EN
The aim of this paper is to analyze the duration of unemployment periods of people using the survival analysis. The hazard model can be a suitable tool for the analysis of unemployment period duration. Estimating the multiepisode risk model with time-varying covariates we calculate the direct risk of leaving the unemployment in the next instant. Hazard models usually comprise not only present duration of the phenomenon as a significant determinant for the probability of its occurrence, but also other observable characteristics of individuals such as gender, age, marital status, place of residence, education level, level of disability. Duration models with time-varying covariates are proper tools for the analysis of the influence of the unemployment benefit received at the risk of leaving unemployment. Additionally, we assess the impact of vocational training on unemployment duration as well. The following hypotheses are put forward by us to be verified using hazard models: old age gives smaller re-employment possibilities; women are less likely to be employed than men; higher education gives greater employment possibilities; there is a greater tendency to leave the unemployment state if the registered person receives no unemployment benefit; vocational training raises the transition rate from unemployment; there is a smaller probability of exiting unemployment state for people entitled to social welfare. To estimate, we use data from the District Labor Office in Słupsk in Poland from January 1990 to July 2007.