II. Why Repentance Matters

Suppose a person commits a crime. He is caught, arrested, tried, and convicted. Let us suppose this criminal regrets being caught, but is not sorry for what he did. He shows no remorse. The judge and jury are more likely to be sympathetic with a truly repentant criminal than with one who merely regrets being caught.

Veronica Gonzales and her husband are on California's death row for the brutal torture and murder of their four-year-old niece. The evidence indicated that while both were active participants in the child's death, the husband was more culpable. Veronica tried to argue that since she, too, had been victimized by her husband's violence, she should not be held responsible. Dan Goldstein, the deputy district attorney who prosecuted their case, offered this opinion as to why the jury had voted for the death penalty for Veronica:

The news recently reported the release of Lionel Tate, a sixteen-year-old Florida boy who had killed a six-year-old child, when he was himself a mere twelve years old. He had kicked, stomped, and beaten a little girl who weighed forty-eight pounds. He said he was imitating what he saw on TV, on World Wide Wrestling.²

The mother of his victim, who stated that she had forgiven him for killing her daughter, nonetheless added that she hoped he would stop saying it was an accident. “We wanted Lionel to step up and say, ‘Look, I have done something wrong here. I'm guilty of this. I'm sorry about this.’ ” Florida governor Jeb Bush didn't mention forgiveness, but made a similar point about repentance: “I do hope that Lionel will express remorse, at a minimum, for what occurred.”

Why did the jury care whether Veronica was lying or not? They were probably using truthfulness as a proxy for repentance. If she thought she was entitled to lie, she was unlikely to be genuinely sorry for what she did. This raises the next question: Why did the jury care whether Veronica was sorry? Why did anyone care whether Lionel was repentant or not?

If a person is not repentant, then we suspect that he is either justifying his offense, or indifferent to it. If a boy can justify killing a six-year-old, he is more likely to commit a similar crime. Indifference is not much better. To kill a child in the brutal way that he did, required him to ignore her cries of pain. It is difficult to imagine the callousness of continuing to kick and stomp a small child until she finally died. Anybody, of any age or intelligence level, who could do that once, could probably do it again. Even if his cause-and-effect thinking was so distorted that he didn't understand that the other child would die, this doesn't much matter. He is highly likely to be a problem to other people. In contrast, if he genuinely understands that what he did was wrong, he will be more likely to restrain himself and less likely to give way to his passions, whatever they might be, that would urge him to stomp another child to death.

With this in the background, let us examine the unrepentant person who uses a variant of economic reasoning to justify himself. I will ultimately show that the rational choice economic paradigm can be amended to account for a rational choice both to ask for and to give apologies. First, however, I want to ask whether the naive version of Homo economicus, or economic man, can ever truly be sorry or repentant for anything. I am not going to examine him in the process of committing spectacular crimes like these. Instead, I am going to follow Homo economicus around the house and watch him try to interact with his family. First, let me outline the key assumptions of the economists' approach to human behavior.

III. The Economists' View of the Person

Economists have a very specific view of personal identity: a person is a utility function. A utility function describes a person's preference ordering of every possible set of goods and services. If we want to know how an individual ranks various bundles of apples and oranges, holding all other things equal, the utility function tells us. These preference orderings are assumed to be stable, at least in comparison with the constraints the person faces when making choices. The major constraints economists analyze are prices and income, broadly defined. Any changed behavior we observe is, in general, not the result of changes in preferences, but of changes in these factors that are external to the person.

Even Gary Becker's pioneering work on preferences assumes a stable “extended preference” function.³

Just as personal identity has a particular interpretation within economics, rational choice also means something specific in economics. Homo economicus or economic man uses all available information about the costs and benefits of various courses of action, calculates which course of action maximizes his utility, and carries out that course of action. The exceptions to this general approach have to do with economizing on the use of information. For instance, economists hypothesize that people might devise some cost-effective strategy for sifting through mountains of data, rather than be stymied and overwhelmed by trying to process every available fact. I will have more to say below about these economizing strategies for using new information. With these words of explanation behind us, let us address the main question.

IV. Can Homo Economicus Be Sorry?

He cannot. A rational economic man, who always maximizes his own personal happiness, taking into account all the relevant costs and benefits of his actions, and using fully all the available information, cannot repent of anything. If an individual chooses to live his life using these textbook assumptions behind Homo economicus, he cannot be sorry, and thus will be almost impossible to live with.

To see how this is the case, make a distinction between regret and remorse (as the jury in the Veronica Gonzales case seemed to do). A person feels regret for two principal reasons. A person might feel regret upon discovering information that he did not have in the first place. He had originally taken a course of action, using all available information, but he was not in possession of some crucial fact that only became available as his actions unfolded. If he had known this information initially, he would have decided and acted differently from the way he did.

