Version classiqueVersion mobile

Understanding History

Jonathan Gorman

Chapter 5. Rational Economic Man

Texte intégral

1Claims to knowledge, including claims to historical knowledge, require rational justification if they are to be admitted. The standard for rational justification which has been conditionally adopted in this book is empiricism, but we have seen that empiricism needs to be supplemented by background metaphysical assumptions — assumptions not themselves based on experience — which tell us what experience (particularly, historical experience) is experience of. In chapter 4 we introduced the background metaphysical assumptions associated with the traditional narrative form of historical writing involving the recovery of meaning from written records. This metaphysics permits at the minimum two kinds of entities: those which are regularly ordered (causally related) in the natural world, and individual persons. Each has its associated theory of explanation: for the natural world, covering law theory, and for individual persons, empathetic understanding. Historical writing, on the traditional view, will be written in the light of such assumptions. It will be a contribution to knowledge, first, insofar as it correctly applies those assumptions, and, second, if those assumptions are themselves rationally justified.

2In chapter 1 we compared and contrasted in an elementary way the traditional approach to history with the econometric approach. In this and the next chapter we shall examine the econometric approach, and do so by considering whether the question “Was slavery profitable?” is to be admitted as a proper question within the traditional metaphysics of human nature now outlined. If this question is wholly admissible within the traditional metaphysics, then this will show that the differences between the traditional and the econometric approaches to history are not metaphysically fundamental: that, through sharing the same assumptions about historical reality and thus complementing each other, both approaches can in principle make contributions to historical knowledge, so long as the traditional metaphysics itself is rationally justifiable. For historians, this final question of the justification of the traditional metaphysics would then only be of abstract interest, since a decision on the matter would not affect the choice of “approach” they actually had to make. On the other hand, if the question “was slavery profitable?” is not to be wholly admitted as a proper question within the traditional metaphysics of human nature, then this will show that the traditional and econometric approaches to history make different metaphysical assumptions. If this is so — as we shall see that it is — then the question of the justification of one’s choice of metaphysical assumptions becomes crucial to historical practice. This last issue will be considered in chapter 7.

3Is the question “was slavery profitable?” to be admitted as a proper question within the traditional metaphysics of human nature? “Slavery” does not name a person or a set of persons, but a practice or social institution: it summarizes in some complex way a number of actions of a similar kind. For the question “was slavery profitable?” to be admissible, we shall need to make general sense of practices and other social institutions within the terms of the traditional metaphysics.

  • 1 Other entities may also be admitted to reality, of course. God, for example.
  • 2 The relationship between experience and background assumptions even in “direct experience” is neve (...)

4One way in which sense can be made of such social matters is by admitting them to reality as a third kind of fundamental entity, in addition to natural objects and individual persons.1 It would then be supposed that such social entities as “slavery” have a life or causal process of their own. While this approach — a major contrast with the traditional approach — is metaphysically possible, our conditional adoption of empiricism precludes it. Given empiricism, we are required to claim no more than is empirically warranted, unless we are forced to. There is a clear sense in which we can directly experience individual people while we cannot directly experience social institutions.2 Empiricism, and the traditional approach to history, require us, therefore, to make sense of social matters in terms of those relevant observable entities which we have already admitted to reality, namely individual persons. Slavery must thus be taken to be a set of related actions undertaken by individual people, and it is to be explained only in terms of the understanding of those individuals.

5Individual persons have been taken to be empathetically understandable. They do not have to be understood in this way; as Dray observed, we give reasons if we can, and turn to empirical laws if we must. However, the traditional metaphysical position may seem to be ambiguous: should we turn to empirical laws only if we must, or not? We shall adopt an interpretation which insists that we turn to empirical laws only if we must, for this seems best to reflect that attitude towards human understanding embodied in the traditional approach. To take human beings seriously as persons requires an empathetic approach. On this view, individuals must be understood, where possible, in empathetic terms, and so, it follows, must the social practices and social institutions which comprise those individuals. An alternative approach to human nature and history will be discussed in chapter 6.

  • 3 Douglass, Narrative, p. 87.

6We have to ask whether “profitability” can be found a place within the traditional metaphysics interpreted in this way. Frederick Douglass noted that “some slaveholders thought it not much loss to allow Mr Covey to have their slaves one year, for the sake of the training to which they were subjected, without any other compensation.”3 There is no doubt that we can readily empathize with a slaveholder who acts for profit or to avoid loss, and so there is no need to fall back on empirical laws here. (To “fall back on empirical laws here” would amount to understanding people mastered by the profit motive in the same way as if they were under the influence of drink.) “Profitability,” empathetieally understood, may thus be straightforwardly admitted to the traditional metaphysics. We may conclude that the question “was slavery profitable?,” if interpreted in the ways just described, is an admissible question within the traditional metaphysics.

7If we insist that social reality be understood, where possible, in empathetic terms, then the profitability of slavery is not to be understood in abstraction from the individuals engaged in it. Slavery was profitable if the individual actions involved were profitable, and the individual actions were profitable insofar as they could be empathized with as such. “Profit,” “loss” and associated notions therefore appear, in the present interpretation, not as existing entities in their own right, but as objects merely of belief or of purpose. Some slaveholders “thought it not much loss” to act in a certain way, and hence they so acted.

8In chapter 1 we suggested how an argument about the profitability of slavery might be used: if slavery was profitable, then “we are to understand why the slaveholders engaged in and continued with slavery until they were forcibly stopped. It was in their interests to do so.” In this chapter we shall introduce one important problem with this explanation. The traditional view holds that individuals are influenced by many “constantly shifting forces” in addition to, or as an alternative to, any desire for profit which they may have, and individuals may well deliberately act otherwise than where they perceive their profit to lie. The mere fact (if it is a fact) that slavery was profitable would not then explain the actions involved, for we would still need a further explanation why the slaveholders all selected profit as a motive rather than acted on the basis of one of the many other possible desires. It is the imperfect rationality and many-sidedness of human beings which makes “slavery was profitable” unacceptable as a complete explanation in terms of the traditional theory of human nature. Knowledge, on this traditional view, involves empathy with real individuals in all their complexity.

9There is no doubt that we may want more than just an explanation of many individual actions in terms of the desires and beliefs the individuals involved actually had. Having empathized, in the way just mentioned, with a slaveholder who acted for profit, we may still want to know why he acted for profit rather than, for example, out of respect for a moral duty to treat his slaves well. We can provide the beginning of an explanation here: via the traditional approach, we merely delve a little deeper into the slaveholder’s character and work through a calculation displaying the appropriateness of his choice of the profit motive, in addition to working through a calculation displaying the appropriateness of the profitable action itself. At the point of empathetic understanding at which we stop, we simply know what kind of person the slaveholder was. We may discover that, in his choice of ends, he is ultimately to be understood as an evil person. On the present view there are moral kinds of people, and no explanation is to be sought for this. The differences between people ultimately lie in themselves, and are apparently not susceptible of further explanation. People do not have to be as they are; they just are what they are.

10Slavery, however, existed as a way of life. Apparently an entire slaveholding class all acted for profit, even though it conflicted with what we may suppose to be a natural morality. The traditional metaphysics requires us to say that we cannot go on asking for a cause of every choice: if some people in certain circumstances act out of respect for profit instead of morality, then that is just the way they are, and the rest of us just differ (we can empathize with them all). But this ultimate appeal to the varied nature of individuals (“it takes all sorts to make a world”) does not fit a situation where large numbers of people behaved in a certain way, in being inclined to act from one principle of action rather than another.

11Either individuals did not, in the slaveholding class of the white South, share the complexity of human nature supposed on the traditional approach (which offends against the metaphysics, and requires further explanation), or they indeed had a complex nature, but their many different desires and purposes were massively overridden by something — something which overwhelmed the usually supposed moral differences between people, and overwhelmed the many different beliefs and motivations which usually guide people in different ways (which again offends against the metaphysics, and requires further explanation). One may blame a slaveholder for acting as he did, but praise and blame are only easily applicable notions when some people deserve the one and some the other. They lose their meaning when, as may be, a whole way of life is at fault. Individual responsibility is then lost; and with it, the existence of the slaveholders as complete and undetermined empathetically understandable moral individuals (which, again, offends against the metaphysics).

12We note that the slaveholders displayed an unindividualistic similarity: they formed a kind of person, with characteristic actions distinctive not of all humankind but distinctive just of a particular time and place and social system. Such kinds of thing, we recognize, need explanation. Perhaps individual actions can be explained by empathetic understanding, but how are we to account for the similarities between slaveholders, which made their actions into a way of life? Why did so many act for perceived profit and not from respect for other human beings?

13We therefore do not yet have a sufficient explanation. “Sufficiency” of explanation, however, is a misleading notion. We must distinguish between an objective and a subjective sense of sufficiency, as follows. Objectively, an explanation of some event or action E specifies a sufficiency if it provides us with another event C, the occurrence of C being sufficient for the occurrence of E. To say that C is sufficient for E, in this sense, is to say that whenever C then E. Demand for a sufficient explanation may then be interpreted as a demand for an explanation involving a regularity — in other words, we have here an insistence upon covering law theory. Thus the desire, which we all share, for a “sufficiency” of explanation may make plausible the view that, fundamentally, we should understand people in mechanistic terms. It may suggest that the point has come to fall back on empirical laws, not just occasionally, but for all explanation and understanding.

14But sufficiency is better understood in a subjective sense, as follows. The demand that an explanation be sufficient is, in a subjective sense, no more than a demand that it provide whatever is sufficient for us, the hearers, to understand the thing explained. We understand something just insofar as our puzzlements are removed. Against covering law theory, we often are satisfied with a “calculation” covering a person’s action. A “calculation” is often sufficient. But, against the point drawn from Dray’s Laws and Explanation in History, more than his notion of empathetic understanding is required for the present problem of why so many acted from a desire for perceived profit rather than from respect for others.

  • 4 There is a tradition, deriving from the work of Wittgenstein, that philosophical problems only ari (...)

15There are two ways of dealing with this question. One way is to provide some further level of explanation which accounts for why or how some motives come to dominate others for so many individuals. Empathetic understanding, which rests with given motives, cannot succeed here, and yet the required further level of explanation cannot be causal, for, if we were caused to have the motives we have, this would characterize persons as essentially deterministic entities, which would offend against the traditional metaphysical assumptions. There is little prospect of success in this causal direction within the traditional approach. A second way of dealing with the question is to dissolve it, so that it does not need an answer. Such dissolution would remove our puzzlement, while we would retain the view that individuals must be understood in empathetic terms. A digression is in order here, to show how questions can come to be dissolved. It is an important move in philosophical discussion.4

16Newton’s first law of motion is as follows: every body perseveres in its state of rest, or of uniform motion in a straight line, except insofar as it is compelled to change that state by forces impressed on it.

17Suppose we fire, from our space station, a rocket, the fuel of which is soon expended. We observe its path through space. It may well be in a straight line, and we shall suppose that it is very far from anything which might exert some force upon it. The rocket is behaving in accordance with Newton’s first law. Let us ask why it is travelling in a straight line, once its fuel is expended. This is not a proper question: given the law, it is simply natural that it do so. There is no explanation to seek. Suppose, however, that the rocket’s path is a curve. Given Newton’s first law, we know that some force must be impressed upon it. It is proper to ask what that force is. When a body travels in a straight line in space, there is nothing to ask. When it travels in a curved path, there is something to ask.

18Return to a pre-Newtonian approach, in particular one which understands the earth to be the centre of the universe, and the heavens, in an astronomical sense, to be the domain of God. It was commonly believed, partly on the basis of observation, that the natural path of a heavenly body, being naturally perfect, was a perfect circle. This being so, there is no question to ask as to why some heavenly object should travel in a circular path: it is simply natural that it do so. On the other hand, should such an object travel in a straight line (for example, a harbinger of evil like a comet), then we must seek an explanation. On this Ptolemaic understanding, when a body travels in a straight line in space, there is something to ask. When it travels in a curved path, there is nothing to ask.

19In this simple example we can see how a question can depend for its existence on some presupposed belief, and we can see also how a question can be dissolved by moving from one system of beliefs to another. A child may be puzzled by both questions about curved and straight motion, but whichever system he learns will show him either how to answer a question, or how to dissolve it. In that sense each astronomical system of beliefs is minimally complete.

20Our metaphysics of human nature must either permit an answer to the question why so many acted for profit rather than morality, or show that the question can be dissolved. Otherwise we find that the question “was slavery profitable?” is indeed not wholly admissible within the traditional metaphysics. Underlying the econometric approach which was mentioned in the first chapter is a claim that people always, in a sense, act for profit. If we adopt a theory of this sort, then we will be able to dissolve the question. We shall explain and examine the foundations of this econometric approach by referring to one of the examples presented by the economist Milton Friedman. It should be clear that we shall be referring to Friedman’s philosophy of economics, and not to his economic theories themselves.

  • 5 Milton Friedman, “The Methodology of Positive Economics,” originally in his Essays in Positive Eco (...)
  • 6 Friedman, pp. 30, 32. For a detailed discussion of Friedman’s views, and an excellent introduction (...)

21Friedman asks us to think of a tree, with leaves, in the sunlight. Consider the density of the leaves around different parts of the tree. “I suggest,” says Friedman, “that the leaves are positioned as if each leaf deliberately sought to maximize the amount of sunlight it receives, given the position of its neighbours, as if it knew the physical laws determining the amount of sunlight that would be received in various positions and could move rapidly or instantaneously from any one position to any other desired and unoccupied position.”5 This suggestion — Friedman stresses that it is no more than that, a mere hypothesis — does not state that leaves deliberately act in these ways, but rather, as he again stresses, as if they did. It is only a hypothesis, but it is a plausible one: perhaps leaves do behave “as if” they had such knowledge and skill. Given this, we can predict how the leaves will move as the sunlight alters in intensity or direction, and we will be right. It is in such a sense that we may treat individual people as so-called “rational economic men.” Thus we have the “economic hypothesis that under a wide range of circumstances individual firms behave as if they were seeking rationally to maximize their expected returns (generally if misleadingly called ‘profits’) and had full knowledge of the data needed to succeed in this attempt.”6 A rational economic man is, first of all, rational. “Rationality” here is understood in a special way, so that the following are true:

