Communities overcome collective-action problems in part through norms against free-riding, but a paradox lies at the heart of theories about how those norms are enforced: Since punishing free riders is costly, no one has an incentive to pay the costs of punishing them. A second-order free-riding problem can therefore unravel a norm if no one is willing to sanction those who violate it, or to sanction those who fail to sanction those who violate it, and so on (see Douglass D. Heckathorn, “Collective Action and the Second-Order Free-Rider Problem,” Rationality and Society 1 [July 1989]: 78–100). That norms are enforced at all is therefore puzzling. A related puzzle is that there is considerable variation in how effectively norms are enforced, with many norms (such as foot-binding, the practice of mutilating girls' feet in China that persisted from the tenth through the twentieth centuries) strongly enforced despite few apparent practical benefits.
Christine Horne, a sociologist, addresses both puzzles with her work on metanorms—norms that concern sanctioning and that therefore play a central role in how other norms are enforced. Her thesis is that metanorms are stronger among members of a group, producing higher levels of enforcement for other norms, the more those members depend upon one another in other ways. People enforced norms favoring foot-binding, for example by shunning parents who did not bind daughters' feet, not because the practice itself produced benefits for them but, rather, because people wanted to make themselves appear to be stalwart members of a community at a time when maintaining strong community ties yielded large individual benefits.
An important implication of the theory is that relationships are more important than individual perceptions in determining how norms are enforced. People will enforce a norm even if they do not see a benefit to the behavior being reproduced—or even, as with foot-binding, if they see it as harmful—as long as they feel a need to look good themselves.
Horne's method is experimental. She recruited undergraduate students to participate in a series of games in which the students, working in isolation at a computer, played with and against one another. The games were highly abstracted but clear; in a thorough appendix, Horne details the precise instructions she gave subjects, how their choices were mapped out for them, and so on. Test subjects played for cash, with their scores determining the ultimate payoff. Three kinds of games, each with several variations, tested propositions implied by the theory. A common theme running through most of the games is that test subjects had an opportunity to work together to generate joint value, sacrificing immediate individual gains in order to contribute to a collective good, but they also had chances to shirk or to steal from one another. (In some instances, the test subjects unknowingly interacted with a computer-simulated player that stole.) Subjects could punish shirkers or thieves at a cost, and could also voluntarily share some of their winnings with other players. Each of the games was repeated many times, giving test subjects a shadow of the future.
In general, Horne shows that when mutual dependence was greater, represented in the experiments as larger gains from joint contributions to a collective good, the test subjects were more likely to reward (by voluntarily sharing some of their points) those players who took costly steps to punish cheaters. In variations on the games, she found that large benefits to a collective good led subjects to sanction perceived violations even under conditions that would seemingly make sanctioning unlikely: when sanctioners themselves did not receive a direct benefit from the collective good, when sanctioning someone was costlier, and when the violations being sanctioned were inconsequential. With high interdependence, metanorms were stronger since subjects had a longer-term interest in looking good by showing that they cared about others; the particulars of the norm violations themselves carried less explanatory weight.
The experiments that make up the empirical centerpiece of the book were all previously published individually in journals, but the book adds substantial value by collecting them in one place where the argument can be developed in detail. Horne is, in particular, to be commended for an extremely well-written book that makes sophisticated theoretical arguments, buttressed with experimental methods, accessible to nonspecialists. The theoretical framework is illustrated throughout with anecdotes from a diverse range of case histories and examples from everyday life. The book should therefore be in easy reach of graduate students outside of the specialty, as well as advanced undergraduates. It is also an excellent resource for scholars working on problems of collective action from other disciplinary perspectives.
Aside from occasional brief suggestions in anecdotes and references to other literatures, Horne does not systematically apply the theory to cases outside of the experimental setting (something future work in this research program may, hopefully, address). Still, the direct applications to mainstream political science research are apparent, especially for those empirical phenomena on which social network analysis has traditionally been focused, such as social movements, insurgent and rebel groups, terrorists, nongovernmental organizations, and networks of policy experts.
Arguments drawn on explicitly sociological theories have not often been used in models of congressional or parliamentary organization, and despite growing acceptance they remain controversial among at least some scholars of international cooperation. Horne's relational theory of enforcement, however, may find acceptance among even political scientists working in a rationalist framework, as she (informally) derives a parsimonious theory that does not require the identities of the actors to change as a result of their interactions. As she concludes one of the theoretical discussions, “People might enforce norms that they would rather not enforce because of their connections to others. This means that even rational people can enforce norms in ways that are damaging. Norms and rationality are not necessarily antithetical” (p. 64).
In short, this is an excellent contribution to our understanding of norm-based cooperation, and a showcase for the way in which experimental work can contribute to the demonstration of a theoretical proposition. Scholars studying situations in which political actors struggle to find ways to overcome collective-action problems, in any context, would be well advised to read this.