A person also might feel regret because some probabilistic event did not turn out as he had hoped. We can easily imagine Homo economicus feeling regret because he bought stock knowing that, as the prospectus states, “the stock market is inherently risky and not suitable for all investors.” He regrets that he didn't sell all his shares of JDS Uniphase in 2000 when it was selling for $150, instead of holding onto it until now when it is trading at $3 per share. But he isn't irrational because things might have turned out otherwise. He isn't and shouldn't be repentant because he didn't do anything morally blameworthy. He just made a mistake.⁴

Feeling remorse means something different. Remorse is the feeling associated with an admission of wrongdoing. The person was not just poorly informed, not just mistaken, but wrong, and he admits it. He has some feelings, both about the wrongdoing itself and about admitting it. A working synonym for “remorse” would be “shame.” Being sorry or repentant means admitting I was wrong. This admission is what someone is looking for when they ask for an apology.

What are the possible responses of the rational person to a request for an apology? The other person says something along the lines of: “I don't like what you did. You hurt me. I want you to apologize.” Here are three general categories of responses:

1. I didn't know I hurt you. I didn't know (in advance) that what I was about to do would be hurtful to you. I regret hurting you.

This is, in effect, a plea of imperfect information. I only have regrets that things did not turn out as I had expected and hoped. I am not truly sorry, because I was only ignorant, not actually wrong or guilty of anything. Regret, rather than contrition, is appropriate, assuming I am telling the truth and not rationalizing. If the other person believed that I was acting out of ignorance, he or she might be inclined to forgive me.

2. I wanted to hurt you. I knew it would hurt you. I am entitled to hurt you if I feel sufficiently aggrieved with you, or with life in general. I am allowed to come home from a crummy job and kick the cat, yell at the kids, and snarl at you. I did what I wanted to do. I didn't expect you to like it. I knew you would probably yell at me. But I was then, and am now, willing to pay the price of your displeasure.

This is Homo economicus all right, rational and calculating. But you would not want to be married to him. I doubt very seriously that his friend or spouse would say compliantly, “Oh well, maximizing your utility involved hurting me. Gee, I understand. That makes everything okay.” The friend is likely to either exit the relationship or punch him in the nose. Nobody wants to be held in so little regard as this “rational” person seems to do in this case.

3. I didn't particularly want to hurt you. I hurt you as a by-product of something else I wanted to do. I do not plead ignorance, or imperfect information. Nor can I honestly say that I wanted to hurt you and felt entitled to do so. In this intermediate case, I simply state that what I wanted to do was more important to me than your feelings about it.

Notice that in none of these cases does the person offer a genuine apology. The unrepentant person is, in each case, using a variant of economic man to get himself off the hook for the need to apologize. In the first case, the rational person might be willing to offer a statement of regret; but contrition is not really appropriate, since he was just ignorant, not malicious. In the latter two cases, the rational person knew there was some probability that his actions would displease the other person, so he cannot plead ignorance. Homo economicus believes he is entitled to do what he wants, even if it hurts the other person. In case 2, the extreme case, he not only believes himself entitled to inflict harm, he actively wants to hurt the other person.

Now a reader might object that no one really behaves like this. No one really uses an economic rationale to relieve himself of the obligation to apologize for bad behavior. I reply that I have known quite a few people like this, and most of them are not economists. These are people whose need to justify themselves and unwillingness to apologize borders on the irrational. And oddly enough, their justification for their behavior is that they are acting in their own interest. They might borrow the title of the best-selling book and say they are “looking out for number 1.”⁵

I want to do two things in what follows. I want to show that economic analysis, properly understood or suitably amended, really cannot be used to justify a flat-out refusal to apologize. I also want to use a variant of economic analysis to show why the whole topic of repentance presents so many difficulties.

To do this, I am going to focus for a moment on the intermediate case, in which one person hurts the other as a by-product of something else he wants to do. This case is most interesting, because it is sometimes understandable. It is easy to imagine situations in which we would feel it was acceptable or even appropriate to hurt someone as a by-product of something else we wanted to do. We can imagine situations in which we would feel fairly treated if someone did this to us. Context is everything. This intermediate case also helps us understand why citing rationality as an excuse for not apologizing is more a rationalization than genuinely rational behavior.

V. A Conflict between Homo Economicus and His Spouse

Suppose a husband refused to accompany his wife to a dinner that was important to her. He knows she wants it, and that she will feel bad if he doesn't go. He chooses to do something else instead. He claims he is making a rational choice, by doing his most preferred activity, all things considered. We, as impartial spectators,⁶

Suppose the dinner is an event honoring the wife for winning the Nobel Prize in economics. Homo Economicus (or HE for short) refuses to go because he does not want to miss his card club, which will be meeting during the trip. His plea of rationality would probably not gain much sympathy, either from us, the impartial spectators, or from his wife. We would probably conclude that there was something skewed about his preferences.

