New directions for research in prevention and treatment of delinquency: a review and proposal.

Sullivan, Richard ; Wilson, Marcia Fenner

INTRODUCTION

The literature describing various approaches to the prevention, control, and treatment of delinquency can be differentiated according to the theories of causation which shaped them; that is, the focus of intervention differs according to the theoretical view of the causes of delinquency. At various times, these theories have emphasized physiological and psychological characteristics of the individual or the structure of the family within the broader social structure itself or within an integrated complex of factors which includes community characteristics.

Over several decades, numerous investigators have found statistically significant relationships between crime and certain inherited and biologically identifiable characteristics such as skull formation, body type, chromosomal abnormalities, and glandular or neurological anomalies (Klein, 1971). One by one these single variable explanations were dropped because they failed to explain the diversity of causal pathways and outcomes among juvenile delinquents.

A contemporary view, including that of the present researchers, favors the concept of an integrated complex of causal factors within which individual, familial, and social structural factors may exert variable influence on a case by case basis. This is a more challenging approach, requiring more rigorous and comprehensive assessment and intervention, but it avoids the limitations and pitfalls of single variable explanations. This article reviews the various approaches which have dominated the field over the last few decades and presents current thinking about the multiple factors which must be considered in research on delinquency; it then concludes with a methodological proposal which may facilitate the systematic consideration of these factors.

Individual and Family Theories

Psychology has been contributing to delinquency research throughout the past century. According to Binder and Binder (1983), the earliest period of exploration emphasized individual psychological characteristics of delinquents, with particular emphasis on intelligence (Goddard, 1915; Burt, 1925; Hathaway & Monachesi, 1957; Conger, 1966). Goddard began by arguing that mental deficiency was the major cause of delinquency. Later developments amended that view to distinguish between deficits in intelligence and specific maladaptive accommodations to learning disabilities. Perlmutter (1987), for example, postulated that delinquent behavior results from the learning disabled student's attempts to compensate for academic failure and frustration.

Social learning theory has contributed the view that social skills deficits are possible causes of delinquent behavior. Long and Sherer (1984) suggest that delinquents behave maladaptively while seeking to attain conventional goals because they lacks the requisite skills to act appropriately.

Besides intellectual and social skills deficits, personality characteristics are often noted as factors. Spergel (1966) describes the root cause of delinquency as a weak ego arising from ineffective or destructive family relationships. These youths are therefore unable to trust and establish productive relationships with adults. They experience low self-esteem, personal conflict, and a high level of anxiety. "Delinquent or predelinquent behavior is ordinarily regarded as neurotic, the effort of a defective ego to 'strike back' at society, i.e., adults and peers, for personal failures. Delinquency is relevant strictly to the individual and the group and the community are only back-drops or peripheral forces contributing to the problem" (Spergel, 1966, p. 55). This perspective turned to the family for an explanation of the conflict.

For some time, researchers and practitioners have assigned a critical role to the family in the development or prevention of delinquency (Tolan, Cromwell, & Brusswell, 1986). based on the tenets of socialization theory, this perspective emphasizes the family's role in helping children adjust to the demands and opportunities of their social environment. If inadequate socialization occurs within the family, delinquency may result (Quay, 1987).

In order to identify the family and parenting variables that influence delinquency, comparisons are sometimes made between the family characteristics of delinquents and nondelinquents (Hetherington, Stouwie, & Ridberg, 1971). Studies using this method typically focus on demographic variables such as family size and composition, social class, and parents' marital and employment status (Carill, Gusmar, & Wolff, 1983). These variables are useful in terms of identifying "at risk" groups but they are less useful for the purposes of formulating intervention strategies because these structural variables are not easy to manipulate.

Variables related to family role functioning also have received attention. In an extensive review of the literature, Rutter and Giller (1984) found that family variables associated with juvenile delinquency included parental criminality, cruelty, passive or neglectful parenting, erratic or harsh discipline, marital conflict, and poor parental supervision. These same variables have been associated with nonadjudicated disturbances of conduct.

Despite these findings, Blomberg and Caraballo (1979) concluded that family variables explain very little of the variance in delinquency. While a number of studies have shown statistically significant relationships between delinquent behavior and family factors, Klein (1971) cautioned that these statistical correlations should not be interpreted as causal relationships. Instead, these variables probably attain their importance through combination with a number of others factors. To consider the factors related to the broader social context beyond the family, we turn to the sociological literature.

