David Moore Commented on A New Version of the Vote Choice Question by David Moore

2008-10-20T18:39:10Z

The problem with this argument (by syvanen) is that it denies the ostensible role of polling - which is presumably to reveal what the voters are thinking. To say that no polls reveal the "truth" and that instead they are used only for their "predictive power" suggests polls can't tell us what the electorate is thinking now. That's certainly not what pollsters claim, nor how pundits and politicians treat the polls. Moreover, pollsters explicitly reject the notion that polls today are "predictive" of what might happen on Election Day, claiming instead that polls are a "snapshot in time." My argument is that the snapshot they take is out of focus, deliberately so, as pollsters refuse to acknowledge the undecided electorate.

Obscuring the undecided vote has implications for campaigns, giving false impressions about which candidates are leading and by how much -- thus influencing fund-raising and media coverage.

When Gallup started this form of questioning 73 years ago, no one really cared what the polls said. Today, polls drive news coverage and profoundly shape the political environment. I believe, therefore, that they should be designed to give as accurate a picture as possible.

(By the way, my vote choice question is no more difficult to ask than the standard question -- the follow-up questions are optional. I could make the argument, in fact, that my basic question is actually simpler for the voter, because it isn't hypothetical but real. Who do you expect to vote for in November (name candidates), or haven't you decided yet? That's the way we talk with our friends. I don't hear people asking for the false hypothetical, "But I don't want to know who you will vote for in November, but how you would vote if the election were held today!")

David Moore's Blog

David Moore Commented on A New Version of the Vote Choice Question by David Moore