Princeton University Press Rational Fools: A Critique of the Behavioral Foundations of Economic Theory Author(s): Amartya K. Sen Reviewed work(s): Source: Philosophy & Public Affairs, Vol. 6, No. 4 (Summer, 1977), pp. 317-344 Published by: Blackwell Publishing Stable URL: http://www.jstor.org/stable/2264946 . Accessed: 21/11/2011 09:10 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. Princeton University Press and Blackwell Publishing are collaborating with JSTOR to digitize, preserve and extend access to Philosophy & Public Affairs. http://www.jstor.org AMARTYAK . SEN Rational Fools: A Critique of the Behavioral Foundations of Economic Theory I In his Mathematical Psychics, published in i88i, Edgeworth asserted that "the first principle of Economics is that every agent is actuated only by self-interest."'1T his view of man has been a persistent one in economic models, and the nature of economic theory seems to have been much influenced by this basic premise. In this essay I would like to examine some of the problems that have arisen from this concep- tion of human beings. I should mention that Edgeworth himself was quite aware that this so-called first principle of Economics was not a particularly realistic one. Indeed, he felt that "the concrete nineteenth century man is for the most part an impure egoist, a mixed utilitarian."2 This raises the interesting question as to why Edgeworth spent so much of his time and talent in developing a line of inquiry the first principle of which he believed to be false. The issue is not why abstractions should be em- This Herbert Spencer Lecture, delivered at Oxford University in October I976, will appear in Scientific Models and Man, ed. H. Harris (forthcoming 1978) and is printed here by kind permission of Oxford University Press. For helpful com- ments on an earlier version, I am grateful to the Editors of this journal, and to Ake Andersson, Isaiah Berlin, Frank Hahn, Martin Hollis, Janos Komai, Derek Parfit, Christopher Peacocke, and Tibor Scitovsky. I. F.Y. Edgeworth, Mathematical Psychics: An Essay on the Application of Mathematics to the Moral Sciences (London, i88i), p. i6. 2. Edgeworth (i88i), p. 104. In fact, he went on to make some interesting remarks on the results of "impure" egoism, admitting an element of sympathy for each other. The remarks have been investigated and analyzed by David Col- lard, "Edgeworth's Propositions on Altruism," Economic Journal 85 (1975). 3I8 Philosophy & Public Affairs ployed in pursuing general economic questions-the nature of the in- quiry makes this inevitable-but why would one choose an assumption which he himself believed to be not merely inaccurate in detail but fundamentally mistaken? As we shall see, this question is of continu- ing interest to modem economics as well. Part of the answer, as far as Edgeworth was concerned, undoubtedly lay in the fact that he did not think the assumption to be fundamen- tally mistaken in the particular types of activities to which he applied what he called "Ceconomicalc alculus": (i) war and (ii) contract. "Admitting that there exists in the higher parts of human nature a tendency towards and feeling after utilitarian institutions," he asked the rhetorical question: "could we seriously suppose that these moral considerations were relevant to war and trade; could eradicate the controlless core of human selfishness, or exercise an appreciable force in comparison with the impulse of self-interest."3 He interpreted Sidg- wick to have dispelled the "illusion" that "the interest of all is the interest of each," noting that Sidgwick found the "two supreme prin- ciples-Egoism and Utilitarianism" to be "irreconcilable, unless indeed by religion." "It is far from the spirit of the philosophy of pleasure to deprecate the importance of religion," wrote Edgeworth, "but in the present inquiry, and dealing with the lower elements of human nature, we should have to seek a more obvious transition, a more earthy pas- sage, from the principle of self-interest to the principle, or at least the practice, of utilitarianism."4 Notice that the context of the debate is important to this argument. Edgeworth felt that he had established the acceptability of "egoism"a s the fundamental behavioral assumption for his particular inquiry by demolishing the acceptability of "utilitarianism" as a description of actual behavior. Utilitarianism is, of course, far from being the only non-egoistic approach. Furthermore, between the claims of oneself and the claims of all lie the claims of a variety of groups-for example, families, friends, local communities, peer groups, and economic and social classes. The concepts of family responsibility, business ethics, class consciousness, and so on, relate to these intermediate areas of concern, and the dismissal of utilitarianism as a descriptive theory 3. Edgeworth, p. 52. 4. Ibid., pp. 52-53. 