Vol. 11 No. 6 (June 2001) pp. 283-287.

LAW AND SOCIAL NORMS by Eric A. Posner. Harvard University Press, 2000. 260 pp. Cloth $39.95. ISBN: 0-674-00156-7.

Reviewed by Jack Wade Nowlin, University of Mississippi School of Law.

Eric A. Posner, a law professor at the University of Chicago, has filled a serious gap in contemporary legal scholarship with a much needed, highly original, and important study of non-legal forms of cooperation and their relationship to law. Posner's book has three essential goals. The first is to show "the value of concepts from game theory for understanding legal issues." The second is "to persuade the reader about the usefulness of [his particular game theory model] for illuminating a variety of legal issues." The third is "to persuade the reader of several substantive claims about the relationship between legal and nonlegal forms of regulation." Posner's substantive claims include the following: (1) that social norms are best understood as "behavioral regularities" that emerge in the "absence of organized, conscious direction by individuals" and that result from "individuals acting in their individual rational self-interest;" (2) that "many legal rules are best understood as efforts to harness the independent regulatory power of social norms;" and (3) that social norms are not necessarily functional or efficient, but rather their value is "mostly a matter of historical accident" (pp. 7-8). Posner's book is generally highly successful in its pursuit of these goals, though a number of his readers may question his strong methodological assumptions grounded in rational choice theory.

As noted, Posner's methodology draws on work in game theory and uses the insights of that approach to construct a general model of nonlegal cooperation based on "repeated game" situations and signaling behavior. Posner contends that even in the absence of a legal system, nonlegal forms of cooperation would emerge from the "rational choices" of individuals, given the benefits cooperation creates for those involved. Since it is often the case that one participant can maximize his short-term gains by breaking an agreement, defection or "cheating" in cooperative agreements, however, remains a constant threat to such endeavors. Even so, Posner maintains that defection is often deterred by fear of reputational injury, which makes participation in future cooperative endeavors more difficult. Once an individual is recognized as someone who cannot be trusted to adhere to cooperative agreements, his opportunities to form such agreements naturally diminish. Therefore, those individuals who possess a low a "discount rate" (i.e., those who place a high value on the potential benefits of future cooperation) are less likely to cheat in cooperative endeavors in order to preserve their reputation (and thus future opportunities) for cooperation. In the simple words of Benjamin Franklin, honesty is the best policy.

Even so, individuals seeking to form cooperative agreements will also need to determine the discount rate (and thus

Page 284 begins here

trustworthiness) of potential partners-as well as make sure that others know about their low discount rate. In a world of imperfect and limited information, individuals must often rely on proxies for trustworthiness and/or the "signals" individuals send out to demonstrate their low discount rate/cooperativeness. "Good types," as Posner calls individuals with low discount rates, are those "[p]eople who care about future payoffs" and who will thus not only "resist the temptation to cheat in a relationship, but will also "SIGNAL their ability to resist the temptation to cheat by conforming to styles of dress, speech, conduct, and discrimination" (p. 5). The signaling of low discount rates can occur in a variety of ways, and Posner uses the term SOCIAL NORM to refer to the resulting behavioral regularities that occur through signaling behavior. In short, then, a social norm is the product of a "signaling game in which people engage in behavioral regularities in order to show that they are desirable partners in cooperative endeavors" (p. 5).

Of course, "bad types" or defectors will also attempt to send out signals of trustworthiness and cooperativeness, confusing potential cooperative partners. In the signaling game, either a "pooling" or a "separating" equilibrium can occur. A pooling equilibrium is one in which "good types" and "bad types" are indistinguishable, and it occurs when signaling costs are low enough that both types can afford to send the signal. A separating equilibrium is one in which "good types" and "bad types" are distinguishable and occurs when the cost of a signal of cooperativeness is sufficiently high that only "good types," who are counting on long-term benefits from the cooperative endeavor to recover their initial investment, can afford to send it. For instance, one can persuasively demonstrate a "low discount rate" or commitment to the future payoffs in a particular relationship by giving expensive gifts. Such an investment would render an early defection from the relationship unprofitable and therefore unlikely as a matter of rational choice. Thus a businessman or suitor who spends a great deal of time and/or money on gifts and entertainments for a prospective partner demonstrates to that individual that he is interested in a long-term business or social relationship, one of a length necessary to recoup the initial expenses and still obtain a surplus of some sort. The prospective partner, then, can trust the businessman or suitor to the extent that person has made "cheating" a money, or other value-losing proposition, through much of the life of the relationship, and a separating equilibrium has occurred.

