Experimental Evidence of Magical Thinking in Public Goods Experiments

Consistent behavior in public goods games is well documented. Typically, participants begin with large contributions to the public good, but contribution rates decline as more iterations of the game are played. However, the impact of magical thinking on contributions to a public goods game has yet to be examined. We hypothesize that, when participants erroneously believe they can use their contribution to set a social norm, contribution rates will be higher. To test this hypothesis, we had participants play a public goods game, and gave them either no additional information or told them that there is a real probability that their contribution would be recorded first. We found no significant difference in contributions between the treatment and control groups. However, magical thinking was ubiquitous across groups, suggesting that magical thinking is a normative feature of behavior in public goods games that is robust to certain manipulations.