Gretchen Sileo

Logo

 gs907@georgetown.edu

 CV |  LinkedIn |  GitHub

Main | Research | Experience | JMP | CV


Works in Progress

Proactive and Reactive Infrastructure Investment (JMP)

Properly functioning infrastructure is maintained through investment. A proactive investment strategy prevents failures but requires expenditures before quality deteriorates. A reactive investment strategy accepts some risk of failure to avoid unnecessary expenditures. I explore proactive and reactive investments in a newly collected dataset on Kentucky water systems to assess the ability of system managers to maintain infrastructure quality. I establish that proactive and reactive investments differentially reduce the probability of a future system failure, and that both managers and consumers are sensitive to system quality. I construct and estimate a dynamic discrete choice model of system manager infrastructure investment decisions incorporating the empirical relationships and investment strategy intuition. Through simulations, I determine that investment is currently too low to successfully prevent the decline of water infrastructure quality. Counterfactual policies that promote only proactive projects lead some systems to make unnecessary investments even as others become vulnerable to extreme quality decline. By contrast, policies that facilitate more effective reactive policies incorporate more equitable levels of risk, reduce overspending, and enable all systems to maintain system quality.

The Evolution of Concentration and Markups in the United States Cement Industry (with Nathan Miller, Matthew Osborne, and Gloria Sheu)

We examine local market concentration and markups in the United States cement industry over 1974-2016. We estimate a model in which buyers use a second-score auction to procure cement from spatially differentiated plants. The model matches aggregated outcomes in the data, and the implied transportation costs and shipping distances are consistent with external sources. We infer local market concentration and markups from the model. At the county-level, the average HHI rises from 1,890 to 2,800 during the sample period. Average markups increase modestly, but prices do not rise. We attribute these changes to a technological innovation— the precalciner kiln—that lowered marginal costs, increased plant-level capacities, and also contributed to an industry shakeout in which many plants closed.

The Price That Inmates Pay (with Nathan Miller and Marleen Marra)

Incarcerated individuals in the United States purchase goods and services from monopoly vendors selected by their correctional authority. We study the price that inmates pay for phone calls, which the Federal Communications Commission has characterized as “exorbitant.” We specify an auction model of procurement and estimate it using data from public records requests. Our results indicate that market power contributes to high prices but that more important are kickbacks (or “commissions”) that providers give to the correctional authority. Regulation that substantially lowers price and eliminates commissions can more than double inmate surplus and simultaneously enable providers to recover their costs.

A Dynamic Discrete Choice Model of Electronic Toll Adoption in the U.S.