ABSTRACT
The predictive strength of substance-using peers may ret1ect differential association or social learning; that is, deviant conduct may be determined in part by normative and interpersonal int1uences, differential reinforcement of deviance, and modeling effects of substance-using peers (Akers, 1994). Such effects are not anticipated in theories of self-control or social bonding. However, the path from substance-using peers to deviant conduct can be read as consistent with control perspectives. It may, first, represent a sorting effect; persons with low self-control tend to t10ck together and share a taste for risk (Hirschi, 1969; Hirschi & Gottfredson, 1995). Substance-using peers may, second, represent exposure to greater opportunity for drug use (Evans et aI., 1997; Kaplan, 1995).