Social Strain Theory

The social strain theory of delinquency is based on the idea that delinquency results when individuals are unable to achieve their goals through legitimate channels (Cohen, 1955; Cloward & Ohlin, 1960). Cloward and Ohlin suggest that a democratic ideology espouses equality of opportunity and universally high aspirations for success, but when there is a discrepancy between aspirations and opportunity, delinquent solutions evolve.

Cloward and Ohlin's monograph has been criticized for the dearth of empirical support for its central theoretical tenets. Agnew (1985) reasoned that if strain theory did explain juvenile delinquency, delinquents would have high aspirations for success while perceiving little prospect of its achievement. Empirical measures of these constructs have produced contradictory findings. Delinquency is highest when aspirations are low (Agues, 1985). Again, the causal pathway may be more complex and convoluted. High aspirations may decline in the face of obstacles to opportunity, fostering a feeling of hopelessness and frustration which leads to delinquency.

Social Control Theory

As presented by Hirschi (1969), social control theory argues that delinquency is inherent in human nature and requires no complex set of variable conditions to emerge. Rather, our bonds to conventional society impose normative constraints that prevent us from acting on deviant impulses. Delinquency emerges when these constraints are substantially attentuated. Unlike strain theory which assumes that individuals are positively socialized and resort to deviance only when they are confronted with inconsistencies between their aspirations and opportunities, social control theory treats the socialization process and commitment to conventional norms and values as problematic (Elliot, Ageton, & Canter, 1979). From this perspective, conventional norms of conduct may be inadequately internalized, or there may be conflict or inconsistency in the rules or social controls (Elliot et al., 1979). This conflict derives from inequities in the distribution of power to define normative conventions and related access to the social good.

Differential Association Theory

Differential association theory derives from social learning theory and assumes that there is no natural impulse toward delinquency, but rather delinquent behavior must be learned and reinforced through the same process as conforming behavior (Thornberry, 1987). Since an individual's characteristic patterns of behavior are usually learned and reinforced within a consistent social context, differential association explains differences in learned social behavior. This theory assumes that delinquency is rooted in normative conflict and seeks to explain how conflicting subcultural patterns are learned and reinforced (Matsueda & Heimer, 1987).

Post-industrial societies contain conflicting structures of opportunity, norms, and definitions of appropriate behavior, giving rise to high rates of crime (Matsueda & Heimer, 1987). At the individual level, norms are translated into individual acts of delinquency through differential association wherein definitions favorable and unfavorable to delinquent behavior are learned through communication, primarily in intimate groups. At the group level, normative conflict is translated into group delinquency through differential social organization. The extent to which a group is organized for or against delinquency determines its rate of law violation (Matsueda & Heimer, 1987). Differential association theory avoids the hegemonic assumption of cultural universality with respect to aspirations and values, but in focusing on the process of how differences develop, it falls short of explaining why certain groups develop deviant or delinquent belief systems.

Labeling and Interactionist Theories

Labeling theory looks at social differentiation itself, particularly differential access to power, for an explanation of how the separations between groups is perpetuated and elaborated through different belief systems. Becker (1974), whose name is closely associated with labeling theory, has decried the application of that term to a theory he prefers to think of as interactionist. The act of labeling or public discrediting is important to the extent that it can foreclose on opportunities to engage in conventional activities, but it does not by itself explain acts of deviant behavior. The distribution of the power to apply labels is indicative of the social differentiation ultimately expressed in different behaviors, attitudes, and beliefs.

Interaction theorists such as Becker (1974), Goffman (1963) and Lemert (1967) are concerned with the comparative power by which some groups define how other groups will be perceived and treated. By controlling what is defined as normal or nondeviant in everyday life, elite groups (generally white, ruling-class men) maintain their power without always having to resort to brute force. Oppression is thereby normalized, routinized, and institutionalized. Interactionist theories help to deconstruct these processes, as does feminist theory.

Feminist theory of delinquency is also interactionist in its attention to the role of power differentials in the adjudication of offenses for which females are charged. Reitsma-Street (1991) notes that girls are "policed to care" for others over themselves and to bear the cost of that caring in restrictions on the development of their own interests and independence. When they rebel against prescribed ways of looking, acting, and loving, retaliation may be expressed first in slurs on their reputation (labeling) and then escalated through probation, fines, and community service. Prior to the 1980s and the introduction of the Young Offenders Act (Revised Statutes of Canada, c. y-1, 1985), up to 60% of girls admitted to Canadian correctional institutions were admitted for status offenses such as running away, incorrigibility, truancy, prostitution, or other sexual immoralities not considered offenses if committed by boys.

