Direct reciprocity among humans

Abstract Direct reciprocity is the tendency to repay others' cooperation. This tendency can be crucial to maintain cooperation in evolving populations. Once direct reciprocity evolves, individuals have a long‐run interest to cooperate, even if it is costly in the short run. The major theoretical framework to describe reciprocal behavior is the repeated prisoner's dilemma. Over the past decades, this game has been the major workhorse to predict when reciprocal cooperation ought to evolve, and which strategies individuals are supposed to adopt. Herein, we compare these predictions with the empirical evidence from experiments with human subjects. From a theory‐driven perspective, humans represent an ideal test case, because they give researchers the most flexibility to tailor the experimental design to the assumptions of a model. Overall, we find that theoretical models describe well in which situations people cooperate. However, in the important case of “indefinitely repeated games,” they have difficulties to predict which strategies people use.