Neighborhood effects in a European city: Secondary education of young people in Helsinki

https://doi.org/10.1016/j.ssresearch.2006.04.003Get rights and content

Abstract

Research on neighborhood effects on youth educational outcomes has been mainly an American tradition. This study reviews the European evidence on this subject and brings new evidence, focusing on Helsinki, the capital of Finland. Register-based data and multi-level statistical analysis are used to assess the existence and form of neighborhood effects on the educational careers of young people in Helsinki. The results indicate that there is no effect on the completion of secondary education generally, but there is an effect on the type of education that is completed. This effect is nonlinear: the affluent neighborhoods differ from the rest of the neighborhoods. The paper concludes by stating that research on neighborhood effects should better recognize the existence of “soft” effects.

Introduction

The concept of a neighborhood effect implies that when trying to understand the behavior of individuals we should take into account the neighborhoods in which they live. This concept has been used in the social sciences especially in regard to the potentially harmful life-course consequences of living in segregated poverty neighborhoods of large cities and is often connected to an analysis of exogenous macro-level processes that produce these neighborhoods (e.g., Wilson, 1987, Gephart and Brooks-Gunn, 1997).

Research on neighborhood effects has been dominated by American studies, and the bulk of the studies have concentrated on adolescents (Ellen and Turner, 1997, Sampson et al., 2002), who are considered to be especially susceptible to neighborhood effects. Jencks and Mayer (1990) conducted an extensive review of existing studies, but could not find enough rigorous empirical evidence to warrant far-reaching generalizations about neighborhood effects on children and adolescents. Many new studies have since been carried out, and the methodology has advanced. American evidence generally confirms the existence of neighborhood effects especially on educational attainment (e.g., Ellen and Turner, 1997, Gephart, 1997, Crowder and South, 2003, South et al., 2003), although it is not clear which neighborhood characteristics matter most (Ellen and Turner, 1997). There are still methodological challenges (Furstenberg and Hughes, 1997, Sampson et al., 2002, Dietz, 2002), alternative explanations for the effects observed (Bauder, 2002) and some conflicting results (Plotnick and Hoffman, 1999).

The results of American studies cannot be directly generalized to the European context, which differs in several respects from the American context. The main difference is the heritage of the racial divide in the USA, in other words the continuing significance of race and its strong connection with residential differentiation (see Wacquant, 1993). The results of American studies on neighborhood effects are also different for different racial groups (e.g., López Turley, 2002, Crowder and South, 2003), which makes it even more difficult to transfer the findings to less ethnically divided contexts. Income differentiation is also higher in the USA, and housing and residential patterns are more closely related to work and income than in the European welfare states (Kintrea and Atkinson, 2001, Ostendorf et al., 2001). In European cities, the redistributive policy of the welfare state and local public administration’s stronger instruments to steer urban development (Bagnasco and Le Galès, 2000, Häussermann and Haila, 2004) are reflected in a lower level of segregation as compared to US cities (Friedrichs et al., 2003).1

These differences suggest that (1) there may not be enough differentiation between neighborhoods in European cities to produce neighborhood effects, and “critical” levels of neighborhood characteristics may not be reached; (2) even if there is considerable differentiation between neighborhood populations, for example in employment and occupational status, the welfare state may level down the consequences, such as income differentiation, and (3) the racial divide in the USA makes it complicated to generalize the results of American research to less divided societies.

Because the most convincing American evidence regarding neighborhood effects is related to youth educational outcomes, this is an obvious choice for searching for evidence of neighborhood effects in European cities. Educational careers are also easily identified, and they are an essential element in social mobility. Nevertheless, there is not much European research on neighborhood effects on the educational outcomes of young people. Some studies exist, however, and their findings suggest that there are neighborhood effects. Research conducted with Scottish data has shown that the deprivation level of neighborhoods had effects on the educational achievements of young people (Garner and Raudenbush, 1991), on the occupational aspirations of boys (Furlong et al., 1996) and on the likelihood of continuing education after compulsory schooling (Furlong, 1996). A study on three Swedish cities (Andersson, 2001, Andersson, 2004) achieved similar results: socio-demographic neighborhood characteristics during adolescence were shown to affect the subsequent educational attainment.

The purpose of this paper is to accumulate the scant European empirical evidence of neighborhood effects on youth educational outcomes, with reference to one European city, Helsinki, the capital of Finland. Although American texts (e.g., Ellen and Turner, 1997, Sampson et al., 2002) have acknowledged for some time that demonstrating the existence or non-existence of neighborhood effects is not enough, the European situation is different. There this basic question still lacks answers. Although studying a single city does not provide a generalizable answer, it helps to accumulate the knowledge on the subject.

The research questions are

  • 1.

    Do neighborhoods affect young people’s completion of secondary education or the type of education?

  • 2.

    If they do, which neighborhood-level factors contribute to the effects?

This paper first describes the residential differentiation in Helsinki, and then moves on to present neighborhood-effect theories, and the hypotheses, data and methods of the study. The results are then presented and hypotheses assessed, and the paper concludes with a discussion of findings.

