Characteristics of Neighborhoods and Schools that Impact the Mental Health and Risk Behavior of Children and Youth

Through the lens of Bronfenbrenner's ecological systems theory (Bronfenbrenner, 1977, 1986), there are likely countless microsystem (directly experienced), mesosystem (cross-context), exosystem (indirectly experienced), and macrosystem (cultural) aspects of neighborhoods and schools that can affect child and youth development. Indeed, a wide variety of neighborhood and school characteristics has been linked empirically with child outcomes. The literature has yet to fully identify the dynamic processes by which neighborhood and school characteristics causally influence child and youth mental health and behavior. Neighborhoods have the potential to influence children and youth as direct, mediated, or moderated effects or through patterns that exhibit age or stage specificity (e.g., adolescents are most likely to experience neighborhoods directly, whereas young children will experience neighborhood effects as mediated through their effects on parents) (Aber, Gephart, Brooks-Gunn, & Connell, 1997). We suggest that school effects on children and youth also are likely to follow such dynamic patterns. In our reading of the literature, we found that many of the important and well-studied neighborhood and school characteristics could be grouped into three overarching categories: disadvantage, violence, and sense of community and support networks. Although we discuss them separately, we wish to emphasize that there are also reciprocal links, akin to Bronfenbrenner's mesosystem, both within categories (e.g., dangerous neighborhoods can make dangerous schools more likely: (Sheley, McGee, & Wright, 1992) and across them (e.g., levels of violence within a school affect its sense of community; Sandoval & Brock, 2002). Whenever possible, we discuss these within and between neighborhood and school influences.

It is also important to note that the vast majority of literature on neighborhood effects on children and youth refers to the neighborhoods in which children live. Clearly home neighborhoods are key environments for children, but we submit that the neighborhoods in which children attend school also have the potential to affect development.


For neighborhoods in particular, the question of how contexts affect child and youth development often hinges on the extent to which the contexts are disadvantaged, with an emphasis on a lack of resources. This has resulted in a bias in the literature toward identifying negative impacts of neighborhood and school contexts on children. Although some of the literature to be reviewed looks to potential contextual buffers, the bias is reflected here in the overwhelming preponderance of negative outcomes. It should also be noted that the majority of research to be summarized was conducted in the United States; European analogs are cited when appropriate.

Disadvantaged Neighborhoods

Characterizing a neighborhood as "disadvantaged" can have one of two meanings. The first is that the neighborhood itself is bereft of crucial institutional and social resources thought to be necessary for an enhanced quality of life; such a lack resources at the neighborhood level constitutes the neighborhood context. The second is that the neighborhood is comprised primarily of individuals and families who themselves are low income and/or disadvantaged; the aggregation of individual-level characteristics to the neighborhood level reflects the neighborhood composition. These two aspects of disadvantage are often inextricably linked: disadvantaged neighborhoods are typically peopled by individuals and families who are themselves disadvantaged, and neighborhoods comprised primarily of low-income families are typically characterized by diminished community resources and restricted opportunities (Sampson, 2001).

To appreciate what it means for a neighborhood to be disadvantaged, it is helpful to consider what constitutes advantage at the neighborhood level. Leventhal and Brooks-Gunn (2003) have identified six institutional neighborhood resources that are most important for child and youth development:

1. Learning activities (such as libraries, resource centers, museums);

2. Social and recreational activities (such as parks, sports programs, art programs);

3. Child care (particularly the availability, quality, and cost of child care);

4. Schools (particularly their quality, norms, and demographics);

5. Health care services (specifically the availability, quality, and affordability of medical and social services); and

6. Employment opportunities (e.g., for adolescents, supply of and access to jobs).

A neighborhood lacking a critical number of these resources could be considered disadvantaged.

Each of these resources can affect children's social-emotional development, either directly or indirectly. Neighborhoods with adequate learning activities will attract children in off-school hours, thereby both engaging them intellectually and keeping them under adult supervision and off the streets where they would be more likely to join in antisocial activities. Similarly, social and recreational activities provide children with typically adult-monitored opportunities to structure after-school and weekend times, again diminishing opportunities to engage in antisocial activities or risk behaviors. Quality child care for preschool children and quality schools and after-school programs for school-aged children foster both social-emotional as well as cognitive development. Adequate health care services can promote social-emotional development by providing mental health counseling and preventive services regarding risk behaviors such as drug use and sexual behavior. Finally, employment opportunities will motivate adolescents to stay in school and do well if doing so will increase their ability to attain and retain good jobs (Connell & Halpern-Felsher, 1997). Part-time or full-time employment for adolescents can be important not just for their own (and often their families') financial stability but also for fostering feelings of self-esteem, self-reliance, and a positive work ethic, attributes that together will decrease the likelihood of experiencing mental health problems or engaging in risk behaviors. (Yet, see recent studies linking hours worked and type of job with lower levels of academic achievement and higher levels of stress, risk behavior, and delinquency: Staff & Uggen, 2003; Weller et al., 2003). In general, the absence of youth-serving organizations, such as boys and girls clubs, Little Leagues, and YMCAs, in disadvantaged neighborhoods leave youth without constructive, organized, attractive ways to spend their time (Connell, Aber, & Walker, 1995).

