The Non-Linear Effect of Wealth on Crime

Although theory suggests the relationship between crime and wealth is ambiguous, most empirical analyses estimate a monotonic relationship and find that wealth has negative effect on crime. Using two proxies for wealth (median income and poverty rate) and two types of crime (property and violent), we find a quadratic relationship is the best fit for our four crime-wealth groups. In general, the expected negative effect of wealth on crime only applies to wealthier counties. In poorer counties, wealth has an unexpected positive effect on crime. This result may be theoretically consistent, or an unintended byproduct of the Uniform Crime Reports data, which do not include unreported crime.