Working papers

“Water-Proofed: Floods and Water Infrastructure Quality in the United-States” (Job Market Paper)

Click here for the latest version.

Rainfall-induced floods impose large economic costs in the United States. These events, which occur throughout the country, are projected to increase in frequency and severity as global temperatures rise. This paper studies how improvements in water infrastructure -- sewer systems -- mitigate the economic impacts of rainfall-induced flooding. The empirical analysis combines a novel statistical model of rainfall-induced flood risk (RIFR) with hourly data on precipitation across U.S. counties from 1996 to 2019. I document how two key elements in the model, the intensity and duration of rainfall, both contribute to predict the occurrence of actual floods and the economic damage from floods. I then study how investments in sewer systems, by improving "drainage flow", reduce the likelihood and severity of flooding for a given level of rainfall. To estimate the causal impact of these investments, I exploit an unanticipated change in federal funding for local water infrastructure, following the 1977 Amendment to the Clean Water Act (CWA). This amendment restricted grant recipients from using CWA funds to finance stormwater systems. My analysis is based on a continuous difference-in-differences framework that compares the relationship between the RIFR and flood occurrence, across counties that received CWA grants shortly before and after the 1977 Amendment. I find that greater investment in sewer system infrastructure led to substantial reductions in local flooding. I estimate that the benefits from these flood mitigation investments exceeded their costs, saving nearly $23 million for the average county. More broadly, these findings highlight how deteriorating public infrastructure may exacerbate the consequences of climate change.

“Measuring the Economic Costs of China’s zero-COVID Policy from Outer Space” (With Jian tang), Submitted

Click here for the latest version.

We quantify the effects of the “zero-COVID” policy using a rich set of county-level lockdown events in China and nighttime satellite imagery. We find that more stringent lockdowns induce a large contemporaneous decline in nightlight followed by a slow recovery, which happens at least two quarters after lockdown enactment. Absent widespread contagions, a county under total lockdown incurs on average a 6% GDP loss compared to those without restrictions. The negative effect is particularly persistent in service-heavy areas as opposed to manufacturing-heavy areas. There exists some evidence consistent with spillover effects near counties under lockdown, but these effects are short-lived.

Work in progress

  • “Centralized Water Systems Adoption in Early American Cities”

  • “Weather shocks and Protectionism in the 20th Century U.S.” (with Andriana Bellou, Emanuella Cardia,and Joshua Lewis)

  • “Trade, Shocks Transmission, and Monetary Policy in a Regional Trade Agreement” (with Fansa Koné)