Peer groups and substance use: examining the direct and interactive effect of leisure activity.

The peer group is the center of the adolescent life-world. Adolescents spend more time with their peers than they do with their parents or alone (Csikszentmihalyi & Larson, 1984; Savin-Williams & Berent, 1990). The emergence of international, commercially driven popular youth cultures makes the peer group particularly salient for today's youth. This youth culture provides global markets that offer many commercial opportunities. These markets have been heavily targeted by the entertainment industry that promotes the culture of fun and excitement and fosters unrealistic expectations. The development of a global youth culture has at the same time provided new opportunities to promote the use of alcohol and drugs among adolescents.

Peers can enforce or challenge the norms and authority of adult society. On the one hand, research indicates that having deviant friends is a major source of risk-taking and delinquent behavior. Thus criminologists argue that association with deviant peers leads to deviance through mechanisms of social learning, peer pressure, and transference of deviant attitudes and values (Akers, 1977; Sutherland, 1947; Warr, 1993; Warr & Stafford, 1991). On the other hand, youth researchers have paid attention to the positive aspects of youth culture, pointing out that participation in youth leisure activities often provides adolescents with valuable experience. Participating in adolescent activities can expand horizons, offer opportunities to develop skills, and foster a sense of acceptance and belonging (Bartko & Eccles, 2003; Dworkin, Larson, & Hansen, 2003; Larson, 2000; Morgan & Sorensen, 1999; Youniss et al., 1999). Research suggests that youth activities can provide a context in which adolescents are emotionally and cognitively engaged in exploring identities (Dworkin et al., 2003; Youniss et al., 1999), enhancing social skills and personal development (Dworkin et al., 2003; Larson, 2000). There is also a body of research which suggests that participation in extracurricular activities is positively related to academic performance, psychological well-being, and self-esteem, but negatively related to substance use (Barko & Eccles, 2003; Thorlindsson, 1989; Thorlindsson & Wilhjalmsson, 1991).

Although drawing upon these different traditions of defining and classifying adolescent cultures and society, we do not attempt to integrate or synthesize them in any systematic way. (We use the concepts of subculutre and leisure activity somewhat interchangably to refer to the same aspect of adolescent society. We draw on subcultural theories, recognizing that the widespread and long-standing use of subcultural theories in social science does not indicate a consensus of the meaning's origins and the influences of subcultures (Fine & Kleinman, 1979; Hagan et al., 1998), but we maintain that subcultural theory helps us capture important aspects of peer society.) We draw from the specific works of specific scholars. Thus our project is framed by the research on adolescent society, but we also draw heavily from Matza's work (1964; Matza & Sykes, 1961), emphasizing how the quest for fun and excitement is a ubiquitous aspect of adolescents' leisure activities and how it can take different forms. From Hagan (1991), we adopt the concept of a party subculture as a central aspect of adolescent leisure activity.

Consequently, this article blends ideas and concepts from criminology and youth leisure research to examine the association between youth peer cultures and substance use. We assume that having fun, seeking excitement, and challenging adult society, are essential characteristic of peer group activities. Following Matza (1964; Matza & Sykes, 1961), adolescents from all walks of life are members of a leisure class characterized by weakened social control, especially on the part of parents and the school. At the same time, adolescents are free from the demands of self-support and the integrative bonds of work and marriage, giving them the freedom to seek leisure without the adult responsibilities. Adolescents draw upon subterranean values that include disdain for hard work, quick success, as well as the seeking of excitement, adventure, and thrills. According to Matza (1964) these values exist side by side in contemporary society with values of hard work, a capacity to endure the mundane day-to-day routine, and standardization that characterizes so much of everyday life.

In line with the youth research literature, we emphasize the diversity of peer culture and peer lifestyle and the importance of the peer group in promoting various kind of behavior. But we argue that some forms of leisure activity may help integrate adolescents into society and enable them to reach shared societal goals, whereas other activities may foster subcultures that challenge the normative consensus of conventional adult society. Classifying and comparing youth leisure activities across domains with reference to specific outcomes may help us gain a better understanding of peer group influence.

We identify three types of highly popular peer activities that offer alternative ways of forming and organizing peer subcultures, having fun, and seeking excitement. The first leisure pattern involves sport participation. In recent years, an expanding group of social scientists have generated a considerable body of knowledge about sport subcultures (Donnelly & Young, 1988; Klein, 1986). Building on this research, we argue that participation in sport can provide a social context that is characterized by distinguishing patterns of values, norms and rituals, and practices that shape the behavior of its participants. Following Hagan 1997, 1991), the second leisure pattern involves activities that consist mainly of hanging out and going to parties. According to Hagan, the "party subculture" is intertwined with contemporary popular culture. It is central in the lives of adolescents and is characterized by the suberranean values that pervade the contemporary popular culture. The third leisure pattern is characterized by participation in adult supervised and/or organized leisure, promoted by social youth clubs. Youth clubs offer the opportunity to have fun with peers, socialize, dance and listen to music in an adult-guided environment.

