"The Long-Term Impact of Oil Spills on Ecosystem Recovery" Job market paper.

Work in progress.

Abstract: Maritime disasters are often appraised using contingent valuation methods, despite their weak connection to measurable ecological damage. This paper uses satellite-based measurements and a synthetic difference-in-differences design to estimate ecological losses at sea in a scalable way. Applying the approach to major oil spills worldwide, I find that large spills generate persistent ecosystem damage: ten years later, the ocean’s capacity to support fish biomass is about 20 percent lower. The implied economic losses reach up to USD 1.84 billion for a single spill. These damages far exceed observed financial compensation, suggesting that existing liability frameworks substantially undervalue marine ecosystem losses.

"Long-Term Environmental Damage: Evidence from European Wildfires" with François Cohen (University of Barcelona and University of Oxford) and Paola Ovando-Pol (Spanish National Research Council).

Work in progress.

Abstract: Economic impact assessments are rarely able to quantify persistent or irreversible losses in natural capital. This paper provides the first continental-scale estimate of long-term environmental damage from wildfires in Europe. Combining satellite data, spatial matching, and ecological modelling, we causally identify how wildfires in 2001–2005 have durably altered forest cover and carbon stocks. We find that 26-percent of burned forests transition to lower-carbon ecosystems. Even after 15 years, areas burnt in 2002-2005 had a 28-percent deficit in stored carbon in 2018 compared to control sites due to recurrent fires and slow recovery. By integrating ecological realism into economic assessment, our results challenge the assumption of natural regeneration. In future versions of this paper, we will use our approach to inform policies on climate mitigation, forest management and forest carbon credits.

"Pollution and congestion in areas near airports: Evidence from a major shock" with Xavier Fageda (University of Barcelona).

Submitted & under review.

Abstract: We examine the closure of Turkey’s largest international airport, taking advantage of granular data collected hourly at the sensor level. The main goal of the paper is to causally assess the impact of the shock on pollution and road congestion. Using a difference-in-difference model, our findings show a decrease in NOX of 27% near the closed Atatürk Airport. With a back-of-the-envelope calculation, the decrease in NOX near Atatürk Airport corresponds to a 3% decrease in excess relative risk (ERR) for cardiovascular mortality and a 1.5% decrease in ERR for hospital admissions owing to respiratory diseases. We also find a 32% decrease in car traffic and a 3% decrease in travel time for every 100 meters of additional proximity to Atatürk Airport compared to the pre-shock rate of congestion on average. The impacts of closure are greater when heterogeneity of road classes is considered. Finally, we conducted a novel analysis on the sources of NOX emissions, car versus air traffic. Our findings suggest that car traffic generated by the presence of the airport represents a greater source of NOX than flights themselves. These findings quantify the heterogenous impacts which airport closure generates.