Is Bad Conduct Always Wrong?
The Ethics of Environmental Effects
I.
In order to do what I hope to do here, I must begin by trying to elicit and clarify what I believe are some widely held intuitions about a wide class of practical situations. I will do this by describing and commenting on some examples and, in order to avoid contaminating the reader's response to them, I will not state, until after some delay, what issue I intend to discuss and what position I will take in regard to it.
The first case should sound familiar to people who work in universities and other bureaucracies. A group of people meet regularly as a committee, to make decisions regarding the place where they work. Because the issues they discuss are important to each of them, each member typically speaks at every meeting in order to place before the group information and arguments which he or she believes is relevant to those issues. Each of them has additional incentives to speak up as well, for speaking affords each with the pleasure of self-expression and the rare joy of hearing a speech which consists entirely of the truth. Because of these additional incentives, each speaks a bit longer than is really necessary to communicate information and arguments effectively. As a result, although no one person speaks for a very long time, the committee's meetings tend to last at least fifteen minutes longer than the two hours officially allowed for each session. All the members find this quite annoying. It happens every time they meet.
What judgments about the behavior of these people do we find it natural to make? One is certainly inclined to say that it would be better if they did not behave this way. It sounds, from what I have said, as if the annoyance each feels at the length of the meetings is greater than the pleasure each feels at speaking too long. Suppose that this is true. What does it signify? It seems to mean that they shouldn't behave as they do, that they ought to do something else. But here some care is necessary to avoid exaggerating the force which this "shouldn't" and this "ought" have. Contrast this judgement with the one we would make if a single individual, one great bore, were to produce the same effect which in the present case the members bring about collectively -- talking at least fifteen minutes longer than is necessary. Such behavior, we would think, is wrong. In the present case, on the other hand, the notion that they "shouldn't" act this way means that it would be better if they didn't, not that what they do is wrong. In this sense, we recognize no strict moral requirement that they do otherwise. This point becomes perhaps more obvious if we contrast the behavior of the members with that of a single member who makes a point by gratuitously insulting other members of the committee.
In the second case, a group of traders are speculating on the price of tulip bulbs. Though they are mostly unknown to one another, they are participants in a web of behavior in which each responds to signals from the others which are received in the form of changes in the price of the good they are trading. Their speculation drives the price of the bulbs higher and higher, until most stand to make a significant amount of money from their enterprise. One day, the price of the bulbs dips and continues to fall for a short while. Each participant has the same notion of what this may well mean. There is some likelihood that the price will drop enough to cost one money in the event that one does not sell out soon. On the other hand, each knows that if one does sell out soon, one might be contributing in some measure to a further drop in the price and consequently to the detriment of those who do not sell. Most of them sell out, the price plunges, and the few who are tardy in selling are ruined.
Once again, one would say that things would plainly have been better if these people had done something other than what they did, and it is possible to think (perhaps by straining a bit) that they in fact should have done otherwise. Further, it does not seem possible to think that they have done wrong. This fact can be brought into sharper relief by contrasting their behavior with that of a speculator who knowingly causes a panic in the tulip bulb market by making a suitably large sale.
For the third case, suppose there is a town in which, for some reason, only two types of automobile are available. One type, the Whippet, is small and, because it uses relatively little fuel, is relatively non-polluting. The other type, the Leviathan, is safer for driver and passengers in the event of a collision because it is larger than a Whippet but, for the same reason, it uses more fuel and is more polluting. Fuel is very cheap and plentiful and no one is very concerned with conserving it. Anyone contemplating buying a Leviathan knows that they will have to breathe a portion of the extra pollution they produce but, even though the town is situated in a valley which is often covered by a temperature inversion layer which tends to trap air pollution inside it, they know that the amount of pollution they impose on any one person is rather small. Most of the townspeople, having taken all these factors into consideration, own Leviathans. The town is often covered with an annoying and unhealthy layer of smog. For each Leviathan driver, in fact, the effects of the extra smog produced by all the Leviathans are more than enough to cancel out the benefits of driving one.
As before, one wants to say that it would be better if these people were to behave differently by, in this case, driving Whippets instead of Leviathans. It is at least plausible to say that this is what they should do. But, once again, this does not seem to mean that what they are doing is morally wrong -- as it would be if one individual (an industrialist, perhaps) were to produce the entire increment of smog single-handedly for reasons similar to those which move the townspeople to produce it collectively.
In making moral judgments about unwelcome social effects we seem to observe an important distinction between certain cases in which a single individual is responsible for the entire effect and certain others in which responsibility is dispersed over a number of individuals. In some of these cases we also observed another distinction. The case of the loquacious committee and that of the polluting drivers are ones in which the individual is responsible for an effect which is repeated or continuous over long periods of time. Here we can be seen to discriminate on the basis of how the effect is dispersed across time. For instance, if the committee in the first example has seven members and each meeting lasts fifteen minutes too long, then each member speaks an average of fifteen minutes too long every seven meetings. If a member were to speak fifteen minutes too long at one meeting out of seven we would be strongly inclined to think that they were violating some requirement which applies to members of committees, even if they did not speak too long at the other six meetings in seven. Yet they would not be wasting more time than members we judge in a categorically more lenient manner. We are prepared to discriminate in the same sort of way in the case of the polluting drivers. If there are sixty Leviathan drivers in town, each produces on the average enough pollution every two months to create, under the right conditions, one smoggy day if it were released in one huge gust. Imagine there is an individual who does not drive a Leviathan but does precisely that, for reasons which are no worse than those which move people to drive Leviathans. Perhaps the individual is an industrialist who is saving some money -- not very much -- on production methods, or perhaps he or she is conducting a sophisticated experiment concerning the effects of eye and sinus irritation on human behavior. This, I am sure we would want to say, is wrong.
Moral philosophers have discussed the possibility that there are actions which are morally meritorious but which nonetheless are not morally required of us. The consensus among most of those who have written on the subject seems to be that such actions do exist.(1) We have just encountered an idea which in an obvious way is analogous to this one but which is apt to be more controversial among ethical theorists, the idea that there are actions which are bad -- bad in terms of consequences which do seem to be morally significant -- but which are not wrong. Whether we want to incorporate this idea into our moral theories or not, it does seem to me a part of our positive morality, the moral principles which determine our actual, pre-theoretical judgments. My examples indicate that this idea, as it is embedded in ordinary moral judgments, covers a wide range of actual situations. In some, the relevant consequences are physical effects such as the transmission of particles and gasses from one place to another, in others they lie merely in the way in which the affected parties interpret symbolic or communicative behavior. In some the group of interacting individuals is large and in others it is small, in some they know each other and in others they do not.
These situations do seem to resemble one another, however, in the nature of the incentives and costs confronting the agent who produces the relevant consequences. Later on, I will clarify the idea we have encountered here by explaining the nature of this resemblance. Eventually, I will try to explain why we have the moral intuitions which we do seem to have about situations of this kind. Finally, I will briefly suggest reasons for suspecting that positive morality is correct in the position it takes concerning such situations, that perhaps these intuitions are as they should be. First, though, I will pause to examine some obstacles that stand in the way of this last suggestion. As we shall see, some extremely respectable theories can be used to argue that positive morality, in these cases, is simply wrong.
II.
Whatever else one might think of them, the three examples I have described represent social problems, since they are clearly instances of people who, whether they are acting wrongly or not, are certainly doing something which is harmful. The idea which I have identified as part of positive morality is clearly compatible with the notion that, when it is possible, something should be done about these problems. It is compatible, for instance, with the notion that institutions should be designed, if possible, which alter the relevant incentives and costs in such a way as to eliminate the unwelcome effects that they elicit. There is an extensive literature, written mainly by economists, in which institutional changes of this sort are defended as solutions to the problem of pollution. These solutions are subject to some of the same objections as the thesis I mean to defend and it should shed some light on the problems I am treating here if we examine them briefly. From the point of view of many economists, the root of the problem of pollution lies in the fact that the normal arrangement of property rights -- in which the atmosphere and large bodies of water are in effect treated as a "commons" -- permit people to use the environment as a free waste-sink, imposing much of the cost of waste disposal on other people. This means, among other things, that some of the total cost of producing an economic good can fail to fall within the scope of the producer's accounting system, and whenever this happens, pollution will be overproduced. There is a level of pollution which would be worth the cost, but human action will fail to find it as long as people need not take account of the entire cost of what they do. One policy that has often proposed by economists as a remedy for pollution resulting from manufacturing is the use of a tax which is variously called a "residuals charge" or an "effluent fee" This policy would impose a per-unit tax on objectionable emissions which would be equivalent to their unaccounted-for, external cost. Other taxes might then be reduced by an amount equal to the pollution taxes which are paid. The market would automatically find the level of pollution which is worth the cost, partly by reducing the consumption of goods that can only be produced by methods that cause pollution and partly by shifting production to less polluting fields. Despite the widespread support among economists for policies like this one, the pollution-control measures actually adopted in this country and elsewhere have almost always represented an entirely different approach. The usual policy is one of "direct regulation," in which pollution, or at least certain forms of it, are declared to be legal offenses and subjected to punitive measures such as fines. Economists have usually explained this striking divergence between expert opinion and actual practice by appealing to the self-interest of the agents who influence actual practice. Some have pointed out that the system actually in operation enables politicians to shift responsibility for controversial decisions to unelected bureaucrats who have no need to garner votes, and who are able to shift the cost of pollution abatement to agents who have little influence over the political system. Others have claimed that, under certain circumstances, the limits which direct regulation places on production actually have beneficial pecuniary effects on manufacturers, the same sorts of effects that cartelization has.(2)
As some of the same authors have pointed out, however, special interests are probably not the only factor that obstructs implementation of the pollution tax policy; there are "widely accepted ethical norms"(3) which stand in its way as well. For our purposes, the most important moral objection to this policy is the one which states that it merely "sells the right to pollute."(4)
As a mere description of the facts, this claim is correct and penetrating. The pollution tax is not a punishment, like a fine, and accordingly does not convey, as part of its expressive function, the conviction that pollution as such is wrong. It is simply a price which is meant to express the value of the environment as a waste sink. To constitute an objection, this claim must assume that pollution is wrong. What reason is there to believe this assumption? Unfortunately, it seems to follow quite naturally from any one of three different widely accepted ethical theories. This means that any of these theories can also be used against the thesis I wish to defend in this paper, at least as it applies to pollution and to any social problem that relevantly resembles it.
