Thursday, 15 March 2018: 10:30 AM
Eurostat's statistics show that in the European Union (28 states, EU28) women earned 16.3% less than men in 2014 based on average hourly earnings. Many researchers have noted that the gender pay gap (GPG) is influenced by many factors and could change in relation to employee group. A large number of studies point out wage differences between men and women with wages depend on the level of remuneration. Such analyses indicated that the GPG in many countries is increasing throughout the conditional wage distribution and accelerating at the top (Arulampalam, Booth and Bryan 2007). In addition, in the top earner group, men are overrepresented. This led to a question regarding the characteristics specific to this group of employees. The main aim of this analysis is to cluster individuals with the highest hourly wages in the 28 EU countries. The study is based on the Eurostat’s Structure of Earnings Survey (SES) individual data for 2014. The SES survey refers to the enterprises with at least 10 employees. Information included in the SES databases is from the enterprises' registers. In the analysis we applied classification and regression tree (CART) analyses (Breiman, Friedman, Olshen, and Stone 1984). Classification trees is a useful method to explore the structure of a set of data. The results of a CART analysis are displayed as a tree diagram when the internal nodes are divided into two succeeding nodes. Nodes are divided until some stopping criterion is reached. We will explore the structure of the top earner group in selected countries. Preliminary results indicate two obvious factors differentiating the level of wages: the profession of employees and economic activity of employers. We will compare the structure of the analyzed groups in relation to these factors. We will test the hypothesis that the structure of top earners is very similar in European countries, and dissimilar compared to all employees.