Phil Arena professes to be puzzled (or is he offended?) by Dani Rodrik's claim that rational choice political economy is "paradoxical." Rodrik has discovered apparently that our belief in our ability to use knowledge to alter policy outcomes is inversely proportional to our ability to explain policy outcomes. Phil asks, "where exactly is the paradox" in this?
I'll give Phil the benefit of the doubt and assume that he realizes that the paradox for Rodrik lies in the unstated assumption that led him (and other economists) to political economy in the first place: the belief that through science we gain knowledge that allows us to control our world in order to produce better outcomes. This seems true in physics and even (in theory) economics (e.g., we did send men to the moon, and "we are all Keynesians now"). Rodrik believes this is untrue in political economy, however, where the entire enterprise is devoted to endogenizing policy choice.
I understand why Phil analogizes Rodrik's confession to a conversion to mysticism. I even pretty much agree with Phil that the apostasy is rather odd coming from a person of Rodrik's stature. So, yeah, you had me at hello. But don't understand the subsequent effort to defend political economy against Rodrik's major point. In brief, Arena argues that ideas do too matter, that people do so change the world, and that scholarly inquiry does so have an impact. Phil seems to believe that this implies something about Rodrik's argument, but it seemed to me completely unresponsive to Rodrik's argument. So, I though it might be useful to outline where Phil's criticism of Rodrik left me a bit confused.
First, the issue isn't whether political economists think ideas matter . The issue is whether ideas are independent of the processes we model and thus can be manipulated by outsiders to any greater extent than the actions of the people whose behavior Rodrik hoped to change but now realizes are responsive to vested interests. Rodrik seems to believe that ideas can be manipulated by outsiders to produce better outcomes. Hence, he seems to believe that ideas are outside of the political economy processes that otherwise shapes behavior and outcomes. One might wonder if Rodrik is about to repeat the quest that he has just abandoned.
Second, the issue isn't whether human agency can bring about change or not. The issue is whether we can derive useful policy advice from a (general equilibrium) empirical model that accurately captures how the interactions among incentivized agents produce policy outcomes. Rodrik's point, and he is absolutely correct, is that we cannot. Such a model can produce redundant advice (do what you have an incentive to do and are thus already doing) or it can offer advice that the agents whose behavior we have modeled have no incentive to accept (do something that you have no incentive to do). Hence, the better we are at explaining policy choices as a function of political incentives, the less our models can generate useful policy advice. Even in models based solely on human agency.
Third, the issue isn't whether scholarly inquiry has an "impact on the world." The issue is whether positive political economy generates knowledge that can be applied usefully to promote policy change. Rodrik argues, and reasonably so, that the answer is no.
In short, Rodrik makes two points. One, positive political economy is useless for policy advice. Two, we can affect policy advice by changing how people conceptualize their identities and interests. For my money, this position defined by these two points is utterly incoherent.
Thus, although I agree with Phil on the first point--if you are going to practice social science, you have to accept the implications and going all mystical isn't really useful--I really dont understand where Phil sits on the fundamental issues that Rodrik raises. Can we derive useful policy advice from empirically accurate general equilibrium models of politics? Are ideas exogenous to the political system in the way Rodrik conceives? Perhaps Phil would take a moment to respond.
Wednesday, March 6, 2013
Paradoxically, you had me at hello.
.
Wednesday, March 6, 2013
Subscribe to:
Post Comments (Atom)
14 comments:
We've gone around on this before, IIRC, and while I agree with you on the basic conclusion I think there is one possible exception: when there are information asymmetries. Ie, things political scientists know about *the general tendency of political outcomes* that politicians do not know.
For example, while I'm sitting here eating my Campbell's soup I saw you'd written a post. I read it and thought "I recall a comp question kind of along those lines" and then tried to remember how I'd answered it. And I believe my answer had something to do with information asymmetry.
Then, I flipped over to the Monkey Cage and saw this post from Gelman:
http://themonkeycage.org/blog/2013/03/06/one-reason-that-summers-and-obama-might-not-have-pushed-for-a-bigger-stimulus-in-2009/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+themonkeycagefeed+%28The+Monkey+Cage%29
Now I don't know if Gelman is right, but let's just assume he is and that this was Obama's thinking. It was clearly influenced by results from political scientists. Moreover, this is not something that Obama's intuition would necessarily have told him, and it's unlikely that he would have spent tons of time studying the question in detail before having to make the decision. So, consuming political science could have influenced policy.
Now what that means, to me, is that political science isn't very good. That is, we have no variable on the right-hand side called "Has the president consumed political science research?" or something. So your conclusion still ends up holding, for the most part: either our research is good -- in which case there is no policy advice -- or else it isn't all that good -- in which case there is no policy advice. But there's a possible middle ground: our research is kind of good, but not perfect. There are central tendencies, but they come with error bars. And those error bars provide a little room for policy advice in some situations. And any individual politician can't/won't know where the error bars are on every question that will come up, so political economy research could help her.
