Application of correspondence analysis to the identification of the influence of features of unemployed persons on the unemployment duration

Abstract

The paper presents the analysis of the influence of selected features of unemployed persons on unemployment duration. Theoretical considerations go along with an empirical study based on individual data from Local Labour Office in Szczecin. The research hypothesis states that features of unemployed persons such as the level of education, sex, work seniority or age have a meaningful influence on unemployment duration. Recent studies usually used event history analysis, inflow-outflow analysis of the labour market or logit models. This research is unique because of the application of one of the methods of multidimensional analysis – correspondence analysis. This method allows analysis of multidimensional relationships between categories of nominal variables. From the results obtained it could be stated that strong relationships between unemployment duration and features of unemployed persons exist (apart from sex).