Fehr–Gächter public-goods experiment

The Fehr–Gächter public-goods experiment tests a basic tension in cooperation: each person can benefit from a shared resource while keeping the personal cost of contributing to a minimum. In the experiment, groups initially contributed despite the incentive to free ride, but contributions fell as participants observed others withholding their share. Making contributions visible and allowing participants to punish free riders reversed that decline. The result shows how social expectations and enforcement can support collective cooperation even when contributing is individually costly.

The game and its incentives

The experiment brought four people together for four rounds. Each began with twenty tokens and could either put tokens into a shared public pot or keep them. A contribution cost its maker one token, but returned 0.4 tokens to each member of the group: the contributor’s personal return was only 0.4, while the group gained 1.6 tokens altogether. If everyone kept their tokens, each left with twenty; if everyone invested all their tokens, each left with thirty-two. The arrangement therefore made the individual temptation and the collective advantage plain: a player could do best personally by keeping their tokens while benefiting from other people’s contributions, but universal free riding left the group with less.

Contributions decline as free riding becomes visible

At first, most participants did not follow the strictly self-interested strategy. Across similar experiments in the developed world, most contributed about half their tokens to the public pot. But contributions dropped from round to round as players saw others free ride. By the end, 70 to 80 percent of participants were free riding, and the group as a whole was poorer than it would have been had cooperation continued.

Fehr and Gächter’s account distinguishes three broad patterns of behavior. About a quarter of participants were consistently selfish in the economic sense, always choosing to free ride. A small minority were altruists who contributed heavily from the start and continued even as others withheld their tokens. The largest group consisted of conditional consenters: they began by contributing at least some of their wealth, but became less willing to do so when they saw others not contributing. By the end of most games, almost all of this group had stopped cooperating. The downward course of the game thus depended not only on the committed free riders, but on the response of people whose cooperation was contingent on whether it seemed reciprocal.

Visibility, punishment, and renewed cooperation

In a modified version, participants learned at the end of each round how much each person had contributed. They could then punish a free rider: spending a third of a token allowed them to take one token away from that person. Punishment carried a cost for the person imposing it, so it was not the individually costless response predicted by a narrow account of self-interest. Nevertheless, participants spent tokens to punish free riders, and those free riders began contributing their fair share.

The change lasted beyond the rounds in which punishment was available. Even in the final rounds, when no further punishment could be imposed, participants continued to contribute. Making free riding visible and punishable had therefore shifted behavior in a way that persisted after the immediate sanction was removed. In the experiment’s terms, enforcement did more than penalize noncontributors: it helped sustain the expectation that others would contribute too.

Significance for collective cooperation

The experiment explains why a public-goods arrangement can fail even when its benefits are shared. The shared return gives each person an incentive to let others pay the cost, but observed free riding can also persuade initially cooperative participants that contributing would leave them at a disadvantage. The conditional consenters are pivotal: when they withdraw, the individually tempting strategy becomes the group’s deteriorating outcome. Allowing people to see and punish noncontribution helped keep that group from settling into universal free riding.

The findings connect experimental cooperation to the broader problem of taxpaying. Taxes fund benefits that people can receive whether or not they personally contribute, making free riding individually tempting; yet those benefits depend on people contributing collectively. The experiment suggests that cooperation depends not just on formal obligations, but on whether people believe others are meeting theirs and whether free riding is subject to credible consequences. It does not establish that punishment alone explains real-world tax compliance, but it demonstrates one mechanism through which costly enforcement and reciprocal expectations can preserve cooperation.