(1) For any individual A and any two options X and Y, one and only one of the following is true: A prefers X to Y; A prefers Y to X; A is indifferent between X and Y.
(2) A’s preferences among options are transitive. (This means that, if X is preferred to Y, and Y is preferred to Z, then X is preferred to Z; and if A is indifferent between X and Y, and indifferent between Y and Z, then A is indifferent between X and Z.)
(3) A seeks to maximize the utility of his or her options, where the utility of an option X is greater than the utility of an option Y for A if and only if A prefers X to Y. The utilities of options are equal just in case the agent is indifferent between them. (In other words, more of what you think is good for you is always preferred to less.)

22One becomes rational economic man rather than just rational man simply by recognizing that among the options available to us are the acquisition of commodities, so that the following is true:

  • 7 There are a number of ways of expressing the assumptions of rational economic man theory. The pres (...)

(4) If option X is acquiring commodity bundle X’ and option Y is acquiring commodity bundle Y’, and if Y’ contains at least as much of each commodity as X’ and more of at least one commodity, then everybody prefers Y to X.7

23Finally, the following assumption is made:

(5) The more you have of something, the more of it you will be willing to give up in order to get something else you want. (An additional bag of gold means less to a millionaire than it does to a beggar, for example.)

  • 8 C. Dyke, Philosophy of Economics, Prentice-Hall, 1981, pp. 35, 51.

24If you imagine the beggar becoming richer bit by bit, then each additional bit of wealth he has means that little bit less to him. Each additional bit of wealth occurs at the “margin” of the wealth that the beggar, at that point, already has. Economists thus call this situation the “diminishing marginal utility” of a commodity. Our assumption (5) may thus be expressed as: the marginal utility of any commodity diminishes as the quantity increases.8

  • 9 Referred to in George J. Stigler and Gary S. Becker, “De gustibus non est disputandum,” The Americ (...)

25Following the argument of Alfred Marshall, we should note an additional tacit assumption which modifies assumption (5). “It is that we do not suppose time to be allowed for any alteration in the character or tastes of the man himself. It is therefore no exception to the law that the more good music a man hears, the stronger is his taste for it likely to become....”9 Parallel assumptions are required for the proper understanding of assumptions (1), (2) and (3), also. Thus (1) holds that it cannot be the case that A both prefers X to Y and Y to X. But we do not wish, by this, to rule out A’s changing his mind, and preferring X to Y at one time, and Y to X at another. Assumptions (1), (2), (3) and (5) take a person’s preferences as given. Nothing here requires us to have some preferences rather than others, in order to be rational, and nothing requires us to be consistent over time in what we prefer.

26The specifically economic assumption (4), however, may be thought to be different from the others. It states that everybody, in certain circumstances, prefers option Y to option X, where Y is the acquisition of a certain commodity bundle Y’. But what if it is, say, illegal to acquire Y’ over X’? Here we should attend to the exact form of assumption (4). It is a conditional statement: it has the form “if P and Q and R, then S.” Such a statement can be true even though P and Q and R are false. Thus this book, if it were a raven, would be black. This statement is true; that this book is not a raven is irrelevant to its truth. Option Y is defined, in assumption (4), to be the acquisition of commodity bundle Y’. If one wishes to state that the situation is one of acquisition of Y’ together with a risk of punishment (owing to the illegality of the acquisition), then one no longer has option Y as defined, and assumption (4) says nothing about the resulting preference.

27If a businessman were to choose X over Y, in such circumstances, then he would not irrationally be preferring commodity bundle X’ to commodity bundle Y’(where X’ contains less than Y’), but rather (on the basis of assumptions (1), (2) and (3)) rationally preferring X’ to Y’, including in the latter the risks associated with its acquisition. The cost of a commodity is to be understood in terms of what has to be given up for it. Maybe it costs $100 (and thus other things costing that amount are foregone), but the price, in terms of our assumptions, is more than that, for it includes also, for example, the risk of punishment if an illegal transaction is undertaken, or even the opportunity cost of the time required to justify to one’s family that it was worth buying. X’ and Y’ thus each express a net evaluation, which is the result of a calculation of the desire for a perceived benefit set against the perceived associated cost. The preference for X’ over Y’ is similarly a net preference, the result of the total calculation.

