Yuya Shimizu

University of Wisconsin-Madison, Department of Economics.

In my Job Market Paper, I develop theoretical foundations for incorporating numerical representations of images and text into econometric models and propose a new bootstrap test for the workflow.

Labor supply elasticity estimates with 90% and 95% confidence intervals for the original BERT representation, 25 random seeds, and median aggregation.
Robustness of estimated values to nonidentification of text representations
Parameter of interest: Labor supply elasticity, with controls for job description text.
Data: Amazon MTurk, an online labor market platform.

Recent Update

  • My Job Market Paper received the Best Student Paper Award at the 2026 Midwest Econometrics Group (MEG) Conference.