Rational Choice, Deterrence, and Identity: Modeling Life Course Transitions and Desistance

This project has four major goals: (1) It estimates models of offender decision-making, including criminal rewards, costs, and opportunities, as well as criminal identities. (2) It develops new methods for analyzing trajectories to estimates the effects of life course transitions work, parenthood, and marriage. (3) It tests whether offender decision-making, changes in peer groups, and changes in criminal attitudes and identities explain the effects of life course transitions on desistance. (4) It uses new qualitative data explore new hypotheses and potential points of intervention. Subjects: Subjects are from the DYS, a nine-wave panel study of 1528 youth from disadvantaged neighborhoods. To check the robustness of our findings, we supplement analyses with data from the NYS, a nine-wave panel of 1725 youth in the U.S., and the Pathways to Desistance Study, an 11-wave panel of 1,354 adjudicated juvenile offenders in Phoenix and Philadelphia. Research Design and Methods: The longitudinal designs allow us to model individual age-crime trajectories using a new curve registration approach from which substantive quantities, such as rates of desistance naturally follow. We will extend our model by incorporating time-varying covariates, measures of desistance, and clustering of trajectories, and compare results to those based on fixed-effects, HLM, hazard, and group-based trajectory models. Analysis: Our analyses take four steps. First, we extend our models of offender decision-making and compare results across datasets. Second, we estimate age crime curves using our curve registration method and then estimate effects of life course events on desistance. Here we explore treatment heterogeneity by age, gender, race, and social context. Third, we test whether the effect of life course transitions on desistance are due to changing perceptions of criminal risks, rewards, and opportunities, or from changing peer groups, attitudes toward crime, or criminal identities. Fourth, we use in-depth interview data from 179 offenders in the DYS to explore desistance and persistence from the viewpoint of the offender, including views of the meaning of crime, changes in risks and rewards, and changes in identity. Products, Reports, and Data Archiving: In addition to semi-annual and final reports, we will publish a number of articles, including methodological papers on extensions to our curve registration model, and substantive papers on offender decision-making, life course transitions, the causal mechanisms producing desistance, and policy implications. We will produce a user-friendly program in R to estimate our UCR models. All datasets will be archived by the end of the funding period.