Recent discussions of the urban underclass have begun to focus attention on the consequences that living in particular neighborhoods may have for their residents. Some have argued that the increasing concentration of minority poor in urban areas may relegate the residents to persistent poverty and “social pathologies.” In his 1987 book, The Truly Disadvantaged, William Julius Wilson emphasizes the role of neighborhoods in shaping the lives of the poor. He argues that the decline of central-city manufacturing, the suburbanization of employment, and the out-migration of middle-class black families from ghetto areas have left behind destitute communities lacking the institutions, resources, and role models necessary for success in post-industrial society. Regardless of the particular definition of the underclass that one accepts, there is evidence that the size of the urban underclass grew substantially after 1970, that it became more spatially concentrated, and that the population is predominately black and Hispanic (Bane and Jargowsky 1988; Mincy 1988; Ricketts and Sawhill 1988; Hughes 1988).1 The conjunction of Wilson’s argument and evidence for the growth and concentration of an urban underclass has led to renewed interest in the role of neighborhoods and communities in the processes that create, maintain, or help to ameliorate the conditions that are typically subsumed under the urban underclass concept.

“During the past two decades, scholars largely ignored the role of neighborhoods and communities in shaping the lives of the poor.” During the past two decades, scholars largely ignored the role of neighborhoods and communities in shaping the lives of the poor. Research on poverty focused on income, and policy focused on bringing cash and in-kind benefits to individuals and families. Current interest in the importance of neighborhoods builds upon older traditions of scholarship, including work of the Chicago school in the 1920s and 1930s and ethnographies of inner-city neighborhoods in the 1950s and 1960s.

Since the Council’s Committee for Research on the Urban Underclass was first appointed in June of 1988, it has given considerable attention to the role of neighborhoods and communities in the processes that help create, maintain, or ameliorate the condition of an urban underclass. Communities and neighborhoods have been conceptualized initially as an intermediate level of analysis between the macro level of cities, states, regions, and nations and the micro level of families and individuals. As such, communities and neighborhoods provide an intellectual and strategic fulcrum for the multilevel analysis which the program seeks to facilitate (Gephart and Pearson 1988). It is at this level that research can analyze the effects of changing labor markets, the roles of other institutions in constructing the social reality that people experience, and the impact of public policy and private philanthropy in creating resources and supporting institutions.

The committee began its consideration of the role of communities and neighborhoods by convening two planning meetings. On October 19–21, 1988, a group met to consider research on the relationship between acutely impoverished urban communities and neighborhoods, family systems, and individual development.2 On January 29-30, 1989, a second group was convened to consider what is known about: (I) the ways in which institutions in neighborhoods and communities shape and mediate the effects of broader changes in society, the economy, and culture; and (2) the ways in which neighborhoods and communities shape individual outcomes (statuses and behaviors), and interactions within and among families and groups, in ways that amplify, maintain, or help to overcome these conditions.3 Both groups were asked to consider what unanswered questions, debates, and controversies exist, for which further research would be helpful.

Discussions at these two planning meetings, and subsequent deliberations of the committee and its working groups have focused on five interrelated themes: (1) the effects of neighborhoods on individual outcomes; (2) the units and characteristics of neighborhoods and communities that mediate broader changes and shape individual outcomes; (3) models of neighborhood change; (4) the intersection of neighborhoods and more macro structures and processes; and (5) relationships between communities and neighborhoods, family process, and individual development. This article considers some of the unanswered questions and challenges for research that have emerged from these discussions.

Effects of neighborhoods on individual outcomes

Most quantitative research on neighborhoods has investigated the effects of neighborhood composition on individual statuses and behaviors. In their review of the consequences of growing up in a poor neighborhood, Jencks and Mayer (1989) review correlational research which examines the effects of the socioeconomic composition and racial mix of schools and neighborhoods on a range of outcomes for adults and children. Most of the research reviewed asks simply whether there is an effect; few studies focus on the magnitude or mechanism of effect. Jencks and Mayer report weak and inconsistent results, and they emphasize the limitations of the research literature that they review, noting a considerable number of shared methodological flaws. The outcomes included in Jencks and Mayer’s review were: educational attainment, cognitive skills, teenage crime, sexual behavior, and eventual labor market success for children; and economic effects, law abidingness, and life satisfaction for adults. Research has also investigated the effects of neighborhoods on school dropout and teen pregnancy. Crane (1989b), for example, using data from the Neighborhood Characteristics File of the 1970 Census, found that neighborhood quality had very large effects on school dropout and teen pregnancy “in the very worst neighborhoods.”

“The central issue is the extent to which a mismatch between the location of jobs and the location of workers explains the lower employment of black males relative to white males.”

Scholars have given renewed attention to the effects of neighborhoods on residents’ chances of finding a good job—frequently referred to as the spatial mismatch hypothesis. The hypothesis is one of several that have been proposed to account for the rising unemployment of inner-city black males. Moreover, the research which has attempted to examine this hypothesis illustrates some of the difficulties and challenges of research on the effects of neighborhoods on individual outcomes. The central issue is the extent to which a mismatch between the location of jobs and the location of workers explains the lower employment of black males relative to white males. Studies based on data from 1970 or earlier failed to provide convincing evidence for a spatial mismatch explanation. Since 1970, joblessness has risen faster among central-city blacks than among suburban blacks, however, and blacks with low levels of education have obtained higher wages in the suburbs (Jencks and Mayer 1988, 1989; Holzer 1989; Kasarda 1989). Kasarda argues that growth in low-skill job opportunities in closer proximity to poorly educated black suburban residents is the most plausible explanation for these facts. However, it is unclear to what extent these facts reflect changes in the characteristics of the people who reside in each place (i.e., the effects of elective migration), in the labor markets that they face (e.g., tight labor market in the suburbs), and/or the extent to which they reflect problems created by location per se for those in central cities.

In their discussion of directions for future research on the spatial mismatch hypothesis, Jencks and Mayer (1989) argue that we need to: (1) assemble times series data that will compare different racial and ethnic groups and subgroups within them; (2) obtain more fine-grained descriptions of specific cities, including maps which depict neighborhoods in terms of earnings, types of employment, and rates of labor-force participation; (3) examine the effects of changes in cities’ residential and employment patterns; and most important, (4) investigate the determinants of residential movements between central cities and suburbs. Scholars also need to clarify and measure what is thought to make a difference in the mismatch between jobs and workers—for example, inadequate access to job networks. Moreover, such mechanisms need to be considered in relation to other barriers to access to employment.

Tienda (1989) identifies several major problems that have clouded research on the effects of neighborhoods on individual behavioral outcomes. Her criticisms imply a number of challenges for future work. Researchers need to include the dimension of social interaction in their definition and measurement of neighborhoods, since such interaction is often the implied mechanism through which neighborhoods are thought to affect individuals. Thus, poor people do not necessarily interact with their affluent neighbors simply because they live in the same census tract. More generally, research needs to specify and measure the mechanisms through which neighborhood effects are thought to operate. Only then will it be possible to distinguish among alternative processes such as contagion processes, socialization, or social comparison processes. Studies also need to measure exposure to the social environments that allegedly influence individual behavior. This is especially important in the light of evidence that chronically poor people change their residence frequently. For individuals who have changed residences, research could include either information about residential mobility, or preferably, information about changes in the characteristics of neighborhoods and neighbors themselves. Finally, it is crucial that research begins to assess and model the nonrandom selection processes that bring together individuals with particular socioeconomic characteristics and behavioral dispositions within spatially defined arenas. This is essential if one is to demonstrate that the characteristics of a neighborhood, rather than elective migration, account for observed associations between neighborhoods and individual outcomes.

“Comparative studies are needed of different neighborhoods, and such research needs to be linked to more quantitative analyses.”

For some time, ethnographic research has argued for the importance of neighborhoods in understanding poverty. Several recent studies have identified distinctive patterns of crime, employment, marriage, resources, and institutions in different urban poor neighborhoods (Anderson 1978; Sullivan 1983; Williams and Kornblum 1985). Most ethnographic accounts also suggest considerable diversity within neighborhoods, with the implication that there are varying social networks in a neighborhood within which a person can embed himself or herself with dramatically different consequences. Anderson (1989) and Williams and Kornblum (1985) emphasize the difference in outcomes between adolescents who are “on the street” and those whose behavior is monitored and supervised by parents and other adults. These accounts suggest that one cannot rely on compositional statistics to understand the way in which neighborhoods affect individual statuses and behaviors. What is lacking in the more qualitative research, however, are systematic analyses of the findings of different studies or any evidence concerning the incidence of patterns identified. Comparative studies are needed of different neighborhoods, and such research needs to be linked to more quantitative analyses.

Neighborhoods and communities: Units and characteristics

What are the agents and units that need to be studied? A key issue for research on neighborhoods is the geographic and/or social unit that is used to define and circumscribe them. Research often measures a neighborhood as a census tract, a block, or even a housing unit, depending on the availability of data. But a neighborhood selected and measured in this way may not correspond with the neighborhood of interaction or of self-location by its residents. Moreover, insofar as neighborhood has a geographical referent, its meaning depends upon context and function. Shifts in technology and in the spatial organization of cities have altered definitions profoundly (Katz 1989). Even today, both the objective and subjective definitions of neighborhood may vary across classes.

A number of characteristics of neighborhoods and communities seem important for understanding how neighborhoods and communities mediate broader changes in society, the economy, and culture in ways that shape outcomes of individuals, families, and other groups. Among them are: the concentration and persistence of poverty; the extent of residential segregation; the extent of social isolation; the quality of the housing stock; the extent of crime and drug use; the nature and effectiveness of formal and informal resources; also, the number and functioning of institutions, especially schools, social welfare and child care organizations, businesses, community-based organizations, and recreational facilities; the number of available jobs; and the hiring preferences and criteria of employers. Less tangible characteristics of neighborhoods such as their competence, social accountability, and collective empowerment seem no less important.

Clearly, the relevant units vary by behavior and domain, and they depend upon the outcome or process of interest. Of particular importance are relationships among neighborhood characteristics over time, the overall climate of a neighborhood, and the dynamics of neighborhood change. Some have argued, for example, that the deteriorating structure of economic opportunity and the decline in resource stocks in inner-city neighborhoods have led to the spread of crime, drug use, and other social problems. These in turn may have diminished the density of social ties and undermined the effectiveness of institutions, social networks, household economies, and family functioning.

Beyond determining the size, boundaries, and characteristics of neighborhoods, an even more difficult issue is how to characterize them in ways that best link their properties to social and personal variation. The measures typically available in studies of census tracts, for example, are relatively remote from perception and action, and therefore make strong linkages unlikely between neighborhood characteristics and outcomes for inner-city poor. More proximal characteristics of the neighborhood, at the level of its social organization and institutional functioning, are more difficult to obtain and often requires community study or ethnographic observation, but they are likely to yield more powerful linkages. Some argue that the strongest linkages between neighborhoods and individual outcomes should emerge—logically—from characterizations of the perceived neighborhood—its norm, opportunities, barriers, dangers, models, controls, pressure, and supports as seen by its residents (Jessor 1968).

“Special attention is needed to understand the social cohesion of units and the mechanisms through which characteristics of neighborhoods affect families and individuals.”

More conceptual work is needed to identify the units for study in a longer-term research agenda. Special attention is needed to understand the social cohesion of units and the mechanisms through which characteristics of neighborhoods affect families and individuals. Perhaps the greatest theoretical challenge is to formulate concepts at the different levels of analysis and to identify the linkages among these levels. Such formulations should help us to specify the logical connections between macro processes, mediating processes, and outcomes (status, behavior) at the level of persons or groups.

Models of neighborhood change

Most of the people who have worked on neighborhood effects have not studied the mechanisms by which neighborhoods get to be the way they are. Yet, understanding the paths and processes of neighborhood changes may be quite important for predicting the pattern of neighborhood effects. Crane, for example, proposes an epidemic model of neighborhood change, using a mathematical model of the spread of infectious diseases to suggest how social problems may increase exponentially within the most disadvantaged neighborhoods (Crane, 1989a). The key implication of the model is that there may be critical points in the incidence of social problems in neighborhoods. If the incidence stays below a critical point, it will tend to remain at some relatively low-level equilibrium. But if it reaches a critical point—”a tipping point” (Schelling 1971), social problems will explode. Another implication of this model is that the pattern of neighborhood effects should take a very specific form. That is, neighborhood effects should be much stronger in the most disadvantaged neighborhoods than anywhere else. Using data from the Neighborhood Characteristics Files of the 1970 Census, Crane recently documented that in neighborhoods with a small proportion of professional jobs the effects of neighborhood characteristics on dropping out of school and teenage childbearing are quite different (and much greater) than when we look at the entire distribution of neighborhoods (Crane 1989b).

Most models of neighborhood change posit a general sequence, or a set of stages through which neighborhoods typically pass (White 1987). Some argue that there are points at which local institutions (e.g., schools, social service systems) actually contribute to the reproduction of social pathologies (Fine 1983). Other models suggest that a community or neighborhood deteriorates because a protective or a corrective mechanism (middle-class role models, recreational facilities) is not present or is inoperative. The absence of such mechanisms, it is proposed, reduces the immunity of neighborhoods to deleterious effects of larger social, economic, or cultural changes. Conversely, communities or population groups which possess these protective capacities may actually benefit from these larger social, economic, or cultural changes. Jencks suggests, for example, that changes in norms toward marriage, divorce, sex, and out-of-wedlock births have further disadvantaged the poor while benefiting the upper middle class (Jencks 1988).

From the standpoint of understanding whether neighborhoods become physically and spatially isolated, or whether they experience renewal, the minimum and necessary conditions that can reverse or halt the process of neighborhood decline need to be ascertained. Specification is needed of the conditions under which neighborhoods experiencing change in the racial and ethnic composition of the population, declining economic bases, or aging housing stocks are destined to become underclass neighborhoods, and under what conditions they are likely to remain viable working-class neighborhoods, even if with lower resource stocks. Such analysis will require multidimensional models of neighborhood change, and a better understanding of the intersection between neighborhoods and broader economic, social, and political structures and processes.

Intersection of neighborhoods with broader structural forces

“Broader economic, social, and political forces affect the patterns and processes of neighborhood change in ways that may lead to beneficial or deleterious outcomes for individuals.”

Broader economic, social, and political forces affect the patterns and processes of neighborhood change in ways that may lead to beneficial or deleterious outcomes for individuals. (For a discussion of research on macro level structures and processes, see Pearson 1989.) These broader forces are also mediated by local institutions in ways that may attenuate or exacerbate their effects. Understanding each of these processes is important for developing models of neighborhood change, and for specifying the role of neighborhoods and communities in a fuller theoretical treatment of the structures and processes which generate, maintain, or help to overcome the conditions of the urban underclass.

Effects on neighborhood change. Scholars have suggested several ways in which broader economic, social, and political changes may have affected neighborhood change in inner cities. The interaction between structural shifts in the distribution of income and patterns of racial segregation has led to increasingly geographically concentrated urban poverty for blacks outside of the West and for Hispanics in the Northeast (Massey and Eggers 1989). An economic restructuring in which jobs requiring lower education have been replaced by knowledge-intensive white-collar jobs may have led to increasing rates of joblessness for inner-city blacks (Kasarda 1989). The out-migration of both white middle-income residents (Kasarda 1978) and of black middle- and working-class families (Wilson 1987) may have led to economic and institutional decline in inner-city neighborhoods.

There are several ways in which these changes may have contributed to the decline of inner-city neighborhoods. Kasarda argues that the loss of blue-collar employment and the exodus of white middle-income residents drained the city tax base and further diminished the number of blue-collar service jobs such as domestic workers, gas stations attendants, and local delivery personnel. Concurrently, many secondary commercial areas of cities withered as lower income levels of minority residential groups that replaced the suburbanizing whites could not economically sustain them (Kasarda 1978). Wilson argues that the out-migration of working- and middle-class blacks from the ghettos left behind concentrations of the most disadvantaged with the least to offer in terms of marketable skills, role models, and economic and familial stability. Under such conditions, ghetto problems magnified. Kasarda suggests that the flight of working- and middle-class blacks from the ghettos may also have led to the closing of black-owned stores and shops, and to the flow of income from black earnings out of the black community The economic decline of inner cities may have been exacerbated by redlining (Bradbury, Case, and Dunham 1989), and by the subsequent decline of schools and other local institutions.