Section snippets

Residential differentiation in Helsinki

Helsinki, the capital of Finland, forms a metropolitan area together with the municipalities of Espoo, Vantaa, and Kauniainen. In the beginning of 2005 the total population of the area was around 970,000 inhabitants (Helsinki Region Statistics), over half of whom (560,000) lived in Helsinki.

The city of Helsinki has been able to steer the development of its urban structure because municipalities have the exclusive right to town planning in Finland and the city owns two-thirds of the land within

An overview of neighborhood-effect theories

Theories of neighborhood effects on young people’s behavior have concentrated on the effects of the social environment in the neighborhood. Two characteristics of neighborhoods have been emphasized: the socioeconomic composition of the population and the social organization within the neighborhood.

Regarding the socioeconomic dimension, the theories are mostly related to the effects of either low- or high-status neighbors (or effects associated with the presence or proportion of either group).

Hypotheses of the study

Four empirical hypotheses were tested in the analysis of the second research question (which neighborhood-level factors contribute to the neighborhood effects?). They represent four possible types of neighborhood effects. The hypotheses are listed below. It is assumed that the non-random selection of adolescents into neighborhoods is controlled for before the associations between neighborhood characteristics and adolescents’ educational outcomes are considered (see Section 6). Only then can the

Individual-level data

The individual-level data set for the study was constructed by Statistics Finland (permission TK-53-80-02). The data came from register-based data, primarily from the longitudinal data file of employment statistics, which contains yearly individual-level data from 1987 onwards for all persons living in Finland. A 50% sample of non-institutionalized 15-year-olds living in Helsinki in the years 1990–1994 was used (n = 10906). This period was selected in order to have as recent data as possible,

Methods

Multi-level logistic regression was applied as the statistical method in the study. SAS Proc Nlmixed, which fits nonlinear mixed models, was used as the statistical procedure (see Wolfinger, 1999).

Multi-level analysis (for an overview, see, for example, Jones and Duncan, 1998, Teachman and Crowder, 2002) has lately become common in studies of contextual effects. The method is suitable when the usual assumption of regression analysis, independence of error terms, does not hold. This kind of a

Results

Table 2 shows the proportions of young people who had completed secondary education either in upper-secondary school or in vocational school and 95% confidence intervals (C.I.) for the proportions. Upper-secondary school was the most common choice and it was more common among the girls than among the boys. Vocational school was more common among the boys. For both sexes, however, the proportion of those with no completed secondary education was higher than the proportion of those with finished

Discussion

The results of this study indicate that there are no neighborhood effects on the probability that young people will complete secondary education in Helsinki. On the other hand, there seems to be a neighborhood effect on a somewhat “softer” outcome: the type of secondary education that is completed. The individual backgrounds explain much more, but there is still something that seems to be accounted for by neighborhoods.

Three neighborhood characteristics were considered as possible factors

Acknowledgments

I appreciate the helpful comments from the anonymous reviewers and all other commentators. The study received funding from the Academy of Finland, project number 53514, from the city of Helsinki, and from the Finnish Youth Research Network.

References (61)

  • N. Buck

    Identifying neighborhood effects on social exclusion

    Urban Studies

    (2001)
  • N. Buck et al.

    Does spatial concentration of disadvantage contribute to social exclusion?

  • J.S. Coleman

    Social capital in the creation of human capital

    American Journal of Sociology

    (1988)
  • J.S. Coleman

    Foundations of Social Theory

    (1990)
  • J. Crane

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

    American Journal of Sociology

    (1991)
  • A.V. Diez-Roux

    Bringing context back into epidemiology: variables and fallacies in multilevel analysis

    American Journal of Public Health

    (1998)
  • G.J. Duncan

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

    American Journal of Education

    (1994)
  • G.J. Duncan et al.

    Conceptual and methodological issues in estimating causal effects of neighborhoods and family conditions on individual development

  • I.G. Ellen et al.

    Does neighborhood matter? Assessing recent evidence

    Housing Policy Debate

    (1997)
  • D.S. Elliott et al.

    The effects of neighborhood disadvantage on adolescent development

    Journal of Research in Crime and Delinquency

    (1996)
  • J. Friedrichs et al.

    Neighborhood effects on social opportunities: the European and American research and policy context

    Housing Studies

    (2003)
  • A. Furlong et al.

    Neighborhoods, opportunity structures and occupational aspirations

    Sociology

    (1996)
  • A. Furlong

    Opportunity structures and the management of transitions

  • F.F. Furstenberg et al.

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

  • G.C. 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)
  • M.A. Gephart

    Neighborhoods and communities as contexts for development

  • M.A. Gephart et al.

    Introduction

  • H. Goldstein

    Multilevel Statistical Models

    (1995)
  • H. Goldstein et al.

    Partitioning variation in multilevel models

    Understanding Statistics

    (2002)
  • Cited by (0)

    View full text