly, but that's more to do with common sense and luck than any science.

However, it is Dicentra's second point that most intrigues me and makes me wonder if it isn't the real answer - both for Obama and Bradley. What if, instead of any sort of racism, the answer has more to do with sloth and enthusiasm? Senator Obama is said to be getting large numbers of youth support, that young people love the guy and are big supporters. The crowds are full of younger voters, his volunteers tend to be younger (when they aren't Che supporting radical freaks). Yet the mythical Youth Vote

Here's a bit I wrote about in February about the Youth Vote:
Here's a classic example of this sort of outlook. remember the Obama Girl? The girl who sang a pop soul song about how she had a crush on Senator Obama? She didn't vote.
“I didn’t get a chance to vote today because I’m not registered to vote in New York,” she said.

So where is Obama Girl registered to vote?

“New Jersey.”

Um, but didn’t New Jersey also hold a primary?

True. The problem, she explained, was that she was sick in New York City and was unable to get back across the Hudson River to the polls in Jersey City.

“I was in Arizona for the Super Bowl — every time I get in the airplane I get sick,” said Ms. Ettinger, who did manage to make it to the Svedka Fembot election returns party at Chinatown Brasserie at Lafayette and Great Jones streets.
At least she got to the post-election party. Miss Ettinger is 26 years old and represents no small slice of society.
The Obama Girl effect is a subsidiary of the Youth Vote myth, in which young people who are big supporters of a candidate are very loud in support but very light in actual turnout when it comes to voting.

Another commenter suggested a pattern for youth when it comes to elections:
Remember back when you were young. The motivations are far more primitive and simple.

* 1) Raise some hell a lot. We’re bored. Maybe get in that cute poll workers scivies. Being where the action is. Party on. Political action. Fuck Yeah!!!111wlwventyonest!!!11

* 2) Voting? Actually going to a polling booth and hanging with “the man”? Ewwww. Thats so what parents and old people do.

- The rites of spring.
-by Big Bang Hunter (pumping you up)
This makes more sense to me than mere racism, which while it might play a part does not really seem to fit this campaign. For the Bradley campaign, a similar effect might have happened: people who supported the black candidate against the cracker republican, but weren't particularly inclined to vote. My vote doesn't matter, it's all rigged anyway, voting is too white, voting is too old, voting is participating in the phallocracy.

The exit polling data is a bit different, but I think that falls more under the basic flaws with exit polls. They're almost always off, partly because most people avoid pollsters if they can unless they are really zealous or want to share their opinions. Another factor is that the people who do exit polling tend to fit a certain profile: young, eager, left-leaning. This will affect who wants to talk to them and who will tend to avoid them. Exit polling almost always leans left, giving the more leftist candidate more points than they actually got. This has something to do with the voters themselves: who's more likely to talk to a pollster after voting, who is more likely to have to get busy or move on, or just does not care to tell someone who they voted for?

There's real racism in America, and it has some effect on elections, but I really do think the primary reason for the polling discrepancies (beyond basic flaws with polling) has little to do with racism.
[technorati icon]

No comments: