1.1. Is anti-free-rider punitive sentiment similar cross culturally?

Punitive sentiments and behavior towards free riders have been observed in both western and nonwestern societies, and in both small-scale and industrial societies, including regions of Japan (Yamagishi, 1986, Yamagishi, 1988), Europe (Decker et al., 2003, Falk et al., 2001, Fehr & Gächter, 2000), the United States (Masclet et al., 2003, Sefton et al., 2002), Latin America (Erasmus, 1977, Price, 2003, Price, in press), and Africa (Erasmus, 1977, Ostrom, 1990, Ostrom, 2000). Ostrom (2000, p. 138) summarizes the cross-cultural evidence:

Extensive fieldwork has by now established that individuals in all walks of life and all parts of the world voluntarily organize themselves [for CA]…Field research also confirms that the temptation to free ride on the provision of collective benefits is a universal problem. In all known self-organized resource governance regimes that have survived for multiple generations, participants invest resources in monitoring and sanctioning the actions of each other so as to reduce the probability of free riding.

The actual biological adaptation, that is, the functional aspect of the phenotype, that produces most punishment in CAs appears to be an emotional system causing anti-free-rider punitive sentiment (Price, Cosmides, & Tooby, 2002). When experimental participants are given opportunities to punish free riders, this sentiment motivates punishment behavior (Fehr & Fischbacher, 2004, Fehr & Gächter, 2002, Masclet et al., 2003, Shinada et al., in press). Even when actual punishment of free riders does not occur, punitive sentiment may still be experienced and may lead contributors to defect:

Without exception, members of ex-collectives I talked with in both Latin America and Africa described the free rider phenomenon as the main cause of failure…[They made statements such as] ‘Some worked more than others, but all were paid the same’ [Venezuela]; ‘Collectives will always fail because some people do not work hard and others then become angry’ [Ghana]. (Erasmus, 1977, p. 309).

While there do appear to be broad cross-cultural similarities in patterns of punitive sentiment in CAs, a lack of precise data about this sentiment from small-scale societies makes it harder to assess the extent to which it may be produced by a functionally specialized, species-typical psychological mechanism. The following study presents such data in hopes of illuminating this issue.

1.2. Punitive sentiment among the Shuar

Study participants were from a Shuar village (population 300) in the Ecuadorian province of Morona Santiago. These villagers are, in some ways, typical Amazonian hunter–horticulturalists, their most important crops being plantains and sweet manioc. Most residents are closely related descendents of the two brothers who founded the village, and the average coefficient of relatedness in the village was .045 (for reference, second cousins are related by a coefficient of .031). Like many Andean/Amazonian groups, they regularly practice traditional CAs called mingas, in which participants work towards some collective goal, for example, clearing a field or building a house. This study focused on an association of sugarcane cultivators (cañicultores), who, once or twice weekly, would hold a minga in which they used machetes to clear their fields of weeds. The expected benefits of membership were an equal share of the profits from sugarcane sales, while the costs were participating in mingas or being fined US$2 per absence (a significant amount, equivalent to what a Shuar could make in one day working as a farmhand for a non-Shuar).

Is the anti-free-rider punitive sentiment experienced by minga participants similar to that experienced by CA participants in industrialized societies? In industrialized societies, punitiveness is mainly experienced by high contributors and targeted at low contributors (Falk et al., 2001, Fehr & Gächter, 2002, Gintis et al., 2003, Masclet et al., 2003), and there is a positive correlation between the punisher's contribution and the extent of his or her punitiveness (Decker et al., 2003, Price et al., 2002, Shinada et al., in press). Not all low contributors are perceived as free riders, however, only those regarded as beneficiaries of the CA who could have contributed highly but who chose not to (Price et al., 2002). That is, the intentionality of the low contribution matters, and only intentional low contributions are considered punishment worthy (Dufwenberg & Kirchsteiger, 1999, Rabin, 1993; review in Gintis et al., 2003). If a goal of punishment were to evoke contributions from would-be free riders (Boyd & Richerson, 1992, Gintis, 2000), then such consideration of intentionality would make sense because punishment effort would be wasted on those unable to contribute.

If punitive sentiment in mingas is similar to that experienced in industrialized CAs, then the following hypotheses should be supported: (1) higher contributors in mingas experience more punitive sentiment towards perceived free riders, and (2) only intentional low contributions are regarded as punishment-worthy free riding.

2.1. Study participants

The cañicultores association consisted of 13 members, and data were collected from all members. Membership was voluntary and limited to official “citizens” (socios) of the community. Because people are usually not elected socios until their twenties, all members were adults, 29.5 to 54.5 years old (x¯=38). Machete work in mingas is traditionally considered men's work, and most members (12 of 13) were male. Although all members were native Shuar speakers, all were fluent in Spanish.

