Baron, J. (1988). Utility, exchange, and commensurability. Journal of Thought, 23, 111-131.

Utility, exchange, and commensurability1

Jonathan Baron, University of Pennsylvania

Western society, particularly the American version, is often criticized for an overreliance on the free market in the allocation of goods and in the division of labor (e.g., Bellah et al., 1985; Schwartz, 1986). For example, many of us extend this mode of behavior to such interpersonal relationships as the selection of friends and mates, the division of household work, and civic duties. Before we ask what we can do for our country (friend, spouse, child, student), we ask what it (he, she) can do for us in return.
Schwartz (1986) has recently argued that social theory, including psychology and economics, has exacerbated this problem by constructing theories in which the market is said to be part of human nature, thus making the situation appear inevitable if not justifiable. It is easy to find support for such a claim. Behaviorist theories of child-rearing and classroom management tell us that children can be induced to behave properly if it is made sufficiently worth their while. Social exchange theory (Homans, 1958) tells us that our everyday social relationships are based largely on market exchange, as does Becker's (1981) theory of the family. Equity theory (Walster, Walster, & Berscheid, 1978) tells us that people see distribution based on `to each according to his contribution' as fair. Most of these theorists do not even bother to ask about other cultures, let alone ask whether the situation is a good one and whether it is changeable. When other cultures are examined, the results that originally supported the theories often do not consistently hold up (see Leung & Bond, 1984).
But Schwartz goes on to argue that some of the very basic concepts of the social sciences are at fault - not just the particular applications of these concepts. One of these is the idea of commensurability of outcomes, as used in economics and the psychology of decision making. According to this idea, it is possible to compare any two outcomes (or changes from one state to another) so as to judge which is better from a certain point of view. As an extreme example, a scientist ought to be able to compare the moral effect of misreporting her data with the personal advancement that might result from doing so.
From the idea of commensurability, it is a short step to the market. If I prefer outcome A to outcome B, you prefer B to A, and I can give you B in exchange for A, we both benefit from an exchange. Thus, we both benefit if I give you use of my laboratory in return for my name on your publications (which I didn't otherwise read), my faithfulness in return for your physical attractiveness (as long as you keep it), my weapons for your hostages, etc. What is crucial here is that each person compares the loss of one sort of thing to the gain of something else. If I trade weapons for hostages, I am comparing the loss in weapons to the gain in hostages. (Even faithfulness amounts to a loss of freedom.)
By implication, Schwartz criticizes the concept of utility as well, for utility has served as the common coin by which outcomes are compared. (The idea of utility is similar to the concept of reward in learning theory.) Utility is, by one account, a measure of the extent to which a person's preferences are satisfied by a particular outcome of an action. Many descriptive theories claim that people seek to maximize utility.
The problem, according to Schwartz, goes beyond descriptive theories of behavior. Normative theories (idealized theories of how people ought to behave - see Baron, 1985, 1986) go on to argue that people ought to maximize utility (or, perhaps, expected utility under conditions of risk - a distinction that is irrelevant here). There are two forms of such theories. Simple decision-theory holds that individuals (and institutions) ought to maximize utility. This concept of utility is utterly neutral about moral questions. There is nothing wrong with preferring moral behavior, or with preferring that others' preferences be satisfied as well as one's own, but such preferences are completely optional. This kind of theory may be accused of condoning selfishness by failing to specify that anything is wrong with it.
