Robert Frank’s high-spirited book is a tract against selfishness. But contrary to what that might suggest, it is not a moral tract urging altruism, benevolence, and brotherly love upon its fallen readers. Frank’s target is not the moral failings of the “average sensual man,” but the intellectual failings of social scientists carried away by the urge to explain all of human behavior as a matter of self-interested calculation. And his message is simple: the view that we are rational, self-interested calculators—the view that economics has popularized and the other human sciences have borrowed—is false in fact, hazardous as the theoretical basis of social science, and dangerous when it is fed to students of economics, psychology, and the other human sciences.

Frank’s robust statement of his case can hardly be improved on:

In recent years, the message from behavioral scientists has been that people are fundamentally selfish. Biologists tell us that behavior is shaped ultimately by material rewards, that the relentless pressures of natural selection will cull out any organism that forgoes opportunities for personal gain. Psychologists affirm this view, noting the pervasive role of material rewards in the learning process. Economists, for their part, point with pride to the power of self-interest to explain and predict behavior, not only in the world of commerce but in networks of personal relationships as well.

Yet, he says,

The plain fact is that many people do not fit the me-first caricature. They give anonymously to public television stations and private charities. They donate bone marrow to strangers with leukemia. They endure great trouble and expense to see justice done, even when it will not undo the original injury. At great risk to themselves, they pull people from burning buildings, and jump into icy rivers to rescue people who are about to drown. Soldiers throw their bodies atop live grenades to save their comrades. Seen through the lens of modern self-interest theory, such behavior is the human equivalent of planets traveling in square orbits.

Frank frequently pauses to list many other obviously non-selfish activities, from tipping generously in restaurants we shall never patronize again to turning out to vote in the foulest weather. Since we all know of such cases, we might wonder why people have been so quick to identify rational behavior with calculatedly selfish behavior. Frank suggests two reasons; the first is the desire to seem smart, to show that one is not naive. “Nice guys finish last” has the ring of worldly wisdom. “Nice guys quite often finish first” is less snappy and anyway sounds sentimental. The second reason is the intellectual predominance of economics among the behavioral sciences; it is this that preoccupies Frank, who himself teaches economics at Cornell. Two hundred years ago economics, thanks to the theories of Adam Smith, emerged as the first successful social science—politically influential, intellectually coherent, its practitioners armed with a conviction that they belonged to a recognizable discipline. One hundred years ago it began to appear as something like a modern science. Upon the assumption that homo economicus is a self-interested and well-informed calculator, economists built mathematical models of economic activity which seemed to endow their subject with a rigor that has so far escaped history, sociology, and political science.

Practitioners of the other social sciences have reacted predictably; many have emphasized the ways in which their own disciplines are different from the austere simplicities of economic theory; others have thought that their disciplines could progress only by adopting the premises and techniques of economics. Richard Posner, a Federal judge and influential conservative thinker, made his name as the author of The Economic Analysis of Law. The American sociologist and economist Gary Becker made his by offering to explain all social behavior as a series of bargains for profit. The urban expert, political scientist, and economist Anthony Downs became well-known for his Economic Theory of Democracy. Nor has the impact of economics been confined to the social sciences. Although Darwin’s account of evolution through the competitive struggle for survival did something to legitimate the economist’s enthusiasm for the competitive struggle of the marketplace, economists had already had a greater impact in preparing the way for Darwin. By the time The Origin of Species was published in 1859, it was an economist’s commonplace that a market economy was a struggle which allowed only the fittest firms, the ablest workers, and the shrewdest consumers to survive.

As Robert Frank says, the intellectual conspiracy whereby the behavioral sciences reinforce one another’s conviction that “rational” behavior is to be identified with the pursuit of self-interest doesn’t survive examination. But because Frank writes with great briskness and a deft polemical style, an unwary reader might think that the object of his book was to show economists the door, and tell them to take their obsession with “material benefits” with them. In fact, he has a much more complex argument to make than that. He wants to show that simple selfishness is not a predominant part of human motivation; but, as an economist, Frank wants to explain how unselfishness can have a material payoff. Simple-minded exaggerations of the economist’s assumptions about self-interest are misleading, but it is still the economist’s habits of thought and the economist’s insights that Frank everywhere relies on.


