Gabriel Lewis

Welcome! I am currently seeking post-graduate jobs. This is my latest C.V., and these are my current affiliations:

  • Ph.D. candidate in Economics at University of Massachusetts Amherst (expected graduation May 2024).

  • Data Research and Statistics Counselor at Smith College.

My Research

My work focuses on labor economics, econometrics (especially robust Bayesian methods), and labor history, although my research interests extend beyond these fields. I am particularly interested in worker voice: the influence that workers may exercise in a firm’s decision-making, such as through a union; I also study how worker voice may interact with technological changes such as the rise of remote work and digital surveillance by employers.

I have been a research assistant for Ina Ganguli and Richard Freeman studying technology, the future of work, and the effects of COVID-19.

Prior to graduate school, I worked at the Stanford METRICS institute, using Bayesian methods to study the reproducibility of medical research.

My Latest Research

For my job market paper, I designed and conducted a survey experiment involving workers who had attended an unemployment agency, eliciting workers’ willingness-to-pay for remote work, (non-)surveillance by their employer, and worker voice: the influence that workers may exercise in their firm’s decision-making.

Using Bayesian and non-Bayesian models, I find that workers in my sample strongly value unions (at about 3% of wages on average), but also strongly value employer-controlled institutions such as employee resource groups (5%), in which workers and managers collaborate to solve workplace issues. However, employer-controlled institutions don’t act as utility substitutes for unions: given an ECI that they value, workers in my sample still want a union.

Workers in my sample value remote work (13% of wages), and strongly dislike digital surveillance by the employer, which is equivalent to a 16% pay cut. Once again, however, these don’t substitute for unions: give workers remote work and non-surveillance, and they still want a union.

The above averages are only part of the story: I estimate substantial variation in the population using my Bayesian model. For example, women and and people of color tend to show substantially greater WTP for most worker voice institutions than men and whites (although all positively value unions); this fits (but does not prove) the theory that workers from discriminated-against groups in particular tend to value worker voice as a mechanism to reduce discrimination. But even within rather narrowly-defined demographic groups, the variance is high: people disagree quite a lot, particularly about the value of unions, surveillance, and remote work.

Personal Details

My hobbies include hiking and guitar, both classical and flamenco (the guitar that is; I hike in various genres).