Lugar: Sala de Reuniones del Departamento de Economía. Edificio Los Madroños, 2ª planta (ECON-2026)

Ponente: Sergio Cappellini (Università degli Studi di Padova)

Título: Optimal Welfare-to-Work Programs with Worker Profiling

Abstract: Profiling plays a crucial role in welfare-to-work programs by categorizing unemployed individuals based on their skills and aligning them with appropriate labor-market policies. This study formulates an optimal profiling strategy by incorporating dynamic learning about recipients' abilities within a principal-agent framework. The optimal profiling approach allows a certain percentage of low-skilled workers to perceive themselves as high-skilled, leading to their inclusion in delegated job searches alongside genuinely high-skilled workers (positive type II error). This occurs when the government prioritizes motivating overly optimistic low-skilled workers to actively search for employment, with lower incentive costs, rather than directing them to passive labor-market policies. Conversely, high-skilled workers are consistently accurately classified as such, ensuring they are not referred to passive policies (no type I error). The implementation of an optimal profiling strategy in the United States is estimated to result in per-capita net-present savings ranging from approximately $26,620 in New Jersey to $28,960 in Oregon.