Nate Silver is Very Good at Something that is Not Terribly Important
[Update: I follow up on this post here, with a bit more about what I mean by ‘policy’.]
There is an enormous brouhaha going on at the moment concerning the innumerate attacks on Nate Silver for relying on mathematics to make probabilistic claims about the state of the election rather than relying on insider knowledge– i.e. asking campaign insiders (who are doubly interested parties, both because they want their side to win and because they need to justify their fees) what they think the state of the race is. The absurdity of this has already been well covered in many places. I don’t have much to add, other than to suggest that when someone is doing something that seems completely unmoored from their supposed purposes the problem might not be in the means chosen but in the validity of the justifications. Telling us that they are having a different conversation about elite concerns would be unspeakable, but this perspective makes the activities of elite pundits sensible in a way that taking them at their word that they are actually trying to describe democratic politics does not. (It’s also worth noting that these same pundits get an enormous amount of undeserved attention from many of the same people who are mocking them now – giving them less attention is the only way to solve that. They will not reform themselves.)
But I want to make a different point. It’s not obvious to me why anyone (outside of political campaigns or those who are using their campaign contributions for purely instrumental ends) should spend much time trying to figure out who is going to win. Our role is supposed to be to make a choice. Journalism, if it was functioning properly, should help us make informed choices, and not just for one race, but all the races top to bottom. I’d go further and suggest it ought to help us hold politicians accountable and push our concerns onto the public agenda, which involves a somewhat more active view of citizenship and democracy than simply making an informed choice between two candidates. Knowing the odds than one candidate will prevail contributes exactly zero to these tasks.
A possible, partial objection might be that while these things are the most important, it’s still essential to know the state of the race. This can be used to mobilize people, who might be more inclined to vote if they think the outcome is close. While this may be true, it also cuts the other way – knowing the likely outcome could lead people not to get involved, and if it’s not close that would presumably be the most rational course. Besides this, that something can be used to some end doesn’t mean it should. You can keep warm by burning books but it would be better to read them and burn wood. What people should use to mobilize voters is policy – policies already enacted, and those proposed. But doing this required actually listening to voters and taking their concerns into account. It requires going beyond the question of support or opposition to a set of elite chosen policies (and frames) and instead tapping into regular people’s top concerns. That sort of election would look very different. But it is possible. as people like Bernie Sanders has shown.
So praise Silver for doing what he does well, and mock those like Joe Scarborough for doing the same thing poorly. But maybe we could turn just a but of our attention on polls and campaign events to supporting active citizenship.