Standard models of human cooperation in economics and biology assume purely self-regarding agents who use repeated interactions (reciprocal altruism) or public reputations (indirect reciprocity) to sustain mutual helping behaviors. While these mechanisms are important, there are many equally important forms of prosocial behavior which cannot be accounted for in the same way (Bowles & Gintis Reference Bowles and Gintis2011; Fehr & Gintis Reference Fehr and Gintis2007; Gintis Reference Gintis2005; Reference Gintis2009). These include: voting in elections, participating in collective actions, being kind to strangers, contributing to community public goods, and behaving morally in anonymous situations, or where the material penalties for immoral behavior are low.

Economic experiments strongly suggest that human prosociality is not limited to calculated selfishness (e.g. Batson Reference Batson1991; Fehr & Gächter Reference Fehr and Gächter2000a; Reference Fehr and Gächter2000b; Reference Fehr and Gächter2002; Fehr et al. Reference Fehr, Gächter and Kirchsteiger1997), but that the presence of free-riders is a key and ever-present threat to sustained cooperation. Strong reciprocity, a behavioral mechanism including both altruistic cooperation and costly punishment of free riders (Gintis Reference Gintis2000) thus helps sustain cooperation over long periods. This work showed that humans have strong and consistent other-regarding preferences that could be enlisted in support of social cooperation. In fact, anthropologists have confirmed that strong reciprocity is indeed routinely harnessed in the support of cooperation in small-scale societies (Boehm Reference Boehm1984; Reference Boehm1999; Henrich et al. Reference Henrich, Ensminger, McElreath, Barr, Barrett, Bolyanatz, Cardenas, Gurven, Gwako, Henrich, Lesorogol, Marlowe, Tracer and Ziker2010a; Wiessner Reference Wiessner2005; Reference Wiessner2009), as stressed in Henrich & Chudek's commentary in this issue.

Guala characterizes the punishment side of strong reciprocity as “uncoordinated.” This is simply incorrect. Collective action is a real-life expression of strong reciprocity (Bowles & Gintis Reference Bowles and Gintis2004, p. 17), and the predisposition to punish “transgressors” is often socially organized and sanctioned. Indeed, individuals are often deterred from carrying out self-initiated sanctions (Boyd et al. Reference Boyd, Gintis and Bowles2010). The experimental evidence for coordinated punishment was laid out in several experimental papers on strong reciprocity (e.g., Cinyabuguma et al. Reference Cinyabuguma, Page and Putterman2005).

Guala claims that costly punishment is rarely observed in the real world, and what punishment is observed is generally not very harsh (e.g., verbal harassment, gossip, ostracism). These observations, even if true, in no way conflict with strong reciprocity models of social cooperation. First, if punishment is effective, it will be rarely carried out. Thus, the absence of frequent punishment is an indication that the threat of punishment has a particularly strong effect. For instance, the average taxpayer in the United States is never penalized for tax evasion, yet no one doubts the importance of prosecuting tax evasion. Similarly, most drivers receive only a few traffic citations in the course of their lives, but many drivers adjust their driving to avoid citations. Second, we stress that most humans are very averse to public criticism of even a verbal form of punishment, and we cite studies that show that verbal criticism alone often leads to conformity (Masclet et al. Reference Masclet, Noussair, Tucker and Villeval2003). In addition, the human emotion of shame serves to amplify social criticism, thereby lessening the need for costly punishment (Bowles & Gintis Reference Bowles, Gintis, Lawrence and Steven2005; Gintis Reference Gintis2004). Moreover, Guala seriously understates the importance of diffuse, uncoordinated, costly punishment in promoting norm adherence.

Guala claims that some punishment is “zero cost.” If so, this would add an interesting dimension to the strong reciprocity model, but it does not conflict with this model.

In sum, we agree with Guala that socially structured punishment is important, but we assert that the predisposition to reward goodness and punish evil underlies the effectiveness of socially structured punishment. We also reaffirm the critical importance of diffuse, unstructured cooperation and punishment in fostering social efficiency and a high quality of life.