Linking School Performance and Delinquency
A great deal of scientific research examines the relationship between poor school performance and delinquency. The direction of the causal link between education and juvenile delinquency is fundamentally complex. Early aggressive behavior may lead to difficulties in the classroom. Such difficulties, in turn, may result in a child’s receiving unfavorable evaluations from teachers or peers. These, in turn, might result in delinquency. Equally, delinquency could be another manifestation of whatever characteristics got the child into trouble with school authorities in the first place.
Some studies have shown reductions in delinquent behavior when a teenager drops out of school. Others have shown increasing rates of delinquency following school dropout. In addition, many studies have shown that family and child characteristics predict both problems in school and an increased likelihood of delinquent behavior.
Despite the ongoing discussion of the direction of causality, the evidence is clear that poor school performance, truancy, and leaving school at a young age are connected to juvenile delinquency (Bachman et al., 1971; Elliott, 1978; Elliott and Voss, 1974; Farrington, 1986; Hagan and McCarthy, 1997; Hawkins et al., 1998; Huisinga and Jakob-Chien, 1998; Kelly and Balch, 1971; Maguin and Loeber, 1996; Mensch and Kandel, 1988; Polk, 1975; Rhodes and Reiss, 1969; Simons et al., 1991; Thornberry et al., 1984). Several factors linked to delinquency, aggression, and violence have been identified. For example, research has found that verbal and reading deficits are linked to victimization (both inside and outside
school), drug use, aggression, and delinquent behavior when students who fall behind in reading become marginalized as failures (Kingery et al., 1996). School failure undermines a student’s interest in and commitment to school and learning. Delinquent peer associations may also be a consequence of school failure when a student comes to reject academic achievement and prosocial behavior as legitimate goals and values. Feelings of isolation and a student’s perception that she is not receiving emotional support from caring adults also may play a role in the etiology of delinquent or aggressive behaviors (Gottfredson, 1997). Research has identified other factors at the community, family, and individual levels that influence the development of delinquent and/or aggressive behaviors, including the availability of criminogenic tools (e.g., weapons), community disorganization, family history of problem behavior, family conflict, and a history of early antisocial behavior (Howell, 1995).
Rolf Loeber, of the University of Pittsburgh Medical Center, Western Psychiatric Institute and Clinic, cautioned that the relationship between delinquency and school performance should not be oversimplified. It may be that progression from delinquent behavior to school failure is contingent on other factors, since not every offending juvenile experiences school failure and not every failing student commits offenses.
In addition, not every act of delinquency affects school performance in the same way. The seriousness of delinquent behavior may determine whether and to what extent school performance suffers. It appears that poor school performance is a more severe problem among serious violent delinquents. In a review of the literature on the predictors of youth violence, Hawkins and his colleagues (1998) concluded that serious and violent delinquents had more school-related problems (e.g., low grades, truancy, suspension, dropping out) than nonviolent children.
Inversely, studies have found that students who do not perform well academically are more likely to be delinquent. The Cambridge Study on Delinquent Development and the Pittsburgh Youth Study have both found that low school achievement predicts adolescent delinquency (Maguin and Loeber, 1996). In a meta-analysis of studies that examined the relationship between academic performance and delinquency and interventions designed to improve school achievement and reduce offending, Maguin and Loeber (1996) found that poor school performance was related to the frequency of delinquent offending, the seriousness of offenses, and persistence in delinquency offending. Findings from this study highlight the importance of examining the effect of poor educational performance on delinquency over time—to think of a child’s development on a trajectory with multiple transition points (e.g., childhood to adolescence) along which key events occur.
There are, however, methodological issues that limit study findings.
Loeber noted that while findings applied equally to boys and girls in some studies, because most studies are conducted with boys, the findings may not in fact be generalizable to the experiences of girls.
While time limitations did not allow for an exhaustive review of the relevant research at the workshop, participants were able to discuss the important role that peers play in the relationship between delinquency and poor school performance. That peers exercise influence on the development of delinquent behavior is a common perception among researchers. Workshop participants discussed three issues related to the effects of peers on delinquency: delinquent peer conversations, peer rejection, and unintended negative effects of grouping high-risk youth together for services or programs.
DELINQUENT PEER ASSOCIATIONS
Studies have found evidence of negative effects attributed to deviant peer associations (Gottfredson, 1987). Many schools include programs designed to improve children’s social behavior. Guided counseling programs, for example, have been mandated in some states. These programs are often administered to students in groups.
Thomas Dishion, of the University of Oregon’s Department of Psychology, described the danger of assuming that all intervention programs are benign. As part of a study designed to measure and code interactions among teenage boys assembled to discuss problems in their relationships with parents and peers, Dishion and his colleagues (1999) found that interactions among the boys were influenced by the content of their conversations. Conversation was classified into two categories: rule-breaking talk and norm-accepting talk. Researchers observed that the nonverbal reactions to rule-breaking and norm-accepting topics and activities communicated either positive or negative reinforcement for the associated behavior (Dishion et al., 1996a). Among nondelinquent dyads, normative talk led to positive reinforcement in the form of laughter. Alternatively, in dyads in which the members had some experience with delinquency, normative talk failed to elicit a positive response; only rule-breaking talk received positive feedback.
