Recently, there has been renewed interest in the contribution of labeling theory to the study of delinquency (Ray & Downs, 1986; Paternoster & Iovanni, 1989; Matsueda, 1992). The basic assumption of the theory is that perceived negative societal reactions lead to the development of negative self-conceptions and greater delinquent involvement (Lemert, 1951; Becker, 1963). Labeling theorists have stressed the importance of both formal and informal labeling (Lemert, 1951). Formal labels are those obtained through contact with social control agencies, whereas informal labels are generated by parents, teachers, and peers.
Contact with social control agencies is believed to "stigmatize" juveniles (Garfinkel, 1956). One of the possible responses to being stigmatized or negatively labeled is involvement in delinquent behavior. The results of numerous studies show that juveniles who are formally processed through the juvenile justice system and have formal contact with other social control agencies report greater subsequent delinquency (Tannenbaum, 1938; Mead, 1964; Gold, 1970; Ageton & Elliott, 1974; Gibbs, 1974; Klein, 1974; Farrington, 1977; Hepburn, 1977; Thomas, 1977; Thomas & Bishop, 1984).
Fewer studies have examined the effects of informal labeling on delinquency. When informal labeling is the focus of research, much of the emphasis is placed on negative parental reactions (Aultman & Wellford, 1979; Matsueda, 1992). This practice overlooks the potentially important role of teachers and peers in the labeling process.
Though empirical research has supported the predictive power of labeling, there continues to be debate in the criminological field regarding the status of labeling as a theory of delinquency (Hagan, 1974; Meade, 1974; Tittle, 1980). The major criticisms of the labeling perspective are that its alleged theoretical imprecision results in insufficient clarity regarding key concepts, and that its propositions are not empirically verifiable (Wellford, 1975; Gove, 1980).
The present study is intended to improve the status of the labeling perspective by first providing an unambiguous rendition of the theory and a methodology for measuring self-concept. Second, the relative explanatory powers of both formal and informal labeling are assessed. Third, many of the previous attempts to confirm labeling theory's secondary deviance hypothesis have been in the form of integrated theoretical models (Menard & Morse, 1984; Simons et al., 1980). In contrast, the present study assesses a model including only theoretically derived labeling variables and appropriate background variables. Finally, this study assesses the robustness of labeling theory variables' predictive powers in terms of explaining involvement in different types of delinquency (i.e., general, serious, and drug-related).
Symbolic Interaction and Labeling Theory
The symbolic interaction perspective provides the conceptual and theoretical foundation for labeling theory's secondary deviance hypothesis. The first question labeling theorists must address is how do juveniles form their self-conceptions for which they attach negative labels? By answering this question, researchers can then proceed to test hypotheses concerning the relationships between perceived negative labeling and delinquency.
Labeling theorists, borrowing from Mead's (1934) social psychology, assume that juveniles, in their routine activities, are bombarded with different cues and clues as to how they are perceived by others in their community. Juveniles, through role-taking (Cooley, 1902) and defining situations (Thomas, 1923), are able to accurately interpret the meanings of symbols and gestures used to project labels upon them.
This ability to make choices and to participate in cooperative interaction via significant symbols suggests that human beings are not passive receptors of negative labels (Manis, 1955; Miyamoto & Dornbusch, 1956; Couch, 1958; Reeder et al., 1960; Maehn et al., 1962; Kinch, 1963; Sherwood, 1965). Labeling theorists acknowledge that some juveniles negotiate labels and indeed attempt to disavow their deviant imputations (Davis, 1961). Thus, the overly deterministic view (Akers, 1967) of labeling theory's secondary deviance hypothesis appears unwarranted.
Mead believed that "to name or define something is never merely an 'idealistic' procedure. It is instead a consequence of an act" (Melossi, 1985, p. 199). This line of reasoning is consistent with Becker's (1963) notion that social groups create deviance by their reactions to known acts. From this perspective, then, "a name, definition, or label designates something which is the product of a successful conversation of gestures" (Melossi, 1985, p. 199). These successful conversations of gestures are what makes the process of labeling the "self" possible.
To summarize, labeling theorists assume that, during real or imagined interactions, individuals project themselves into the role of significant others and make assessments or self-appraisals (Cooley, 1902; Shibutani, 1961; Bern, 1972). The self becomes an object (Mead, 1934) for which the individual attaches labels, both positive and negative. This assumption is guided by the view that humans have the ability to choose among competing labels for their self-conceptions.
Measuring Perceived Informal Labeling of the Self
The major obstacle to empirically verifying labeling theory's secondary deviance hypothesis is the self-concept variable. Most self-concept studies have relied on global measures of self-esteem (e.g., see Rosenberg, 1979) instead of focusing on the adoption of a deviant self-concept. This body of literature adds little support for labeling theory because a juvenile can have a negative self-concept and yet have high …