Compensating Differentials of Work-from-Home (Job Market Paper) link

This paper estimates the wage effect and labor supply response of work-from-home. There was a large increase in work-from-home due to the Covid-19 pandemic that has persisted in the aftermath. I leverage this within-person time variation in work-from-home incidence to estimate the compensating wage differential of working from home by utilizing the panel structure and work-from-home questions added to the 2019-2023 iterations of the Survey of Income and Program Participation (SIPP). I find that increasing remote work leads to slightly lower wages, even though workers in these arrangements typically earn a higher wage than their in-person counterparts. The bulk of the effect comes from those that spend all of their work days working from home. In addition, remote work leads individuals to work more hours overall. This is true in both hybrid and primarily work-from-home arrangements. I explore these findings with a labor supply model that includes differentiated productivity and preferences for work-from-home.

Job Search at Unemployment Benefit Exhaustion: Evidence from Google Trends link

This paper revisits the question of how unemployment benefits impact job search effort by looking to the largest unemployment benefit expansion in US history in response to the Covid-19 pandemic. I estimate the effect of states ending unemployment benefits early on internet job search traffic. Employers were met with a tight labor market as the world opened back up with many pointing to the expanded unemployment benefits as a possible cause. This led to states deciding to opt out of the benefits at staggered times in the summer of 2021. I create a weekly state-level job search index derived from Google Trends and use difference-in-differences to leverage the variation in the dates of expanded unemployment benefit exhaustion across states. When states withdrew from the federal unemployment benefit programs early, search actually declined (as much as 10%) in the weeks following relative to states that did not. I find a small jump (6-7%) in job search at the time of state announcement that quickly dissipates. While we may expect job search to rise as unemployment benefits fall, these findings imply that expanded unemployment benefits were not the primary driver keeping people out of the labor force.

Forecasting in Real-Time with Google Trends (with Daniil Manaenkov) link

This paper explores the usefulness of internet search data from Google Trends in real-time forecasting work. We adapt previous work to an environment in which data is revised multiple times before being finalized, in a comparable manner that many macroeconomic series are. We find no performance gains in forecasting national employment and varied performance effects in forecasting state employment, in contrast to other findings using the full history of data.

The U.S. Economic Outlook for 2023(4)-2024(5,6)” with Jacob T. Burton, Gabriel M. Ehrlich, Owen Kay, Daniil Manaenkov, Tereza Ranosova, Niaoniao You, and Yinuo Zhang (Feb. 2023, May 2023, Aug. 2023, Nov. 2023, Feb. 2024, May 2024, Aug. 2024, Nov. 2024, Feb. 2025) 

Quarterly U.S. Forecast for over three hundred economic variables with commentary describing the current state of the economy and an explanation of our proposed path. Our work is included in the Federal Reserve Bank of Philadelphia’s Survey of Professional Forecasters.

"The Economic and Demographic Outlook for Michigan Through 2050" with Jacob T. Burton, Gabriel M. Ehrlich, Donald R. Grimes, Daniil Manaenkov, and Michael R. McWilliams (July 2022) 

Long-run forecast of Michigan’s fertility, migration, and labor force trends provided to the Michigan Department of Transportation.