Article and Discussion

Youth & Society Volume 41 Number 1 September 2009 53-79 © 2009 Sage Publications 10.1177/0044118X09338343 http://yas.sagepub.com hosted at http://online.sagepub.com 53 Author’s Note: This research was supported in part by grant 96-MU-FX-0017 from the Office of Juvenile Justice and Delinquency Prevention, U.S. Department of Justice. Points of view or opinions in this article are those of the authors and do not necessarily represent official positions or policies of Office of Juvenile Justice and Delinquency Prevention. Please address correspondence to Beverly Kingston, University of Colorado, Institute of Behavioral Science, 1877 Broadway St., Suite 601, Boulder, CO 80302; e-mail: [email protected].

A Test of Social Disorganization Theory in High-Risk Urban Neighborhoods Beverly Kingston David Huizinga Delbert S. elliott University of Colorado although there is a growing body of research based on social disorganization theory that relates the neighborhood context to juvenile crime and delinquency, it is unknown whether neighborhood social processes operate in a similar way across all types of disadvantaged neighborhoods. It is possible that some social processes are unique to economically depressed areas. This research attempts to explain theoretically and test empirically the relationships between neighborhood social structure, social processes, delinquent opportunity structures, and rates of adolescent delinquency among structurally disadvantaged neighborhoods. The hypotheses are tested using neighborhood-level parent and youth data from 44 Denver neighborhoods. a series of regression models are constructed to estimate the effects of the neighborhood on rates of delinquency. The results show that for this high-risk sample the most consistent predictor of rates of problem behavior is youths’ perceptions of limited opportunities for the future.

Keywords: social disorganization theory; juvenile delinquency; neighbor- hood disadvantage S ocial disorganization theory indicates how structurally disadvantaged neighborhoods, characterized by high levels of poverty, single-parent households, racial and ethnic heterogeneity, and residential mobility are likely to have higher rates of juvenile delinquency (Bursik & grasmick, 54 Youth & Society 1993; elliott et al., 1996; Sampson, 1997; Shaw & McKay, 1942). Thus, neighborhood researchers often hypothesize that neighborhood structural disadvantage gives rise to social processes that lead to neighborhood dis- organization and higher rates of youth problem behavior. although these social problems tend to be worse in neighborhoods characterized as highly disadvantaged, some studies find that even among neighborhoods with high levels of structural disadvantage, there is substantial variation on these outcome measures (elliott, Menard, Rankin, & elliott, 2006; esbensen & Huizinga, 1990). For instance, neighborhoods characterized as highly disadvantaged on the basis of their social structure do not always exhibit high crime rates. This finding suggests that there may be significant vari- ation in social processes operating within structurally disadvantaged neighborhoods. although there is a growing body of research based on social disorgani- zation theory that relates neighborhood context to juvenile crime and delin- quency (for reviews, see Leventhal & Brooks-gunn, 2003; Sampson, Morenoff, & gannon-Rowley, 2002), it is unknown whether neighborhood social processes operate in a similar way across different types of disadvan- taged neighborhoods. It is possible that some social processes are unique to particular economically depressed areas. However, most neighborhood stud- ies examining social processes that contain all, or a variety of neighborhood types, do not have a large enough sample of disadvantaged neighborhoods to test these kinds of hypotheses. In this article, an attempt is made to theo- retically explain and empirically test the relationships between neighbor- hood social structure (poverty, single-parent households, racial and ethnic heterogeneity, and residential mobility), social processes, delinquent oppor- tunity structures, and rates of adolescent delinquency among structurally disadvantaged neighborhoods. By examining these issues using a sample of only high-risk neighborhoods, this research has the potential to clarify the relationship between social structure, social processes, delinquent opportu- nity structures, and delinquency for high-risk neighborhoods. These analy- ses may tease out some social structural variables that have the greatest impact on neighborhood social processes, and in-turn, on delinquency in this type of neighborhood. Because these are the neighborhoods with higher concentrations of a variety of social problems, understanding the variation between these neighborhoods that leads to juvenile delinquency may inform future neighborhood-level interventions (Coulton, Korbin, Su, & Chow, 1995; Sampson, 1992). Kingston et al. / a Test of Social Disorganization Theory 55 Neighborhood Social Structure, Social Processes, Delinquent Opportunity Structures, and Juvenile Delinquency Much of the neighborhood literature is couched within the frame- work of social disorganization theory that emerged primarily through the work of Shaw and McKay. In this theory, crime and delinquency are attributed to impaired local controls at the neighborhood level. Research has consistently associated the following structural characteristics of a neighborhood with high rates of crime and delinquency: residential mobility, population heterogeneity, single-parent households, and con- centrated poverty (Sampson & groves, 1989; Sampson & Lauritsen, 1994; Sampson, Raudenbush, & earls, 1997; Shaw & McKay, 1942; Wilson, 1987). Since the early 1990s, studies have attempted to explain the social proc- esses or mechanisms through which neighborhood structure leads to delin- quent behavior. Reviews of this research reveal two complementary types of neighborhood social processes that fit within the framework of social disor - ganization theory: social processes generated by formal and informal net- works of association and informal social control or collective efficacy (Leventhal & Brooks- gunn, 2003; Sampson, et al., 2002). Neighborhoods with weak social networks and low levels of collective efficacy typically lack the resources, social support, and informal social controls that are essential for healthy youth development. Consequently, youth growing up in such neighborhoods may develop weakened social bonds to conventional society ( e lliott, ageton, & Canter, 1979; elliott et al., 2006; Hawkins, 1996).

Neighborhoods are also likely to vary by the opportunities they provide for achieving delinquent and conventional goals (Cloward & Ohlin, 1960; elliott et al., 1996). Disorganized neighborhoods, with low levels of social control, may foster the existence of delinquent opportunity structures, including exposure to delinquent peer groups. For youth with weak bonds to conven- tional society, the social reinforcement provided by the delinquent peer group increases the likelihood of sustained delinquent behavior. The following con- ceptual discussion illustrates the relationships between neighborhood social structure, social processes, and delinquent opportunity structures and describes how these neighborhood-level variables are expected to influence delinquent behavior in disadvantaged neighborhoods. 56 Youth & Society Social Networks, Collective Efficacy, and Juvenile Delinquency Bursik and grasmick (1993) asserted that neighborhood life is shaped by the structure of formal and informal networks of association. These networks of association influence the amount of social support, informal social control, and resources available in the neighborhood. Bursik and grasmick rely on Hunter’s (1985) three-level approach to community social control to explain how social networks are fundamental to the control of crime and delinquency at the neighborhood level. The first level, the private level, consists of the intimate, informal primary groups that exist in the area. Social control at the private level stems from the allocation or threatened withdrawal of sentiment, social support, and mutual esteem that occurs within these networks when the network norms are violated. For example, youth who are bonded to those who hold prosocial beliefs do not want to threaten the bond by behaving in ways that would jeopardize their relationship. The second level, the parochial order, represents broader, local, interpersonal networks and the interlocking of local institutions and organizations such as schools, religious institutions, and youth centers. Ideally, the parochial order provides youth with institutional resources, informal surveillance by adults, and exposure to prosocial community norms. The third level, the public level of social control, focuses on the community’s ability to secure public resources and services by forming connections to agencies and organizations located outside the neighborhood boundaries.