First, at least one sort of hypothetical contract theory of morals could easily be used to argue that the polluting drivers in my example are doing something morally wrong in driving Leviathans. Such a theory would state something to the effect that people are morally obligated to follow whatever rules it would be in their interest to agree on if agreements, supposing that were possible and that everyone's compliance with the rules were somehow guaranteed. Whippet drivers and non-drivers would clearly have an interest in agreeing to a rule prohibiting the use of Leviathans, and Leviathan-drivers would as well, since the aggregate undesirable effects of all the Leviathan drivers are greater for them than the differential benefits of driving one of the more-polluting cars. Thus, everyone has good enough reason to agree to a rule prohibiting the use of Leviathans, even if there is no veil of ignorance preventing them from knowing to which section of the population they belong. This would mean that this sort of hypothetical contract theory would imply that driving the more-polluting type of automobile is morally wrong.
The same result would follow if one begins with a utilitarian account of what is morally required. If we make the necessary assumption that the utilities of different individuals can be compared, it is impossible to avoid thinking that, even though the negative effect a Leviathan driver (as compared to a Whippet drive and a non-driver) has on any one individual is small, the sum of these negative effects taken together is greater than the benefits which the driver could only get from driving a Leviathan. In the choice between driving a Leviathan and not driving one, then, the utility of everyone affected is maximized by not driving one. An act-utilitarian moral theory would therefore require one to not drive a Leviathan. The same result would follow from a rule-utilitarian theory.
Utilitarianism and hypothetical contract theories could also be used to show that the members of the loquacious committee are acting wrongly. They could also, with some plausibility, do the same thing for the individuals who sold out and precipitated the tulip bulb panic. A third moral theory, the neo-Lockean natural rights view, is probably more limited in its application, but it seems to have even stronger implications for cases of pollution.(5)
In this view, each individual enjoys a morally inviolable domain consisting of the individual's person and property. Anyone who invades one's domain without one's consent, at least if it is an invasion which one would prevent if one could, is violating one's rights and thus is doing something which is morally wrong. One can commit such a wrongful invasion by bringing one's person into the domain of another, as in cases of trespass. One can do so also by initiating a process in which some physical entity enters the domain of another. This would include launching a bullet, hurling a burning torch, and carelessly parking a truck on a hill with its brake off. It would also include doing things which send particles, fluids, gasses, or vibrations into the domain of another individual -- at least if the individual would have prevented their entry if he or she could. This means that it would include all activities which we would normally regard as polluting. Thus, although a neo-Lockean natural rights view may not regard the behavior of the people in my first two cases as morally wrong -- because they might not involve invasions of this sort -- it would certainly seem to imply that the behavior of the Leviathan drivers, along with all polluting behavior, is morally wrong.
III.
Despite the implications of these influential ethical theories, positive morality contains no rules that imply that the bad conduct in the cases I have described is wrong. I will now try to explain why this is so.