In addition to a lack of institutional resources, disadvantaged neighborhoods are also afflicted with multiple environmental stressors. Typical among environmental stressors are population density, noise, housing problems, toxins and pollution, and violence (Evans, 2004; Evans & English, 2002; Perera et al., 2003). There is some evidence that the negative associations between family poverty and children's mental health and social behaviors are at least partially mediated by exposure to the stressors present in low-income neighborhoods (Evans & English, 2002; Evans, Lercher, Meis, Ising, & Kofler, 2000). The presence of neighborhood safety hazards (e.g., drive-by shootings, gangs, drug dealing) has been associated with increases in depression and anxiety as well as in oppositional defiant disorder and conduct disorder (Aneshensel & Sucoff, 1996). Adolescents in the lowest socio-economic neighborhoods report the highest rates of stress (Allison et al., 1999).

Neighborhood disadvantage has been consistently linked with increased levels of child and youth externalizing behaviors, including criminal and antisocial behaviors and drug use (e.g., Beyers, Bates, Pettit, & Dodge, 2003; Dubow, Edwards, & Ippolito, 1997; Eamon, 2001; Loeber & Witkstrom, 1993; Simons, Johnson, Beaman, Conger, & Whitbeck, 1996). One mechanism for this link may be through social cognitive processes: perhaps through associations with violence exposure, neighborhood disadvantage has been linked with adolescents' positive beliefs about aggression and in turn with their perpetration of violence (Halliday-Boykins & Graham, 2001).

Much of the research on neighborhood effects is completely correlational in design. But findings from two uniquely powerful research designs support the notion that neighborhood disadvantage causally influences children and youth. Research with two separate samples of twins has found that neighborhood deprivation significantly increases the risk that a child will display aggression, emotional problems, inattention problems, or conduct disorder over and above genetic contributions (Caspi, Taylor, Moffitt, & Plomin, 2000; Cleveland, 2003). In addition, experimental studies, in which families are randomly assigned to move to more advantaged neighborhoods, have demonstrated that adolescents who remain in high-poverty neighborhoods are more likely to be arrested for violent crimes than their peers who moved to low-poverty neighborhoods (Leventhal & Brooks-Gunn, 2000).

Neighborhood disadvantage, indexed as having high numbers of disadvantaged families (thus an aspect of neighborhood composition), has also been linked with child and youth outcomes. There is some evidence that relatively advantaged neighborhoods (i.e., middle-class) can buffer children against the negative impact of multiple risk factors (e.g., being a minority, being poor, and having a single parent) on their levels of aggression (Kupersmidt, Griesler, DeRosier, Patterson, & Davis, 1995). Possible explanations for this finding are that aggression is relatively more adaptive in poor and stressful neighborhoods or that these children have fewer successful adult and peer role models (Kupersmidt et al., 1995). Having affluent neighbors also has been associated with decreased rates of adolescent childbearing over and above socioeconomic characteristics of individual families (Brooks-Gunn, Duncan, Klebanov, & Sealand, 1993; Ku, Sonenstein, & Pleck, 1993).

The fact that neighborhood-level disadvantage is often confounded with family-level disadvantage does suggest that family or individual characteristics might account for what is typically chalked up to a neighborhood effect. For example, one multilevel analysis found no between-neighborhood differences (with neighborhoods characterized by disadvantage, stability and collective efficacy) on adolescent prosocial behavior or behavior problems, and rather that any neighborhood-level differences could be attributed to the clustering of individual-level factors within neighborhoods (Rankin & Quane, 2002). However, other studies have confirmed neighborhood effects distinct from family disadvantage effects. Sampson et al. (1997) found that collective efficacy at the neighborhood level was associated with less perceived violence in the neighborhood, over and above individual-level demographic risk factors. A multilevel study in the Netherlands determined that levels of child behavior problems were highest when families lived in deprived neighborhoods, over and above family socioeconomic status (SES; Kalff, Kroes, Vles, Hendriksen, Feron, Steyaert, van Zeben, Jolles, & van Os., 2001). It is most likely that both neighborhoods and families have separate and combined effects on children and youth. However, it is clearly important that neighborhood- and family-level disadvantage should always be analyzed conjointly in order to determine the unique contributions of each to child development.

Disadvantaged Schools

As with neighborhoods, school-level disadvantage also can be conceptualized as having both contextual and compositional features. At the contextual level, a key determinant of school advantage or disadvantage is per-student spending. Across the country, per-student expenditures are the result of a complicated formula that combines federal, state, and local tax-based funding. As a result, the range in per-student expenditures nationally is quite large. In the 1999-2000 school year, the national median total per student expenditure by school districts was $7,463; however, individual states show great variation in spending, with the median total per student expenditure in the highest-spending state, Alaska ($14,320), being more than twice that of the median expenditure in the lowest-spending state, Arkansas ($5,624; National Center for Education Statistics, 2003). Although some differences can be attributed to differences in cost of living (i.e., Alaska is a more expensive state in which to live or operate a school than Arkansas), these discrepancies also reflect the extent of state and local commitment to school funding.

At the compositional level of disadvantage, a human marker for under-resourced schools is the percentage of the student body that qualifies for free school lunch (in other words, those students whose families earn below the poverty line). It is a sad fact of the structure of school financing in the United States that high numbers of low-income children and low per pupil spending go hand in hand. Young children whose families are poor or low income are more likely to attend schools with higher proportions of inexperienced teachers and with fewer technological resources for learning compared than are non-poor children (Mayer, Mullens, & Moore, 2000). Schools comprised predominately of children from low-income families suffer disproportionately from disrepair, forcing children to attend class in schools with such problems as inadequate plumbing, leaky roofs, or dysfunctional heating and cooling systems (National Center for Education Statistics, 2000).