The second form of normative utility theory is in fact the moral theory of utilitarianism, which holds that the morality of decisions be judged according to the extent to which they maximize utility over all relevant individuals.2 Utilitarianism justifies the use of methods based on utility maximization for social decision-making, including the setting of medical policies and social decisions about risky technology (e.g., Brown, Kahr, & Peterson, 1974; Keeney & Raiffa, 1976). Utilitarianism may also be used to justify reliance on the free market, on the basis of the same kind of exchange argument given above.3
If utilitarianism is at the root of the problem Schwartz finds, his argument is of interest beyond the confines of philosophy departments. Utilitarianism in the form at issue lies behind most of economics, and other social sciences that attempt to prescribe as well as understand. An attack on utilitarianism is thus an attack on applied social science as we know it.
Three assumptions (among others - see Sen & Williams, 1982) are commonly made in utilitarianism:
  1. Utility scale. Preference may be described in terms of a scale of utility along which all outcomes of a choice are ordered.
  2. Interval scale. Differences on such a scale are also ordered, allowing us to compare (for example) the gain to one-person with the loss to another from some decision. Note that this assumption includes the claim that differences may be compared across individuals or across domains of activity within an individual.
  3. Consequentialism. When evaluating choices, only consequences need be considered. The idea here is to exclude from consideration whether a choice conforms to various moral rules. Utility maximization is supposed to be the justification for actions or for rules, so it cannot take such rules into account.
At issue is an implication of these assumptions, the principle of exchange: two individuals ought to make an exchange whenever the utility of what each gets is higher than the utility of what each must pay. The utility-scale assumption plus consequentialism are sufficient to yield this principle in the form stated. When we add the interval-scale assumption, we may justify exchanges even when one person loses utility, provided the utility lost to one is more than compensated by the utility gained by the other. This conclusion is relevant only later in this paper.
Schwartz's claim now amounts to an argument that the principle of exchange conflicts with very basic intuitions about what is morally right and about how society ought to be organized. (At least, it would conflict with such intuitions if we were not brainwashed by social theory.) Thus, Schwartz argues that utilitarianism is part of the problem, not part of the solution.
It is argued here that utilitarianism, properly understood, is innocent of the charges that Schwartz has made against it. First, the argument just outlined interprets the idea of a `consequence' too narrowly, to include overt behaviors but not the inner states that behaviors often signify. Second, many of Schwartz's objections can be handled by adoption of a distinction between critical and intuitive moral thinking, as proposed by Hare (1981). Third, the idea of commensurability need not be given up, and it is in fact required if we are to make any sort of arguments about social justice. Aside from these three points, the positive case for utilitarianism is put aside (for that, see Hare, 1981).
It may seem that I am attempting to salvage utilitarianism by narrowing its scope and weakening its claims, and that further weakening will follow further attack, until nothing is left. Although this may be true (and without consequence except for the terms we use), I would argue that utilitarianism has indeed suffered through its association with certain behaviorist doctrines in economics and psychology. I thus agree with Schwartz that there is a serious theoretical problem in these disciplines, but I think the problem is a form of behaviorism, not utilitarianism.

1  Consequences and inner states

As Schwartz notes, there is something wrong with the idea that human relationships should be based on exchange. By this idea, it is perfectly fine for Dick and Jane to form a relationship based on a certain kind of contract, explicit or implicit. Dick will provide a steady income, do some household chores, participate in child rearing, and show certain outward signs of affection, such as bringing flowers home occasionally. In return, Jane will sleep with Dick (except when she truly has a headache), do most of the housework, raise the children, and provide similar signs of affection. Similarly, the relationship between employer and employee, between teacher and student, and between citizen and state (or fellow citizen) may be described in terms of this sort of contract. Such contracts may indeed be appropriate in the world of commerce, between consumer and producer, but they distort these other relationships, to varying degrees.
Why is this? Schwartz claims that there is a fundamental flaw in utilitarianism, which condones this sort of contractual relationship of exchange. Alternatively, let us ask ourselves what we really want out of these sorts of relationships, what our true preferences (in the sense relevant to utilitarianism) are. Do we want a spouse who performs spousely duties only because of the benefits we in turn bestow? Or do we want a spouse who acts out of love and honor (if not obedience!)?
Many of our true preferences concern not just the outward acts of others but also the motives behind those acts. We want a spouse who acts out of love. In fact, the outward acts are valued mostly because they are indications of this inner state. Literature is full of stories of lovers united by a love that transcends their mistreatment of one another. Of course, if claims of love are not sometimes supported by actions, they become empty and unbelievable.
Similarly, employers want employees who care about the enterprise they are engaged in, employees want employers who treat them as if they were this kind of employee, teachers want students who want to learn, students want teachers who want to convey their knowledge, the government wants truly loyal citizens, and the citizens want a truly caring government. If each person in such relationships is convinced that he is getting what he wants (true loyalty, etc.), he will be more willing to tolerate occasional lapses in behavioral manifestations. People do not like to have their behavior closely monitored; they would rather be trusted to carry out their role from a true sense of duty or loyalty. Hence the idea that professionalism, citizenship, etc., are important virtues.
Thus, I am arguing that the real problem is behaviorism, the confusion of inner motives with their outward expressions. Behaviorists think that what we really want from one another are expressions of love, loyalty, etc. They are wrong. What we want is the real thing. Where theories of human nature have distorted human nature itself is in telling us that what we want is just the behavior (or even more extremely, just its products). If human nature has changed as a result of these theories, it may well be in the acceptance of this idea.
It is not important here what loyalty, love, caring, etc., really are. What is important is that each is not identical with any particular list of behaviors - although it is of course true that we know the mental states and potentialities of others only through their behavior (this fact being the ball that behaviorists have grabbed and run with).
Because of the distinction between inner states and their behavioral manifestations, it is possible to speak of deception and of not knowing the true feelings of others. We do not want our spouses to pretend to love us, nor our students to pretend to want to learn. We cannot speak of such deception in the case of acts that are only overt behaviors. A pitcher cannot pretend to throw a strike.
Admittedly, some are satisfied with a show of affection (e.g., from a prostitute) or loyalty (from a sycophant), but the satisfaction they receive from this is ordinarily incomplete, or else the recipient has deceived himself - has made himself forget that the prostitute's signs of affection are occurring because he paid for them. Admittedly, behavior may matter on its own. People prefer a loving husband who cooks dinner to an equally loving husband who doesn't. Finally, there is surely some truth in the saying, `Where the hand leads, the heart follows.' What begins as a polite show of affection in order to justify physical intimacy might turn into the real thing. However, these admissions recognize, even require, a distinction between the pretense and reality of certain inner states, which are often the true objects of our preferences and desires.
My argument so far implies that we could apply utilitarian calculus, and the exchange principle, to the satisfaction of these true desires, so long as they are not confused with their behavioral manifestations. Utilitarianism might justify market-like transactions in which Dick gives Jane his true love in return for hers.
Once again we come up against certain facts about human nature. Just as our true preferences concern inner states rather than their outward manifestations, so it is a fact that these states cannot be turned on and off entirely at will. Some of these states are what Elster (1983) calls `essentially bi-products.' If we want to bring them about, we cannot aim directly at them. Most of the time, the best we can do is to set up the conditions under which they might occur (e.g., dating as a condition for falling in love). Arguably, this limited control is a necessary property of the states we desire in others. If control were perfect, these states would not be as desired as they are (perhaps because the supply would exceed the demand, if for no other reason).
Once again, behaviorism has misled us, this time by assuming that everything involved in a transaction or exchange is ours to give or take freely, just as a pigeon may choose to peck a key or not. (Skinnerians would say that the relevant behavior is `under operant control.') Nor can it be argued that behaviorism combines naturally with utilitarian assumptions as opposed to others. For example, a rights-based legal system could ignore inner states in decisions about sanctions. Such a system would be wrong for the same reason that the economic view Schwartz criticizes is wrong.
I conclude that the principle of exchange is useless for many of our most important concerns. It does not follow from the basic assumptions of utilitarianism because it ignores, first, the fact that our preferences often involve inner states, not behaviors, and, second, the fact that our control over these states is limited. Because of these limits, the exchange principle is of limited utility, so to speak, in maximizing utility. Market exchange may best be limited to those relatively few goods and behaviors where the behaviorist assumptions apply, that is, consumer goods and services of the usual sort. If the market is to be defended at all as a basis of other human relationships, it must be by showing that it is better than the alternatives, despite these difficulties.

