Incoherent Criticisms of Political Science
On the one hand Stevens claims that political science is not too good at predicting. That is true, but neither new nor fatal as a criticism. On that point she might have read this Op Ed by my University of Rochester colleagues Kevin Clarke and David Primo, also published in The Times, a month or so ago. As they make plain (both there and in the paper and book that Stevens also seems to have overlooked) models are not terribly useful for making empirical predictions. So what? As Stevens herself intimates, we might judge performance by ability to explain rather than ability to predict. (This move does not deflate the claim of political scientists to be doing science. After all, are evolutionary biologists great predictors? How good are geologists who study plate tectonics at predicting specific earthquakes?) In other words, Stevens takes the wind out of her own sails.
On the other hand, Stevens seems to be unhappy that political scientists take money from the government at all. Except that she really is not. If work by her preferred sort of political scientist were funded by some government agency, the money, it seems, would somehow miraculously be sanitized. Research "especially [by] those who use history and theory to explain shifting political contexts" to "challenge our intuitions and help us see beyond daily newspaper headlines" merits governmental support. I admit that I really do not know what that means.
Stevens, of course, mostly objects to military funding of social scientific research. I share her skepticism. But where precisely does she propose to draw the line? At what point does the source of research funding become acceptable? By what criteria? Public, but not military? Are only non-governmental sources acceptable? The latter restriction, it seems, is too much even for Stevens. That is a good thing. Private right wing foundations - Olin, Smith Richardson, etc. - have funded all sorts of social science research. Do they pass muster? The endowment at the private, liberal Russell Sage Foundation (for whose generosity I myself am grateful) originated, I believe, from the fortune of a conservative railroad tycoon.* Does research funded there pass muster? How about if the research in question is grounded in quantitative methods and focused on matters of political inequality?
Finally, Stevens offers a truly astounding remedy for the dire situation she sketches. We should use a lottery system to discriminate between the sort of "contrived data sets" about which she complains and the "thorough demographic, political and economic data collection" she thinks the government should fund. That proposal is breathtaking in its shortsightedness. What we lack is judgement and discrimination. And we should use randomization to remedy that shortcoming?
I will stop there.
* Foundations, after all, are largely institutionalized ways of dodging taxes. So, if we are concerned, as Stevens intimates we should be, with fairness and equality ...
P.S.: A comment pointed me in the direction of this blog post in which Stevens follows up on the Op Ed.