Speed enforcement in Norway: Testing a game-theoretic model of the interaction between drivers and the police

This paper probes the relationship between changes in the risk of apprehension for speeding in Norway and changes in the amount of speeding. The paper is based on a game-theoretic model of how the rate of violations and the amount of enforcement is determined by the interaction between drivers and the police. This model makes predictions both about how drivers will adapt to changes in the amount of enforcement (the more enforcement, the less violations) as well as how the police will adapt to changes in the rate of violations (the less violations, the less enforcement). The paper attempts to test the game-theoretic model empirically. Testing the model rigorously is difficult, mainly because some of the relevant variables are not reliably measured and are endogenous. Two models were developed: one to identify sources of changes in the rate of violations, one to identify sources of changes in the amount of enforcement. The predictions of the game-theoretic model were supported, although the results were not statistically significant in the model of how the police adapt enforcement to changes in the rate of violations.