One of the main tenets of economics is that incentive systems need to be designed in order to minimize the difference between private and social interests. Legal rules, taxes, subventions, and other external interventions are regarded as necessary to motivate prosocial behaviors. The sanctions are intended to lower the return on self-interested behavior and increase the incentive to behave pro-socially. 

Frequently observed types of sanctions that impose tangible costs of money on those who do not cooperate are non-deterrent or mild (Engel 2014). Kahan (2000) has written about the difference between “gentle nudges” versus “hard shoves” in legal regulation. The latter correspond to sufficiently severe sanctions, which are meant to directly change behavior, while the former present an indirect inducement to cooperate: mild monetary sanctions may activate norms of cooperation (Cooter 1998). This norm-mediated effect of mild sanctions is often called “expressive”. Tyler and Feld (2006) suggest a direct and an indirect effect of mild monetary sanctions. First, the enactment of such sanctions may directly activate cooperation norms because it might suggest that free-riding is unacceptable even for one willing to incur the sanction. Second, they may improve cooperation because of its indirect effect on the expectations about others' behaviors. The common knowledge aspect of the sanction may indeed have a “multiplier” effect. But mild monetary sanctions have also been shown to crowd-out prosocial behavior (Bohnet et al. 2001, Gneezy and Rustichini 2000). The effect of mild/expressive monetary sanctions is therefore contested.

As in Dal Bo and Dal Bo (2014), expressive policy instruments to increase prosocial behavior can also be nonmaterial in nature. The authors exposed subjects to different moral messages and investigated their effect on contribution levels in a public good game. As with mild sanctions, there are two mechanisms through which moral suasion operates. The first is by directly affecting preferences. The second is by indirectly affecting expectations about the behavior of others.

The relative effectiveness of these two types of expressive motivations – mild monetary sanctions versus moral suasion – is however unknown.

In this article, we report the results from a simple experiment that demonstrates, first, that the mild monetary sanction has a positive and significant effect on contributions. However, the effect is transitory. While contributions in the first few periods are higher for the treatment with the mild sanction than for the treatment with no-incentive at all (and no-message), they are not significantly higher for the last periods. Second, our results show no significant effect on contributions from the moral message. We observe a decay in contributions in the treatment with the moral message as well as in the one with no-message and no-incentive at all. Interestingly enough, the decay is not significant in the latter treatment, but it is significant in the former (i.e. with a moral message). We conclude that a policy that seeks to avoid being too severe would better rely on mild monetary sanctions than on moral messages (alone).

Finally, our design allows us to compare the effect of mild monetary sanctions before and after a moral message was in place. We observe, however, no order effects. This suggests that monetary incentives, even mild, may be effective even in an environment where a moral norm of cooperation has been explicitly activated. That is, we do not observe a crowding-out effect from the mild monetary sanction, even when morality is salient.

The context within which we study the power of these two expressive policies is the Voluntary Contributions Mechanism (VCM), a game extensively used to study the conflict between self- and group-interest. The game is analogous to many real world scenarios. As such, this study will be of interest to researchers and decision-makers in a wide range of domains.