Abstract: This study tests the applicability of an integrated model of deviance - social bonding and learning theories - to drug use among a representative sample of U.S. adolescents (12-17 years old). A structural equation model (SEM) was estimated across all subgroups (age, race, and gender) as well as the overall group. The relationships between exogenous variables (social bond and delinquent peer) and endogenous variables (delinquent peer and drug use) were significant and in the hypothesized direction for the overall group and for each subgroup. The results also showed some differences and similarities across demographic groups. The explained variance in substance use ranged from 0.27 to 0.48. Applications for future study are also discussed.
Keywords: social bonding theory; learning theories; drug use; structure equation model.
The adolescent life-stage is a period of high risk for engaging in many different kinds of problem behaviors, such as substance use (e.g., cigarettes, marijuana, alcohol) and delinquency. Involvement in these acts can place youth at increased risk of future criminal involvement or social maladjustment. Some studies (Elliott 1994; Moffitt 1993; Nagin and Paternoster 1991; Sampson and Laub 1993) have documented that early involvement in antisocial behavior is strongly related to criminality in adulthood.
Among juvenile deviance, drug use is a common phenomenon. A substantial body of research has suggested that involvement in drug use has become a national concern, whether it is alcohol (Barnes 1984; Wechsler et al. 1984), or marijuana use (Smith 1984). Ellickson, Collins, and Bell (1999) have suggested that the use of "hard" drugs (e.g., heroin, cocaine) commonly follows the onset of "gateway" drug use, such as alcohol and marijuana. Moreover, the Office of National Drug Control Policy (ONDCP 2003) has found that youth substance use or abuse can cause many negative consequences, including deviant acts (e.g., early sexual initiation and suicide) and delinquency. For these reasons, identifying and understanding the dynamics underlying youths' drug use are important.
The present study seeks to assess important social factors in understanding adolescent drug use. A theoretical model of adolescent drug use that integrates central ideas from social control theory (Hirschi 1969) and learning theories (Sutherland and Cressey 1966; Akers 1973) is formulated and tested. This research does not aim to compare the usefulness of both theories. Rather, it is hoped that by combining the important concepts of learning theory to social control theory, more insights into juvenile substance use can be obtained. Although many studies have employed the same idea to study juvenile drug use (Aseltine 1995; Ellickson, et al. 1999; Marcos, Bahr, and Johnson 1986; Massey and Krohn 1986), this study departs from previous studies in an important way in that the present study applies this integrated model across different demographic groups (e.g., gender, race/gender). In so doing, this study provides insights of the differences of drug use across demographic groups and adds to the information from previous studies which consider important demographic variables as control variables.
Social Control Theory
Hirschi's (1969) social control theory argued that adolescents who had no strong bond to conventional social institutions were more likely to commit delinquency. Many empirical studies that follow Hirschi's theory have found general support that juveniles who have strong social bonds are involved in fewer delinquent acts (Agnew 1985; Costello and Vowell 1999; Erickson, Crosnoe, and Dornbush, 2000; Hindelang 1973; Hirschi 1969; Junger-Tas 1992; Sampson and Laub 1993; Thornberry et al. 1991). Some studies that specifically employed social control theory to explain juvenile drug use have also found support for this theory (Ellickson et al. …