2.2. Variables

The dependent measures included each member's degree of punitive sentiment towards free riders (PUNISH), number of days worked in mingas (DAYS PRESENT), and unexcused absences from mingas (UNEXCUSED ABSENCES). All variables passed the Shapiro–Wilk test for distribution normality.

PUNISH was assayed as each member's response to the question: “I want to know your opinion about how much the fine should be if a member is inexcusably absent from a minga. And, let's suppose that the opinions of the other cañicultores don't matter; this is your opinion, nothing more” (translated from Spanish). The value of PUNISH was how much each member said the fine should be lowered or raised from its actual US$2.00 amount (range=−US$1.20 to 2.00, x¯=US$.51, S.D.=US$1.12).

DAYS PRESENT and UNEXCUSED ABSENCES were based on minga attendance records kept by the association secretary. At the outset of the study, the association had completed 22 mingas over 5 months. For each minga, members were recorded “present” (presente), “inexcusably absent” (falta), or “excusably absent” (justificado). Justificados are due to sickness or injury and, unlike faltas, are not punished. DAYS PRESENT was the total number of presentes recorded for each member (range=10 to 19, x¯=15.46, S.D.=2.88), and UNEXCUSED ABSENCES was the total number of faltas (range=2 to 7, x¯=4.31, S.D.=1.60). The reason for calculating unexcused absences was as follows. The above hypothesis predicts a positive correlation between DAYS PRESENT and PUNISH, but this correlation might be due not to people high in DAYS PRESENT being more punitive, but rather to people high in UNEXCUSED ABSENCES being less punitive: DAYS PRESENT and UNEXCUSED ABSENCES should be negatively correlated, and it would not be surprising if free riders were relatively unenthusiastic about harsh punishment of free riding. Thus, to measure the direct effect of DAYS PRESENT on PUNISH, it is necessary to control for the relationship between UNEXCUSED ABSENCES and PUNISH.

4.1. Why might punitive sentiment be similar cross culturally?

The above results suggest that anti-free-rider punitive sentiment in one small-scale society is similar to that in industrialized societies. These results also seem broadly consistent with the way that punitive sentiment is described by the cross-cultural researchers quoted above. However, more data are needed, especially because the present study's sample size was small due to limited membership in the cañicultores association. Future cross-cultural examinations would provide welcome additional evidence about this sentiment's universality.

As discussed above, some researchers believe that administering punishment would have been individually fitness damaging in ancestral CAs, and therefore, that individual-level adaptations specialized for punitive sentiments and behaviors are unlikely (e.g., Boyd & Richerson, 1985, Gintis et al., 2003). This belief has led some of these researchers to propose biological group selectionist explanations for punishment (Gintis, 2000, Sober & Wilson, 1998), but theoretical difficulties associated with such selection have led many to agree that some gene–culture coevolutionary scenario is more plausible (e.g., Boyd et al., 2003, Fehr & Fischbacher, 2003). However, if the nature and design of punitive sentiment is similar cross culturally, then invoking cultural evolution as part of the explanation for its existence may be superfluous. Before invoking culture, we should thoroughly consider whether adaptations for altruistic punishment could have been individually fitness enhancing in ancestral environments.

There are several ways in which altruistic punishment could have benefited punisher fitness, five of which are discussed in Price (2003). Punishment, for example, might be regarded by observers as a costly signal of some underlying desirable quality (Gintis, Smith, & Bowles, 2001), such as commitment to group interests (Fessler & Haley, 2003), which could result in the punishing individual being a preferred partner in social interactions. Alternatively, punishment ability itself might be a desirable quality in a CA partner (at least to those who do not intend to free ride) because CAs involving effective punishers should tend to involve less free riding and to thus be relatively productive (Price, 2003). Finally, and most speculatively, genes for punishment may have evolved by a “greenbeard” process (Dawkins, 1976, Haig, 1996, Hamilton, 1964, Queller et al., 2003; see Price, in press, for a discussion of CA from a greenbeard perspective), whereby engagement in the punishment of free riders was a reliable label of a gene for punitiveness. Such a gene may have produced resources preferentially for itself and thus promoted its own replication, by instructing CA participants to engage in reciprocal altruistic punishment according to the decision rule “punish free riders to the extent that you expect your average coparticipant to punish free riders.”

4.2. Conclusion

If sentiments and behavior in CAs are products of a human nature specialized for cooperation, then they should be fundamentally similar cross culturally. The data presented above suggest that one kind of cooperativeness, punitive sentiment towards free riders, may meet this criterion. Future research could further investigate the degree to which this and other kinds of cooperativeness in CAs are similar cross culturally. If similarities are convincingly demonstrated, and seem most parsimoniously explained as functionally specialized aspects of human nature, then researchers will have the challenge of explaining how human nature came to be this way.