In order to analyze this mechanism, we build an overlapping generation growth
model where the links among crime, growth and labor market imperfections are explicitly formalized and endogenously rendered although highly stylized to keep the model tractable. As we mentioned, the literature underlined both a growth and a level effect of crime. We therefore explore both theoretically and empirically this issue. In fact, the model is considered in two versions, an endogenous growth version and an exogenous one. According to both versions of the model, an economy presenting a highly imperfect labor market could end up in a lower equilibrium with high unemployment, high crime, low growth and/or low income. The empirical part tests the model implications through the Italian regional data according to the exogenous or endogenous version of the model using the homicides rate as measure of crime. In fact, given the high rate of unreported property crimes in many southern Italian regions, alternative measures of economic crimes result to be highly unreliable due to the well known underreporting and underrecording bias. From an empirical viewpoint, the distinction between short-run and long-run effects on growth is of primary importance in this context. The more so, because of the presence of unemployment and by the crime measure we use, the homicides rate, which is used for alternative measures of crime, especially economic crimes, result to be highly unreliable in many southern Italian regions, given the high rate of unreported property crimes. 3 Both variables are correlated with income growth in the business cycle. In order to avoid misleading conclusions, in this paper we employ the Pooled Mean Group estimator proposed by Pesaran et al. (1999) which allow to separate the longrun effects from the short-run ones by using an ECM approach. The paper is organized as follows. Section 2 formalizes the individual choice concerning the legal-illegal supply of time, and the inter-temporal consumption choice considering the technology of the economy and the labor demand by firms in an exogenous set-up and in an endogenous context as well. Section 3 analyzes some policy issues concerning the fight against crime. Section 4 tests the theory through the Italian regional data. Section 5 summarizes and concludes.