The estimation of neighborhood effects in the social sciences: An interdisciplinary approach☆
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)
Testing for strategic interaction among local governments: the case of growth controls
Journal of Urban Economics
(1998)Neighborhood influence and technological change
Regional Science and Urban Economics
(1992)The evolution of conventions with mobile players
Journal of Economic Behavior and Organization
(1999)- et al.
Coordination and local interaction: experimental evidence
Economics Letters
(1998) - et al.
Segregation, the concentration of poverty, and the life chances of individuals
Social Science Research
(1991) Using sibling data to estimate the impact of neighborhoods on children’s educational outcomes
Journal of Human Resources
(1998)A theory of social custom, of which unemployment may be one consequence
The Quarterly Journal of Economics
(1980)- et al.
Teen out-of-wedlock births and welfare receipts: the role of childhood events and economic circumstances
Review of Economics and Statistics
(1993) Spatial Econometrics: Methods and Models
(1988)Heterogeneity, stratification, and growth: macroeconomic effects of community structure
American Economic Review
(1996)
Network effects and welfare cultures
The Quarterly Journal of Economics
Structural Equations with Latent Variables
Race differences in sexual activity among adolescent women: the role of neighborhood characteristics
American Sociological Review
Interactions-based models
Neighborhood Poverty: Context and Consequences for Children
Do neighborhoods influence child and adolescent development?
American Journal of Sociology
Hierarchical Linear Models: Applications and Data Analysis Methods
The association between men’s economic status and their family and community origins
Journal of Human Resources
Community level factors and child maltreatment rates
Child Development
The epidemic theory of ghettos and neighborhood effects on dropping out and teenage childbearing
American Journal of Sociology
Are ghettos good or bad?
The Quarterly Journal of Economics
Effects of community and family background on achievement
The Review of Economics and Statistics
Stochastic learning and the evolution of conventions
Constitutional Political Economy
Community influences on the relation of family statuses to adolescent school performance: differences between african americans and non-hispanic whites
American Journal of Education
Families and neighbors as sources of disadvantage in the schooling decisions of black and white adolescents
American Journal of Education
Using multilevel models to model heterogeneity: potential and pitfalls
Geographical Analysis
A theory of persistent income inequality
Journal of Economic Growth
Does neighborhood matter? Assessing recent evidence
Housing Policy Debate
The effects of neighborhood disadvantage on adolescent development
Journal of Research in Crime and Delinquency
Learning, local interaction, and coordination
Econometrica
School and neighborhood characteristics associated with school rates of alcohol, cigarette, and marijuana use
Journal of Health and Social Behavior
School leaving: a longitudinal perspective including neighborhood effects
Child Development
Competition between private and public schools, vouchers, and peer-group effects
American Economic Review
Altruists, egoists, and hooligans in a local interaction model
American Economic Review
Measuring peer group effects: a study of teenage behavior
Journal of Political Economy
The modifiable areal unit problem in multivariate statistical analysis
Environment and Planning
The geography of metropolitan opportunity: a reconnaissance and conceptual framework
Housing Policy Debate
Neighborhood effects on educational attainment: a multilevel analysis
Sociology of Education
Cited by (324)
Neighborhood Trip. How do guests’ neighborhood preferences shape the touristification process?
2024, Tourism ManagementSpatiotemporal-aware machine learning approaches for dissolved oxygen prediction in coastal waters
2023, Science of the Total EnvironmentNeighborhood-level housing affordability and maternal depression
2023, SSM - Mental HealthNeighborhood deprivation and depressive symptoms in Chinese adolescents: Mediated by parenting styles and moderated by relative family status
2023, Journal of Affective DisordersMeasuring small-area level deprivation in Belgium: The Belgian Index of Multiple Deprivation
2023, Spatial and Spatio-temporal Epidemiology
- ☆
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.