28It is easy, therefore, to hold assumption (4) to be true, for it does not imply that there ever is a choice between options which are nothing but the pure acquisition of different commodity bundles. It just says that, if there were such a choice, then it would go to more rather than less. Nevertheless, while one might wish to stress that real life options are more complex than those of pure acquisition (and include assessments of risk or reputation or love, for example), the point of (4), when used by an economist, is to hold that indeed we may, without loss of significant truth, treat many choices as if they were just matters of pure acquisition or disposal. It is plausible for the economist to say this because we live in a society where there is a large element of freedom to trade. In a free market, constraints or options other than the simple choices of acquisition or retention or disposal of commodity bundles may be ignored as irrelevant. This defines, in fact, what a free market is. This is why we described assumption (4) as “specifically economic,” for it determines the scope of the subject matter of economics.

29It should be noted that, in current political debate, the notion of a “free market” is less clear than this. Strictly, it means freedom to trade where the only options are between commodity bundles. In practice, there is no such thing. Every exchange carries consequences of a different sort, and the law regulates carefully the existence and activities of firms and their agents. Even without the law, matters like reputation and perceived generosity and revenge may determine preference, independent of the size of the commodity bundles involved. A situation is described, politically, as a “free market” when there is minimum government interference with the freedom of exchange of commodities. The economist claims to be able to model such exchanges using the assumptions now specified.

30There are, at present, a number of attempts by Western governments to extend the assumptions specified by the concept of “rational economic man” into areas beyond the free exchange of commodities, into, for example, law or health care or education or the provision of information. Whether health care, say, should be treated as a commodity is a topic outwith the scope of this book, but it should be noted that the notion of “freedom” disappears, if we understand a free market to be one where every option is freely exchangeable or chooseable (not, that is, merely those options concerned with the acquisition of commodity bundles in the ordinary sense). Every state of affairs is then a free market. Thus, in a central State regulated economy, such as used to exist in the Soviet Union, there is a free market, in that one always has a free choice whether to follow instructions, or to trade in a capitalistic manner and be imprisoned for it. In effect, political freedom — the absence of unjustified State interference — is here assimilated to mere freedom of the will, where man is free in everything he does, even in a labour camp or a slave plantation. This is not a proper interpretation of the notion of a free market. What ought to count as a “commodity” — slaves? — is a political issue, which is not determined by assumption (4).

31Following Milton Friedman, the hypothesis is that people behave as if they were rational economic men. The concept of “rational economic man” is specified by our assumptions (1) to (5). Thus the hypothesis is that people behave as if assumptions (1) to (5) were true. Thus people behave, it is hypothesized, as if they preferred one thing to another, in accordance with assumptions (1) to (5). But what is it to behave as if you preferred one thing to another?

32The formulation “A behaved as if he preferred X to Y” implies nothing about whether A really did prefer X to Y, We would use such a formulation, appropriately filled in, in ordinary conversation, to imply that we did not know whether A really preferred X to Y. But in ordinary conversation we would be supposing, at least, that A was the kind of entity which had a preference, or was indifferent, about the matter. The formulation, however, exactly understood, does not imply that A is such an entity. A could, after all, be one of Friedman’s leaves.

33But even if we leave open the question what kind of entity A is, the reference to preference in the above formulation is not empty. A firm link between preference and behaviour is implied by the formulation, namely this: if A prefers X to Y, then A will choose X over Y.

34Notice the conditional form of this statement. In conditional statements, which have the form “if P then Q,” nothing is said about whether P or Q are true. Only if we know, in addition, that (say) P is true, can we deduce that Q is true. Formally, it is correct logic to move from “if P then Q” and “P” to “Q.” But we cannot rationally move from “if P then Q” and “Q” to “P.” It is a deductively invalid step. So, in the present example, we may know that, if A prefers X to Y (being the kind of entity that can have preferences), then he will choose X over Y. Thus “A behaved as if he preferred X to Y” means that A chose X instead of Y. But it does not follow from the fact that A chose X instead of Y that A preferred X to Y. Nor does it follow that A is the kind of entity which can have preferences between options.

35So it can be true even of an inanimate entity which does not have preferences between options that, if it preferred X to Y, then it would choose X over Y. Logically, this has the same status as the earlier remark that this book, if it were a raven, would be black.

