Legal Commentary

Just after the New Hampshire Democratic primary in January, there was some discussion on the legal blogs of the Bradley Effect, when Hillary Clinton won despite exit polls indicating a sizeable lead for Barack Obama. The Bradley Effect (also called the "Wilder Effect") refers to people's reluctance to admit their own biases on issues of race and gender. It is the phenomenon of people answering polling questions about their votes or preferences one way to show public open-mindedness and liberality on issues of race, while actual voting behavior in the privacy of the booth is quite different, in part based on some racial bias. The principle is named for former L.A. Mayor Tom Bradley, who lost the 1982 election despite exit polls showing him with a big lead, and former Virginia Governor L. Douglas Wilder, who won what turned out to be a close race despite exit polls showing him with a big lead.

But could we generalize the principle to the notion that people do not want to admit to holding what the question suggests is the "negative" or "bad" position on some issue? Everyone wants to look open-minded, honest, fair, and to favor "the good" and will not admit to an illiberal or bad view. So they will be less than accurate in responding to most polls.

I wonder if this generalized notion explains polls and behavior with respect to negative campaigning and negative campaign ads. Now, let me add that the term "going negative" has become so broad as to be meaningless, used by the media to mean everything from personal character attacks (think Swift Boat) to baiting or demagogic mischaracterizations that unfairly tag the candidate with something (think Willie Horton) to ads that, legitimately, point out and criticize the failures of an incumbent's policies or the unwisdom of a candidate's votes and proposals. Too often, anything beyond "vote for me because I will do X and X is good" is deemed "negative." In any event, polls strongly show that people do not like negative ads--and certainly not ads that are "negative" in the narrowest sense of getting blatantly personal and defamatory. Yet candidates continue to use them, as do advocacy groups. And they use them because they work--they do not turn voters off to the candidate or to the side running the ad (as polls might suggest). Rather, they seem to affect how voters vote and view the candidate who was attacked.

So could it be that something like a non-racial, generalized Bradley/Wilder Effect is at work? People are not going to admit to liking negative ads--that would make them seem illiberal or like they favor something that is unpopular and widely seen as a "bad" thing. So they tell pollsters that they do not like negative ads. But their voting and other behavior suggests they genuinely respond to such ads. And thus such ads continue.