Consider the nature of the situations which such rules would govern. In each situation, there are no psychologically salient rights which the bad conduct would violate, other than the rights which are specified by the rule itself, if it exists. Whether such a rule exists or not, however, the individuals involved have other reasons for acting. They know that they are in a situation in which (1) all have a choice to make and (2) definite results, which in one degree or another are desirable or undesirable, will follow for each of them from the choices that all of them make. Situations which have these two characteristics are what game theorists call "games." In each of my cases, the "players" have two options and the results of the game are related to their options in definite ways. In each case, there is one option which is such that (a) if all or most of them do not take it, the results for all are undesirable, (b) if all or most of them do take it, the results for all are desirable, while © for any set of choices on the part of the others, each achieves a better result by not taking it than by taking it, and (d) those who do not take it diminish to some extent the desirability of the results obtained by those who do. Rather obviously, this description also fits the familiar Prisoners' Dilemma, and games which have characteristics (a) through (d) have been called "generalized Prisoner's Dilemma structured situations."(6)
A rule strictly prohibiting the bad conduct in my examples would bring everyone involved into the desirable state of affairs described in (b). This fact, however, certainly does not distinguish between rules which don't and shouldn't exist and ones which undoubtedly do and should. It has been suggested -- very plausibly, I think -- that all or most of the basic duties of morality resolve situations which have the same four characteristics by bringing everyone involved into a state of affairs which fits the description in (b).(7)
Nonetheless, such duties do differ with regard to the motives one has for following them from the class of rules I am discussing here and the difference is closely relevant to the point I wish to make. One can see the difference, in fact, precisely by considering some implications of the fact that behavior which conforms to basic moral duties can represent a benign solution to a Prisoner's Dilemma situation. As an example of a basic moral duty I will take the familiar idea that one owes aid to those in need. As would be the case with the class of rules I have been discussing here, the behavior which the rule requires is something to which the beneficiary has no salient rights other than those which are specified by the rule itself. Nonetheless, there are motives which are external to the rule which tend to support behavior which follows it. We can see that this is so if we first note how situations in which people are in need of help can present a Prisoner's Dilemma structure.(8)
Suppose that two men, R and C, regularly have lunch at the same cafeteria. From time to time, one of them finds that he hasn't enough money on him to buy lunch when the other is able to help him out by giving or lending him the needed money. Rendering assistance would cost either of them a definite amount of well-being: two units. But receiving the same assistance when one is in need is worth considerably more than what one loses by helping, and I will suppose this benefit amounts to six units. The situation becomes a recurring Prisoner's Dilemma if we suppose that R and C alternate in coming up short, with first one and then the other needing help. The first two emergencies and all succeeding pairs are described by the following matrix:
C1: C2:
R1: 4,4 -2,6
R2: 6,-2 0,0
The strategy marked 1 is giving assistance and the one marked 2 is not giving assistance. Each occurrence of this Prisoners' Dilemma game places both R and C in one of the cells of the matrix. If both have given (and received) help at the end of one play, both get a result of 4 (= 6 - 2). If neither of them incur the costs of helping, neither of them will receive the benefits of being helped and both will get 0.
As the off-diagonal cells indicate, this situation makes it possible for one person to gain at another's expense: it is possible to accept help without returning it. But a comparison between the diagonal cells also indicates that it presents the possibility of gains from cooperation: mutual help is at all events better for both parties than mutual indifference. Moreover, it has a characteristic which can promote cooperation, in that it is a situation of mutual power, since the results each party gets depends on what both of them do.(9) Each player, to promote his own self-interest must care about what the other does, and since this is an iterated game in which each party knows what the other has done, each must care about the other's response to his own behavior. To some extent, each has an interest in accepting help without giving it, since this would guarantee oneself a result of 6 in one occurrence of the game; but this would tend to work against one's interests in future plays. This is because it would provide the other player with evidence that such unhelpful behavior will continue -- a prospect which would mean that it is in his self-interest to "punish" one's unhelpfulness by being unhelpful himself, since he would thereby expect a result of 0 instead of -2. In the long run, each has a motive to avoid discouraging the other from being helpful, and one can only do this by being helpful oneself. If this motivational feature of the situation is known to both players and known to be sufficiently strong, mutual help will follow.
Thus there is at least one motivational feature of situations in which people are in need that exists independently of a duty to render aid and supports behavior which conforms to it. Viewed in the light in which I have shown it so far, helping behavior presents the appearance of a mere insurance-scheme. Of course, this is not the appearance it presents in the world around us, but this should not be surprising. At least sometimes, people in the world around us help each other in part from a sense of duty, and this helps to differentiate such behavior from an insurance scheme. But rules which specify duties are not the only source of actual helping behavior which distinguishes it from a sort of insurance. Human beings have a certain tendency, which is sometimes very strong, to take an interest in the interests of others. It includes trying on occasion to alleviate the perceived sufferings of others and appears to exist to some extent independently the requirements of positive morality. Without implying any psychological speculations about the source of this tendency, we can follow Hume and Adam Smith in calling the tendency "sympathy."(10) When a player's turn comes up in the game I have described, he knows that the other player is in need, that he is in a position to relieve it, and that if he does not do so the need will persist until it is relieved in some other way. This is a motive for taking option 1 which does not appear in the payoff structure, and it obviously can be a powerful motive.