In the present study, we use data from a nationally representative sample of Icelandic adolescents to study the statistical relationships among adolescent substance use and involvement' in the three leisure patterns, while controlling for social control stemming from strong ties to parents and the school (Hirschi, 1967). We expect involvement in sport and youth clubs to be negatively associated with substance use while involvement in the party subculture should be positively associated with substance use (see Bernburg and Thorlindsson, 2001; Hawdon, 1996; Osgood et al., 1996). Aside from the direct effects of leisure pattern involvement, we argue that differential leisure pattern involvement may condition the well-documented relationship between adolescent deviance and associating with deviant peers. Deviance and crime researchers have shown that social bonding to conventional institutions, such as the family, the school, and religion, weaken the effect of deviant peer association on adolescent deviance (Elliott et al., 1985; Thornlindsson & Bernburg, 2004; Warr, 1993). We expect involvement in sports and youth clubs to have similar protective effects. Thus, we hypothesize that the statistical influence of substance-using peers on substance use should decrease with a higher level of involvement in sports and youth clubs. Conversely, we expect that the statistical influence of having substance-using friends should increase with a higher level of involvement in the party subculture.

CONCEPTUALIZING PEER GROUPS

Having fun is an important part of adolescent peer group activity. However, the search for risk, excitement, and new challenges may result in participation in different leisure or peer group activities. Building on the works of Matza (1964), Campbell (1969), and Hagan (1991), we stress that although teenage culture may in many cases undermine scholastic effort and the dominant adult culture, it may also offer "sufficient inherent satisfactions to attract and maintain" many adolescents who "otherwise would be vulnerable to the appeals of delinquency" (Campbell, 1969:840-41). Involvement in various adolescent activities, even those that border on deviance, can play a part in adolescent socialization, providing supporting networks and a sense of belongingness. Matza (1964) suggests that teenage culture may actually reduce deviancy, because less serious subcultural involvements may have the virtue of curbing more deviant leanings.

Matza's theory rests in part on the notion that the search for fun and the celebration of values such as excitement, toughness, and disdain for work are in fact typical of leisure values held by the adult population. These values are presented in competitive games, drinking parties, gambling, and "concealed deviance." According to Matza (1964) and Matza and Sykes (1961), the delinquent brings out the "subterranean values" of society--hedonism, disdain for work, aggression, violence, and masculinity.

Matza (1964) argued that adolescents who do not have a strong sense of agency and direction may drift between crime and convention if they experience a loosening of social control. He suggests that while some adolescents may be driven or directed by parents and schools to acquire "cultural or social capital" through their extracurricular activities, adolescents with less parental and educational support are more likely to drift into subcultural pursuits that are less supportive of school work and conventional values (see also Bordieu and Passeron, 1977). Following Matza it may be suggested that adolescents who are controlled less by conventional institutions can be lured into subcultural involvements or extra-curricular activities which will reduce their attraction to more seriously deviant cultures.

Building on Matza's ideas we suggest that "less serious" subcultural activities may provide an alternative to more serious delinquency by providing excitement, fun, and opportunities to explore the borders of deviance and convention. Hence participation in a particular activity may act as protection against deviance. Accordingly, we postulate that adolescents involved in subcultures that provide conventional bonding, such as sports and youth clubs, are less likely to be lured into substance use by their substance-using peers.

We recognize that subcultures do not exist in isolation. On the contrary, adolescents participate in different activities and with different groups (Fine & Kleinman, 1979; Hagan, 1991). Hagan (1991) points out that many teenage subcultures which have various meanings and consequences for their adult lives may exist side by side. Adolescents mix with different groups and often come in contact with various subcultures. They often belong to more than one subculture with friends who belong to these others. Fine and Kleinman (1979) point out that youth membership in a subculture is never totally bounded or finite. They suggest that while multiple group membership may play an important role in spreading cultural lore among youth groups, there are boundaries that define networks and appropriate behavior, providing the parameters that characterize various subcultures. Participating more in one subculture than another may be the result of several factors including age, physical proximity, and lifestyle similarities (Maxwell, 2002); or it may happen by chance as a result of being at a certain place at a certain time.

Sport

In recent years researchers have given attention to sport-specific subcultures (Donnelly & Young, 1988; Klein, 1986). While some studies have focused on sport as a whole, others have dealt with subcultures confined to specific sport disciplines (see Klein, 1986). Sport provides a good example of the expression of Matza's subterranean values in a nondeviant subculture. According to Matza (1964), it is one of the few acceptable ways of displaying the values of adventure, excitement, and daring in a leisure social context. It is also a typical leisure activity in the sense suggested by Veblen (1899/1994). It provides a time out from the mundane world of work. But the subculture of sport differs from the party subculture identified by Hagan (1991) and the subculture of delinquency identified by Matza (1964) in the sense that it emphasizes the values of hard work, self-discipline, and accomplishment in a conventional way. Sport may therefore provide a subculture that allows adolescents to express their subterranean values but in a different social context from that of the delinquent and party subcultures. As an alternative it may curb the tendency toward delinquent behavior and the use of drugs and alcohol for adolescents. In one of the earliest studies of sport and deviance, Schafer (1969) suggested that sport may provide an alternative to deviance since it offers adolescents the opportunity to escape from boredom and to experience thrills, adventure, and excitement. Shafer points out that athletes who are interacting primarily with like-minded friends theoretically are shielded from anti-establishment and deviant subcultural values, norms, and behaviors.