These disparities in resources are reflected in student achievement: Schools with the highest percentages of students eligible for free lunch have the lowest academic achievement (Bickel, Howley, Williams, & Glascock., 2000; Wirt et al., 2003). Students attending high poverty schools are least likely to report a positive attitude toward school, and parents at high poverty schools are least likely to participate in school activities (Wirt et al., 2003). Even more troubling is the finding that students who were not themselves eligible for free lunch had lower achievement if they attended schools in which more than 50% of their peers were eligible for free lunch than if they attended schools with 25% or fewer students eligible (Wirt et al., 2003). This finding suggests that a resource-poor school environment exacts a toll even among children who are otherwise relatively advantaged.

The financial resources of a school strongly determine its human resources, the most important of which are teachers. Available funds constrain both the quality of those teachers who can be hired (with quality indexed as their education level and amount of experience) and the number of teachers who can be hired (which affects student-to-teacher ratios). Indeed, schools with high proportions of low-income students, as well as those with high proportions of minority students, have a disproportionately high share of inexperienced and academically unprepared teachers (Ingersoll, 1999; Mayer et al., 2000). Not only do children learn better from and achieve more with teachers who have strong academic training and more teaching experience (Mayer et al. , 2000), but they behave better as well. Better educated and more experienced teachers constitute a human capital resource for students to draw upon and their students tend to have fewer behavior problems in students (Parcel & Dufur, 2001).

Outside of the quality of teachers, the sheer number of teachers in a school, and their ratio to students, has important implications for individual students' social-emotional behaviors. In order to adapt to a lack of adequate financing, schools and school districts often increase the sizes of schools as well as of individual classes. This trend is likely to continue: The U.S. Department of Education has forecast fairly steady increases in the number of children enrolled in elementary and secondary schools throughout the next decade (Gerald & Hussar, 2002). Without enhanced resources, such increases in enrollment can result in overcrowded schools and classrooms, a shortage of teachers, and aging and unsafe school buildings (U.S. Department of Education, 2000).

School size is implicated in how schools are organized, the curriculum that is offered, and how school members interact (Lee, 2000), each of which can impact students' mental health and behavior. Students at large schools have a more difficult time feeling part of a school community, while administrators at large schools have a harder time monitoring potentially disruptive student behavior (Flaherty, 2001). School size has been found to be particularly important for disadvantaged students, with positive attitudes toward school and achievement decreasing among disadvantaged students who attend very large or very small schools (Flaherty, 2001; Lee & Smith, 1997). Students at small schools (less than 500 students) report higher levels of social support and caring within their schools, whereas students at large schools (more than 1,000 students) report feeling a lack of community within their school (Lee, Smerdon, Alfeld-Liro, & Brown, 2000). However, there is a caution: Students may find it difficult to respect the requests and deadlines of teachers with whom they have come to know well, suggesting that even in small schools students need direction and advocacy from adults (Lee et al., 2000). Taking such various factors into account, statistical models have suggested that schools with enrollments of 600 to 900 students have the highest levels of achievement (Lee, 2000).

The effects of having more students in a school, and thus larger class sizes, has direct impacts on individual student behavior. From an economics perspective, classrooms with large numbers of students in relation to teachers result in resource dilution (Parcel & Dufur, 2001). Overcrowding increases the difficulty teachers have in establishing meaningful relationships with their students, in offering individualized attention to students, and in monitoring and disciplining student behavior (Flaherty, 2001; Flannery, 1997). Not surprisingly, the overcrowding of classrooms and schools has been linked with an increased likelihood of school violence (Flannery, 1997).

Exposure to Violence

Children's exposure to violence in its many forms, including in neighborhoods and schools, from families, and through the media, has become an increasing concern to parents, researchers, and policymakers, and indeed to children and youth themselves, over the past few decades. Exposure to non-familial violence is typically thought of as occurring at the community, or neighborhood, level, such as that in the form of observed muggings, attacks, and murders. However, a large portion of the non-familial violence that children and youth experience both as witnesses and as victims occurs in schools. In their landmark study of Washington, D. C., children, Richters and Martinez (1993c) found that a majority of witnessed violence occurs near children's homes but a substantial minority occurs in schools (68% and 22%, respectively); in contrast, violence victimization occurred roughly equally near home (48%) or at or near school (52%). In a more recent study in New York City, youth aged 12 to 20 reported that 57% of instances of seeing someone slapped, punched, or hit occurred at or near their schools and 47% of instances in which they themselves were slapped, punched, or hit occurred at or near their schools (Gershoff & Aber, 2003). These studies and others confirm that exposure to violence in schools should be of equal importance to parents, researchers, and policymakers as exposure to violence in neighborhoods.

Family-level disadvantage and violence and neighborhood-level disadvantage and violence are often closely intertwined (Aber, 1994). At the neighborhood level, poverty has been consistently linked with violent crime (Hsieh & Pugh, 1993). Accordingly, community violence is often thought to be a problem primarily for low-income, urban adolescents, a perspective reflected in the predominance of such samples in studies of exposure to violence (Flannery, 1997). Yet, national studies have confirmed that violence witnessing and victimization are facts of life for children and adolescents throughout the country (Addington, Ruddy, Miller, & DeVoe, 2002; Barton, Coley, & Wenglisky, 1998; Boney-McCoy & Finkelhor, 1995). Sadly, although exposure is more common among adolescents, exposure to violence is common in the lives of many young children as well (Osofsky, 1995).

One of the most important characteristics of violence exposure for child and youth development is its chronicity: children in violent schools or in violent neighborhoods are often exposed to multiple repeated violent or even traumatic events. Thus, violence becomes something that children assimilate into their views of themselves and others and that they adapt to through self-protective behaviors. Such chronic violence is more likely to have pervasive effects on child development than a single event such as a natural or terrorist disaster (Kupersmidt, Shahinfar, & Voegler-Lee, 2002).

This has indeed been found relative to the events of September 11th, 2001. In our study of New York City adolescents, exposure to the single-event of the terrorist attacks had relatively few effects on their mental health and behavior problems up to 2 years later relative to their exposure to community violence. Specifically, media exposure to the event was associated with heightened symptoms of post-traumatic stress, direct exposure to the event (e.g., seeing the planes hit, the towers fall, or smelling smoke) was associated with increased social mistrust, and having a family member at, injured in, or killed in the World Trade Center was associated with increased social mistrust (Aber, Gershoff, Ware, & Kotler, 2004). In contrast, chronic witnessing and being a victim of community violence had pervasive and very strong effects. Victimization by community violence was strongly associated with adolescents' symptom levels of depression, anxiety, conduct disorder, and post-traumatic stress disorder (PTSD); witnessing community violence was associated with symptom levels of depression, conduct disorder, and PTSD as well as with increases in hostile attribution bias and social mistrust (Aber et al., 2004). We concluded from these findings that chronic violence exposure has a greater impact on urban adolescents' symptomatology and social attitudes than exposure to the single traumatic event of September 11th. Although the services directed at children and youth in the wake of the terrorist attacks were important, clearly greater efforts are needed to assist urban children and youth in dealing with chronic violence exposure (Aber et al., 2004).

Processes By Which Violence Exposure May Affect Children and Youth

There are multiple ways in which exposure to violence may affect children's mental health as well as the likelihood that they will engage in antisocial or risk behaviors. One is through the impact of exposure on mental health at the individual level. Witnessing violent events can be traumatic particularly if the victim is a friend or family member, and thus can affect children's levels of anxiety, fear, and stress. As would be expected, victimization can have more direct links to individual trauma and stress. Exposure to violent interpersonal events, such as muggings, stabbings, shootings, rapes, murder, and family violence, can have tremendous negative effects on youth's mental health if they themselves are witnesses or victims (e.g., Allen, Jones, Seidman & Aber, 1998; Farrell & Bruce, 1997; Gorman-Smith & Tolan, 1998; Horn & Trickett, 1998).

Another avenue for the impact of exposure is through social information processing. At the individual level, internal cognitive processes likely play an important role in transforming the experience of violence at the community level into behavioral differences at the individual level. Coie and Dodge (1998) suggest that context can affect aggressive behavior by altering cognitive processes, such as by making children more likely to perceive threat or experience fear, to access aggressive responses to solve problems, and to view aggression as a desirable and effective means of achieving positive consequences. By providing models of violence as an effective problem solving strategy, and thus encouraging cognitive justifications for violence, or by requiring violence as a self-protective mechanism, violent neighborhoods increase the likelihood that youth will themselves engage in violence (Attar, Guerra, & Tolan, 1994). Indeed, exposure to severe community violence (e.g., shooting, stabbing, murder) has been associated with increased levels of hostile attribution bias and approval of aggression (Guerra, Huesmann, & Spindler, 2003; Shahinfar, Kupersmidt, & Matza, 2001) and in turn with increased child aggression both directly and mediated through positive beliefs about aggression (Guerra et al., 2003; Colder, Mott, Levy, & Flay, 2000). In an interesting example of domain specificity, witnessing and victimization have been linked to aggressive behavior through distinct mediational processes: witnessing through social-cognitive biases, and victimization through emotion dysregulation (Schwartz & Proctor, 2000).

Violence in Neighborhoods

Characterizing Neighborhood Violence. Since the appearance of Richters and Martinez's (1990) exposure to violence measure, the majority of research on exposure to violence has distinguished violence that is witnessed from violence that is directly experienced as a victim. The pervasive nature of violence in many neighborhoods is such that exposure through witnessing tends to be strongly correlated with exposure through victimization (e.g., Allen et al., 1998; Schwartz & Proctor, 2000), although children are more often witnesses than victims (Richters & Martinez, 1993b). While the potentially traumatic nature of victimization is obvious, it is perhaps less understood that witnessed violence also can be quite traumatic and have mental health consequences. Indeed, witnessed violence is often referred to as "covictimization" (Shakoor & Chalmers, 1991).

The one common thread through the literature on neighborhood violence exposure in the United States is that rates of exposure among children have been and continue to be distressingly high. Homicide is the second leading cause of death for youth aged 15-24 and the fourth leading cause of death for children aged 1-14 (Minino & Smith, 2001). Nationally, 35% of children and adolescents reported being a victim of violence at some point in their lives and an additional 5% reported an attempted victimization (Boney-McCoy & Finkelhor, 1995). On Chicago's South Side in 1987-1988, 24% of 10- to 19-year-olds had witnessed a murder, 45% reported seeing more than one type of violence, between 40% and 56% of witnessed crimes involved a familiar victim, and 47% had been a victim of a violent crime in the previous year (Uehara, Chalmers, Jenkins, & Shakoor, 1996). In their study of children in Washington, D.C., Richters and Martinez (1993b) found that among first and second graders, 61% had witnessed violence and 19% had been a victim, while among fifth and sixth graders, 72% had witnessed violence and 32% had been victimized. In New Haven in 1994 and 1996, 36% of adolescents had experienced at least one type of violent act, while on average, each had witnessed nearly three different types of violence (Schwab-Stone et al., 1999).

