Turning from Crime: A Dynamic Perspective

This paper presents a dynamic model of criminal choice in which an offender's probability of arrest depends on his level of criminal activity as well as the actions of the criminal justice system. In addition, current choices affect future outcomes through a stock accumulation process. We introduce social capital stock into the preference structure to represent the flow of services associated with a good reputation and social acceptance. Social capital is used to account for the influence of social norms on the crime decision by assuming that the stigmatism associated with arrest depreciates an individual's social capital stock. Euler equations exclusive of the value function cannot be obtained from the first-order conditions of our dynamic model since the probability of arrest depends on the individual's own actions. This becomes an issue for estimation because the value function is not specified completely by parameterizing the utility function. We overcome this issue using a nested fixed-point algorithm within the estimation procedure to numerically approximate the value function.