The joint annual congress of the European Economic Association and Econometric Society European Meeting took place in Lisbon this year. We were super excited to place our paper in the main program. The paper proposes a new way to estimate gender wage gaps, which allows to account for the fact that the labor markets may be segmented. If access to "better jobs" is somehow gendered, the standard estimators of the gender wage gap mix altogether the actual wage gap, the gender bias in the access to the "better jobs" and the difference between wages in "better jobs" and other jobs. Our method allows to capture that, which is of interest to policy-making: actual gender wage gap, accounting for individual characteristics. Analyzing data from a variety of European countries, we find that indeed, mixing up everything makes us wrongly informed about the actual prevalence of gender wage inequality.