Positive Affirmations Youth
Figure 25.1 Percent of New York City youth reporting witnessed community violence (2002-2003) From Gershoff, Pedersen, Ware, and Aber (2004).
2003). From Gershoff, Pedersen, Ware, and Aber (2004).

Startlingly, rates of community violence exposure do not appear to be diminishing. In our study of over 900 primarily minority, low-income New York City adolescents from 2002 to 2003, 98% reported witnessing at least one form of violence and 87% reported being a victim of at least one form of violence (Gershoff, Pedersen, Ware, & Aber, 2004). Figure 25.1 displays the rates of specific forms of witnessed violence; most youth reported seeing someone slapped, hit, or punched or seeing someone arrested, but a distressingly high 14% of youth reported seeing someone shot at and 12% reported seeing someone killed. Victimization rates are displayed in Figure 25.2; over half had been slapped, hit, or punched themselves, a third had been asked to sell, buy, or use drugs, and a fifth had been threatened with serious bodily harm. Thirteen percent had been themselves beaten up or mugged and 6% had been sexually assaulted, molested, or raped. By any measure, the rates at which youth in New York City continue to be exposed to community violence as witnesses or victims remain alarmingly high and demand increased efforts at both prevention of violence in communities and mental health service intervention to witnesses and victims.

Several demographic characteristics of children have been identified as covariates of violence exposure, although it is important to keep in mind that several of these covariates are potentially confounded with family and neighborhood socio-economic status:

• Children aged 12 to 17 years old are twice as likely to be victims of serious violent crime and three times as likely to be victims of assault than adults (Office of Juvenile Justice and Delinquency Prevention, 2000).

• Minorities are at greater risk for violence exposure than Whites (Eitle & Turner, 2002; Lauritsen, 2003; Salzinger, Feldman, Stockhammer, & Hood., 2002; Stein et al., 2003).

• The majority of violent events experienced by adolescents (53%) occur within 1 mile of their homes (Lauritsen, 2003).

• Boys tend to report more violence exposure than girls, except in the case of sexual violence (Eitle & Turner, 2002; Kuo, Mohler, Raudenbush, & Earls, 2000; Lauritsen, 2003; Myers & Thompson, 2000; Salzinger et al., 2002; Stein et al., 2003).

• The afternoon, particularly after school, hours are the most dangerous for children (Snyder & Sickmund, 1999).

• Living in urban and/or disadvantaged neighborhoods puts children at increased risk of violence exposure (Aneshensel & Succo, 1996; Lauritsen, 2003; see Salzinger et al., 2002, and Stein et al., 2003, for reviews).

• Youth living in single-parent homes are 50% more likely than youth living in two-parent homes to be victims of violence in general and 200% more likely to be victims in their neighborhoods, most likely because youth in single parent families are more likely to live in impoverished and violent neighborhoods (Lauritsen, 2003).

Identifying Impacts of Neighborhood Violence. At the psychobiological level, a review of the literature found that some studies report hypoarousal among children exposed to violence whereas others report hyperarousal (Lynch, 2003); this inconsistency remains to be resolved but has important implications for whether internalizing or externalizing symptoms might be expected as outcomes of violence exposure. Until it is resolved, both types of outcomes might be expected.

Exposure to violence generally, and victimization and witnessing in particular, has been linked with a range of internalizing mental health problems. Foremost among these is PTSD, with several studies finding links between exposure to violence and increased PTSD symptomatology (Aber et al., 2004; Berman, Kurtines, Silverman, & Serafini, 1996; Berton & Stabb, 1996; Boney-McCoy & Finkelhor, 1995; Ceballo, Dah, Aretakis, & Ramirez, 2001; Hoven et al., 2004; Lynch, 2003; Myers & Thompson, 2000; Schwartz & Proctor, 2000). A range of other internalizing mental health problems have been linked to violence exposure, including anxiety, depressive symptoms, somatization (Aber et al., 2004; Hoven et al., 2004; Lynch, 2003; Moses, 1999; Richters & Martinez, 1993a; Schwab-Stone et al., 1999), suicidality (Elze, Stiffman, & Doré, 1999), hopelessness, and a heightened belief that they would have a violent death (Hinton-Nelson, Roberts, & Synder, 1996).

Several of these psychological reactions to violence exposure may act in combination, in other words be comorbid with each other. In particular, the associations of exposure to violence to both depressive symptoms and suicidal ideation have been found to be mediated by PTSD symptoms in children (Mazza & Reynolds, 1999), and comorbid separation anxiety symptoms and PTSD symptoms have been associated with exposure to the violent events of September 11th, 2001 (Hoven et al., 2004). Similarly, intrusive thinking has been found to partially mediate the association between victimization and internalizing (Kliewer, Lepore, Oskin, & Johnson, 1998).

The bulk of research on exposure to neighborhood violence has repeatedly found links between levels of exposure and levels of externalizing problem behaviors such as aggression or antisocial behavior (for reviews see: Overstreet, 2000; Salzinger et al, 2002; for examples see: DuRant, Cadenhead, Pendergrast, Slavens, & Linder, 1994; Eitle & Turner, 2002; Lynch, 2003; Schwab-Stone et al., 1999). In a study of youth on Chicago's South Side, 86% of youth who reported perpetrating violence were themselves witnesses and victims of violence (Uehara et al., 1996). Several studies have determined that neighborhood violence predicts externalizing problem behaviors including violence perpetration, over and above demographic and family psychosocial risks (such as parent depression and marital conflict) (Ceballo et al., 2001; Greenberg, Lengua, Coie, & Pinderhughes, 1999).

