Confusion and Learning in the Public Goods Game

We test if confusion and learning could potentially explain all the decay of contributions in the repeated public goods games by implementing a limited information environment to mimic the state of confusion. A comparison shows that the rate of decline is more than twice as high in a standard public goods game. Furthermore, we find that simple learning cannot generate the contribution dynamics, which are commonly attributed to the existence of conditional cooperators. We conclude that cooperative behavior observed in public goods games is not a pure artefact of confusion and learning.