On Conservatives’ Ambivalence toward Social Science

If you spend enough time in conservative intellectual circles (as I have) you’ll eventually hear someone complain about “social science.” It would be a useful exercise to taxonomize the different versions of this complaint. However, my suspicion is that the source of most of these worries is the idea that the social sciences involve the “pretense of knowledge,” the title of F.A. Hayek’s Nobel Prize address. In brief, Hayek argues that social science faces significant limits in developing predictive explanations of human behavior. In its weaker forms, this version doesn’t say where these fundamental limits to scientific explanation are — only that they exist. In a stronger form, it makes more specific claims about where scientific understanding is likely to fall short. As a specific example, the economist Timur Kuran argues that the phenomenon of “preference falsification” (the tendency of people’s publicly expressed preferences to diverge from their privately held ones) limits the predictive power of social science.

The “pretense of knowledge” view undergirds several distinct versions of complaints about social science:

  1. Perverse Unintended Consequences. Because the object of social scientific analysis is complex and analysis is always predicated on a simplified model of society, policy informed by social science are likely to have unintended consequences. This is what Albert Hirschman calls the “perversity thesis”: social change is likely to produce perverse effects that undermine or entirely negate the value of the change.
  2. The Futility of Social Engineering. Another thesis Hirschman identifies is the “futility thesis”: Because society is a complex whole that cannot be fully explained by social scientific efforts, policy informed by social science will often be ineffective. As the joke from the Cold War goes, “Under capitalism, man exploits man. Under socialism, it is the other way around.”
  3. Social Science as Ideology. Because social science imitates the methods and forms of the natural sciences, it confers the appearance of objectivity and impartiality to its findings. This feature allows political actors to dress up their preferred political outcomes as scientifically demonstrable. Sometimes this complaint also includes the claim that social science is “anti-democratic” because it functions to exclude ordinary citizens from political discussion.

The point I want to make here is that there is a basic tension in conservatives’ complaints about social science which this three-part distinction (I hope) makes clear. Note that all three views rely on particular factual claims about human societies. According to Futility, society exists in a stable equilibrium that resists change. On Perversity, society is a complex, interdependent whole such that changes to one part unexpectedly affect other parts. Ideology is a complaint about the social discourse social scientific research creates.

A natural question for in light of these factual premises in different versions of the conservative’s complaint is whether the premises are true. How would we know? It seems that an important way to answer that question would be… social science. Now granted, perhaps we would need different methods or research practices from the ones currently in use. But arguments and competition between different research programs is a standard part of any scientific field. If the complainers deny that social scientific methods are appropriate to assess their own complaints about the ubiquity of unintended consequences, for example, they owe us an explanation of why we should accept their factual premise.

As a more casual observation, it seems that conservatives are entirely apt to make social scientific arguments about anti-poverty programs or the welfare state, for example. But why not bring the same concerns about social science toward this research? If the conservative’s objection is to the possibility of social science as such, then they should treat ideologically favorable results with the same skepticism. A natural reply is that conservatives are against “junk science” and bad research, not carefully reasoned and methodologically sound work. But in that case the conservative might simply be drawing attention to the need for higher-quality research rather than the possibility of social science guiding public policy. And that naturally invites the question of how to improve both the discourse around social scientific research and how to bring high-quality research to bear on public policy.

Or consider the complaint that the discourse around social science is “anti-democratic” because it excludes people who have other, non-scientific contributions to contribute to political discussion. Consider this excerpt from a recent interview with National Review’s Michael Brendan Dougherty:

“Political commentary has shifted in this direction where you have to know social science, inside and out, to even begin opining — which is itself a kind of anti-democratic spirit, be that as it may…. it’s unhealthy for democracy, that’s fundamentally where my contempt for this type of talk colonizing all of politics comes from.”

What quickly follows this complaint, however, is reassurance that the complainer is not somehow “against science” or against the idea that policy shouldn’t be made using the best available public knowledge:

“Now of course we should try to be informed by the best social science that’s out there, and someone like [labor economist] George Borjas is actually admirably frank about what can be known from the type of work he does, and I think he’s one of the best ones. But I find the “studies show” discourse to be anti-democratic. It’s a kind of priest-craft that forbids the common person from saying what they want to say.”

This raises the question, again, of whether the complaint is with the quality of “bad” social science, and whether the expectation that people inform themselves of the best social science before contributing to political discussion is anti-democratic and objectionable. My suspicion is that the objection is really to the quality of research, not an objection to some “anti-democratic” property of social science itself. But because Dougherty equivocates between social science and the discourse of “social science,” this obvious question goes unanswered.