The “Bradley Effect” has been getting a lot of attention lately – most recently from folks who say it’s wrong because Tom Bradley lost the 1982 California Governor’s race for a lot of reasons other than race, and because the polling accurately predicted the outcome.
The real problem with the Bradley Effect is that it’s not about Tom Bradley at all. The "Bradley" effect doesn't matter - the effects of race on poll results do.
By focusing on the personification of the theory that the race of the pollster and the person being polled, and the race of the subject of the poll, matters rather than on the theory itself, the conversation has shifted to the Bradley campaign rather than the role race plays in surveys. The explanatory metaphor (in this case the 1982 Bradley campaign) has replaced that for which it is a metaphor (race effects in surveys). Instead of talking about what the Bradley Effect claimed to talk about, we’re now talking about Tom Bradley.
That the race of the interviewer matters in surveys has been discussed since at least 1971 when Howard Schuman and Jean Converse published an article in Public Opinion Quarterly titled "The Effect of Black and White Interviewers on Black Responses.” Race may matter differently now, and pollsters may have creative ways for compensating for the effect, but saying that the effect of race in polling has vanished because polls in the ’82 Bradley race didn’t account for early voting or a ballot initiative dealing with guns is silly. Analysis of the underlying issue has been replaced with an analysis of an election that took place 26 years ago. In semiotics jargon, we have replaced the signified with the signifier.
Unfortunately this sort of transference happens all the time. We find a noun or event that seems to capture a phenomenon, and then treat the noun or event as if it were the thing we were trying to describe. At some level this is inevitable, language cannot fully capture and convey anything. Language, to quote Kenneth Burke, is necessarily a selection, reflection and deflection of reality. But the connection can also be undone; the metaphor stands in for the concept, but it is not the concept itself. As such, we can return to the concept and craft new metaphors that do a better job at getting at what we think is the point.
Smart advocates know this, and select language that presents a view of the issue that helps advance their cause (the death penalty as fairness and innocence rather than evil killers; nuclear power as about energy independence and inexpensive power rather than as about massive environmental destruction). Smart advocates also know when to step away from their metaphor and refocus on the thing being discussed.