Integrated Theories

Among the integrated approaches, feminist criminology advances interactionist approaches by including gender, class, and power in the study of the regulation of social interaction. It also obviates the fact that delinquency theory and research, including this present review, is largely the study of male delinquency.

Among other attempts to revise and integrate the various causal theories of delinquency, Thornberry (1987) advances an interactional theory that attempts to examine reciprocal causal structures. Thornberry asserts that since most human behavior occurs in social interaction, it is best explained by a model that focuses on interactive processes. Whereas Reitsma-Smith (1991) suggests that adjudicated female delinquency is the consequence of excessive social restraints, Thornberry suggests that the fundamental cause of delinquency lies in the weakening of social restraints (as per strain theory and social control theory), but that this attenuation of controls does not lead directly to delinquency. For delinquency to occur it must be learned and performed in situations which provide a high probability of reinforcement (learning theory). This then requires attention to both sociological and situational factors.

Elliot et al. (1979) attempted to formulate a model that integrated juvenile delinquency theories based on the premise that multiple causal paths lead to delinquency. The "proposed integrated theoretical paradigm begins with the assumption that different youths have different early socialization experiences, which result in variable degrees of commitment to an integration into conventional social groups" (Elliot et al., 1979, p. 9).

Other integrated theories of delinquency incorporate environmental and community variables. Rutter and Giller (1984) draw on Clark's (1978) finding that predisposing factors are not themselves predictive of delinquency. Whether a delinquent act is committed also depends on current crises and stresses, the presence of situational opportunities, and cognitive and motivational factors such as the perception of risk and state of temperament at the time. Rutter and Giller draw on Rutter's (1979) own earlier work illustrating that causative influences on any specific behavior involve four different sets of factors: individual predisposition, ecological predisposition, current circumstances, and situational opportunities. It is therefore all four of these dimensions which must be explored and for which measures must be determined and intervention strategies developed.

Elaborating on the concept of ecological predisposition or risk, others have stressed the importance of interventions at the level of the community as well as the individual (Sundeen, 1982). Spergel, (1976) argued that the effectiveness of delinquency control programs is determined in significant measure by differences in community structure. Programs that emphasize services directed at the person and ignore the community may compound the problem in a community characterized by a high degree of control by outside agencies and a weak or fragmented local or horizontal system.

Sundeen also refers to the classic work of Warren (1963) who identified six types of neighborhoods based on identity, internal organization, and external linkages. The roles appropriate to the change agent seeking to improve the functioning of these communities are different with each of these types. To illustrate this point, diversion and deinstitutionalization efforts require the community to have the resources to reintegrate the incarcerated delinquent back into its community life. Their effectiveness rests in some measure of consistent social control operating beth formally and informally to sustain the person's links to a structure of integrative resources. Integrative structures and social control are inextricably linked through a set of community characteristics that determine the effectiveness of that linkage. Sundeen summarizes those characteristics to include: socio-demographic characteristics (including income and age distributions, ethnic/racial homogeneity, population transiency/stability); the extent and seriousness of crime patterns; relevant physical characteristics (including the density, type, and condition of housing); community values and attitudes toward delinquency and youth-serving agencies; and the organizational environment. Sundeen also draws on the findings of Coates, Miller, and Ohlin (1978) to support the importance of these community characteristics as determinants of program effectiveness in juvenile correction facilities. These authors found that the extent and quality of a program's community linkages are statistically associated with client measures including recidivism, referral rates, past offense record, family self-sufficiency, attitudes toward public officials, self-image, and perceptions of primary groups.

The theme of interorganizational linkages as a dimension of community integration was also central to the work of Spergel (1976b) who concluded that effective integration within the youth-serving agency's external organizational environment must include linkages with the legal system, the neighborhood, the school, and the family. In a more recent survey of the literature on juvenile correctional treatment between 1975 and 1984, Lab and Whitehead (1988) affirmed this finding. They found, for example, that diversion programs that were formally administered by the juvenile justice system held greater promise for diminishing recidivism rates as compared with those with less accountability and looser connections to the source of their referrals. They found that approaches that are community-based and rely on extensive social support systems and use well-trained counselors reported the best results.

Adding to Spergel's analysis of the community context of organizational behavior and Sundeen's perspective on the community context of individual behavior, Bartol and Bartol (1989) suggest that the analysis of social networks and their effects on the behavior of participants can serve as a bridge between levels of explanation. This approach follows the logic of Simmel (1922) who deduced that adolescents' beliefs in and conformity to conventional norms is dependent upon their attachments to people and institutions. The more links with these entities, the stronger the bond to conventional society.

Bartol and Bartol delineate four concepts central to network analysis that help establish the utility of this approach as a bridge between individual and systemic approaches. These are: social network; personal network; multiplexity; and density.

Social networks are comprised of groups or organizations linked by a web of social relations such as overlapping memberships or personal affiliations. The personal network emphasizes the individual participant and his or her connections with other people. Network multiplexity refers to the number of contexts (microsystems) in which the same people interact, and network density refers to the extent to which those in the same social network know and interact with each other. Network density is usually measured as the ratio of actual ties in a network to the number of possible ties. Density reaches a maximum when everyone in a network knows everyone else.

In essence, this approach assumes that greater multiplexity and density in social relationships fosters consistency in behavior and more individual conformity. In a dense, multiplex network, participants are more likely to conform to the social conventions that prevail in that network. It follows from this reasoning that neighborhoods characterized by such networks will have low levels of delinquency when the neighborhood itself is well-integrated into the broader society.

There is an implicit assumption in this approach which makes it vulnerable to the same criticism that has been leveled at earlier adherents of the theory of "social strain" to explain delinquency (Cohen, 1955; Cloward & Ohlin, 1960; Agnes, 1985; Gibbons, 1986). Both approaches assume that the network or subculture of juveniles at risk for delinquency conforms to dominant social conventions. The network approach does, however, admit the possibility of conflict in values and standards among different networks.

Bartol and Bartol suggest that in the case where the density of involvement in the primary network implies socialization to a set of behaviors and standards at variance with the broader society, that involvement may preclude introduction to a corrective alternative outside of that network. This is consistent with the argument advanced by Broderick and Pulliam-Krager (1979) in describing the family as a variably permeable system whose influence on the developing child is conditioned, in part, by the extent, nature, and compatibility of its transaction with other systems. The intense demands of the family, peer group, or any other primary group may inhibit participation in the broader community. The values and standards of the smaller and larger social aggregates in such instances may or may not be congruent and where they are not, the formation of alternative perspectives may be precluded by intense participation in the narrower network. The adolescent whose socialization to the norms of the broader, dominant society is thereby truncated may experience no "role strain" or conflict because the norms of the dominant society are not internalized as part of his or her social identity.

This perspective differs from both social strain theory (Cohen, 1955) and social control theory (Hirschi, 1969). The former assumes that the delinquent is socialized to ends he does not have the legitimate means to achieve, and the latter assumes that the delinquent's socialization was faulty. A network analysis, on the other hand, may suggest that on a day-to-day basis, the adolescent does not experience any conflict between the norms of his or her primary network and those of the broader society because the latter do not influence him or her in any familiar or meaningful way. Socialization to the norms of the youth's primary network may be thorough and highly influential. Conflict, on those occasions when it becomes evident, is thus better understood in terms of social dominance in controlling the process of public labeling in a pluralist and poorly integrated society.

Friday and Hage (1976) advance a "role relationship theory" which attempts to deal with both formative socialization and social integration. This theory has close parallels to network theory in its attention to the web of relationship patterns in which youths' socialization and social integration transpires and to the points of potential conflict in the interstices of competing systems. These authors suggest five critical social systems for youths: kin, community or neighborhood, school, work, and peers. The risk of delinquency depends on the degree of enmeshment and role congruence among these five. In order for community, school, and work to be meaningful arenas of social interaction, youths' attachments to these systems must be stable, unconflicted, and secure. Detachment from any one of these systems may not compel a youth to delinquency where there is close integration among the remaining systems and where the youth is securely attached to those systems. Friday and Hage found that the greater the number of relationships across these five systems, the less the risk of delinquency.

The integrative capacity of neighborhoods may deteriorate when they are propelled into demographic and economic instability by four main factors associated with alterations in the structure of industrial society: disinvestment, demolition and construction, demagoguery, and deindustrialization (Skogan, 1986). The subsequent physical deterioration of neighborhoods serves to signal the loss of control that residents may feel.