This hypothesis is in part supported by considerable research which indicates that participation in organized sport is associated with less involvement in a variety of delinquent activities (Bjarnason, 2000; Fejgin, 1994; Snyder & Spreitzer, 1983; Thorlindsson, 1989; Thorlindsson & Vilhjalmsson, 1991). Some studies have shown that adolescents who participate in sport are less likely to use alcohol, tobacco, and illegal drugs (Burmann, 1997; Hastad et al., 1984; Snyder & Spreitzer, 1983; Thorlindsson, 1989; Thorlindsson & Vilhjalmsson, 1991).

Snyder (1994) acknowledges that while there is some evidence to support the claim that sport deters deviance, the evidence is not conclusive. His case study of athletes who became involved in delinquency illustrates how the risk taking and the quest for thrills experienced in sport can be extended to delinquency. Some scholars, such as Brower (1979) have concluded from his research into youth sports that there is no valid evidence that sports build character. Eccles and Barber (1999) found that involvement in team sport is associated with an increase in both drinking and academic achievement. McCaul et al. (2004) also found that adolescents who participate in intense physical activity were actually more likely to drink alcohol than were adolescents who did not. They conclude that the relationship between physical activity and alcohol use is more complex than previously believed.

Party Subculture

Hagan (1991) has argued that the search for excitement, adventure, and thrills is manifested in the "party subculture" that exists alongside other youth subcultures. The party subculture consists of hanging out and going to parties and dances. It values idleness, shows disdain for school and the Protestant values of hard work. Having fun is the opposite of work. The party subculture is perhaps the one that is most directly intertwined with contemporary popular culture. It is heavily promoted by the entertainment industry and celebrated in popular music and movies. Hagan (1998) has suggested that the party subculture is characterized by the subterranean values that pervade contemporary popular culture. The centrality of the party subculture in adolescents' lives means that it overlaps in various degrees with other subcultures. Thus Hagan et al. (1998) suggest that the party subculture in some cases may even overlap the smaller but "more ominous delinquent subculture." Drinking and the use of mood-enhancing drugs are integral parts of the party subculture and there is evidence that the use of illegal drugs is increasingly promoted and distributed through the entertainment industry (Measham et al., 1998).

Organized Social Club Activities

A considerable effort has been made in Iceland to organize the social life of youth and direct it into a more accepted form of leisure. One such effort is the establishment of organized social youth clubs. These clubs run programs and social events geared for youth of different ages. They draw youth from the local community and are financed by the municipalities. They offer club houses or centers in every neighborhood in the country for social events. Their main purpose is to offer youth the opportunity to have fun with their peers, socialize, dance, and listen to music but without the use of alcohol and drugs.

The programs are run in an adult-guided environment, organized by professional youth workers who oversee the daily activities and make sure that the rules are followed. Sometimes the adolescents organize their own events, play their own music, put on their own shows, but under the supervision of youth workers. No alcohol is allowed and smoking is forbidden. The youth workers are knowledgeable about young people, and about drug prevention.

METHOD

The data are derived from a national survey of Icelandic adolescents conducted by the Icelandic Institute for Educational Research (Thorlindsson et al., 1998). The original sample consisted of all students born in 1981 and 1982 (15 and 16 years old), attending the compulsory ninth and tenth grade of the Icelandic secondary school. Anonymous questionnaires were administered to all students present in class on one day in March, 1997. Two versions of the questionnaires were distributed, both of which contained the same core questions, but each focused in more detail on different topics. By random assignment, the respondents were given one of two versions of the questionnaires. Questionnaires were administered by teachers and research assistants (for methodological considerations, see Bjarnason, 1995). Students who were not in school on the day of the survey were not included. Valid questionnaires were obtained from 91% of all 7,785 students in these two cohorts. Hence the attrition rate from the original sample was only 9%. Only one of the versions contained the variables of interest in this study, which were answered by approximately half the respondents--3,913, of whom 48% were females. After a listwise deletion of missing values we had 3,431 cases for analysis.

Measures

Leisure activity patterns. Subjects were asked the frequency with which they participate in various types of leisure activities. Principal components factor analysis was used to confirm that the interrelations among various types of leisure activities reflected three underlying constructs: sport, party lifestyle, and involvement in social clubs. Standardized factor scores were used to create uncorrelated indecies of these three leisure activity patterns. Table 1 reports eigenvalues and factor loadings for the items forming the three factors. The internal reliability coefficients (Cronbach's alpha) for sports, party lifestyle, and clubs were .80, .68, and .60, respectively.

