Giacomo Opocher - University of Bologna

Abstract

Empirical research shows that individuals’ responses to treatments vary along latent characteristics, such as innate ability or motivation. Therefore, a policymaker seeking to maximize social welfare may consider assigning treatments not only based on observed characteristics, but also on estimated latent traits. I characterize how the accuracy of these estimates affects the worst-case performance of policies, deriving new regret bounds for assignment rules that include or exclude them, and illustrate new trade-offs with the complexity of the policy space. I then study how a policymaker can mitigate this effect either through better estimates (via empirical Bayes, or leveraging heterogeneity in the estimation error across units), or by collecting more data before solving the decision problem. In an empirical application in development economics, I show that including a proxy for entrepreneurs’ business skills in policy recommen-
dations increases welfare by 6%, and halves the probability of generating welfare losses.

Additional information:

• Speaker: Giacomo Opocher
• Time: Thursday, 22.01.2026; 16:00
• Location: Faculty Lounge, Room 0.036
• Further links:
  • Speaker's webpage
• Organizer: Statistics Group
• Contact:
  • Almut Lunkenheimer
  • +49 228 73-9228
  • ifs@uni-bonn.de