Heterogeneous Effect of Unemployment on Crime in Iran: Hierarchical Panel Bayesian- Poisson Approach

Many empirical studies have analyzed the relationship between economic and social behavior. Unemployment is considered as a phenomenon which is not desirable. The effect of unemployment on crime, may have heterogeneous impact on distinct regions (province or country). The technical approach in econometric literature indicates that, due to high integration of economic and social activities, the independency assumption-in this study, provinces- in the sample cannot be accepted, especially if the sample is small. In order to show the difference of crime effect on unemployment rate in various provinces, the random coefficients Panel Bayesian Conjugate Poisson model (Hierarchical) is used. To do this, the data of five provinces of Iran are used during the period of 2009-2014 to demonstrate such heterogeneous effects. Results show that for an increase in unemployment, there is a positive and varying effects on crime in the provinces for which the data were collected. In other words, in this study, for a %1 increase in unemployment, the greatest impact is in Tehran and the lowest is in Semnan province.