Peer substance use. Subjects were asked two questions about alcohol use among their friends: "How many of your friends drink alcohol?" "How many of your friends are drunk at least once a month?" Responses ranged from "1" (none) to "5" (all of them). We created a scale for peer alcohol use by averaging the answers to these two questions (Cronbach's alpha = .92). Subjects were also asked, "How many of your friends smoke hashish?" Responses also ranged from one to five.

Positive peer attitude toward substance use. An alternative measure of peer pressure is the perceived level of positive attitude toward substance use in the peer group. Subjects were asked: "How would your friends react if you would get drunk?" and "How would your friends react if you would smoke hashish?" The response categories ranged from "1" (they would be very opposed to) to "5" (they would very much agree).

Social bonding to family and school. The following statements measure school bonding: "I feel my studies have no meaning," "I'm bored in school," "I feel Sad in school," "I want to quit school," "I don't get along with my teachers," "I want to go to another school," "I am not prepared for classes," "I do not engage in my studies." Response categories ranged from "1" (almost always applies to me) to "5" (almost never applies to me). The answer-scores to these eight questions were averaged to create a scale (Cronbach's alpha = .86). Parental support was measured by averaging the answers to the following questions (Cronbach's alpha = .83); "How easy or difficult would it be for you to get the following from your parents?" (1) warmth and caring. (2) Discussion about personal matters. (3) Advice regarding school. (4) Advice regarding other issues. (5) Help with various things. The answer categories ranged from "1" (very difficult) to "4" (very easy).

Substance use. We created an indicator of alcohol drinking by averaging answers to two questions (Cronbach's alpha = .92): "How many times did you drink alcohol of any kind during the last 30 days?" and "How many times did you get drunk during the last 30 days?" The response categories ranged from "1" (never) to "7" (40 times or more). Subjects were asked one question about lifetime hashish use, also ranging from "1" (never) to "7" (40 times or more).

RESULTS

Table 2 reports the range of values, sample means, and standard deviations for the variables used in the present study.

Table 3 presents multiple regression models where alcohol drinking is regressed on different sets of independent variables. Model 1 examines the simultaneous effects of the three leisure patterns on drinking without controls. As we have hypothesized, the leisure patterns are all significantly related to alcohol drinking. As expected, involvement in sports has a significant, albeit modest, negative effect on drinking. Also, involvement in social clubs has a significant, negative effect on drinking, although quite weak. In contrast, involvement in party subculture has a strong, positive effect on drinking. Jointly, the three leisure patterns explain about 34% of the total variation in alcohol drinking in the present sample. Controlling for gender and social bonding to parents and the school in Model 2 reduces the effect of leisure patterns on drinking somewhat, but the effects of the leisure patterns observed in Model 1 hold in the main.

We have argued that involvement in leisure activities such as sport and organized social clubs should moderate the influence of contact with peers who drink. In Models 3 and 4 we tested this hypothesis by adding multiplicative product terms for the proposed conditional effects of peer drinking and positive peer attitudes toward drinking behavior, respectively. The results in Model 3 show that the association between peer drinking and drinking behavior is in fact significantly contingent on all three lifestyle indicators, net of controls. Thus, as predicted, the relationship between peer drinking and drinking behavior is significantly proportional to involvement in party subculture; the relationship is stronger as adolescents are more involved in this leisure pattern. By contrast, the effect of peer drinking on an individual's drinking is significantly reduced as adolescents are more involved in the other two leisure patterns--sports and social clubs. Model 4 reveals a similar substantive pattern of contingent relationships. The positive effect of positive peer attitudes toward drinking is significantly moderated by leisure patterns; the effect is stronger with increased involvement in the party subculture but weaker with increased involvement in sport and social clubs.

The contingent effects of peer behavior and attitudes on drinking are illustrated in Table 4. The sample is divided into three roughly equal subgroups based on their involvement in different leisure patterns. Considering sports first, the table shows that the unstandardized coefficient for the effect of peer drinking on drinking is .56 for the group that has low involvement in sports but is reduced to .42 in the group that has high involvement in sports. Similarly the coefficient for the effect of positive peer attitude is about 24% lower in the group that has little involvement in sports compared to the high involvement group. Similar findings are reported for social club activity. Finally, the coefficient for the effect of peer drinking is more than three times larger among those who have high involvement in party subculture compared to those that have low involvement. The coefficient for the effect of positive peer attitudes is more than twice as high as the former group.

In Tables 5 and 6 we replicated the analysis for hashish use. As before, the three leisure patterns all have significant statistical effects on hashish use. Model 1 shows that involvement in sport and social clubs have significant negative effects on hashish use while party lifestyle has a positive effect. Jointly the leisure patterns explain about 10% of the total variation in hashish use in the present sample. Model 2 shows that the effects of the leisure patterns are marginally reduced when the control variables are added to the equation.

