Main | Research | Experience | CV

---

Works in Progress

Proactive and Reactive Infrastructure Investment

Maintaining infrastructure requires investment. Faced with uncertain quality degradation, managers choose to invest proactively to prevent failure or reactively to address problems. Using a new dataset on drinking water systems, I estimate a dynamic discrete choice model of infrastructure investment. Simulations indicate that tightening regulations without financial support increases failures and raises costs. Reactive projects enable timely intervention as systems approach noncompliance, reducing disparities stemming from income and size. Efficiently restoring compliance for all systems requires expanded proactive investment to maintain quality and an even greater increase in reactive funds to address unexpected failures.

The Commission Paradox: Evidence from Prison Telecommunications with Nathan Miller and Marleen Marra

When an intermediary procures services for end users and receives a commission from the provider, a tension arises: procurement designs that more effectively extract provider rents can simultaneously raise prices and reduce total surplus. We formalize this commission paradox in a first-score auction model and estimate the model using procurement data from the prison telecommunications industry. The paradox emerges empirically, as commissions reduce provider rents but raise call rates. We also find that when commissions enter the scoring rule, increased competition among providers raises payments to the correctional authority but does not lower rates. Banning commissions restores the link between competition and lower rates. Our estimates indicate that recent federal regulations banning commissions and capping rates would lower costs for incarcerated individuals while preserving provider profitability.

Technology and Market Power in the Cement Industry with Nathan Miller, Matthew Osborne, and Gloria Sheu

We examine the evolution of market power in the cement industry over more than four decades using a structural model of procurement. The model matches aggregated outcomes in the data and implies transportation costs, shipping distances, and demand elasticities that are consistent with external sources. We find significant increases in local market concentration, but markups increase only modestly, and real prices do not rise. We attribute these patterns to a technological innovation—the precalciner kiln—that lowered variable costs, increased plants’ capacities and economies of scale, and contributed to an industry shakeout in which many plants closed.

The United States Paid Plasma Industry with Peter Jaworski and Kaitlyn Wilson

The Dynamics of Emissions Pricing and Technology Adoption with Sarah Armitage, Nathan Miller, and Matthew Osborne

Published Work

New Frontiers in Research on Industrial Decarbonization with K. Gillingham, L. Barrage, S. Armitage, D. Burtraw, J. Colmer, L. de Preux, J. Hawkins-Pierot, C. Holt, V. J. Karplus, A. Lofgren, R. Martin, N. Miller, M. Muuls, M. Osborne, E. Severnini, W. Shobe, V. Smirnyagin, T. Stoerk, A. Tsyvinski, K. Wagner, U. J. Wagner, and X. Wu. 2025. Science, 390(6771): 338-340.