Abstract |
How does South Africa’s extremely concentrated income inequality affect the incidence of property crime? Studies based on developed countries with much lower inequality levels show that property crime increases monotonically with inequality; but this is not the case for South Africa. We use 2011 South African census data and property crime data locally disaggregated to the police precinct level. The best fitting model is a flexible one including non-linear inequality and income effects as well as an interaction between these two variables.We link this result to extreme inequality signalling both that local elites should invest in security but also relative credit-constraints for potential criminals. Our results are robust to seven different inequality measures, but the precise form of these results varies based on how sensitive measures are to the top or bottom of the income distribution. We conclude that the usual monotonic relationship between property crime and inequality is not robust in high-inequality contexts like South Africa and that measurement of inequality matters in order to correctly specify this relationship. |