Elsevier

Health & Place

Volume 13, Issue 4, December 2007, Pages 788-798
Health & Place

No neighborhood is an island: Incorporating distal neighborhood effects into multilevel studies of child developmental competence

https://doi.org/10.1016/j.healthplace.2007.01.006Get rights and content

Abstract

The purpose of this study was to examine whether incorporating information regarding neighborhoods which were more distal to the child's neighborhood added any explanatory power to models of child cognitive competence. Participants included a sample of young African-American children living in an urban setting in the northeast United States. Spatial geographic methods were used to estimate the concentration of economic disadvantage, population instability, and crime in the neighborhoods surrounding the child's residence, and multilevel modeling methods were used to estimate the contribution of these factors to between-neighborhood variance in child cognitive scores. Results indicated that the conditions of distal neighborhoods were related to cognitive scores among the preschooler-age children in this sample.

Introduction

In the last 10–15 years, there have been numerous reports in the literature documenting that neighborhood characteristics contribute significantly to cognitive outcomes in children over and above the variance explained by differences in family characteristics (Brooks-Gunn et al., 1993; Caughy and O’Campo, 2006; Chase-Lansdale and Gordon, 1996; Chase-Lansdale et al., 1997; Kohen et al., 2002; Leventhal and Brooks-Gunn, 2004; Shumow et al., 1998). A number of theorists have hypothesized about the processes by which neighborhoods affect the families and children living within them with early published reports focused primarily on community role models as the mechanism underlying the association between neighborhood socioeconomic characteristics and child cognitive outcomes. Brooks-Gunn et al. (1993) reported that high levels of affluent neighbors were associated with higher IQ scores among three-year olds and lower high school drop out rates among adolescents. These researchers theorized that the presence of affluent neighbors provided a positive role model for parents and children in the neighborhood. Sampson (1992) proposed that neighborhood structural and population characteristics influence the functioning of families and children by altering levels of community cohesion and informal control. As a sociologist studying the roots of crime and delinquency, Sampson built upon the long tradition of social disorganization theory which posits that neighborhood poverty, ethnic diversity, and population instability contribute to the breakdown of social cohesion in the neighborhood and, in turn, results in higher rates of juvenile delinquency. Sampson hypothesized that neighborhood collective efficacy affected child development by reducing levels of nurturing and supportive parenting within the family (Sampson, 1992). There is an extensive literature documenting that sensitive and responsive parenting combined with the provision of opportunities for age-appropriate cognitive stimulation and consistent firm discipline is associated with more optimal child development outcomes (see, e.g., Brooks-Gunn et al., 1999; Collins et al., 2000; Guo and Harris, 2000).

In this paper, we examine the contribution of three different dimensions of neighborhoods—concentrated economic disadvantage, population instability, and crime—to the cognitive functioning of young urban-dwelling African-American children. As described above, Sampson (1992) lays out a theory whereby concentrated economic disadvantage and population instability contribute to higher levels of community social disorganization which in turn compromises nurturing and supportive parenting as well as family management processes such as child monitoring. If population instability undermines social organization in the community, it might be more difficult for parents to access resources in the community, such as after school programs or extracurricular activities that might foster the development of cognitive skills in their children. In a qualitative study of adolescent outcomes in a diverse set of poor neighborhoods in Philadelphia, Furstenberg (1993) reports that in neighborhoods with few social resources, parents had to be exceptionally motivated “to cultivate the sparse opportunities within their community and to search out opportunities beyond the confines of their local area” (p. 243).

The impact of community crime and violence on child development has been more extensively studied in relation to socioemotional outcomes than cognitive outcomes. Garbarino et al. (1991) provided a summary of effects of exposure to extreme violence, such as when a child grows up in a war zone, as well as exposure to chronic violence such as community crime. The empirical evidence is strong linking high levels of crime in the neighborhood with adjustment problems in children (Aneshensel and Sucoff, 1996; Ceballo et al., 2001; Shumow et al., 1998; Martinez and Richters, 1993; Cicchetti and Lynch, 1993; Simons et al., 2002). Given the strong link between child behavior problems and academic outcomes, one would hypothesize that neighborhood crime would also negatively affect cognitive development. Shumow et al. (1998) provide one of the few studies that has examined the relation between neighborhood crime and cognitive outcomes. Utilizing a composite index of neighborhood risk which included rates of violent crimes, Shumow et al. (1998) reported that neighborhood risk was associated with poorer academic achievement in fifth grade. Another way that neighborhood crime could compromise cognitive development is by limiting the child's access to outdoor play opportunities. Qualitative data suggest that parents living in high risk neighborhoods are less likely to allow their child to play outside (Jarrett, 1999; Burton and Price-Spratlen, 1999; Furstenberg, 1993; Furstenberg and Hughes, 1997), and these limitations might also function to limit opportunities for exploration that would foster the development of cognitive skills.

