Have you ever seen a keynote drinking a liter of beer during their speech? Well, this is what welcomed us in Brno, during the Young Economists' Meeting (2018). The conference, organized by the Masaryk University, gathers experimental and labor economists in a relaxed environment to discuss their ongoing research. We went there to present our research on wage dispersion within occupations.

Our analysis builds on the models developed in the literature by testing for one (often implicit) assumption: that occupations less susceptible to automation present higher dispersion of wages. This relation can be explained both by the dynamic aspects of the theory (reallocation of workers towards these occupations creates worse matches over time), but also due to static differences in the type of tasks performed by workers. We argue that tasks not susceptible to automation are more sensitive to individual productivity.

By and large, our research confirms the assumptions from the theory. A one standard deviation increase in the share of tasks susceptible to automation is related to a 10% higher dispersion (7% if we account for differences in workers and firm characteristics). More importantly, the relation is robust to the inclusion of alternative mechanisms, such as the presence of winner-takes-all markets and changes in employment.

Of course, my gratitude towards participants for their insightful suggestions and comments, and to the organizers for accepting my research.