Abstract
I study how changes in the skill composition of workers affect the college wage gap in the US. In a range of wage determination models I show that skill may be measured as log wages subtracting any factors that shift wages for all workers within a firm or over time, such as differences in technology or labor demand. Using LEHD data I find that the average skill of the college workforce declined by 16% between 2002 and 2019, consistent with lower skill for marginal college graduates over time. Average skill for non-college workers changed little. I estimate that growth in the college wage gap would have been four times greater absent changes in skill composition. Through the lens of a supply and demand model, the college wage gap continues to rise despite narrowing skill differences because college workers sort to firms with much faster relative growth in labor demand.
Job Market Paper
Working Papers
Bank Expansion, Firm Dynamics, and Structural Transformation: Evidence from India's Policy
Experiment with Dian Jiao and Marshall Mo
Last updated in October 2024
Social Preferences in the Field: Evidence from a Nationwide Tipping Study with Uri Gneezy,
John A. List, and Ian Muir
Last updated in October 2024
When and Why Defaults Fail to Change Aggregate Outcomes: Field Experimental Evidence
from 40 Million Tipping Observations with Uri Gneezy, John A. List, and Ian Muir
Last updated in October 2024
Service Quality on Online Platforms: Empirical Evidence about Driving Quality at Uber with
Susan Athey and Juan Camilo Castillo.
Forthcoming at Management Science.
Last updated in September 2024
Design and Analysis of Cluster-Randomized Field Experiments in Panel Data Settings with Ali
Hortacsu, John A. List, Ian Muir, and Jeffrey Wooldridge.
Reject and Resubmit at Journal of Econometrics.
Last updated in July 2019
Work in Progress
Measuring Firm-Level Skill: Effects on Performance and Productivity
Abstract
I derive a method for measuring worker skill at the firm level. I use it to study the effect of skill on firm
performance and productivity using the LBD and LEHD from the US Census. I document three key empirical findings.
First, average worker skill strongly predicts firm survival and employment growth. Second, firms that pay higher
wages per unit of skill employ more skilled workers on average, but this correlation nearly disappears when
conditioning on industry. Third, I show that accounting for skill leads to meaningfully different conclusions about
firm TFP. Manufacturing firms in high education counties have experienced faster TFP growth, a fact masked by
conventional TFP measures that adjust for skill using hours or the wage bill.
Coercive Sterilization in India with Ruchi Mahadeshwar
Resting Papers
Publications
Order My Book
Talks and Workshops
Contact