Predictors of and Spatial Relationships Among Neighborhood Social Organization in Chicago

Background and Purpose: Residential neighborhoods contribute to individual and family outcomes including mental and physical health and later-life earnings. Theory suggests that neighborhood social organization (NSO) is important for why neighborhoods matter and how macro social workers can intervene. NSO encapsulates how neighborhoods differentially organize to exert social control over their community. It can be both formal and informal. Formal NSO consists of established organizations such as social services and block groups. Informal NSO captures residents acting to address neighborhood problems. 

NSO can inform macro social work interventions, but we need to understand what promotes and deters it. Researchers have found both positive and negative effects of neighborhood disadvantage on formal organizations. Neighborhoods of color have fewer organizational resources across types of organizations. Collective efficacy, a component of informal NSO, is higher in less disadvantaged and more stable neighborhoods. Little is known about how NSO is spatially distributed across neighborhoods. We therefore investigate whether neighborhood advantage, disadvantage, and racial composition relate to formal and informal NSO. Next, we test relationships between types of NSO. Finally, we examine whether spatial relationships contribute to social organization.

Methods: We used data from the 2001-2003 Chicago Community Adult Health Study, a face-to-face survey of 3105 adults (ages 18-92). We created a neighborhood-level file of aggregated survey responses, InfoUSA data on organizations, and the 2000 US Census. Neighborhoods (N=343) are diverse, ranging from 0% to 94% white (M=27%).

Measures. Formal NSO was measured in two ways: organizational resources was the rate of InfoUSA social service organizations and organizational participation was aggregated resident-reported participation. Informal NSO is measured using collective efficacy (Sampson et al., 1997). We generated a queen contiguity matrix for spatial analysis using geographic information systems (GIS).

Analysis. We conducted global Moran’s I to examine spatial autocorrelation. Next, we conducted Lagrange Multiplier (LM) tests to test spatial lag and spatial error models against OLS models to check for neighborhood-level spatial diffusion and unobserved spatial relationships, respectively.

Results: Moran’s I results indicated that each measure of NSO was significantly spatially clustered in Chicago. Based on LM test results, we fit spatial lag models for organizational resources and participation and a spatial error model for collective efficacy. Neighborhood demographic characteristics were significant, but inconsistent predictors of each aspect of NSO. Organizational resources and participation were both associated with more affluence while collective efficacy was associated with lower disadvantage, lower affluence, and higher percent white. Organizational resources were positively associated with organizational participation (Coef=0.62, p<.05) but negatively associated with collective efficacy (Coef=-1.04, p<0.05). Organizational participation was positively associated with organizational resources (Coef=0.02, p<.05) and collective efficacy (Coef=0.35, p<0.01). Both had significant spatial lag in the final models. Collective efficacy was higher in neighborhoods with fewer organizational resources (Coef=-0.01, p<0.05) and more organizational participation (Coef=0.11, p<0.01). Spatial error remained significant in the final model.

Conclusions and Implications: Practitioners in place-based interventions should consider how 1) components of NSO within the neighborhood promote or supplant each other and 2) NSO outside the neighborhood influences an intervention’s outcomes.