To the extent that the principle that one must render aid to those in need governs conduct in Prisoners' Dilemma situations, the nature of these situations appeals to both self-interest and sympathy in ways which support behavior which conforms to the principle. It is obvious that both factors have an important influence on actual conduct in which people follow this principle and that both of them help to explain why such behavior exists. The same things can be said, with approximately the same force, for most of the other principles which specify the basic duties of positive morality, including the rules which require truth-telling and promise-keeping, and the ones which forbid theft and aggressive violence.
The same things, however, could not be said of the rules which would forbid the bad conduct which occurred in the three situations I cited at the outset. These situations did not appeal either to self-interest or to sympathy in ways that would support such rules. This is because, unlike the game of helping, these situations are not ones of mutual power, but of mutual powerlessness. This in turn is in large part due to the fact that while the helping game is a two-person game, these situations affect the individuals involved as N-person games. A player is someone who both receives payoffs and determines the payoffs of any one occurrence of the same game jointly with the other players, so that the results of an occurrence of a two-person game are determined by two agents and those of an N-person game by more than two.
It is widely recognized that the difference between these two types of game can have very important consequences.(11) In my three troublesome cases, the power of each player is such that the effect that any one player has on any other is not significant, in that neither the agent nor the patient clearly perceives it as affecting the patient's well being: the effects produced by any one player are diffused among all the other players. But the aggregate effect produced by all the players for any one player is significant. This further limits each player's exercise of power in several important ways. In the helping game, each player has two possible moves, and it was crucial to the way in which self-interest led to mutual aid that: 1) each player had a self-interested motive for making the move that is less beneficial to the other player if that is what the other does, and 2) doing so is perceptibly damaging to the other player. In the three N-person situations both these conditions are entirely absent. None can make a move which is perceptibly damaging to any one player because none has any significant effects on any one player at all. And the (insignificant) effects one does have fall on all the others regardless of how they play. Further, no one has a motive to invent some way to damage those who make the less beneficial choice because no one feels that any one person is harming or failing to benefit them in a significant way. In addition -- again, because no one has significant effects on any one person -- the situation does not appeal to sympathy as a motive for mutually beneficial behavior. There are no salient damaging effects with which to sympathize. In general, the promptings of self-interest are entirely on the side of making the less socially beneficial move oneself, and do nothing to motivate one to impel others to do anything else, while sympathy is unable to compensate for this insufficiency of self-interested considerations.
There is an easy way to describe this sort of N-person situation more generally and more clearly. The individual who makes the more beneficial choice is providing a good which is scarce (in the sense that everyone would prefer to have more of it at low enough cost) and costs the individual something to produce. The fact that this good is "diffused" across the other players of the game means that, in addition to its scarcity and costliness, it has the other characteristics of what economists call a "public good": 1) once it is provided for someone, the additional cost of providing it for someone else is zero, and 2) there is no way to exclude anyone from receiving it. In each of the three cases I discussed at the outset, the benefit produced by the more beneficial action is a public good. Other examples of public goods production are refraining from dropping litter in a public place, refraining from contributing to traffic congestion by not driving during rush hour, and refraining from running toward the exit when someone yells "Fire!" in a crowded theater. If there is an amount of some public good which is such that, if it were produced by enough people, the benefits would exceed the costs for everyone concerned although one's own share of the good that one produces is less than the cost one incurs in producing it, the resulting situation will have characteristics (a) through (d) of the generalized Prisoner's Dilemma structured situation.(12) It will also be a situation in which one cannot encourage production of the beneficial results involved by denying them to those who failed to produce -- as was done in the helping game -- because there is no way to exclude only non-producers. If, in addition, the share of the good which any person would supply to any one person is not significant, then no one has a motive to punish non-producers in some other way, and for the same reason no one will be motivated to provide it by considerations of sympathy. This means that situations in which individuals have an opportunity to produce some public good, where the amount of the good that any individual can provide to any one individual is insignificant, differs in an obviously crucial way from a generalized Prisoners' Dilemma situation of the sort modeled in the helping game. In the production of public goods which are diffused in this way, a rule requiring people to do the benign actions described in (b) (in these cases, producing the good) is left unsupported by the incentives which in the other cases are provided by the situation itself. Behavior which produces such diffused public goods is not the sort of behavior we have learned to expect from others: it is, at any rate, a far less reliable feature of our social world than behavior which conforms to the basic duties of positive morality. I submit that what we have seen here must be part of the reason why this is so. Compliance with the basic duties of positive morality is relatively reliable and predictable because of the incentives which are built into the situations which the rules govern: Since compliance is clearly beneficial to some individual or other, it is supported by the promptings of sympathy. In addition, the individuals who could thus be benefitted have a motive to punish non-compliance, and they can do so by means of the easy and natural method of denying to those who fail to comply the benefits of their own compliance.
