Yes, I'll do it: a large-scale experiment on the volunteer's dilemma

In many real-life situations people face a simple decision whether to volunteer or not to provide some benefit for themselves and also for others. This research investigates the effects of the group size and the magnitude of the volunteering cost in a controlled large-scale laboratory experiment, where subjects play the volunteer's dilemma only once. The experiment varies group sizes ranging from groups of 3 to about 100, and 2 different cost/benefit ratios. Results show that high cost reduces volunteering probability only in the smallest groups, but not for other group sizes. Furthermore, non-monotonic group size effect is found on the individual volunteering decisions. These findings are not in line with the mixed-strategy Nash equilibrium prediction. Subjects volunteer more often in most treatments than the Nash prediction which benefits them on average compared to the Nash prediction.