2  Critical and intuitive thinking

So far, I have argued that the real problem is not in the utilitarian justification of the exchange principle as a full normative theory of human relationships, for the exchange principle cannot provide a full theory. (Only a behaviorist could think that it might.) But we might apply utilitarianism directly to justify what we might call prostitution in the most general sense. Could it not be argued that people should make the kinds of exchanges at issue (e.g., weapons for hostages, lab space for credit) when utility is maximized? If so, this conclusion conflicts with our intuitions and argues against the utilitarian use of the commensurability (interval scale) assumption. To deal with this problem, we need to step back a bit.

2.1  Critical vs. intuitive thinking

Many of the conflicts between intuition and utilitarianism can be understood in terms of Hare's (1981) distinction between critical and intuitive moral thinking (see also Baron, 1985, 1986). Critical thinking is the application of utilitarian calculus to individual acts and to the evaluation of moral rules to be followed at the intuitive level. Intuitive thinking is the following of moral rules and intuitions.
Some of the rules to be followed at the intuitive level concern when to revert to the critical level. In general, the critical level should be avoided when miscalculation is likely and when it could lead to errors with serious consequences, or when there is strong temptation to rationalize action in one's self-interest. For example, those contemplating political assassinations would do well to avoid critical reasoning; even though such acts may appear justified, the moral track record of people who have made such calculations is quite poor. On the other hand, the critical level may be useful in the dispassionate analysis of public-policy questions in which one does not have a particular stake.
Act-utilitarianism is defensible only at the critical level (Hare, 1981). Assuming such a defense, the question of what rules should be followed at the intuitive level is still a question of design or invention. Different personal, material, and cultural circumstances may require different moral rules, different ways of living. We may invent new moral rules, just as we invent new methods of communication or education. The success of our moral inventions are to be judged by utilitarianism applied to them at the critical level, but utilitarianism provides no procedure for invention itself. Most moral philosophies are meant for the intuitive level. They concern how we actually ought to live.
The critical-intuitive distinction is illustrated in its handling of the classic dilemma in which you have an inescapable choice between killing one person or letting ten others be killed by someone else. Here, killing through an act of commission violates a very strong intuitive principle. One's role as a causal agent, one's agency (Sen, 1985), is intuitively relevant. This intuitive principle conflicts with the principle that one should save as many lives as possible, leading to the dilemma. The intuitive prohibition against active killing is clearly a good one to hold. It prevents killing, and if everyone held it, this kind of dilemma could not arise. Because it is a good intuitive rule, most of us have strong preferences not to violate it, and any violation would make us feel guilty. In this light, a person who declines to participate in killing on grounds of agency, when the preferences of others would be better satisfied if he did participate, is indulging a personal preference at the expense of others. The situation is complicated because this personal preference arises from intuitive moral rules that are themselves worth following. None the less, a person who refuses to `get his hands dirty' for a good cause may be seen as self-indulgent. In the same way, so might the professor who refuses to pass the quarterback.
An important feature of the critical-intuitive distinction is that critical thinking must deal with hypothetical examples like this one, while intuitive principles need not, exactly because they are rare. Intuitive principles are meant to serve us in our ordinary lives, where such situations do not often arise. Hence, it is easy to find conflicts between our moral intuitions and the dictates of critical moral thinking for hypothetical examples.