Moreover, Models 3 and 4 confirm that leisure pattern significantly conditions the effect of peer hashish use attitudes on an individual's hashish use; peer influence decreases with rising involvement in sport and social clubs, but increases with rising involvement in the party lifestyle. Again, the strength of the interaction effects is illustrated in Table 6.

DISCUSSION AND CONCLUSION

Our findings highlight the fact that peer groups formed on the basis of leisure activity can function as agents that deter alcohol and drug use even if the individual has close contact with drug-using peers. The findings reveal that adolescents who engage in leisure activities such as sports and organized club activity are less likely to use alcohol and drugs. More importantly, the findings indicate that the influence of associating with alcohol- and drug-using peers on substance use becomes significantly weaker with greater involvement in sports and clubs. These findings hold when other important factors related to family and school are controlled for.

These findings underline the importance of leisure activity for adolescents in terms of prevention in contemporary society. Both our findings and the wider social context strongly suggest that promoting youth sport and organized leisure as alternatives to association with alcohol- and drug-using peer groups should help reduce alcohol and drug use among adolescents. Sport provides a particularly interesting avenue for preventive work. A majority of adolescents at one time or another participate in organized sport. It is highly structured and is practiced in an institutionalized context in clubs or in relation to school. The existing sport networks among youth are quite large and could be activated for prevention purposes. Given the importance of sport in the life of young people, it is somewhat surprising how little attention it has received in the effort to reduce adolescents' alcohol and substance use. Alcohol producers have certainly recognized the opportunities for influencing adolescents through sport. They have targeted sport events, promoting the sport and advertising beer and liquor at the same time.

As the adolescent society becomes more independent from the adult society of parents and school, it becomes easier for market forces to influence adolescents directly. They can create a youth culture that promotes fun and consumption in one form or another. The entertainment industry is not very likely to promote parental or educational goals unless it benefits them directly. They are, in fact, more likely to encourage the emergence of subterranean values among adolescents. A more salient international and commercially driven youth culture that is becoming more independent of the local adult society, combined with the fact that adolescents already spend more time with their peers than they do with parents, is a trend that preventive work needs to take seriously.

On a more general level these findings strongly indicate the importance for anyone interested in the welfare of young people to pay close attention to adolescent peer culture and lifestyle. If the present trend continues, adolescent leisure activities and peer culture will probably be even more important in the near future. The weakening of social control leaves today's youth more free to seek excitement and fun, form subcultures, and develop lifestyles which are in contrast to the dominating adult culture. The emergence of international, commercially driven popular youth cultures makes the peer group particularly salient in the life of young people. Social changes that have occurred in the last fifty years have shifted the balance of power from traditional institutions such as the family, church, and school to the market (Messner & Rosenfeld, 1994). The entertainment industry, heavily geared toward peers, offers adolescents more freedom to choose lifestyles, values, and norms outside the direct guidance and control of family and school, strengthening the peer group's position at the center of the adolescent life-world. It provides global markets that have been heavily targeted by the entertainment and fashion industry that promotes the culture of fun and excitement and fosters unrealistic expectations.

As peers become more autonomous and influential, parents and educators become more concerned about the power of the peer group in promoting deviant and risk-taking behaviors. At the same time, attention should be turned toward the diversity that exists in peer groups. It seems inadequate to deal with the wide range of leisure activities within this age group as if it represented homogeneous peer cultures and lifestyles.

REFERENCES

Akers, R. L. (1977). Deviant behavior: A social learning approach. Belmont, CA: Wadsworth Pub. Co.

Bartko, W. T., & Eccles, J. S. (2003). Adolescent participation in structured and unstructured activities: A person-oriented analysis. Journal of Youth and Adolescence, 32, 233-241.

Bernburg, J. G., & Thorlindsson, T. (2001). Routine activities in social context: A closer look at the role of opportunity in deviant behavior. Justice Quarterly, 18, 543-567.

Bjarnason, T. (2000). Grooming for success? The impact of adolescent society on early intergenerational social mobility. Journal of Family and Economic Issues, 21, 319-342.

Bourdieu, P., & Passeron, J. (1977). Reproduction in education, society and culture. Beverly Hills, CA: Sage Publications.

Brower, J. (1979). The professionalization of organized youth sport: Social psychological impacts and outcomes. Annals of the American Academy of Political and Social Science, 445, 39-46.

Buhrman, H. G. (1977). Athletics and deviance: An examination of the relationship between athletic participation and deviant behavior of high school girls. Review of Sport and Leisure, 2, 17-35.

Campbell, E. (1969). Adolescent socialization. In D. A. Coslin (Ed.), Handbook of socialization theory and research. (pp. 861-884). Chicago, IL: Rand McNally.

Csikszentmihalyi, M., & Larson, R. (1984). Being adolescent. New York, NY: Basic Books.

Donnelly, P., & Young, K. M. (1988). The construction and confirmation of identity in sport subcultures. Sociology of Sport Journal, 5, 223-240.