The mediating role of local institutions. A variety of institutions at the local level have been shown to attenuate or exacerbate the effects of broader structural forces. Among them are community-based organizations, local labor markets, and schools. Stories can be told in many communities about the dramatic changes effected by the leadership of grass-roots community organizations. Some have suggested, however, that dramatic instances of neighborhood renewal may depend upon the dynamism of the local economy. Some evidence concerning the effects of tight vs. slack local labor markets is provided by recent analyses of data on Boston. Freeman and Osterman report that local labor market shortage conditions in Boston have improved the employment of disadvantaged youths, increased their earnings, and reduced their poverty during a period when each of these measures of economic well-being deteriorated nationally for disadvantaged youths (Freeman 1989; Osterman 1989). Osterman notes, however, that conditions for Hispanics have barely changed. Unfortunately, we do not know to what extent these economic effects may have ameliorated other social ills.

Schools are particularly important institutions that mediate the effects of broader structural forces. In terms of the number of dollars spent, schools use more societal resources than other institutions at the neighborhood and community level. Moreover, when one asks what the likelihood is of a child’s transition out of the underclass, the potential role of schools comes immediately to mind. Formal schooling increasingly sorts individuals into classes of service workers and professional employees, and the changing nature of the economy suggests that the power of schools as sorting mechanisms in our society is likely to increase.

Existing research suggests that the most successful programs, the most effective instructional approaches, and the family and community supports that have been demonstrated to improve educational outcomes, are least likely to be found in large urban public schools with high proportion of disadvantaged students. Moreover, there is an unarticulated, implicit school-family partnership, for which the intersection of class and ethnicity seems to produce substantial inequities. Ogbu emphasizes the importance of cultural models which children and families of different class and ethnic backgrounds bring to the experience of schooling (Ogbu 1988). Clark argues that poor black families vary significantly in ways that affect school success (Clark 1983). Some have suggested that poor families are more sensitive than affluent families to neighborhood and school characteristics. Yet, studies evaluating experimental interventions in schools typically fail to consider how the social structures of neighborhoods and communities and the dynamic of their change may reinforce, mediate, condition, or work at cross purposes to such interventions. Successful programs of school restructuring have often focused on social relationships, on building self-esteem, and on improving relationships between families, schools, and communities (Comer 1988).

Increasingly, researchers are examining schools as organizations, schooling as a process, and schools as differentiating environments. Such research offers promise for beginning to answer the key question of why some schools are much less likely than others to produce disadvantaged students, even in the face of deprived family and community settings (Boyd 1989). Future research needs to focus on differences between schools in the context of their relationships to families and communities.

Communities and neighborhoods, family process, and individual development

Understanding the processes by which neighborhood characteristics affect individual outcomes is complicated. The effects are likely to be indirect, with their impact depending upon the interaction of neighborhood characteristics with those of families, households, social networks, and individuals. Moreover, understanding the nature of effects on individuals, and their consequences, often requires a focus on development and developmental trajectories. Current discussions of neighborhood effects often imply that incentives operate on a person at the time of an outcome. Yet there is considerable evidence that some outcomes of interest—such as dropping out of school—are accurately predicted by events in childhood (i.e., early grade failure). In addition, research has demonstrated that fundamental developmental constructs such as efficacy, competence, and self-esteem, mediate individual outcomes. Developmental issues may also be critical in relation to the early age of exposure to things “on the street” in inner-city neighborhoods.

“Little is known about the basic family functioning that may be affected by neighborhoods and communities.”

Discussions in the planning meetings emphasized the need to think about the expectable sequences and forms of influence between communities and neighborhoods, parents and families, and developmental trajectories. Little is known about the basic family functioning that may be affected by neighborhoods and communities. Do families function differently in resource-rich neighborhoods? What are the effects of danger in the environment, of the comparative opportunities and constraints that crime and drugs or low wage jobs provide to inner-city residents, of the role of gangs, of the quality of schools?

Much of the work on disadvantaged families has focused on family structure, as measurable by census categories. But these may not be the most important distinctions. By relying on marriage rates in the census, researchers may have misstated the “single parent family” problem. Moreover, because so much of the knowledge base about families derives from white middle-class populations, its relevance to minority and poverty populations is in question. Current theory and research on family process illustrate this point. Traditional studies of family process have been restricted to intra-familial interactions and exchanges. Parenting style in relation to the extra-familial context may be equally consequential for child behavior and development. This suggests that the extra-familial environment has not been taken to be particularly problematic, but for families in poverty the external environment is by no means benign and supportive. How families have managed such environments—dangerous neighborhoods, inadequate and demoralizing schools, problem-prone peer groups, scarce resources for creative self-enhancement for children—are certainly critical aspects of any assessment of parenting style (Jessor 1988). Moreover, even within the family, existing research has not engaged those aspects that would more directly reflect the social context of poverty such as the unemployment of parents, the perception of blocked opportunity, socialization into ethnic and class identities, feelings of inefficacy, and so on. There is a need to reconceptualize the concepts of household, kinship, social network and family—to get beneath the labels, and to investigate the roles and functioning of kin and non-kin in relation to the political economy of households and the rearing of children.