3I9 Rational Fools of behavior does not leave us with egoism as the only alternative. The relevance of some of these considerations to the economics of negotia- tions and contracts would be difficult to deny. It must be noted that Edgeworth's query about the outcome of economic contact between purely self-seeking individuals had the merit of being immediately relevant to an abstract enquiry that had gone on for more than a hundred years already, and which was much discussed in debates involving Herbert Spencer, Henry Sidgwick, and other leading thinkers of the period. Two years before Edgeworth's Mathematical Psychics appeared, Herbert Spencer had published his elaborate analysis of the relation between egoism and altruism in The Data of Ethics. He had arrived at the comforting-if somewhat unclear -conclusion that "general happiness is to be achieved mainly through the adequate pursuit of their own happinesses by individuals; while, reciprocally, the happiness of individuals are to be achieved in part by their pursuit of the general happiness."5I n the context of this relatively abstract enquiry, Edgeworth's tight economic analysis, based on a well-defined model of contracts between two self-seeking individuals, or between two types of (identical) self-seeking individuals, gave a clear answer to an old hypothetical question. It appeared that in Edgeworth's model, based on egoistic behavior, there was a remarkable correspondence between exchange equilibria in competitive markets and what in modern economic terms is called "the core" of the economy. An outcome is said to be in "the core" of the economy if and only if it fulfills a set of conditions of unimprov- ability. These conditions, roughly speaking, are that not only is it the case that no one could be made better off without making somebody else worse off (the situation is what is called a "Pareto optimum"), but also that no one is worse off than he would be without trade, and that no coalition of individuals, by altering the trade among them- selves, could on their own improve their own lot. Edgeworth showed that given certain general assumptions, any equilibrium that can emerge in a competitive market must satisfy these conditions and be in "the core." Thus, in Edgeworth's model the competitive market equilibria are, in this sense, undominated by any feasible alternative 5. H. Spencer, The Data of Ethics (London, 1879; extended edition, I887), P. 238. 320 Philosophy & Public Affairs arrangement, given the initial distribution of endowments. More sur- prising in some ways was the converse result that if the number of individuals of each type were increased without limit, the core (repre- senting such undominated outcomes) would shrink towards the set of competitive equilibria; that is, the core would not be much more extensive than the set of competitive equilibria. This pair of results has been much elaborated and extended in the recent literature on general equilibrium with similar models and with essentially the same behavioral assumptions.6 Being in the core, however, is not as such a momentous achievement from the point of view of social welfare. A person who starts off ill- endowed may stay poor and deprived even after the transactions, and if being in the core is all that competition offers, the propertyless per- son may be forgiven for not regarding this achievement as a "big deal." Edgeworth took some note of this by considering the problem of choice between different competitive equilibria. He observed that for the utilitarian good society, "competition requires to be supplemented by arbitration, and the basis of arbitration between self-interested con- tractors is the greatest possible sum-total utility."7 Into the institu- tional aspects of such arbitration and the far-reaching implications of it for the distribution of property ownership, Edgeworth did not really enter, despite superficial appearance to the contrary. On the basis of the achievement of competition, however limited, Edgeworth felt en- titled to be "biassed to a more conservative caution in reform." In calculating "the utility of pre-utilitarian institutions," Edgeworth felt impressed "with a view of Nature, not, as in the picture left by Mill, all bad, but a first approximation to the best."8 I am not concerned in this essay with examining whether the approximation is a rather remote one. (This I do believe to be the case even within the structure of assumptions used by Edgeworth, but it is not central to the subject of this paper.) I am concerned here with the view of man which forms part of Edgeworth's analysis and 6. See, especially, K.J. Arrow and F.H. Hahn, General Competitive Analysis (San Francisco, 1971). 7. Edgeworth, p. 56. 8. Ibid., p. 82. 32I Rational Fools survives more or less intact in much of modem economic theory. The view is, of course, a stylized one and geared specifically to tackling a relatively abstract dispute with which Spencer, Sidgwick, and several other leading contemporary thinkers were much concerned-namely, in what sense and to what extent would egoistic behavior achieve gen- eral good? Whether or not egoistic behavior is an accurate assumption in reality does not, of course, have any bearing on the accuracy of Edgeworth's answer to the question posed. Within the structure of a limited economic model it provided a clear-cut response to the abstract query about egoism and general good. This particular debate has gone on for a long time and continues to provide motivation for many recent exercises in economic theory to- day. The limited nature of the query has had a decisive influence on the choice of economic models and the conception of human beings in them. In their distinguished text on general equilibrium theory, Arrow and Hahn state (pp. vi-vii): There is by now a long and fairly imposing line of economists from Adam Smith to the present who have sought to show that a decen- tralized economy motivated by self-interest and guided by price sig- nals would be compatible with a coherent disposition of economic resources that could be regarded, in a well-defined sense, as superior to a large class of possible alternative dispositions. Moreover, the price signals would operate in a way to establish this degree of coherence. It is important to understand how surprising this claim must be to anyone not exposed to the tradition. The immediate "common sense" answer to the question "What will an economy motivated by individual greed and controlled by a very large num- ber of different agents look like?" is probably: There will be chaos. That quite a different answer has long been claimed true and has indeed permeated the economic thinking of a large number of people who are in no way economists is itself sufficient ground for investi- gating it seriously. The proposition having been put forward and very seriously entertained, it is important to know not only whether it is true, but whether it could be true. A good deal of what follows is concerned with this last question, which seems to us to have considerable claims on the attention of economists. 