Dworkin, J. B., Larson, R., & Hansen, D. (2003). Adolescents' accounts of growth experiences in youth activities. Journal of Youth and Adolescence, 32, 17-26.

Eccles, J. S., & Barber, B. L. (1999). Student council, volunteering, basketball, or marching band: What kind of extracurricular involvement matters? Journal of Adolescent Research, 14, 10-43.

Elliott, D., Huizinga, D., & Ageton, S. (1985). Explaining delinquency and drug use. Beverly Hill, CA: Sage.

Fejgin, N. (1994). Participation in high school competitive sports: A subversion of school mission or contribution to academic goals? Sociology of Sport Journal, 11, 211-230.

Fine, G. A., & Kleinman, S. (1979). Rethinking subculture: An interactionist analysis. American Journal of Sociology, 85, 1-20.

Hagan, J. (1991). Destiny and drift: Subcultural preferences, status attainments, and the risk and rewards of youth. American Sociological Review, 56, 567-582.

Hagan, J. (1997). Defiance and despair: Subcultural and structural linkages between delinquency and despair in the life course. Social Forces, 76, 119-134.

Hagan, J., Hefler, G., Classen, G., Boehnke, K., & Merkens, H. (1998). Subterranean sources of subcultural delinquency beyond the American dream. Criminology, 36, 309-341.

Hastad, D. N., Segrave, J. O., Pangrazi, R. P., & Petersen, G. (1984). Youth sport participation and deviant behavior. Sociology of Sport Journal, I, 366-373.

Hawdon, J. E. (1996). Deviant lifestyles: The social control of routine activities. Youth & Society, 28, 162-188.

Hirschi, T. (1967). Courses of Delinquency. Berkley, CA: University of California Press.

Klein, A. (1986). Pumping irony: Crisis and contradiction in body building. Sociology of Sport Journal, 3, 112-133.

Larson, R. W. (2000). Towards a psychology of positive youth development. American Psychologist, 55, 170-183.

Matza, D. (1964). Delinquency and drift. New York, NY: John Wiley & Sons.

Matza, D., & Sykes, G. (1961). Juvenile delinquency and subterrranean values. American Sociological Review, 26, 12-20.

Maxwell, K. A. (2002). The role of peer influence across adolescent risk behaviors. Journal of Youth and Adolescence, 31, 267-277.

McCaul, K., Baker, J., & Yardley, J. K. (2004). Predicting substance use from physical activity intensity in adolescents. Pediatric Exercise Science, 16, 277-289.

Measham, F., Parker, H., & Aldrige, J. (1998). The teenage transition: From adolescent recreational drug use to the young adult dance culture in Britain in the mid-1990s. Journal of Drug Use, 28, 9-32.

Messner, S. F., & Rosenfeld, R. (1994). Crime and the American dream (second edition). Belmont, CA: Wadsworth Publishing.

Morgan, S. L., & Sorenson, A. B. (1999). Parental networks, social closure, and mathematics learning: A test of Coleman's social capital explanation of school effects. American Sociological Review, 64, 661-682.

Osgood, D. W., Wilson, J. K., O'Malley, P. M., Bachman, J. G., & Johnston, J. L. (1996). Routine activities and individual deviant behavior. American Sociological Review, 61, 635-655.

Savin-Williams, R. C., & Berent, T. (1990). Friendship and peer relations. In S. S. Feldman & G. K. Elliott (Eds.), At the threshold: The developing adolescent (pp. 277-307). New Cambridge, MA: Harvard University Press.

Schafer, W. (1969). Some sources and consequences of interscholastic athletics: The case of participation and delinquency. International Review of Sport Sociology, 4, 63-79.

Snyder, E. (1994). Interpretations and explanations of deviance among college athletes: A case study. Sociology of Sport Journal, 11, 231-249.

Snyder, E. E., & Spreitzer, E. A. (1976). Correlates of sport participation among adolescent girls. Research Quarterly, 47, 804-809.

Snyder, E. E. (1983). Social aspects of sport (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall.

Sutherland, E. H. (1947). Principles of criminology (4th ed). Chicago, IL: Lippincott.

Thorlindsson, T. (1989). Sport participation, smoking, and drug and alcohol use among Icelandic youth. Sociology of Sport Journal, 6, 136-143.

Thorlindsson, T., & Bernburg, J. G. (2004). Durkheim's theory of social order and deviance: A multi-level test. European Sociological Review, 20, 271-285.

Thorlindsson, T., Sigfusdottir, I. D., Bernburg, J. G., & Halldorsson, V. (1998). The social context of drug use among Icelandic youth. Reykjavik: Institute for Educational Research.

Thorlindsson, T., & Vilhjalmsson, R. (1997). Factors related to cigarette smoking and alcohol use among adolescents. Adolescence, 26, 399-418.

Youniss, J., McLellan, J., Su, Y., & Yates, M. (1999). The role of community service in identity development: Normative, unconventional, and deviant orientations. Journal of Adolescent Research, 14, 248-262.

