Research

Publications:

Experience Rating as an Automatic Stabilizer (with Mark Duggan and Audrey Guo) Forthcoming at Tax Policy and the Economy [SSRN link] [NBER link] 

Unemployment insurance taxes are experience-rated to penalize firms that dismiss workers. We examine whether experience rating acts as an automatic stabilizer in the labor market. We exploit the fact that penalties for layoffs vary by state using detailed data on state tax schedules, and we measure whether firms react less to labor-demand shocks in the presence of greater layoff penalties. The average penalty for layoffs reduces firm adjustment to negative shocks by 11 percent. The results imply experience rating has a stabilizing influence on labor markets. Experience rating saved, for instance, nearly a million jobs in the Great Recession. 

Broadening the Unemployment Insurance Tax Base (with Mark Duggan and Audrey Guo) forthcoming at American Economic Association: Papers & Proceedings, (2022) [SSRN link] [IZA link] 

The tax base for state unemployment insurance (UI) programs varies significantly in the U.S., from a low of $7,000 annually in California to a high of $52,700 in Washington. Previous research has provided surprisingly little guidance to policy makers regarding the tradeoffs associated with this variation. In this paper, we use 37 years of data for all 50 states and Washington, D.C. to estimate the impact of the UI tax base on labor-market outcomes. We find that the low tax base that exists in California and many other states (and the necessarily higher tax rates that accompany these) negatively affects labor market outcomes for part-time and other low-earning workers. 

Unemployment-Insurance Taxes and Labor Demand: Quasi-Experimental Evidence from Administrative Data, American Economic Journal: Economic Policy, (2021) [journal link] [SSRN link] [IZA link (ungated)] 

To finance unemployment insurance benefits, states raise payroll taxes on employers who engage in layoffs. Since tax rates increase in response to layoffs, taxes are highest for troubled firms after downturns, potentially hampering labor demand and employment during recoveries. Using full-population administrative records from Florida, I estimate the causal effect of these targeted tax increases on firm behavior leveraging a regression kink design in the tax schedule. UI tax hikes reduce firm hiring and employment substantially, with no effect on layoffs or worker earnings. Analysis of heterogeneity and timing suggests the role of cash constraints in explaining the magnitude of the estimates. The results imply unanticipated costs of the financing regime which, once accounted for, reduce the optimal benefit calculation by a quarter.

Is Compassion a Good Career Move?: Evidence on Nonprofit Earning Differentials from Employer Changes (with Carla Johnston), Journal of Human Resources, (2021) [journal link] [SSRN link] [IZA link (ungated)]

We explore the nonprofit earnings penalty. To separate the influence of demand and supply, we leverage workers who change employers in administrative tax data. The average nonprofit worker earns 5.5 percent less than the average for-profit worker. Supply-side factors (worker selection) contribute 80 percent of the nonprofit differential. The remaining 20 percent is from demand (a nonprofit penalty). Within-worker nonprofit variation generates several insights about the influence of nonprofits on the labor market. Nonprofits compress the wage distribution and reduce inequality among earners. Nonprofit penalties are much more pronounced in classic charities than in “commercial” nonprofits, which sometimes exhibit nonprofit premia. 

The Finance of Unemployment Compensation and its Consequence (with Audrey Guo), Public Finance Review, (2021)

For every payment, there is an equal and opposite tax. Economists have contributed important theoretical and empirical findings to the study of unemployment insurance benefits, but a deliberate study of UI taxation's unique structure remains undone. While presenting new results on the distributional impacts of UI taxes, this paper serves as a primer for researchers new to the field. We summarize available evidence on UI taxation, describe the history and institutions of experience rating, and outline important lines of inquiry for future work. As unemployment has risen, so has the need for a body of policy-relevant knowledge about the function and financing of UI systems. 

A Privacy-Oriented Deferred Multi-Match Recommender System for Stable Employment (with Amar Saini and Florin Rusu) Recommender Systems (2019)

Coordination failure reduces match quality in the job market. Centralized matching can improve match quality, but it requires all participants to inflexibly accept assigned matches. In this paper, we present a matching recommender system that generates stable pairings using rank preference data. We explore a series of adaptations of the deferred acceptance algorithm which combine the flexibility of decentralized markets with the intelligence of centralized matching. Experimental results using real and synthetic data confirm the benefits of the proposed algorithms across several quality measures. Over the past year, we have implemented a prototype and deployed it in an active job market. Using the gathered real-user preference data, we find that the match recommendations are superior to a typical decentralized job market.

Potential Unemployment Insurance Duration and Labor Supply: Evidence from a Benefit Cut (with Alexandre Mas), Journal of Political Economy, (2018) [journal link] [local link] [SSRN link] 

We examine how a 16-week cut in potential unemployment insurance (UI) duration in Missouri affected search behavior of UI recipients and the aggregate labor market. Using a regression discontinuity design (RDD), we estimate a marginal effect of maximum duration on UI and nonemployment spells of approximately 0.45 and 0.25 respectively. We use the RDD estimates to simulate the unemployment rate assuming no market-level externalities. The simulated response, which implies almost a one percentage point decline in the unemployment rate, closely approximates the estimated change in the unemployment rate following the benefit cut. This finding suggests that, even in a period of high unemployment, the labor market absorbed this influx of workers without crowding-out other jobseekers.

The Effect of Unemployment Benefits on the Duration of Unemployment Insurance Receipt: New Evidence from a Regression Kink Design in Missouri, 2003-2013 (with David Card, Pauline Leung, Alexandre Mas, and Zhuan Pei), American Economic Review: Papers & Proceedings, (2015) [journal link] [manuscript link] 

We provide new evidence on the elasticity of unemployment insurance weekly benefit amount on unemployment insurance spells based on administrative data from the state of Missouri covering 2003-2013.  Identification comes from a regression kink design that exploits the quasi-experimental variation around the kink in the UI benefit schedule.  We find that unemployment durations are more responsive to benefit levels during the recession and its aftermath, with an elasticity of about 0.9 as compared to 0.35 pre-recession.