Wednesday, May 28, 2008

The role of peers

The interesting difference between the role of parents with the role of peers is that parents are almost exclusively conceptualized as barriers to criminal activity, be it indirectly through the development of self-control via parental socialization (Gottfredson and Hirschi, 1990) or directly via social control (Hirschi, 1969?) or monitoring (i.e., Osgood and Anderson, 2004?), while peers are considered a motivating influence that encourages involvement in criminal or otherwise deviant activities.

The importance of peers first came into consideration with Sutherland's (1947) theory of differential association. Differential association argues that people become criminal through their associations (i.e., friendships) with other criminals. Sutherland argues that these relationships will provide the individual with an excess of delinquent or antisocial definitions; in other words, these relationships will cause the individual to develop the same antisocial attitude as their friends. It is through these relationships that individuals learn the techniques necessary to engage in criminal activities as well as the proper reasons for doing so. Sutherland notes that these relationships vary in frequency of contact, duration, priority, and intensity, so merely knowing someone who is engaged in delinquent or deviant behavior does not automatically mean that the individual in question will also become delinquent.

Sutherland's theory was expanded upon by Burgess and Akers (1966), who incorporated the notion of differential association into their concept of social learning theory. Social learning theory is based around four central concepts: differential association, differential reinforcement, definitions, and imitation. Differential association is used here similar to how Sutherland (1947) originally conceptualized it: it is association with delinquent peers leading to an excess of definitions of behavior favoring delinquency; it is an environment where deviant behavior is normative and prosocial behavior non-normative. Differential reinforcement is the balance between the anticipated rewards and consequences of a particular act and the actual rewards and consequences; this suggests that a lack of suitable punishments will reinforce the individuals delinquent identity. This is related to the third concept, definitions. Here, the authors argue that the individual attaches general and specific meanings to the behavior they are engaging in, both from a religious or moral standpoint (general) and a specific standpoint (this may also be considered a symbolic interactionist definition, such that the individual is attaching specific meanings to these behaviors ((Katz, ??)). Finally, the individual must also imitate the behaviors they are exposed to in order to become delinquent themselves. While these propositions may seem intuitive, they do expand about Sutherland's original conceptualization in important ways -- namely, Sutherland argues that merely making criminal behavior normative will cause individuals to engage in these acts. Burgess and Akers include the actual process of becoming a delinquent that Sutherland takes for granted.

Naturally, social learning theory, differential association, and the general perspective that peers influence criminal behavior has evolved since Burgess and Akers first introduced social learning theory in 1966. Recent research on the role of peers has taken three distinct approaches: (1) an assessment of demographic differences, (2) an expansion of the definition of "friends" to include romantic peers, "best friends", and siblings, and (3) reassessing the idea of differential definitions of delinquent behavior.

The work of Haynie and Payne (2006) and Mears et al. (1998) are recent examples of research that has focused on the importance of race and/or gender in regard to the relationship between peer associations and crime. Haynie and Payne (2006) focus explicitly on the ability of friendship network composition to explain racial differences in participation or involvement in violence. The authors also show significant differences between white, black, and Hispanic youth in regard to the effect of friendship network characteristics on propensity for violence. Perhaps most notably, these authors find that network heterogeneity is negatively associated with involvement in violence for black adolescents, but positively associated with involvement in violence for white adolescents.

Gender, specifically the gender gap in involvement in delinquency, is examined by Mears et al. (1998). These authors argue that men and women are effected differently by exposure to similar criminogenic conditions, in this case, association with delinquent peers. Their analysis supports their hypothesis that the gender gap in delinquency exists partially because of men and women attach different moral evaluations to delinquency, in this case, the behavior of their peers. At the very least, these findings qualify the basic premise of social learning theory and differential association; being surrounded by delinquent peers does not have the same effect for all persons, above and beyond Sutherland's (1947) allowing for variation based on frequency of contact, intensity, duration, and priority.

The expansion of the definition of "friends" is another part of differential association and social learning theory that has come into consideration in recent years. Specifically, recent research has looked at the way relationships with romantic partners (Haynie et al., 2005) and siblings (Haynie and Mchugh, 2003) influence involvement in delinquent activity as well as whether network placement, specifically the difference between friends and best friends, impacts involvement in delinquency (Weerman and Smeenk, 2005). Both Haynie et al. (2005) and Haynie and Mchugh (2003) expand upon the traditional conceptualization of friends to include other peers that adolescents normally and reasonably spend time with in varying capacities. Not surprisingly, both sets of authors find that relationships with romantic partners and siblings during adolescence can contribute to the individual's overall participation in delinquency; Haynie et al. (2005) show that romantic partners' delinquency has a unique effect on the individual's participation, net of the roles of their other friends. While romantic partners have more of a direct influence on participation in delinquency, Haynie and Mchugh (2003) show that siblings have an indirect effect on involvement in delinquency. Here, the authors show that siblings have an indirect effect on delinquency via their own friends; that is, if you have a sibling who has delinquent friends, you may be more likely to become delinquent as well via your knowledge of and interactions with this third party. A third approach is taken by Weeman and Smeenk (2005), who ask whether network placement -- i.e., friends versus best friends -- makes a difference in the likelihood that an individual will become involved in delinquency. Here, contrary to what may seem to be an intuitive relationship, these authors find no support for the idea that best friends will have a stronger impact on participation in delinquency than ordinary friends; there is no difference whatsoever. This further complicates the role of peers; romantic partners (Haynie et al., 2005) have a direct impact, siblings have an indirect impact (Haynie and Mchugh, 2003), but best friends do not play a special part.

Finally, a third area of research on the role of peers has focused on reassessing the idea of "definitions" first discussed by Burgess and Akers (1966). This approach, exemplified by the work of Rebellon (2006) and Kreager (2007), argues that delinquency may be a means to an end for some adolescents, instead of being the end result from their existing relationships. Specifically, both authors argue that in some cases, involvement in delinquency is a way for adolescents to attract positively-valued attention from their peers. Of course, this does not apply to all adolescents; Kreager (2007) finds that it is most applicable among young males with low levels of education. Rebellon (2006) also shows that engaging in delinquency to attract peers doesn't necessarily work the way hypothesized by Kreager (2007); here, Rebellon finds no support for a direct effect, but he does find what Burgess and Akers (1966) might call an "imitation" factor, in that those persons the individual is trying to attract socially are more likely to engage in delinquency themselves, indirectly reinforcing the individual's idea that these types of behaviors will make them popular. That being said, participation in delinquency is not always a means of attracting peers; Kreager (?) shows that isolated individuals who have problematic encounters with their peers are more likely to become involved in delinquency. In this case, delinquency is the byproduct of negative relationships with peers, rather than something that is a result of positive relationships with them.

Of course, while research on the peer effect does show that peers influence delinquency in a variety of ways, Haynie and Osgood (?) caution us about taking the peer effect too far. Their research indicates that the effect of antisocial behavior becoming normative through peer interactions is not as significant as differential association (Sutherland, 1947) would have us believe; that peer interactions do not fully account for the importance of opportunity; and that peer influence does not mediate other influences, including those of age, gender, family, or school. Therefore, the argument can be made that the role of peers is no different from the role of parents -- this group does have a real influence on involvement in delinquency, but not to the extent that they were originally believed.

No comments: