Modeling Unemployment Duration in Slovenia using Cox Regression Models

摘要

This paper applies the semi-parametric Cox regression approach to model unemployment duration in Slovenia. The empirical analysis is based on a comprehensive dataset that consists of all unemployment spells in Slovenia from 2002 to 2005. The impact of the variables age, gender, level of education, and region on the hazard ratio is discussed. It takes longer for female and older unemployed persons to find a job and on average the duration of unemployment decreases with increasing level of education. Surprisingly, unemployed persons with a professional college degree or a bachelor's degree are better off than unemployed persons with a master's degree. A comparison of the Cox proportional hazards model and the Cox regression model with a time-dependent covariate reveals that the more appropriate model with a time-dependent covariate places greater emphasis on higher levels of education.