Suppose instead that the dinner is at her mother's house. Her mother is a shrew who can't shut up and continually criticizes everyone in sight, including and especially her son-in-law, whom her daughter never should have married in the first place. The husband says he would prefer to do just about anything rather than go to this dinner and endure his mother-in-law's criticism. His wife is Sly Homo Economica (SHE for short). SHE proclaims that she is hurt, demands that he always act to avoid hurting her, and pronounces him an inconsiderate lout.

This set of facts makes the case seem quite different. As impartial spectators, it is difficult to sympathize with the wife. We might even conclude that SHE is using her pain to manipulate her husband.

But one thing is for sure about both of these cases: “rational choice” is not a stand-alone justification for the behavior. No one is obligated to accept the husband's calculation that his card club is more important than honoring his wife's accomplishments. Likewise, her claim to be hurt is not an open-and-shut argument that her husband must instantly comply with her wishes. In each case, the offender (or purported offender) needs to supplement the cost-benefit calculus with a justification for the preferences themselves. The wife in the first case does not think that traveling to Sweden should count as a cost. The husband in the second case does not think that visiting his mother-in-law should count as a benefit. The argument is over these issues, not over whether anybody calculated accurately. In fact, pleading rationality sounds a lot like changing the subject.

In personal situations like these, the rationality argument is a distraction from the main issues of the worthiness of my preferences, and my willingness to accommodate you in the grand scheme of my preferences. The argument is over whether HE is entitled to place such a high weight on the card club, or whether SHE is entitled to place such a great weight on visiting her mother. The hurt that each one feels comes from knowing that he or she is not really very important to the other person.

Here is another sure thing: A person who habitually pleaded the rationality argument, in the absence of these other contextual considerations, would be extremely difficult to live with. In fact, it would be difficult to commit to a relationship with such a person. You could never know in advance how highly your feelings would count in the person's calculus. You would always be wondering whether HE really wanted that card game, or whether HE just did not want to be inconvenienced. You would wonder whether SHE really was hurt, or whether SHE was manipulating you. Because you could not predict all these preferences and choices in advance, you would almost certainly feel aggrieved some of the time.

The person claiming that this is rational behavior could stand to read a few articles on rational altruism or economic theories of social interaction.⁷

VI. The Rational Choice among Mates

As an example of such thinking, one might argue that in choosing to enter a serious relationship with a person, a rational person would consider all of the other person's preferences as part of the decision-making process. If you cannot live with the other person's preferences, you shouldn't get involved and you certainly should not get married. It is plausible to argue that a person could and should screen out completely unsuitable people in this way. One can easily conclude that members of a street gang have preferences so different from one's own that there would be no point in pursuing their friendship.

If we are imagining the choice of whom to marry, however, it is not credible to suppose that one could accurately predict the details of every ranking one's potential spouse might ever make among all possible goods, services, activities, and moral values over the course of an entire lifetime. This approach would press the rationality assumption to an absurd extreme. Homo economicus would not only need to know his own preferences; he would also need to be perfectly informed about the preferences of his spouse and friends. More than that, he would need to be this thoroughly informed about each potential spouse, otherwise, we could not say that he made a rational choice among them.

In effect, this argument for the rational choice among potential spouses takes the economist's view of personal identity to an extreme in the following way. The person is a utility function: he can be perfectly described by all of his preference rankings among every possible combination of goods, services, and moral values. When he goes looking for a spouse, he is shopping among other individuals, who also can be perfectly described by a set of preference orderings. He is shopping among utility functions, with which he can interact. To make a rational choice among spouses, he must be perfectly informed about each potential spouse's utility function.

Friedrich A. Hayek's argument about the limits of knowledge would certainly apply to this situation.⁸

In point of fact, one of the charming facets of lifelong friendship is the element of surprise and serendipity. A friend can bring ideas and activities into my life that I would never have even considered. I cannot possibly have a preference ordering among goods I have never thought of. Since he might bring completely new things to my attention, I cannot very well argue that the decision to be friends with this particular person entailed a tacit decision to honor all of his possible choices equally. I will be delighted by some of the things my friend or spouse brings into my life, disgusted by others, and utterly indifferent to still others.

