Rewards and cooperation in social dilemma games

Rewards are identified as a mechanism to sustain cooperation in standard Public Good games, but have been found less effective in Common Pool Resource games. Both paradigms are important for environmental and resource economics as they capture the essence of real-world environmental and resource problems - the provision of pure public goods, and overextraction of common pool resources. This paper aims to understand why rewards are effective in one paradigm and not in the other. We hypothesize that this is because of an important difference between the two; the marginal per capita return is uncertain in the Common Pool Resource game because subjects can undo cooperative actions of others. This is just one of many differences between the two paradigms and hence we test our Hypothesis by introducing the option to reward in the Claim game, a game identical to the standard public good game except that it allows for both giving and taking. This feature causes the marginal per capita return to be uncertain. We find that while rewards are effective in sustaining cooperation in the Public Good game, they are much less effective in the Claim game. We identify the underlying mechanism causing this differential impact.