"Leaping into the future of labor economics: the research potential of linking employer and employee data" is the title of a paper by Daniel S Hammermesh published in Labour Economics in 1999. I quote it here, since it captures much of my motivation for the work included in this thesis. Considering applied micro econometrics and labor economics my main elds of interest, the development of linked employer-employee data that took place in Denmark around the time of the new millennium, marked new and exciting possibilities. For some years Danish researchers have had access to very detailed information on all people living in Denmark, but at the beginning of this century also data on all companies linked to these persons was being made available for research. Combined with modern computer technology this meant access to a linked database following all employers and all employees in Denmark over time. I had no doubt that this should be the centerpiece of my Ph.D. The result has been two lines of research, one studying the e¤ect of globalization on labor demand in Denmark, and one studying sorting, that is, how and why employers meet employees in the labor market. In the summary I treat each line of research independently although I would like to emphasize, that studying a labor market where fi rms and workers reacts to one another is the corner stone in both.