The Effects of Poverty on the Susceptibility to Crime in South Africa

D. Mark Anderson, University of Washington

This paper examines empirically the effects of household-level poverty, measured by household expenditures per capita, on the susceptibility to crime in South Africa. The paper uses an instrumental variables strategy combined with community fixed effects to account for potentially endogenous expenditures and unobserved between-community heterogeneity. Across all model specifications the probability a South African household is robbed increases with expenditures. When using instrumental variables, the positive effect of expenditures on the susceptibility to robbery increases substantially. In addition, the effect of expenditures remains positive and significant for “nonwhite” areas. This suggests that robberies are a problem not only for the rich who live in gated communities and hire private security, but also for the relatively wealthy who reside in poorer neighborhoods. Finally, this paper fails to find a statistically significant relationship between expenditures and the susceptibility to violent crimes such as murder, rape and assault.

  See paper

Presented in Session 46: Crime and the Labor Market