I got into this with Noam Scheiber on the night of last week’s debate, and now I see David Frum making the same point:
bq. Political science proclaims, “debates don’t matter.” After this election, we may need to retire a lot of political science.
So that’s just wrong. In my piece about debates, I was careful to separate two things:
1) Whether debates move the polls enough to determine the winner of the election. This may have happened in 1960, although the polling data is thin and so any inference is tenuous. This may have happened in 2000 as well.
2) Whether the debates move the polls at all. They can! Not always, and not always by a lot, but they can. Look, I even posted the data.
Yes, political science finds a smaller role for presidential campaigns than reporting and commentary often suggests or implies. So our position gets caricatured as “campaigns don’t matter.” But that’s just a convenient straw man, and it’s not what the political science research says.