Journal Article

No. 2009-9 | April 02, 2009
Economists, Incentives, Judgment, and the European CVAR Approach to Macroeconometrics PDF Icon

Abstract

This paper argues that the DSGE approach to macroeconometrics is the dominant approach because it meets the institutional needs of the replicator dynamics of the profession, not because it is necessarily the best way to do macroeconometrics. It further argues that this “DSGE theory-first” approach is inconsistent with the historical approach that economists have advocated in the past and that the alternative European CVAR approach is much more consistent with economist’s historically used methodology, correctly understood. However, because the European CVAR approach requires explicit researcher judgment, it does not do well in the replicator dynamics of the profession. The paper concludes with the suggestion that there should be an increase in dialog between the two approaches.

JEL Classification

B4

Citation

David Colander (2009). Economists, Incentives, Judgment, and the European CVAR Approach to Macroeconometrics. Economics: The Open-Access, Open-Assessment E-Journal, Vol. 3, 2009-9. http://dx.doi.org/10.5018/economics-ejournal.ja.2009-9

Assessment

Downloads: 2690 (Journalarticle: 1416, Discussionpaper: 1274)
external link Search this article at Google Scholar



Comments and Questions


Claude Hillinger - Critical Comment on the Article
April 06, 2009 - 21:22

I have rather different views regarding both the proper methodology for and the history of macroeconomics than David Colander. These views have been stated much more fully than is possible here, and with references to relevant literature, in the references at the end of this comment.

I am much ...[more]

... more skeptical about macroeconomics than he is. I have seen research programs in macroeconomics come and go. At the onset of my career, macroeconomics in introductory textbooks was essentially the Keynesian cross. In advanced texts it was the IS-LM model. Empirical macroeconomics took the form of adding ever more equations to the hundreds that already constituted the large-scale macro-econometric models. This was succeeded by two distinct approaches: a. Monetarism which taught that in empirical applications a single equation is plenty. b. Time series analysis which produced an orgy of testing: for co-integration, for unit roots, for non-linearity, for chaos, for causality. Economic theory was almost wholly absent. During subsequent decades, the new Keynesians and new classicals engaged in spirited combat and the proponents of real business cycles, allied with the new classicals joined the fray. Now DSGE models in heavy mathematical armor are strutting on center-stage.

Each of these research programs claimed and believed itself to be the future of macroeconomics; none was. That this record will be improved upon by either DSGE or CVAR stretches credulity. Ultimately there is only one criterion for judging success or failure of a research program: did it produce significant and replicable knowledge. All the research programs mentioned have failed this test. Of course, before a program can yield results it is carried forward on the basis of hope. So let us look at what the hope is based on in the case of DSGE and CVAR.

While Colander is critical of DSGE, he has not focused on the Achilles heel of that, as well as other approaches that claim to have provided micro foundations for macroeconomics. The Achilles heel is the assumption of representative agents. Extensive literatures exist both on the representative consumer and the representative producer. Essentially the conclusion is that a representative agent can be assumed if all agents are identical in all respects. If the utility or production functions are identical and homothetic, then the agents may have budgets of different size. That these assumptions are totally unrealistic is not even the only problem. The literature on representative agents deals with the problem of aggregating across agents. Empirical applications involve an aggregation across commodities that is equally problematic, but almost never discussed. On the side of production, all types of labor are aggregated to just one homogeneous unit; similarly for all types of capital. The myriad consumer goods are similarly aggregated into broad categories such as food or housing. The only assumption that allows such aggregates to be entered in utility or production functions is Hicks aggregation, i.e. that all prices within a group always move in the same proportion. For broad aggregates this assumption is absurd.

DSGE advocates and other users of representative agent models claim superior rigor for their approaches, as extensively discussed by Colander. This may be true for the mathematics that they use. Conceptually, I find that the approach cannot be taken seriously.

Regarding the CVAR approach I have no independent knowledge, so I concentrate on the description by Campos et al. the Colander cites approvingly. They describe an economy as:
A complicated, dynamic, nonlinear, simultaneous, high-dimensional, and evolving entity [in which] social systems alter over time; laws change and technological innovations occur.
As is clear from many of Colanders remarks, the assumption is that this complexity must be taken into account in all aspects.

I regard the view that complex reality precludes simple explanations as being pre-scientific. It is an argument frequently advanced by politicians. In The Double Helix, James Watson describes how the discovery of the double helix as an essentially simple mechanical explanation of heredity was opposed by traditionally oriented biologists who were convinced of the irreducible complexity of life. That simplicity is a primary consideration for the acceptability of a theory in the natural sciences is stressed in many books on the philosophy of science. Modeling complexity directly was tried once before in macroeconomics with the large-scale macroeconomic models.

I agree with Colander that over the past decades the training and socialization of economists has gone more and more in the direction of creating streamlined and uncritical suppliers of whatever the mainstream of the moment is demanding. Thus the history of economic thought, once a core subject in any graduate program, has all but disappeared from the curricula. By acquainting students with alternative theories and approaches, such courses enabled them to think more critically about the fashions of the moment.

The philosopher and historian of science Ravetz has pointed to what I think is the deepest level cause of this development. He argued that the social ‘sciences’ cannot really be called sciences because they have not agreed on criteria for determining what is factually the case. In other words, they have no agreed upon and stable criteria for determining the truth or falsity of claims, either with regard to stylized facts or with regard to explanatory theories. The consequence is that each new research program develops the criteria that tend to support it. In economic theory the prime criterion has been that the theory should be mathematical, the mathematics should be sufficiently difficult to inspire respect, and the proposed model should imply the optimality of a market economy. Economics thus joined ranks with the dominant neoliberal ideology. SDGE is the dominant macro-theory of the moment because it best fits this criterion.

References:
Hillinger, Claude (2008). Science and Ideology in Economic, Political and Social Thought.
Economics: The Open-Access, Open-Assessment E-Journal, Vol. 2, 2008-2.
http://www.economics-ejournal.org/economics/journalarticles/2008-2
(2005), Evidence and Ideology in Macroeconomics: The Case of Investment Cycles (March 2005). Available at SSRN: http://ssrn.com/abstract=814527
Ravetz, Jerome R. (1971), Scientific Knowledge and its Social Problems, Oxford, Clarendon Press.