This is why it is not plausible to claim that a rational choice among spouses makes a person who habitually pleads the rationality argument any easier to live with. To be specific, there is no way that I can make a “rational choice” among spouses that would completely obviate any need to consider the impact my behavior has on my spouse, or that his behavior has on me. I am not necessarily entitled to insist that he always act in a way that pleases me, just because he made a “rational choice” to marry me, and he knew perfectly well what this would mean. Nor can I claim that I am entitled to do whatever I want, without regard to his feelings about it, since he knew what sort of person I was when he married me. I cannot reasonably claim that he has no cause for legitimate complaint because he knew, or should have known, that I was the sort of person to do that sort of thing.

Even if both people accepted the basic premise of the rational choice of spouses, that still would not get them very far. Neither could trust the other to truthfully reveal their preferences if it inconvenienced them to do so. The rationality argument—“I did it because I wanted to, all things considered”—does not help much if you are trying to be in a relationship, trying to understand the other person, trying to have genuine intimacy with the other person. Hovering around the rationality argument is an implied clause: “when I considered all things, you and your feelings didn't get a very high ranking.”

An economist might attempt to bring Homo economicus out of his isolation and into a more social mode of being, by explicitly including friendships and other relationships among the things that give a person happiness. One way to do this would be to enter my friend's well-being or happiness into my utility function. His happiness makes me happy. Alternatively, we might try to capture the friendship's impact on my happiness by saying that spending time together, doing anything or nothing in particular, is a good in itself. Being together gets a weight of its own in my decision making, independently of whatever it is that we are doing. Either of those alternatives allows another person to matter enough to the rational person that he takes the other's responses into account in deciding what to do. Both friends say that they no longer get to do anything they want without considering other people.

Neither of these reformulations of utility completely prevents the genuinely unrepentant person from looking for a rationalization to refrain from apologizing. They just push the problem back a step. “I did what I wanted, all things considered. And by the way, my dear, you know perfectly well that you and your feelings were among the things I considered.” The “by the way” clause is certainly a major improvement over “I did what I wanted and you don't matter.” But I still might owe you an apology on occasion, and I still might want to ask for one from you. If I really do consider your feelings about my actions and take them into account in my calculations about what to do, that does not necessarily imply that you will never be aggrieved with me or disappointed in me.

In a way, the argument that I have already taken you into account makes the problem deeper, because it becomes harder to understand how and why you could ever get mad at me. How could you justifiably be angry, when I have already taken you and your feelings into account? Do I owe you an apology, just because you are mad at me, and have asked for an apology? What exactly would an apology mean in this context?

We are right back to the three choices with which we began: I didn't know it would hurt you. I knew and I wanted to hurt you. I knew and I didn't care.

VII. Is the Demand for Repentance Rational?

Let us turn this question around for a moment to shed some different light on it. Instead of asking whether a rational person could ever be sorry, let us ask ourselves whether a rational person could ever ask for an apology. What might the request, or demand, for an apology mean? Suppose I want you to apologize for something. What is it that I want you to say or admit? And is my desire reasonable?

One possibility is that I want you to admit that you were irrational. You didn't really mean it when you assigned an inappropriately low weight to me and my feelings. Or you didn't really know the impact your actions would have on me. This interpretation of an apology corresponds to the argument from ignorance. If I accept an apology based on ignorance, I am in effect letting the person off the hook.

Another, deeper possibility is that I want you to change. You admit knowing what you were doing; and I agree with that, but it doesn't make me feel any better. It actually makes me feel worse. My problem with you is that I don't like what you wanted. I want you to apologize, because that would tell me you agree that there is something wrong with you and your preferences. Either you assigned too low a weight to my feelings, or your values are so different from mine that I cannot accept or respect yours. The wife in our example might say, “Thinking your card club is more important than my Nobel Prize is beyond the pale.” Or the husband might say, “Telling me that I don't love you unless I put up with your mother's nagging is telling me that your mother is more important to you than I am.”

Each spouse can say to the other, with some justification, “I cannot accept your preferences. I don't have to change my preferences in this particular matter, and you can't make me. I think I am right, and I am not going to be bullied out of my position by your claim to autonomy, or your claim to be hurt, or your right to choose, or anything else.”

The request for an apology might have a third possible meaning: Work with me. Help me understand. I want to understand. Offer an account of yourself, beyond “I did it because I wanted to.” Give me an explanation of your preferences, your values. I would really like to be on good terms with you, and not be mad at you. But if you can't offer me an explanation that satisfies me in some form or fashion, I give myself permission to be mad at you.

This third case illustrates that justification is really the heart of the issue. If I ask you for an apology, I am asking you to justify yourself. If you can't, then I am justified in being angry at you because you did something I didn't like.

