Spatial Dimensions of Social Capital

Background: This study examined the impact of spatial dynamics on social capital formation among a sample of public housing residents. Neighborhood-based research has relied on administrative boundaries to measure neighborhood effects and their impact on social capital formation. However, emergent thinking calls into question the saliency of these boundaries and invites us to consider the various spaces – cognitive, geographic, social, and virtual – in which social capital is generated beyond the neighborhood. Envisioning space as a multidimensional concept accounts for the ways in which individuals perceive their community contexts, the dynamic nature of the spaces we occupy, and the role of social interactions in determining the amounts and types of social capital available.

Methods: Data were collected using a cross-sectional survey design. Participants were selected using proportionate random sampling from housing authority provided lists of resident addresses in two public housing communities (N=135). Most participants were African American (98.5%) females (88.9%) who identified as poor and had lived in the neighborhood an average of 6 years. Social capital was assessed using Sampson and Graif’s (2009) multi-component scale consisting of 14 unique dimensions. Participants were asked to draw their neighborhood boundaries on a paper map; those boundary polygons were digitized in ArcGIS 10. Average size of cognitive neighborhoods was 0.4 square miles (SD=.34) similar to median size Coulton and Jennings (2013) found.

We created a composite social capital variable by summing the collective efficacy, local network, organizational involvement, and conduct norms constructs Sampson and Graif (2009) found in their factor analysis. We conducted an OLS regression to examine how the composite social capital variable predicted the size of one’s cognitive neighborhood, controlling for demographic variables. Because this social capital construct is a summation of 14 individual dimensions, and because these dimensions have rarely been used to examine the size of one’s cognitive neighborhood, it was important to explore further whether or not individual dimensions were better predictors of the size of one’s cognitive neighborhood than was the composite social capital construct. We conducted 14 different OLS regression analyses.

Results:  Composite social capital was not a significant predictor of the size of one’s cognitive neighborhood. Of the four factors, collective efficacy (cohesion, social control, moral cynicism, neighborhood attachment, and police efficacy) was predictive of the size of one’s cognitive neighborhood (β = 0.23; p < .01). Qualitative analysis of the cognitive neighborhood maps reveals larger polygons included resources (religious institutions, medical centers, grocery stores, services) necessary to achieve instrumental and expressive returns. 

Implications: These findings suggest that individuals expand their neighborhood space to build larger, more diverse ties that have an effect on bonding social capital. Findings challenge conventional wisdom that strong ties are necessary to govern neighborhood processes; instead, weak ties are also important in the development of neighborhood-level social capital. Social capital is built in the various social and professional spaces we occupy. Continuing to focus on narrowly defined neighborhood effects may bias our estimates and prevent a more robust understanding of how individuals build social capital.