Date(s): June 18, 2013, 2:00-3:15pm

Location: Room 6124, EPA West Building, 1301 Constitution Ave., NW, Washington, DC

Contact: Carl Pasurka, 202-566-2275

Presenter(s): Dietrich Earnhart and Lana Friesen, Department of Economics, University of Kansas

Description: According to the standard model of enforcement (Becker, 1968), deterrence generated by enforcement stems from two components of expected penalty imposition: likelihood and size. Penalty likelihood reflects the certainty of punishment; penalty size reflects the severity of punishment. This model predicts the relative efficacy of certainty versus severity at deterring crime. For example, based on risk preferences alone, risk neutral individuals are deterred equally by equivalent increases in the probability and severity of punishment, yet risk averse individuals are deterred more by increases in punishment severity than equivalent increases in punishment likelihood, while risk lovers are deterred more by increases in punishment probability. Other factors also influence relative efficacy.

Despite much theoretical analysis (Polinsky and Shavell, 2000), the relative efficacy of certainty versus severity remains an unresolved question. Our empirical study contributes to this literature. First, it examines separately the likelihood and severity of punishment, compares their relative strengths, and tests whether one effect dominates the other. Second, our study measures the likelihood of punishment based on the presence of actually imposed sanctions. Third, our analysis adjusts the enforcement and inspection measures with respect to facilities’ compliance history preceding regulatory interventions. This adjustment improves the measures’ abilities to capture facilities’ perceptions of the expected likelihood and severity of future punishment since punishment is driven by facilities’ compliance. Preliminary results appear to reveal that the certainty of punishment and the severity of punishment are comparably effective at inducing better compliance with discharge limits, i.e., lower discharge ratio, for both pollutants. If these results stand, regulators should utilize both features comparably.

Seminar Category: Environmental Economics