2.2  Purity of practices

Schwartz's recommendations draw heavily on MacIntyre's (1981) idea of practices. A practice is a culturally-defined domain of activity that sets its own standards. Examples are cooking, chess, sports, education, science, music, commerce, citizenship, and friendship. Once people are initiated into a practice, they take the goals of that practice as their own goals. When they import the goals of some other practice, the original practice is weakened. Thus, the scientific pursuit of truth (a goal of science) is weakened when a scientist distorts his results for material gain (a goal of some other sort of practice). The practice of education is weakened when professors give passing grades to athletes who don't otherwise deserve them.
Schwartz argues that the purity of practices ought to be preserved. His criticism of modern U.S. society is that certain standards of the market - particularly the goal of getting the most gain for the least sacrifice - are infiltrating other practices and weakening them.4 Utilitarianism also weakens practices because it permits their rules to be violated when utility is maximized by doing so.
This argument is asserts that we should follow `lexical rules,' that is, rules in which one consideration is given absolute priority over another. Such rules lead to ludicrous consequences when applied in hypothetical probabilistic situations or situations in which many people are involved (Baron, 1986). For example, suppose that the issue were not whether to fail a particular quarterback but whether to give professors enough discretion so that some professor, some year, might give a free point or two in order to see the team win. If we were to give academics absolute priority over sports, we would make sure that professors are prohibited from reading the campus newspaper lest they find out what their students do outside of their classes.
We may conclude that Schwartz's principle of purity cannot be maintained at the critical level. Let us therefore treat is as an intuitive principle. Here, there are strong arguments for it. Practices are institutions (like the idea of promise keeping) that benefit us all. Impurity sets a precedent for future violation (by the violator or others), and the long-run benefits of the practice may be weakened or jeopardized. The disutility of such weakening is only rarely compensated by other gains, so the purity principle usually coincides with the dictates of critical analysis. Moreover, violations of the purity principle may occur in just those cases in which critical thinking should be avoided: where self-interest is involved, etc.
However, these arguments might not be decisive. An alternative intuition is that some of our practices are, in a sense, too pure. Baron (1986) proposed an idea of simple moral systems, in which people wall off the moral part of their life from the rest. Outside of a few specifically moral habits such as giving to charity and going to church, many people limit moral considerations to a list of prohibitions, e.g., those against cheating or stealing. Such prohibitions are often found within practices such as commerce or science. In these contexts, however, they represent only a minimal effort to incorporate moral standards into the practices. It may be argued, for example, that scientists ought to pay more attention to the potential consequences of their work when they choose a problem to work on. For example, the successful study of lasers seems likely to lead to mixed blessings, while the study of immune reactions is likely to lead to pure ones.
The decision between these two intuitions is a question of which does better in our world. It has been argued (against my view) that the importation of moral considerations into science will open the floodgates to all sorts of other `non-scientific' considerations such as individual financial gain, governmental use of science policy to remedy racial injustice, etc. Ultimately, it is argued, the institution of science will crack or crumble, so that it is unable to serve its most unique purpose, the search for truth. Such a consequence would be bad by any theory. At issue is whether this risk is sufficiently great.
A reply to this argument (following Baron, 1986) is that it is impossible in today's society to convince people to maintain the purity of institutions at all costs except through a kind of brainwashing, unless they themselves understand the moral justification of our rules at the critical level. Given such understanding, the importation of moral considerations into the standards of other practices is easily justified at the critical level, and it poses little danger of inspiring the importation of other destructive considerations, insofar as it will also be clear that such importation cannot be critically justified.
If Schwartz is correct in his argument about appropriate intuitive rules, it may be best to assume as an intuitive precept that there is no useful common coin, no commensurability, and no such thing as utility, even if, at the critical level, people might admit that these precepts are wrong. Indeed, we admire people who seem to adopt precisely such precepts: the explorer who perseveres despite the risk to his men, the professor who fails the quarterback (who is one point short of passing). Such people typically do not defend their actions in terms of utility maximization (although they might, if they were to consider weakening an institution as a consequence of giving in). Rather, they argue that academic standards, for example, are not commensurable with the standards of sports, and therefore these standards must not under any circumstances be traded off with any other standards. Even if such an argument is wrong, it might be best for all if people believed it.
My reply is, again, that such an argument is not believable. Incommensurability does not imply that one consideration should be given priority over another. It implies that it is impossible to say which of two considerations should be prior in general or in a given case. Those who argue that a human life cannot be given a monetary value must give some further justification if they want to argue for the preservation of a particular life at great cost. They cannot argue - as the utilitarian can - that the life in question is worth more than the money. (A utilitarian can even maintain that any life is worth more than all the money in the world, if he really believes it.)
In sum the critical arguments against the intuitive principle of purity are apparent to thoughtful people in our society. Hence, it is essentially impossible to convince people that they ought to follow lexical rules or preserve perfect purity. Any attempt to build an intuitive moral theory on the belief in utter incommensurability of different values will not work. Moreover, even if people could be convinced, should educators withhold critical arguments in order to insure that their students have appropriate (but possibly wrong) beliefs? Such a stance would seem to contaminate the practice of education. Put another way, the utility of the open search for enlightenment is too high for it to be intentionally subverted.

