washington, dc

The Democratic Strategist

Political Strategy for a Permanent Democratic Majority

Race and “Effects”

Well, the cat’s out of the bag. As both Noam Scheiber and Chris Bowers have written about today, it no longer much matters whether a Bradley/Wilder Effect tilted NH to Clinton; discussion of that possibility in NH and beyond has taken on a life of its own. Scheiber calls it the “Bradley Effect Effect.”
In case you’re just tuning in, the Bradley Effect (or Bradley/Wilder Effect) is an insider term for the phenomenon of voters telling pollsters they’ll vote for an African-American candidate, and then pulling the lever against that candidate in the privacy of the voting booth. In the absence of any other completely convincing explanation of why all the polls in NH were wrong, it’s become a very popular theory, for obvious reasons. And that’s particularly true among African-American political observers, for equally obvious reasons having to do with the last few centuries of world and American history.
If you think about it, there are three different racially-motivated “Bradley Effects” that could theoretically have been in play in NH, and could be in play down the road.
There’s the classic “Bradley Effect” of voters lying about their preferences and then indulging racist impulses in the privacy of the voting booth (which is why, presumably, it didn’t happen to Obama in IA, where there’s no voting booth and no privacy). There’s the second-order “Bradley Effect” of lying to pollsters out of fear of being perceived as racist, followed by a more honest actual vote. And then, particularly in a Democratic primary (unlike the actual Bradley and Wilder elections), there’s a third-order “Bradley Effect” of a vote cast out of fear of other voters’ racism–i.e., concerns about an African-American candidate’s electibility (though this theory is undermined by the NH exit polls in which a lot of HRC voters deemed Obama the most electible candidate). This last “effect,” of course, has been widely reported as prevelant among African-Americans prior to IA, and particularly in SC, where for a long time Clinton was running ahead of or even with Obama among black voters.
We could now be about to witness a fourth-order “Bradley Effect,” if African-American voters in SC and elsewhere react to discussion of this issue by uniting behind Obama to counter-act perceptions of white voter semi-secret racism.
It’s all pretty complicated, eh? Maybe it’s good to get the race issue on the table and deal with it now rather than later; it’s inevitably going to be a factor in the decisive phases of the nomination contest, and if Obama wins, in the general election as well. But it is a little ironic that it’s come to the fore at this moment. After all, an African-American presidential candidate has just finished first and then a close second in two of the whitest states in the country. Thanks to the expectations game going into the second of these exceptionally honkified electorates, we’re having to face race straight up, and right away.

3 comments on “Race and “Effects”

  1. edkilgore on

    First-order means secret racists who don’t want to admit it. Second-order means non-racists who don’t want to appear racist by admitting preference for white candidate. In practice, there’s not much difference.
    I agree that the first-order effect doesn’t make much sense, but that’s the way the whole phenomenon has been interpreted in the past, and why it’s so explosive.
    Thanks for the comment.
    Ed Kilgore

  2. Joseph Clarke on

    Just to be clear, what is the distinction between “first-order” and “secord-order” Bradley Effect, in your argument? They sound the same to me.
    For my part, I don’t see why a secretly racist Democrat would feel pressured to lie to pollsters in a Democratic primary, in which, obviously, there are plenty of plausible non-race-related reasons for a Democrat to support a white candidate.
    But you’re right to identify various levels of meta-Bradley Effect at work in this campaign.

  3. nfreeman on

    I’m skeptical of the Bradley/Wilder effect’s impact on the NH primary. Most of the polls I saw over the weekend and on Monday had Obama at around 38%-39% and Clinton at about 30%. In the end, Obama got 36% and Clinton got 39%.
    The 2-3% drop for Obama doesn’t seem like a really huge Bradley effect, especially when one considers that the margin of error in most polls is 4%-5%. Moreover, even if the Bradley effect did happen in this race, it only explains a fraction of Clinton’s 9% jump.
    If anyone has seen a polls with different numbers than these, please say so, but without any more supporting evidence, it seems farfetched to me.


Leave a Reply

Your email address will not be published. Required fields are marked *

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.