Competing norms and social evolution: is the fittest norm efficient?

• Jurisdiction: United States
• Date: 01 June 2001
• Author: Mahoney, Paul G.

An influential theme in recent legal scholarship is that law is not as important as it appears. Social control, many scholars have noted, is often achieved through social norms--informal, decentralized systems of consensus and cooperation--rather than through law. This literature also displays a guarded optimism that social evolutionary processes will tend to favor the adoption of efficient norms. Using concepts from evolutionary game theory, we demonstrate that efficient norms will prevail only in certain settings and not in others: survival of the fittest does not imply survival of the efficient. In particular, we show that in many games of interest to legal scholars--games describing fundamental interactions in property, tort, and contract--evolutionary forces lead away from efficiency. We also describe how law rights this trend.

INTRODUCTION

Economists are paying increasing attention to the role of institutions--notably legal institutions--in the operation of economic systems. Both economic historians and financial economists contend that legal rules protecting investment are essential for markets to flourish.(1)

At the same time, an influential theme in legal and political science scholarship is that law is not as important as it appears. Social control, many scholars note, is often achieved through social norms--informal, decentralized systems of consensus and cooperation--rather than through law.(2) Many discussions of norms draw on a separate literature on cooperation in game theoretic settings, particularly in repeated games. Using formal or informal models of norms as strategies in repeated games, scholars, such as Robert Ellickson, express a tentative optimism that socially desirable, or efficient, practices can arise from the interactions of decentralized, utility-maximizing individuals.(3) Moreover, Randal Picker and Robert Axelrod argue not just that efficient norms are the possible outcomes of decentralized interactions, but that such norms have, by virtue of their efficiency, a natural advantage in the social evolutionary processes that determine norm adoption.(4)

These seemingly contradictory positions may be reconciled so long as each is kept within its proper domain. Using concepts from evolutionary game theory, we demonstrate that efficient strategies will prevail in evolutionary settings for some games, but not for others. In particular, we show that many games of interest to legal scholars--games that are analogous to fundamental interactions in property, tort, and contract--are precisely those in which evolutionary forces lead away from efficiency, and institutions are required to right the trend.

At the heart of the distinction between evolutionary "fitness" and economic efficiency lies the concept of mismatch risk. Whether a norm is efficient depends entirely on how well it does when matched against itself. Norm A is Pareto superior to norm B if everyone has higher payoffs when norm A is generally adopted. But whether a norm is selected by evolutionary processes depends also on how well it does when mismatched with other norms. On the evolutionary battlefield, the successful norm, like the successful weapon, must also fare comparatively well against those not similarly armed. A norm that is Pareto optimal but does relatively poorly against disparate norms is said to have mismatch risk. When a norm's mismatch risk is larger than its efficiency advantage, evolution will tend to select against it.

We argue that mismatch risk is a plausible metaphor for many important real-world problems that arise because efficient practices require investment and inefficient practices may involve appropriating that investment or destroying much of its value. Thus, while the investment-oriented norm may be efficient, it is likely to be the relative loser when the other side reneges or appropriates. The most basic legal rules--those providing protection for property rights, enforcing contracts, and requiring reasonable care in the conduct of risky activities--can be seen as responses to just such situations.

When mismatch risk is great in relation to the potential gains from playing efficient strategies, completely decentralized repeat play is unlikely to produce efficient norms. To assure efficiency, we must introduce an external party that can alter payoffs. We note that in well-functioning societies, law acts as this third party. It redistributes part of the defecting party's payoff to the cooperating party, thus mitigating mismatch risk and improving the evolutionary fitness of efficient strategies. Therefore, while norms scholars are correct to point out that law is not the only, or even the primary, means of social control, economists are also right to point out that law and efficiency are usually found together.

