Rewards facilitate human cooperation under natural selection

Evolution of cooperation -- or how to suppress free riders

University of Vienna has an admirable tradition of investigating conditions promoting the evolution of cooperation. For last decades, research advances at University of Vienna have shed new light on various aspects for sustaining high levels of cooperation. As is well known, it has been an interdisciplinary issue to understand how cooperation can thrive under Darwinian natural selection. Usually, voluntary contribution is costly, and thus, those who free ride on others' contribution are often advantageous over those who contribute.

Social optimum or otherwise collective failure

In the paper, Tatsuya Sasaki and the coauthor Uchida Satoshi (RINRI Institute, Tokyo) investigate how voluntary reward funds can help coordinate cooperators in threshold public good games. In the game a certain amount of public goods is provided if the number of cooperators exceeds a critical point, or otherwise, contribution to the public-good provision ends up extinction. But when and how can high levels of cooperation evolve? The new research, in terms of evolutionary game theory, examines a bypassing policy to overcome the bad end through rewards for cooperative behaviors, such as subsidies and bonus programs. As these rewards are also costly, thus it is very important to investigate under which situations voluntary rewarding can emerge and help achieve a group goal. However, how can we prevent the reward system from being eroded by those who take a free ride on the voluntary fund raising ("second-order free riders")?

Through rise and fall of reward funds

"Finally, the voluntary reward fund falls," says Tatsuya Sasaki. "But ironically, the rewards, which are vanishing through invasion of the second-order free riders aforementioned, complete the evolution towards a stable society of certain cooperation." "Imagine a law for official subsidy which often includes its own expiration conditions. In our model, with achievement of a sufficient fraction of cooperators, the reward system is evolutionarily abolished." "Moreover, even under the strictest condition that the public-good provision requires unanimous agreement among all members, the reward fund considered can establish 100% cooperation."

Rewards and punishments are ones of the most commonly used tools to shape human behaviors. Previous studies in evolutionary game theory have mainly focused on punishment. Punishing is more costly for its initiator than rewarding. Thus the evolution of cooperation by punishment often needs support from other mechanisms. By contrast, the novel theoretical result is obtained by numerically examining a very simple model. Tatsuya Sasaki says: "It is crucial to understand how we can achieve cooperation without heavily relying on punishment, which can potentially lead to decreasing openness and escalating conflicts among groups. Therefore, investigating effects of rewards on transforming social norms would be fruitful for future research."