Multivariate: Control for Community Factors
In our introductory remarks, we also pointed to comments by Hay et al. (2006) about the nexus between poverty and community context. Although there are too few studies in our sample that test the interaction terms implied by Hay’s discussion, there were many multivariate studies where indicators of community factors were included in models with poverty or resource deprivation and we can evaluate whether controlling for community factors matters in our evaluation of the evidence on the differential etiology of violence.
Bellair (2000), mainly interested in informal surveillance, reported that burglary and robbery/stranger assault were both significantly, positively associated with concentrated disadvantage, controlling for neighborhood characteristics such as residential stability, and whether the community was “downtown" Although Hannon and DeFronzo (1998) were looking at the impact of welfare and its interaction with resource deprivation in their models, they still found that resource deprivation was significantly associated with both violent and property crime, controlling for a variety of community factors such as “percent one-person household" residential mobility, divorce, percent youth and urbanism. Their county-level analysis, however, may miss more localized effects of neighborhood. Krivo and Peterson (1996) found that both violent and property crime were associated with high levels of poverty in models including many indicators of neighborhood disorganization. In these models, they found stronger effects of disadvantage on violent crime. In models controlling for indicators of community structure such as ethnic heterogeneity and residential stability, poverty was positively and significantly associated with assault and burglary in an analysis of Seattle census tracts (Warner & Rountree, 1997). This was true even in models which included interactions between poverty and heterogeneity and poverty and stability. In another analysis of Seattle census tract data, Wilcox et al. still found that concentrated disadvantage was significantly associated with violence and burglary, controlling for a series of indicators of community characteristics and disorder (Wilcox, Quisenberry, Cabrera, & Jones, 2004).
Thus, the studies included in our review suggest that poverty/resource deprivation is not spuriously associated with violent crime due to a common association with other community factors.