Almost exactly a year ago, The Society Pages published my feature, “The Sociology of Silver,” about statistics, public discourse, and pop statistician Nate Silver. Around the time of the 2012 Presidential Election, Silver received a great deal of attention (and generated controversy in some parts) for the stunning accuracy of his predictive models. This week, he once again set the chattering class achatting for his big move away from The New York Times and to ESPN.
For those who have followed Nate Silver’s work closely, it was not terribly surprising to learn last week that he had decided to return to realm of sports where he began. If there was a surprise, it was that he willingly decided to step away from the prestige of working for The New York Times, the nation’s paper of record. But then, ESPN has its own merits including an increasingly sociological perspective in some of its programming (maybe I’ll say more about that in another post).
In the wake of his departure, many media watchers have been trying to describe his legacy at The Grey Lady. One common trope is that he challenged the values of political journalism and was sometimes resented for it. In a rather gossipy column, The Times’ Public Editor Margaret Sullivan writes, “[Silver] was, in a word, disruptive … A number of traditional and well-respected Times journalists disliked his work.” If political journalists depend on toss-up elections, narrated with daily rundowns of unimportant events magnified into “game-changers,” then Silver’s model, which showed a clear Obama win beginning in June 2012, made them look foolish. As Silver told NPR’s Morning Edition today, “At some point, I began to push back and to kind of launch a counter-critique of some of the conventions of horse-race journalism and punditry. I have less of a critique of traditional shoe leather journalism.”
Here, Silver, distinguishes among 1) data-driven journalism, 2) horse-race journalism, and 3) traditional shoe leather journalism (a.k.a. investigative journalism). To distill a bit, he’s saying that he’s smart, the shoe leather journos are hard workers, and the race watchers are insipid and bow to market demand for spectacle. Such a typology is partially about drawing status distinctions within the journalistic field, but it also recognizes a growing divide not so much between Nate and everybody else, but among the types of methodologies used by journalists. As I wrote last summer, the development of advanced statistics, widespread availability of fast computers and readily accessible data, and a smaller, but more technocratic marketplace of ideas paved the way for journalists to adopt data-driven methods. “Silver may have been the first to post Stata output on the New York Times’ web site, but if he hadn’t, someone else surely would have.” In his eulogy to Silver’s tenure at the Times, Ezra Klein gets it exactly right when he says that Silver wasn’t doing anything unique statistically. His main achievement is a journalistic one: “What Silver figured out was how to make data-driven election journalism into a daily product that could satisfy political obsessives.”
So, those (including Silver himself) who have said that he attacked the idea of horse race coverage of the election or emphasized data over narrative are dead wrong. Silver’s coverage was day-to-day coverage of ups and downs in a horse race where one horse was pretty consistently winning. Moreover, his coverage was desirable to a particular consumer base precisely because it animated data with narrative. In some ways, Silver embodies the very journalistic values he claims to critique. His primary innovation was to integrate a new type of methodology into the journalistic field and such “paradigm shifts” are not usually well received. But just watch as isomorphism takes hold. The New York Times will hire a new Nate Silver (Drew Linzer maybe?). The Washington Post will hire a stats guy for Ezra Klein’s wonkblog. And eventually, the methodological innovation becomes commonplace and no longer threatening.
As Silver noted on Morning Edition, the sportsworld largely accepted statistics as an essential analytical tool over a decade ago (See: Moneyball). If he fits in better at ESPN than the Times, it won’t be because his new colleagues are invulnerable to market demand or as enlightened as he about not falling into horse race rhetoric (hell, sometimes they cover actual horse races!), it will be because his methodology is more accepted.