Thursday, 15 March 2018: 9:50 AM
The paper deals with the social and economic factors of crimes in the Czech Republic. In order to quantify the significance of independent variables that theoretically influence criminality, we estimated a model of panel data of the Czech Republic. Socioeconomic data related to crimes, police data and criminality come from the Czech Statistical Bureau, European Commission and Czech police statistics. We use annual data of eight Nomenclature of Territorial Units for Statistics (NUTS) 2 regions of the Czech Republic for the 11 year period 2005 – 2015. The dependent variable of crime rate is modelled through one-equation model where social and economic variables are independent variables. The list of the potential determinants of crime is created according to papers by Phillips, J., Land, K. C. (2012) and Ha, O. K., Andresen, M. A. (2017). Further, the paper also verifies the main idea of the Cantor-Land model methodology (originally Cantor, D., Land, K. C., 1985). According to the model, crimes are affected by unemployment through two (contradictory) effects: the (negative) opportunity effect and (positive) motivation effect. Panel data are tested through Levin-Lin-Chu unit root test. Further, relationship between criminality and specific determinants of crimes are examined in a pair base, in order to identify character of dependency. As a following step, we estimate the model that include all the considered independent variables. The most appropriate model is chosen using calibration where the least significant variables are gradually eliminated. Finally, the results of the model are interpreted with an emphasis on identifying the most important crime variables.