Friday, October 03, 2008

On the Debate(s)

I watched last night's debates with some interest. For the Canadian debate I participated in a great community event with a discussion of the debate at the end. The discussion eventually turned to whether the Green Party should actively throw its support behind the Dion Liberals. It was a great discussion about the moral legitimacy and/or imperative of strategic voting. It's a more muddled question for me now than it was before the conversation, which I take as a sign of a great discussion.

As for the debates themselves, I am not keen to pick a winner or a loser, because I think we all see different things. And it's quite possible for every leader to do well among their respective constituencies and thus discussion of who won and who lost is rather fruitless. Indeed, the most significant recent research on debates suggests that debates do just this. To review the findings of Blais and Perella (two colleagues and friends) consider this abstract:

Almost an entire generation of election survey data was pooled together from the United States and Canada to assess the systemic effects of televised debates. Four questions were posed: (1) Is there a general tendency for evaluations of candidates to improve or deteriorate after a debate? (2) Do evaluations of one candidate negatively correlate with changes in evaluations of opponents? (3) Do debates disadvantage incumbents? (4) Do debates advantage less popular candidates? Using "feeling thermometer" items to measure voter evaluations, four patterns are revealed. First, candidates generally gain points.The supposed mudslinging that characterizes a debate appears not to feed into any notion of cynicism. Instead, voters appear to gain an appreciation for the debaters. Second, a candidate's gain is not earned at the expense of those deemed to have "lost" the match. Third, a debate does not disadvantage an incumbent. A candidate with a record to defend stands about as much chance of benefiting from a debate as a challenger.And fourth, any evaluation gaps before a debate become narrower following a debate. This final effect, which is particularly true of American presidential debates, may reflect a debate's ability to raise awareness of less popular candidates.

Think about that before you prognosticate on who "won" the debate.

No comments: