Essays on remote work and public economics

Abstract

This dissertation consists of three chapters that present an empirical and structural analysis of remote work through wage differentials and examine the implications of remote work for public policy from a public finance perspective. The first essay analyzes wage disparities between work-from-home (WFH) and onsite employees using data from the American Community Survey (ACS). I document a substantial and historically persistent wage gap favoring WFH workers. A decomposition of this differential reveals that over half can be attributed to human capital characteristics and occupation fixed effects, leaving an unexplained premium of 8–11% in log terms. I show that this premium persists across demographic groups such as education and age, and cannot be accounted for by a standard Roy model. To explore the underlying mechanisms, I develop a partial equilibrium model with discrete workplace choice that incorporates productivity losses associated with WFH. The model attributes part of the wage differential to selection into WFH, suggesting that the direction and strength of selection depend on the curvature of utility over consumption and the magnitude of productivity loss. The second essay studies optimal income taxation in the context of remote work. Workplace choice is modeled discretely, incorporating heterogeneity in preferences and productivity losses associated with WFH. Using a Ramsey taxation framework and a parametric income tax function, the analysis highlights how remote work affects key public finance margins—specifically, the elasticity of taxable income, the degree of consumption smoothing, and the level of income inequality. I compare optimal tax schedules with and without the option to work remotely. The results show that allowing for WFH systematically lowers the optimal marginal income tax rate—a finding that is quantitatively significant and robust across alternative calibrations. In a calibrated economy, optimal marginal tax rates are approximately 2%–3% lower for low-income groups and 4%–5% lower for high-income groups when WFH is incorporated. For an income increase of $10,000, the optimal policy with WFH reduces the tax burden by $280 to $520 across incomes ranging from $40,000 to $500,000, relative to a no-WFH benchmark. The third essay examines optimal income taxation when tax policy can condition on workplace type. Using a Ramsey framework, I compare a uniform tax schedule with a tagging policy that assigns separate marginal tax rates for remote and on-site workers. This flexibility allows the planner to internalize productivity losses and selection effects associated with WFH. The analysis shows that the planner imposes higher tax progressivity on remote workers, discouraging WFH entirely despite its amenity value. This workplace-contingent policy achieves higher welfare than a uniform schedule, particularly when heterogeneity in WFH preferences is modeled with a log-normal distribution. In that case, the difference in optimal progressivity between WFH and No-WFH workers reaches nearly 12 percentage points, compared to about 5 points under a uniform distribution. These findings highlight the efficiency gains from conditioning tax policy on observable workplace choice.