Tama Leventhal1 and Jeanne Brooks-Gunn
National Center for Children and Families, Teachers College, Columbia University, New York, New York
Neighborhoods are increasingly studied as a context where children and youth develop; however, the extent of neighborhoods' impact remains debatable because it is difficult to disentangle this impact from that of the family context, in part, because families have some choice as to where they live. Evidence from randomized experiments, studies using advanced statistical models, and longitudinal studies that control for family characteristics indicates that neighborhoods do matter. In nonexperimental studies, small to moderate associations were found, suggesting that children and adolescents living in high-income neighborhoods had higher cognitive ability and school achievement than those living in middle-income neighborhoods, and children and adolescents living in low-income neighborhoods had more mental and physical health problems than those living in middle-income neighborhoods. The home environment has been shown to be partly responsible for the link between neighborhood and children’s development. For adolescents, neighborhood effects are partially accounted for by community social control. Experimental studies in which families were randomly assigned to move to low-poverty neighborhoods from housing projects found larger neighborhood effects than nonexperimental research, particularly for boys’ outcomes. Additional issues reviewed are relevant neighborhood characteristics, theoretical models explaining the pathways underlying neighborhood effects, methods for research assessing neighborhood processes, and policy implications.
neighborhood; community; achievement; health; income/socioeconomic status; policy
Historical trends document the declining economic conditions in which children grow up. Compared with their predecessors, children today are more likely to be raised in poor families (i.e., those whose incomes fall below a federally established threshold), as well as to live in poor neighborhoods (i.e., 20% or more of residents poor). Almost half of poor families reside in urban neighborhoods that are increasingly marked by concentrated poverty. Both family and neighborhood poverty are rooted in demographic shifts in family composition and labor-force participation, changes in migration and residential patterns, declines in industrialization, and housing segregation (Massey & Denton, 1993; Wilson, 1987). Responding to these developments, both academic scholars and policymakers developed a rising interest in the contexts in which children are reared, including larger social environments beyond the family, notably neighborhoods. By the mid-1980s, questions such as the following were raised: Does neighborhood residence influence children’s well-being? How do neighborhoods affect children and youth? What can be done to alleviate neighborhood disadvantage and its potentially harmful effects on children’s development? In response to these questions, scholars from various disciplines—economics, epidemiology, demography, sociology, and psychology—launched a field of study that has become known as neighborhood research (Brooks-Gunn, Duncan, & Aber, 1997).
Most neighborhood research has used census-based measures of neighborhood structural or sociodemographic characteristics in conjunction with data collected on children and families, often from large national data sets (e.g., Panel Study of Income Dynamics, National Longitudinal Survey of Youth-Child Supplement), to examine associations among neighborhood residence and child and adolescent outcomes. Data from the U.S. Census come from the forms the population fills out on April 1 of the 1st year of every decade. Thus, census information is limited to structural characteristics, such as median household income, percentage of residents with a high school diploma, racial composition, and percentage of homeowners. The census tract is the most frequently used definition of “neighborhood” in these studies. Tract boundaries are identified by local communities working under Census Bureau guidelines and reflect salient physical and social features that demarcate neighborhoods, such as major streets, railroads, and ethnic divisions; census tracts contain approximately 3,000 to 8,000 individuals.
Neighborhood income or socioeconomic status (SES)—a combination of social and economic indicators—is the most commonly investigated neighborhood characteristic. In these studies, researchers often use two separate measures of SES, because the presence of poor and affluent neighbors may have differential associations with child and adolescent outcomes (Jencks & Mayer, 1990). High-SES/affluence measures may take into consideration indicators such as income, percentage professionals, and percentage of residents who are college educated; low-SES/poverty measures may take into consideration indicators such as percentage poor, percentage of households headed by females, percentage on public assistance, and percentage unemployed.. Other structural characteristics frequently examined include racial-ethnic mix (e.g., percentage Black, percentage Latino, and percentage foreign-born) and residential instability (e.g., percentage moved in last 5 years, percentage of households in their current home less than 10 years, and percentage renters).