3  Commensurability: further issues

Some further arguments for commensurability at the critical level may be made. There are two issues: commensurability across people and commensurability across domains of activity such as practices, i.e., commensurability across different goals, within or across individuals. If inter-domain comparison is impossible, then interpersonal comparison is generally impossible as well, since different people may be involved in different domains of activity.
Clearly, inter-domain comparison is difficult. It is difficult to compare the harm done to sports by failing the quarterback and the harm done to the academic enterprise by passing him. But it need not be impossible.
We may put this difficulty in perspective by considering the fact that many other comparisons are difficult although possible. It is somewhat difficult to compare the difference in brightness between a 75 watt bulb and a 100 watt bulb to the difference between a 25 watt bulb and a 40 watt bulb, and it is more difficult to compare the difference in brightness of two lights with the difference in loudness of two tones. However (practical problems aside), so long as it is possible to elicit repeated judgments, so that the variability averages out, and so long as it is possible to define a relevant context for eliciting the judgment, comparisons like these may be meaningfully made.
Note that the difficulty of inter-domain comparisons is often exaggerated by the use of poorly specified examples. For example, it may be argued that it is difficult to decide whether to give $50 to charity or go out to dinner (Schwartz, 1986, p. 154). This example does not specify which charity, how needy that charity is at the time, how much we have given to charity recently, which restaurant, how long it has been since we have gone out to dinner, etc. There are many ways of filling in these details so that the choice would not be difficult at all. The appearance of difficulty in comparison may result largely from ambiguities in the states (or differences) being compared.
The fact is that we make inter-domain comparisons all the time through our decisions. We do, for example, make choices between personal relationships and careers when we decline (or accept) a job because we want to stay with (or are willing to leave) a potential spouse. Often such choices are painful and difficult, both because they involve uncertainties and because of the essential difficulty of inter-domain comparison. We worry that if we take the job, it will not turn out well, or that if we decline it, the relationship will not work out. But we do decide, sometimes easily.
The cases in which people feel most resistant to making inter-domain tradeoffs seem to be those in which they see one of the two domains as much more important. Some opponents of abortion regard human life as so important that they are unwilling to trade it for any benefit even when the life is itself of questionable status. Although such unwillingness may be defended through some sort of argument about incommensurability, this is not the only defense. An alternative defense (as noted earlier) is that the domains are in fact commensurable and that the life domain has so much more importance than any other (in terms of the common measure) that no amount of benefit on other dimensions is equal to the value of a single life. This is because the other dimensions are inherently bounded on the scale of importance. The critical test of such a belief would involve the choice of actions that might cause an abortion with some very low probability, e.g., a medical treatment for a dangerous brain infection in the mother, which would increase the probability of miscarriage by .00001. If an opponent of abortion assents to such a treatment, he is implicitly accepting commensurability, and the possibility of a tradeoff (Baron, 1986).5
If inter-domain comparison is possible, we may ask whether interpersonal comparison is possible as well. Many arguments for such comparisons have been made (e.g., Hare, 1981, ch. 7; Parfit, 1984, ch. 15; Baron, 1986), but perhaps one more thing needs to be pointed out. If we did not implicitly make such comparisons, we would not be able to judge that any distribution of goods is too extreme. We would not be able to say that one person is so much richer than another that the world would be a better place if the former were $100 poorer and the latter $100 richer. Such a claim involves an interpersonal comparison of differences. In fact, we make these kinds of statements all the time. Although interpersonal comparison is difficult, it does seem to be possible.

