Marina Ligato (Université libre de Bruxelles)
Sébastien Fontenay and Ilan Tojerow
Social security systems feature complex tax and benefit rules and limited scope for individualized counseling. We distinguish between awareness of rules (general, static information) and computational complexity (personalized, interactive information). We develop an online benefit simulator and test its effects in a randomized field experiment. Providing generalized information has no impact on behavior, while reducing computational complexity through the sim ulator increases returns to work among claimants by 3.75 percent after one year. Simulation evidence shows that work incentives are difficult to compute and highly heterogeneous across claimants. Our results suggest that personalization is a more effective lever than salience in complex policy environments, pointing to a high return role for digital counseling tools in social insurance.