Is the difference in wages between men and women (the gender wage gap) due to less labor market experience of women compared to men, or is it due to discrimination against women, for example, because labor market experience of women is valued less than labor market experience of men? How much of the gender wage gap can be "explained" by differences in endowments such as education, skill, or experience? How much do changes in educational attainment and general trends in earnings inequality contribute to the change in the wage gap over time? How would test scores of pupils with and without migration background compare if there would be no differences in average socio-economic status? How much did de-unionization and the decline in real minimum wages contribute to rising wage inequality? How high would the mortality rate in country A be if it had the demographic composition of country B? Decomposition methods can help finding answers to such and other questions by providing insights into the mechanics of group differentials (such as earnings differences between men and women). Based on methodological developments mostly in labor economics (and some parallel developments in demography), these methods are increasingly popular in various fields of the social sciences. The seminar introduces the statistical concepts of decomposition methods, provides an overview of various approaches, and makes students familiar with the application of the methods and the interpretation of their results. Theoretical inputs and practical exercises (using Stata) will be alternated throughout the course.