The researchers concluded from this study that delinquent peer groups are organized around rule-breaking talk (Dishion et al., 1996a). Positive reinforcement for rule-breaking talk is referred to as “deviancy training.” Dishion and his colleagues (1996b) found that, controlling for past behavior, deviancy training observed at ages 13 and 14 predicted an increased probability of escalating addictive substance use, delinquency (self-reported), and police-reported violent behavior in the next two years. These findings have been replicated among delinquent and nondelin-
quent girls. Although adolescent girls differed from adolescent boys in terms of the topics they discussed and the rules they broke, the deviancy training process was similar.
At the workshop, Dishion argued that these findings point toward needed changes in school policy. If it is the case that deviant peers exert a strong influence on the development of delinquent behavior, one way to discourage this is to reintegrate at-risk children and adolescents into the educational mainstream. By doing this, children who would traditionally be grouped together because of problem behavior (or school failure) would benefit from the prosocial influence of peers who exhibit more normative conduct.
Workshop participants also discussed how peer rejection influences delinquency. Research findings in this area are contested, however, and mechanisms through which peer rejection leads to delinquency are not at all clear. Aggression has been suggested to explain the connection between peer rejection and delinquency. Participants noted that it is equally reasonable that aggression leads to peer rejection as it is that peer rejection leads to aggression. While the research proposes a link between peer rejection and aggression (see Coie et al., 1990), it may be that the popular literature overstates the relationship. On closer inspection of this body of research, it appears that only children who are both aggressive and victimized are rejected by their peers. In other words, it may not be aggression that leads to peer rejection. On the contrary, by adolescence many aggressive children are admired, and in some settings delinquents are popular. Furthermore, not all peer-rejected adolescents perceive themselves as being rejected. These observations undermine support for the assertion that peer rejection is causally related to delinquency and aggression (Cairns and Cairns, 1994; Graham and Juvonen, 1998).
Workshop participants were in agreement that while the research may not be able to identify how peer rejection relates to delinquency, peer rejection has a meaningful impact on students’ commitment to school and learning. Peer rejection can occur in many different contexts, some amenable to school intervention, others not. Workshop participants noted that a great deal of peer rejection occurs in classrooms—a context in which teachers have considerable influence. Some teachers do a good job of organizing the classroom environment so that children do not feel rejected. Other teachers do a poor job of controlling peer rejection in their classrooms or, worse yet, encourage it.
POTENTIAL HARM OF GROUPING HIGH-RISK YOUTH
Several times during the course of the workshop participants stressed how programs that aggregate high-risk youth (e.g., anger management classes, alternative schools) should be considered with caution. Even when researchers observe prosocial effects and skill improvement in subjects who participate in these programs, such groups may nonetheless facilitate the formation of deviant peer associations. This can happen even when clinicians are careful that interactions that can lead to negative outcomes do not occur in their presence.
Thomas Dishion, professor of psychology at the University of Oregon, described the danger of assuming that all intervention programs are beneficial. An example of the contrary comes from the Guided Group Counseling Programs, which have been mandated in some states. When evaluated using random assignment to such a program, negative effects were found for high school students (Gottfredson, 1987).
Dishion has also documented this phenomenon in his research. In a study conducted by Dishion and Andrews (1995), young adolescents (ages 11 to 14) and parents received an intervention designed to reduce problem behavior. Participants were placed in one of four groups— teenagers only, parents only, parents and teens, and self-directed —and administered curricula designed to improve communications skills, facilitate better family management, and encourage prosocial behavior. Groups met for 90 minutes each week for 12 weeks, and individuals participating in groups were also visited once by a therapist. Subjects in the self-directed group received curriculum material through the mail and were not visited by a therapist. Findings indicated that while subjects in the teenager-only group experienced a reduction in antisocial behavior over the short run, all delinquency-involved teenagers in the group showed increases in antisocial behavior (e.g., smoking and teacher-reported problem behavior) on follow-up. Adolescents in the self-directed group and those in the mixed parent-teenager group did not show these same negative changes in behavior over time. Dishion and Andrews theorized that the delinquency-involved youth in the teenager-only group received subtle forms of positive feedback (e.g., head turning and attention) and approval for their antisocial behavior and that this may have accounted for the increase.