Neighborhoods with strong organization at the private and parochial levels are more likely to access funding opportunities and other public resources that support the developmental needs of youth. While Bursik and grasmick (1993) discussed how social networks and their associated resources impact the ability of a community to regulate the behavior of its youth, Sampson and his colleagues (1997) contended that social networks are a necessary—but not necessarily sufficient—condition for the informal control of crime and delinquency. They explain that collec- tive efficacy is the emergent process by which neighborhood social ties are activated within social networks to enhance social control and demonstrate that rates of violence are lower in neighborhoods characterized by high levels of collective efficacy. Collective efficacy is defined as mutual trust and solidarity among neighbors combined with the willingness of local residents to intervene on behalf of the common good (i.e., neighbors inter- vening if kids from the neighborhood are getting into trouble). Residents tend to intervene for the common good when there are conditions of mutual Kingston et al. / a Test of Social Disorganization Theory 57 trust and solidarity among neighbors, but they are unlikely to intervene when the rules are unclear and people mistrust or fear each other. Juvenile delinquency is likely to be lower in neighborhoods with high levels of col- lective efficacy due to greater levels of informal social control.

Relationship Between Neighborhood Social Structure and Social Processes Neighborhood social structure influences the levels of social support, collective efficacy, and informal social control, and the availability of resources generated at the neighborhood level. The structural characteristics of disor- ganized neighborhoods (e.g., population heterogeneity, residential mobility, single-parent households, and concentrated poverty) are likely to impede the formation of strong prosocial networks; informal social controls; edu- cational, recreational, and health resources that are essential for healthy youth development; and the prevention of delinquent behavior. Population heterogeneity, involving a diversity of values, cultural back- grounds, and often language barriers, affects social networks by impeding communication and decreasing the likelihood that residents will share com- mon values. Likewise, residential mobility is expected to have an adverse affect on the formation of social networks because it takes time to develop strong social ties that contribute to community social organization.

Neighborhoods with a high percentage of single-parent households may have fewer adults physically available to provide surveillance for the behavior of their children or other children in the neighborhood. In addi- tion, single working parents may have little time or energy left to invest in relationships with neighbors. These structural conditions affect the ability of neighborhood residents to form social relationships essential for devel- oping mutual trust and solidarity, which are the prerequisites for the activa- tion of collective efficacy. Concentrated poverty at the neighborhood level also influences the for- mation of social networks and the availability of resources in the neigh\ bor- hood. Poverty is likely to be detrimental to the formation of social networks due to the negative feelings and experiences it fosters in individuals (Williams & Collins, 1995), which may in turn affect the formation and development of social relationships within the neighborhood. High-poverty neighborhoods also have limited resources to support the educational, rec- reational, and health needs of youth residing within these communities (Wilson, 1987). For instance, the poorest health services are often found in lower-income, minority, and transient areas (Bronfenbrenner, Moen, & 58 Youth & Society garbarino, 1984; O’Loughlin, Paradis, gray-Donald, & Renaud, 1999).

Typically, impoverished neighborhoods lack such stable institutions as recreational facilities, stores, libraries, banks, convenience stores, and child care centers. Schools that serve impoverished neighborhoods are often deficient of basic material resources. These schools are rundown, over- crowded, lack teaching supplies and textbooks, and employ less qualified teachers than more affluent neighborhood schools (Kozol, 1991). Schools in wealthy areas can provide high-quality education that increase students’ interests in academic pursuits and enhance their chances of future success, whereas schools in impoverished neighborhoods may impart a lower qual- ity of education due to their limited resources.

Delinquent and Conventional Opportunity Structures Neighborhoods are also expected to vary according to the opportunities they provide youth for achieving delinquent or conventional goals (Cloward & Ohlin, 1960; De Coster, Heimer, & Wittrock, 2006; Haynie, Silver, & Teasdale, 2006). The social structure (poverty, single-parent households, racial and ethnic heterogeneity, and residential mobility) and social proc- esses (weak social networks and low levels of collective efficacy) that characterize disorganized neighborhoods also contribute to the presence of delinquent opportunity structures and an absence of conventional opportu- nity structures for youth. While limited research exists on the relationship between neighborhood social structure, social processes and delinquent opportunity structures, recent studies examining these variables suggest that neighborhood disad- vantage influences delinquency indirectly by increasing exposure to crimi- nogenic street context (De Coster et al., 2006) and opportunities for involvement with delinquent peer groups (Haynie et al., 2006). Structurally disadvantaged neighborhoods that lack the resources to effectively monitor children and provide few sanctions for inappropriate behavior are likely to have a higher number of delinquent peer groups avail- able to youth (Cloward & Ohlin, 1960; Rankin & Quane, 2002; Sampson, 1997; Sampson & groves, 1989). also, the social networks between parents and the parents of their children’s friends who allow parents to monitor their children’s activities tend to be deficient in structurally disadvantaged neigh- borhoods. Poorly monitored youth are more likely to socialize with deviant peers and to engage in misconduct (Henry, Tolan, & gorman-Smith, 2001; Larzelere & Patterson, 1990; Patterson & Stouthamer-Loeber, 1984). Due to their limited resources, disadvantaged neighborhoods are also expected to provide limited conventional opportunities, including exposure Kingston et al. / a Test of Social Disorganization Theory 59 to social networks modeling conventional success. Youth in advantaged neighborhoods usually have opportunities that poor children lack, such as summer camps, music lessons, sports training, home computers, and spe- cial tutoring, whereas disadvantaged neighborhoods have few public resources that support the developmental needs of youth and their families. Often these neighborhoods lack the presence of community learning activities such as parks; family resource centers; and literacy programs that promote school readiness, physical health, and socioemotional well-being (Leventhal & Brooks-gunn, 2003). Furthermore, the social isolation fostered by neigh- borhood structural disadvantage is also likely to deprive youth of cultural learning from mainstream social networks modeling conventional success (Wilson, 1991). Youth residing in these areas are often uninformed about how to access potential jobs or educational supports that could provide them with the personal competencies essential for a successful transition to adulthood.

adolescents living in poverty may recognize the limitations of their circumstances and have little hope for their future. Youth living in these environments are likely to cognitively understand that their schools and communities are inadequate and that they lack access to the same opportu- nities as their wealthier counterparts. as youth from these neighborhoods consider their future options, they can assess the limited likelihood of attaining success through legitimate means such as by going to college or by getting a good job. Research on self-efficacy demonstrates that an indi- vidual’s beliefs about his or her future success affect behavior (Henderson & Dweck, 1990; Skinner, 1995). Thus, youth residing in impoverished environments, who feel a sense of hopelessness about their future, may act in ways that are counterproductive to their healthy growth and development (e.g., dropping out of school, engaging in delinquent behavior, and using alcohol and drugs). a visual representation of the full conceptual model illustrating the rela- tionship between neighborhood social structure, social processes, delin- quent opportunity structures, and rates of delinquent behavior is given in Figure 1. 1 Social Disorganization Theory in Structurally Disadvantaged Neighborhoods although prior neighborhood research suggests the model diagrammed in Figure 1 predicts rates of delinquent behavior, it remains unknown whether neighborhood social processes and delinquent opportunity structures operate in a similar way across different types of structurally disadvantaged 60 Youth & Society neighborhoods. In fact, there are different possibilities that may explain variation in crime and delinquency between structurally disadvantaged neighborhoods, including differences in the levels and types of structural disadvantage that occur between neighborhoods. Most neighborhood stud- ies treat socially disorganized areas as if they uniformly possess the same structural characteristics. However, these areas may vary on specific social structural characteristics linked to the social processes operating within the neighborhood. For example, one set of structurally disadvan- taged neighborhoods could have high poverty rates but lower rates of single-parent households, which may contribute to variation in the levels of social processes and, in turn, rates of juvenile delinquency between neighborhoods. elliott and colleagues, in their study of Denver and Chicago neighborhoods found some high-poverty neighborhoods with above-average levels of social organization and positive normative orientations. also, delinquent opportunity structures are not found in all structurally disad- vantaged neighborhoods (Cloward & Ohlin, 1960; Kobrin, 1951; Kornhauser, 1978; Short & Strodtbeck, 1965; Smith & Jarjoura, 1988; Figure 1 Conceptual Model of Neighborhood-Level Influences on Juvenile Delinquency Weak social networks • Pr ivate networ ks • Parochial networ ks • Public networ ks Disorganized neighborhoods • Low economic status • Population heterogeneity • Residential mobility • Single parent households High rates of delinquent behavior Existence of delinquent or low levels of conventional opportunity structures Weak social networks • Pr ivate networ ks • Parochial networ ks • Public networ ks Low levels of collective efficacy • Mutual trust • willingness to inter vene Kingston et al. / a Test of Social Disorganization Theory 61 Spergel, 1964). On the other hand, it is also possible that neighborhood social processes may operate independently of these neighborhood social struc- tural characteristics. Duncan, Duncan, Okut, Strycker, and Hix-Small (2003) found no relationship between neighborhood demographic indica- tors (residential mobility, percent White, and percent below poverty) and perceived collective efficacy in the neighborhood. although some studies suggest that neighborhood structural characteristics directly affect crime and other social outcomes, the results are usually modest (Morenoff, Sampson, & Raudenbush, 2001; Peterson, Krivo, & Harris, 2000). Thus, overall, there appears to be room for variation in structurally disadvan- taged neighborhoods and reason to believe that the quality of the neigh- borhood is not entirely determined by its level of advantage or disadvantage (elliott et al., 2006). The goal of this article is to empirically test the conceptual model pre- sented in Figure 1 among a sample of structurally disorganized neighbor- hoods and thereby also examine whether sufficient variation exists between such neighborhoods to permit useful analyses. The following questions are addressed: What is the relationship between each of the social structural characteristics (poverty, single-parent households, racial and ethnic heterogeneity, and resi- dential mobility) and the neighborhood social processes? In other words, what is the relative strength of each of these structural predictors on social processes?

What is the relationship between neighborhood social structure, social proc- esses, and delinquent opportunity structures?

What is the combined effect of neighborhood social structure, social processes, and delinquent opportunity structures on rates of juvenile delinquency? Methods Sample The Denver Youth Survey (DYS) is a prospective longitudinal study, with major funding from the Office of Juvenile Justice Delinquency Prevention and National Institute on Drug abuse, designed to improve the understanding of the causes and correlates of serious delinquency, vio- lence, and drug use by examining how youth develop within the context of family, school, peers and community. The DYS is based on a probability sample of households in high-risk neighborhoods of Denver, Colorado. 62 Youth & Society High-risk neighborhoods were defined as areas characterized by structural disorganization and high official crime rates (esbensen & Huizinga, 1990).

To determine structural disorganization, 35 variables were initially selected from the 1980 census representing seven conceptual domains: family struc- ture, ethnicity, socioeconomic status, housing, mobility, marital status, and age composition. a factor analysis of the variables in each of the seven domains resulted in the identification of 11 distinct factors. Using these factors, a cluster analysis was conducted to identify and combine similar block groups of the city. Seven clusters, or groupings of block groups emerged, and three of these clusters were identified as being disorganized based on their structural characteristics. Using Denver Police Department arrest data, Denver neighborhoods in the upper one third of the crime dis- tribution were identified. To meet the DYS high-risk criteria, a block group had to be identified as both structurally disorganized and be in the upper one third of the crime distribution. 2 Selection of survey respondents was based on a probability sample of households drawn from the block groups meeting the high-risk criteria. a full household enumeration of these areas was conducted and a sample of more than 20,000 households was selected. a listing of household members was obtained from the sampled households. Youth aged 7, 9, 11, 13 and 15, and one of their parents/primary caretakers were the eligible respondents of the study. In the first wave of the DYS study, 1,528 youth between the ages of 7 and 15 in 1987 (85% of estimated total number of eligible youth) com- pleted interviews. The sample is almost equally divided by gender (53.3% are men) and is ethnically diverse (10.4% White, 32.6% african american, 44.4% Hispanic, and 12.6% report Other or mixed ethnicity). Data were gathered by confidential, in-person interviews with youth and their primary caretaker beginning in 1988 (Wave 1). all participating respondents, youth, and parents, provided informed consent before being interviewed. Because the present research is concerned with adolescent delinquency, only youth aged 12 to 17 in 1987 (N = 875) are included in the sample for the analyses of this report. 3 Parent and youth data between 1989 and 1990 (Waves 2 and 3) of the DYS are used to test the hypotheses in this study.

These analyses make the theoretical assumption that the most proximate neighborhood influences on adolescent delinquency are current influences in time, and the effect of the neighborhood on delinquency is most likely to be observed during the year the delinquency is performed (aber, gephart, Brooks-gunn, & Connell, 1997; elliott et al., 2006). each wave of the DYS measures the respondents’ current beliefs and attitudes about Kingston et al. / a Test of Social Disorganization Theory 63 their neighborhood, whereas respondents’ delinquent behavior was reported for the past 12 months. Using these two consecutive waves of data main- tains the correct causal order by placing the measurement of neighborhood social processes at the start of the year corresponding to the year of delin- quency measurement.

as this study examines the influence of the neighborhood on adolescent delinquency and drug use, study participants who moved out of their origi- nal neighborhood between Wave 1 and Wave 2 were removed from the sam- ple. Out of 875 youth, 134 were removed from the sample because their family moved; the remaining 741 are included in the study. Because the DYS sample was originally selected for the study of individual-level data, transforming the data into neighborhood-level block groups required consid- eration of several factors. although sample respondents were from 89 dif- ferent block groups, some of the block groups had too small a population to provide reliable neighborhood-level measures. Previous research sug- gests that sampling 20 persons per neighborhood produces interrater relia- bilities for neighborhood information ranging from .70 to .90 (Raudenbush & Sampson, 1999). To address this block-group size issue, contiguous block groups, usually located within the same census tract, were combined to form 44 block-group clusters, treated here as neighborhoods. On aver- age, the sample of youth from each resulting neighborhood was 17, with a range between 5 and 46 youth. Measures Two types of neighborhood measures are used in this study: (1) struc- tural measures, using census data, and (2) neighborhood social process and delinquent opportunity structure measures, using aggregations of individual survey responses by neighborhood. Using Wave 2 (1989) data, aggregate- level neighborhood measures were developed by combining the responses of survey respondents living in the same neighborhood and use multiple reporting sources (adult and youth residents). These measures were con- structed by calculating the within-neighborhood mean of each measure across subjects. 4 Measures of neighborhood social structure were taken from concepts of social disorganization theory (elliott et al., 1996; Sampson et al., 1997; Shaw & McKay, 1942) and constructed from 1990 decennial census block group data. 5 Poverty is measured as the percentage of families within the 64 Youth & Society neighborhood that are living below the federal poverty level. Residential mobility is comprised of the proportion of families within the neighbor- hood that were living in a different house in 1985. Single-parent house- holds are measured as the percent of single-parent families with children under the age of 18 within the neighborhood. a scale with a range of 1 to 4 is used to measure the racial mix of the neighborhood. The scale increases by 1 for each racial group comprising 10% or more of the population. Five scales were constructed to measure private and parochial social networks and collective efficacy. For each scale, a higher score corresponds to stronger social networks or higher levels of collective efficacy. Private social networks are operationalized with two scales from the parent survey:

Social Network Size and Social Network Involvement. Social Network Size is an additive scale that measures the size of private social networks. This includes the number of good friends and family members who live in the neighborhood. It also reflects the number of other adults who residents talk with on a regular basis and how many children and teenagers in the neigh- borhood are known by name. 6 Social Network Involvement is a 7-item additive scale that reflects the levels of interaction and support provided by these private networks. It asks how many times in the past month neighbor- hood residents have been to each others house, gone out for an evening together, invited someone from the neighborhood to their house, talked to their neighbors about personal problems, or asked if they could borrow money (α = .67). Institutional effectiveness is a 7-item scale constructed to measure the quality of the parochial networks that exist in the neighbor- hood. It asks parent respondents to rate whether the neighborhood social institutions (e.g., schools, transportation services, police, and medical serv- ices) are effectively serving the neighborhood (α = .87).

The measure of collective efficacy is derived from the work of Sampson and colleagues (1997) and combines two scales that capture the neighbor- hood’s capacity for informal social control and social cohesion. The infor- mal social control measure in the present study includes four statements asking parents the likelihood that their neighbors would intervene if some- one were breaking into their house, selling drugs on the street, fighting in front of their house, or if their kids were getting into trouble (α = .87). The DYS does not have the same measure for social cohesion that Sampson and colleagues used in the construction of their collective efficacy scale. They describe social cohesion as a sense of trust between neighbors and shared norms and values. The Social Cohesion scale in this study taps the con- struct of shared norms and values by measuring neighborhood consensus on norms and values regarding delinquent and conventional youth behavior.

a high score indicates that a respondent believes that almost everyone in Kingston et al. / a Test of Social Disorganization Theory 65 their neighborhood would agree with their view on the norms and values regarding youth behavior. In contrast, a low score indicates that a respond- ent believes almost no one in his or her neighborhood would agree with his or her perception of the norms and values regarding delinquent and conven- tional youth behavior (α = .67).

Three scales are constructed to measure youths’ perceptions of the exist- ence of delinquent and conventional opportunities in the neighborhood. as observed in the conceptual model, a lack of conventional opportunities and the existence of delinquent opportunity structures in the neighborhood are expected to influence juvenile delinquency. For each of these scales, a higher score corresponds to greater perceived existence of delinquent opportunities and lower perceived existence of conventional opportunities at the neighborhood level. The Illegal Opportunities scale consists of 6 items asking youth to describe how often they have the chance to get away with criminal acts such as stealing, damaging property, or beating some- body up (α = .88). The Limited general Opportunity scale consists of 8 items designed to measure youths’ expectations about their future. For example, youth were asked whether they agreed with statements such as the following: “The world is usually good to people like you”; “a ll you see ahead are bad things not good things”; and “ as you get older, things will get better” (α = .65). The Limited Neighborhood Opportunity scale con- sists of 5 items relating to youths’ perceptions about the opportunities available in their neighborhood and includes items such as, “There isn’t much chance that a kid from your neighborhood will get ahead,” and “You’ll never have as much opportunity to succeed as kids from other neighborhoods” (α = .67).

The dependent variables in this study are derived from Wave 3 (1990) youth data and include two measures of the frequency of self-reported delin- quency aggregated to the neighborhood level. The Property Offending scale consists of 15 items ranging from minor theft (theft of less than $5, $5-50, $50-100) to more serious property damage such as purposely setting fire to a house, building, car, or other property and theft of more than $100. Violent offending is composed of 8 items that range from hitting someone with the idea of hurting them to more serious violence such as attacking someone with a weapon or with the idea of seriously hurting or killing them. See Table 1 for an overview of the neighborhood-level descriptive characteristics. Analysis Plan Ordinary least square (OLS) and backward stepwise regression are employed to construct a series of regression models to answer the questions 66 Youth & Society proposed in this research. The full models provide a standard examination of model parameters. Backward stepwise regression is utilized for choosing the variables to include in the regression models because it allows determi- nation of the variables most statistically salient and addresses the possibil- ity of suppression, where one variable may have a significant effect only when another variable is controlled (agresti & Finlay, 1997; esbensen, Huizinga, & Menard, 1999; Wofford, elliott, & Menard, 1994). an impor- tant issue in backward elimination is the possible exclusion of important variables from the model. Bendel and afifi (1977) contended that setting .05 as the criterion for statistical significance is too low and likely to result in missing a relationship that actually exists and suggest the statistical sig- nificance criterion for inclusion be set between .15 and .20 to reduce the likelihood of this occurring. This problem is magnified in these analyses as statistical significance is more difficult to detect with a small sample. Due to the small sample (N = 44), a significance level of .20 was set for the inclusion of variables in the backward elimination models presented here. For extended discussions of stepwise techniques, see agresti and Finlay (1997), and also Menard (2002). Table 1 Neighborhood-Level Descriptive Characteristics Variable Name Minimum Maximum M SD Neighborhood structure Percentage poverty 4 84 36 0.18 Percentage mobility 26 75 57 0.13 Percentage single-parent households 2 56 20 0.12 Racial mix 1 3 2.25 0.69 Neighborhood social process Social network size 9.56 78.60 25.31 12.73 Social network involvement 3.57 43.25 12.30 7.38 Institutional effectiveness 12.85 19.33 16.90 1.63 Social control 8.00 15.17 12.87 1.51 Social cohesion 7.01 11.17 9.03 0.95 Delinquent opportunity structures Illegal opportunities 6.44 14.18 10.44 1.74 Limited general opportunities 15.41 20.44 17.91 1.05 Limited neighborhood opportunities 9.69 15.29 12.16 1.15 Delinquency Property offending 0.14 156.88 6.50 23.52 Violent offending 0.09 57.74 3.78 9.81 Kingston et al. / a Test of Social Disorganization Theory 67 as specified in the conceptual model, there is an assumed causal order to the relationships between social structure and social processes.

Neighborhood social structure is assumed to influence social processes and social processes are assumed to influence delinquent opportunity struc- tures. To assess the relationship between the neighborhood social structure and social processes, each social process measure is regressed on the four social structure measures. Then, each delinquent opportunity structure measure is regressed on the social process and social structure measures.

