Report

The Competitive Advantage of Sanctioning Institutions

See allHide authors and affiliations

Science  07 Apr 2006:
Vol. 312, Issue 5770, pp. 108-111
DOI: 10.1126/science.1123633

Abstract

Understanding the fundamental patterns and determinants of human cooperation and the maintenance of social order in human societies is a challenge across disciplines. The existing empirical evidence for the higher levels of cooperation when altruistic punishment is present versus when it is absent systematically ignores the institutional competition inherent in human societies. Whether punishment would be deliberately adopted and would similarly enhance cooperation when directly competing with nonpunishment institutions is highly controversial in light of recent findings on the detrimental effects of punishment. We show experimentally that a sanctioning institution is the undisputed winner in a competition with a sanction-free institution. Despite initial aversion, the entire population migrates successively to the sanctioning institution and strongly cooperates, whereas the sanction-free society becomes fully depopulated. The findings demonstrate the competitive advantage of sanctioning institutions and exemplify the emergence and manifestation of social order driven by institutional selection.

The uniqueness of human cooperation necessitates investigations that reach beyond the explanations of cooperative behavior of nonhuman animals (15). Profound empirical evidence shows that the possibility of sanctioning norm violators stabilizes human cooperation at a high level, whereas cooperation typically collapses in the absence of sanctioning possibilities (611). Would a sanctioning institution deliberately be adopted when individuals can choose between a sanctioning and a sanction-free institution? The considerable payoff losses in the process toward stable cooperation—for both the punishers and the punished individuals—as well as natural resentments against punishment caused, for example, by its detrimental effects (12) might guide individuals' choice toward the sanction-free institution.

The argument that higher cooperation levels in sanctioning institutions “automatically” lead to their prevalence—because rational individuals choose the institution with the higher payoff (13)—is often brought forward as an affirmative argument for the competitive advantage of sanctioning institutions. The force of this argument can be questioned, however, because it displaces rather than solves the evolutionary puzzle of human cooperation. The reason for this is that stable cooperation requires a positive share of individuals who carry personal costs for cooperation and punishment to the benefit of the entire group (1416). These individuals have a clear payoff disadvantage compared to cooperators who free-ride on the punishment acts. Recent research shows that a positive share of strong reciprocators—cooperating individuals who are willing to reward fair behavior and to punish unfair behavior even when they cannot gain materially from doing so—can be evolutionarily stable (17, 18). But what happens if the population is perfectly mobile and is permanently invaded by outsiders from a noncooperative environment who are attracted by high payoffs from cooperation? Is the fraction of strong reciprocators who choose the sanctioning institution sufficiently large to keep up the cooperative culture? These arguments cast serious doubt on the prevalence of sanctioning institutions.

However, several affirmative arguments for the competitive advantage of sanctioning institutions also come to mind, e.g., the large number of institutional frameworks that facilitate the sanctioning of norm violators in human societies (1921) and the recent finding that humans derive satisfaction from punishing defectors (22). Additionally, theories of cultural and institutional selection (2326) that are grounded on the exceptional human ability of social learning support the competitive advantage of sanctioning institutions. They suggest that individuals preferentially migrate to groups with higher payoffs and imitate the decisions prevalent in these groups. Hence, group members punish, because it is common to do so. When cooperation is sufficiently widespread, the payoff-disadvantage from punishing is relatively small, and only a weak tendency for conformist behavior suffices to stabilize the punishment of noncooperators.

We inquire into the competitive advantage of sanctioning institutions in a laboratory experiment in which we implement permanent competition between a sanctioning and a sanction-free institution through endogenous choice. It allows one to study the evolution of the different institutions over time as well as the changes in behavior in the same individual when participating in different social settings.

In our experiment, 84 participants anonymously interact in a social dilemma situation in 30 repetitions. Each repetition consists of three stages: An institution choice stage (S0), a voluntary contribution stage (S1), and a sanctioning stage(S2). In stage S0, the participants simultaneously and independently choose between a sanctioning institution (SI) and a sanction-free institution (SFI) in which neither positive sanctioning (rewards) nor negative sanctioning (punishment) is possible. In stage S1, each participant interacts in a public goods game with all other participants who have chosen the same institution in S0; each player is endowed with 20 money units (MUs) and may contribute between 0 and 20 MUs to a public good. Each group member equally profits from the public good, independent of his or her own contribution. The MUs not contributed to the public good are transferred to the participant's private account. The diametrically opposed individual and collective interests constitute the social dilemma in public good provision: It is always in the material self-interest of any subject to free-ride on the contributions of others and to keep all MUs for the private account, whereas the collective interest demands full contribution of all group members. After the players have simultaneously made their contribution decisions, they are informed about the contributions of each member in their institution. In stage S2 each player in SI may positively or negatively sanction other members of SI by assigning between 0 and 20 tokens to other members. Each token used as a negative sanction costs the punished member 3 MUs and the punishing member 1 MU. Each token used as a positive sanction yields the receiving member 1 MU and costs the member who uses it 1 MU. At the end of the period each participant receives detailed (but anonymous) information about each of the other participants from both institutions (27).

