Specific Deterrence and Speed of Punishment: Causal Evidence from high-frequency Micro Data

• Applicant: Professor Dr. Christian Traxler
• Subject Area: Economic Policy, Applied Economics
• Term: from 2017 to 2021
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 323565767

This project follows an economics of crime approach to study specific deterrence, i.e., the effect from actually experiencing punishment on an individual's risk of re-offending. Our institutional setting -- a radar-based enforcement of speeding tickets in a Prague suburb -- offers quasi-experimental variation in the probability of being punished, in the magnitude as well as in the delay of punishment. Moreover, we implement a randomized control trial in cooperation with the enforcement authority to generate truly experimental variation in the speed of punishment. Exploiting the different sources of exogenous variation and high-frequency micro data on roughly 22 million observations, we first provide causal evidence on specific deterrence. The data allow us to identify if drivers (a) adjust their speed and avoid certain speed levels; (b) whether drivers' responses are short-lived or long-lasting and (c) whether they change their driving frequency or circumvent specific roads (extensive margin responses) after receiving a ticket. We can also exploit a second discontinuity in the penalty structure to evaluate (d) whether drivers' responses depend on the magnitude of the sanctions communicated in speeding tickets.Beyond that, we will also investigate -- both theoretically and empirically -- how the specific deterrence effects depend on the delay between the offense and the experience of a sanction. The formal analysis will thereby consider traditional learning models in economics (reinforcement learning) and approaches from behavioral economics (mental accounting). Based on the theoretical analysis we will derive testable predictions which ideally allow us to capture the underlying mechanism behind the observed treatment effects.Finally, we evaluate the overall impact of the radar system and discuss revenue and welfare implications.

• DFG Programme: Research Grants
• International Connection: Czech Republic
• Cooperation Partner: Dr. Libor Dusek