Two of the major limitations of the existing literature on neighborhood effects on child competence, however, are the narrow definition of neighborhoods which is often utilized and the statistical methods used to analyze multilevel neighborhood associations.

Most investigators utilize data from the census, and consequently, they have used a geographic unit based on census boundaries to represent the neighborhood, such as census tract boundaries or census block group boundaries. Although there has been considerable debate in the neighborhood literature regarding the best geographic unit to represent neighborhoods (O’Campo, 2003; O’Campo and Caughy, 2006), it is likely that none of these municipally defined neighborhoods perfectly approximates the boundaries of neighborhoods as they are perceived by residents.

Conceptually, people are influenced not only by their immediate surroundings, but by areas further away from their homes. For example, shopping areas may be close but not in an immediate neighborhood. Similarly, schools and parks may be nearby but not within a few blocks of a family's home. Yet, to date, all of the studies of neighborhood effects on child well-being have modeled neighborhoods as if they exist independently of one another and without respect to the conditions of the neighborhoods surrounding them. Traditional multilevel models ignore the spatial aspect of neighborhoods in that they cannot address the spatial scale of variation (Chaix et al., 2005). Autocorrelation models are one approach toward capturing the spatial distribution of individual outcomes. Initial attempts have been made to examine spatial models for mental health outcomes (Chaix et al., 2005), and we capitalize on autocorrelation models for modeling spatial effects on child outcomes.

By not capturing the spatial nature of neighborhoods, we neglect the very real possibility that the effects of the immediate neighborhood environment might be further moderated by the effects of more distal neighborhood environments. For example, one would hypothesize that living in a poor neighborhood surrounded by poor neighborhoods might have qualitatively different effects on children than living in a poor neighborhood surrounded by non-poor neighborhoods. Indeed, researchers such as Wilson (1987) and Jargowsky (1997) have focused on the deleterious effects of high-density urban poverty in which poor families, primarily ethnic minorities, live in neighborhoods which are increasingly more isolated. However, none of the extant research on neighborhood effects on children has captured these more spatial characteristics of urban poverty by attempting to incorporate characteristics of more distal neighborhoods in modeling the effects of neighborhoods on children. Sampson et al. (1999) approached this issue but examining whether there was spatial autocorrelation between measures of neighborhood social processes relevant to childrearing in the immediate neighborhood and social processes in surrounding neighborhoods. Although the results of their analyses did support the hypothesized spatial relations, these findings did not extend to an examination of whether these relations made any difference in the development of children.

There are methods drawn from geography that can be used to incorporate effects of more distal neighborhoods when examining neighborhood effects on child competence. Because our study utilizes spatial data, Geospatial Information Systems (GIS) is used in the context of address-matching, variable creation, and exploratory analyses to examine the effects of spatial clustering, through the development of our multilevel regression models. The geographic correlation of neighborhood events and social characteristics often cause problems for traditional statistical techniques because one of the primary assumptions that researchers must make in order to use traditional methods is the assumption of independence (McClendon, 1994; Anselin, 1995, Anselin, 1996; Anselin and Kelejian, 1997). That being said, multilevel models allow researchers to control for some types of clustering, such as the presence of multiple children in the same neighborhood (Bryk and Raudenbush, 1992). However, multilevel models, as they currently are developed, are unable to simultaneously control for the effects of spatial autocorrelation (i.e., the relationship between independent variables at the second level) (Anselin, 1988, Anselin, 1998; Anselin and Kelejian, 1997; Robinson, 1998; Ding and Fotheringham, 1992).

Spatial autocorrelation can be checked for using tests such as the Moran's I statistic (Moran, 1948), which is a univariate statistic designed to test the null hypothesis of the absence of spatial clustering (Cliff and Ord, 1981; Baller et al., 2001). In simple terms, Moran's I measures the deviation from spatial randomness, or the concentration of an attribute over space. Moran's I is similar to a Pearson correlation coefficient and is scaled to be less than one in absolute value. If locations are close together and tend to be similar in attributes, this will be reflected with a positive spatial autocorrelation score (contagion, spillover, externalities) and underestimated regression coefficients (Robinson, 1998). Conversely, if locations are proximate but instead have very dissimilar values, this is reflected as a negative spatial autocorrelation score (competition, revulsion) and overestimated regression coefficients (Robinson, 1998). Larger absolute values indicate higher levels of spatial autocorrelation in the data. When values are independent of their location, then zero autocorrelation is present. (Ding and Fotheringham, 1992; Baller et al., 2001).

