L'explication de la delinquance prolifique: L'importance de l'interaction entre le risque individuel et le risque contextuel.

摘要

The purpose of this dissertation is to explain the prolific delinquency of certain offenders. We suggest that prolific delinquency is explained by the formation of a greater number of criminogenic situations. A criminogenic situation makes reference to situations wherein an offender may interact with a criminal opportunity in an environment which is conducive to crime. More precisely, a criminogenic situation is the moment when an offender faces this opportunity, but the crime has not yet been committed. The formation of criminogenic situations facilitated by the interaction and interdependence of three elements: criminal propensity, criminal social environment and deviant lifestyle. Thus, prolific delinquency cannot be adequately explained without accounting for the interaction between individual and contextual risk.;The overall objective of this dissertation is to demonstrate the importance of a model based on the interaction between individual and contextual risk to explain the prolific delinquency of some offenders. To accomplish this objective, one hundred and fifty-five offenders, under the responsibility of Services Correctionels du Quebec and four Centres Jeunesse, completed an evaluation through self-administered questionnaires.;The first objective of this study was to describe and compare, in Chapter Three, the delinquent nature of the offenders in our sample. Our results revealed that our sample of offenders is similar to that of other samples of delinquents; we found this with respect to the nature of their delinquency, and in particular, the volume, diversity and severity of their crimes. Indeed, the majority of the participants reported a small volume of crimes against a person and property. A small group distinguished themselves with a very high lambda (13.1% of offenders in the sample are responsible for 60.3 % of all crimes reported). Additionally, more than four out of five offenders reported having committed at least one crime against a person and one crime against property. Moreover, 50 % reported having committed crimes in at least four subcategories. Finally, although the offenders in our sample have a relatively low IGC (gravity scale) mean (median = 77), nearly 40 % of the offenders reported having committed at least one of the two most serious crimes identified in this study (discharging firearm and robbery).;The second specific objective was to explore, in Chapter four, the interaction between personal characteristics, social environment and the lifestyle of offenders which may lead to criminogenic situation. People with a higher propensity to crime tend to be surrounded by other criminalized people and have a more idle lifestyle. The criminalized social environment tends to also influence the lifestyle of these offenders. Thus, the interdependence between these three elements can lead to criminogenic situations and can create a climate conducive to the emergence of prolific delinquency.;The last specific objective of this dissertation, covered in Chapter Five, is to analyze the impact of factors leading to situational crime on the nature of the delinquency. Analyses of multiple linear regression and regression trees highlighted the contribution of personal characteristics, social environment and lifestyle in explaining the nature of the crime. On the one hand, regression analyses (additive models) suggest that all the elements leading to situational crime make a unique contribution to the explanation of delinquency. However, on the other hand, regression trees allowed us to better understand the interaction between the elements which underlie prolific delinquency. For example, a lower position on certain items may be offset by a higher position on others. Moreover, the accumulation of risk factors which lead to situational crime does not happen in a linear fashion. These conclusions are supported by the proportions of explained variance which were higher for the regression trees than for the multiple linear regressions.;In conclusion, focusing simply on one element (the person and their criminal propensity or the context and its opportunities) or on their combination in a simply additive manner does not represent the reality of criminal phenomenon. This dissertation therefore serves to highlight the importance of considering the interaction between the individual risk and the contextual risk in explaining prolific delinquency.;Keywords: individual risk; contextual risk; lifestyle; interaction effect; classification and regression trees.