Deterrence: A Panel Analysis

Recently, two research groups have tested the deterrence model using longitudinal designs (Saltzman, et al. 1982; Minor and Harry 1982). They suggest that most cross-sectional survey research on deterrence is so methodologically flawed as to be arguably irrelevant. Moreover, using a research strategy which is clearly more appropriate to test the model, they uncover no evidence of deterrent effects. This paper represents a partial replication and extension of these studies, using two-wave panel data from a sample of 2,147 adolescents. Unlike the prior panel research, the findings here support the viability of the deterrence model. They also indicate that there are significant differences in responsiveness to the perceived threat of sanctioning in terms of sociodemographic characteristics of potential offenders. The methodological and theoretical implications of the research are discussed.