Since the literature on the effect of the unemployment rate as reflection of economic fluctuations on crime shows an empirically ambiguous effect, the purpose of this paper is to argue that a new way of modeling the dynamics of unemployment and crime by focussing on the transitory and persistent effect of unemployment on crime helps resolve this ambiguity.
Panel data for US states from 1965 to 2006 are examined using the Mundlak (1978) methodology to incorporate the dynamic interactions between crime and unemployment into the estimation.
After decomposing the unemployment effect on crime into a transitory and persistent effect, evidence of a strong positive correlation between unemployment and almost all types of crime rates is unearthed. This evidence is robust to endogeneity and the controlling for cross-panel correlation and indicators for state imprisonment.
The paper is the first to examine the dynamics of the interaction of crime and economic fluctuations using the temporary and persistent effects framework of Mundlak (1978). In one set of estimates, one can evaluation both the short- and long-run effects of changes of unemployment on crime.
