Does gender segregation at the labor market exist in Poland?

We would like to answer the question if gender pay gap and segregation at the labor market exist in Poland, applying decision trees. Therefore for the grouping variable i.e. net income from the main place of employment we construct clusters of respondents that are created due to such features as: gender (dichotomous variable), level of education (- 9 variants of variable), employment contract (- 6 variants of variable), economic activities NACE (- 23 variants of variable), occupational groups (- 10 variants of variable), working time (full-time or part-time), additional job, age (- 5 variants of variable), size of enterprise measured by the number of employees (- 6 variants of variable) and job seniority in years(- 7 variants of variable).

Investigation is provided applying data from the Polish Labor Force Survey in the years 2001, 2003, 2006, 2009 and 2011. Employing data from selected years is due to assumption that the structure of the labor market change essentially in the longer period than one year (thus we do not compare situation at the Polish labor market year by year). In our research we use classification trees that are a powerful alternative to the more traditional statistical models. This model has the advantage of being able to detect non-linear relationships and showing a good performance in presence of qualitative information. We apply classification tree QUEST with pruning.

The results of our experiments show that the most important factors that influence clustering are: level of education, size of company, gender and job seniority. It is also proved that in Poland gender pay gap and segregation at the labor market are observed.