Greenberg, Roke and Williams (1985) argue similarly that low-income residents feel less in control over what happens in their environment, less effective in doing anything about it, less choice about being there, less overall involvement, less consensus, and more fear of crime. Low income areas that do develop strong informal social control tend to be characterized by one dominant ethnic group and greater network density and multiplexity such that social control is exercised through all of the systems with which families and youths interact.

Schuerman and Kobrin's (1986) longitudinal, ecological study of Los Angeles County's highest crime areas examined the developmental process whereby neighborhoods declined into high crime areas. They were able to delineate a series of stages in community deterioration. In the first stages, land use changes included an increase in the number of multiple-family dwellings and commercial establishments, followed by gradual disinvestment and abandonment. Changes in income distribution and subcultural changes followed in later stages. In advanced stages of decline, there was a greater increase in the proportion of unattached people, increased mobility, a breakdown of social controls, and normative ambiguity.

This conceptualization of the stages of community deterioration provides a clue to the more complex relationship between delinquency and the structure of family life. If we conceive of the structure of family life as comprised of sets of role relations through which the family seeks to make self-sustaining transactions with other social systems in its environment, then the importance of that environment as a determinant of family structure becomes evident. Families are linked to other systems through the participation of their members. If a family is isolated for sociocultural reasons or by its demographic isolation in a neighborhood where there are declining numbers of similar families, and where changes in land use patterns, disinvestment, and other indicators of deterioration have emerged, then the family's transactions with its environment will be less effectively self-sustaining and less integrative. Under these circumstances, family members may become more vulnerable to anomic behaviors including delinquency.

CONCLUSION

There are multiple causal pathways to delinquency. Psychological, familial, sociological, and community factors may interact in various ways to produce such behavior. These factors are dynamic and interactional in nature (Thornberry, 1987). Effective intervention rests on comprehensive assessment that can measure the relative and interactive effects of contributory factors in these various domains and produce a plan of intervention that is tailored to the problem profile of the youth, family, and community. Treating all delinquents with the same set of intervention techniques and resources is doomed to partial success at best. Such monolithic approaches confound evaluation, underestimating the effectiveness of intervention with those youths for whom that particular service mix "fits" and provide no explanation for the lack of success in other instances. Further, any developing person's problem profile is not static.

Effective assessment and intervention must be comprehensive and have the flexibility to respond both when family and psychological factors predominate in early adolescence and when community factors complicate the youth's emancipation process a year or two later. Program evaluation must therefore be conceived as part of the overall design of comprehensive intervention, and the evaluation models applied to such efforts must themselves be flexible and robust enough to examine differences and interactive effects among various program components.

Where comprehensive approaches to prevention and treatment of delinquency are established, evaluation methodology can move beyond the unidirectional explanations that have characterized the field. It is increasingly clear that delinquency is a complex phenomenon requiring more complex causal analysis. Structural equation modeling may meet that challenge.

Structural equation modeling allows researchers to explore various constructs and their inter-relationships as well as their causal contributions to other constructs. These constructs can be measured by multiple variables. This statistical technique allows for a more thorough and inclusive analysis of the causal paths which can lead to delinquency.

As noted, the causes of delinquency can be divided into four general categories: individual, family, community, and sociological characteristics, including subcultural affiliations (e.g., gangs). All of these are interactional in nature and must be measured using multiple measured variables.

Some of the individual characteristics that have been identified as possible causes of delinquency include: low self-esteem; learning disabilities, social skill, educational and problem-solving deficits; and socialization problems with respect to conventional norms and values. Each of these variables can be measured using standardized tests and scales, and when taken together they represent the latent variable or construct of "individual characteristics." Those variables that prove insignificant to delinquency causation with specified samples, are dropped statistically from the model. By implication, interventions focused at the individual level would not be targeted to this sample.

Research on family contributions to delinquent behavior has explored several relevant family characteristics. Although these variables have not been established as "causes" of delinquency, they may be significant through their interaction with other variables. Structural equation modeling allows for an analysis of those interactions. Important family variables may include both demographic and process variables. Demographic variables, for example, may include family size, income, parental composition, and marital status. Process variables may include parental affection, family conflict, level of supervision, style or harshness of discipline, and parental deviance. Again, each of these can be measured and the insignificant variables dropped from the model.

