This paper studies the question of how to target aid after a natural disaster. