The Dynamics of Neighborhood Structural Conditions: The Effects of Concentrated Disadvantage on Homicide over Time and Space

Abstract

Several recent spatial analyses conclude the strong positive association typically found between neighborhood concentrated disadvantage and crime in cross‐sectional studies significantly differs across neighborhoods. It is possible this spatial variation is due to within‐neighborhood dynamics of continuity and change, as suggested by ecological theories of neighborhood crime. Using ordinary least‐squares and geographically weighted regression models, I explore the role of within‐neighborhood change on the disadvantage‐homicide relationship across Chicago neighborhoods and find that controlling for historical changes in disadvantage within neighborhoods reduces—but does not eliminate—spatial variation in the cross‐sectional relationship. Within‐neighborhood changes in concentrated disadvantage from 1970 to 2000 are positively related to homicide rates, net of the level of disadvantage in 2000. This suggests the relationship is influenced to some degree by temporal continuity or change in the neighborhood ecological structure, consistent with the dynamic conceptualization of neighborhoods inherent to ecological theories of crime like social disorganization.

Within-Neighborhood Dynamics: Disadvantage, Collective Efficacy, and Homicide Rates in Chicago

“…1 Theoretically, structural stability should have a positive effect on neighborhood organization, net of its level of disadvantage (or some other condition), producing lower-than-expected crime rates. Prior work supports this argument; Jacob H. Becker (2016) found that within-neighborhood changes in concentrated disadvantage between 1970-2000 partly accounted for spatial variation in the crosssectional association between disadvantage in 2000 and homicide in 2001-2005, and the results further suggested that changes in disadvantage were disruptive to the neighborhood and predictive of higher homicide rates. Within-neighborhood dynamics appeared to play an important role in the relationship between structural conditions and crime (at least homicide).…”

“…After the GWR procedure estimated the entire set of local regression parameters and statistics, a geographical variability test indicated there was significant spatial variation in the disadvantage parameter estimates across neighborhoods. This signified the level of disadvantage influenced homicide rates differently across NCs, as suggested by prior research (Becker 2016;Cahill and Mulligan 2007;Graif and Sampson 2009;Light and Harris 2012). Dummy variables of within-neighborhood changes in disadvantage were added to the model to examine if and how structural stability accounted for this spatial variation.…”

“…An important assumption of nonspatial cross-sectional models is spatial invariance-that the relationship between a given predictor and outcome is the same across geographic units, or "stationary" (Fotheringham, Brunsdon, and Charlton 2002). Work that directly tests this assumption generally finds it does not hold; associations between independent and dependent variables have been significantly different across counties (Light and Harris 2012), census tracts (Graif and Sampson 2009), block groups (Cahill and Mulligan 2007), and neighborhood clusters (Becker 2016). If the mechanisms linking structural conditions to neighborhood crime are embedded in the neighborhood's developmental trajectory, it is possible that within-neighborhood historical dynamics explain these results.…”

“…Some of the earliest work on collective efficacy discusses the "destabilizing potential" of neighborhood change (Sampson et al 1997:919) and points out that structural stability contributes to the construction of social ties and efforts to maintain social control. However, the influence of within-neighborhood stability on collective efficacy-and thus crime-has largely been ignored in recent work, and while Becker's (2016) findings are suggestive, the analyses do not account for collective efficacy. The effect of structural (in)stability on mediating social processes like collective efficacy represents a critical and understudied dimension of neighborhood effects and crime research.…”

“…The theoretical influence of neighborhood conditions emerges over time as a function of dynamic processes, and research that focuses on between-neighborhood differences in the level of a particular condition-like economic disadvantage-overlooks the possibility that neighborhoods that appear similar in cross-section (e.g., have comparable levels of disadvantage) may have substantially different structural histories. The current research builds on prior work that found both the cross-sectional level of disadvantage and historical changes in disadvantage were significant predictors of neighborhood homicide rates (Becker 2016) by incorporating a measure of one mechanism theorized to mediate the disadvantage-crime relationship: collective efficacy. Structural changes are not expected to directly impact homicide rates.…”