Slippery fish: enforcing regulation when agents learn and adapt | Mushfiq Mobarak (Yale)

Mushfiq Mobarak will be discussing the paper Slippery Fish: Enforcing Regulation when Agents Learn and Adapt which they co-authored with Andres Gonzalez-Lira.

Abstract

Attempts to curb illegal activity through regulation gets complicated when agents can adapt to circumvent enforcement. We present a model of enforcement where agents can learn about patterns and weaknesses of the auditing process over time. Conducting audits on a predictable schedule, and (counter-intuitively) at high frequency, can undermine their effectiveness as agents learn, adapt, and take advantage of loopholes. We conduct a large-scale randomized controlled trial with the Chilean government to test these ideas by auditing vendors selling illegal fish. We test the model’s specific predictions on learning and adaptation by tracking vendors daily using mystery shoppers. We cross-randomized a information campaign to consumers, to test whether simple demand-side interventions alleviate the need for complex monitoring and enforcement. Low-frequency, unpredictable enforcement is most effective at curbing illegal fish sales, but accounting for vendor adaptation, the information campaign proves to be almost as cost-effective, and easier
to implement.