Whether the person offers an explanation of his behavior, or something closer to a genuine apology, the demand for an apology has placed some costs on him. This helps explain why the victim would go to the trouble of asking for an apology. The hard work of apologizing functions a bit like posting a performance bond for good behavior. The offender bears some cost of apologizing. By bearing this cost, the offender signals that he is willing to take responsibility for his behavior and the impact it has on others. He is not going to keep trying to see if he can wear the victim down, and eventually get away with it. Nor is he counting on the victim to continually monitor and correct him. An apology, particularly a costly apology, helps the victim feel that the offender is really serious about improving his behavior.

So we can understand why someone might ask a normal person for an apology. Does it make sense, however, to ask Homo economicus for one? Rational economic man has his justification prepared: he has already optimized. He has considered all costs and benefits, including the impact of his actions on other people. He cannot really be sorry for anything. He has already done the best he can do. A lame sort of “I am sorry you were hurt” is as close to an apology as we're going to get from him.

So asking Homo economicus for an apology is not rational, since we know in advance what he is going to say. The rational choice is to save our breath and not bother. We are forced to conclude that the naive version of rational economic man can neither give nor request an apology. We need some explanation for the fact that people do request and receive apologies.

The philosophers and theologians who are skeptical of the rational choice model altogether are perhaps chuckling at this point. They might think the obvious solution is to throw the whole rational choice apparatus overboard, along with Homo economicus and all his relatives. But this economist is not quite ready to admit defeat. There may be a way to account for this apparently irrational behavior.

VIII. Willful Blindness

We can get some insight by turning the question of repentance around once again: Why should repentance be so difficult? My spouse didn't like what I did. I didn't know he would be so upset about it. If I had really known all about the impact my actions would have on me and others, I would have chosen otherwise. When he asks for an apology, why do I resist? Why not just say, “Gee, I didn't realize this would bother you so much. I am sorry.”

And of course, in many cases, this is exactly what we do. Oftentimes, the issue at hand is not a big enough deal that the offender has any problem offering an apology. In some situations, it really is a matter of new information: I honestly didn't know this was going to upset you.

But these “no big deal” cases do not exhaust the full set of situations in which people either ask for or give an apology. When a person has resisted giving an apology, and finally pleads ignorance, her partner might still be dissatisfied. He might very well respond, “What do you mean you didn't know how much it would bother me? How could you have not known? I don't believe you didn't know.” The truth may very well be that the other person did not want to know.

This is the real source of the problem, both for our hypothetical married couple and for rational choice theory. My spouse has a different, and perhaps more realistic, picture of my behavior than I do. He wants me to change because I am giving him trouble in some way and he wants to be more comfortable. If I saw myself the way he saw me, I might agree that it is somehow in my interest to change. But I resist knowing this, especially if my husband is in my face about it. Whether his assessment of the situation is correct or not has little bearing on how easy it is to listen to.

This refusal to see is quite often at the source of the demand for an apology. One person feels wronged and the other sees no need for repentance. The demand for an admission of guilt can be a demand that the other person stop his willful blindness and see, really see, the impact that his actions have had on the person who has been wronged.

This phenomenon of self-deception also helps us to see why the plea of ignorance can be so infuriating to the person who feels wronged. If somebody is clueless about his own behavior, the claim of ignorance is not really satisfying. “I didn't know this bothered you” is an explanation that can easily be worn out by overuse. After hearing this a few times, the other person doesn't care whether you didn't know. Not knowing is what they want you to get over.

An example of this phenomenon took place during the 2004 Super Bowl, the major championship game in American professional football. During the halftime show, broadcast on the CBS network, singers Janet Jackson and Justin Timberlake “accidently” had a “wardrobe malfunction” that resulted in Jackson's breast being exposed on nationwide television. Brent Bozell, president of the Parents Television Council, made precisely this point that a plea of ignorance is no substitute for an apology: “We do not accept the apology of CBS, nor do we accept the statements of regret by MTV [which produced the halftime show]. It is absolutely reckless for CBS to claim that it had no prior knowledge that such activity was likely to take place. MTV is known for exactly this type of conduct. MTV and CBS are both owned by Viacom. Whether or not CBS executives did or did not know in advance is irrelevant; Viacom executives had the duty to know.”⁹

Willful blindness presents a problem for rational choice theory as well. Rational choice theory would have us believe that I would always welcome accurate information about myself. But I think every adult knows what I am talking about when I say that I sometimes resist knowing the truth about myself. If my husband is in my face about an irritating habit, knowing he is right does not make me feel any better. In fact, I would rather he bugged me about something that wasn't true, so I could dismiss him as an idiot. (How a Rational Economic Woman chooses to marry an idiot, or convinces herself that she has chosen to marry an idiot, is an interesting problem I shall not pursue.)

