Poll-Tested Questioners

John Judis makes several interesting points about the oddity of selecting a debate audience composed of people who (a) tell pollsters they’re thus-far “uncommitted” and (b) are demographically representative of the country:

I remember having similar questions about the rules that limited juries in infamous and well-known cases to people who hadn’t even heard of the cases and had no opinion about them. Wouldn’t the result be that you might limit the pool of questioners – or jurists – to the less informed parts of the population, or to the more quirky and less representative. OK, suppose that 13 percent of the questioners are African-Americans, which would roughly fit the population. Where are the debate chieftains going to come up with genuinely uncommitted African American voters?

Along these lines, given Obama’s lead in the polls a group of uncommitted voters is going to be a disproportionately right-of-center group of people.