Conference Agenda

This article evaluates the effectiveness of military interventions in deterring illegal deforestation in Colombia’s largest national parks. We focus on Operation Artemis, an intervention conducted between 2019 and 2022, which used military units to target illegal deforestation hotspots in the Amazon. To identify causal effects, we use a novel approach that exploits geographic and temporal variation of hotspot policing strategies, along with exogenous restrictions on the legal authority to enforce illegal deforestation. We show that the military interventions were able to reduce illegal deforestation. However, our evidence suggests that the rural hotspot policing strategies may have limited effectiveness in controlling the environmental crime. Results are robust to different types of specifications and controls.

This study examines whether exposure to residential greenness helped buffer the health impacts of COVID-19 lockdowns, using unique longitudinal survey data from Quebec collected between 2020 and 2022. We estimate two-way fixed-effects models that exploit within-individual variation in lockdown exposure over time and link respondents’ locations to a satellite-derived measure of local vegetation density. Lockdowns led to modest but statistically significant declines in mental health, concentrated among urban residents, while physical health was largely unaffected. These declines persisted after adjusting for regional COVID-19 dynamics and household-level economic stressors, suggesting a direct mental-health burden associated with mobility restrictions. We also find that greener residential environments were associated with smaller declines in mental health during lockdowns. Among urban respondents, the implied monetary cost of lockdown-related mental-health losses ranges from \$659 to \$1,034 per person, with exposure to residential greenness offsetting up to half of this loss.

This paper estimates the impacts of protected area downsizing and degazettement (PADD) on land-use dynamics in the Brazilian Amazon. Analyzing PADD events implemented between 2009 and 2015, we compare estimates from standard difference-in-differences methods to synthetic difference-in-differences, which addresses violations of parallel trends arising from selective treatment assignment. We show that conventional difference-in-differences estimates yield null effects, consistent with prior literature. Synthetic difference-in-differences estimates, however, show that PADD increases deforestation by approximately 23% relative to pre-treatment baselines, driven by a 40% increase in pastureland expansion and a 3,471% increase in mining area growth. The divergence in results suggests that earlier null findings reflect methodological limitations rather than the absence of actual effects. Our findings underscore the importance of legal protection for environmental outcomes, especially in politically or economically contested areas.

This study presents a methodologically rigorous cost-benefit analysis of Precision Agriculture (PA) and Enhanced Efficiency Fertilizers (EEF) in the Želivka catchment, Czech Republic, serving 1.3 million people. We develop a dynamic probabilistic framework combining high-resolution LPIS field data (52,845 hectares across 729 farms), SWAT+ hydrological modeling outputs for nitrogen transport pathways, and recently derived health cost valuations (EUR 7.25 per mg/L per capita) with biophysical effectiveness parameters from meta-analysis of 403 field experiments. The heterogeneous agent adoption model accounts for farm size distribution, risk aversion, and learning costs, calibrated using Czech agricultural structure. Monte Carlo simulation (10,000 iterations) propagates uncertainty across all key parameters over a 10-year horizon at 3.38% social discount rate. The analysis reveals that indirect public health benefits dramatically outweigh direct private costs and benefits: while farmers gain EUR 23.7 million from yield improvements, society captures EUR 99.6 million in avoided health costs from reduced nitrate exposure. Technology adoption reaches 44.7% (PA) and 7.3% (EEF) by Year 10, reducing nitrogen loads by 65,843 kg annually and decreasing reservoir nitrate concentration by 2.96 mg/L. Sensitivity analysis confirms that SWAT-calibrated nitrogen transport parameters and health cost valuation drive results more than technology costs, highlighting the critical importance of proper externality valuation in agricultural environmental policy design.

US regulators employ summer-only gasoline volatility restrictions to attenuate the formation of ozone. This paper reexamines the impacts of such regulations on air quality and welfare, advancing prior literature by studying particulate matter effects alongside ozone. We leverage high-frequency, spatially detailed data and apply parsimonious polynomial seasonal adjustments and flexible non- parametric controls for weather, featuring temperature, precipitation, lags, and historical patterns. Complementing whole-year OLS regressions, we introduce a two-stage event-study approach that de-seasonalizes pollution time series before analyzing discontinuities around annual policy phase-ins and phase-outs, with robust controls. For the RVP Phase II program, whole-year regressions indicate summer volatility restrictions reduce ozone concentrations by 2.9%, while event-study type models suggest a 2.1% decline. In a possible adverse unintended consequence, we find the same regulations increase PM10 concentrations by 3.6% (whole year models) to 11.6% (event study models). Net mortality effects are subject to uncertainty, but are projected to be adverse most likely. We recommend these policies be further evaluated.

Demographic disparities in exposure to various sources of environmental stress are well documented. While specific stressors are linked to adverse health outcomes, there is uncertainty about how they interact. A first step towards evaluating these joint impacts is to understand the degree to which the same individuals are exposed to extreme simultaneous exposure to each stressor. Here, we adapt the Alkire-Foster multi-dimensional poverty measure to rank exposures to multi-dimensional environmental harm in a way that accounts for the frequency, breadth, and severity of exposure among the extremely exposed. This measure can be used to normatively compare distributions of extreme simultaneous exposure for a given group both across time and in comparison with other groups. We use publicly-available data on air toxics, PM2.5, and land surface temperature to identify trends in the multi-exposure index for socio-demographic groups based on race/ethnicity and poverty status for 168 U.S. cities. Controlling for city-characteristics, we find persistent, yet narrowing, multi-exposure gaps on the basis of race/ethnicity between people of color (POC)

and non-Hispanic White. Within each group, individuals in households above the poverty line fare better than those below, but above-poverty POC have similar multi-exposure to below-poverty non-Hispanic White.

Wind turbines are essential for decarbonization but often face local resistance due to noise, visual disruption, and landscape change. These burdens are spatially concentrated and materialize in reduced property values, yet existing compensation policies rarely align with where these losses occur. Here, we assess how well different financial compensation designs match the local disamenities of wind energy. Using over 680,000 housing listings in Germany, we estimate distance-sensitive property value losses from turbine exposure using causal forests and extrapolate these effects nationally via spatial smoothing. We then simulate eight compensation schemes, defined by their assessment basis (e.g., generated power, number of houses) and apportionment rule (by land area or house count). We find that compensation performance varies widely by design: to offset half of the disamenities, the least efficient Power-Area scheme requires 118% more funding than the most precise alternative, House-House. Schemes tend to overcompensate rural municipalities and undercompensate larger cities. Current designs, like Germany’s Power-Area policy, cover less than 10% of disamenities. Linking payments to housing patterns, rather than turbine output, yields significantly better alignment. Our framework offers a scalable tool to improve the targeting precision of compensation schemes in wind deployment.