I somehow doubt that this refusal to see the truth about oneself is unique to the Morse household. Nor is it confined to married couples. I imagine that every reader can picture someone who matches one of these descriptions of the willfully blind. The drunk in denial refuses to see the harm he does to himself, as well as to others. An abusive person thinks the other person has it coming, or that she or he really doesn't mind. The neglectful parents believe their child is really okay, even though he is alone watching TV in his room all the time. The self-important snob talks constantly about himself without noticing that no one is listening and everyone is avoiding him. The indulgent parents do not see that the complaints of their child's teacher are valid. Instead, they stick up for their precious little darling who has never done anything wrong in his whole life. The control freak can never manage to see that other people might possibly have something to contribute, and cannot understand why no one feels like cooperating with her.

The phenomenon of self-deception, then, is not an obscure or insignificant one. People are capable of bearing substantial costs rather than embracing certain kinds of information about themselves. The challenge for rational choice theory is to offer an account of why rational people would resist absorbing accurate information, freely offered. This is a particular challenge when the information is about something as intrinsically important as how we affect the people closest to us.

IX. The Wish to Think Well of Oneself

Putting this phenomenon of self-deception into a rational choice framework requires specific assumptions about the content of people's preferences. We can account for the refusal to see the negative impact we have on others by hypothesizing the existence of a particular good: the wish to think well of oneself. Virtually everyone has this good in his or her utility function to some extent. Positing the existence of a preference for something like this good helps account for our resistance to updating our beliefs about ourselves. For convenience, I suggest a shorter, more compact term for this thing we seem to value so much: pride.

We can see how pride plays a part in generating willful blindness by looking back to my short list of deluded individuals in the previous section. The alcoholic does not want to surrender his image of himself as a sober, upstanding citizen who can hold his liquor as well as anyone else. The abusive person does not want to surrender either the power he has over his victim, or his fantasy that his violence is somehow the other person's fault. The neglectful parents, who are perhaps absorbed in themselves, do not want to see that their child needs something from them. Nor do they want to surrender their view of themselves as good parents because they are good providers. The snob especially enjoys his good opinion of himself. He is totally deluded about the fact that other people lost interest in him long ago. The indulgent parents view their child as an extension of themselves. When the child causes problems at school, these parents prefer to blame the school rather than correct the child. Not only is blaming the school easier than correcting a difficult child, but this helps them preserve their image of themselves as good parents. All of the control freak's self-esteem depends upon being in control of every detail of every situation. Giving way to another person, or even giving credit to another person, feels like an intolerable loss of status.

In all of these cases, the cost of accepting negative information about the self is that we must forgo some part of our good opinion of ourselves. We must bear this cost in addition to any intrinsic difficulties of changing our behavior. This is a rational choice approach to accounting for the self-deception we see all around us. The cost of embracing accurate, but negative, information about ourselves is that we must surrender our pride.

The rational choice approach would say that a person will only embrace negative information about himself when the cost of remaining ignorant exceeds the benefits of clinging to his good opinion of himself. This would account for the common perception that an alcoholic has to “hit bottom” before he will be willing to sober up. The alcoholic must overcome more than the physical costs of going through withdrawal. If the physical costs were the only issue, alcoholics would talk about their situation quite differently from the way they typically do. They would say something along the lines of: “I recognize that I am addicted to alcohol. I choose not to give up drinking because quitting is too hard.” The addict more often claims he is not addicted, that he can quit anytime he wants to, that other people are exaggerating the problems, and so forth. We surmise that the greater cost to the typical alcoholic is the cost of admitting that he is not the upstanding citizen he imagines himself to be. It is a testimony to the power of pride that the costs of continuing to drink have to be enormous before the alcoholic will make a serious effort to change.

The economists' way of looking at things also helps us see why the sin of pride has always been considered a deadly sin. The wish to think well of oneself is not wrong in itself. If that wish induces a person to ignore or resist accurate information, however, then it has morphed out of something legitimate and understandable and into something more serious, and even sinister. It is always irrational to refuse to use accurate information, when that information is presented to us without search costs or processing costs. Therefore, an economist would conclude that pride at the cost of self-deception is not rational. And that is really saying something, since “irrational” is the only word economists have that means “bad.”