The few studies that have examined positive developmental outcomes have found that exposure to neighborhood violence tends to have a dampening effect. In the few studies that have examined competence, neighborhood violence exposure was associated with lower social competence (Cooley-Quille, Turner, & Biedel 1995; Greenberg et al., 1999; Schwartz & Proctor, 2000) and with declines in academic performance (Lynch, 2003).

Given the amount of violence in their communities and the potentially drastic effects it can have on their development, how do children cope with exposure to this violence? Youth growing up in violent neighborhoods are repeatedly exposed to challenges that are outside their control and that overwhelm their coping abilities, such that coping relying on either aggression or depression function as viable short-term tactics (Blechman, Dumas, & Prinz, 1994). In response to exposure to and fear of community violence, adolescents have reported engaging in such protective behaviors as avoiding certain parks, carrying weapons as self-defense, staying home from school, ceasing to engage in a particular activity or sport, or finding someone to protect them (Arnette & Walsleben, 1998).

Of course, it is possible that some third factor, correlated with neighborhood violence, is the causal factor in the found associations between exposure to violence and negative outcomes for children and youth. Similarly, it could be that poor and violent neighborhoods "attract" more vulnerable children and families to be residents. These issues of causality and direction of effect have plagued the neighborhood literature for decades. Only social experiments that randomly assign neighborhoods in experimental designs (e.g., the Moving to Opportunity study, discussed below: Katz, Kling, & Liebman, 2003; Ludwig, Duncan, & Ladd, 2003) or advances in the causal analysis of non-experimental data (e.g., instrumental variable analyses) hold promise in untangling various links in the causal chains. Of course, this is not simply a matter of scientific or scholarly concern. Understanding the underlying causal reality of neighborhood effects has profound implications for the design and implementation of preventions, interventions, and policy strategies.

Violence in Schools

Characterizing School Violence. The extent of violence in schools has been a persistent concern for parents, administrators, and policymakers in recent decades. The most recent data on crimes in schools (from the School Crime Supplement to the National Crime Victimization Survey) indicate that 6% of 12- to 18-year-olds reported being victims of nonfatal crimes at school during the previous 6 months (DeVoe et al., 2003). Boys, middle school students, and students at public schools were more likely to report being victims at school. Tellingly, during the period of 1992 to 2001, adolescents aged 12 to 18 were 31% more likely to be victims of theft and 300% more likely to be victims of a violent crime at or on the way to school than away from school (DeVoe et al., 2003).

One important source of violence in many schools (and neighborhoods, for that matter) is gangs. Students who report gangs at their school are almost twice as likely to experience violent or property victimization than students who do not report gangs in their school (18.4% vs. 10.8%; Addington et al., 2002), despite the fact that schools with gangs have the greatest number of security measures (Howell & Lynch, 2000). Gang presence is most likely in schools with the lowest family incomes and tends to go hand in hand with the availability of drugs in a school (Howell & Lynch, 2000).

The availability and selling of drugs at school and school violence are often co-occurring problems. As antecedents to problem behavior, the consumption of alcohol and drugs can lead to an overall harmful school environment (Fagan & Wilkinson, 1998). Twenty-nine percent of students report that someone has offered them drugs while at school (Centers for Disease Control and Prevention, 2002a). Five percent of students report drinking on school property and 5% report using marijuana at school (DeVoe et al., 2003). Another pernicious, and often overlooked, source of violence and victimization is bullying, which involves a powerful person repeatedly oppressing a less powerful person through psychological or physical means (Farrington, 1993; Olweus, 1995). Eight percent of 12- to 18-year-olds nationally reported having been bullied in the last six months; in a trend similar to that of overall victimization cited above, boys, middle school students, and public school students were most likely to be bullied (DeVoe et al., 2003). Bullying is reflective of, and contributes to, a school environment characterized by fear and intimidation (Arnette & Walsleben, 1998).

The majority of schools throughout the country have problems with violent victimization. Seventy one percent of public schools reported one or more violent incidents at their schools (DeVoe et al., 2003). Somewhat surprisingly, students at rural and suburban schools were as likely to be victims of theft or violent victimization as students in urban schools in 2001 (DeVoe et al., 2003). What does make a difference is school size: Large schools (those with 1,000 or more students) are 46% more likely to report violent incidents than schools with less than 300 students (DeVoe et al., 2003).

Of course, violence against students in schools tends to be perpetrated by other students. Eighteen percent of boys and 7% of girls in a national sample reported having been in a physical fight on school property (DeVoe et al., 2003). On average, 7% of students carried a weapon to school and 16% got into physical fights on school property (Centers for Disease Control and Prevention, 2002b). Such fights occur even when students are aware of their dangers: In a sample of middle school students, 20% reported getting into regular physical fights, despite the fact that nearly three quarters of the sample thought that fighting was a risky behavior (St. George & Thomas, 1997).

