How Selfish Are People—Really?
There is, however, also a reexamination of rationality going on inside the economics business. This effort seeks not so much to overturn the idea of universal competition as to take it to a new and subtler level of understanding. If history is any guide, this is the development to watch, for as Paul Samuelson likes to say, economics will be changed by its friends, not its critics.
Change there certainly is. Efforts to produce a theory of cooperation or of altruism suggest that much of the certitude about the nature of man that economists have advanced these last 100 years may have been misleading.
There may be a good and logical foundation for doctrines of loyalty and sympathetic understanding after all.
Perhaps the best-known book to have opened up new avenues in the study of human behavior (at least along the economic axis) is Robert Axelrod’s The Evolution of Cooperation. From its beginnings nine years ago as a report published in the Journal of Conflict Resolution on a computer tournament among diverse strategies, the argument grew to become a highly successful article in Science magazine (
it won the Newcomb Cleveland prize in 1981), then a book published to wide acclaim in 1984, then a paperback issued a year later. Since then, it has been extensively discussed, taught in business schools, employed in arms limitation talks, consulted by labor negotiators.
Axelrod begins his analysis with the familiar prisoners’ dilemma, an illustrative exercise that has been one of the dominating features of the landscape since game theory first brought considerations of strategic behavior to economic theory 40 years ago.
In this situation, two prisoners are accused of a crime, which they did in fact commit. The jailers structure the payoffs to encourage each prisoner to confess: if neither prisoner confesses, both are given light jail sentences of, say, one year. If one prisoner confesses while the other remains silent,
the first goes free while the other receives a heavy sentence of, say, ten years. If both prisoners confess, both get the heavy sentence, but with time off for good behavior—say, five years. Neither one knows what the other is going to do.
Clearly, each player does better by confessing than by remaining silent: if he confesses and his partner doesn’t, he goes home immediately, while if he and his partner both confess, they each get five years instead of ten.
So the question is, why would either ever stand pat and say nothing? How is it that cooperation ever gets started?
The answer, it turns out, lies in repeated play. Researchers before Axelrod had noted that the tendency to cooperate in prisoners’ dilemma games increased dramatically whenever a player was paired repeatedly with the same partner. In these circumstances,
a strategy called Tit for Tat quickly emerged: cooperate on the first move, then follow suit on each successive move; cooperate if your partner cooperates, defect if he defects, at least until the end of the game is in sight (then defect no matter what).
This strategy has, of course, been known at least since Biblical times as “an eye for an eye, a tooth for a tooth.”
In a recent survey of the work since the publication of his book, Axelrod wrote that cooperation based on reciprocity had been noted in everything from vampire bats to vervet monkeys to stickleback fish, and that advice based on the theory had been offered for problems in breaches of contract, child custody arrangements, superpower negotiations, and international trade.
We were constantly gaining a better understanding of the conditions in which cooperation would arise, he said; light had been cast on the significance of variations in the number of players, the payoff structure, population structure and dynamics, and the “shadow of the future,” meaning the prospect of retaliation.
The study of cooperation was well established and growing, Axelrod said; cooperative behavior could be taught.
What Axelrod forcefully contributed was the much-prized quality of robustness. He showed that Tit for Tat players in reiterated games would find each other and accumulate higher scores than meanies who always defected. He demonstrated how clusters of Tit for Tat players might invade an evolutionary game and win.
He generalized the strategy and found that Tit for Tat worked well against a wide range of counterstrategies simulated on computers as well as in biological systems from bacteria to the most complex species. He published his computer tournament results and proofs of his theoretical propositions.
For nonexperts, the real persuasive power of Axelrod’s argument lay in the variety of real world situations he found to which Tit for Tat applied. Businesses really did cooperate, extending each other reciprocal credit, until liquidation loomed.
Then trust fell apart, and even old associates vied with each other to see who could file the quickest writs. Elected representatives really did learn to cooperate, for if they didn’t learn to produce legislative results through logrolling, they weren’t reelected.
But the dramatic centerpiece of Axelrod’s book is a long analysis of the live-and-let-live system that evolved in between the large battles of World War I. Generals could force soldiers into battle whenever they could directly monitor their behavior; but when headquarters wasn’t watching, the soldiers restored tacit truces.
