ODIS JOHNSON, JR. AND VON E. NEBBITTINTRODUCTION
INSTITUTIONS SUCH AS PUBLIC HOUSING developments have ecological structures, features, and functions similar to those that define neighborhoods. Like neighborhoods, they possess a structural composition, have a built environment, and inspire social processes. They constitute represented communities, places of social organization, and mechanisms that bring about social outcomes. The major question addressed in this chapter extends from this reality and requires us to conceptualize these two ecological contexts and their interrelated components, as well as how they work in tandem to determine the developmental outcomes of underrepresented children.
The importance of the question extends from the vast differences in the conditions of the neighborhoods in which public housing communities are situated. Inequality between poor and affluent areas in the United States has grown since the early 1980s (Massey & Fischer 2003), adversely affecting public housing developments, which are more likely to be located in areas of concentrated poverty (Newman & Schnare 1997). Neighborhood variations may contribute to pronounced differences in the quality of life of public housing communities, and by extension to disparities in the developmental outcomes of children. Interest in neighborhood effects has produced a healthy body of research that focuses on the developmental outcomes of children (Leventhal & Brooks-Gunn 2000) and adolescents (Duncan 1994; Halpern-Felsher et al. 1997) and relates neighborhood disadvantage to their academic performance (Sampson, Sharkey, & Raudenbush 2008). Research has also investigated the relationship between public housing residency and education (Currie & Yelowitz 2000; Jacob 2003). However, few studies explore quantitatively the connection of neighborhood conditions to the educational experiences of public housing residents in a way that incorporates the influence of their public housing community. Understanding how neighborhoods and the institutions embedded within them function in concert to shape youth’s educational experiences is a first step in producing related research.
The understanding of neighborhood–institutional relationships presented in this chapter is mostly informed by an exhaustive review of the neighborhood effects research. With this literature in mind, we first provide a diagram of neighborhood–institutional relationships within an ecological context, drawing heavily from Bronfenbrenner’s ecosystems theory (Bronfenbrenner 1979). We further explain the model by summarizing research related to each ecological component’s influence on the development of youth. Next, we review urban relocation studies to assess the relative benefits of residing in public housing communities and neighborhoods for children and adolescents. The chapter concludes with an assessment of the field’s progress and the challenges awaiting the next generation of studies.
INSTITUTIONS WITHIN THE NEIGHBORHOOD CONTEXT
The lack of distinction between neighborhood qualities and those of the institutions embedded within them is commonplace within neighborhood studies (Johnson 2010), extending from the more homogeneous social background characteristics of individuals populating those institutions and the fact that residency often determines who is served by those institutions. Therefore, to the extent either domain is responsible for youth development, their influence may be more similar than unique. Regardless of whether neighborhoods and institutions function separately to produce similar social outcomes for young people, there needs to be a way to discuss and explore the relationships through which the similarity emerges, especially if features of either domain can be changed to support the development of children and adolescents.
In an effort to define neighborhood–institutional relationships, we present a diagram of an ecological system that represents the primary components discussed within the neighborhood effects research. The diagram’s features are aligned with Bronfenbrenner’s (1979) ecosystems theory, which features nested layers of ecological contexts (i.e., from macro to micro) that operate in support of child development. Rather than depicting each nested layer in concentric zones of influence, as has been the case in other representations (Simbeni & Allen-Meares 2002), figure 2.1 lays out each stratum so that more detail can be provided about their components and interactions. Beginning at the macroecological level, broader social forces, such as changes in urban labor markets or migratory shifts within metropolitan areas, define the system of relationships at lower levels.
Venkatesh’s (2002) historical exploration of the Robert Taylor Homes in Chicago described in detail the relationships between the macroecological and exoecological strata. He reported that greater demand for public housing in the late 1960s and 1970s, likely due to the macroeconomic concerns of the time, led to policy changes at the exoecological level enacted by the Department of Housing and Urban Development (HUD) and the Chicago Housing Authority. These policies gave housing preference to families with the most financial need, effectively changing the Robert Taylor Homes from a mixed-income dwelling to one primarily for the impoverished. Pattillo’s (2007) account of a neighborhood’s efforts to fight the return of public housing to a Chicago neighborhood emphasized the influence of courts and judicial decrees in defining the neighborhood population, the construction of large public housing structures, and the presence of public institutions. In fact, federal and state policies assisted the continued ghettoization of African Americans through the provision of public housing. HUD built large housing projects in largely black, low-income communities and imposed rules that required public housing recipients be assigned to areas that matched their racial classification (Hirsch 2000). Doing so confined the growing black population within its traditional ghetto boundaries (Massey 2007; Polikoff 2006; Yinger 2001). The exoecological level then represents the organizational and policymaking communities that influence neighborhoods and their institutions at the mesoecological level.
The components at the mesoecological stratum are defined as structural (i.e., demographic composition), of the built environment (i.e., manmade surroundings), institutional (e.g., schools, public housing, policing), and social (e.g., crime, social cohesion). Here we discuss each of these components in turn. Structure refers to the population characteristics of a particular area and is often presented within research as an average or “concentration” of certain social and behavioral characteristics. The most frequently investigated structural effects are related to the classic indicators of social stratification: social class, race, and labor market opportunity. This focus has come to characterize much of the neighborhood effects research, owing largely to Wilson’s early work (1987), which speculated about the pernicious effects of macroeconomic change in increasing the concentration of poverty in African American neighborhoods. Hence, the mesoecological level can be influenced directly by the macroecological level’s impact on population characteristics as well as indirectly through the exoecological in which it is embedded. Because the macroecological level at times directly influences the neighborhood components and in turn prompts policymaking organizations to respond, the figure has bidirectional arrows in which meso-level dynamics can induce actions at the exo-level.
The built environment (i.e., manmade surroundings) is another neighborhood feature of the mesoecological stratum and is often related in research to the development of children. For example, studies have linked the presence of dilapidated housing (Limbos & Casteel 2008) and the number of rental housing units in the neighborhood (Thompson et al. 2006) to school crime and student feelings of school connectedness. Other studies looked for possible connections between the location of toxic sites (Margai & Henry 2003), public schools (Schlossberg et al. 2006) and private schools (Barrow 2005), opportunities for community center involvement (Anderson, Sabatelli, & Kosutic 2007), and the quality of signage in the area (Celano & Neuman 2001) to children’s possibility of being identified as learning disabled, achievement, transportation needs to school, and literacy development.
The structural and built environment collude to shape the character of life in affordable housing communities, as indicated by the arrows in figure 2.1. For example, Hunt (2009) and Venkatesh (2002) both revealed that a significant number of large apartment units (aspects of the built environment) within public housing high-rises led to a disproportionate representation of children (an aspect of the structural environment) in those communities. Malfunctioning elevators and less adult control over peer interactions resulted, eventually contributing to an inability to police those communities and a growth in gang activities.
The diagram in figure 2.1 also identifies institutions or entities provided and managed by governmental agencies as part of the neighborhood context, primarily because institutions such as public housing heavily influence the nature of residential areas. In addition to the impact of public housing’s presence on a neighborhood’s poverty levels, research has established linkages between the location of public housing and a neighborhood’s property values (Ellen et al. 2005; Galster et al. 1999; Green, Malpezzi, & Seah 2002), the desires and social activism of homeowners (Pattillo 2007; Williams 2005); neighborhood crime rates (McNulty & Holloway 2000), and school safety (Limbos & Casteel 2008). As such, neighborhood components influence and also are influenced by institutions—a fact that is reflected in figure 2.1 by the bidirectional arrows between differing aspects of neighborhoods.
Finally, the mesoecological level also supports the development of social processes among neighboring individuals in the public housing and neighborhood contexts. Social processes such as peer relationships, role modeling/mentoring, social disorganization, and social cohesion are thought to better represent how neighborhood and institutional effects occur. Ainsworth’s (2002) analysis revealed that social processes account for approximately 40 percent of the total neighborhood effect on education outcomes. Although many social process effects have been proposed (Gephardt 1997; Jencks & Mayer 1990), this review found seven of them to be the subject of multiple investigations and likely to arise in public housing and neighborhood contexts, as briefly discussed in the following sections.
Collective Socialization
Collective socialization acknowledges the influence adults within a public housing community may have on the behavior of children other than their own. Role model effects therefore emphasize the impact of characteristics of older community members over the decisions of younger members (Durlauf 2001). There are at least two ways in which role model effects bring about the collective socialization of area youth. In the first, behaviors are modeled for children unbeknownst to the adult. In contrast to this more passive form of collective socialization, adults may collectively and actively monitor youth in the community (Leventhal & Brooks-Gunn 2000). This kind of role modeling may have more pronounced effects because it presumes adults are active agents of influence. In the literature, however, more passive role model effects are measured, largely at the neighborhood or school level (Ainsworth 2002; Chase-Lansdale & Gordon 1996; Darling & Steinberg 1997), although reports of declining educational outcomes of affordable housing recipients in New York cite the lack of role models for children as a primary cause (Furman Center 2008).
Epidemic processes identify how youths influence each other’s behavior through imitation and peer pressure and lessen children’s reliance on parents and adults for guidance and sustenance (MacLeod 1995; Rainwater 1970). After a certain prevalence or threshold of youth behaviors within a neighborhood or housing community is reached, the likelihood that other children will adopt those behaviors increases at a faster rate (Clark 1992; Crane 1991). Research has sought to differentiate between types of peer effects, separating the outcomes of group decision-making and behaviors from the individual behaviors that may develop in response to group compositional qualities (Bobonis & Finan 2008; Durlauf 2001, Graham 2003; Manski 1993), although most of these distinctions have been made about school-based peers rather than public housing or neighborhood peers. Investigations of peer effects have linked the relationship of student assessments of their neighborhood peers (Darling & Steinberg 1997; Nash 2002; South, Baumer, & Lutz 2003), the educational characteristics of neighbors (Duncan, Boisjoly, & Mullan-Harris 2001), the number of friends known by parents (Lopez-Turley 2003), and the random assignment of roommates (Sacerdote 2001) with the education outcomes of children and adolescents.
Social Networks and Capital
The composition of an environment may determine the quality of the neighbors that public housing residents must rely on for information and advice about employment opportunities, child care, and educational decisions. A few studies focus on the benefits of social ties within the neighborhood context to educational outcomes (Lundberg & Startz 2000), whereas others suggest that social ties can lead to limited success or even greater disadvantage if one’s network contains neighbors and peers with fewer beneficial connections (Pattillo-McCoy 1999; Portes & Landolt 1996). Three measures of social networks explored in the literature include the level of social integration or whether there is overlap between the social networks of adults and youth in a neighborhood (Darling & Steinberg 1997); intergenerational closure, being the number of close friends known by a child’s parents (Ainsworth 2002); and the occurrence of neighborhood social interaction between adults and children (Caughy et al. 2006).
Groups are considered socially cohesive when their members’ interpersonal interactions operate to maintain identifiable and positive group-level membership attitudes, behaviors, attractions, and attachments (Friedkin 2004). In contrast, public housing communities and neighborhoods facing high family mobility rates would appear less cohesive. A few studies estimated the effects of social cohesion on child outcomes by considering the ability of neighbors to recognize a stranger in the neighborhood (Plybon et al. 2003), the willingness of adults to take action on behalf of their neighbors (Kohen et al. 2002), and the impact that residential stability may have on the development of relationships and norms in the area (Ainsworth 2002; Pebley & Sastry 2003; Woolley & Kaylor 2006).
Social Disorganization/Violence
Social disorganization is not only suggested by a neighborhood’s inability to realize shared values and achieve desired outcomes, but also with the establishment of area norms among adults and children that work to create less safe environments (Shaw & McKay 1942). The fear associated with the observation or experience of crime, for example, is thought to reduce a resident’s feelings of safety, mutual trust, and the willingness to supervise, report on, or intervene in the behavior of youth. Several studies in this area of research relate school attendance and grades to children’s perception of crime (Nash 2002), their residency in areas with criminal activity (Madyun & Lee 2008), personal experiences with crime and perceptions of delinquent behavior among their peers (Bowen & Bowen 1999), and their feelings of safety and perception of neighborhood support (Bowen, Bowen, & Ware 2002; Bowen et al. 2008).
Social Control/Collective Efficacy
Collective efficacy is described as the presence of mutual trust and the shared willingness to intervene for the public good (Sampson, Raudenbush, & Earls 1997), which is essential to the effective daily management of behaviors in public housing communities and neighborhoods. As suggested in qualitative work, crime and violence can inspire families to organize to demand more police protection and assume more collective responsibility for the safety of children (Venkatesh 2002). Hence, in contrast to social disorganization theory, crime may inspire tenants to become more socially organized and to assume more social control of their surroundings. Research in this area has considered children’s opinions of whether an adult would tell other adults if they had behaved badly as an indicator of social control (Nash 2002) and the strength of peer group effects as an indicator of a neighborhood’s lack of social control (Ainsworth 2002).
Racial Socialization
Ethnic/racial socialization demonstrates the endogenous quality of identity through the neighborhood’s role in identity development (Borjas 1995; Durlauf 2001). A neighborhood may influence one’s racial socialization and learning to the degree that its racial composition colors a child’s perceptions of opportunity, his or her chances of success, and the economic returns on a personal investment in education. Research in this area has explored how various racial socialization orientations relate to differences in the cognitive performance of children given their neighborhood’s racial and socioeconomic structure and its social processes (Caughy et al. 2006).
Although institutions are considered a part of the neighborhood context and as having relationships with aspects of the neighborhood, they are also depicted in figure 2.1 at the microecological level as possessing internally those same components with which they interact externally. For example, institutions also possess a structural composition. Although there is a dearth of studies about public housing population effects on children, the examination of composition effects within other institutions, such as schools, has a relatively long history (Alexander et al. 1979; Crain 1971; Crain & Mahard 1978). Institutions also possess a definitive built environment. On this point, Venkatesh’s (2002) account of public housing in Chicago suggested that one of the most significant contributors to the quality of the day-to-day experiences of public housing residents was the infrastructure of the housing itself and the larger housing complex. Studies have also detailed the social organization of public housing communities, including the social activism of mothers, their practices of informal social control, and the consequences of isolation (Venkatesh 2002; Williams 2005). Altogether, these institutional ecological components function to influence the well-being and development of youth. Examinations of public housing residency effects, for example, have investigated whether living in them leads to better educational outcomes for low-income children than other housing arrangements (Furman Center 2008; Jacob 2003; Lubell & Brennan 2007) and if children in public housing are more likely to experience grade repetition (Currie & Yelowitz 2000).
COMPLICATIONS OF NEIGHBORHOOD–INSTITUTIONAL RELATIONSHIPS
Ideal research in this area would link neighborhoods to the institutions that influence the development of youth because they spend a considerable amount of time in these environments. In fact, the activities of children living in large public housing developments rarely require them to leave the housing community (Shlay & Holupka 1991). There are, of course, challenges to the joint consideration of neighborhood–institutional relationships. Much has been written about these methodological challenges as they complicate the estimation of neighborhood effects (Duncan & Raudenbush 1999; Leventhal & Brooks-Gunn 2000; Manski 1993). Less has been written about factors that complicate the estimation of neighborhood and institutional relationships in particular. In this section, we explore issues that complicate the modeling of neighborhood–institutional relationships.
Transaction Underestimation
Work that attempts to take into consideration neighborhoods and institutions often includes institutional-level covariates in the statistical models. Considering institutions helps to limit some of the institutional bias that may arise when important institutional-level factors are unobserved. However, there is reason to believe that including institutional-level controls in models of neighborhood effects leads to an understatement of the true effect of neighborhoods on cognitive outcomes. Consider, for example, that 43 percent of units in public housing family developments are in neighborhoods where 40 percent or more of the residents live in poverty (Newman & Schnare 1997). Neighborhood poverty rates (and their racial composition) were factors that led HUD to locate public housing in those areas. Inasmuch as the measure of a public housing characteristic (e.g., poverty) is determined by neighborhood dynamics, once included in the analysis, that characteristic may deliver part of the neighborhood’s influence and thereby depress the neighborhood effect estimate.
Indirect pathways such as these also work to the opposite effect and possibly overstate the neighborhood’s impact. Recall the bidirectional flow noted in figure 2.1 between the location of affordable housing and the residential decision-making of families (Pattillo 2007) and property values (Ellen et al. 2005; Galster et al. 1999; Green, Malpezzi, & Seah 2002). To the degree that public housing influences perceptions of a neighborhood’s quality, public housing will have an influence on the residential makeup of the neighborhood. Neighborhood measures will subsequently appear more strongly related to the developmental outcomes of public housing residents than they should and public housing characteristics less so. The cross-currents of causality among social domains and the resulting underestimation of those domains’ effects is called transaction underestimation and has been tested as it relates to linkages between neighborhoods, families, and the home environment (Duncan, Connell, & Klebanov 1997).
FIGURE 2.1 System of nested ecological relationships.
Unobserved Institutional Effects
Discussions about how an institution functions often consider its social organization (Rosenholtz 1991; Venkatesh 2002; Weick 1976; Williams 2005) or isolation (Bronfenbrenner 1979). For instance, tenants of the Robert Taylor Homes had to compensate for a lack of policing, governance, and maintenance by city officials (Venkatesh 2002). The lack of formal social control within the housing community stood in stark contrast to its presence in the neighborhood context across the city. From a more positive vantage, institutions also function somewhat independently from neighborhoods due to their social and governance organization and more definitive built environment. Quantitative research in neighborhood and public housing effects has not kept pace with thinking about the possible semi-autonomous functioning of public housing and neighborhood effects, leaving a void in research about neighborhood–institutional relationships.
Not only are few institutional factors included in models of neighborhood effects on child and adolescent outcomes, social process effects occurring within institutions also remain largely unobserved in quantitative neighborhood studies. A look at school effects research makes this point clear. Some of the largest institutional effects within that body of literature have been of the social process kind, not the traditional factors that are frequently the subject of educational policy. School process measures in the school effects literature, for example, have demonstrated that constructs such as teacher learning opportunities, teacher certainty, and teacher commitment are between three and seven times larger in magnitude than the traditional measures of teaching experience, school socioeconomic status, and the prestige of the teacher’s undergraduate institution (Rosenholtz 1991). Another study found that low-performing schools were more likely to experience increased academic productivity with improvements in relational trust—that is, mutual respect, competence, a personal regard for others, and integrity among school actors—even after controlling for teacher background and school composition (Bryk & Schneider 2002). Similarly Bryk, Lee, and Holland (1993) revealed that the large advantages Catholic schools have over public ones—teachers’ enjoyment of work, staff morale, and students’ interest in academics—are due almost entirely to their communal social organization.
These studies imply that social process measures are just as instrumental in accounting for institutional influences as they are to the consideration of neighborhoods. The subsequent task for neighborhood effects research is to model institutional social processes well enough to account for their effects. This will prevent those effects from being erroneously thought of as neighborhood variation in learning as opposed to variation in learning across neighborhood institutions.
RELATIVE EFFECTS OF HOUSING COMMUNITIES AND NEIGHBORHOODS
Mobility presents an opportunity to weigh the costs and benefits for youth of leaving a family housing development context, joining a neighborhood context, and often times switching schools in the process. Three of the most notable studies on this subject have produced very mixed results. The first of these studies, the Gautreaux Assisted Housing Mobility Program that was sponsored by HUD, transitioned underrepresented families from large public housing dwellings into Section 8 housing within racially mixed urban neighborhoods or largely white suburban areas. Evaluations of the Gautreaux program typically compare the experiences and outcomes of those who moved to the largely white suburbs and those who moved to racially mixed urban areas and their schools (Rosenbaum, Kulieke, & Rubinowitz 1988; Rosenbaum 1995). Rosenbaum (1995) reported that after experiencing difficulty in adjusting to the higher standards of their new schools, children in suburban Gautreaux families generally did better in school than their counterparts who relocated to urban areas. Although college-going was valued equally among city and suburban movers, suburban students were more likely to take college-track courses and to attend a 4-year college. Although the program outcomes of Gautreaux are considered generally positive, one must use caution in concluding that certain neighborhood contexts are more beneficial for children than family housing developments.
First, the absence of a control group in the Gautreaux design prevents a comparison of segregated environments to predominantly white or racially mixed ones. It is unsafe to assume that the positive effects found for the Gautreaux children who relocated to predominantly white areas means that children in those areas did better than those who stayed behind. Consider the scenario, for example, where a control group in a segregated area, a comparison group in a racially mixed area, and an experimental group in a suburban area all experience educational gains, yet children in the racially mixed area make the least amount of progress. A significant difference in achievement might result from a comparison of the movers to the suburban and racially mixed areas, as was the case in the Gautreaux study. Both, however, might be insignificantly different from the group that did not move, whose average achievement growth is nested between that of the experimental and comparison groups. Therefore, leaving segregated areas has insignificant effects. Although purely hypothetical, this scenario demonstrates what we would know about segregation if the Gautreaux program had a control group and what we cannot rule out because it is does not.
Second, the Gautreaux findings were not replicated in two comparable studies of movers in Cincinnati’s Special Mobility Program (Fischer 1991) and Yonkers (Fauth, Leventhal, & Brooks-Gunn 2007). In the first program, the Cincinnati residents who moved to the suburbs were less satisfied with the suburban schools than the residents who relocated to other parts of the city. In contrast to the results of the Gautreaux program, Fischer (1991) reported that more than one-quarter of the suburban children generally did worse in school after relocation, compared to 10 percent of the city children. The Yonkers program revealed that moving was negatively related to the reading and math performance of children age 8 to 18 years. A difference was also found in how children in neighborhoods and those who remained in the public housing community rated their performance in school, with the latter group rating themselves well above average whereas movers rated themselves slightly above average. Finally, for both younger and older adolescents, as the age of movers increased, their school engagement decreased until it was lower than that of the youth who remained in public housing (Fauth, Leventhal, & Brooks-Gunn 2007).
The Moving to Opportunity Study (MTO), authorized by the Housing and Community Development Act of 1992, was fielded in Boston, Baltimore, Los Angeles, Chicago, and New York. In contrast to the Gautreaux program, MTO participants were randomly selected and assigned to three residential opportunities: a voucher that could be redeemed only in a low-poverty environment, Section 8 housing that had no residential restrictions, and reassignment to their current residential status. For elementary school children, the results from the Baltimore site were generally positive, with children assigned to the experimental group exceeding the scores of their counterparts in public housing in both reading and math by approximately one-quarter of a standard deviation (Ladd & Ludwig 1997).
The assessment of the New York MTO site produced results that were much different than the Baltimore study. The estimates of achievement in the New York site were of adolescents and were measured on two occasions after the families relocated (Leventhal, Fauth, & Brooks-Gunn 2005; Ludwig, Ladd, & Duncan 2001). Leventhal, Fauth, and Brooks-Gunn (2005) reported that the cognitive advantage that was found for the experimental group relative to urban ones 2 years after relocation disappeared by year 5. In some instances, the African American males who were relocated to the suburbs performed less well than the African American males who remained in public housing communities (Leventhal, Fauth, & Brooks-Gunn 2005; Sanbonmatsu et al. 2006). Two analyses of the pooled data from all five demonstration sites show no significant effects of relocation on test scores for any age group among more than 5,000 youth up to 7 years after relocation (Orr et al. 2003; Sanbonmatsu et al. 2006). In sum, the removal of children from family housing developments does not assure their cognitive outcomes will improve by residing in low-poverty neighborhoods.
Apparently, there are lasting consequences for having at any time been exposed to resource-poor environments (Quillian 2003). Moving to a better neighborhood may interrupt the former environment’s suppression of cognitive development, but the effects of the new neighborhood would have to be unusually strong to undo the influence of an adolescent’s past residency in public housing and inhibit its effect on future learning. To this point, a Chicago study found that having previously resided in areas of concentrated disadvantage reduces the later verbal ability of African Americans by an amount equivalent to 1 or more years of schooling (Sampson, Sharkey, & Raudenbush 2008). The implication is that a neighborhood’s social class still matters, but residential advancement becomes unable to counter its firmly rooted effects later in one’s development. The greater educational performance of MTO’s Baltimore children over that of its adolescents in Baltimore and New York supports this view. Programs that relocate families to higher income or more diverse areas may prove to be more beneficial for younger than older children, provided that program incentives can be put into place to compel parents to reside for longer periods of time in their new neighborhoods and enroll their children in new schools.
CONCLUSION
Quantitative inquiry in the area of ecology has a relatively short history, with the first studies of neighborhood effects (Fernandez & Kulik 1981) and cognitive outcomes (Datcher 1982) appearing in the early 1980s. The framework and research presented in this chapter suggest that ecological research has leaped toward social and scientific relevance when small steps might have been expected. First, the research that has been generated since the 1980s has informed Bronfenbrenner’s (1979) ecosystem theory in child and adolescent development. This chapter recast the ecological theory in a way that organizes and summarizes the field of neighborhood and public housing effects while elaborating on the definition and function of its components as they relate to learning. In this chapter, we also briefly identified some of the conceptual challenges that arise in the measurement of neighborhood–institutional relationships, including transaction bias and the continued problem of unobserved institutional dynamics in research. Last, we juxtaposed institutional and neighborhood contexts in their effects on children and adolescents who moved in comparison to children who did not move. Moreover and surprisingly, research in this area has not proven that neighborhood environments are better for youth development than public housing communities.
The next generation of studies benefits from the momentum of the progress underway as it faces the challenge of continuing a relentless investigation into these topics. We place the investigation of embedded ecological institutions at the forefront of the challenge to understand how neighborhoods work to influence learning. Also a challenge is the development of ecological approaches to the study of public housing developments. Although there are clear methodological obstacles to the analysis of small residential environments with fairly homogenous populations, new methodologies will meet these demands, while our imagination will find dimensions on which public housing residents vary. Hence, a primary goal of public housing research is to identify the social machinery that is hiding behind its influence on residents. Finally, the framework presented in this chapter is a heuristic for investigators to manipulate as better ideas and more research become available. This chapter will hopefully serve as a point of departure for future conceptualizations of interactions between these important domains.