Finally, a model estimating the effects of social structure, social processes, and delinquent and conventional opportunities is constructed to estimate the effects of neighborhood characteristics on rates of property and violent offending. all the independent variables are included in the estimation of these models and backward elimination is employed to determine the strongest predictors of delinquency. as a rule of thumb, when discussing the strength of the standardized coefficients for the aggregate-level data, regression coefficient sizes of 0 to .15 were considered weak, .15 to .30 moderate, .30 to .50 moderately strong, and effects above .50 are consid- ered strong. Results Relationship Between Neighborhood Social Structure, Social Processes, and Delinquent Opportunity Structures Table 2 presents the results of the stepwise analysis of the relationship between neighborhood social structure, social processes, and delinquent opportunity structure measures. 7 explained variance of the effects of social structure on social processes ranges from 4% to 29%. The social structure measures appear to explain institutional effectiveness fairly well. Poverty and residential mobility together explain 29% of the between-neighborhood variance in institutional effectiveness. Higher poverty neighborhoods have significantly less effective social institutions. Contrary to social disorgani- zation theory, neighborhoods with higher levels of residential mobility have more effective institutions. Racial mix and the percentage of single-parent households in the neighborhood together explain 18% of the variance in social control. as expected, neighborhoods with a greater racial mix and a higher percentage of single-parent households have lower levels of social control. Likewise, higher residential mobility is related to lower levels of social cohesion in the neighborhood, but explains only 7% of the variance 68 Table 2 Relationships Between Neighborhood Social Structure, Social Processes, and Delinquent and Conventional Opportunity Structures a Delinquent and Conventional Social Process Measures Opportunity Structure Measures Social Social Limited Limited Network Network Institutional Social Social Illegal general Neighborhood Size Involvement effectiveness Control Cohesion Opportunities Opportunities Opportunities Neighborhood structure Poverty — — –4.24*** — — — 1.72** 4.32*** (–.47) (.29) (.67) Mobility — 10.92 3.06* — -1.87* — — — (.20) (.25) (-.26) Single-parent households — — — -2.70 — — — -3.30* ( -.22) (-.36) Racial mix -5.45* — — -0.76** — — 0.63** — ( -.29) (-.35) (.41) Neighborhood social process — — — — — — — — Social network size — — — — — — — — Social network involvement — — — — — — — — Institutional effectiveness — — — — — — — — Social control — — — — — — — — Social cohesion — — — — — — 0.25* — (.23) R 2 .09* .04 .29*** .18** .07* 0 .31*** .28*** Note: The standardized regression coefficients are in the parentheses below each unstandardized coefficient.

a. Only variables with p < .20 are considered for inclusion in each model.

*p < .10. **p < .05. ***p < .01. in social cohesion. Racial mix is the only social structure variable signifi- cantly related to social network size, explaining 9% of the variance between neighborhoods. as predicted, neighborhoods with greater racial mix have smaller social networks. Social structure and social processes explain between 0% and 31% of the variance in the delinquent opportunity structure measures. Poverty, racial mix, and social cohesion together explain 31% of the between- neighborhood variance in limited general opportunity. Neighborhoods with higher poverty levels have significantly higher rates of limited opportuni- ties as perceived by youth at the .05 level. Neighborhoods with greater racial mix also have higher rates of youth perceiving limited opportunities for their future at the .05 level. Unexpectedly, neighborhoods with higher levels of social cohesion have higher rates of youth perceiving limited opportunities for their future. Poverty and single-parent households together explain 28% of the vari- ance in Limited Neighborhood Opportunity scale. as predicted, higher poverty neighborhoods have significantly higher rates of youth perceiving limited opportunities for their future due to their neighborhood. Contrary to expectations, when controlling for poverty, neighborhoods with higher percentages of single-parent households have lower perceptions of limited neighborhood opportunities. None of the neighborhood social structure or social process measures significantly relates to illegal opportunities at the neighborhood level.

Neighborhood Effects on Rates of Property and Violent Offending The relationships between neighborhood-level measures and rates of property and violent offending are presented in Table 3. Included in this table are the regression coefficients for the full and final set of predictors for each outcome measure. 8 For the Property Offending scale, the full model shows youths’ perceptions of limited neighborhood opportunities to be a strong significant predictor. Social network involvement, institutional effectiveness, and social cohesion approach the .20 significance level and their regression coefficients are in the hypothesized direction in the full model. In the final model, social cohesion and limited neighborhood opportunity explain 23% of the variance in rates of property offending. Youths’ perceptions of limited neighborhood opportunities (β = .44) remain a moderately strong significant predictor of property offending. at the .20 level, lower levels of social cohe- sion relate to higher levels of property offending in the final model. Kingston et al. / a Test of Social Disorganization Theory 69 70 Youth & Society Table 3 Results of Regression Analyses of Neighborhood-Level Measures on Rates of Property Offending and Violent Offending Property Offending Violent Offending Neighborhood Structure Full a Significant Final b Significant Full a Significant Final b Significant Poverty -25.83 .45 — — 20.11 † .17 16.70* .09 ( -.20) — — — (.37) — (.30) — Mobility 15.98 .62 — — -5.45 .69 — — (.09) — — — (-.07) — — — Single-parent -9.54 .82 — — 19.77 .27 20.21 † .15 households (-.05) — — — (.25) — (.26) — Racial mix -4.14 .51 — — -1.63 .54 — — ( -.12) — — — (-.11) — — — Neighborhood social process Social -0.05 .86 — — -0.12 .37 — — network size (-.03) — — — (-.15) — — — Social network -0.63 .23 — — 0.37* .10 0.30 † .11 involvement (-.20) — — — (.28) — (.22) — Institutional -3.19 .24 — — 0.80 .49 — — effectiveness (-.22) — — — (.13) — — — Social control -0.20 .95 — — 0.35 .77 — — ( -.01) — — — (.05) — — — Social cohesion -5.40 .25 -4.92 † .16 -0.24 .90 — — ( -.22) — (-.20) — (-.02) — — — Delinquent opportunity structures Illegal -0.70 .74 — — 0.06 .95 — — opportunities (-.05) — — — (.01) — — — Limited general -2.73 .55 — — -0.17 .93 — — opportunities (-.12) — — — (-.02) — — — Limited 10.53** .01 8.40*** .01 0.94 .57 — — neighborhood (.51) (.41) (.11) opportunities R 2 .34 .23*** .32 .27*** Note: The standardized regression coefficients are in the parentheses below each unstandard- ized coefficient.

a. Results from backwards stepwise regression with all variables included in the full model.

b. Only variables with p < .20 are considered for inclusion in the final model.

†p < .20. *p < .10. **p < .05. ***p < .01. The full model for violent offending shows poverty (β = .37) as a mod- erately strong predictor of violent offending in the hypothesized direction, revealing neighborhoods with higher rates of poverty have higher rates of violent offending. Social network involvement is a significant moderate Kingston et al. / a Test of Social Disorganization Theory 71 predictor of violent offending in the opposite direction hypothesized, indi- cating that neighborhoods with higher social network involvement have higher rates of violent offending. Single-parent households (β = .25) approaches the .20 significance level, suggesting neighborhoods with higher rates of single-parent households have higher rates of violent offending. In the final model, poverty, single-parent households, and social network involvement explain 27% of the variance in rates of violent offending.

Poverty (β = .30) becomes a moderate significant predictor of rates of vio- lent offending. While not statistically significant, the significance levels for single-parent households and social network involvement are below .20 in the final model. The direction and strength of the coefficients for these variables are similar to the parameters identified in the full model. Discussion One of the major goals of this article was to examine whether there was sufficient variation among a set of neighborhoods, all of which were identi- fied as being socially disorganized, to permit a test of a conceptual model specifying neighborhood level structural, social, and delinquent and con- ventional opportunities as predictors of delinquency. That is, are socially disorganized neighborhoods all the same? given the findings presented above, the answer appears to be no. as hypothesized, neighborhood social structure appears to affect social processes and delinquent/conventional opportunity structures in the sample of neighborhoods used in this study. Of the social process measures, social structure measures explain 29% of the variance in institutional effectiveness and 18% in social control. Social structure explains less than 10% of the variance in social network size, social network involvement, and social cohesion. Consistent with the results of previous research suggesting that limited public resources are available to support the educational, recreational, and health needs of residents in disadvantaged communities (Bronfenbrenner et al., 1984; Wilson, 1987), residents from poorer neighborhoods in the sample perceive less effective social institutions. Racially mixed neighbor- hoods had smaller social networks and lower levels of social control. These results suggest that residents in racially mixed neighborhoods have fewer close relationships with their neighbors and are less likely to intervene on behalf of the common good of the neighborhood. Sampson and colleagues (1997) explained that low levels of social control in a neighborhood may be due to a sense of mistrust between residents. It is possible that racially 72 Youth & Society mixed neighborhoods impede communication between residents, foster distrust, and lead to the existence of competing value systems.

Social structure and social process measures explain 31% of the vari- ance of limited general opportunities and 28% of the variance of limited neighborhood opportunities in the sample. However, these social structure and social process measures explain none of the variance for illegal oppor- tunities. Conceivably, all of the disorganized neighborhoods included in the sample provide substantial illegal opportunities for residents. Poverty posi- tively affects both limited general opportunities and limited neighborhood opportunities as perceived by youth suggesting that adolescents living in Denver’s poorest neighborhoods recognize the limitations of their circum- stances and have less hope for the future. although the majority of the relationships between neighborhood social structure, social processes, and delinquent opportunity structures are in the expected direction, three relationships appear contrary to the hypotheses.

First, when controlling for poverty, neighborhoods with higher mobility are perceived by residents to have more effective social institutions. One expla- nation for this finding may be related to the way residential mobility is measured and the nature of the DYS sample. Residential mobility is com- prised of the proportion of families within the block group that were living in a different house 5 years prior to the 1990 Census. Within the sample of high-risk neighborhoods, this could relate to the more upwardly mobile proportion of the population. These are families who have moved but into a better neighborhood with more effective social institutions than they were living in prior to their move. In fact, Furstenberg, Cook, eccles, elder, and Sameroff (1999) found that geographic mobility was a primary family management strategy once parents’ perceived high dangers in the neighborhood. They report that the more stable families migrated to better neighborhoods. Second, social cohesion has a positive effect on limited general opportunities, implying that parents perceive greater consensus on norms and values regarding delinquent behavior in neighborhoods in which youth perceive limited opportunities for their future. This finding may also relate to the manner in which social cohesion is measured. Though the scale for social cohesion measures neighborhood consensus, it does not reveal whether this consensus supports conventional norms and values or delin- quent values. It could be that high levels of consensus on values in the DYS refer to agreement with regard to antisocial values. Third, when controlling for poverty, neighborhoods with higher percentages of single-parent fami- lies have lower rates of youth who perceive limited opportunities for their future. This result seems to describe a unique neighborhood combination of Kingston et al. / a Test of Social Disorganization Theory 73 high single-parent households and lower poverty. Contrary to the typical expectations of single-parent households, these neighborhoods may have higher percentages of single parents who are succeeding economically and thus modeling sources of hope for their children’s future.

This research also examines the effects of neighborhood social struc- ture, social processes, and delinquent opportunity structures on property offending and violent offending. The level of explained variance ranges from 23% to 27% in these outcome measures. Different variables appear to influence these two types of delinquent behavior. For example, social structure influences violent offending but not property offending. Of the social structure measures, poverty is the only significant predictor of vio- lent offending, revealing neighborhoods with a higher percentage of resi- dents living in poverty have higher rates of violent offending. Other aggregate-level studies supports this finding (Farnworth, Thornberry, Krohn, & Lizotte, 1994; Jarjoura, Triplett, & Brinker, 2002; Ludwig, Duncan, & Hirschfield, 1998). Longitudinal research by Jarjoura and col- leagues showed that the persistent poor, as distinguishable from the short- term poor, are more likely to engage in delinquent behavior. They explained that “the degree to which a cross-sectional measure of poverty will tap persistently poor families is dependent on the nature of the sample” (p.

181). as described by Jarjoura’s research, samples like the DYS are spe- cifically designed to capture a larger sample of economically disadvan- taged youth. even though the data used here measure one point in time, it is likely that the highest poverty neighborhoods in the DYS sample include a high proportion of the persistent poor. The strongest predictor of rates of property offending was youths’ per- ceptions of limited opportunities for their future because of the neighbor- hood in which they live. This finding suggests that higher rates of property offending may be a response to a seemingly hopeless situation. For exam- ple, youth report believing that they will never have as much opportunity to succeed as kids from other neighborhoods and that there is not much chance that a kid from their neighborhood will ever get ahead. For youth perceiving limited opportunities because of their impoverished environ- ment, crime may be seen as the only opportunity for achieving a higher socioeconomic status. Whereas objective measures of social structure are likely to contribute to youths’ perceptions of their future, this finding implies that what is most important in predicting rates of property offend- ing in high-risk neighborhoods may be youths’ perceptions of the neighbor- hood structure and the ensuing options available to them that support their future success in conventional society. 74 Youth & Society The social process measures had little impact on both property offending and violent offending in this research. This may be a result of the homoge- neity of the high-risk neighborhoods included in the DYS sample. Several researchers have noted that a restricted range in the type of neighborhood samples can lead to an underestimation of neighborhood effects (Duncan, Connell, & Klebanov, 1997; Duncan & Raudenbush, 1999; Furstenberg et al., 1999). Duncan and Raudenbush (p. 29) stated, It is important to realize that effects may turn out to be small because the degree of variation is small, rather than because the setting is irrelevant.

Correlational research based on naturally occurring variation can identify plausible consequences only if that variation currently exists.

With little structural variation between neighborhoods, the capacity to identify subtle relationships between social structure, social processes, and the outcome measures is seriously restricted in this research. Therefore, it is likely that the estimated effects of neighborhood context understate the significance of neighborhoods for youth and their families residing in the DYS neighborhoods. There are several limitations in this study and the analyses presented.

First, the neighborhood-level sample (N = 44) is fairly small, restricting the types of analyses that could be utilized to examine the influence of the neighborhood-level predictors on delinquency. The small sample may also affect the ability of this study to detect significant relationships between the variables. Second, the normative orientation for the measures of Social Network Size, Social Network Involvement, and Social Cohesion scales is unclear, making the interpretation of the study results difficult for these variables. Though this research assumed that these networks are prosocial, this may not be the case. In fact, greater social networks and higher levels of social network involvement could refer to antisocial rather than proso- cial relationships. Third, whereas a nonrecursive relationship between the neighborhood characteristics and adolescent delinquency is specified in the conceptual model, it is possible that the real relationship between these meas - ures is recursive. Researchers have argued that the relationship between crime and social disorganization may be reciprocal in nature (Bursik & grasmick, 1993; Sampson et al., 2002). However, this research cannot provide information about the true causal relationship of the measures, because behavior is measured at a single point in time. Finally, results from this study can only be generalized to high-risk Denver neighborhoods dur- ing the time period of the study. Though the data set is from the period Kingston et al. / a Test of Social Disorganization Theory 75 1989-1990, it is a test of the neighborhood social processes first described by Shaw and McKay in the 1920s. The confirmation of their theory, in this and other research, suggests that these relationships apply over time (Leventhal & Brooks-gunn, 2003; Sampson et al., 2002). However, these methodological limitations suggest caution in the interpretation and gener- alization of the findings. More definitive conclusions await further studies and replication of findings. Conclusion Whereas most neighborhood studies treat socially disorganized areas as if they uniformly possess the same structural characteristics, this research aimed to tease out which social structural variables have the greatest impact on neighborhood social processes, delinquent opportunity structures, and, in-turn, on adolescent delinquency. as hypothesized by social disorganiza- tion theory, this study shows that in a sample of structurally disadvantaged neighborhoods, variation in neighborhood poverty accounts for variation in the perceived effectiveness of social institutions and youths’ perceptions of limited opportunities even when the variation in the sample is quite restricted. Furthermore, neighborhoods with greater racial mix have smaller social networks and lower levels of social control. The poorest Denver neighborhoods also have significantly higher rates of violent offending.

