doi:10.1016/j.healthplace.2006.09.001 
Copyright © 2006 Elsevier Ltd All rights reserved.
Neighborhood social capital and adult health: An empirical test of a Bourdieu-based model
Richard M. Carpiano
, a, 
aDepartment of Sociology, University of British Columbia, 6303 Northwest Marine Drive, Vancouver, British Columbia, Canada V6 T 1Z1
Received 21 December 2005;
revised 2 August 2006;
accepted 12 September 2006.
Available online 3 November 2006.
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Abstract
Drawing upon Bourdieu's [1986. The forms of capital. In: Richardson, J.G. (Ed.), Handbook of Theory and Research for the Sociology of Education. Greenwood, New York, pp. 241–258.] social capital theory, I test a conceptual model of neighborhood conditions and social capital—considering relationships between neighborhood social capital forms (social support, social leverage, informal social control, and neighborhood organization participation) and adult health behaviors (smoking, binge drinking) and perceived health, as well as interactions between neighborhood social capital and individuals’ access to that social capital. Analyzing Los Angeles Family and Neighborhood Survey data linked with tract level census data, results suggest that specific social capital forms were directly associated with both positive and negative health outcomes. Additionally, residents’ neighborhood attachment moderated relationships between various social capital forms and health. Future studies should consider social capital resources and the role of differential access to such resources for promoting or compromising health.
Keywords: Social capital; Neighborhoods; Bourdieu; Health inequalities
Fig. 1. Proposed conceptual model of neighborhood social capital processes on individual health outcomes.
Fig. 2. Average perceived health associated with informal social control by each level of individual neighborhood attachment.
Fig. 3. Average perceived health associated with neighborhood organization participation by each level of individual neighborhood attachment.
Table 1.
L.A.FANS items used for computing neighborhood-level social cohesion, social capital, and individual-level neighborhood attachment variables
Indicates that the scale was originally coded in the reverse direction.
Table 2.
Bivariate correlations of neighborhood-level variables (n=65)
*p
0.05; **p0.01; ***p0.001.
Table 3.
Descriptive statistics of individual-level variables (n=2076)
Note: SD=Standard deviation; Range=minimum and maximum values of the variable.
Income values represent the average mean and standard deviation of the five separate samples created from multiple imputation procedures (see text for details).
Table 4.
Results for individual-level daily smoking, binge drinking twice or more in the past 30 days, and perceived health regressed on neighborhood-level social conditions and individual-level neighborhood attachment and sociodemographics
†p0.10; *p0.05; **p0.01.
Note: Neighborhood-level N=65; Individual-level N=2076; Models estimated with robust standard errors. Each model controls for individual-level neighborhood attachment and sociodemographics (see Appendix B for the individual-level estimates).
a Odds ratios (95% confidence intervals).
b Slope coefficients (Standard errors).
c All neighborhood-level variables are standardized.
Table 5.
Results for cross-level interaction models: Individual-level daily smoking, binge drinking twice or more in the past 30 days, and perceived health regressed on neighborhood-level social conditions and individual-level neighborhood attachment and sociodemographics
†p0.10; *p0.05; **p0.01.
Note: Neighborhood-level N=65; Individual-level N=2076; All models estimated with robust standard errors and control for individual-level variables (see Appendix B for the individual-level estimates).
a Odds ratios (95% confidence intervals).
b Slope coefficients (Standard errors).
c All neighborhood-level variables are standardized.
d Moderate and high neighborhood attachment compared to No/Low neighborhood attachment.
Table A1.
Results for all individual-level variables entered into the models shown in Table 4
†p.10; *p.05; **p.01.
Note: Neighborhood-level N=65; Individual-level N=2076; Models estimated with robust standard errors.
a Odds ratios (95% confidence intervals).
b Slope coefficients (Standard errors).
c Low neighborhood attachment=referent.
d White/Caucasian non-Hispanic=referent.
e Married=referent.
Table A2.
Results for all individual-level variables entered into the models shown in Table 5
†p.10; *p.05; **p.01.
Note: Neighborhood-level N=65; Individual-level N=2076; Models estimated with robust standard errors.
a Odds ratios (95% confidence intervals).
b Slope coefficients (Standard errors).
c Low neighborhood attachment=referent.
d White/Caucasian non-Hispanic=referent.
e Married=referent.
Abstract
Purchase PDF (219 K)
E-mail Article
Add to my Quick Links
Cited By in Scopus (6)
Purchase PDF (394 K)
