Intelligence promotes cooperation in long-term interaction : experimental evidence in infinitely repeated public goods games

A growing body of literature in experimental economics examines how cognitive ability affects cooperation in social dilemma settings. We contribute to the existing literature by studying this relationship in a more complex and strategic environment when the number of partners increases in an infinitely repeated public goods game. We designed four treatments with different continuation probability under two conditions: whether cooperation can be sustained as risk dominance or not. We asked participants to decide whether to cooperate in every period in the first five rounds. They were further asked to decide if they should elicit their strategy at the beginning of each super game using the strategy method in the last five rounds. We found that participants with greater cognitive abilities cooperated more (less) when cooperation could(not) be sustained as risk dominance. A similar trend was observed in the frequency of fully cooperative strategies. We also found that participants with greater cognitive abilities employed lenient and forgiving strategies more frequently when the continuation probability was far higher than the risk dominant threshold level.