We might rescue this individual from the charge of irrationality by invoking the concept of satisficing, rather than maximizing. This hypothesis about information management suggests that people are not in a continual frenzy of gathering more and better information. Rather, people make a rational choice about how much information will suffice for the purpose at hand. In the present example, the individual does not continually update his beliefs about himself and his own preferences as a result of the input from other people. Instead, he waits until he has accumulated a critical mass of vital information to make an adjustment. He is willing to absorb some costs associated with bothering other people and losing their esteem in the meantime.¹⁰

This argument does not entirely rescue our self-deceiving friend. It only moves the behavior from the category of irrational and into the category of rational satisficing. But satisficing is a rational strategy only if there are substantial costs of gathering and processing information. In these cases of self-deception, gathering the information is not the problem. Mrs. Homo Economicus is literally right in her husband's face with more information than he wants to hear. It is the cost of processing the information that the person does not want to bear. His self-deceptive behavior is incomprehensible without seeing that forgoing pride counts as a substantial cost. Plainly, the individual would be better off with a lower cost of processing negative information about himself.

X. The Costs and Benefits of Apologies

So this is, in the end, what is so difficult about apologizing. Apologizing requires us to admit that we were wrong. The cost of that admission is surrendering some of our good feelings about ourselves. We give up some of our pride. Seeing this helps us to see what a person is looking for when he asks for an apology.

When someone asks for an apology, he is asking the other person to concede that there was something wrong about the other's preferences. If I ask you for an apology, I want you to admit that you were wrong in what you wanted, in the weights you assigned to different things, and in particular, in the weight you assigned to me and my feelings. This is exactly why giving an apology is so difficult. We have to admit we were wrong in what we wanted, as well as in how well we pursued our own interests.

The economists' perspective actually helps us to see just how acute this difficulty is. Economists define the person by his preferences as a matter of analytical convenience. This is not the same thing as hypothesizing that individuals literally identify themselves with their preferences. But if people choose to identify themselves in this way, they are less likely to surrender their preferences. Admitting to some wrong in the preferences is admitting to some substantial disorder in the soul. If you tell me there is something wrong with what I want, we are not just having a minor difference of opinion. You are telling me there is something intrinsically wrong with me. In other words, an attack on a person's preferences amounts to an attack on the person's identity. It is a lot more problematic than somebody saying they don't like my hairstyle. I can change my hairdo more easily than I can change my preferences.

To get a better idea of the point, think about someone for whom hair is identity. Many teenage girls have their identity all wrapped up in their appearance. If somebody tells her she has bad hair, a teenage girl becomes hysterical. The same comment to a more mature woman, or to a man of any age, is likely to be a big yawn. The teenage girl says to herself, “I am my appearance.” The extreme version of economic man says, “I am my preferences.”

Some features of the problem of repentance suggest that this view of the self has some plausibility. For instance, this helps account for the intense conflicts that sometimes arise over what ought to be simply matters of taste. We might think that arguments over preferences would be fairly tame. You want what you want; I want what I want. Live and let live.

But this turns out to be not quite true for the person who identifies himself with his wants. A complaint about tastes is more like a personal attack than a simple difference of opinion. If you make such a complaint, I have to put a stop to this assault on the interior of my soul. I can either sever my relationship with you, so I don't have to listen to you and your nonsense about my preferences being bad, or I can argue with you about your views, and hope that I can bludgeon you into submission. In neither case is the “live and let live” position quite stable.

Taking the economists' approach seriously, then, helps us to zero in on a couple of features about repentance we might not otherwise have noticed. The more reluctant a person is to receive negative information about himself, the less likely he will be to be repentant or to apologize. Self-deception is almost synonymous with a reluctance to absorb negative information about the self. A person's level of self-deception is related to two other factors: his level of pride and the extent to which he identifies himself with his preferences. In particular:

1. The greater the person's pride, the less open he will be to negative information about himself. Therefore, increases in pride will be correlated with increases in self-deception and other forms of irrational behavior. Increases in pride will also be correlated with a greater reluctance to apologize.

2. The more closely the person identifies himself with his preferences, the less open the person will be to negative information about those preferences. He will therefore be less willing to apologize. The more someone identifies himself with his preferences, the more defensive he will be about any confrontation regarding the content of his preferences.

Changes in preferences really are the heart of repentance. When we ask someone to apologize, we want him to do more than say he made a mistake. We want him to have new preferences so that he will be less likely to want to do the things that were hurtful to us. We want him to monitor himself, so we don't have to confront him every time. And if we do have to confront him, it will be a lot easier if he already understands that he needs to change. If he adopts new preferences, we don't have to convince him each and every time that there was something twisted about his original preferences.

We can put this into economic terms by observing that both excessive pride and excessive attachment to one's preferences operate as increases in the costs of processing new and accurate information about oneself. Higher processing costs make a person worse off than lower processing costs, precisely because these costs act as a barrier to using accurate information.

This is why it is a mistake for our unrepentant friend to claim he is using rational choice theory to justify his refusal to apologize. He is unwilling to say that his pride is more important than the relationship, that his good opinion of himself is more important than his spouse's feelings about him. He is unwilling to update his preferences to include significant information about the impact his behavior has on other people. He is unwilling to offer an account of his preferences. He expects the people in his life to “take his preferences as given,” the way an elementary economics textbook might do. Hiding behind rational choice theory is a rationalization, not a reason.