This research shows that variation in the levels and types of structural dis- advantage influence neighborhood social processes and delinquent oppor- tunity structures in some unexpected ways. For example, when controlling for neighborhood poverty, the effects of residential mobility and single- parent households appear contrary to expectations. Neighborhoods with higher mobility rates were perceived by residents to have more effective social institutions and neighborhoods with higher percentages of single- parent families have lower rates of youth who perceive limited opportuni- ties for their future. Unlike prior studies that test social disorganization theory using a full range of neighborhoods, this research shows social proc- ess measures to have minimal effects on delinquent opportunity structures and delinquency. This is likely a result of the lack of variation between neighborhoods on these measures. The strongest predictors of rates of delinquent behavior in high-risk neighborhoods are objective poverty and youths’ perceptions of limited oppor - tunities for their future suggesting the importance of implementing interven- tions aimed at improving both real opportunities and youths’ perceptions about their future opportunities. Typically, youth from disadvantaged neighborhoods 76 Youth & Society are exposed only to their parents and neighborhood friends and relatives who live in poverty. Intervention strategies that include matching youth with successful role models serving in a mentorship capacity and cultural enrichment activities introducing them to opportunities and lifestyles that exist outside of their neighborhood may affect rates of delinquency in high- risk neighborhoods. For example, the Quantum Opportunities Program is designed to serve disadvantaged adolescents by providing education, serv- ice, and development activities, as well as financial incentives, over a 4-year period, from ninth grade through high school graduation (Lattimore, Mihalic, grotpeter, & Taggart, 1998). This program aims to compensate for both the perceived and lack of real opportunities that are characteristic of disadvantaged neighborhoods. Likewise, the I Have a Dream Program helps children from low-income areas become productive citizens by providing a long-term program of mentoring, tutoring, and enrichment, with an assured opportunity for higher education through partial financial assistance to help pay for college. Because influencing youths’ perceptions about their future without improving real opportunity structures may increase frustration levels, interventions aimed at addressing both perceived and real opportuni- ties may have the strongest impact on delinquency rates in disadvantaged neighborhoods. Notes 1. See Kingston (2005) for an extension of this model explaining how the neighborhood is expected to affect individual-level bonding, association with delinquent peers, and delin- quent behavior. 2. Details of the factor and cluster analyses included in this process are described in Huizinga, esbensen, and Weiher (1991), and esbensen and Huizinga (1990).

3. Because the sampling procedure resulted in slightly disproportionate sampling within the three clusters, analyses were weighted to provide estimates representative of Denver’s high-risk neighborhoods. 4. For a complete list of the scale items, contact the corresponding author.

5. Because the census data used in this study were collected in 1990, the social structure measures were collected in the year following the social process and delinquency measures.

Consequently, this research is assuming some stability from the prior year’s social structure characteristics. 6. For this scale and the scales measuring self-reported delinquency, examining internal consistency is inappropriate. For example, engaging in one type of delinquent behavior does not necessarily mean the respondent will engage in another type. Likewise, residents who report having many good friends in their neighborhood may not necessarily have many family members living in their neighborhood. 7. a bivariate correlation matrix was not included in this manuscript due to space limita- tions but is available through the corresponding author. 8. The R 2 is not statistically significant in the full model due to the restricted degrees of freedom in the sample. Kingston et al. / a Test of Social Disorganization Theory 77 References aber, J. L., gephart, M. a., Brooks- gunn, J., & Connell, J. P. (1997). Development in con- text: Implications for studying neighborhood effects. In J. Brooks- gunn, g. J. Duncan, & J. L. Lawrence (eds.), neighborhood poverty (Vol. 1, pp. 44-61). New York: Russell Sage.

agresti, a. & Finlay, B. (1997). Statistical methods for the social sciences (3rd ed.). Upper Saddle River, NJ: Prentice Hall.

Bendel, R. B., & afifi, a. a. (1977). Comparison of stopping rules in forward regression.

Journal of the American Statistical Association, 72, 46-53.

Bronfenbrenner, U., Moen, P., & garbarino, J. (1984). Child, family, and community. In R. Parke (ed.), Review of child development research: Volume 7: The family (pp. 283-328).

Chicago: University of Chicago Press.

Bursik, R. J., Jr., & grasmick, H. g. (1993). Neighborhoods and crime. Lexington, Ma:

Lexington Books.

Cloward, R., & Ohlin, L. (1960). Delinquency and opportunity. New York: Free Press.

Coulton, C. J., Korbin, J., Su, M., & Chow, J. (1995). Community level factors and child maltreatment rates. Child Development, 66, 1262-1276.

De Coster, S., Heimer, K., & Wittrock, S. M. (2006). Neighborhood disadvantage, social capital, street context, and youth violence. The Sociological Quarterly, 47, 723-753.

Duncan, g. J., Connell, J. P., & Klebanov, P. K. (1997). Conceptual and methodological issues in estimating causal effects of neighborhoods and family conditions on individual develop- ment. In J. Brooks-gunn, g. J. Duncan, & J. L. aber (eds.), Neighborhood poverty:

Context and consequences for children (Vol. 1, pp. 219-250). Thousand Oaks, Ca: Sage.

Duncan, T. e., Duncan, S. C., Okut, H., Strycker, L. a., & Hix-Small, H. (2003). a multilevel contextual model of neighborhood collective efficacy. American Journal of Community Psychology, 32, 245-252.

Duncan, g. J., & Raudenbush, S. W. (1999). assessing the effects of context in studies of child and youth development. Educational Psychologist, 34(1), 29-41.

elliott, D. S., ageton, S. S., & Canter, R. J. (1979). an integrated theoretical perspective on delinquent behavior. Journal of Research in Crime and Delinquency, 16(1), 3-27.

elliott, D. S., Menard, S., Rankin, B., &. elliott, a. (2006). Good kids from bad neighbor - hoods: Successful development in social context. New York: Cambridge University Press.

elliott, D. S., Wilson, W. J., Huizinga, D., Sampson, R. J., elliott, a., & Rankin, B. (1996).

The effects of neighborhood disadvantage on adolescent development. Journal of Research in Crime and Delinquency, 33, 389-426.

esbensen, F., & Huizinga, D. (1990). Community structure and drug use: From a social disor- ganization perspective. Justice Quarterly, 7, 693-709.

esbensen, F., Huizinga, D., & Menard, S. (1999). Family context and criminal victimization in adolescence. Youth & Society, 31, 168-198.

Farnworth, M., Thornberry, T., Krohn, M., & Lizotte, a. (1994). Measurement in the study of class and delinquency: Integrating theory and research. Journal of Research in Crime and Delinquency, 31, 32-61.

Furstenberg, F., Cook, T. D., eccles, J., elder, g. H., & Sameroff, a. (1999). Managing to make it: Urban families and adolescent success. Chicago: University of Chicago Press.

Hawkins, J. D. (1996). Delinquency and Crime. Cambridge University Press.

Haynie, D. L., Silver, e., & Teasdale, B. (2006). Neighborhood characteristics, peer networks and adolescent violence. Journal of Quantitative Criminology, 22, 147-169.