These explanations of some familiar features of human behavior also suggest explanations of certain facts about human beliefs: namely, the intuitions I have been assuming we have concerning the three examples with which I began this essay. In each case we fail to judge that people are acting wrongly when they refrain from producing public goods of the diffused type I have just been describing, and we do so in spite of the fact that everyone who is able to produce the good would be better off if they all did so. The most obviously and directly relevant explanatory factor in this context is the fact that each person's failure to produce the relevant good does no salient damage to anyone, so that the promptings of sympathy provide no motive for producing it. Our ability to sympathize with the harm our actions cause for others provides us with a reason to believe that such actions are wrong, as well as a reason for not doing them. These reasons for thinking an action is wrong are entirely lacking when we refrain from producing public goods like the ones in my three examples, and this lack seems to be a significant one. It seems to be very difficult for us to see an action as wrong, merely on the basis of its being harmful, if it is not clearly harmful to any specific individual.
The fact that there can be relatively little punishment for failure to produce public goods of this kind also suggests some explanations for our intuitions about such failures, though their explanatory relevance is, I think, less direct than is the case with the insufficiency of sympathy. By itself, the relatively remote chances of being punished simply mean that, so to speak, the cost of cooperative behavior is relatively high: the cost of producing the good is not offset by the thought that by cooperating one has avoided a relatively likely penalty for not producing it. I do not think that this by itself can explain why we do not believe that there is some moral requirement to the effect that we must produce it. Of course, there are some people who possess and effectively use a certain all-too-human ability to make themselves believe that they are not morally required to do whatever would be expensive or inconvenient, but there is no reason to think that this cognitive skill is sufficiently widespread to explain phenomena as nearly universal as the ones we are considering here.
Nonetheless, the high cost of cooperating (if one wants to put it that way) does seem to be part of the reason why people in fact do not cooperate very consistently in contexts like the present one, and this in turn can explain why we tend not to believe that we are obligated to cooperate in such situations. For instance, if people are generally not very reliable when it comes to refraining from littering already littered streets or refraining from speaking a bit too long at meetings then, unless I am obviously different in this respect, no one will expect me to behave in these ways. This means they will not come to depend on me to do such things, and consequently will not be disappointed if I should fail to do them. Further, since these same people are unlikely to do such things very consistently themselves, they are unlikely to feel that I am taking advantage of their good behavior without contributing something to them in return. This would seem to mean that though they might believe -- as we very often do believe -- that such good behavior is just the sort of thing that we ideally ought to be doing, they will probably lack a strongly personal, emotionally charged motive to reproach me for my less than ideal behavior. They will be unlikely to feel that I have disappointed them and exploited their good nature. This means that, though my behavior is actually harmful to other people, I will not expect that whatever advice I might get from them to the effect that I should change my ways will have that peculiar biting or stinging quality which characterizes the reproach we direct at people who violate strict duties toward others.
Furthermore, if I do not observe that such ideal conduct forms anything like a consistent pattern in the world around me, then all the things I have just said of other people will also be true of me. I will not depend on them to behave that way, nor will I be disappointed if they do not, nor will I feel that they are taking unfair advantage of me. I will therefore have no motive to bitterly reproach them for their behavior. This, of course, is quite different from the way I would view the behavior of others when they violate strict duties toward other people.
These facts all indicate in various ways that a belief to the effect that we are obligated to behave ideally in these contexts cannot rest on the same sort of social and psychological support that sustains those strict duties toward others which actually are part of our positive morality. The fact that the behavior of the people around me displays a relatively consistent pattern of refraining from robbing and cheating one another, for instance, makes it quite obvious to me that I would face their angry reproaches if I were to break out of this pattern and do such things myself. This salient and enduring possibility of bitterness and blame serves as a very strong reminder to me that -- as I believe to be the case -- one really should not do such things. It makes my belief so much the more difficult to forget, ignore, or rationalize away. The same is true of the fact that I am disappointed and aggrieved when others violate this pattern of behavior. The enduring possibility of my own bitterness and blame gives to this belief a psychological presence well beyond that which is normally possessed by the more remote and refined ideals. Perhaps most importantly, the mere presence of the pattern of behavior itself in the world around me serves as a powerful reminder that this is how one should act: it is compelling evidence that others believe that this is how things should be and that they are willing to act on this belief. All these facts serve to explain why the strict duties toward others which are part of positive morality enjoy the widespread and stable acceptance as guides for conduct which in fact they do have. The social context in which we come to believe them also reinforces that belief by making them unforgettable and, psychologically, difficult to evade.
