Prosocial Behavior and Policy Spillovers: A Multi-Activity Approach

Observing that people who wish to engage in prosocial behavior are often presented with more than one means to the same end, we develop a theory in which agents may contribute to a single public good through a range of different activities. Our aim with this extension is twofold. First, we deliver positive results. Noting that effort on one activity has been argued to crowd out ("moral licensing") as well as in ("moral consistency") effort on other activities, we predict that for a large set of plausible cases, policy to facilitate one activity reduces effort on other activities. However, this negative spillover effect is incomplete in the sense that overall public-good production still increases. Second, we revisit prominent results from single-activity models in the published literature and show that they grow rather more ambiguous, or even fall apart, when they are extended to our multi-activity setting. This is not due to dimensionality per se, but to the fact that single-activity models implicitly assume that agents use narrow mental accounts to categorize activities. By contrast, our model imposes broad accounting.