Elsevier

Social Science Research

Volume 31, Issue 4, December 2002, Pages 539-575
Social Science Research

The estimation of neighborhood effects in the social sciences: An interdisciplinary approach

https://doi.org/10.1016/S0049-089X(02)00005-4Get rights and content

Abstract

This paper critically examines the interdisciplinary research of neighborhood effects. Neighborhood effects are community influences on individual social or economic outcomes. Examples include labor force activity, child outcomes, criminal behavior, and other socioeconomic phenomena. The existing theoretical and empirical literature is reviewed. Conceptual definitions from sociology are linked and contrasted with economic models. Early studies are criticized for failing to account for a number of endogeneity concerns. Recent empirical studies are also explored. The results of these papers demonstrate that although neighborhood characteristics are important, their influences are much smaller than suggested by previous research. An additional emphasis is dedicated to linking the neighborhood effects literature to other research of local or spatial activity within economics, sociology, and geography. In particular, the potential employment of local interaction game theory and spatial econometrics in neighborhood effects research is discussed. Additionally, a catalogue of existing empirical research is assembled.

Introduction

The debate over the magnitude and even the existence of neighborhood effects is a controversial area of inquiry in the social sciences. Naturally, part of this controversy arises from the policy implications that result from influential studies. However, the study of neighborhood effects is also an interdisciplinary topic. Economics, sociology, geography, and the other social sciences possess their own accepted research tools and terminology. Therefore, nominally similar research questions regarding neighborhood effects are addressed by remarkably dissimilar techniques. In some cases, researchers do not recognize these disparities because they are unaware of studies within another branch of social science.

What is a neighborhood effect? The definition varies according to the researcher, but in general a neighborhood effect is a social interaction that influences the behavior or socioeconomic outcome of an individual. Neighborhood effect research includes, but is not restricted to, models of endogenous preference effects, peer effects, and compositional effects. Neighborhood effects also include influences on individual behavior or outcomes due to the characteristics of an individual’s neighbors and neighborhood. For example, population sorting may result in exposure to dissimilar institutional influences among neighborhoods. There is also an explicit spatial aspect to neighborhood effect research. Typically, this spatial relationship is defined with respect to location of residence. However, a measure of social distance may also be appropriate.

There are several notable literature reviews on neighborhood effects. Jencks and Mayer (1990) wrote an influential early survey of the field. In the succeeding decade, estimation techniques have developed considerably. Haveman and Wolfe (1995) review empirical findings involving children. Galster and Killen (1995) pay particular attention to geographic and spatial contexts of neighborhood variation. Brooks-Gunn et al. (1997) review many studies, mostly in sociology, while presenting some original findings. Ellen and Turner (1997) examine the importance of neighborhood influence and review some empirical findings. Brock and Durlauf (2001) provide a thorough examination of the econometric aspects of all types of social interactions, including neighborhood effects.

This paper seeks to contribute to the field by examining prominent techniques and results, with an emphasis on finding linkages among disciplines and related areas of inquiry. I show that the skepticism exhibited by some authors with respect to the claims of early studies of neighborhood effects is well founded. This caution is due to the methodological challenges in estimation. In some cases, statistical problems invalidate the results of the study.

Section snippets

Classifications of neighborhood effects

The first major distinction in the study of neighborhood effects is with respect to the geographic scope of the effect. The impact of the neighborhood effect may be within or among neighborhoods. In almost all cases, the existing research examines within neighborhood effects. This review reflects this emphasis. In studies of within neighborhood effects, no interaction occurs among the neighborhoods; that is, the neighborhood possesses no spillover characteristics. Thus, neighborhoods with

Sociological models

Sociologists are responsible for identifying the mechanisms by which a neighborhood effect may arise and operate. Jencks and Mayer (1990) classified these ideas into several types. Contagion theories primarily investigate peer influences that are responsible for the spread of social ills, as the negative connotation of the word contagion implies. A large number of existing studies examine contagion theories. Examples include Case and Katz (1991), Evans et al. (1992), and Corcoran et al. (1992).

Empirical estimation challenges

There are a number of serious empirical complications that researchers of neighborhood effects must confront. The first two concerns, endogenous and omitted independent variables, result in biased parameter estimates. This is a critical problem because tests for the existence of neighborhood effects typically examine the sign and significance of regression parameters for neighborhood characteristics. A third problem with the estimation of neighborhood effects is that identification of the

A critique of selected empirical studies

A number of prominent studies of neighborhood effects are subject to the above concerns regarding estimation. This section of the review examines some of these studies and discusses problems with their results. A catalogue of existing empirical studies appears in Table 1. Datcher (1982) is one of the first empirical tests for neighborhood effects. This longitudinal study of Panel Study of Income Dynamics (PSID) data examined earnings and years of schooling for males, ages 23–32. From age 13–22,

A critique of selected studies accounting for sources of bias

Evans et al. (1992) offer evidence that highlights the concerns raised in this review. The authors recognize that peer group sorting is an endogenous process. Furthermore, they correct the problem by using instrumental variables. Like other studies, they are concerned with the determinants of teen pregnancies and school dropouts. First, they estimate their model using a single equation probit specification. For this estimation, peer group formation is assumed predetermined. They find that peer