Similar to the finding that crime in neighborhoods is concentrated at certain locations, or "hotspots" (Sherman, Gartin, & Buerger, 1989), researchers have identified specific physical locations within schools that are prone to violence. School violence is often time- and place-specific. Not surprisingly, it tends to occur before or after school or during transitions between classes, and to occur in either crowded places (such as hallways, cafeterias, or playgrounds) or in unsupervised places (such as bathrooms or staircases) (Flaherty, 2001; Garofalo, Siegel, & Laub, 1987; see Astor & Meyer, 2001, for review). In addition, violence is more likely in schools that are large, are overcrowded, are poorly organized, or have poor resources (Astor & Meyer, 2001; Flannery, 1997; Warner, Weist, & Krulak, 1999). Schools characterized by fear of violence and crime will have difficulty recruiting and retaining good teachers, and the teachers who remain will be less likely to confront misbehaving students out of fear of their own safety, leading to disruptive classrooms with fewer opportunities for learning (Flannery, 1997).

Although we know that children are more likely to be exposed to violence if they live in violent and disadvantaged neighborhoods, we know little about the extent to which violence exposure depends on the characteristics of the neighborhoods in which children attend school. There is some evidence that community violence begets school violence (Sheley et al., 1992); the overlap of neighborhoods and schools, particularly with regard to exposure to violence, will be discussed further below.

Identifying Impacts of School Violence. As with community violence, exposure to school violence either as a witness or a victim can exact a toll on students' mental health. Among both elementary and high school students, violence victimization in schools has been linked with greater anxiety, depression, PTSD symptoms, and general trauma, as well as to violent behavior (Berton & Stabb, 1996; Singer, Anglin, Song, & Lunghofer, 1995). Students who report that their schools are dangerous have a decreased sense of school-based self-efficacy (Bowen, Richman, Brewster, & Bowen, 1998).

Exposure to violence in schools has often been linked with a particular form of anxiety: fear and avoidance of school. In 2001, 6% of students reported that they were fearful about their safety at school either sometimes or all of the time (DeVoe et al., 2003). Feeling that school or the neighborhood around school is unsafe led 2% of students skip school out of fear of victimization, a rate that increases to 4% among previously victimized students and 7.5% among bullied students (Addington et al., 2002). Students who report more violence and drug activity in their schools are also likely to have heightened fear of victimization and to be more vigilant about places in and around school where violence may occur (Mayer & Leone, 1999).

Sense of Community and Support Networks

The need to belong to a group or community motivates much of human activity, with acceptance by a group yielding positive emotions and experiences and rejection by a group often precipitating stress, anxiety, depression, behavioral problems, and drug abuse (Baumeister & Leary, 1995). For children and youth, two of the most salient communities to which they belong are their home neighborhoods and their schools (sports teams, school clubs, and religious organizations constitute other important communities for children and youth). The extent to which children feel a part of these communities, the level at which they feel supported by them, and the norms endemic to these communities can each impact children's mental health and the extent to which they engage in risk behaviors.

Adolescents in particular are at the stage of early development most conducive to neighborhood and school effects. Increased autonomy that comes with age as well as preparation for work and life independent from parents propel adolescents into new and varied contexts within their neighborhoods. As they age out of childhood and toward adulthood, adolescents spend less time with parents in the home and more time with peers out of the home (Connell & Halpern-Felsher, 1997); whereas when they were younger much of neighborhood effects were mediated through their parents, adolescents' exposure to their neighborhoods becomes more direct (Aber et al., 1997). We continue to discuss neighborhood and school impacts on all children, while keeping in mind that much of the literature to be reviewed focuses on adolescents.

Neighborhood as Community

We have discussed how tangible aspects of neighborhoods, such as resources and levels of violence, affect children. Now we turn to the social context of neighborhoods. Beliefs, attitudes, and norms that are held by a majority of individuals in a community can influence the developing beliefs, values, and norms held by children and youth. These, in turn, can directly affect child and youth behavior. For example, community norms regarding crime and violence, in addition to overall levels of crime and violence, determine whether youth engage in antisocial behavior (Williams, Ayers, & Arthur, 1997). Youth living in high-poverty urban neighborhoods are less likely to participate in community service, have less civic knowledge, and are less politically tolerant than youth living in low-poverty neighborhoods (Atkins & Hart, 2003), findings which suggest that adolescents in disadvantaged neighborhoods are likely to feel estranged from their neighborhoods and communities. We will focus here on two important ways in which neighborhood norms can affect children, namely through feelings of neighborhood collective efficacy and through peers' behavior and norms (Leventhal & Brooks-Gunn, 2003).

Collective efficacy is a relatively new rubric for understanding the processes by which neighborhoods affect children and families. Collective efficacy grows out of mutual trust among neighbors and their willingness to intervene and exercise social control, such as through the supervision of children and youth and the maintenance of public order (Sampson, Raudenbush, & Earls, 1997). It thus extends the notion of social capital, or relationships among persons that facilitate action (Coleman, 1988). Neighborhoods high in social capital/collective efficacy are better able to realize common values and to maintain social controls (Sampson, 2001).

Collective efficacy is distinct from, though potentially related to, the demographic characteristics of the neighborhood's population: concentrated disadvantage, immigration concentration, and residential stability have been found to account for 70% of the neighborhood variation in collective efficacy (Sampson et al., 1997). There is some evidence that the effects of neighborhood disadvantage on adolescent behavior are mediated through the levels of neighborhood-level informal control (Elliott et al., 1996). Children who live in disadvantaged neighborhoods and who report low levels of collective socialization (a similar concept to that of collective efficacy) in those neighborhoods are more likely to associate with deviant peers (Brody et al., 2001).

