The opportunities in the labor market and formation of human capital (education) affect the implicit costs of engaging in illegal activities. In this article we empirically investigate the effect that the unemployment rate, which real income in the labor market and which school atandence rate for young people (15 to 24 years) exert on the prevalence of homicides in Brazilian municipalities. We also analyze the heterogeneities related to the specialization of work and the life cycle in youth. In order to overcome the problems of endogeneity, in our identification strategy we used a instrumental variable estimation of panel data model for all Brazilian municipalities (between 1980 and 2010), closely following the approach developed by Gould et al. (2002). The instruments its relationship to long-term structural changes in industry that affect the demand for labor in cities. We found the following results over homicide rate: i) positive effect of men's unemployment rate, which is greater for young adults (25-29 years); ii) no effect of real hourly labor wage ; and iii) the negative effect of the school attendance rate of 15 to 17 years.
labor markets crime homicide unemployment rate school attendance rate