UNEMPLOYMENT DURATION IN THE CZECH REPUBLIC

The article deals with the problem of modelling individual unemployment duration in the Czech Republic using data from the CZSO's Labour Force Surveys (LFS). Standard approach to the interval censored data is used and the hypotheses about different survival distributions for the different groups are tested, thus identifying the more and less endangered strata of the society. The main aim is to use standard techniques and available software for interval censored data that is outcome from several LFS datasets. Results indicates, that there may be a connection between unemployment duration and region, number of persons in household, age, sex, education, marital status and handicap status, but the data were not adjusted to avoid problems with relationship between explanatory variables. On the other hand, the data does not prove connection between unemployment duration and cooperation with employment bureau, previous job experience, status at last job (employee or employer) or activity before seeking a job.