His argument is this: in many cases people do not act on the basis of a rational calculation of different possible payoffs, but on the basis of uncalculating emotions such as love, anger, hunger, resentment at injustice, moral outrage, guilt, and sympathy. Even if South African fruit is much cheaper than other fruit and even if an apparently convincing argument is made that a boycott will not help end apartheid, I may still not buy a South African melon. I simply may not want to buy products made under such unjust conditions. More interestingly, Frank argues, there is every reason to be glad that we act on the basis of such feelings. Calculating, selfish creatures of the “me-first” persuasion would chronically do much worse than people like most of us impulsive and emotional people. Passions Within Reason is an apt title; we have good reason to be passionate.

Still, this seems to lead to a paradox; we are asked to believe that people do better for themselves in the long run by being disposed to act in ways that do them no immediate good—that impulse is better than calculation. But Frank’s most striking example of our readiness to act without looking out for ourselves suggests that we may do very badly indeed if we don’t calculate consequences. For a quarter of a century, at least some New Yorkers have felt vicarious shame at the mention of Kitty Genovese, the young woman from Queens who in March 1964 was raped and murdered by Winston Moseley, within earshot of at least thirty-eight people who had heard what was happening and had not even telephoned the police. They conclude that it shows how selfish and indifferent to others we really are. But in the same city in the spring of 1982 three CBS technicians, Leo Kuranski, Robert Schulze, and Edward Benford, tried to rescue Margaret Barbera from an assailant who was brandishing a long barreled pistol in a parking lot on the west side of Manhattan. All three were shot and killed. Their behavior certainly seems to show that some New Yorkers at any rate are unselfish, indeed heroic. On the other hand, that they were killed casts doubt on the thought that unselfishness yields dividends. Their impulse to help Margaret Barbera did them no good. If Kitty Genovese’s neighbors were selfish, their selfishness did them less harm than Kuranski, Schulze, and Benford’s bravery did them.

A moment’s thought shows the danger of generalizing from one case or another. Unselfishness and impulsiveness may get people into trouble that selfishness would have kept them out of, but a society of selfish people will be much worse off than a society of impulsive altruists. Kitty Genovese’s neighbors escaped a little inconvenience; she got killed. If they had been in her place, they would have wished for different neighbors. The very existence of law and its enforcement depends on people not being selfish calculators. If every policeman, asked to prevent a crime at some cost to his own safety, calculated that it was not worth the risk, how long could any form of government survive? And how long could the policeman’s willingness to do his duty survive if the rest of us felt no impulse to assist him? Nor is it only the most approved emotions—sympathy, outrage at lawlessness, and a sense of fair play—that do us a good that calculation cannot. The impulse to revenge ourselves may get us into trouble, but it can also do good, as a little reflection on nuclear deterrence reveals.

Setting its morality to one side, nuclear deterrence in the shape of Mutually Assured Destruction works because each side possesses enough weapons to deliver a devastating retaliation against anyone who launches a first strike against it. The danger that the other side will retaliate is enough to stop either side starting trouble. But it has often been said that there is a logical flaw in deterrence:

Rationalists have argued that this form of deterrence makes no sense. Once we know we are the victim of a first strike, they reason, it is obviously too late to deter anything. At that point, our interests clearly dictate that we not retaliate, for to do so would only increase the likelihood of world destruction.

This wholly undercuts the deterrent; our opponent reasons that once he has launched his attack, our only “rational” response is to do nothing, in which case he has nothing to fear by attacking us.


As many writers have argued, the deterrent will only deter if our opponent believes there is a good chance that after being attacked we will behave irrationally, and revenge ourselves regardless of what good it does. Or, to put it in the paradoxical fashion that theorists find attractive, it is rational before the event to predispose ourselves to be irrational afterward. The case of the nuclear deterrent is only a large-scale example of a universal issue. Almost all punishment costs society more to inflict than the offense did for which it is imposed; but the threat of punishment would have no deterrent effect unless people were convinced that it would be inflicted without regard for the cost. The vengeful urges that lead us to support the authorities who inflict the punishment do us good precisely because they stop us asking what good the punishment does once the crime has been committed.

