Thoughtful Economic Man: Essays on Rationality, Moral Rules, and Benevolence.

Levy, David M.

This collection of papers originated as seminar papers at Gay Meeks's Cambridge course on philosophical issues of economics. The papers cluster into related groups with a positive intersection. There is an exchange between Frank Hahn ("Benevolence") and A. K. Sen ("Beneconfusion") over the question on whether human choice can be described in terms of a unitary preference order. David Collard's ("Love is not enough") addresses issues raised by altruism. There is a cluster of papers related to choice under uncertainty: John Broome ("Rationality and the sure-thing principle") and two essays on Keynes's ideas on investment, one by Robin Matthews and another by the editor. Broome's paper addresses the question of preferences for pieces of language, e.g., preferences for propositions. Two other papers do not cluster as easily. Alan Ryan ("Exploitation, justice and the rational man") writes a considerable essay on property, justice and self-interest. Tony Cramp ("Pleasures, prices and principles") attempts to Christianize economics by in part questioning the utilitarianism of Jeremy Bentham and J. S. Mill upon which welfare economics seems to depend.

Suppose we wish to introduce pieces of language (propositions) into the standard economic calculus, we might wish to ask why we accept one proposition and not another. Can we describe preferences for propositions and constraints for propositions? The kind of constraints which rational expectations place upon belief have been discussed extensively in the literature; consequently, I shall focus on the question of what sort of constraints there are on metapreferences. Can we define a corresponding metaconstraint for a metapreference? (I learned to think about metaconstraints from James Buchanan.)

The link between language and metapreference is natural because we know about metapreferences from language. Some of us say we want to have different preferences than we have. We tell others, and perhaps even ourselves, that we want to be better people. Is truth a constraint to metapreferences or to the rational acceptance of theory?

Perhaps because of historical reasons of proximity the exchange on the nature and significance of metapreferences between Sen and Hahn is elliptical and inconclusive. Their papers read like formulaic statements summarizing a disagreement long carried on in private. Hahn's objection to metapreferences, if I read him correctly, is that human nature is pretty much a constant quantity so whatever we would wish to become does not influence what we do in fact choose. When we are practicing benevolence, Hahn points out, we can be seen working out the usual maximizing conditions. What we see in models of economic behavior is what we get in the real world. Thus, if we want to explain choice, we need not deal with metapreferences. It seems to me that Hahn's objection ignores the link of economics with politics because we do see metapreferences working themselves out in totalitarian politics.

Let us think about metaconstraints. Is the fact that a proposition is false a sufficient metaconstraint? Let me bluntly assert that the falsity of a proposition does not imply uselessness in an optimizing context so its falsity is not a compelling constraint to its acceptance. (I think Broome believes it would be.) It is straightforward to construct an optimization setting in which false constraints of the sort which might be carried in language can be likelihood or utility enhancing [2]. A false constraint would not correspond to how people actually choose - it would not be false if did - but it might be a help in hard optimization problems. Thus, there seems to be no reason to preclude rational economic agents from metapreferences which deny the deep mathematical structure of their own preferences. Thus, agents with standard preferences might well have lexicographical metapreferences.

Suppose that ordinary people do have lexicographical metapreferences. We want to be better people and damn the consequences. Could we then vote to implement these metapreferences? Could we vote to make ourselves into New Model Persons? If "yes," then the question is what happens when we try to make ourselves into something which we cannot be? If we agree that the New Soviet Man is not feasible, what happens when policy tries to implement the illusion that it is? One answer is Stalin.

This issue was glanced at long ago by David Hume in History of England when he pointed out how the gross personal immorality of puritans, when it became common knowledge, discredited puritanism itself. (Hume himself might have been influenced by Shaftesburg's pragmatic doctrine that a true statement cannot be ridiculed.) The gross conflict between professed belief and behavior of Marxists opponents of birth control was used with great effect by their opponents to discredit Marxism itself [1]. This suggests that the gross conflict between professed beliefs and behavior can serve as a metaconstraint.

Obviously, an effective metaconstraint would require the free dissemination of ideas, information and ridicule. It is germane, therefore, that elsewhere Sen has found that democracy and a free press are the most effective anti-famine policy. The Soviet Union's constitutional crisis and demise may have been brought about in pan by the exposure of the great Soviet falsification of history [3].

Needless to say, I believe that the interrelation between among belief, language and choice are as interesting a cluster of problems as an economist can find. On the other hand, as an unrepentant student of George Stigler it is truly annoying in a book devoted to urging economists to look to the higher things to find violations of the duty that scholars owe to the past. Some sin by omission. Ryan's essay on property and justice manages not to mention David Hume's theory of property and justice; Cramp's consideration of utilitarianism does not look at those with whom the utilitarians quarreled - the spectator school of Adam Smith and his associates. Some sin by commission. Cramp writes that "the English Poor Law of 1834 ... put the unemployed into workhouses for the sake of general utility ...." The general excellence of the papers and the problems they confront, however, offers absolution for such peccadillos.

References

1. Holden, Christine and David M. Levy, "Birth Control and the Amelioration Controversy." History of Political Economy 25 (1993): 285-313.

2. Levy, David M. Economic Ideas of Ordinary People. London and New York: Routledge, 1992.

3. ----- "The Public Choice of Data Provision." Accountability in Research 3 (1993): 157-63.

David M. Levy George Mason University