My Email to Charles Lane

Jun 6 '12

I wrote Charles Lane a brief note in response to his piece and he was kind enough to respond.  In responding to him further, this is what I wrote (adapted slightly for the blog format).  Lane had observed to me that his critics had a material interest in this debate and that illustrates his point about the challenges of remaining objective and the perils of politicization (I’m paraphrasing).  So my response proceeds from there.

On the question of politicization.  Yes, it’s true that when the government funds any program, that program becomes politicized, in the sense that there’s political debate about it.  In that context, beneficiaries of the program will usually defend it.  That strikes me as politics-as-usual.  Political scientists have to engage in this like anybody else.  As one of my friends who studies lobbying once said, “If you’re not at the table, you’re on the menu.”

The issue I thought Lane raised regarding politicization was whether the political beliefs of social scientists somehow contaminated their research.  As Seth Masket pointed out, one’s values might drive your research questions or maybe inform a hypothesis, but any reputable scientist will let the data speak.  If they don’t confirm the hypothesis, you report this and then think harder.  In my experience, social scientists are as honest about this as any other scientist.  Lane tweeted the story of the psychologist Diederik Stapel’s falsifications and wrote “read this and tell me whether psych is just like physics.”  So does falsification happen in physics?  The answer is yes. In medicine too — where double-blind randomized controlled trials are the gold standard — the same thing happens.  See also this piece in Nature.

This speaks to the broader point: I’m not sure the distinctions Lane draws between the social sciences and “hard” sciences — in terms of methodology, values, etc. — are as sharp as he suggests.

One other question that I’d ask is: isn’t there some value — scientific, for policy, etc. — in research that isn’t experimental?  Take a subject not particularly amenable to randomization, like civil war.  And take this NSF-funded piece of research, which appeared in the flagship political science journal.  Political scientists James Fearon and David Laitin find that civil wars are not more likely in countries that are ethnically heterogeneous or autocratic, but are more likely in countries that have mountainous terrain and low levels of economic development.  Some of those findings are not intuitive.  All of them are very useful to know if, for example, you work for the National Security Council.  This is one reason why the Defense Department and others have doubled down on funding for the social sciences.  Here is a longer discussion of federal funding of the social sciences and national security.

Another question concerns the support of the NSF for large-scale social science survey projects, such as the American National Election Study and the General Social Survey, which now span 60 and 40 years, respectively.  These projects do include experiments within them — e.g., splitting the sample at random and asking different versions of a question — but that is not their primary purpose.  Their primary purpose is to provide much more detailed information about public opinion, with significantly higher response rates, than can be obtained in a traditional phone poll.  And, obviously, having Americans’ answers to questions asked the same way over a 60-year span of time is rare.

A final question concerns the NSF’s extensive funding of social science education — e.g., via research projects that employ undergraduate and graduate students as research assistants, or via dissertation grants to graduate students.  Is that a valuable investment?  Do we want to train good social scientists?

As I hope is clear, I believe that social science has no inviolable right to federal funding.  Instead, I think it important to show that there is public value to the work that we do, and thus that the government’s investment in our research is worthwhile.  Political scientists or other social scientists can undoubtedly do that better.  I hope our critics are willing to listen to the defense, even if our motives are, in part, self-interested.