PRINCIPAL COMPONENTS ANALISIS FOR MODELING RISKS OF CRIMINAL CRIME

Abstract

In this work, an attempt was made to identify the main factors that determine significant individual characteristics of prisoners and increase the risk of committing repeated criminal offenses. The empirical analysis was conducted on the basis of a unique set of official criminal records of 13,010 convicts serving sentences in penitentiary institutions in Ukraine. It was established that the main factors determining the generalized profile of convicts are the risk factor of committing repeated criminal offenses and the age factor consolidating the age characteristics of prisoners (real age, age at the time of the first conviction before the real and/or suspended sentence). It has been proven that with an increase in the number of convictions to the actual degree of punishment, the risk of committing repeated criminal offenses increases. A direct relationship between the number of early releases and the propensity for criminal recidivism was revealed. It was determined that the earlier a person was sentenced to real and/or suspended punishment, the greater the risk of criminal recidivism; with increasing age, this risk decreases. A linear multivariate regression model was constructed to assess the influence of the factors included in the factor model on the risk of repeated criminal offenses by convicts. It was found that the number of suspended sentences and early releases exerted the greatest influence. The obtained estimates confirm the results of the conducted factor analysis. Reducing the term of punishment and replacing the real term with a suspended one encourages convicts to commit new crimes. The results of the conducted analysis can provide important knowledge to law enforcement agencies and justice authorities regarding the optimization of crime prevention and combating strategies.