Studies investigating neighborhood effects on children’s development also account for family characteristics, such as income, composition, and parents' education, age, and race or ethnicity, to demonstrate whether neighborhood effects go “above and beyond” family influences. Because families have some choice as to where they live, adjusting for these background factors also minimizes the possibility that unmeasured individual and family characteristics associated with neighborhood residence (i.e., selection bias) might account for observed neighborhood effects. Some researchers also have addressed selection problems by using various advanced analytic strategies, such as comparisons of siblings or first cousins, which hold family characteristics constant; instrumental variable analyses, which minimize unmeasured correlations between neighborhood characteristics and child outcomes; and behavior genetics models; which differentiate between genetic and environmental influences.
We recently conducted a large-scale review of the neighborhood research (Leventhal & Brooks-Gunn, 2000). Findings reported in that review as well as subsequent work revealed consistent patterns of neighborhood effects on children’s and adolescent’s development; comparable finding have been documented in U.S. and Canadian samples and in cross-sectional and longitudinal studies that control for family characteristics. Across these studies, neighborhood effects were small to moderate in size. For preschool and school-age children, the presence of affluent or high-income neighbors was positively associated with children's verbal ability, IQ scores, and school achievement. In contrast, the presence of low-income or low-SES neighbors was associated with children’s mental health problems. For adolescents, living in a high-income neighborhood was also associated with high school achievement and educational attainment, particularly for males. Residence in low-income neighborhoods was associated with adverse mental health, criminal and delinquent behavior, and unfavorable sexual and fertility outcomes for adolescents. For both children and adolescents, these patterns of results also have been found in studies employing advanced statistical techniques, although effect sizes are typically reduced.
Although nonexperimental neighborhood research has yielded fairly consistent patterns of findings, it has been criticized on the grounds that families have some choice as to the neighborhoods in which they live, resulting in selection bias (even after accounting for family characteristics). Experimental and quasi-experimental studies that randomly assign families to live in certain types of neighborhoods overcome the selection problem in nonexperimental research. Although such designs may seem implausible, it has been possible in the context of housing programs that randomly select families for assistance in relocating from public housing to less poor neighborhoods (e.g., they may receive vouchers to rent housing in the private market or be offered public housing built in nonpoor neighborhoods).
The oldest and most well known quasi-experimental study is the Gautreaux Program, enacted following a court order to desegregate Chicago’s public housing. Families were given vouchers to move, and assignment of families to neighborhoods was random, based on housing availability (see Rubinowitz & Rosenbaum, 2000, for a review). A 10-year follow-up found that youth who moved to more affluent suburban neighborhoods fared better academically than youth who moved to poor urban neighborhoods.
In 1994, partially in response to positive findings reported in the Gautreaux Program, the U.S. Department of Housing and Urban Development initiated the Moving to Opportunity Program in five cities across the country. Approximately 4,600 families were randomly assigned vouchers to move out of public housing in high-poverty neighborhoods into private housing of their choice or into private housing in low-poverty neighborhoods (with special assistance); a subset of families remained in public housing. Although initial evaluations were conducted independently in each city, there is some overlap in the outcomes examined in different cities, so that it is possible to draw some preliminary conclusions from this research (Goering, in press). Findings from these experimental studies revealed that several years into the program, children and youth who moved to less poor neighborhoods had higher educational achievement and superior physical and mental health than their peers who remained in high-poverty neighborhoods (Katz, Kling, & Liebman, 2001; Leventhal & Brooks-Gunn, 2002, in press). In addition, arrests for violent crime were lower among male youth who moved to less poor neighborhoods compared with peers who stayed in high-poverty neighborhoods (Ludwig, Duncan, & Hirschfield, 2001). Neighborhood effects in these experimental studies were large. In addition, larger effects were generally seen for children and youth who moved to low-poverty neighborhoods than for those who moved to moderately poor neighborhoods , and effects were more pronounced for boys than girls.
The experimental and nonexperimental studies we have reviewed illuminated specific neighborhood characteristics that were associated with particular outcomes. These studies, however, do not address the mechanisms through which neighborhood effects occur. We have proposed several theoretical models to explain potential pathways of neighborhood influences (Leventhal & Brooks-Gunn, 2000). These models highlight different underlying processes operating at various levels (individual, family, school, peer, and community). This work draws heavily from a review and analysis by Jencks and Mayer (1990); from research on family stress, economic hardship, and unemployment; and from literature on community social organization and urban sociology.