4  Summary

The principle of exchange seems to be limited in its application, and it cannot serve as a link between utilitarianism and the idea of a market for interpersonal relations. Our preferences concern the inner states of other people as well as their overt behavior. The neglect of this aspect of our preferences is a result of the coupling of utilitarianism with behaviorism. The problem is thus behaviorism, not consequentialism.
It might be argued that commensurability is wrong because it sanctions impure practices - aside from any principle of exchange. In reply, it may be argued that the principle of purity is an intuitive one, which must be justified at a higher level. At the intuitive level, the purity principle competes with the view that practices are in some cases too pure already, and that blind purity cannot be maintained without educational deception.
The apparent difficulty of making tradeoffs among different kinds of consequences is a problem in practice, not a problem in principle. The difficulty may indeed justify certain prescriptive rules, but a blanket prohibition against impure practices would seem to require considerable defense.

5  References

Baron, J. (1985) Rationality and intelligence. Cambridge: Cambridge University Press.
Baron, J. (1986) Tradeoffs among reasons for action. Journal for the Theory of Social Behavior, 16, 173-195.
Becker, G.S. (1981) A treatise on the family. Cambridge: Harvard University Press.
Bellah, R.N., Madsen, R., Sullivan W.M., Swidler, A., & Tipton, S.M. (1985) Habits of the heart: individualism and commitment in American life. New York: Harper & Row.
Brown, R.V., Kahr, A.S., & Peterson, C.R. (1974) Decision analysis for the manager. New York: Holt, Rinehart, & Winston.
Elster, J. (1983) Sour grapes: studies of the subversion of rationality. New York: Cambridge University Press.
Foot, P. (1967) The problem of abortion and the doctrine of the double effect. Oxford Review (5).
Hare, R.M. (1963) Freedom and reason. Oxford: Clarendon Press.
Hare, R.M. (1981) Moral thinking: its levels, method and point. Oxford: Clarendon Press.
Homans G.C. (1958) Social behavior and exchange. Journal of Sociology, 63, 597-606.
Keeney, R.L., & Raiffa, H. (1976) Decisions with multiple objectives. New York: Wiley.
Leung K, & Bond, M.H. (1984) The impact of cultural collectivism on reward allocation. Journal of Personality and Social Psychology, 47, 793-804.
MacIntyre, A. (1981) After virtue. Notre Dame: University of Notre Dame Press.
Parfit, D. (1984) Reasons and persons. Oxford: Clarendon Press.
Schwartz, B. (1986) The battle for human nature: science, morality, and modern life. New York: Norton.
Sen, A. (1985) Well-being, agency and freedom: The Dewey Lectures, 1984. Journal of Philosophy, 82, 169-221.
Sen, A., & Williams, B., Eds. (1982) Utilitarianism and beyond. Cambridge: Cambridge University Press.
Walster, E., Walster, G.W., & Berscheid, E. (1978) Equity: theory and research. Boston: Allyn & Bacon.


1This is a pre-publication version of Baron, J. (1988). Utility, exchange, and commensurability. Journal of Thought, 23, 111-131. It is similar to the published version.
2Utility is not always defined in terms of preference satisfaction, although most recent versions of utilitarianism have adopted this view, e.g., Hare (1981). I shall assume here that other versions of utilitarianism may be translated into this one.
3A free market will generally fail to maximize utility, for a variety of reasons, e.g.: it does not provide a mechanism for helping the poor (who might gain much utility at a small cost to others), it fails (by itself) to insure cooperative behavior (from which all might benefit), and it provides no protection against irrational neglect of future preferences of participants or their progeny. However, these issues are not in dispute, but rather the claim that utilitarianism justifies the market, other things equal.
4Interestingly, the practice of the market itself is undermined by excessive reliance on this goal, for it has other standards, such as accurate representation of one's goods and adherence to one's word.
5Some would allow abortion only when it is an unintended or undesired risk (Foot, 1967). As a basis for making decisions (as opposed to a basis for evaluating the decision maker), the role of intention seems justifiable only prescriptively (not normatively). See Baron (1986).

File translated from TEX by TTH, version 3.59.
On 13 Apr 2005, 05:18.