The rest of this paper is organized as follows. In Part I, we summarize and connect the relevant strands of the legal literature on norms, emphasizing the optimistic arguments about norm evolution. Part II presents a nontechnical analysis of the parallel literature on evolutionary game theory. Here, our emphasis is on the surprisingly general agreement within that literature that efficiency and fitness are distinct concepts. We also discuss that relationship between fitness and efficiency is affected when agents can choose with whom they interact. In Part III, we provide several elementary legal examples in which mismatch risk plays an important role. Part IV explores the consequent necessity of centralized institutions in guiding evolutionary processes toward efficient outcomes.

1. NORMS: COOPERATION WITHOUT LAW

   Norms, as the term has come to be used in legal scholarship, are rules of conduct that constrain self-interested behavior and that are adopted and enforced in an informal, decentralized setting.(5) They arise in the course of individuals' repeated interactions, which sometimes produce a consensus about desirable and undesirable conduct.(6) Enforcement of the resulting rules takes place outside the legal system through mechanisms such as gossip or ostracism. If a norm becomes sufficiently strong, individuals may internalize, or self-enforce, the norm by feeling guilt when they contemplate its violation.(7) The norms literature has produced a detailed taxonomy of sanctions, including "first-party" sanctions such as feelings of guilt or shame, "second-party" sanctions such as retaliation by a victim, and "third-party" sanctions such as social disapproval.(8)

   A central issue in the norms literature is whether norms will generally be efficient. We would characterize the dominant theme among scholars who have studied the issue as guarded optimism,(9) although a few pessimistic notes have been sounded as well.(10) We briefly survey the case for the efficiency of norms.

   1. Norms as Repeated Game Equilibria

      The leading theoretical approach to the efficiency question (indeed, to the entire law and economics literature on norms) is the theory of repeated games. Norms are modeled as equilibrium strategy choices in a particular repeated game.(11) Depending on the situation, other authors would narrow the definition by limiting it to cooperative strategies(12) or to collectively enforced strategies.(13) These refinements are unnecessary for our purpose, which is to describe the relative evolutionary fitness of cooperative and uncooperative strategies. Thus, for the sake of simplicity, we use a broad definition.

      To provide some real-world motivation, we can imagine a pair of neighbors. An attractive tree straddles their property line. The tree requires some care--pruning, mulching, and so on. That care generates a positive externality. Part of the benefits of each individual's effort accrues to her neighbor; and while the joint benefits from individual effort exceed private costs, private benefits do not. To gain tractability at the expense of some verisimilitude, we limit each neighbor to two possible actions chosen in isolation from the other--pay $50 for care or pay nothing. When only one neighbor pays $50 for tree care, each neighbor individually enjoys a $40 benefit, while if both pay $50, each enjoys an $80 benefit. The resulting payoffs are shown in Table 1. Throughout, we will show only the row player's payoffs in the matrix representations. All of the games we will use are symmetrical in the sense that the row and column players are interchangeable. We indicate pure strategy Nash equilibria by shading.

      Table 1: Tree Maintenance as a Prisoner's Dilemma Pay Don't Pay Pay 30 -10 Don't Pay 40 0 The game is a prisoner's dilemma and the dominant strategy for each neighbor is to pay nothing. The outcome is Pareto inefficient, as each neighbor could be made better off if each paid $50. Although the situation is bleak if the players meet each other only in a one-shot interaction, the well-known folk theorem of repeated games suggests that cooperation in caring for the tree is a possible outcome if the neighbors play this game repeatedly (more generally, are engaged in an ongoing relationship). Specifically, if each individual can condition current actions on her neighbor's past behavior, and cares not just about current payoffs but, to a sufficient degree, about future outcomes, then cooperation can be maintained as a Nash equilibrium of the infinitely (or indefinitely) repeated game.(14)

      One salient and compelling mode of cooperation is the "tit-for-tat" strategy. According to this strategy, each neighbor pays to maintain the tree in the first year and from thereon imitates her neighbor's prior-year behavior. Provided the neighbors care enough about future payoffs, tit-for-tat is an equilibrium of the repeated game. It is also clear that if both neighbors play this strategy, they will enjoy the higher payoffs of perpetual cooperation.

      We should note here that we are side-stepping yet another definitional issue that has plagued the norms literature.(15) Some scholars use the term...