The community variables that contribute to delinquency have been less extensively researched, but a rich body of theory, particularly in sociology, specifies the importance of some community variables. Standardized measures of these variables may not be as available, but their development will constitute an important contribution to the field. Bartol and Bartol's network analysis is most promising in this regard. For example, measures of the multiplexity of service networks could serve as important outcome measures in evaluating the community development aspects of comprehensive intervention. They might also operationalize an important intervening variable in the evaluation of client outcomes. Other variables to be explored at the community level should include: the availability of community resources; level of community organization; sociodemographic characteristics; community values, norms and attitudes about delinquency; level of devaluation of the neighborhood; and general living conditions.

The existence of a "gang subculture" or delinquent subculture necessitates the systematic consideration of this variable in terms of its influence on individual youths, and also in terms of its interaction with community-level variables. Particularly in destabilized or fragmented neighborhoods, the gang represents an alternative social network with its own norms, values, and social controls. A high level of involvement in a delinquent subculture or gang is apt to produce higher levels of delinquent activity. A youth's level of subcultural involvement can be measured by the frequency of interaction, density, and quality of gang-related social support, delinquent attitudes and internalized norms, and comparative perceptions of acceptance and involvement in the delinquent subgroup and the general community.

Figures A through E present a conceptual model of the constructs which must be captured in a research design which employs structural equation modeling in the investigation of interaction causal factors across these various domains.

These diagrams summarize the potential sources of influence reviewed here and propose a model of causal and interactional relationships that may contribute to delinquency. Though complex, this model presents a more comprehensive, more realistic approach to the multifaceted problem of contemporary delinquency. The model is intended to enable practitioners, theorists, and researchers to conceptualize the multiple causal pathways that can lead to delinquency and therefore explain the variation within the population of delinquent youths. Implicit in the model is the recognition that interventions with this population must be wholistic and address many areas rather than be limited by the assumptions of any single-focus programs. Delinquency manifests itself as a community problem, and no single agency has either the resources or the mandate to presume to represent an entire community. Conceived as a problem with multiple interactive causes, delinquency requires multiple interaction solutions. In the process of developing those solutions, multiple agencies and formal and informal community groups and organizations will strengthen the ties among themselves, and in so doing remediate one of the potential sources of the problem.

REFERENCES

Agnew, R. (1985). A revised strain theory of delinquency. Social Forces, 64, 151-67.

Bartol, C. R., & Bartol, A. M. (1989). Juvenile delinquency: A systems approach. Englewood Cliffs, NJ: Prentice-Hall.

Becker, H. (1974). Labeling theory reconsidered. In P. Rock, & M. MacIntosh (Eds.), Deviance and social control (pp. 41-66). London: Tavistock Publications.

Binder, A., & Binder, V. L. (1983). Juvenile diversion. The Counselling Psychologist, 11, 69-77.

Blomberg, J. G., & Carballo, F. L. (1979). Accelerated family intervention in juvenile justice: An exploration and recommendation for constraint. Crime and Delinquency, 25, 497-502.

Broderick, C. B., & Pulliam-Krager, H. (1979). Family process and child outcomes. In W. Burr, R. Hill, F. Nye, & I. Reiss (Eds.), Contemporary theories about the family, Vol. 1. (pp. 604-615). New York: Free Press.

Burt, C. (1925). The young delinquent. New York: Appleton & Co.

Carill, V. C., Gusmar, L. L., & Wolff, L. (1983). The role of the family in resocialization of juvenile offenders. Journal of Comparative Family Studies, 13, 63-74.

Clarke, R. E. (1972). Reference group theory and delinquency. New York: Behavioral Publications.

Cloward, R., & Ohlin, L. (1960. Delinquency and opportunity: Theory of delinquent gangs. New York: The Free Press.

Coates, R. B., Miller, A. D., & Ohlin, L. E. (1978). Diversity in a youth correlational system: Handling delinquents in Massachusetts. Cambridge: Ballinger.

Cohen, A. (1955). Delinquent boys: The culture of the gang. Illinois: The Free Press.

Conger, J. J. (1966). Personality, social class and delinquency. New York: Wiley.

Elliot, D., Ageton, S., & Canter, R. (1979). An integrated theoretical perspective on delinquent behavior. Journal of Research in Crime Delinquency, 16, 3-27.

Friday, P., & Hage, G. (1976). Youth crime in postindustrial societies: An integrated approach. Criminology, 14, 347-368.