This in turn would explain why we tend not to believe that we have similar duties to produce public goods of the sort I have been discussing. Such beliefs would have to survive without the support that the social context gives to the duties which are actually in force: the explanation would be that they are apparently not able to survive without such support.
IV.
This completes my explanation of the intuitions I attempted to elicit in the reader at the outset. I have also claimed that this explanation suggests why there should be no rules which run to counter these intuitions. Obviously, no one has to believe this further claim without additional argument to back it up. If an act utilitarian were to agree to my explanation, he or she might just say that I have merely pointed out a flaw in human nature itself. I have focused on a subset of a certain type of situation -- namely ones in which there are no salient rights involved -- in which people normally do make decisions on the basis of utility. This is just what we should always do, according to the act-utilitarian. The flaw lies in the fact that, as I have indicated, we do not -- at least not always -- make decisions on the basis of utility by summing the utilities of different people. According to the act-utilitarian, if the social benefits of an action exceed their costs, that act should be done, even if no significant benefit falls on any one individual and even if the costs are focused on the one who provides the benefits. This is not a way in which it is natural for us to think. This fact is no doubt related to the fact, which I referred to early on, that in evaluating the utility effects of one individual we do not do so by summing those effects across time. We can fail to think that it is wrong of a committee member to speak slightly too long in each of seven meetings even though we would think it decidedly wrong to waste the same amount of time in one speech. Sometimes we evaluate the utility effects of actions upon persons one at a time -- one act at a time and one person at a time. The act-utilitarian would say that at least part of this is a mistake: we ought to sum the utilities of individuals and do what is most socially beneficial on balance, and such judgment ought to be part of ordinary morality. It should not matter that, if I break the relevant rule, there are not salient injuries with which to sympathize. We should accept the rule in spite of that fact, and in spite of the fact that other people typically do not seem to be willing to follow the rule, and in spite of the fact that neither they nor I will experience strong emotions if someone violates it. More generally, any of the moral theories which are widely accepted could be used to say that it is at least sometimes wrong to fail to provide the sorts of public goods I have described and, accordingly, they would also imply that facts like these should not be insuperable obstacles in the way of believing the relevant rules. If human nature stands in the way, then human nature is faulty.
Ordinarily, it is easy to say something like "well, people ought to believe this, and if they don't, they're just wrong." In the present case, though, I don't think it should come quite so naturally. After all, if we accept my explanation, the obstacles which stand in the way of accepting these rules are different in kind from those which usually prevent the masses of human beings from believing the insights of enlightened minorities. It is not that people are being led astray by incorrect ideas, or by the mindless pressure of tradition, or by mere emotional prejudice. The obstacles seem to be less meliorable than that. Because of the insufficiency of natural sympathy in these situations, people are not able to see, vividly, the point of the onerous behavior the rules enjoin. The social context in which the rules are to be learned and applied does not naturally lend its influence, as it does in other cases, to making them memorable and impressive enough to form a permanent part of the stock of ideas which are readily available for planning our everyday lives. Given this, the claim that people should believe them anyway seems a hopeless one: it could be that most people never will believe, or not consistently.
Of course, it is still possible to have hope here if one has enough confidence in the influence which ideas and theories have over human behavior. One can still think that we should exert more ideological pressure on those who do not consistently accept these rules. For many centuries human beings have looked to moral leaders and their ideas for guidance which, in fact, does sometimes win out over supposedly intransigent human nature. Isn't it possible that such methods would work in this case?
The only honest answer I can give to this question is that, although I have given reason to be pessimistic, I cannot deny that such a thing is possible. In that case, one might well ask, why not persist in resisting the intuitions with which I began this essay? If all available ethical theory agrees that at least some of these rules are right, it is surely not foolish to persist if there is a real chance that most people will be converted in their hearts to the true faith at last. My reply to this is that the explanations of human behavior that I have offered here suggest a theoretical reason why such persistence might actually be wrong.
The explanations I have given suggest that to rely on moral leadership in this way and in this context would represent a more profound departure from the way things presently are than one would think. As I have depicted it here, the part of positive morality which consists of strict duties toward others is not a collection of abstract ideas which people learn from moral leaders and then apply to their own conduct. It is a social phenomenon of a very particular sort: Insofar as it is a collection of ideas at all, these ideas are rooted in patterns of behavior in which people's interests are related in such a way that the pattern is self-maintaining. Although the rules of positive morality require people to do things which are in the interests of others, people are normally able to adhere to them without sacrifice, since their own compliance is normally in their long-run self-interest. The system of incentives which guarantee that this is so also makes these rules memorable and impressive, so that they are not experienced as abstractions but as everyday realities. We should expect these facts to greatly enhance the stability of the rules and the pattern of behavior in which they are rooted. Given that our continuing to survive together in peace depends on these rules, their stability is a virtue of great importance. Rules which are only in force because of the influence exercised by moral leaders, and because of our conscientious and deliberate efforts to apply their ideas, are radically different in kind and will inevitably have much less of this important virtue.