Geographic and spatial issues

A recurring concern in the study of neighborhood effects is the operational geographic definition of neighborhood. Some economists and sociologists possess a reluctance to approach this issue with anything more than a conceptual explanation of its complications. Given data set constraints, this is a natural position. On the other hand, geographers have given this topic considerable attention. I briefly identify a few research tools and issue here. However, this summary of issues is not

Conclusion

The search for neighborhood effects involves contributions from the different social sciences and is an issue that is likely to receive increased policy attention due to a renewed focus on decentralized social organization and governance. The initial supportive evidence for the existence of neighborhood effects has been called into question in recent years. This caution concerning the interpretation of empirical results is well founded due to both the empirical challenges in estimation and the

References (82)

  • M. Bertrand et al.

    Network effects and welfare cultures

    The Quarterly Journal of Economics

    (2000)
  • K. Bollen

    Structural Equations with Latent Variables

    (1989)
  • K. Brewster

    Race differences in sexual activity among adolescent women: the role of neighborhood characteristics

    American Sociological Review

    (1994)
  • W. Brock et al.

    Interactions-based models

  • J. Brooks-Gunn et al.

    Neighborhood Poverty: Context and Consequences for Children

    (1997)
  • J. Brooks-Gunn et al.

    Do neighborhoods influence child and adolescent development?

    American Journal of Sociology

    (1993)
  • A. Bryk et al.

    Hierarchical Linear Models: Applications and Data Analysis Methods

    (1992)
  • Case, A.C., Katz, L.F., 1991. The company you keep: the effects of family and neighborhood on disadvantaged youths....
  • M. Corcoran et al.

    The association between men’s economic status and their family and community origins

    Journal of Human Resources

    (1992)
  • C.J. Coulton et al.

    Community level factors and child maltreatment rates

    Child Development

    (1995)
  • J. Crane

    The epidemic theory of ghettos and neighborhood effects on dropping out and teenage childbearing

    American Journal of Sociology

    (1991)
  • D. Cutler et al.

    Are ghettos good or bad?

    The Quarterly Journal of Economics

    (1997)
  • L. Datcher

    Effects of community and family background on achievement

    The Review of Economics and Statistics

    (1982)
  • T. Dieckmann

    Stochastic learning and the evolution of conventions

    Constitutional Political Economy

    (1998)
  • Dietz, R., Haurin, D., 2002. The social and private micro-level consequences of homeownership....
  • S.M. Dornbusch et al.

    Community influences on the relation of family statuses to adolescent school performance: differences between african americans and non-hispanic whites

    American Journal of Education

    (1991)
  • Dumanovsky, T., Fagan, J., Thompson, P., 1999. The neighborhood context of crime in NYC’s public housing projects....
  • G.J. Duncan

    Families and neighbors as sources of disadvantage in the schooling decisions of black and white adolescents

    American Journal of Education

    (1994)
  • C. Duncan et al.

    Using multilevel models to model heterogeneity: potential and pitfalls

    Geographical Analysis

    (2000)
  • S.N. Durlauf

    A theory of persistent income inequality

    Journal of Economic Growth

    (1996)
  • I. Ellen et al.

    Does neighborhood matter? Assessing recent evidence

    Housing Policy Debate

    (1999)
  • D. Elliott et al.

    The effects of neighborhood disadvantage on adolescent development

    Journal of Research in Crime and Delinquency

    (1996)
  • G. Ellison

    Learning, local interaction, and coordination

    Econometrica

    (1993)
  • S. Ennett et al.

    School and neighborhood characteristics associated with school rates of alcohol, cigarette, and marijuana use

    Journal of Health and Social Behavior

    (1997)
  • M.E. Ensminger et al.

    School leaving: a longitudinal perspective including neighborhood effects

    Child Development

    (1996)
  • D. Epple et al.

    Competition between private and public schools, vouchers, and peer-group effects

    American Economic Review

    (1998)
  • I. Eshel et al.

    Altruists, egoists, and hooligans in a local interaction model

    American Economic Review

    (1998)
  • W.M. Evans et al.

    Measuring peer group effects: a study of teenage behavior

    Journal of Political Economy

    (1992)
  • A.S. Fotheringham et al.

    The modifiable areal unit problem in multivariate statistical analysis

    Environment and Planning

    (1991)
  • G. Galster et al.

    The geography of metropolitan opportunity: a reconnaissance and conceptual framework

    Housing Policy Debate

    (1995)
  • C.L. Garner et al.

    Neighborhood effects on educational attainment: a multilevel analysis

    Sociology of Education

    (1991)
  • Cited by (324)

    View all citing articles on Scopus

    I thank Donald Haurin and Bruce Weinberg for their invaluable assistance. I also thank Morton O’Kelly and the participants of The Ohio State University’s Urban and Regional Roundtable. In addition, this paper benefited from the comments of two anonymous referees. Any remaining errors or omissions are my own.

    1

    The author is a Ph.D. candidate in Economics and a research associate at the Center for Urban and Regional Analysis at The Ohio State University.

    View full text