322 Philosophy & Public Affairs The primary concern here is not with the relation of postulated models to the real economic world, but with the accuracy of answers to well-defined questions posed with preselected assumptions which severely constrain the nature of the models that can be admitted into the analysis. A specific concept of man is ingrained in the question itself, and there is no freedom to depart from this conception so long as one is engaged in answering this question. The nature of man in these current economic models continues, then, to reflect the particu- lar formulation of certain general philosophical questions posed in the past. The realism of the chosen conception of man is simply not a part of this inquiry. II There is another nonempirical-and possibly simpler-reason why the conception of man in economic models tends to be that of a self-seek- ing egoist. It is possible to define a person's interests in such a way that no matter what he does he can be seen to be furthering his own interests in every isolated act of choice.9 While formalized relatively recently in the context of the theory of revealed preference, this ap- proach is of respectable antiquity, and Joseph Butler was already arguing against it in the Rolls Chapel two and a half centuries ago.'0 The reduction of man to a self-seeking animal depends in this ap- proach on careful definition. If you are observed to choose x rejecting y, you are declared to have "revealed"a preference for x over y. Your personal utility is then defined as simply a numerical representation of this "preference," assigning a higher utility to a "preferred"a lterna- tive. With this set of definitions you can hardly escape maximizing your own utility, except through inconsistency. Of course, if you choose x and reject y on one occasion and then promptly proceed to do the exact opposite, you can prevent the revealed preference theorist 9. If a person's actions today affect his well-being in the future, then under this approach his future interests must be defined in terms of the way they are assessed today. In general, there is no reason to presume that the future interests as assessed today will coincide with those interests as assessed in the future. This adds an additional dimension to the problem, and I am grateful to Derek Parfit for convincing me of the conceptual importance of this question. Io. J. Butler, Fifteen Sermons Preached at the Rolls Chapel (London, 1726); see also T. Nagel, The Possibility of Altruism (Oxford, 1970), p. 8I. 323 Rational Fools from assigning a preference ordering to you, thereby restraining him from stamping a utility function on you which you must be seen to be maximizing. He will then have to conclude that either you are incon- sistent or your preferences are changing. You can frustrate the re- vealed-preference theorist through more sophisticated inconsistencies as well.1' But if you are consistent, then no matter whether you are a single-minded egoist or a raving altruist or a class conscious militant, you will appear to be maximizing your own utility in this enchanted world of definitions. Borrowing from the terminology used in connec- tion with taxation, if the Arrow-Hahn justification of the assumption of egoism amounts to an avoidance of the issue, the revealed prefer- ence approach looks more like a robust piece of evasion. This approach of definitional egoism sometimes goes under the name of rational choice, and it involves nothing other than internal consistency. A person's choices are considered "rational" in this ap- proach if and only if these choices can all be explained in terms of some preference relation consistent with the revealed preference defi- nition, that is, if all his choices can be explained as the choosing of "most preferred" alternatives with respect to a postulated preference relation.'2 The rationale of this approach seems to be based on the idea that the only way of understanding a person's real preference is to examine his actual choices, and there is no choice-independent way of understanding someone's attitude towards alternatives. (This view, by the way, is not confined to economists only. When, many years ago, I had to take my qualifying examination in English Literature at Cal- cutta University, one of the questions we had to answer concerning A Midsummer Night's Dream was: Compare the characters of Hermia and Helena. Whom would you choose?) I have tried to demonstrate elsewhere that once we eschew the curious definitions of preference and welfare, this approach presumes both too little and too much: too little because there are non-choice sources of information on preference and welfare as these terms are iI. See H.S. Houthakker, "Revealed Preference and the Utility Function," Eco- nomica 17 (1950); P.A. Samuelson, "The Problem of Integrability in Utility Theory," Economica I17 (1950). 