Research

"The Job-Preservation Effects of Paycheck Protection Program Loans" (R&R at The Journal of Finance, with Michael Faulkender and Stephen Miran)  [Draft, April 2023]

Abstract

The Paycheck Protection Program (PPP) supported over 60 million jobs through August 2020. How many of those jobs would have otherwise been lost? We estimate the number of jobs saved by leveraging the relationship between local banking markets and the average speed to loan approval. With county-level weekly unemployment insurance (UI) data, we estimate that a 10 percentage point increase in PPP payroll coverage at sub-100 employee businesses led to a 1.0 percentage points suppression of initial UI claims. That same increase suppressed the insured unemployment rate (IUR) by 2.4 percentage points.  In aggregate, we estimate that PPP loans saved 10.5 million jobs at sub-100 employee businesses and 13.8 million overall.

"Learning Speeds and Labor Market Readjustments" (Job Market Paper, with Elena Mattana) [Draft]

Abstract

As the economy changes, some workers confront a labor market that no longer demands their set of skills. While these workers can adapt by learning new skills, not all workers learn at the same rate. This paper asks how heterogeneous learning rates influence labor market outcomes and occupational choice, particularly under economic volatility. First, we document heterogeneous responses to job loss from establishment closures using Danish administrative data. For workers who earned high grades in secondary education, annual earnings rebound to those of their peers in the years following job loss. Workers with lower grades suffer a permanent earnings loss relative to their peers, showing no recovery whatsoever. Second, we study occupational choice in a dynamic Roy model where workers learn on the job at varying speeds. We estimate these learning speeds alongside the evolution of unobserved skills using a non-linear Kalman Filter, corrected to accurately estimate posterior covariances. High-type learners acquire new cognitive-based skills 65% more quickly than low-type learners, however that pattern is flipped for manual skills. High-type learners choose occupations they are initially unsuited to, especially early in their careers. This early mismatch delivers mid-career benefits in the form of higher skills and earnings. A higher learning rate also allows for quicker adjustment and recovery in the face of adverse shocks to occupational prices.