EMPIRICAL EVIDENCE ON THE DETERMINANTS OF CRIME IN DEVELOPING COUNTRIES

Abstract

This study empirically tested Becker’s hypothesis that by increasing the opportunity cost of crime, criminals could be retrained from doing criminal activities. This hypothesis is empirically tested by employing panel data analysis for 26 developing countries for the time period 2003 2020. The independent variables of this study have been categorized into four categories: economic, demographic, governance and deterrent variables. The economic variables include GDP per capita, inflation, unemployment and expenditure on education; demographic variable include the urbanization rate; governance indicators include the rule of law, control of corruption and political stability and the deterrent variable include the total number of convicted persons. First of all, the standard panel data estimations techniques of fixed effects and random effects were applied. The choice between these two was made by Hausman specifications test, and the fixed effects model was found to be appropriate. As this study used a short panel where N (26) is greater than T (18), applying Generalized Methods of Moments (GMM) estimations technique was appropriate which tackles the possible presence of Autocorrelation, Heteroscedasticity and endogeniety 1and is also used to check the consistency of results. The results of both techniques showed that the inflation rate, urbanization rate, and expenditure on education are positively associated with crime, while governance and deterrent variables are negatively associated with crime, while unemployment and GDP per capita have asymmetric results from the two methods.