Could one make the argument that ideas are to political economy what "tastes" are to consumer theory? Economists regard consumer tastes as important and agree that they can cause shifts in a demand curve but don't have whole lot that is particularly insightful to say about how tastes are formed. This incidentally reminds me of a criticism Krugman made about the micro foundations of New Keynesian economics, namely micro ultimately needs to have psychological/sociological foundations. Now how do you actually explain utility in political economy? Surely a sense of justice is a part of politics and ecomomics. If my sense of justice changes how I conceptualize interests will change and thus policies I support. Are questions like this really outside of the purview of social science or political economy? Or are they just difficult to mathematize.....
Yes, we have gone around on this before. Re the Gelman thing, this seems to pertain to the information set: what happens if I do x and what happens if I do y. This is subsumed into a model based on complete information and possibly absent in a model with incomplete information. Hence, our models could offer no better advice than that which is possible given the information set.
As for the error bars, isn't that uncertainty? And isn't uncertainty the absence of knowledge? And so isn't this precisely what Rodrik is arguing--we believe we can offer useful advice when our models contain lots of uncertainty. As we reduce the uncertainty, our belief in our ability to offer useful advice diminishes. But notice that what changes isn't the usefulness of the advice; its our belief in its usefulness.
Vladimir--I tend to think of ideas as technology rather than as tastes. Tastes seem rather individualistic and idiosyncratic. Ideas seem more social and shape the production of policy, much as technology shapes the production of goods and services.
As for whether questions re justice and such are outside the purview of political economy, the question in my mind is whether the formulation of these conceptions are endogenous to the very political interactions that generate other outcomes. I would imagine that they are.
Hi Thomas. Good post. I should have been clearer about these things.
I think you are right that Rodrik is arguing that no useful policy advice can be derived from PE models, and I think he's right about that. But I don't see that as any sort of argument against continuing to analyze such models.
(I don't think he's right that better models limit our ability to give advice, btw. I think better models, if they actually are better, uncover how little room there was for giving useful advice in the first place. They don't change the state of affairs, only our understanding of them.)
As for ideas, I don't think they are exogenous to the political system, no. I'm not sure Rodrik's position requires that we believe that they are, though, since we as scholars are part of the system, not "outsiders". The important question is whether they are less difficult to change than are material incentives. And I'm willing to grant that they are...in the same sense that 0.000001 is a larger probability than 0. Trying to change the way people define their interests seems to me a little bit different than trying to tell them not to pursue their interests.
Simply put, though, I don't think that we as scholars are particularly likely to affect change through *any* channel, and I don't think that's where scholars should be focusing their energy. I think Rodrik is essentially right to characterize PE the way he does. I don't see why that should lead anyone to abandon PE. I think it borders on absurd to declare that you don't want to understand the world better because once you do, you'll see that you never had much chance of changing it.
Re Phil's
I think it borders on absurd to declare that you don't want to understand the world better because once you do, you'll see that you never had much chance of changing it.
This comes close to saying you think it's absurd for a scholar to want to improve the world. (You don't say that explicitly, but that's one implication, it seems to me.) But that is *precisely* the motive that led (past tense deliberate, maybe not so true now) many people to study politics and IR. I confess I haven't read Rodrik's piece but I can't see his apparent desire to improve the world as being anything other than admirable. (I'll forgo quoting the 11th thesis on Feuerbach.) Not everyone has to be focused on improving the world, but we'd better have some scholars who are, and how they choose to label their efforts and package them intellectually is somewhat secondary, IMO.
See, btw, for some remarks relevant to this issue, Michael J. Smith and Linda B. Miller, "Reflections on an Ideal Influence," in Ideas and Ideals: Essays on Politics in Honor of Stanley Hoffmann (Westview Press, 1993), esp. pp.57-59.
I'm sympathetic to the desire to live in a better world, LFC. And I don't need to be told that most people study politics because they hope to make a difference. But it simply will not do to pretend that obstacles to change don't exist because we don't like that they exist. When Rodrik says that the problem with PE is that he doesn't want to live in the world it describes, that's basically what he's doing. And I'm pretty comfortable calling that absurd.
I'll refrain from reply b.c, as I mentioned, I haven't read the Rodrik piece.
I've now read the Rodrik piece albeit quickly.
On what seems to be, as I gather, a central pt at issue betw Rodrik and T.Oatley -- is the formation of ideas exogenous to the "very political interactions that generate other outcomes"? -- I think I am more w Rodrik (though I might not put the case quite as strongly as Rodrik does).
One way to look at this is that it is arguably a rehash of liberalism (or Weber, if you prefer that) vs. Marxism at its least nuanced: are ideas (consciousness) a product of "social being," or do ideas have some independent force (as Weber's "switch men" metaphor implies)?
I shd have said: independent origin and independent force
Hi Phil--thanks for the clarification. I appreciate that.
I do wonder if Rodrik has become a social constructivist. His assertion that identities and interests are not fixed and exogenous, but can be changed seems to indicate that.
This might mean his conversion is somewhat less a shift to mysticism and somewhat more a shift from neo-positivism toward something like critical realism.
This seems less absurd, on the face of it.
"I do wonder if Rodrik has become a social constructivist"
Before reading this remark it had occurred to me that while Rodrik perhaps/probably has not read Wendt, there may be some similarity in how they think about ideas and interests, though they would use different language to make their point. (Not that Wendt-by- himself=constructivism, of course, but his discussion in STIP is what came to mind. A blog thread is prob. not an esp. good place to hash this out, however.)
Post a Comment