Still, as the example of Kitty Genovese suggests, her selfish neighbors did better than Margaret Barbera’s unfortunate rescuers. The theory of evolution suggests that people do what pays, i.e. what helps them survive better; does this not mean that the altruists will be wiped out by the struggle for survival? This is an old question, and one that evolutionary theorists have long struggled to answer; nature is full of “self-sacrificing” behavior that seems to contradict Darwin’s belief that the struggle for survival takes place between selfish individuals. To show how the impulsive might beat the calculating, and the cooperative might defeat people who cheat, Frank borrows at one point from evolutionary theorists such as Richard Dawkins, who suggest that when a species prospers it does so because the average members of the species have genes that cause them to do well, not because of the superior genes of particular individuals. But this view works best for behavior that benefits close kin who share a great many genes with one another—which is why J.B.S. Haldane famously offered to sacrifice himself for two siblings or eight first cousins—and that is why it makes better sense of ants, who are virtual clones of one another, than of ourselves, who are not.

A better thought, and it is pursued for all it is worth by Frank, is that evolution selects what one might call a readiness to join in cooperative enterprises, a disposition to sympathize with the aims of others, which encourages cooperation, inhibits defection from agreements, and provides the raw material for a sense of fair play. Raw selfishness encourages its possessor to make an agreement, get the benefit, and then fail to perform his or her side of the bargain—I take your check, but then I abscond with the money and fail to hand over the car I’ve agreed to sell you. This looks like the “rational” thing for the selfish person to do. Why then don’t absconders and cheats flourish more than they do? Roughly, because the people who cooperate with one another can detect them and confine their cooperation to their trusting and trustworthy colleagues, a fact that explains why a certain number of cheats and absconders get away by defeating our detecting abilities. But this mechanism also gives the selfish some reason to be honest. When buying a used car, we look out for salesmen who will cheat us; from the point of view of the salesman this makes it worthwhile having an honest reputation at least. If it is at all hard to get an undeserved reputation for honesty, then honesty in fact will turn out to be the best policy.

What role do the emotions play in all this? Frank devotes much entertaining discussion to this question. At one point he argues that since emotional symptoms of dishonest intentions such as blushing or looking away are hard to suppress, their absence is a good indicator of reliability. Our candid appearance gives some guarantee to others that we will play the game fairly.

That is not all, of course. The label “the emotions” covers an enormous and diverse range of states, from hunger at one, non-complex end to a sense of injustice at the other, elaborate end. Frank speculates that some of our reactions are, in the jargon of the behavioral scientist, “hard-wired,” that is so tightly linked to the physiological system that they are all but impossible to suppress. Hunger is a preeminent example. If we had to decide whether to look for food, we might well starve; having a fierce appetite as it were wired in solves that problem:

When food is scarce, finding something to eat requires great effort, no mean feat for someone who is already weakened by lack of nourishment. For such a person, feelings of intense hunger can mobilize reserves of energy in a way that mere rational calculation cannot.

Other answers are more elaborate, and it is here that Frank makes telling criticisms of the moral effects of education in economics. Take the elaborate emotions involved in resenting an unfair deal. People who are selfish in the ordinary sense make bad employers; they try to impose unfair deals on their employees, and in the process arouse feelings of indignation and resentment that usually make for unhappy companies and poor productivity. Orthodox economic theory implies that such bosses are entirely rational and their employees entirely irrational. If you are a boss in a powerful position, it is irrational to offer your workers one cent more than you need to; if you are a worker in a weak position it is irrational to reject any deal that gives you something more than you currently get. Frank suspects that people who are brought up on a diet of orthodox economics come to believe that it’s only reasonable to be selfish, and thus to turn into the sort of bad boss his view predicts.

Psychologists at Yale ran a laboratory experiment to see how far students behave like the rational egoists of the economist’s theory. The experiment pairs off the students they used as subjects and gives one of them a sum of money to divide with his or her partner; the partner can only accept the division or reject it—if he or she rejects it, neither student gets anything. Rejection is thus a way of punishing one’s partner for an unfair offer, but it is expensive because in depriving him or her of an unfair share of the money, you deprive yourself of any part of it too. Orthodox self-interest theory says that it is irrational to offer your partner anything beyond the barest minimum; equally, it is irrational for him or her to reject any offer at all—since even a cent is better than nothing. The result that interests Frank is that students of commerce and economics are much more likely than students of other subjects to try to get away with an unfair offer; conversely, they are much slower to walk away from palpably unfair offers. They seem to lack other students’ passion for fairness.

