Thanks for having me, and a quick thought about Nate Silver’s effect on Washington chatter
posted by David Schleicher
I want thank Danielle and all the other great people here for having me. I apologize for the late start and it may be a few more days before I really get going. But once I do, I’ll be blogging about a few election law and election related things, some local government law and land use things and probably a few random thoughts.
But I want to start by talking about something truly important: Washington gossip. One of the great joys of being an election junkie used to involve collecting political information. You could monitor local newspaper to find poorly reported on polling data, you could constantly check Charlie Cook’s and Larry Sabato’s predictions (Is NY-13 in play? MA-06? Wow!). Or, even better, if you were enterprising enough, you could sometimes score unreleased newspaper or campaign insider polling. We scholars had another advantage; we could knowledgably discuss models that showed the importance of economic factors in Presidential elections and the uselessness of polls early in election cycles, revealing our superior remove from the vagaries of the news cycle. All of this would allow you to have something to add when conversation turned to the campaign, a necessary survival tool for living through election season in D.C. People here talk about elections like their jobs depend on it, and this type of insider chatter played an important role. As partisan rhetoric heated up, predictions and related gossip was a topic about Democrats and Republican could talk about safely over cocktails, as you could be analytical without actually having to argue.
This, though, has been totally ruined by Nate Silver. (To be fair, there are few others who contribute, like the excellent Nate Cohn and a few of the great new political science blogs like The Monkey Cage and Mischiefs of Faction.) While he’s been in the news a great deal since he broke onto the scene in 2008, the innovations in Silver’s modeling techniques this cycle really deserve notice. Particularly, he has successfully and innovatively incorporated economic fundamentals into his poll-of-polls modeling techniques, drawing on a long political science literature as a method for improving the model’s predictive force early in the cycle and reducing the sensitivity of the model to short-run spikes in opinion or sampling error. The result? His model is better than your intuitions and, in aggregate, probably better than even carefully sourced insider information. Particularly if you pair it with the prediction markets and Vegas odds, there’s not much new one can add to these conversations any more, as everyone reads Nate Silver, and he provides a clear, easy to understand number reflecting the best information about each candidate’s likelihood of winning. Want to sound off about the effect of a convention speech on the public? He’s got the data on when these things are bounces and when they are real and how you can tell the difference. Want to talk about debates? He’s got you covered. Want to know much to freak out? He let’s you know – you should freak out exactly 7% more today than yesterday. Even on Twitter, where overreaction to each drip and drab of polling data is the expectation, the authority with which Silver’s analysis speaks makes the early responses to polls look extremely small, even when they come from leading journalists and political insiders. Of course, you can always can note that everyone else is overreacting, but that’s nothing more than a pose, and no fun besides.
While many people have written about Silver since 2008 – check out this great recent New York Magazine profile — almost no one has noted how he has single-handedly ruined political conversations in D.C. In a post-Five Thirty Eight world, the only way to have one up in an election-based conversation is if you’ve checked the site since the last time he updated it and the person you’re talking to hasn’t. Outside of that, if you’re still talking about this stuff, you’re almost certainly doing something wrong.
The effects of the death of this type of chatter are only beginning to be felt. It might lead to, I don’t know, substantive conversations about the issues. And who knows where that could end up?