Victor David Bernal Pavas

University of Houston · Economics · 2026–27 job market

Fields

Urban & Regional, Public, Environmental & Energy, Applied Micro

Job market paper

Less Traffic, Higher Tolls: Optimal Congestion Pricing under Remote Work and Traveler Heterogeneity

This paper studies how a persistent shift toward remote work changes optimal freeway congestion prices. I combine hourly station traffic from the California Performance Measurement System with tract-level commuting patterns, work-from-home shares, industry composition, population, and wages for five large California counties. Station traffic is allocated to residential tracts using fixed pre-pandemic commuting-route exposure, producing an estimated tract-level freeway VMT measure that aggregates exactly to observed PeMS VMT. A shift–share design based on predetermined local industry shares and national leave-California-out changes in industry remote work implies that a 10 percentage point increase in working from home reduces tract freeway VMT by 1.43 percent. A capacity-normalized BPR technology estimated by Poisson pseudo-maximum likelihood yields a congestion curvature of 3.27. These inputs produce three main pricing findings. First, holding the value of time fixed, the demand shift lowers average optimal tolls by 2.4–4.7 percent. Second, allowing the value of time to move with local wages reverses the result: optimal tolls rise by 8.0–11.4 percent. Third, income heterogeneity changes the meaning of distributional comparisons. Lower-income tracts supply 77–84 percent of modeled traffic and bear most of the marginal external damages in the common Pigouvian calculation. Under a constrained policy that imposes the same time-equivalent charge on both groups, however, the monetary rate paid by higher-income travelers is 1.6–2.0 times larger because their value of time is higher. Distributional conclusions therefore depend on whether the object of interest is a group's role in and exposure to external delay or the monetary rate it pays.