Earlier today, over at New Republic, Leon Wieseltier wrote, The Emptiness of Data Journalism. It is yet another attack on Nate Silver for his comments about the superiority of what he does versus what opinion journalists do. And it is by far the best that I’ve read. I’ve always liked Silver and I was totally in agreement when he attacked the insider-based campaign coverage. It was true that if you want to know how people are going to vote, the best thing to do is to ask them. Asking the campaigns might make for entertaining reading but it doesn’t provide useful information. But now Silver is claiming that opinion journalism writ large is similarly useless. Wieseltier calls him on this.
Paul Krugman also called out Silver—in two posts. The latter is more interesting, Further Thoughts on Hedgehogs and Foxes. In that, he noted that data alone don’t explain anything. They must be applied in a context with a model of what’s going on. We don’t, for example, care about the inflation rate because any inflation rate is good or bad. In an under-performing economy, higher inflation is usually a good thing. In a booming economy, higher inflation is usually a bad thing. What’s more, different levels of inflation affect people differently. The data alone mean nothing and I dare say someone like Silver is likely to apply unstated assumptions about such data.
What I was most struck with in Silver’s interview was his description of opinion columnists, “It’s people who have very strong ideological priors, is the fancy way to put it, that are governing their thinking.” The implication is that Silver is not burdened by such “ideological priors.” I know from experience—and this is one of the biggest problems with American journalism—that anyone claiming to be objective should be watched closely. What such claims mean is, “I’m totally unaware of the huge amount of baggage that I bring to everything I do.”
Krugman linked to an amazing article from a year and a half ago by climate scientist Michael Man, FiveThirtyEight: The Number of Things Nate Silver Gets Wrong About Climate Change. It’s a relatively long article and very much worth the time. But the basics are that Nate Silver may have just looked at the data, but his analysis was way off because he did what supposedly non-ideological centrists always do: he figured that the bad global change work by a marketing professor (You read that right!) was equal and opposite to the work of thousands of climate scientists of the past three decades.
I’ve been wondering what the new Five Thirty Eight was going to do. I understand statistical analysis of sports. And I understand political polling. But what about the rest? And if Silver really thinks he can get good data on important policy issues without understanding the subjects, he’s a far sillier man than I had thought. Wieseltier may have it right in his conclusion:
I’m afraid that Wieseltier nailed it.
Update (19 March 2014 3:29 pm)
I never said that I was original. Ryan Cooper wrote, Nate Silver’s FiveThirtyEight and the Dangers of Being Ideologically Neutral. He said very much what I said: