Wednesday, May 28, 2014

With One Hand Tied Behind My Back

Some economists sometimes say that neoclassical economics will be abandoned when a better theory is available. And they use this as an excuse for not pursuing, say, heterodox economics.

I am tempted to respond that a better theory already exists, for example, some combination of Post Keynesian and institutional economics. But two properties of most varieties of heterodox economics make it difficult for this answer to register for mainstream economists. Before I get to to those properties, let me caveat my answer.

Economists have a number of applied fields. Is heterodox economics developed to such a state that one can say economists would (or should) know how it should impact their work in every applied field? I cannot say how my concerns would apply to, say, transport economics, although I know that work exists there that draws on behavioral economics. Anyways, I doubt that many applied economists are aware of or concerned with how there work fits better with some approaches in heterodox economics. Much applied work seems to me agnostic between paradigms. For example, would you say that most work applying linear programming is Sraffian, even though the historical origins of linear programming are entwined with John Von Neumann's Sraffa-like, classically inspired growth model? How about work with National Income and Product Accounts (NIPA), which, to me, fit comfortably with the classical focus on the generation, distribution, and use of the surplus? Furthermore, I know of some applied fields to which heterodox economists have contributed. I doubt that I can count Walter Isard as a heterodox economist, but I think of economic geography and regional analysis as compatible with Sraffian economics. Wassily Leontief, as far as I am concerned, was a heterodox economist, although, as far as I know, he never commented on the theoretical controversies in which I am interested. I am fairly sure that Bertram Schefold has work in energy economics inspired by Sraffian theory. Is not Amartya Sen's work on the capabilities approach a heterodox contribution to welfare economics? Given Paolo Sylos Labini's importance to the field, I suspect that many models exist in Industrial Organization (IO) that are compatible with a broad definition of Post Keynesianism. To conclude, the connection between applied fields and debates about heterodox economics are not clear to me.

Anyways, heterodox economists tend to emphasize a need for open systems models in economics. This emphasis is the first property of heterodox economics that I want to mention that makes it difficult for mainstream economists to accept the superiority of much heterodox work. Tony Lawson has provided lots of elaboration on the need for an open systems approach to studying human society. But I do not need to go into such philosophy. My favorite approach to price theory takes the distribution of income as given for the most abstract theory of the prices of commodities. Likewise, the composition of final uses is exogenous in Sraffa's theory. The structure of the theory is open to elaborations at lower levels of abstractions that include inputs from other social sciences than economics. Consumption might be explained by a substantive theory, in contrast to the empty formal theory of neoclassical utility-maximization.

Second, heterodox economists tend to be skeptical of the possibility of economic laws applying to all societies across all of human existence. By contrast, Lionel Robbins' definition of economics as the study of the allocation of scarce means among alternative uses is often claimed to be universal. There is nothing in the mainstream about capitalism or how social norms might differ among societies. Contrast, for example, with Luigi Pasinetti's work on structural dynamics, in which the theoretical structure builds in a place for variations in institutions. Or look at suggestions in Sraffa's book that taking one of wages or the rate of profits as the independent variable might be more appropriate at different times or places. Or look at the emphasis on conventions in Keynes' analysis of investment. I could go on.

So, by adopting theories appropriate for the problem domain, heterodox economists have developed theories that are superior to many orthodox theories. But, because of this very appropriateness, mainstream economists are socialized to fail to perceive this superiority.


Matthieu Montalban said...

Interesting post, but I think you miss one point: the difference of mainstream and heterodox approaches is that you have got an integrated framework for all objects in one side, and approaches more fragmented and not always compatible together in another side. All the method of neoclassical economics (i.e methodological individualism, rationality and utility maximization), concepts and types of reasoning can be applied to macro, industrial organization, health economics, urban/spatial economics, development economics, labour economics, consumption theory, economics of environment, law and economics, public choice, institutional economics, economics of marriage (!), sex (!!?) etc... This is the real (explanatory and "political") power of neoclassical framework. We have no such a thing in heterodox economics, instead perhaps in old institutionalism, but far less developed (and then, we are using old institutionalism, we are considered as sociologists... that's what happens to me each time). A lot of econometric models are based on neoclassical models. For example, when some economists are trying to measure the value of a landscape and the willingness to pay to preserve it (this is done in environmental economics), it is based on neoclassical economics. In the same vein, when economists are trying to select which drugs or treatment has the best cost/benefits ratio, it is also neoclassical economics (you compare the cost with the utility, even if the utility is approximated with the gain in life expectancy). When in spatial economics, we study the role of location of firms, of houses or the role of proximity with the center on the value of rents, it is largely based on neoclassical models etc... I'm convinced that what we say in heterodox schools is sometime more relevant than neoclassical economics, but we would never have such an integrated framework unfortunately.

Robert Vienneau said...

Good point. I have heard claims that in applications one can relax various assumptions of neoclassical economics. Thus, the aggregate of applied neoclassical economics may be internally inconsistent. Yet, I agree, the image is as you say.

And I thought my comments in the post on applied economics resembled too much a list.

valter said...

"the aggregate of applied neoclassical economics may be internally inconsistent"

but probably the aggregate of applied X economics will be internally inconsistent for all X because there is no realistic prospect of a Grand Unified Theory of economics or social science.

So that's not a very damaging indictment.

Robert Vienneau said...

I guess I am not reporting the claim I recall from this discussion on a discussion list for the history of economic thought.

The idea was that, in practice, all assumptions of neoclassical economics "float" when you consider the sum of applied fields. Here they are willing to disregard this assumption, there they are willing to discard another. And this nebulousness in their theory just does not bother mainstream economists.

I cannot find a link, and this recasting may still not seem very damaging to you.