The initial choice of institution provides a clear picture: Only about one-third of the participants (mean = 36.9%; SE = 4.0%) prefer SI to SFI in the first period. The revealed institution preference correlates with different types of behavior (28, 29). Participants who initially join SI contribute on average 12.7 MUs (SE = 0.79) in the first period, while on average only 7.3 MUs (SE = 0.54) are contributed in SFI (Wilcoxon signed rank matched pairs test, z = –2.366, P = 0.016, two-tailed). Almost half the subjects (mean = 48.4%; SE = 8.5%) who opt for SI in the first period are “high contributors” in that they contribute at least 15 MUs. Almost three-fourths (mean = 73.3%; SE = 17.0%) of these high contributors exert punishment tokens to discipline low contributors and thus try to enforce and establish a norm of high cooperation. These subjects amount to 13.1% (SE = 4.0%) of the total subject population and can clearly be classified as “strong reciprocators,” i.e., subjects with a predisposition to make high contributions and to punish norm violators. In contrast, 16.1% (SE = 5.2%) of the subjects in SI contribute 5 MUs or less (“free-riders”) in the first period. The situation is completely different in SFI, where in the first period almost half of the subjects are free-riders (mean = 43.4%; SE = 3.4%), whereas high contributors arerare(mean = 11.3%; SE = 4.3%). A subject who chooses SFI in the first period with a contribution of more than 15 MUs and uses negative sanctions immediately after having switched to SI may also be classified as a strong reciprocator. We observed two subjects with this behavior in our subject population (2.4%), so that 15.5% (SE = 5.6%) is a lower bound for the proportion of strong reciprocators in the subject population. Initially, the significantly higher contributions in SI do not result in higher payoffs in SI: Average payoffs in the first period of SI (mean = 38.1 MUs; SE = 2.05) are significantly lower than in SFI (mean = 44.4; SE = 0.32) (Wilcoxon signed rank matched pairs test, z = –2.047, P = 0.047, two-tailed). Due to frequent punishment activities, free-riders earn significantly less in SI (mean = 30.2; SE = 4.51) than in SFI (mean = 49.7 MUs; SE = 0.86) in the first period (Wilcoxon signed rank matched pairs test, z = –2.366, P = 0.016, two-tailed).

Although subjects are initially reluctant to join SI, it becomes predominant over time; eventually, nearly all participants (mean = 92.9%; SE = 3.4%) choose SI and cooperate fully (Fig. 1) (30). Simultaneously, contributions in SFI decrease to zero. In period 10 the contributions in SI are on average 89.9% (SE = 10.3%) of the endowment and from there on they steadily increase. In the last period the difference between the two institutions is almost as extreme as it can be with average contributions of 19.4 MUs (SE = 0.714) in SI and 0 MUs (SE = 0.0) in SFI. Averaged over all periods, subjects in SI contribute 18.3 MUs (91.4% of the endowment; SE = 5.0%), whereas subjects in SFI contribute only 2.9 MUs (14.4% of the endowment; SE = 3.0%) (Wilcoxon signed rank matched pairs test, z = –2.366, P = 0.016, two-tailed).

Fig. 1.

Subjects' choice of institution and their contributions. The average contributions in both institutions over the 30 periods of the interaction are measured as the percentage of endowment contributed to the public good.

What causes this dramatic change of mind? Pure imitation of the successful behavior would lead to an increase of free-riders in SFI because they earn the highest average payoffs in the first period. This is actually observed in period two. Consequently, the payoffs of free-riders in SFI decrease and over the periods, participants in SFI experience the typically observed collapse of cooperation in repeated social dilemma interactions (Fig. 1). A comparison of the payoffs of the two predominant behavioral patterns—free-riding in SFI and high contributions in SI (Fig. 2)—shows that from period five onward a high contributor in SI achieves a higher payoff than a free-rider in SFI (Wilcoxon signed rank matched pairs test, z = –2.366, P = 0.016, two-tailed). It therefore pays for a monetary payoff maximizing participant to switch from free-riding in SFI to contributing in SI. This triggers an amplifying effect; namely, the greater the number of cooperators in SI, the higher their payoffs. Indeed, from period 10 onward, 86.1% (SE = 13.1%) of all members of SI contribute fully (20 MUs) and 86.0% (SE = 8.6%) in SFI contribute almost nothing (2 MUs or less). The finding that players apparently choose institutions according to payoffs indicates that stochastic forces play only a minor role in determining switching behavior (31).

Fig. 2.

Payoffs of the two predominant behavioral patterns, “free-riders” (contributions between 0 and 5 MUs) in the sanction-free institution (SFI) and “high contributors” (contributions between 15 and 20 MUs) in the sanctioning institution (SI). The highest attainable payoff (under full contributions of all subjects and no punishment) is 52 MUs and the payoff from complete free-riding and no punishment is 40 MUs.