Henderson, V. L., & Dweck, C. S. (1990). Motivation and achievement. In S. S. Feldman & g. R. elliott (eds.), At the threshold: The developing adolescent (pp. 308-329). Cambridge, Ma: Harvard University Press. 78 Youth & Society Henry, D. B., Tolan, P. H., & gorman-Smith, D. (2001). Longitudinal family and peer group effects on violence and nonviolent delinquency. Journal of Clinical Child Psychology, 30, 172-186.

Huizinga, D., esbensen, F. a., & Weiher, a. W. (1991). are there multiple paths to delin- quency? Journal of Criminal Law and Criminology, 82, 83-118.

Hunter, a. (1985). Private, parochial and public school orders: The problem of crime and incivility in urban communities. In g. D. Suttles & M. N. Zald (eds.), The challenge of social control: Citizenship and Institution building in modern society (pp. 230-242).

Norwood, NJ: ablex.

Jarjoura, g. R., Triplett, R. a., & Brinker, g. P. (2002). growing up poor: examining the link between persistent childhood poverty and delinquency. Journal of Quantitative Criminology, 18, 159-187.

Kingston, B. (2005). The effects of the neighborhood on adolescent delinquency and drug use.

Unpublished doctoral dissertation, University of Colorado, Boulder.

Kobrin, S. (1951). The conflict of values in delinquency areas. American Sociological Association, 95, 922-963.

Kornhauser, R. (1978). Social sources of delinquency. Chicago: University of Chicago Press.

Kozol, J. (1991). Savage inequalities. New York: HarperCollins.

Larzelere, R. e., & Patterson, g. R. (1990). Parental management: Mediator of the effect of socioeconomic status on early delinquency. Criminology, 28, 301-324.

Lattimore, C. B., Mihalic, S. F., grotpeter, J. K., & Taggart, R. (1998). Blueprints for violence prevention. Book Four: Quantum opportunities program. Boulder, CO: Center for the Study and Prevention of Violence.

Leventhal, T., & Brooks-gunn, J. (2003). Children and youth in neighborhood contexts. Current Directions in Psychological Science, 12(1), 27-31.

Ludwig, J., Duncan, g. J., & Hirschfield, P. (1998). Urban poverty and juvenile crime:

Evidence from a randomized housing-mobility experiment. Unpublished manuscript, georgetown Public Policy Institute.

Menard, S. (2002). Applied logistic regression analysis. Thousand Oaks, Ca: Sage.

Morenoff, J. D., Sampson, R. J., & Raudenbush, S. W. (2001). Neighborhood inequality, col- lective efficacy, and the spatial dynamics of homicide. Criminology, 39, 517-560.

O’Loughlin, J. L., Paradis, g., gray-Donald, K., & Renaud, L. (1999). The impact of a community-based heart disease prevention program in a low-income, inner-city neighbor- hood. American Journal of Public Health, 89, 1819-1826.

Patterson, g., & Stouthamer-Loeber, M. (1984). The correlation of family management prac- tices and delinquency. Child Development, 55, 1299-1307.

Peterson, R. D., Krivo, L. J., & Harris, M. a. (2000). Disadvantage and neighborhood violent crime: Do local institutions matter? Journal of Research on Crime and Delinquency, 37, 31-63. Rankin, B. H., & Quane, J. M. (2002). Social contexts and urban adolescent outcomes: The interrelated effects of neighborhoods, families, and peers on african-american youth.

Social Problems, 49, 79-100.

Raudenbush, S. W., & Sampson, R. J. (1999). ecometrics: Toward a science of assessing ecological settings, with application to the systematic social observation of neighbor- hoods. Sociological Methods and Research, 29, 1-41.

Sampson, R. J. (1992). Family management and child development. In J. McCord (ed.), Facts, frameworks, and forecasts: Advances in criminological theory (Vol. 3, pp. 63-93).

New Brunswick, NJ: Transaction Publishing. Kingston et al. / a Test of Social Disorganization Theory 79 Sampson, R. J. (1997). Collective regulation of adolescent misbehavior: Validation results from eighty Chicago neighborhoods. Journal of Adolescent Research, 12, 227-244.

Sampson, R. J., & groves, W. B. (1989). Community structure and crime: Testing social dis- organization theory. American Journal of Sociology, 94, 774-802.

Sampson, R. J., & Lauritsen, J. (1994). Violent victimization and offending: Individual, situa- tional, and community level risk factors. In a. J. Reiss, Jr. & J. a. Roth (eds.), Understanding and preventing violence: Social influences (Vol. 3, pp. 1-114). Washington, DC: National academy Press.

Sampson, R. J., Morenoff, J. D., & gannon-Rowley, T. (2002). assessing “neighborhood effects”: Social processes and new directions in research. Annual Review of Sociology, 28, 444-478. Sampson, R. J., Raudenbush, S. W., & earls, F. (1997). Neighborhoods and violent crime:

a multilevel study of collective efficacy. Science, 277, 918-924.

Shaw, C. R., & McKay, H. D. (1942). Juvenile delinquency and urban areas. Chicago:

University of Chicago Press.

Short, J. F., & Strodtbeck, F. L. (1965). Group process and gang delinquency. Chicago: University of Chicago Press.

Skinner, e. a. (1995). Perceived control, motivation, and coping. Thousand Oaks, Ca: Sage.

Smith, D. a., & Jarjoura, g. R. (1988). Social structure and criminal victimization. Journal of Research in Crime and Delinquency, 25, 667-703.

Spergel, I. a. (1964). Street-gang work: Theory and practice. Reading, Ma: addison-Wesley.

Williams, D., & Collins, C. (1995). U. S. socioeconomic and racial differences in health: Patterns and explanations. Annual Review Sociology, 21, 349-386.

Wilson, W. J. (1987). The truly disadvantaged. Chicago: University of Chicago Press.

Wilson, W. J. (1991). Studying inner-city social dislocations: The challenge of public agenda research. American Sociological Review, 56, 1-14.

Wofford, S., elliott, D. S., & Menard, S. (1994) Continuities in marital violence. Journal of Family Violence, 9, 195-225.

Beverly Kingston is the Project Director of the adams County Safe Schools/Healthy Students Initiative, a 4-year $8 million federal grant serving children, youth and families northeast of Denver, Colorado. Her research and professional interests focus on bridging the gap between research and practice to create and sustain social and physical environments that support healthy child and youth development. She has published on the theory of differential oppres- sion and adolescent problem behavior, Denver’s Child/Youth Friendly City Initiative, and the impact of playground renovations on children’s physical activity.

David Huizinga is a Senior Research associate at the Institute of Behavioral Science at the University of Colorado. His research interests include methodology and the substantive areas of child and adolescent delinquency, crime, violence, drug use/abuse, mental health, home- lessness, domestic violence, mental health care utilization and the comorbidity of these prob- lems. He is the co-author of five books, several book chapters and numerous journal articles on issues surrounding the development of delinquency, drug use and mental health. Delbert S. Elliott is a Distinguished Professor emeritus in the Department of Sociology and the Director of the Center for the Study and Prevention of Violence at the University of Colorado at Boulder. His research interests include the epidemiology and etiology of youth violence, crime, and substance use/abuse; crime prevention and intervention strategies; longi- tudinal survey research methodology. Professor elliott and Dr. Huizinga have recently pub- lished Good Kids from Bad Neighborhoods: Successful Development in Social Context.