36Statements like these are known as counterfactual conditional statements. “Counterfactual” does not mean “false,” for they are commonly true; it means just that the first part, or “antecedent,” of the conditional is false. What makes a counterfactual conditional statement like “if this book were a raven it would be black” true is that, simply, all ravens are black. We noted in chapter 3 the distinction between the “accidental” and the “universal” senses of a generalization. We saw that the universal sense of the conditional “whenever C then E” covers possibilities as well as actualities. A counterfactual conditional is thus an example of the application of a universal conditional to a non-actual logical possibility. If we can rationally demonstrate that “all ravens are black” is true, then we can conclude that the counterfactual conditional statement “if this book were a raven then it would be black” must also be true. In spite of the problem of induction, we showed, in chapter 3, on the basis of the assumption that the world is a regular place, how we might rationally come to accept the truth of a universal generalization. It is indeed reasonable for us to hold that all ravens are black.

37A parallel argument to the above has to be produced to support the claim that, if this entity (not being one which has preferences between options) preferred X to Y, then it would choose X over Y. This counterfactual conditional has to be supported by the truth of the following generalization: whenever an entity prefers X to Y, then it will choose X over Y. This generalization could be assumed directly as a metaphysical assumption, but our conditional adoption of empiricism does not allow us to do this unless we are forced to, so we must proceed by finding an empirical foundation for it. Because of the problem of induction, we cannot observe such a regularity for every possible entity at every possible time. We have to confirm the generalization by checking it against actual situations of preference accompanied by actual situations of choice — testing it — and finding nothing to disprove it. Thus it would be disproved by just one case of preference not being followed by choice. Suppose, however, we find no such disproof in our actually experienced cases. We may then come to believe the generalization. Given this, we may say of the leaves around Friedman’s tree that, if they prefer X to Y, then they will choose X over Y. We may then plausibly hypothesize — without, that is, offending against logic or intelligibility — that the leaves behave as if they prefer X to Y.

38It should be noted that a counterfactual conditional cannot enter into the testing process of a universal generalization. Thus, here, we cannot test “whenever an entity prefers X to Y, then it will choose X over Y” by looking at what Friedman’s leaves do. Testing a universal generalization, which has the form “whenever C then E,” requires us to look for cases of the form “C and not E.” But the claim about the leaves being a counterfactual claim amounts to saying that the “C” in the “whenever C then E” is not true of them. It is thus impossible to find the “C” in the “C and not E,” either, and therefore such a counterfactual case cannot possibly provide a test of the generalization. We must necessarily refer to real cases for testing, actual cases, cases which provide independent characterization of preference, and independent characterization of choice. Then we can examine how regularly they are associated. Leaves cannot provide this test, but people can.

39In fact, only people, as entities, do provide us with evidence bearing on the question whether it is true that “whenever an entity prefers X to Y, it will choose X over Y.” It follows that this last generalization cannot apply counterfactually to people, even if it may do so to leaves. The generalization cannot apply counterfactually to every entity, or it is empirically empty, and would be empirically meaningless even as a metaphysical assumption. Some entities must provide a test, or at least a meaningful reference, and persons are the only candidates for the position.

40When we say of a person, rather than of leaves, that “A behaved as if he preferred X to Y,” then we are assuming that A is the kind of entity which has preferences between options. Assumption (1) then turns out to be stronger than a mere hypothesis which people behave “as if” they followed. When we are dealing with an entity (a person) which is characterized as having preferences between options, then, given assumption (1), one and only one of the following is true of it: A prefers X to Y; A prefers Y to X; A is indifferent between X and Y. This exhausts the logical possibilities of preference. We have seen that we cannot derive, from the fact that A chooses X over Y, that A prefers X to Y, because of the logical nature of a conditional statement. But given assumption (1), we can deduce that it is not the case that A preferred Y to X, for if he did he would have chosen Y over X. We can conclude, from A choosing X over Y, therefore, that either A prefers X to Y, or A is indifferent between X and Y. The rational economic man hypothesis is no longer to be understood as a hypothesis that people behaved “as if” they followed; rather, it is a hypothesis about the preferences people actually have. The “as if” has disappeared. It is still a hypothesis, of course. There is sufficient “ifness” in that fact. There is no need to double the degree of abstraction.

  • 10 The rational economic man theorist may not be committed to interpreting morality solely in terms o (...)

41It follows, on the rational economic man hypothesis, that a person never does what he prefers not to do. There is therefore a sense in which preference is demonstrated in behaviour. We may often come to know what a person prefers, in a given situation, simply by observing the choices he actually makes. Thus if a person says he prefers running to cycling, but whenever he is given the choice he opts for cycling, then we will insist that he really prefers cycling to running after all. Nevertheless, it remains the case that, in common-sense understanding or ways of speaking, we sometimes allow that a person may do what he prefers not to do. Thus he may do his duty instead. It may seem as if this fact falsifies the hypothesis that a person never does what he prefers not to do, but it is not quite so simple. The thesis is, rather, that people attempt to satisfy to the maximum their desires as such, whatever those desires or preferences are. A person who acts morally may be taken to demonstrate a preference for moral against immoral action: he prefers morality to immorality, and acts so as to maximize his return, measured in the satisfaction of that preference.10

42Note, however, that acting morally does not prove a preference for morality; as shown earlier, “A behaved as if he preferred X to Y” means that A chose X instead of Y, but it does not follow that A preferred X to Y. He might be indifferent between them, or the rational economic man assumption may be false. The “demonstration” of preference in behaviour is a mere common assumption, which the economist calls “revealed preference.” For real-life economic problems, revealed preference often provides the only evidence for what actual preference is. It is evidence for preference, however, not proof. To hold revealed preference to prove preference would be to hold it to define preference, and this would make the rational economic man assumptions collapse into an empty definition of words.

