Evolution of Fairness: Rereading the Data

Abstract

J. Henrich et al. 's Research Article (“Markets, religion, community size, and the evolution of fairness and punishment,” 19 March, p. [1480][1]) is a valuable addition to the growing literature testing behavioral hypotheses through careful cross-cultural measurement. However, the data they report falsify their theory. The authors propose that increases in third-party punishment of unfairness drove an increase in fairness norms, enabling the emergence of large-scale market economies. Critical to this theory is their hypothesis that exposure to third-party punishment actually elicits an increase in fairness. Relevant to evaluating this hypothesis, the authors conducted two near-parallel economic games: a Dictator Game (in which a “dictator” unilaterally divides a windfall gain with another person—a measure of fairness) and a Third-Party Punishment Game (the Dictator Game with the addition of exposure to possible third-party punishment). Their central hypothesis requires that adding punishment to the Dictator Game will increase fairness, but their data show that the addition of punishment decreases fairness (p. 1483). This finding unambiguously refutes their central hypothesis. Henrich et al. also assert that their data—showing patterned cultural variability in cooperation—can determine whether modern levels of generosity and altruism are driven by an evolved social psychology or by cultural processes. The authors claim that their data decisively favor the cultural processes hypothesis. Yet nothing in their data can test (even in principle) whether it is psychological or cultural processes (or both) that cause these cross-cultural differences. Only long-abandoned instinct-as-reflex theories expect invariant responses in the face of different social inputs. By contrast, modern adaptationist theories predict that our evolved social psychology will be calibrated by relevant environmental inputs. Many of these inputs—such as the local value of long-term cooperative relationships and the fidelity of reputations—are likely to covary with market integration, making it at least as likely that an evolved, context-sensitive social psychology is driving the results that the authors observe. That is, psychological and cultural theories both predict cross-cultural variation. [1]: /lookup/doi/10.1126/science.1182238