The first model, institutional resources, posits that neighborhood influences operate by means of the quality, quantity, and diversity of learning, recreational, social, educational, health, and employment resources in the community. The second model, relationships and ties, highlights families as a potential mechanism of neighborhood effects. Important variables in this model include parental attributes (e.g., mental and physical health, coping skills, and efficacy), social networks, and behavior (e.g., supervision-monitoring, warmth, and harshness), as well as characteristics of the home environment (e.g., learning and physical environments, family routines, and violence). The last model, norms and collective efficacy, hypothesizes that neighborhood influences are accounted for by the extent of formal and informal social institutions in the community and the degree to which they monitor or control residents' behavior in accordance with socially accepted practices and the goal of maintaining public order (Sampson, Raudenbush, & Earls, 1997). This model includes influences such as peer groups and physical threats in the neighborhood (e.g., violence, availability of illegal and harmful substances).The models are intended to be complementary rather than conflicting, with the utility of each model for explaining neighborhood effects on children's well-being depending on both the particular outcome studied and the age group examined.
An emerging body of research, focused largely on adolescent problem behavior, substantiates the norms and collective efficacy model. At both the community and the individual levels, mechanisms of social control have been found to account, in part, for associations among neighborhood structure and rates of problem behavior among adolescents (e.g., Elliott et al., 1996; Sampson et al., 1997). There has been scant research relevant to the other models, although several studies of young children support the relationships-and-ties model; quality of the home environment was found to partially account for associations between neighborhood structure and children’s achievement and behavioral outcomes (Klebanov, Brooks-Gunn, McCarton, & McCormick, 1998).
Evidence from randomized experiments, studies employing advanced statistical models, and longitudinal studies controlling for family characteristics indicates that neighborhoods, and particularly their socioeconomic composition, do matter. The size of neighborhood effects reported in the nonexperimental literature has typically been small to modest (after background characteristics are accounted for). However, neighborhood effects reported in the limited set of experimental studies were large, likely because the changes in neighborhood conditions were substantial ( when families initiate their own moves, the changes are usually not so large, particularly among low-income families). To determine whether nonexperimental research is underestimating neighborhood effects, it will be necessary to try to replicate the experimental findings by undertaking natural studies where radical changes in neighborhood economic conditions have occurred. In addition, suggestive evidence from nonexperimental studies reveals that neighborhood residence may be differentially associated with outcomes for Latinos compared with European and African Americans, pointing to acculturation as a potentially important and unexplored variable moderating the effects of neighborhood structure.
The impact of neighborhood residence is also likely to vary across development; however, because much of the neighborhood research is cross-sectional or based on neighborhood residence at a single point in time, this issue has not been adequately addressed. This relatively static view of neighborhood influences extends to neighborhood conditions. Researchers often ignore the fact that families move across neighborhoods and that even when families do not move, neighborhood structure changes, for example, through gentrification and immigration (Leventhal & Brooks-Gunn, 2001).
What else remains unclear is how neighborhoods matter. Our proposed models—institutional resources, relationships and ties, and norms and collective efficacy—provide a framework intended to aid empirical investigations of theoretically driven neighborhood research. To test theoretical models, it is necessary to move beyond census measures of SES and directly assess underlying processes. Alternative data sources are required to measure neighborhood-level processes, in particular. Useful data will come from (a) city, state, and federal records (e.g., vital statistics from health departments, crime reports from police departments, school records from education departments, and child abuse and neglect records from social service departments); (b) systematic social observations by trained observers using a structured format to characterize neighborhoods along a range of social and physical attributes; (c) community surveys in which non-study participants (i.e., an independent sample) are interviewed about their neighborhoods (usually about their neighborhoods' social organization); and (d) neighborhood-expert surveys in which key community leaders are interviewed about neighborhood political and social organization (see Leventhal & Brooks-Gunn, 2001, for additional details). Recent studies designed specifically with neighborhood influences in mind are collecting longitudinal and process-oriented data on children, families, and neighborhoods and are well suited for exploring mechanisms through which neighborhoods influence child well-being ,as well as addressing dynamic models of neighborhood influences.