Veblen, T. (1899/1994). The theory of the leisure class. New York: Penguin Books.

Warr, M. (1993). Parents, peers and delinquency. Social Forces, 72, 247-264.

Warr, M., & Stafford, M. (1991). The influence of delinquent peers: What they think or what they do? Criminology, 29, 851-866.

The authors gratefully acknowledge the support of the Department of Sociology and Anthropology at North Carolina State University where the first author was a visiting professor during the writing of this paper.

Jon Gunnar Bernburg, Adjunct Professor, University of Iceland.

Send reprint requests to Thorolfur Thorlindsson, Faculty of Social Science, University of Iceland, 101 Reykjavik, Iceland. E-mail: [e-mail]

Table 1
Principal Components Factor Analysis for Leisure Activities

                                                      Factor Loading
Leisure Pattern                                    (Varimax Rotation)

Sport (Eigenvalue = 2.6)
Participation in sports and body training               .89
  outside of school gymnastics (a)
Participation in sport clubs (a)                        .87
Sports or outdoor activity with friends (a)             .80

Party (Eigenvalue = 2.2)
Being downtown in the evenings or on                    .77
  weekends (b)
Going to parties (b)                                    .74
Hanging out at malls or in candy shops (b)              .72
Playing computer games in game halls (b)                .55

Social Club (Eigenvalue = 1.5)
Participation in organized social activities            .73
  within the school (b)
Participation in organized youth clubs (b)              .69
Participation in leisure clubs organized by the         .68
  school (b)
Going to a youth center (b)                             .53 (a)

(A) Answer categories range from 1 (never) to 6 (almost every day).

(b) Answer categories range from 1 (almost never) to 5 (four times a
week or more).

Table 2
Descriptive Statistics

                               Range
                             of Values
                                                    Standard
Variable                   Min.     Max.    Mean    Deviation

Female = 1 (Male = 0)        0       1       .48       .50

Leisure Patterns
Sport                      -2.07    1.82     0          1
Social clubs               -2.06    3.13     0          1
Party                      -1.87    4.24     0          1

Social Bonds
Parental support             1       4      3.36       .59
Parental monitoring          1       4      2.45       .71
School commitment            1       5      3.59       .69

Peers
Peer drinking                1       5      2.56      1.28
Positive peer attitudes      1       5      2.92      1.28
  toward drinking
Peer hashish use             1       5      1.38       .74
Positive peer attitudes      1       5      1.54       .97
  toward hashish use

Substance Use
Alcohol use                  1       7      2.09      1.40
Lifetime hashish use         1       7      1.27       .96

Table 3
Alcohol Drinking Regressed on Indicators of Lifestyle, Conventional
Social Bonds, and Peer Drinking: Additive and Nonadditive Models

Independent Variables           Model 1              Model 2

Female                       --         --       .21 **     (.08)

Leisure Patterns
Sport                      -.27 **    (-.20)    -.19 **    (-.13)
Social clubs               -.08 **    (-.06)    -.04 *     (-.06)
Party                       .75 **     (.54)     .64 **     (.46)

Social Bonds
Parental support             --         --       .03        (.01)
Parental monitoring          --         --       .14 **     (.07)
School commitment            --         --      -.49 **    (-.24)

Peers
Peer drinking                --         --        --         --
Positive peer attitudes      --         --        --         --
  toward drinking

Interaction Terms
Peer drinking * Sports       --         --        --         --
Peer drinking * Party        --         --        --         --
Peer drinking * Clubs        --         --        --         --
Peer attitudes * Sports      --         --        --         --
Peer attitudes * Party       --         --        --         --
Peer attitudes * Clubs       --         --        --         --

[R.sup.2]                   .34                  .39

Independent Variables           Model 3              Model 4

Female                      .09 **     (.03)     .20 **     (.07)

Leisure Patterns
Sport                       .07 *      (.05)     .07        (.05)
Social clubs                .10 **     (.07)     .08 *      (.06)
Party                      -.05       (-.04)     .12 **     (.09)

Social Bonds
Parental support            .02        (.01)     .02        (.01)
Parental monitoring         .11 **     (.05)     .10 **     (.05)
School commitment          -.29 **    (-.14)    -.36 **    (-.18)

Peers
Peer drinking               .48 **     (.44)      --         --
Positive peer attitudes      --         --       .37 **     (.34)
  toward drinking

Interaction Terms
Peer drinking * Sports     -.08 **    (-.17)      --         --
Peer drinking * Party       .15 **     (.34)      --         --
Peer drinking * Clubs      -.05 **    (-.11)      --         --
Peer attitudes * Sports      --         --      -.07 **    (-.16)
Peer attitudes * Party       --         --       .12 **     (.28)
Peer attitudes * Clubs       --         --      -.04 **    (-.09)

[R.sup.2]                   .57                  .50

Note. The table reports unstandardized coefficients (standardized
coefficients in parentheses) from ordinary least squares regression
analysis. * p < .05, ** p < .01 (two-tailed tests).

Table 4
Regressing Drinking Behavior on Peer Drinking
and Positive Peer Attitudes for Different Subgroups

                                                  Unstandardized
                Level of      Unstandardized     Coefficient (SE)
               Involvement   Coefficient (SE)    for the Effect of
Leisure        in Leisure    for the Effect of     Positive Peer
Pattern          Pattern       Peer Drinking         Attitudes

Sport
Group 1            Low           .56 (.03)           .41 (.03)
Group 2          Medium          .49 (.03)           .38 (.03)
Group 3           High           .42 (.02)           .31 (.02)

Social Clubs
Group 1            Low           .52 (.03)           .38 (.03)
Group 2          Medium          .52 (.03)           .39 (.03)
Group 3           High           .43 (.03)           .33 (.03)

Party
Group 1            Low           .26 (.02)           .22 (.02)
Group 2          Medium          .54 (.02)           .43 (.03)
Group 3           High           .67 (.03)           .48 (.03)

Note. All reported coefficients are statistically significant
(p < .01). Drinking behavior is regressed on peer drinking and
positive peer attitudes separately. The following variables are
controlled in the equations: female, family support, attachment
to school, family commitment, and leisure patterns.

Table 5
Lifetime Hashish Use Regressed on Indicators of Lifestyle,
Conventional Social Bonds, and Peer Drinking: Additive and
Nonadditive Models

Independent Variables           Model 1              Model 2

Female                       --         --      -.06       (-.03)

Leisure Patterns
Sport                      -.14 **    (-.15)    -.11 **    (-.12)
Social clubs               -.07 **    (-.08)    -.04 *     (-.04)
Party                       .27 **     (.28)     .21 **     (.22)

Social Bonds
Parental support             --         --      -.04       (-.02)
Parental monitoring          --         --       .03        (.02)
School commitment            --         --      -.23 **    (-.17)

Peers
Peer drinking                --         --        --         --
Positive peer attitudes      --         --        --         --
  toward drinking

Interaction Terms
Peer drinking * Sports       --         --        --         --
Peer drinking * Party        --         --        --         --
Peer drinking * Clubs        --         --        --         --
Peer attitudes * Sports      --         --        --         --
Peer attitudes * Party       --         --        --         --
Peer attitudes * Clubs       --         --        --         --
[R.sup.2]                   .10                  .13

Independent Variables           Model 3              Model 4

Female                     -.08 **    (-.04)    .01         (.01)

Leisure Patterns
Sport                       .18 **     (.19)     .08 **     (.08)
Social clubs                .06 *      (.06)     .04        (.04)
Party                      -.18 **    (-.19)    -.14 **    (-.14)

Social Bonds
Parental support           -.01       (-.01)     .00        (.00)
Parental monitoring         .02        (.02)     .01        (.01)
School commitment          -.11 **    (-.08)    -.15 **    (-.11)

Peers
Peer drinking               .50 **     (.39)      --         --
Positive peer attitudes      --         --       .23 **     (.23)
  toward drinking

Interaction Terms
Peer drinking * Sports     -.17 **    (-.29)      --         --
Peer drinking * Party       .18 **     (.33)      --         --
Peer drinking * Clubs      -.04 *     (-.07)      --         --
Peer attitudes * Sports      --         --      -.10 **    (-.19)
Peer attitudes * Party       --         --       .18 **     (.38)
Peer attitudes * Clubs       --         --      -.03 *     (-.06)
[R.sup.2]                  .40                   .27

Note. The table reports unstandardized coefficients (standardized
coefficients in parentheses) from ordinary least squares regression
analysis. * p < .05, ** p < .01 (two-tailed tests).

Table 6
Regressing Hashish Use on Peer Hashish Use
and Positive Peer Attitudes for Different Subgroups

                                                  Unstandardized
                Level of      Unstandardized     Coefficient (SE)
               Involvement   Coefficient (SE)    for the Effect of
Leisure        in Leisure    for the Effect of     Positive Peer
Pattern          Pattern     Peer Hashish Use        Attitudes

Sport
Group 1            Low           .85 (.04)           .36 (.03)
Group 2          Medium          .67 (.03)           .36 (.03)
Group 3           High           .40 (.03)           .23 (.02)

Social Clubs
Group 1            Low           .75 (.04)           .34 (.03)
Group 2          Medium          .70 (.03)           .34 (.03)
Group 3           High           .56 (.03)           .29 (.03)

Party
Group 1            Low           .51 (.03)           .15 (.02)
Group 2          Medium          .53 (.03)           .23 (.02)
Group 3           High           .79 (.04)           .44 (.03)

Note. All reported coefficients are statistically significant
(p < .01). Hashish use is regressed on peer hashish use and
positive peer attitudes separately. The following variables
are controlled in the equations: female, family support,
attachment to school, family commitment, and leisure patterns.