Presented By: Department of Economics
Michael Beauregard Seminar in Macroeconomics: What is a Labor Market? Classifying Workers and Jobs Using Network Theory
Jamie Fogel, PhD U-M
Abstract:
This paper develops a new approach to classifying heterogeneous workers and jobs and demonstrates that traditional methods may understate the effects of labor market shocks on workers. Our key innovation is a new method for identifying high degrees of latent worker and job heterogeneity directly from data, without relying on covariates like education or occupation. Building upon tools from network theory, we classify workers and jobs into a large number of latent “worker types” and “markets,” re-spectively, by exploiting the network structure of worker–job links in linked employer-employee data. Intuitively, two workers belong to the same worker type if they have similar probabilities of being employed by particular jobs, and two jobs belong to the same market if they have similar probabilities of hiring particular workers. We use discrete choice methods to infer the productivity of each worker type when matched with each market using the logic that worker–job matches that pay more and occur more frequently in equilibrium reveal themselves to be more productive. We embed this method within a general equilibrium model with sectoral demand shocks to perform counterfactuals. We show that our worker types and markets are better able to predict wage changes in response to the 2016 Olympics than are occupations. Finally, we show that traditional ways of defining worker heterogeneity and exposure to shocks may underestimate the effect of shocks on workers by as much as a factor of 4.
* To join the seminar, please contact at econ.events@umich.edu
This paper develops a new approach to classifying heterogeneous workers and jobs and demonstrates that traditional methods may understate the effects of labor market shocks on workers. Our key innovation is a new method for identifying high degrees of latent worker and job heterogeneity directly from data, without relying on covariates like education or occupation. Building upon tools from network theory, we classify workers and jobs into a large number of latent “worker types” and “markets,” re-spectively, by exploiting the network structure of worker–job links in linked employer-employee data. Intuitively, two workers belong to the same worker type if they have similar probabilities of being employed by particular jobs, and two jobs belong to the same market if they have similar probabilities of hiring particular workers. We use discrete choice methods to infer the productivity of each worker type when matched with each market using the logic that worker–job matches that pay more and occur more frequently in equilibrium reveal themselves to be more productive. We embed this method within a general equilibrium model with sectoral demand shocks to perform counterfactuals. We show that our worker types and markets are better able to predict wage changes in response to the 2016 Olympics than are occupations. Finally, we show that traditional ways of defining worker heterogeneity and exposure to shocks may underestimate the effect of shocks on workers by as much as a factor of 4.
* To join the seminar, please contact at econ.events@umich.edu
Co-Sponsored By
Livestream Information
LivestreamSeptember 15, 2021 (Wednesday) 4:00pm
Joining Information Not Yet Available
Explore Similar Events
-
Loading Similar Events...