Matt Yglesias is kind enough to promote political science:
One feature of the American political media that I’ve oft had occasion to the lament is the lack of influence by the field of political science. It’s generally taken for granted that some familiarity with economists’ research is relevant to writing about economic issues, but people seem very comfortable making broad, sweeping assertions about the American political system that are totally uninformed by research into it. It’s true that political science isn’t really science like physics that’s going to definitively answer every question you might have, but empirical and theoretical inquiry by political scientists can and does shed a lot of light on a lot of important issues.
But this compliment still has a back-handed flavor, as he repeats this often-heard analogy to physics, which can “definitively answer every question” while political science (and, presumably, economics, sociology, etc.) can only “shed light.” I’ve always found this a strange conception of physics or any other “hard” science, for that matter. It doesn’t take long to think of many important questions that physicists haven’t answered. Or you can turn to medicine and think about how little we still understand about the human brain, despite initiatives like this one. Etc., etc.
It’s not that I don’t appreciate this compliment. And Yglesias himself is particularly alert to ideas and findings in political science. But I think a more accurate depiction of “the sciences” wouldn’t establish some hierarchy with physics at the top and the social sciences further down.
{ 8 comments }
A postulate: The more familiar non-scientists are with a topic of scientific study, the less credence they will give the scientific study of that topic, because they will be more familiar with outliers and counter-examples to which they will give (because of their familiarity) disproportional weight. Conversely, the less comprehensible scientific claims are on their face — and the less individuals can compare those claims to their regular experiences — the more likely they are to accept them as ‘scientific.’
In other words, basic familiarity with the political system means many people think they know how politics work, and that scientific investigations of these mechanisms either cannot provide precise answers(Yglesias) or are not worth investing in (Coburn). When scientific claims are made that have no easy reference to normal experience, people are least aware of outliers and cases that contradict these claims and therefore assume greater ‘scientific’ validity.
So what are political scientists to do? One approach might be to promote political science that contradicts conventional wisdom. While many claims that are counter intuitive are also wrong, political science has produced a host of empirical findings that have contradicted common preconceptions:
– Campaign contributions don’t buy floor votes (Ansolabehere, de Figueiredo and Snyder)
– Voter turnout has risen, not fallen, once one looks at the right common demonminator. (McDonald and Popkin)
– Partisanship has become more, not less important in structuring voter behavior over the past decades (Bartels)
– Rich states vote Democrat but rich people vote Republican (Gelman et al)
– Cable tv entertainment choices have contributed more to the polarization of the electorate than choices over cable news programming (Prior)
– TV campaign ads can be effective at changing voter preferences … but only for a few days. (Gerber, Gimpel, Green and Shaw)
And that’s just a few off the top of my head. Surely there’s enough valid but ‘counter-intuitive’ political science to produce a ‘FreakPoliticalScience” type book for a lay audience. Such a book might be useful at pushing back at the notion that political science (1) isn’t scientific and (2) doesn’t tell us anything new.
correction: looks at the right *denominator*
John – oh, come on.
Of course physics is a different type of science that the social sciences. I don’t think we have any finding in the social sciences whose predictions we trust enough to bet our live on them (even if we don’t have to).
E.g. my mother, a physics teacher, in her class on the law of conservation of energy used to stand against a wall, hold a 5 pound pendulum against her forehead and let it go – standing perfectly still as the pendulum swung back, usually up to less than an inch of her face.
Once again – any polisci finding that you’d trust in a similar way?
I think what we have to fight against is the idea that this means polisci doesn’t have any real findings or isn’t a real science. We can point out that it’s incredibly hard to do (because of the complexity of the system – physicists can’t deal with either the weather or a drop of cream in a coffee cup very well either), and that it’s interactive – there is no chance that the pendulum may have read your paper and decides to swing a little furter – but we do no, for example, that campaign managers are aware of a lot of the political science research on campaigns (as an example).
Kevin: Thanks for your comment. I think that problems arise in this debate because people have different working definitions of what constitutes “useful” science. People like Tom Coburn think it needs some clear practical application. Yglesias wants definitive answers. Others want Freakonomics-y counterintuition. I am a bit ambivalent about going the counterintuition route. It makes for good publicity but does it make for good science? We still need to do the spadework necessary to prove the intuitions.
Sebastian: I didn’t say that physics and political science were identical. But I think Yglesias’ distinction downplays how much of the physical world doesn’t led itself to definitive answers. You make this point yourself: a lot of physics doesn’t involve something as simple as a pendulum — hence the “complexity of the system.” This is why I don’t think it’s that useful to set up physics as the One True Science while political science, sociology, economics, psychology, et al. trudge along behind like ugly stepchildren.
And are “campaign managers aware of a lot of political science research”? They certainly are not. Campaign management is at best a dollop of something approximating social science, combined with oceanfuls of intuition and shamanism.
To clarify, my suggestion is not that counter-intuitive political science are the most important works in the field, but rather that by publicizing this subset of political science research, we can challenge the oft-held beliefs of non-political scientists that the popular perceptions of political processes are undeniably correct and that the scientific study of those processes have little to add.
So no, I’m not suggesting that counter-intuitivity makes for better or worse science. Rather, I’m suggesting that it makes for exactly the kind of publicity that is needed.
Kevin: If publicity is what political science wants (or needs), then you’re most likely correct. Incidentally, there have been books that have tried to do something like this — Kathleen Hall Jamieson’s Everything You Know about Politics is Wrong and, more recently, Kauffman, Petrocik, and Shaw’s Unconventional Wisdom.
Part of the problem here is that everybody and their pet mouse thinks that we are all using the same definition for science when we banter about it. If you ask what is science on this website:
http://www.askphilosophers.org/ you’ll get much disagreement and maybe some agreement that it’s not that useful of a line of demarcation when used on its own.
E. O. Wilson’s “Consilience: The Unity of Knowledge” (1998) in Chapter 9 “The Social Sciences” starts:
“People expect from the social sciences-anthropology, sociology, economics, and political science-the knowledge to understand their lives and control their future. They want the power to predict, not the preordained unfolding of events, which does not exist, but what will happen if society selects one course of action over another.
“Political life and the economy are already pivoted upon the presumed existence of such a predictive capacity. The social sciences are striving to achieve it, and to do so largely without linkage to the natural sciences. How well are they doing on their own? Not very well, considering their track record in comparison with the resources placed at their command.” (Chapter 9 begins at page 197 and ends page 228. It is a good read. Query whether there has been improvement since 1998?)
Comments on this entry are closed.