43A person may well prefer to do his duty, but equally well he may prefer not to (but do it anyway), and the hypothesis that a person never does what he prefers not to do could be held falsified by many cases. Many more cases plausibly falsify the hypothesis of rational economic man, where preferences are taken to be for the acquisition of commodity bundles, and no economic theorist will think that the complexities of real life can be fully represented by this ideal model. But it does not matter, for the rational economic man hypothesis is a hypothesis, and it is the task of the theorist to discover just when this hypothesis is applicable. In a situation which approximates that of a free market in commodities, it works. In a monastery, it doesn’t. It is open to scientific test, as outlined in chapter 3, just how far the hypothesis is applicable between these two extremes.

44If we suppose that the slaveholders in the Southern states were indeed, as Fogel and Engerman suggest, acting “profitably” in an efficient system, then we are to understand them as attempting to maximize the satisfaction of their preferences, in accordance with the assumptions we have specified and discussed. The question asked earlier in this chapter: why did so many act for perceived profit and not morality? may then be dissolved, for we have now seen the sense in which “morality” has been assimilated into “profit,” as being just one preference among others. Nevertheless, there remains a clear sense in which this move simply forces the question one step back, only to reappear: why did so many people prefer material improvement in their circumstances, rather than preferring not to enslave other human beings? But this question is not a difficult one for rational economic man theory.

45Against what might be expected on the “traditional” view, slaveholders displayed an unindividualistic similarity, with characteristics distinctive not of all humankind but distinctive just of a particular time and place and social system. On the economic approach now presented, the preference to engage in and continue with slavery was a net preference: the outcome of a calculation which took into consideration a range of understandable desires, including moral considerations and those for material well-being and improvement, together with estimates and expectations of associated costs. What distinguishes us from the slaveholders, even the evil ones, is not that we have noble desires which they didn’t, but that they had costs which we don’t. Our net preferences differ from theirs, because our own cost-benefit calculations differ from theirs, and these calculations differ not in terms of our having different desires, but in terms of our having different costs. We don’t have to enslave people to achieve the normal desires for material improvement which we share with the slaveholders. It was thus the social system that was distinctive and not the characteristics of the people involved.

46The problem we found with the “traditional” approach was that it seemed necessary to explain why people had the desires they did, and yet impossible to provide the explanation. The economic approach now outlined, while consistent with the supposition that individuals be understood solely in empathetic terms, shows that we can properly leave desires unexplained, for these no longer matter when it is a difference of cost, and not a difference of desire, which carries the weight of the explanation. There may be variations in preferences between different people at different times, but it need not be the case that it is differences of desire which provide the explanatory foundation. Similar desires with varying costs will explain equally well. The economic approach does not preclude explanation in terms of distinctive desires, however, if the historical evidence indicates that these were the cause of some state of affairs. To this extent the economic approach can duplicate the empathetic approach. It can also improve on it. On the economic approach we do not have to explain why people had the desires they did, and the problem can be dissolved, as we have seen. It may be thought that we nevertheless have to take desires as given, whether we adopt the earlier empathetic approach or the economic version of it, but rational economic man theory, while it can assume desires to be given, can also explain why we have the desires that we do.

  • 11 Stigler and Becker, p. 76. Economists have investigated “human capital” at length in many publicat (...)

47The economic theorists George J. Stigler and Gary S. Becker11 have shown that we may hold that desires or “tastes neither change capriciously nor differ importantly between people.” They argue that examples of what purport to be simple changes of desire can be understood as cases of stable desire. One such example was quoted earlier from Alfred Marshall, that “the more good music a man hears, the stronger is his taste for it likely to become.” This, when discussing the diminishing marginal utility assumption (5), was interpreted as a simple change of taste not covered by the assumptions specifying the rational economic man hypothesis. Such a change in taste is supposedly not explainable, but Stigler and Becker suggest an economic mode of explanation for it.

  • 12 Ibid., pp. 78, 79.
  • 13 Ibid., pp. 79, 80.