Several policy implications may be drawn from the existing neighborhood research. The research suggests that it would be beneficial to develop programs that foster moving poor families out of poor neighborhoods. In line with this goal are efforts to reduce housing discrimination in nonpoor neighborhoods. What Moving to Opportunity and other such programs have demonstrated is that without special assistance, poor families who are given vouchers do not necessarily move out of poor neighborhoods. A complementary approach is to build scattered-site public housing in nonpoor neighborhoods, as was done in Yonkers, New York. However, if the most advantaged poor families move out of poor neighborhoods, what remains are concentrations of poor families, and possibly of those with the most mental and physical health problems, poor coping skills, and low literacy—all barriers to economic self-sufficiency.
An alternative strategy is to move nonpoor families into poor neighborhoods to change the mix and reduce poverty concentration and segregation. Gentrification also typically entails providing services (e.g., good-quality schools) and jobs in these neighborhoods. It is still unclear if poor families (or which poor families) benefit from this transformation or if they are forced out.
In summary, future research will likely lead to better answers to the original questions posed by academic scholars and policymakers, as well as to the design of more effective policies.
Brooks-Gunn, J., Duncan, G.J., & Aber, J.L. (Eds.). (1997). (See References)
Burton, L.M., & Jarrett, R.L. (2000). In the mix, yet on the margins: The place of families in urban neighborhood and child development research. Journal of Marriage and the Family, 62, 1114-1135.
Goering, J. (Ed.). (in press). (See References)
Leventhal, T., & Brooks-Gunn, J. (2000). (See References)
Sampson R.J., Raudenbush S.W., & Earls, F. (1997). (See References)
1. Address correspondence to Tama Leventhal, National Center for Children and Families, Teachers College, Columbia University, New York, NY 10027; e-mail: TL91@columbia.edu.
Brooks-Gunn, J., Duncan, G.J., & Aber, J.L. (Eds.). (1997). Neighborhood poverty (2 vols.). New York: Russell Sage Foundation Press.
Elliott, D., Wilson, W.J., Huizinga, D., Sampson, R., Elliott, A., & Rankin, B. (1996). The effects of neighborhood disadvantage on adolescent development. Journal of Research in Crime and Delinquency, 33, 389-426.
Goering, J. (Ed.). (in press). Choosing a better life? How public housing tenants selected a HUD experiment to improve their lives and those of their children: The Moving to Opportunity Demonstration Program. Washington, DC: Urban Institute Press.
Jencks, C., & Mayer, S. (1990). The social consequences of growing up in a poor neighborhood. In L.E. Lynn & M.F.H. McGeary (Eds.), Inner-city poverty in the United States (pp. 111-186). Washington, DC: National Academy Press.
Katz, L.F., Kling, J.R., & Liebman, J.B. (2001). Moving to Opportunity in Boston: Early results of a randomized mobility experiment. Quarterly Journal of Economics, 116, 607-654.
Klebanov, P.K., Brooks-Gunn, J., McCarton, C., & McCormick, M.C. (1998). The contribution of neighborhood and family income to developmental test scores over the first three years of life. Child Development, 69, 1420-1436.
Leventhal, T., & Brooks-Gunn, J. (2000). The neighborhoods they live in: The effects of neighborhood residence upon child and adolescent outcomes. Psychological Bulletin, 126, 309-337.
Leventhal, T., & Brooks-Gunn, J. (2001). Changing neighborhoods: Understanding how children may be affected in the coming century. Advances in Life Course Research, 6, 263-301.
Leventhal, T., & Brooks-Gunn, J. (2002). A randomized study of neighborhood effects on low-income children’s educational outcomes. Manuscript submitted for publication.
Leventhal, T., & Brooks-Gunn, J. (in press). Moving to Opportunity: An experimental study of neighborhood effects on mental health. American Journal of Public Health.
Ludwig, J., Duncan, G.J., & Hirschfield, P. (2001). Urban poverty and juvenile crime: Evidence from a randomized housing-mobility experiment. Quarterly Journal of Economics, 116, 655-679.
Massey, D.S., & Denton, N.A. (1993). American apartheid: Segregation and the making of the underclass. Cambridge, MA: Harvard University Press.
Rubinowitz, L.S., & Rosenbaum, J.E. (2000). Crossing the class and color lines: From public housing to white suburbia. Chicago: University of Chicago Press.
Sampson, R.J., Raudenbush, S.W., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277, 918-924.
Wilson, W.J. (1987). The truly disadvantaged: The innercity, the underclass, and public policy. Chicago: University of Chicago Press.