Almost half of Posner's book involves the legal applications of his signaling theory of social norms in the area of gift-giving, family law, stigma in the criminal law, voting, political symbols, racism and nationalism, and contract law. Two or three examples will illustrate Posner's main points. For instance, Posner applies his theory to voting behavior, noting that "rational choice theorists have not produced a satisfactory explanation of why people vote," often simply positing an arbitrary "taste" for voting (p. 122). Posner reframes the question as follows: "Why does a social norm require people to vote?" (p. 122). Posner argues that voting should be analyzed as symbolic or "signaling behavior," the signal being the "costly" act of voting itself, because an individual's vote seldom or never determines the outcome of an election and thus provides no plausible non-signaling benefit. Why would a "rational" person motivated by a rational cost-benefit analysis take the trouble to engage in the symbolic behavior of voting?

Page 285 begins here

Although the act of voting is "costly," it has potential benefits. The "good type" uses the act of voting to signal his low discount rate to those who associate voting with cooperativeness. Therefore the "good type" expects to "recover his costs through repeated rounds of cooperation with another person" made possible in part precisely by signaling behavior such as voting (p. 123). It follows, then, that voting is rational in precisely the same way that gift-giving to a prospective business partner is rational: The expected benefits of future cooperative endeavors outweigh the investment in signaling costs that make such cooperation possible. This argument also illustrates one of Posner's most important points. Adherence to costly social norms may often appear "irrational" on its face from a traditional rational choice perspective, but once one takes into account the necessity of signaling behavior designed to promote future cooperative endeavors, such behavior is in fact quite rational.

Posner also applies his game theory model of social cooperation to the use of "shaming" or "stigmatizing" penalties in the criminal law, which use social stigma to punish and deter crime. Posner notes that social shaming is essentially a "nonlegal" sanction, imposed on some members of the community by other members. In fact, there are two reasons why individuals may avoid dealing with "shamed" parties. First, the shamed individual is revealed as a likely bad cooperative partner. Second, other individuals themselves wish to signal their "low discount rate" by avoiding shamed persons. Thus the often- costly behavior of shaming and shunning are in part examples of signaling behavior. Moreover, this second motivation for stigmatizing, the desire to signal one's low discount rate by shaming someone of revealed "bad type," is potentially dysfunctional. Indeed, such shaming behavior can turn into a "cascade" perhaps resulting in some instances in complete ostracism or even lynching. Posner compares this potential "collective disaster produced by individually rational [signaling] behavior" to a bank run (p. 92). As Posner notes, this example reveals the potential pathologies of shaming, including potentially extreme over-punishment, because the "chief motive" of shaming is "to enhance reputation [of those engaged in shaming behavior], not to do justice" to the shamed individual (p. 93). Governments which impose shaming penalties do so in order to exploit the power of social norms, to "facilitate ostracism by the community," as a form of punishment (p. 94). The problem with this policy is "that the government cannot control the level of ostracism that it provokes," nor its effect on the shamed individual and the individual's family (pp. 95-96). For similar reasons, shaming punishments also promote the formation of deviant or criminal sub-communities by diminishing the shamed individual's ability to form cooperative partnerships outside of such sub-communities. Posner concludes that these concerns about over-punishment and creation of deviant sub-communities must be carefully weighed against both the desire to punish by promotion social stigma and by the community's need for information about past criminal activity. The potential pathologies of shaming punishments also reveal the dysfunctional aspect of social norms and the possible unintended negative consequences of social policies meant to exploit such norms.

Finally, Posner tackles some of the normative implications of his analysis, arguing in particular that (1) social norms (as we have seen) are often dysfunctional; (2) that individuals routinely in engage in putatively "principled" behavior for strategic

Page 286 begins here

reasons; and (3) law and markets do not interfere with communities and that there has been no real decline in community in the U. S. (p. 7). The second of these claims is of particular interest, given the larger questions it raises about Posner's understanding of the basis of social norms. It is Posner's contention that while individuals routinely claim to make decisions on the basis of "principle," as opposed to, say, expediency, they do not in fact do so, and that they make statements of principle strategically. In Posner's view, we do not in fact place an "absolute" value on things such as family life, the environment, or religious faith. Nor do we value these things in some "special" or "different" way that cannot be easily translated into the common metric of money. In short, virtually everyone and everything has its price, despite conventional pieties to the contrary. Moreover, one of Posner's main points is that an unwillingness to openly challenge these conventional pieties is rational from an individual perspective because endorsement of conventions such as these is a form of signaling behavior. Posner thus attributes what he views as the common-place persistence in misdescribing the basis of decision-making to principle-rather than, say, money-to an instrumental or strategic imperative: The need to "signal" an adherence to principle in order to facilitate future cooperative endeavors. Everyone has a price, but (almost) everyone denies it, and, everything else being equal, the individuals who care enough about future cooperative relationships to make the effort of denial ("I won't cheat on you the moment it becomes sufficiently profitable!") are the safer bet for such relationships.

