Does Promoting One Behavior Distract from Others? Evidence from a Field Experiment

Impact evaluations of behavioral interventions typically focus on target outcomes. However, if effort and attention are costly, generic interventions might induce negative spillovers on other behaviors. I run a large field experiment in which individuals receive combinations of messages and incentives promoting two behaviors, meditation and meal logging. I find that the interventions reduce completion rates of the opposite behavior by 19-29%. Decomposing spillovers using a new taxonomy, I find that spillovers act in part as a fixed cost: interventions with larger target effects will not necessarily generate larger negative spillovers. I discuss implications for cost-effectiveness analysis