Other intervention experiments also suggest that peer group interactions may explain some deleterious effects. For example, Feldman (1992) evaluated an intervention that provided group-level behavior modification treatments to two groups, one of all antisocial youth and one in which several antisocial youth were included in a group of prosocial peers. The observed misbehavior of boys in the mixed groups declined,
but that of the boys assigned to unmixed groups did not. In other work, McCord (1992) analyzed the effects of the Cambridge-Somerville Youth Study, which used a matched pair design (that is, each boy in the treatment group was matched to a particular boy in the control group) so that a variety of interventions could be evaluated. Boys in the treatment group were sent to summer camp, about half of them for one summer and the remainder for more than one summer. Those sent to summer camp more than once turned out considerably worse than their matched pairs in terms of crime convictions, early death, alcoholism, and several mental health disorders.
Workshop participants noted that it might be useful for publicity to be given to harmful as well as beneficial effects and that special care is needed in the evaluation of programs that put misbehaving young people together in groups.
Research findings support the existence of common factors that may cause both delinquency and poor school performance. These factors include intelligence and attention deficit hyperactivity disorder (Maguin and Loeber, 1996). According to Maguin and Loeber’s (1996) meta-analysis of intervention programs, the consensus of studies found low intelligence and attention problems to be common causes of both delinquency and poor academic performance. Child and family risk factors, peer group influences, socioeconomic status, low school motivation, and early conduct problems were also causes of school failure and delinquency and, in combination, increased the risk of both.
Studies suggest that changing educational performance and behavior simultaneously is more likely than either alone to result in durable positive outcomes. For example, evaluations of interventions designed to address delinquency and poor academic performance have found that educational programs that teach self-control and social skills and provide parental training (Arbuthnot and Gordon, 1986; Gottfredson, 1990; Tremblay et al., 1992) were more successful in improving education outcomes than those that provide only remedial educational assistance (Maguin and Loeber, 1996). What this suggests is that addressing behavioral and cognitive deficits (i.e., factors that arguably operate as common causes of both delinquency and poor school performance) may do more to improve academic performance and to decrease or prevent delinquency than either providing remedial academic support or imposing punitive criminal sanctions alone. The available evaluations of programs that focus on cognitive or behavioral deficits alone find that the effects are either equivocal or of a short-lived, positive nature (Maguin and Loeber, 1996).
Disabilities may also operate as common factors in the etiology of both poor school performance and delinquency. Such disabilities include language and speech problems, learning disabilities, behavioral problems (e.g., attention deficit hyperactivity disorder), and emotional problems (e.g., severe emotional disturbance) (see Meltzer et al., 1986; Perlmutter, 1987). More research, especially research using longitudinal study designs, is needed to examine how disabilities operate as common causes of both delinquency and poor school performance.
Workshop participants also noted the importance of family risk factors as common to both poor school performance and delinquency. Research has shown that family management skills training can disrupt trajectories toward school failure and crime. By reducing negative family interactions and conflict and improving parental supervision of children and parent-child relationships, family management skills training reduces risk factors associated with delinquency and increases parental monitoring of a child’s activities and school progress (Maguin and Loeber, 1996).
Families not experiencing problems can also benefit from increased parental involvement. Dishion noted how the parents’ role in education and delinquency prevention is often left out in devising prevention and intervention strategies. He suggested the need to reverse this trend and to do more to reengage parents, particularly at important transition points in children’s development. Transition into middle school is an especially critical time.
Workshop participants noted that there are serious structural and motivational barriers to parental involvement in prevention and intervention programs. Poor and working parents may find it difficult to attend meetings consistently. More difficult are the issues that arise in working with crime- and/or drug-involved parents. Not only does the behavior of the parent run counter to program goals and objectives, but work with these parents and children also is more complex and requires a level of services often not available through traditional prevention and intervention programs. The needs of special populations (e.g., those who are poor, single parents or substance abusing) should be anticipated and addressed if prevention and intervention programs are to assist such parents in supporting their children’s development.
Many children exhibit aggressive behavior at an early age, but most do not persist in this behavior as they mature. Researchers have a great deal of difficulty in discriminating between children displaying problem behavior in the preschool period who will desist and those who will become persistent adult offenders. Researchers and policy makers must
keep in mind that the period from early childhood to adolescence is a dynamic one, accompanied by complex and often unpredictable behavior.
Research findings argue against adopting a point of view that portrays delinquency as the result of a discrete event in a child’s life. Loeber pointed out that research consistently shows that individual delinquency is a gradual process. He noted that serious instances of school violence do not erupt without some prior signs of trouble. The most serious incidents of violence occur among individuals who for years have displayed minor forms of aggression, including physical fighting, gang fighting, or frequent arguing and bullying. These findings should inform program design and the selection of interventions. Programs should be targeted and designed for different stages in the escalation process. Loeber asserted that it is naive to think that serious violence can be completely prevented in schools. What school prevention and intervention programs can do well is deal with lower-level forms of acting out (e.g., serious bullying and physical fights), which, if left unchecked may evolve into more serious instances of delinquency and violence.