A closer look at individual behavior immediately before and after migration from one institution to the other confirms the bipolar pattern of behavior induced by the two institutions. Indeed, 80.3% (SE = 5.0%) of subjects increase their contribution when migrating from SFI to SI in two consecutive periods. Moreover, 27.1% (SE = 5.3%) of subjects even “convert” from being a complete free-rider (contributing 0 MUs) to a full cooperator (contributing 20 MUs) when switching from SFI to SI. The migration behavior in the opposite direction, i.e., from SI to SFI, is similarly extreme. Roughly 70% (mean = 70.9%; SE = 4.9%) of subjects reduce their contribution when switching from SI to SFI and about 20% (mean = 17.0%; SE = 4.7%) switch from full cooperation to free-riding.

Individual payoff maximization cannot explain why new members in SI follow the second norm established by the strong reciprocators who joined SI in early periods, i.e., the norm to punish low contributors. The most successful behavior would be to contribute in SI (and hence avoid being punished), but refrain from the costly punishment of others. Because punishment of defectors constitutes a second-order public good (in which defection cannot be sanctioned in our setting), individual payoff maximization would rule out punishment. However, only a minority of subjects follow this payoff-maximizing behavior. The overwhelming majority of 62.9% (SE = 8.5%) of the subjects immediately conforms to and adopts the prevailing norm of punishment in SI, i.e., they always use punishment immediately after they switch to SI. This results in a quite stable proportion of ∼40% (mean = 42.1%; SE = 5.9%) of subjects who both contribute highly and punish during the last 20 periods (Fig. 3). Figure 3 also shows that the payoff difference between high contributors who punish and those who do not constantly diminishes over time because punishment becomes ever more unnecessary. Additionally, because the absolute number of punishers increases, the individual burden from effectively punishing free-riders becomes smaller over time (32). Toward the end, subjects who both contribute highly and punish exhibit a payoff disadvantage of less than 2%; hence, the “selection pressure” against strong reciprocators becomes quite weak (33). This leads to a continuous increase in efficiency gains in SI up to 95.8% (SE = 4.6%) in the final period, whereas efficiency gains in SFI converge to zero (mean = 0; SE = 0.0).

Fig. 3.

Payoffs and percentages of punishers and nonpunishers among the “high contributors” (contributions between 15 and 20 MUs) in the sanctioning institution (SI). The highest attainable payoff (under full contributions of all subjects and no punishment) is 52 MUs and the payoff from complete free-riding and no punishment is 40 MUs.

Although the use of both positive and negative sanctions per individual decreases over time, the ratio in which they are used is rather stable; on average, 1.66 negative sanction points (SE = 0.60) are allocated per positive sanction point. A Tobit regression of the combined effect of positive and negative sanctions exhibits a clear positive impact of punishment on subsequent contributions, whereas positive sanctions have a slightly negative but rather insignificant effect (Table 1). It seems that positive sanctions are not perceived as an unambiguous encouragement to increase the contribution; perhaps they are taken as an indication that the contribution has been higher than expected by others and hence may be lowered. These observations reflect the asymmetry between negative and positive sanctions. Positive sanctions are addressed to those who already abide by the social norm and, to preserve the approval of cooperation, a continuous application of the instrument is required. Negative sanctioning, by contrast, is an instrument for disapproving of norm-violating behavior and need only be exerted if the norm is not followed. If an individual abides by the norm, punishment is not necessary. The threat of punishment alone is able to support cooperation.

Table 1.

Results of a Tobit regression, independent variable: Contribution (t + 1) – Contribution (t). Tobit regression for subjects who opted for SI in period t and (t + 1) with a robust estimation for the standard errors using the independent observations as clusters. The values in parentheses denote the robust standard errors.

Independent variableCoefficientz value
Negative sanctions in t 0.444 (0.085) 5.24View inline
Positive sanctions in t -0.148 (0.102) -1.45
Constant 0.000 (0.053) 0.00
  • View inline* Denotes significance at the 1% level.

  • Our results show that the sanctioning institution is the undisputed winner in a “voting-with-one's-feet” competition with a sanction-free institution. The results provide profound empirical evidence for the existence and importance of strong reciprocators, as well as a form of conformist behavior, as described in models of cultural selection. The initial establishment of the “norm to cooperate and punish free-riders” is mainly driven by the steadfastness of the strong reciprocators to punish noncooperative subjects, despite severe individual losses (34). Although strong reciprocators are a minority, they manage to establish and enforce a cooperative culture that attracts even previously noncooperative individuals and thus resolves the social dilemma. The predominant tendency to punish norm violators after a migration from the noncooperative environment of the sanctioning-free institution to the sanctioning institution provides support for the assumption that humans adapt to the common behavior although it deviates from the payoff-maximizing behavior. This tendency for conformism raises sanctioning activities at a high level such that cooperation can be stabilized.

    Supporting Online Material

    www.sciencemag.org/cgi/content/full/312/5770/108/DC1

    Materials and Methods

    Figs. S1 and S2

    Table S1

    References

    References and Notes

    View Abstract

    Navigate This Article