48One’s desire for music depends upon the time allocated to it, and also on the training and other characteristics which are conducive to the appreciation of music. This training and so forth Stigler and Becker call “music capital.” An increase in music capital increases the productivity of any time spent listening to music; “the marginal utility of time allocated to music is increased by an increase in the stock of music capital.”12 The appreciation of music increases with exposure to music because the marginal utility of the time spent on music rises with exposure, even though there is no change in desires. “The effect of exposure on the accumulation of music capital might well depend on the level of education and other human capital.... This would explain why educated persons consume more ‘good’ music (i.e., music that educated people like!) than other persons do.”13 Differences between people here, to put it simply, depend, not on different tastes, but on the “human capital” involved — training, knowledge, and so forth — resulting in different costs. It is the cost of appreciating “good” music which makes some choose it and others not, a cost which is lower and thus more easily met by those with educated “human capital.” It is the cost of freedom for others that makes a man or keeps a man a slaveholder, not a desire for slavery. Costs are a function of capital, and it is not just the absence of machinery or the quality of land or the distance of markets that are relevant here, but human capital, which may be represented in tradition or education. The cost of new information permitting a change in what is believed from the traditional may be prohibitive. The cost of moving from a false belief to a true one may be prohibitive. It is painful to change a tradition or a habit. We may have exactly the same desires as the slaveholders (and the slaves), but different costs. It may be that what counts as a desire for the purpose of one cost-benefit calculation is itself the net preference outcome of a prior cost-benefit calculation, and therefore explainable in terms of that calculation. Empathetic understanding perhaps works because, at bottom, we have exactly the same desires as others.

49With the help of the rational economic man assumptions we find we have been able to provide further analytical detail for Dray’s notion of a “calculation.” We can empathize with another person, for we can work through a calculation which displays the net preference and consequent choice as appropriate given that person’s desires and beliefs as to his circumstances. These desires and beliefs characterize the calculation as being a cost-benefit analysis. In these terms we can make sense of the question “was slavery profitable?,” and we have thus outlined a version of the economic approach which is consistent with the traditional approach to history. In chapter 6 we shall continue with a different problem for the consistency of the two approaches.


1 Other entities may also be admitted to reality, of course. God, for example.

2 The relationship between experience and background assumptions even in “direct experience” is nevertheless problematic, and will be discussed in chapter 7.

3 Douglass, Narrative, p. 87.

4 There is a tradition, deriving from the work of Wittgenstein, that philosophical problems only arise through a misunderstanding of the use of words, and are dissolved by a proper attention to their correct use. We eschew this position here: fundamental philosophical problems do not arise through such inattention, and the dissolution which we speak of here is not sought through linguistic analysis but through attention to presupposed beliefs.

5 Milton Friedman, “The Methodology of Positive Economics,” originally in his Essays in Positive Economics (University of Chicago Press, 1953). Quotations here are taken from the sections appearing in Frank Hahn and Martin Hollis eds., Philosophy and Economic Theory (Oxford University Press, 1979).

6 Friedman, pp. 30, 32. For a detailed discussion of Friedman’s views, and an excellent introduction to the philosophy of economic theory, see M. Blaug, The Methodology of Economics, Cambridge University Press, 1980.

7 There are a number of ways of expressing the assumptions of rational economic man theory. The present version is an expansion of that appearing in Daniel M. Hausman, “Are General Equilibrium Theories Explanatory?,” in J. C. Pitt ed. Philosophy in Economics (D. Reidel, 1981), p. 18.

8 C. Dyke, Philosophy of Economics, Prentice-Hall, 1981, pp. 35, 51.

9 Referred to in George J. Stigler and Gary S. Becker, “De gustibus non est disputandum,” The American Economic Review, March, 1977, p. 78.

10 The rational economic man theorist may not be committed to interpreting morality solely in terms of preference, although this is usual; and see M. Smith, “The Humean Theory of Motivation,” Mind 96, 1987, pp. 36–61, and D. Lewis, “Desire As Belief,” Mind 97, 1988, pp. 323-332. Still, while we are assumed to be desiring beings who can have preferences, we are also assumed to have beliefs as to what our circumstances are, and it is possible in principle to regard morality as a “factual” matter. Such a view is presented in chapter 7. For a discussion of the general relationships between ethics and economics, see A. Sen, On Ethics and Economics, Oxford: Basil Blackwell, 1987. There is a useful bibliography.

11 Stigler and Becker, p. 76. Economists have investigated “human capital” at length in many publications in their discipline’s learned journals.

12 Ibid., pp. 78, 79.

13 Ibid., pp. 79, 80.

© Les Presses de l’Université d’Ottawa | University of Ottawa Press, 1992

Conditions d’utilisation :

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.


Volume papier
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search