big data comes to the social sciences

Gary King, director of Harvard’s Institute for Quantitative Social Science, has written a manifesto entitled Restructuring the Social Sciences. I have mixed feelings about it, but it’s a useful statement of influential trends in academia. King begins:

The social sciences are in the midst of an historic change, with large parts moving from the humanities to the sciences in terms of research style, infrastructural needs, data availability, empirical methods, substantive understanding, and the ability to make swift and dramatic progress.

King is highly enthusiastic about these trends, asserting that “the social sciences are undergoing a dramatic transformation from studying problems to solving them.” Solving problems certainly sounds like a good thing. One important reason is that social scientists are moving from statistical models based on samples (for instance, surveys) to the analysis of comprehensive datasets, such as all the job announcements posted in a set of newspapers over many years, or all the votes cast in the 2012 election. Social science thus merges with the kind of research conducted by firms like Google and Facebook, government agencies like the NSA, and political campaigns. Disciplinary boundaries are blurred, as some of the most interesting basic research on society now comes from computer science and business rather than the liberal arts.

In practical terms, King advocates the creation of centers like his own that can provide a shared infrastructure and a meeting place for diverse social scientists who use the new techniques. He claims that qualitative methods will retain an important role, because the masses of data that ethnographers and interviewers collect can also be mined by data analysts.

He suggests that centers for social science can become dramatically more efficient and effective if they apply their findings about organizational psychology to their own operations. For instance, they need lots of IT support, and they can provide that in ways that mimic the best-practices of IT firms. Finally, King would make a place for theorists, arguing that their insights can be helpful. “Moreover, theorists don’t cost anything! They require some seminars, maybe a pencil and pad, and some computer assistance.”

I am left with several questions:

  1. What does King mean by the humanities? He repeatedly describes the social sciences as moving away from the humanities, but what does he think they are leaving behind? Solo research? Unsystematic research? Unproductive research that doesn’t solve problems? (See my post on “What are the humanities? Basic points for non humanists” and also “Stop problematizing–say something“)
  2. How successful are these new techniques, really? In particular, are they generating new general knowledge and frameworks, or simply ad hoc answers to very particular problems? King cites a study that used massive data to demonstrate discrimination against people with stereotypically African American first names. I think that is an important finding. But does it tell us anything about the underlying reasons for racial prejudice or general strategies that we might use to defeat it? (Cf. “Bent Flyvbjerg’s radical alternative to applied social science” and my “critique of expertise, part 1″)
  3. What are the ethical pitfalls of increasing our power to track, predict, and influence human behavior? To put it another way, if the social sciences move from studying problems to solving them, are the “solutions” ethically acceptable in terms of their means, their ends, and the ways that they engage the affected populations? (See my “qualms about Behavioral Economics” and “the new manipulative politics: behavioral economics, microtargeting, and the choice confronting Organizing for Action.”) This, of course, is why the humanities remain so important in an era of big data.

About Peter

Associate Dean for Research and the Lincoln Filene Professor of Citizenship and Public Affairs at Tufts University's Tisch College of Civic Life. Concerned about civic education, civic engagement, and democratic reform in the United States and elsewhere.
This entry was posted in academia, civic theory. Bookmark the permalink.