Henry Manne on Behavioral Overreach

Henry G. Manne is Dean Emeritus at George Mason University School of Law

Behavioral Economics, like so many efforts previously to upend the hegemony of the neo-classical market model, will leave some footprints on the intellectual sands of time.  However, there is no way that it can accomplish what many of its disciples seem, subliminally at least, to believe:  that we should abandon the traditional model with (because of?) all its implications about private property, competitive markets and individual freedom.  That dream is, of course, ridiculous for one obvious and frequently mentioned reason: Behavioral Economics does not even attempt to offer an integrated theory of resource allocation, the ultimate and necessary mission of any economic theory.  All it does is putter around some select edges of the traditional theory (mainly the very weakly held – and, as we shall see, unnecessary – rationality assumption) and, again like its predecessors in the intellectual history of economics, it claims far more damage to the received model than it actually delivers.

My first observation about the present state of affairs in Behavioral Economic theory is that it builds too ambitiously on the findings of psychology and does not pay enough attention to what economists already well understand.  I don’t know of any respectable economist of the last 80 years or more who has pushed a fundamentalist notion of the rationality assumption in descriptive or analytical economics.  To the extent that some of the more dedicated Behavioralists do accuse devotees of the market model of something like this, they are clearly misunderstanding the heuristic nature of the perfect competition model.  I won’t take the opportunity to try to reeducate them in what economics is all about and what economists do.  Suffice it to say that the model is valuable enough if it does nothing more than paint an idealistic picture of a perfect market – else what’s a heaven for?

What I would like to point out, however, is the irrelevance of much of the substance of Behavioral Economics for “doing” economics.  My principal (as a matter of fact my sole) authority for this proposition (though some of Gary Becker’s work also comes to mind) is the magnificent classic article by Armen Alchian, Uncertainty, Evolution, and Economic Theory, 58 JPE 211 (1950). In this work Alchian is himself taking on the “full information” assumption of the classical model, which is actually broader than the rationality assumption attacked by the Behavioralists.  The basic conclusion of that work is that, even if individuals or firms make totally uniformed choices (to say nothing of merely somewhat irrational ones) the end result as far as the allocation of resources is concerned will be the same as in the traditional model.  This is so because the theory of competition in the classical model is itself a survival theory, and the survival mechanism will operate to winnow out the less efficient uses to which resources will unknowingly or irrationally be put, even if the human actors don’t understand the process or their role in it.  Of course, an economy based on this (again) purely heuristic assumption of perfect ignorance would not be as productive as one in which information and rationality play their usual assigned roles, but the difference may not be so great as would first appear and there is nothing peculiar or earth-shattering about finding that there are transactions costs in the world and that in equilibrium they will be accounted for.  And, as one begins to add notions of imitation and improvement, as Alchian does, one gets very close to a highly descriptive model of the real economy, and one which has plenty of room in it for all sorts of irrational behavior but without throwing the received theory out with the bath water.

So far this sounds as though everyone has made one huge error in taking Behavioral Economics seriously, but I think that is not really the case. There is a real weakness in traditional economic methodology, and I think that, perhaps unknowingly, the Behavioralists have taken fair advantage of this problem, while clearly overstepping the logical limits of their work. The problem in traditional economics style (I think “style” a more appropriate word here than “methodology”) is a general failure to keep separate that part of economic assumptions which pertains to individual behavior and that which properly pertains to aggregate behavior.  About the latter, BE has very little to say (here I am talking exclusively abut microeconomics, as I am really not sure how much relevance BE, aka psychology, has to macro areas), and Alchian simply carries the day.

But many economists do in fact use the same language and the same assumptions and the same analytical tools when discussing individuals’ behavior that they do when discussing aggregate behavior, and I think that without sufficient caveats, this is a big mistake.  The behavior of individuals, unless one is abstracting likely or probabilistic individual behavior from aggregate data, is the realm of psychology not of economics. Thus it is gross error to suggest that economics teaches us that individual’s demand curves are downward sloping if by that we mean that we can predict how a given individual will respond to a change in price.  Again economics can tell us what some idealized individual, whose assumed psyche is perhaps derived from aggregate data, will do, or perhaps, putting it slightly differently, economics may be able to tell us what the probability is that an individual’s behavior will conform to the aggregate data.  But never can economics (or economists, qua economists) tell us what any particular individual will do.  That is psychology, and any theory that assumes that some “normal” behavior is always functioning in individual choices is doomed to failure.  Alas, this caveat about the proper reach of economics is very often observed in the breach (it is rather cumbersome to be constantly making this distinction), and economists quite regularly commit this disaggregation fallacy.

This is why I do believe that there is a valuable role for BE in economics and further why I am willing to predict that this proper role will be assimilated into the conventional model but without all the ideological and political overtones.  This assimilation will not require any “correction” of the conventional theory, as I noted above.  It merely will be reflected by the greater care economists will take in their pronouncements to clarify whether they are speaking as economists (deriving conclusions from aggregate behavior) or psychologists.