In the Project on Human Development in Chicago Neighborhoods, Sampson and colleagues (1997; Sampson, 2001) have found that collective efficacy was negatively associated with community violence and hypothesize that collective efficacy plays a key role in monitoring and controlling adolescent peer groups (Sampson & Groves, 1989). Adolescents in neighborhoods characterized by low collective efficacy have more unconventional friends and, in turn, are less prosocial and more likely to have behavior problems (Rankin & Quane, 2002).

By virtue of the fact that children are most likely to be interacting with peers in neighborhood contexts, peers are a main way that children and adolescents experience the "community" of a neighborhood. Peers constitute social support networks for other children; however, they are not always "supportive" of positive adjustment. Negative peer norms, or peer culture, can dispose children to engage in antisocial or delinquent behavior. Aspects of negative peer culture include the extent to which youth in one's neighborhood get in trouble with the police, use drugs or alcohol, or join a gang (Bowen & Bowen, 1999). Children's perceptions of a negative neighborhood peer culture and reports of personal threats received in neighborhoods both are associated with their trouble behavior in school (e.g., being sent out of class for behavior, having notes sent to parents about behavior, being suspended or expelled; Bowen & Bowen, 1999). When neighborhood-based institutions and norms fail to regulate peer group behavior, peer influence is exacerbated and is typically negative, increasing delinquent and problem behavior (Elliott et al., 1996; Sampson & Groves, 1989). This is sharply illustrated by the finding that in high-risk neighborhoods, peer support increases antisocial and risky behavior, whereas in low-risk neighborhoods peer support decreases such negative behaviors (Dubow et al., 1997; Gonzalez, Cauce, Friedman, & Mason, 1996).

School as Community

The major components of a school community are students' relationships with teachers as well as their relationships with fellow students (Osterman, 2002). Having a sense of belonging to a supportive school community has been associated positively with emotional well-being, intrinsic motivation, and prosocial behavior, as well as with lower levels of emotional distress and suicidal behaviors (Osterman, 2002; Resnick et al., 1997). In contrast, feelings of rejection or not "fitting in" at school are associated with behavioral problems in school, school avoidance, and school drop-out as well as with mental health problems including anxiety, distress, substance use, sexual activity, violence, and suicide (Osterman, 2002). If students perceive a mismatch between the resources and characteristics of their schools and their own needs (such as those for support or for autonomy), they may feel angry and resentful and act out as a result (Roeser, Eccles, & Strobel, 1998). Unfortunately, those students most in need of a supportive school community may be least likely to experience it; students at low-income schools are more likely to experience them as uncaring (Watson, Battistich, & Solomon, 1997).

Teachers' and administrators' attitudes and expectations about ability and achievement can affect students' performance in school as well as their mental health. Students who attend schools that have challenging curricula and that have high expectations for their students in turn have high engagement in and achievement at school (Mayer et al., 2000). When students are engaged in and value school, they will be less disposed toward disruptive and maladaptive behaviors (Dishion, Patterson, Stoolmiller, & Skinner, 1991). A classroom climate characterized by an emphasis on internal control and personal responsibility has been associated with greater individual student self-esteem (Ryan & Grolnick, 1986). School and classroom characteristics such as teachers' use of praise, a comfortable environment for students, enhanced opportunities for student responsibility and participation, and an academic emphasis together have been found to promote positive student behavior (Rutter, 1980). Absenteeism and drop-out rates are lower in schools characterized by an emphasis on academics, an orderly environment, and the presence of faculty more engaged with students (Bryk & Thum, 1989). In schools that emphasize innate ability as a determinant of academic ability, children without such ability may develop internalized feelings of low self-esteem or anxiety (Roeser et al., 1998). Middle school students who perceive that their school emphasizes competition and ability and that their teachers treat students differentially based on race or gender also report declining mental health over time (Roeser, Eccles, & Sameroff, 2000).

At the classroom level, teacher characteristics, the instructional climate, social relations within the classroom, and student aggregate characteristics are thought to affect children's adjustment (Perry & Weinstein, 1998). Teachers' levels of expectations, teachers' modeling of appropriate behavior, and teachers' reinforcement of appropriate student behavior are key to encouraging and managing appropriate student behavior, but if these norms are not consistent throughout the school or are not accepted by the students, their potential positive impact is diminished (Rutter, 1980). The social climate of a classroom, including average levels of disruptiveness, adjustment, and achievement, can affect individual behavior and achievement. Furthermore, individual students in classrooms with greater numbers of students with behavior problems are more likely to exhibit shy behaviors (Werthamer-Larsson, Kellam, & Wheeler, 1991).

General attitudes and norms of the student body can also affect individual mental health and risk behavior. Adolescents' perceptions of their school's norms regarding drug use are predictive of their use of illegal substances (Allison et al., 1999). Similarly, aggregated at the school level, students' endorsement of a subculture of violence is associated with more, and their endorsement of academic values with less, interpersonal violence, theft, and delinquency (Felson, Liska, South, & McNulty, 1994).

How To Win Your War Against Anxiety Disorders

How To Win Your War Against Anxiety Disorders

Tips And Tricks For Relieving Anxiety... Fast Everyone feels anxious sometimes. Whether work is getting to us or we're simply having hard time managing all that we have to do, we can feel overwhelmed and worried that we might not be able to manage it all. When these feelings hit, we don't have to suffer. By taking some simple steps, you can begin to create a calmer attitude, one that not only helps you feel better, but one that allows you the chance to make better decisions about what you need to do next.

Get My Free Ebook

Post a comment