Common Trends in the US State-Level Crime. What Do Panel Data Say?

Abstract

This paper aims to investigate the long-run relationship between crime, inequality, unemployment and deterrence using state-level data for the US over the period . The novelty of the paper is to use non-stationary panels with factor structures. The results show that: i) a simple crime model well fits the long run relationship; ii) income inequality and unemployment have a positive impact on crime, whereas deterrence displays a negative sign; iii) the effect of income inequality on crime is large in magnitude; iv) property crime is generally highly sensitive to deterrence measures based upon police activities. KeywordsCrime, deterrence, inequality, unemployment, panel cointegration, cross-section dependence This paper aims to investigate the long-run relationship between crime, inequality, unemployment and deterrence using state-level data for the US over the period . The novelty of the paper is to use non-stationary panels with factor structures. The results show that: i) a simple crime model well fits the long run relationship; ii) income inequality and unemployment have a positive impact on crime, whereas deterrence displays a negative sign; iii) the e↵ect of income inequality on crime is large in magnitude; iv) property crime is generally highly sensitive to deterrence measures based upon police activities. JEL Codes C33, E20, K40