PolitiPorn™

Can we once and for all stop talking about the supposed “Bradley Effect” and it bearing on the current primary battle between Sens. Barack Obama and Hillary Clinton. The latest in a long line of chatter comes to us from Susan Estrich, Fox News contributor and (from the appearance of the picture associated with her column) antithesis of Dorian Gray.

While not cackling like a witch on crack, Estrich offers the ‘liberal” view on Fox News which under the strictures of Fox News programming permits her to be slightly less sycophantic toward the current administration than other contributors.

She posits that the reason Obama was polling well ahead of Clinton in one California poll and with a smaller margin in other polls, but ultimately lost by a significant margin to Clinton has to do with the “Bradley Effect.”

To pundits the Bradley Effect is taken as gospel and can be used to explain other occurrences. It refers to a California gubernatorial election in which an African American candidate with a significant lead in the polls was ultimately defeated in actual voting. The logic behind the supposed effect is concise: When in the privacy of the polling booth individuals will succumb to racial prejudice and vote for a white candidate instead of the politically correct choice they told the pollsters.

However, we are well advised to remember that complex problems will have simple, easy-to-understand, wrong answers. Who decides that the “Bradley Effect” has taken place? Pundits? Pollsters?

What biases might those two groups bring to the table?

Pollsters believe that they are executing a scientific procedure that elicits infallible results (of course their use of the term “margin of error” demonstrates a slight problem with this position). If the voters do not act the way the pollster has predicted it must be a problem with the voters. It couldn’t possibly be that the polling was inherently flawed. Could it?

Pundits are by their nature inclined to believe that polls are infallible. After all, without polls how would they fill six hours of chatter every day talking about the race? If they said, “We don’t know what is going to happen,” the shows would last 15 minutes and repeat, like those AM stations used for national park information. Instead we are treated to a poll showing one candidate at 43 percent, the other at 42 percent and being told that the former is leading even though the margin of error is 5 percent.

So what explains the differences between polls and returns? Who knows, but here is a list of suspects.

• Bad polls - perhaps the model being used to make sense of the responses is outdated or cannot reflect the dynamics of the current race (changes in youth turnout, demographic dynamics, geographic dynamics).
• Early Voting - In California a significant portion of the electorate voted before Super Tuesday and before Edwards dropped out offering Clinton a significant margin in these votes
• The “Who gives a ****” Effect - Some voters may not bother voting if they believe that the candidate they support is going to win anyway. Here the polling margins themselves create an effect on voters and lead to apathy on the part of supporters. Though this effect should be cancelled out by dispirited supporters of the other candidate it could be that someone who relies heavily on a group or groups that are disinclined to go out of their way to vote may suffer disproportionately (e.g. the youth/slacker vote).

We don’t know why the polls show outcomes differing from what the election returns report. To claim that there is some overarching “effect” is just more useless nonsense from the punditocracy.

Note: Politiporn is not sexist about the Reverse Dorian Gray thing. Another example of such a phenomenon can be observed with the official pictures of MSNBC’s Chris Mathews (who also appears to have had the spittle airbrushed from his lip).