THEORETICAL AND EMPIRICAL SPECIFICATION OF A MODEL OF INFORMAL LABELING

Recently criminologists have begun to explore the importance of disagg regating frequency measures of self-reported delinquency into the sepa rate decisions of initiation and continuation. Given that labeling mak es predictions concerning continuation, the purpose of this paper is t wofold. The first is to test the ability of an informal labeling model to predict the decision to continue delinquent behavior once it is in itiated. The second purpose is to address the broader question of whet her disaggregation matters. The findings support the idea that the inf ormal labeling model is predictive of the decision to continue delinqu ent behavior. The findings also suggest that, at least for a measure o f general delinquency, there are some differences to be found by choos ing the appropriate sample and form of the dependent variable.