Gibbons, D. (1986). Juvenile delinquency: Can social science find a cure? Crime and Delinquency, 32(2), 186-204.

Goddard, H. H. (1915). The criminal imbecile: An analysis of three remarkable murder cases. New York: MacMillan.

Goffman, I. (1963). Stigma. Englewood Cliffs, NJ: Prentice Hall.

Greenberg, S. W., Roke, W. M., & Williams, J. R. (1985). Informal citizen action and crime prevention at the neighborhood level: Executive summary. Washington, DC: U.S. Government.

Hathaway, S., & Monachesi, E. (1957). The personalities of pre-delinquent boys. Journal of Criminal Law, Criminology, and Police Science, 48, 148-163.

Hetherington, E. M., Stouwie, R. J., & Ridberg, E. (1971). Patterns of family interaction and child-rearing attitudes related to three dimensions of juvenile delinquency. Journal of Abnormal Psychology, 78, 160-173.

Hirschi, T. (1969). Causes of delinquency. Berkeley: University of California Press.

Klein, M. W. (1971). Street gangs and street workers. Englewood Cliffs, NJ: Prentice-Hall.

Lab, S. P., & Whitehead, J. T. (1988). An analysis of juvenile delinquency correctional treatment. Crime and delinquency, 34, 60-83.

Lemert, E. (1967). Human deviance, social problems, and social control. Englewood Cliffs, NJ: Prentice-Hall.

Long, S., & Sherer, M. (1984). Social skills training with juvenile offenders. Child and Family Behavior Therapy, 6(4), 1-11.

Matsueda, R. L., & Heimer, K. (1987). Race, family structure, and delinquency: A test of differential, association and social control theories. American Sociological Review, 52(6), 826-840.

Perlmutter, B. (1987). Delinquency and learning disabilities: Evidence for compensatory behaviors and adaptation. Journal of Youth and Adolescence, 16, 89-95.

Quay, H. (1987). Handbook of juvenile delinquency. New York: Wiley.

Reitsma-Street, M. (1991). Girls learn to care: Girls policed to care. In C. T. Baines, D. M. Evans, & S. M. Neysmith (Eds.), Women's caring: Feminist perspectives on social welfare (pp. 106-137). Toronto: McClelland and Stewart.

Rutter, M. (1979). Changing youth in changing society. London: Nuffield Provincial Hospitals Trust (1980), Cambridge, MA: Harvard University Press.

Rutter, M., & Giller, H. (1984). Juvenile delinquency: Trends and perspectives. New York: Guilford Press.

Sampson, R. J. (1986). Crime in cities: The effects of formal and informal social control. In A. G. Reiss, Jr., & M. Tonry (Eds.), Communities and crime. Chicago: University of Chicago Press.

Schuerman, L., & Korbin, S. (1986). Community careers in crime. In A. G. Reiss, Jr., & M. Tonry (Eds.), Communities and crime. Chicago: University of Chicago Press.

Simmel, G. (1922). Conflict and the web of group affliction. New York: The Free Press.

Skogan, W. (1986). Fear of crime and neighborhood change. In A. J. Reiss, Jr., & M. Tonry (Eds.), Communities and crime. Chicago: University of Chicago Press.

Spergel, I. A. (1966). Street gang work: Theory and practice. Reading, MA: Addison-Wesley.

Spergel, I. A. (1976). Interactions between community structures, delinquency, and social policy in the inner city. In M. Klein (Ed.), The juvenile justice system (pp. 55-99). Beverly Hills: Sage.

Spergel, I. A. (1976b). Community-based delinquency prevention programs: An overview. In J. Monihan (Ed.), Community mental health and the criminal justice system (pp. 35-50). Elmsford, NJ: Pergamon.

Sundeen, R. A. (1982). The community connection: Locally based criminal and juvenile justice programs and their community setting. Los Angeles. School of Public Administration, The Public Policy Institute of the Center for Public Affairs, USC.

Thornberry, G. (1987). Toward an interactional theory of delinquency. Criminology, 25, 863-91.

Tolan, P., Cromwell, R., & Brusswell, M. (1986). Family therapy with delinquents: A critical review of the literature. Family Process, 25, 619-649.

Warren, R. (1963). The community in America. 3rd ed. Chicago: Rand-McNally.

Marcia Fenner Wilson is a Doctoral Student, School of Social Work, University of Southern California, University Park, Los Angeles, California 90089-0411.