Many readers, of course, will object to Posner's strong methodological assumption that individuals make decisions largely or purely in conformity to their "rational" self-interest rather than from motivations of moral or political principle. Take the question of "honesty." Honesty is not only routinely a good policy, but it is also, of course, a moral principle. The thrust of Posner's methodological commitment is that persons who act honestly do so simply out of an instrumental concern for maximizing their participation in future cooperative endeavors. Therefore, such a person, a "good type," faced with the (not so rare) hard choice between acting honestly but appearing dishonest or acting dishonestly but appearing honest, would choose the latter, favoring appearance over reality when the two happen to conflict. In fact, Posner's analysis suggests that those persons who maintain that they are "principled" rather than "instrumental" in this regard and would thus choose the substance of honesty over its mere appearance are by that very statement actually engaged in the reverse, the favoring of the appearance of honesty over its substance: They are strategically "lying" about their basis of decision-making in order to signal "honesty" to potential cooperative partners and thus to maximize their opportunities for future cooperative endeavors. One may be forgiven for finding that conclusion more than a little dubious-as well as false to the common self- understanding of most individuals.

What, then, of motivations rooted in moral principle, moral virtue, or "natural" emotions of love, kinship, and solidarity? Are these really largely irrelevant to an inquiry into social norms? Are social norms properly understood as "nothing but" simple behavioral regularities driven by some narrow cost-benefit form of rational self-interest? Consider the case of voting. Absent from Posner's discussion of the (ir)rationality of voting, is the idea that voting might be a rational act not (merely)

Page 287 begins here

because it is viewed as a sign or signal of "cooperativeness" or "low discount rate" in many social circles, but because it is the exercise of a civic or moral virtue. One might well imagine that substantial voter turnout tends to promote legitimacy and political stability in democratic regimes and therefore a civic obligation to vote is grounded in the duty to provide the government with the pre-conditions for its flourishing. Of course, one particular vote is unlikely to make a difference in either a particular election result or to the overall health of the regime, but moral principles of universalism (the first formulation of Kant's categorical imperative) and reciprocity (the Biblical "golden rule") suggest that one has a civic obligation to vote rather than "free ride" because IF everyone were to abstain from voting, given its costs, ultimately the regime would be severely damaged. Voting then might be thought quite rational, and the common habit of voting explained by a combination of the operation of practical reasoning about civic obligation and the internalization of norms practiced habitually and often held to by ordinary persons in inarticulate, unreflective, or under-theorized forms. This explanation fits more with common sense and with the lived experience of individuals in democratic communities than does either Posner's signaling theory or the common "rational choice" expedient of simply positing a non-rational "taste" for voting

Notably, Posner does NOT claim, as an empirical matter, that, "rational choice theory can offer a complete explanation of social norms or of cooperation" (p. 46). He acknowledges, for instance, that "[c]ognition and emotion are not irrelevant" to an inquiry into social norms, but he also contends that "[t]hey are just not well enough understood to support a theory of social norms, and repeated but puzzled acknowledgments of their importance would muddy the exposition of the argument without providing an offsetting benefits" (p. 46). Perhaps, but perhaps not. Posner here expresses a controversial preference for a methodologically "fruitful," if sometimes misleading, theoretical "explanation" of social phenomena over a methodologically "sterile" but more accurate (or at least more candid) admission that there is much about human behavior we simply cannot explain in theoretically satisfying terms, at least if that is thought to require formal modeling and empirical verification through quantitative analysis. It is non-obvious why we should prefer the first of these approaches to the exclusion of the second, and Posner largely ignores the larger implications of this foundational methodological question. In fact, one obvious "offsetting benefit" of a recognition of the humane, qualitative, and moral aspects of social norm formation, even as it "muddies" a game-theory exposition, is precisely a dilution of the potentially corrosive effects of rational choice theories on social relations and moral idealism. Posner's "cost-benefit" analyses of the motivations for honesty and of the rationality of voting are cases in point.

In any event, LAW AND SOCIAL NORMS is very well written, analytically rigorous, and highly sophisticated in its use of signaling models to illuminate a wide range of issues. It is an important book, and one that provides a powerful and highly innovative treatment of an under-explored aspect of legal theory. It will be of great interest to anyone working in the areas of law and economics, rational choice, game theory, social norms, the social-norm context of legal issues, or law and public policy more generally.


Copyright 2001 by the author, Jack Wade Nowlin.