The key to the system was that soldiers in the trenches rarely moved; they got to know each other, and became, in essence, partners in an oft-repeated prisoners’ dilemma game. When one player “defected,” the common penalty response was an exchange of two-for-one or three-for-one.
A French soldier explained, “We fire two shots for every one fired at us, but we never fire first.” This brief historical excursion is a convincing proof that cooperation could evolve among even the most desperate of egoists, those who had been issued rifles and ordered to kill.
For humanists, however, and those scientists who are troubled by the conviction that there is more to human nature than the purely selfish, even this description of cooperation through reciprocity is disappointing.
Axelrod’s work is built firmly on the foundation of self-interest. In a sense, his prisoners’ dilemma is no dilemma at all to those who see human choice as strictly rational. There is no divided loyalty here, no painful choice, just a simple calculation.
Choose the course with the bigger payoff now: cooperate if you think you are going to play again, stiff your partner if you think you won’t see him again. There is no reason to feel embarrassment; cheating is the rational thing to do as long as you don’t expect to be caught.
The trouble is that there is a wide range of familiar, everyday behavior that we all know doesn’t square with this logic. Travelers still leave the requisite tip in restaurants in cities to which they will never return. Citizens vote in elections even though they know that their vote is extremely unlikely to make a difference.
People help strangers in trouble. They willingly bear costs in the name of fair play. They remain married in situations in which it would clearly pay to cut and run.
A highly imaginative approach for dealing with such instances, and for extending economics to the realm of the emotions in general, is proposed in a new book by Robert H. Frank.
Frank, a Cornell University professor, spent ten years performing the comparatively humdrum duties of a teacher before going to Washington, D.C. as Alfred Kahn’s chief economist at the Civil Aeronautics Board.
Kahn moved on to serve as President Jimmy Carter’s “anti-inflation czar” and Frank remained behind to help close up the CAB. When he returned to Cornell, a couple of remarkable books tumbled out, sufficient to place Frank on leading lists of the half-dozen most interesting mid-life economists working in the United States today.
Choosing the Right Pond: Human Behavior and the Quest for Status is an exploration of status fairly bursting with novel ideas about why people tend to organize themselves into leagues. It is the kind of book that any reader, perhaps especially readers of this magazine, can pick up and browse with pleasure.
Now, with Passions Within Reason, Frank has written a somewhat tighter and more demanding book. But it is the one that is destined to help change the way we think about the basis of ethical behavior.
Frank’s starting point is to take emotions as a given. They exist, he says. They’re probably not the “fuzzy thinking” that most economists believe them to be. We see a homeless person, we are moved to pity; we see a child in danger, we are moved to help; we see a sterling baseball play, we are stirred and excited;
we imagine our mate with another person, we burn with jealously and rage; we contemplate stealing from an unattended change box, we blush with shame. Thinking as an evolutionist, Frank asks, what useful purpose might these feelings serve?
The answer he gives is that the highly useful function of the emotions is precisely to short-circuit narrowly self-interested behavior, because honest and helpful people are those whom everyone wants for partners, and because nobody messes with people who get angry when they are crossed.
It is well known that the ball hog doesn’t make the team, that, in the end, the utter egoist doesn’t win at romance; the existence of mitigating emotions is evolution’s way of making us more “fit” partners.
For Frank, emotions are a way of solving the “commitment problem”—the fact that, for society to work, people have to make binding commitments that can later require otherwise rational actors to behave in ways that seem contrary to their self-interest.
There are any number of everyday situations where common sense dictates that it helps to have one’s hands tied by emotional predispositions.
If you want people to trust you, it helps, not hurts, to blush when you tell a lie. If you want people not to take advantage of you, it helps, not hurts, to be known as someone who will fly into an irrational rage if you are cheated.
The self-interest model counsels that opportunists have every reason to break the rules when they think no one is looking. Frank says his commitment model challenges this view “to the core,” because it suggests a compelling answer to the question, “What’s in it for me if I’m honest?”
Frank writes, “I am still annoyed if a plumber asks me to pay cash; but now my resentment is tempered by thinking of (my own) tax compliance as an investment in maintaining an honest predisposition.
Virtue is not only its own reward here; it may also lead to material rewards in other contexts.”
The trick here is that, in order to work, your emotional predisposition must be observable; in order for evolutionary processes to produce the kind of emotionally based, altruistic behavior that interests Frank,
cooperators have to be able to recognize each other. Moreover, an emotional commitment must be costly to fake; the Quakers grew rich on the strength of their reputation for honest dealing, partly because it takes just too much time and energy to become a Quaker in order to take advantage of the opportunity to cheat. Any Quaker you meet is almost bound to be honest.
The same principle applies to the rich set of linkages between the brain and the rest of the body, according to Frank. Posture, the rate of breathing, pitch and timbre of the voice, facial muscle tone and expression, eye movement—all these offer clues to a speaker’s emotional state.
An actor can fake them for a few minutes, but not more. Even a baby can discriminate between a real smile and a forced one. Humans have evolved this complicated signaling apparatus because it is useful in communicating information about character. And forming character and recognizing it is what emotions are all about.
For Frank, moral sentiments are like a spinning gyroscope: they are predisposed to maintain their initial orientation. Nature’s role is to provide the gyroscope, in the form of “hard wiring” between the body and brain; culture’s role is to provide the spin.
In the end, Frank sees his commitment model as a kind of secular substitute for the religious glue that for centuries bound people together in a compact of mutuality and civility. To the question, “Why shouldn’t I cheat when no one is looking?”
Frank notes that religion always had a compelling answer: “Because God will know!” But the threat of damnation has lost much of its force in the last century or so, and “Smith’s carrot and Darwin’s stick have by now rendered character development an all but forgotton theme in many industrial countries.”
The commitment model offers a way back to good behavior based on the logic of self-interest: gains will accrue almost immediately to those who become trustworthy characters. In this view, no man is an island, entire of himself, for each is a part of the other fellow’s utility function, thanks to the biological adaptation of the emotions.
Does this make sense? Of course it does. What Axelrod and Frank have in common is that each has offered an account of how “nice” people survive and thrive in the economic world—why they aren’t automatically competed out of existence by persons who are more relentlessly self-seeking.
What makes Frank’s approach more appealing is that it treats emotions as observed facts of life and attempts to account for them rather than immediately rationalizing them away as a regrettable imperfection of the spirit. He gets at what we really mean by “honest”—as opposed to merely prudent behavior.
There are still other explanatory approaches to this situation, in some cases even more promising. Herbert Simon, for example, has proposed a trait he calls “docility”—meaning susceptibility to social influence and instruction—that would contribute to individual fitness and so explain altruism within the framework of natural selection.
Such evolutionary approaches may yield more understanding of the rise of the complex organizations that populate the modern world economy than reasoning about the equilibrium of the firm.
Whatever way you cut it, the “news” from economics is beginning to confirm what most working people know in their bones: that integrity and fellow-feeling are highly effective forms of individual fitness.
When you consider the amount of time and effort that goes into the moral education of the child, the claim of the economists that there is self-interest and only self-interest is preposterous.
In general, children learn the Golden Rule in kindergarten. Religious traditions introduce them to the absolute prohibitions of the Ten Commandments. In families they learn the role of the conscience and are introduced to many forms of cooperation, including frequent self-sacrifice in the interest of the group.
In schools they learn to be members of cliques, dividing their loyalties between friends inside and outside their gangs. In sports they learn teamwork, including the lesson that nice guys finish all over the standings; as spectators, they learn that fan loyalty may pay off, as may the lack of it.
In love and war they learn sympathetic understanding, and they return constantly to the narrative arts (TV, movies, talk shows, novels, and biographies) to exercise and replenish their understanding.
They may even go to military academies or business schools to learn more intricate forms of cooperation before going out into the world of large organizations to practice it.
Character development, in other words, is far from “forgotten” in industrialized countries. Instead, it is simply ignored by most economists while practiced by nearly everyone else—including most economists.
If practitioners may now turn to economics to learn that the conscious pursuit of self-interest is often incompatible with its attainment, so much the better—for economics.
Most of us will continue to disregard the utterly premature claims of economics to “scientific” certainty about the intricacies of human nature.
We will continue looking to the humanistic tradition for our instruction in ethics, as we have all along.