Economics serving society

The effect of past enforcement on current cooperation

Permanent short link to this page:

Roberto Galbiati, Emeric Henry and Nicolas Jacquemet

Social dilemmas are situations in which the collectively optimal outcome differs from the outcome arising from decentralized, uncoordinated, individual decisions. This kind of situation typically requires policy interventions aimed at realigning private interests with social incentives. The usual tools to achieve such a goal are control - forcing particular kinds of decisions - and punishment - changing individual benefits, so that collectively optimal decisions become individually rational (for instance fines for smoking in public places or taxes on the use of plastic bags). The prisoner’s dilemma is the workhorse game to represent such social dilemmas: two prisoners are caught and are given the choice between remaining silent (“cooperate”), or denouncing the other’s crimes. Any prisoner who has not been denounced is released, and a bonus is offered to an innocent who denounces the crimes of the other. In this context, both prisoners are better off remaining simultaneously silent (the optimum of the small society made of them both), but are tempted to deviate and be the only one to denounce the other. Since both players face this same incentive, the predicted outcome is that both prisoners choose to denounce. A vast amount of evidence based on laboratory experiments has shown that imposing fines on non-cooperation improves the outcome, and promotes cooperation.

In this research, Roberto Galbiati, Emeric Henry and Nicolas Jacquemet study the dynamic effects of such fines on cooperation: punishing deviations with fines not only change the current willingness to cooperate but also the future willingness to do so, even in environments where fines are no longer in place. These results suggest that if fines are costly to implement, a one-time investment in an enforcement policy can have future effects. The social returns of enforcement can well be higher than the current benefits of cooperation, as future behavior adjusts to this experience. The study also shows that two different channels could explain such behavioural spillovers of policy interventions to promote cooperation: past fines not only increase the current willingness to cooperate (through “direct spillovers”) but also increase cooperation among people met in the past, which in turn results in higher willingness to cooperate today (“indirect spillovers”). The two kinds of spillovers have drastically different consequences on the persistence of the effect of policy interventions on subsequent behaviour: while direct spillovers are short living, indirect spillovers induce a snowball effect, whereby more cooperation today induces more cooperation tomorrow, which in turn promotes cooperation in the long future. The authors show that indirect spillovers dominate direct ones. As expected, such indirect spillovers induce a persistent effect of policy interventions put in place in the past: individuals respond to past environments with fines that they themselves never experienced.


Original title: Dynamic effects of enforcement on cooperation.

Published in: Proceedings of the National Academy of Sciences, Vol. 115 (49), pp.12425-12428

Available at: