Top income shares and crime

Abstract

This article estimates the relationship between income inequality approximated by the top 10% and 1% income shares derived from income tax return data and nine crime categories for the US states between 1979 and 2003. We fail to find a positive relationship between inequality and crime, except for the case of the top 10% share and motor vehicle theft.

Keywords:

• inequality
• top income shares
• crime
• panel data

JEL Classification:

• C23
• D63
• K42

Notes

¹ The use of top income shares calculated from tax return data has been increasingly popular in recent distributional analyses, especially in those requiring data covering longer time periods (see, Atkinson and Piketty, 2007, 2010; Atkinson et al., 2011). Leigh (2007) has shown that there is a strong positive relationship between the top income shares and other conventional inequality measures (e.g. including the Gini index) in a panel of rich countries.

² Accounting for the inherently dynamic nature of crime, we include 1-year lagged crime in all FE specifications. In the GMM modelling, we use a two-step estimator with the Windmeijer correction.

³ We have also estimated all models with all independent variables lagged by 1 year to account for the fact that the impact of inequality on crime may be delayed in time. Similar to the presented models, the coefficients on inequality variables in models with lagged independent variables were either negative or insignificant. Results are not reported for the sake of brevity, but are available on request.