12. For the main analytical results, see M.K. Richter, "Rational Choice," Pref- erence, Utility and Demand Theory, ed. J.S. Chipman et al. (New York, 1971). 324 Philosophy & Public Affairs usually understood, and too much because choice may reflect a com- promise among a variety of considerations of which personal welfare may be just one.13 The complex psychological issues underlying choice have recently been forcefully brought out by a number of penetrating studies dealing with consumer decisions14 and production activities.15 It is very much an open question as to whether these behavioral characteristics can be at all captured within the formal limits of consistent choice on which the welfare-maximization approach depends.16 III Paul Samuelson has noted that many economists would "separate economics from sociology upon the basis of rational or irrational be- havior, where these terms are defined in the penumbra of utility 13. A.K. Sen, "Behaviour and the Concept of Preference," Economica 40 (1973). See also S. Korner's important recent study, Experience and Conduct (Cambridge, I97I). Also T. Schwartz, "Von Wright's Theory of Human Welfare: A Critique," forthcoming in P.A. Schlipp, ed., The Philosophy of Georg Henrik von Wright; T. Majumdar, "The Concept of Man in Political Economy and Eco- nomics," mimeographed (Jawaharlal Nehru University, New Delhi, 1976); and F. Schick, "Rationality and Sociality," mimeographed (Rutgers University, Phi- losophy of Science Association, 1976). I4. See T. Scitovsky, The Joyless Economy: An Inquiry into Human Satisfac- tion and Consumer Dissatisfaction (London and New York, I976). See also the general critique of the assumption of "rational" consumer behavior by J. Kornai, Anti-Equilibrium (Amsterdam and London, I971), chap. ii; and the literature on "psychological choice models," in particular, D. McFadden, "Economic Ap- plications of Psychological Choice Models" (presented at the Third World Econ- ometric Congress, August I975). I5. See H. Liebenstein, "Allocative Efficiency vs. x-Efficiency," American Eco- nomic Review 56 (I966). Also critiques of the traditional assumption of profit maximization in business behavior, particularly W.J. Baumol, Business Behavior, Value and Growth (New York, I959); R. Marris, The Economic Theory of Mana- gerial Capitalism (London, I964); 0. Williamson, The Economics of Discretion- ary Behavior (Chicago, I967); and A. Silberston, "Price Behaviour of Firms," Economic Journal 8o (1970), reprinted in Royal Economic Society, Surveys of Applied Economics, vol. i (London, I973). i6. On the required conditions of consistency for viewing choice in terms of a binary relation, see my "Choice Functions and Revealed Preference," Review of Economic Studies 38 (I97I); H.G. Herzberger, "Ordinal Preference and Ra- tional Choice," Econometrica 4I (1973); K. Suzumura, "Rational Choice and Revealed Preference," Review of Economic Studies 43 (I976); S. Kanger, "Choice Based on Preference," mimeographed (Uppsala University, I976). 325 Rational Fools theory." 17 This view might well be resented, for good reasons, by soci- ologists, but the cross that economists have to bear in this view of the dichotomy can be seen if we note that the approach of C"rationabl e- havior," as it is typically interpreted, leads to a remarkably mute theory. Behavior, it appears, is to be C"explainedin terms of prefer- ences, which are in turn defined only by behavior." Not surprisingly, excursions into circularities have been frequent. Nevertheless, Samuel- son is undoubtedly right in asserting that the theory "is not in a tech- nical sense meaningless."18 The reason is quite simple. As we have already discussed, the approach does impose the requirement of internal consistency of observed choice, and this might well be refuted by actual observations, making the theory "meaningful" in the sense in which Samuelson's statement is intended. The requirement of consistency does have surprising cutting power. Various general characteristics of demand relations can be derived from it. But in the present context, the main issue is the possibility of using the consistency requirement for actual testing. Samuelson speci- fies the need for "ideal observational conditions" for the implications of the approach to be "refuted or verified." This is not, however, easy to satisfy since, on the one hand, our love of variety makes it illegiti- mate to consider individual acts of choice as the proper units (rather than sequences of choices) while, on the other hand, lapse of time makes it difficult to distinguish between inconsistencies and changing tastes. There have, in fact, been very few systematic attempts at test- ing the consistency of people's day-to-day behavior, even though there have been interesting and useful contrived experiments on people's reactions to uncertainty under laboratory conditions. What counts as admissible evidence remains unsettled. If today you were to poll economists of different schools, you would almost certainly find the coexistence of beliefs (i) that the rational behavior theory is unfalsi- fiable, (ii) that it is falsifiable and so far unfalsified, and (iii) that it is falsifiable and indeed patently false.19 17. P.A. Samuelson, The Foundation of Economics (Cambridge, Mass., 1955), p. 90. i8. Ibid., p. 9I. I9. The recent philosophical critiques of rational behavior theory include, among others, M. Hollis and E.J. Nell, Rational Economic Man (Cambridge, 1975); S. Wong, "On the Consistency and Completeness of Paul Samuelson's