A growing number of researchers have been exploring how disadvantaged youth in high-risk neighborhoods navigate their way out of poverty. Williams and Kornblum, in their account of the diverse ways of managing disadvantage, suggest that the successful youth find and use the meager resources available in disadvantaged neighborhoods (William and Kornblum 1985). The intriguing question raised by their study is how and why certain youth gravitate to safe niches. Part of the explanation for successful youth lies in the individual characteristics of youth or the “resiliency” of certain children (Garmezy and Rutter 1983). However, it is clear that resourcefulness or intelligence may equip youths for success on the streets as well as in the classroom. Other scholars have pointed to the importance of family influence. Clark (1983), for example, spells out the particular qualities of family interaction required to sponsor independence in the ghetto.

“Of special interest is how parents and other adults locate safe niches within dangerous environments and how they use community resources to foster the acquisition of positive social roles.”

Some children are shielded by protective parents and mentors from the despair, alienation, and rejection which generally characterize areas of concentrated poverty. Of special interest is how parents and other adults locate safe niches within dangerous environments and how they use community resources to foster the acquisition of positive social roles (Furstenberg 1988). It seems likely that poor neighborhoods are heterogeneous in the extent of institutional resources available to create “safe niches” for adolescents. Youth clubs, recreational centers, health agencies, public libraries, family planning clinics, and social services are unevenly distributed within disadvantaged districts. It might be hypothesized that the greater the density of these institutions designed to shield youth from the diversions of “street life,” the pull of unconventional peer relations, and the allure of the underground economy, the more likely it is that youth will find mentors or guides to conventional careers. Needed are studies of how community institutions help support the family’s efforts to instill prosocial values and behavior.

Discussion of developmental trajectories imply a concern with how well children, adolescents, and their parents are doing, how this changes over time, and how being raised in communities of varying degrees of economic and social disadvantage influences development (Brooks-Gunn et al. 1989). Developmental trajectories need to be examined in specific domains, including cognitive and academic well-being, physical and mental health, and socioemotional well-being. Increasingly, developmentalists have emphasized a more process-oriented conceptualization of child functioning, which focuses on competence, autonomy, relatedness, self-regulation, identity, and engagement/disengagement. All of these are believed to underlie successful adaptation. Moreover, all may be influenced by growing up in poor neighborhoods, attending low-quality schools, being labeled as poor and minority, and growing up in families with limited resources.

Conclusions: Committee projects and working groups

To address the unanswered questions and controversies that have been identified will require considerable conceptual and empirical work. The task is likely to be aided by the augmentation and reanalysis of existing data. But it will also require the design of new research and the collection of new data. Multilevel data sets—for example, those which include characteristics of neighborhoods, schools, families, and individuals—will be needed. Ethnographic work in different neighborhoods must be combined with census- and survey-based data, and with micro-observational studies of development and family process.

The Committee for Research on the Urban Underclass has initiated a series of projects and working groups to begin to address some of these needs.4 It has commissioned conceptual and empirical papers, and reviews of existing research. It has also cosponsored a conference to examine the evidence for the arguments which Wilson advances in The Truly Disadvantaged. The committee is planning a workshop on research on neighborhood compositional effects; it also expects to establish focal groups on the economic and social ecology of drugs and crime in American inner cities, on differences between schools in the context of their relationships to families and communities, and on housing.


Martha A. Gephart is currently an adjunct associate professor of education at Columbia University and she worked at the SSRC until 1993. She served as director of the Council’s Program on Persistent Urban Poverty from 1988 to 1993.

This essay originally appeared in Items Vol. 43, No. 4 in December of 1989. Visit our archives to view the original as it first appeared in the print editions of Items.

Posted on April 18, 2017