Perhaps the rules which are rooted solely in such influence of moral leaders do not belong in the part of positive morality which consists of strict duties toward others. Abstractions and their originators obviously have powerful virtues of their own, but they are also extremely unpredictable. What I would like to suggest -- and at this time I only know enough to offer it as a suggestion -- is that the rules for which they are the sole basis should be reserved for problems which clearly require their powerful virtues and which do not require extremely stable and predictable results.(13)
1. For comments on some contributions to this literature, see Shelly Kagan, "Does Consequentialism Demand Too Much? Recent Work on The Limits of Obligation," Philosophy and Public Affairs, September 1984, Vol. XIII No. 3, pp. 239-254.
2. J. M. Buchanan and G. Tullock, "Polluter's Profits and Political Response: Direct Controls Versus Taxes," The American Economic Review, March 1975, Vol. LXV No. 1, pp. 139-147.
3. Buchanan and Tullock, p. 143.
4. See Vincent Barry, Moral Issues in Business (Belmont: Wadsworth, 1983), p. 331.
5. The most familiar current version of this view is to be found in Robert Nozick's Anarchy, State, and Utopia (New York: Basic Books, 1974). For a sophisticated and very interesting defense of direct anti-pollution regulation which is based on neo-Lockean natural rights theory, see Richard A. Epstein, "Nuisance Law: Corrective Justice and Its Utilitarian Constraints," The Journal of Legal Studies, January 1979, Vol. VIII, pp. 49-102. Nozick favors treating pollution as a tort (Anarchy, State, and Utopia , pp. 79-81). Such a policy would not differ from direct regulation in a way that is relevant to my present purposes because it, too, involves treating pollution as a legal offense.
6. Edna Ullmann-Margalit, The Emergence of Norms (Oxford: Oxford University Press, 1977), pp. 25-27. In case the relationship to the prisoners' dilemma is not obvious, consider this typical Prisoners' Dilemma matrix:
C1: C2:
R1: 3,3 0,5
R2 5,0 1,1
Characteristic (a) correctly describes the R2-C2 outcome and (b) describes the R1-C1 outcome. Applied to this matrix, (c) states that each player gets better results from strategy 2, whatever the other one does, and (d) states that the result for a player who takes 1 when the other takes 2 are worse than would have resulted had the other taken 1 as well.
7. Ullmann-Margalit, pp. 38 fn. 5. For a classic statement, see David P. Gauthier, "Morality and Advantage," The Philosophical Review, October 1967, Vol. LXXVI No. 4, pp. 460-475.
8. The ideas in the following two paragraphs are mainly taken from Robert Axelrod, The Evolution of Cooperation (New York: Basic Books, 1984), passim, and from Robert L. Trivers, "The Evolution of Reciprocal Altruism," Quarterly Review of Biology, March 1971, Vol. XXXVI No. 1, pp. 38-39. The example I use is my own.
9. This point tends to be pushed into the background in discussions of Prisoners' Dilemmas, many of which treat it as virtually the paradigm of all social evils. See, for instance, Gordon Tullock, The Social Dilemma (Blacksburgh: University Publications, 1974), esp. Ch. I. The relationship between a capitalist and a labor union has Prisoners' Dilemma characteristics which admittedly can cause problems, but to understand such situations fully we should compare such relationships with others, such as those between masters and slaves, which lack such characteristics and for that reason contain fewer possibilities of cooperation to mutual advantage. A Prisoners' Dilemma is not a situation of exploitation with impunity.
10. By assembling a number of ingenious examples, Smith argues persuasively that sympathy is such a primitive human response that it must not depend on morality for its existence. See The Theory of Moral Sentiments (Oxford: Oxford University Press, 1979), I; Sect. 1, Ch. 1.
11. See Robyn M. Dawes, "Social Dilemma," Annual Review of Psychology, Vol. XXXI, pp. 169-193.
12. From the fact that everyone is made better off if the good is produced by a large enough number of those who can do so without their costs exceeding their benefits, characteristics (a) and (b) immediately follow. From the fact that the individual's share of his or her own product exceeds the cost of producing it, it follows that one is always better off not producing it and merely consuming the amount of the good (if any) which is produced by others, and this gives characteristic (c). This would leave those (if any) who do produce it worse off, and this is (d).
13. Part of the first draft of this paper was written during a stay at the Center for Study of Public Choice at George Mason University which was made possible by a generous grant from the Center. I owe James M. Buchanan and all the people at the Center my thanks.