Poster Paper:
Occupational Licensing, Skills, & Labor Market Spillovers
*Names in bold indicate Presenter
Using data from O*NET on each occupation's skill, ability, and broad educational requirements, I cluster occupations together using various unsupervised machine learning techniques to select occupations that act as the counterfactual to a licensed occupation. In order to reduce dimensionality in the set of skills and abilities over which to cluster, I use random forests to identify the most "important" wage-determining characteristics of each occupation, which I then factor into my cluster analysis.
I also supplement this clustering technique by grouping occupations together using occupation to occupation transitions from the Survey of Income and Program Participation. After adjusting for time-varying equilibrium trends in skill prices, I then employ difference-in-differences and interrupted time series designs paired with data from the CPS Outgoing Rotation Group to examine the labor market effects of introducing a new occupational licensing requirement on within-occupation wages and employment; overall wage dispersion; within-cluster wages and employment; and overall changes in inequality across skill groups.