In this study, we draw upon spatial analytic methods from geography to examine neighborhood effects on the developmental competence of children. Using data from a sample of young African-American children living in an urban setting, we address the question of whether or not incorporating information from neighborhoods more distal from the one in which the child lives adds explanatory power for between neighborhood differences in cognitive functioning. The extant literature regarding neighborhood effects on child cognitive development provide little guidance regarding theories of spatial effects. As previously stated, Wilson (1987), in his seminal work documenting accelerated rates of concentrated economic disadvantage in the US, focused on the isolation of poor individuals in the newly emerging urban ghettos of the 1980s. In positing how spatial effects of neighborhood structural effects affect child cognitive outcomes, one approach would be to extend theories of how the more proximal neighborhood affects children. For example, if concentrated neighborhood economic disadvantage in the immediately surrounding neighborhood limits the resources available to families with young children, then concurrent economic disadvantage in more distal neighborhoods may act to exacerbate the limitation of resources for families and children in poor neighborhoods. If high crime in a neighborhood results in parents limiting their children's access to outdoor play, living in a high crime neighborhood surrounded by more high crime neighborhoods could further limit outdoor play by increasing the distance parents would have to travel to reach safe play alternatives.

Section snippets

Participants

Data for the spatial analyses were drawn from a study of African-American families living in Baltimore with preschool-age children between 3 and 412 years of age. Recruitment methods for both studies were similar. Participants were recruited from Baltimore City neighborhoods through door-to-door-canvassing, targeted mailing lists, and referrals from other participants. Neighborhoods were defined as census block groups and were stratified by average household wealth and racial composition to

Results

The characteristics of the study sample are displayed in Table 1. The majority of the respondents were mothers of the target children. Almost half of the participating families lived below the federal poverty line, and approximately a quarter lived above 180% of the poverty level.

Differences in child cognitive competence by neighborhood characteristics are displayed in Table 2. High concentrated economic disadvantage and population instability were associated with lower cognitive scores. High

Discussion

The purpose of this study was to examine the contributions of distal neighborhood environments to the developmental competence of young children. Both the proximal neighborhood as well as the more distal neighborhood contributed unique variance to cognitive competence in this sample of African-American preschoolers living in an urban setting. High levels of concentrated economic disadvantage in the immediate neighborhood as well as high levels of economic disadvantage in the surrounding

Acknowledgments

This research was supported by Grant #MCJ-240731-01-1 from the Maternal and Child Health Bureau and Grant #RO1HD4041901A1 from the National Institute of Child Health and Development. The authors would like to thank Deborah Brothers and Bennette Drummond-Fitzgerald for conducting interviews, and Kimberly Lohrfink for providing project management. Data management and analysis support was expertly provided by YiHua Chen, Crystal Evans, Patricia Gwayi-Chore, and LiChing Lee. Finally, we would like

References (47)

  • Y. Ding et al.

    The integration of spatial analysis and GIS

    Computers, Environment, and Urban Systems

    (1992)
  • C.S. Aneshensel et al.

    The neighborhood context of adolescent mental health

    Journal of Health and Social Behavior

    (1996)
  • L. Anselin

    Spatial Econometrics

    (1988)
  • L. Anselin

    Local indicators of spatial association—LISA

    Geographical Analysis

    (1995)
  • Anselin, L., 1996. GIS research infrastructure for spatial analysis of real estate markets. Paper presented at the...
  • Anselin, L., 1998. GIS research for spatial analysis of real estate markets. Paper presented at the Regional Research...
  • L. Anselin et al.

    Testing for spatial error autocorrelation in the presence of endogenous regressors

    International Regional Science Review

    (1997)
  • R.D. Baller et al.

    Structural covariates of US county homicide rates: incorporating spatial effects

    Criminology

    (2001)
  • J. Brooks-Gunn et al.

    Do neighborhoods influence child and adolescent development?

    American Journal of Sociology

    (1993)
  • J. Brooks-Gunn et al.

    Are socioeconomic gradients for children similar to those for adults? Achievement and health of children in the United States

  • A.S. Bryk et al.

    Hierarchical Linear Models: Application and Data Analysis Methods

    (1992)
  • L.M. Burton et al.

    Through the eyes of children: an ethnographic perspective on neighborhoods and child development

  • M.O. Caughy et al.

    Neighborhood impoverishment, social capital, and the cognitive development of African American preschoolers

    American Journal of Community Psychology

    (2006)
  • R. Ceballo et al.

    Inner-city children's exposure to community violence: how much do parents know?

    Journal of Marriage and the Family

    (2001)
  • B. Chaix et al.

    Comparison of a spatial perspective with the multilevel approach in neighborhood studies: the case of mental and behavioral disorders due to psychoactive substance use in Malmö, Sweden, 2001

    American Journal of Epidemiology

    (2005)
  • P.L. Chase-Lansdale et al.

    Economic hardship and the development of five- and six-year-olds: neighborhood and regional perspectives

    Child Development

    (1996)
  • P.L. Chase-Lansdale et al.

    Neighborhood and family influences on the intellectual and behavioral competence of preschool and early school-age children

  • D. Cicchetti et al.

    Toward an ecological/transactional model of community violence and child maltreatment: consequences for children's development

    Psychiatry

    (1993)
  • A.D. Cliff et al.

    Spatial Processes: Models and Applications

    (1981)
  • W.A. Collins et al.

    Contemporary research on parenting. The case for nature and nurture

    American Psychologist

    (2000)
  • F.F. Furstenberg

    How families manage risk and opportunity in dangerous neighborhoods

  • F.F. Furstenberg et al.

    The influence of neighborhoods on children's development: a theoretical perspective and a research agenda

  • J. Garbarino et al.

    What children can tell us about living in danger

    American Psychologist

    (1991)
  • Cited by (32)

    • Using population surfaces and spatial metrics to track the development of deprivation landscapes in Glasgow, Liverpool, and Manchester between 1971 and 2011

      2018, Computers, Environment and Urban Systems
      Citation Excerpt :

      The UGCoP highlights that whilst an individual may reside within a particular areal unit, this is very unlikely to be the only place they encounter; there are other places, indeed multiple spatial scales, exerting influence on them. This has been highlighted as particularly pertinent to studies examining relationships between health and deprivation (Cummins, 2007) and led to the suggestion that the smallest and most local area possible may not be the most appropriate scale of analysis (Caughy, Hayslett-Mccall, & O'campo, 2007; Cummins, 2007). Perhaps the best way of dealing with this problem is to somehow undertake analyses which are able to examine multiple spatial scales from very fine to whole city.

    • Moving to opportunity and mental health: Exploring the spatial context of neighborhood effects

      2016, Social Science and Medicine
      Citation Excerpt :

      Despite these limitations, our analysis contributes a unique examination of the effects of cumulative spatial adversity over time. It also improves our understanding of neighborhood effects on health by focusing on neighborhood concentrated disadvantage in addition to poverty – a broader indicator or neighborhood quality that is consistent with prior insights on concentration effects (Wilson, 1987), the social disorganization literature (Sampson and Groves, 1989), a recent analysis of delinquency among MTO youth (Graif, 2015), and research on the neighborhood effects on other health outcomes in varying social groups and countries (Allender et al., 2012; Caughy et al., 2007; Sridharan et al., 2007; Zhang et al., 2011, 2013). Our results show that exposure to residential neighborhood disadvantage is associated with worse mental health, as expected.

    • Differential effectiveness of Head Start in urban and rural communities

      2016, Journal of Applied Developmental Psychology
    • Elucidating the changing socio-spatial dynamics of neighborhood effects on adult obesity risk in Taiwan from 2001 to 2005

      2010, Health and Place
      Citation Excerpt :

      Thus, the potential of placing neighborhoods in a larger system (environment) has generally been overlooked. The present study extends the frameworks of Morenoff et al. (2001), Morenoff (2003), Caughy et al. (2007), and Chen and Wen (2010) on the spatial patterns of neighborhood effects. Its design was based on the assumption that individuals often work, shop, and attend school far from home.

    • Socio-spatial patterns of neighborhood effects on adult obesity in Taiwan: A multi-level model

      2010, Social Science and Medicine
      Citation Excerpt :

      Integrating spatial analysis technique for displaying the connection of social and spatial process is important for progress in neighborhood-effects research (Sampson et al., 2002). Following the framework provided by Caughy et al. (2007), Morenoff (2003), and Morenoff et al. (2001) on the spatial patterns of neighborhood effects, this study examines the multi-level factors associated with individual obesity by incorporating individual-level, within-neighborhood, and among-neighborhood characteristics in multi-level hierarchical regression models. In this framework, the characteristics of nearby neighborhoods are potentially crucial to understanding factors associated with obesity in a given neighborhood.

    View all citing articles on Scopus
    View full text