In this chapter, we have examined the literature to see if poverty is a differential predictor of violence. Specific comparisons vary, but the body of comparisons is very clear. We conclude that poverty is consistently related to violent crime and that it is more consistently related to violent crime than nonviolent crime and the effects are often larger in magnitude. We acknowledge that the effect is strongest and most consistent when aggregate indicators of resource deprivation or concentrated disadvantage are used. When standard indicators of poverty were employed, the findings are less consistent and this begs the question of why gross measures of poverty are insufficient. Our own analysis of Add Health data and a brief report of NYS data presented here (Table 9.2), suggest that, in those data sets, self-report violent offenders have lower income and socioeconomic levels than nonviolent offenders.
In addition, we must return to case evidence which indicates that severe poverty is very common among violent offenders and there is consistent evidence that violent, serious, and chronic offending is associated with poverty, whereas the association is less consistent for less serious forms of crime. Individual-level analyses are telling in this regard; some of them suggest that nonviolent offenders have higher SES than nonoffenders.
The body of evidence points in the same direction. However, the evidence also strongly suggests that poverty is associated with nonviolent offending, so it is clear that if we wish to distinguish violent offenders from less serious offenders, or from frequent nonviolent offenders, we will need to learn more about the relevant dimensions of poverty to make that prediction. Krivo and Peterson (1996) found dramatic differences when they used various operationalizations of poverty, suggesting that “extreme” poverty has a special relationship with violence, but that is the limit of our understanding from this group of studies.