Why are economists so different?” Frank asks. One possibility he admits is that the students who took part in the experiments saw them as puzzles in economics, and therefore reacted as they thought economic theory demanded. Frank prefers a less reassuring answer:

Another possibility is that economists are a distinct personality type to begin with. If some people in the population are, by nature, much more materialistic, much less given to emotion, than others, it is easy to imagine many of them being attracted by the teachings of economics

If the message being fed to future MBAs is that ratting on your colleagues is the height of rational behavior, it is hardly to be expected that Ivan Boesky will be the last of the big stealers.

Much of Frank’s account has a slippery quality. He slides back and forth between arguing that “material benefit” is exaggerated as a motive and arguing that we gain our material benefits by a more indirect route than the simplest self-interest theories can account for. But there is a good deal of difference between throwing “material benefit” out of the window and giving it a different place in an argument that nevertheless depends on material benefit for its driving force. Frank never explicitly recognizes the difference. In much the same way he slides between accepting the economist’s definition of “rational” behavior as the behavior of a calculating egoist—and therefore opting for irrational behavior himself—and broadening the definition of the rational by pointing out how rational it is for us to have nonrational responses. Again, his argument lacks rigor in failing to confront the difference.

Still, it is worth remembering that he is writing for economists who too often mistake the axioms of a useful model for deep insights into the human psyche. His book will justify itself if fewer economists continue to think that their discipline “shows” that human beings are fundamentally selfish or that policies that rely on selfish motives will necessarily work better than ones that rely on altruism—a view that the contrast between European health services and those of the United States ought to have demolished many years since. Frank is also writing for students who think that economics is a “tough” discipline, suitable to a tough world, and an appropriate education for rugged individualists out to get what they can for themselves. Reading Passions Within Reason, a few more of them will understand how German and Japanese industry runs rings around American industry by encouraging moral commitment rather than unadorned greed. Above all, he’s writing for the lay reader who might understandably wonder sometimes whether he’s foolish to pay his taxes, to take the trouble to vote, or to give blood. For him, the message is that in general he’s not foolish; the people who are altruistic and cooperative have opportunities the narrowly selfish never get, and there is no intrinsic reason why by doing good we will not also do well.

Diego Gambetta’s collection of disparate essays on the theme of Trust takes off from the issues Frank identifies. It boasts a most distinguished set of contributors, with the past and present provosts of King’s College, Cambridge, among the most prominent. Bernard Williams contributes an acerbic little essay that raises the question whether there is any reliable set of motives that will make people behave in a trustworthy fashion; not surprisingly, he concludes that there isn’t—altruism is always threatened by the risk that it will be too costly to act altruistically, while selfishness is just too hard to keep track of and punish, even, as he drily notes, when we invoke divine sanctions to persuade people that if the law doesn’t get them, God will. Patrick Bateson provides a very dense and austere account of the technical issues that underlie Robert Frank’s invocation of evolutionary theory, distinguishing, as Frank himself hardly bothers to, between simple forms of cooperation and those that require genuine trust, and thus between the explanations appropriate to each.

A grimly satisfying piece is contributed by the editor himself, on the Camorra of Naples and the Mafia of southern Italy. It forms an apt counterpart to Frank’s account of how trust and cooperation are sustained, for it is an account of what happens when they mostly aren’t sustained, and when the only source of reliable commercial information is a criminal conspiracy. Where government is corrupt and the law is not enforced, what are you to do if you wish to buy a reliable horse? You must go to the dealer the Camorra points out to you; left to himself, he might cheat you, but he dare not humiliate the members of the secret society by cheating you after they have introduced you to him. Why does the Camorra not settle down to running a sort of private state, then? Because it is in its interests to keep up a general atmosphere of distrust in which its guarantees become so much the more valuable. This example of selfish isolation, Gambetta plausibly argues, explains much of the spectacular record of poverty, violence, and misery that has been the economic history of the mezzogiorno for the past two hundred years. It implications for the economics of the American inner city will occur to every American reader.

This Issue

May 18, 1989