For those interested in yesterday's post about Nate Silver's blogging about early polling on presidential nominations as a good predictor, I have two recommendations to follow up. I supplied the logic for why Silver missed the boat on this one; Brendan Nyhan ran the numbers to demonstrate the same point quantitatively. But don't miss the comment political scientist Matt Jarvis left on my post yesterday (bottom two paragraphs; he's wrong about the baseball stuff).
Look, the bottom line here is we don't know. We don't know going in which way causality runs (do some candidates do well in the polls because other things are going well? Do some candidates acquire other valuable resources because they're doing well in the polls?). We suspect (strongly suspect?) that there are some important variables here that are difficult to observe having to do with internal party network signaling and conversations. We don't know going in whether to use 1972-2008, or 1980-2008, or some other cut at it; we don't know whether to consider Democrats and the GOP together or not.
Meanwhile, I'm aware that there are some political scientists who aren't big fans of Nate Silver, but I'm not one of them. So I'm happy to be able to send you over to his most recent post, on Barack Obama's recent rally-round-the-flag poll bounce. I'm not sure he's right about the long-term effects, but he makes an solid argument and, most important, gets the plausible magnitude basically right, which is 90% of the game as far as I'm concerned. But the really important part is the first half of the post, which is makes a very important point about causality that traditional reporters typically butcher; if every reporter would read and understand the pretty pictures and the accompanying text there, campaign coverage (and other reporting) would be dramatically improved.
It strikes me that, to some degree, academics and journalists have opposite natural biases when it comes to the interaction of events and trends in producing an election result.
ReplyDeleteGreatly simplifying, but journalists must have a natural bias toward the breaking story. It is their bread and butter, and also their mythology. ("Stop the presses! We're remaking Page One!")
Academics must have a similar bias toward underlying forces and the long term. Their bread and butter is papers meant to be cited, and their mythology centers on classics, works with durable relevance.
It is sort of fun (for certain values of 'fun') to watch this dialectic play out.