XI. Put on the New Man

The issue of repentance raises the question of whether it is safe to assume that preferences remain unchanged. Some cases of remorse and apology are straightforward; others are life-changing. The repentant person is a new person.¹¹

Recall Lionel Tate, the Florida teenager who killed a six-year-old girl. Part of Lionel's process of repentance will have to include a changed understanding of himself. If his wish to think well of himself requires him to justify his every act, no matter the consequences to others, his behavior is unlikely to change. However, he may be able to change his behavior, if he is able to say, “I killed someone. I was wrong to do it. My reasons were not good enough. My reasons were excuses.” He will, at least temporarily, think badly of himself and what he did. He has to find a different way of thinking well of himself. If he is really going to change, he has to both see what he did and develop a desire to behave differently.

The unrepentant soul claims that he is entitled to have preferences that are completely self-contained and self-referential. The problem of self-deception gives a hint that there is something mistaken about this. If a person places too great a weight on pride, and if he is too strongly attached to his preferences, he will behave in an utterly irrational manner, in exactly the economists' sense: he will ignore accurate information. That accurate information concerns the impact he has on other people. Our subjective feelings may well be facts, but they are not primal, nonnegotiable facts. There comes a point at which our feelings are just not the most important fact about a situation.

A person who is truly committed to rational behavior will be willing to accept all accurate information, even if that information implicates his preferences in some way. This person is more rational, but less apt to identify himself closely with any one set of preferences. This person, if he thinks in terms of preferences at all, will be more likely to be looking outside himself for guidance about how to behave. He may very well end up with a higher level of utility than would be possible with any one set of closely held preferences. In plain English, he is more likely to be happier more of the time, if he is less attached to the details of his preference ordering and more attentive to the information he can obtain from outside himself.

Once the person has made this kind of commitment, it seems that he can reasonably say, “I like my new preferences better.” The vantage point from which one can say, “I like my new preferences better,” has to include a substantial component of input from the world outside the individual himself. We need somehow to make that statement from a position outside, or above, both our current and our previous preferences. In other words, we need a God's eye view.¹²

Now, one need not claim to have a personal telephone line to God Himself in order to make the claim that the God's eye view would be superior to a purely subjective view of our preferences. Indeed, we need not even believe that such a God exists, in order to understand the idea that if He did exist, His perspective would be superior. God knows more than I do. God has a longer time horizon than I do. He can see how my actions will unfold over the long term, for a larger number of people, far better than I can. Finally, the God's eye perspective takes the good of more people truly into account. Since I am a social being, I actually do care about others. Since I am limited and finite, I am unable to fully comprehend or apprehend the impact I have on others. Therefore, God's way will be superior.

We could intelligibly argue that the more closely I align my preferences with the preferences God has for me, the happier I will be. I can meaningfully say, “I like these better.” Under this approach, rationality means something quite different from successfully maximizing utility, using a completely subjective, self-derived set of preferences. A person is more rational, the more closely he conforms himself to the standard that exists outside himself.

Even without a specific religious tradition, we might still be able to discover something about the God's eye view, or at least, a view broader than our own preferences. Repentance is a key part of this process. For the process of repentance compels a person to recognize that he is not God. Part of being repentant is admitting that I don't know everything, that I am not always right, and that I do not always care as much about others as I would like to think. It is in my interest to listen to what other people have to say.

That willingness to take in new information is part of what allows a person to detach from a particular set of preferences, and change. Being detached from particular preferences makes a person more rational, in exactly the economists' sense of using more information at lower cost. Being detached from our high opinion of ourselves also reduces the costs of processing negative information about ourselves. This openness to new information is also part of what allows people to be in genuine relationship with others. We can be less self-referential and more able to actually grow. One need not believe in an omniscient God to see that having some detachment from the self and openness to others is superior to trying to maximize utility based on a completely self-contained, self-defined set of preferences.

In fact, any moral code based upon a “living exemplar” (a virtuous person used as a model) offers a way out of the closed loop of personal preferences. Aristotle enjoins us that the way to become virtuous is to practice virtuous acts, under the tutelage of a virtuous person.¹³

Of course, we must make a decision about whom to follow; and of course, we might pick out a knave or worse. There is no escaping the possibility of self-deception in the service of self-gratification. But if we choose to focus entirely on the self, we excuse ourselves from even trying to choose a guide to follow. The choice of an exemplar forces some process of self-examination and self-reflection that pulls people out of their natural preoccupation with themselves.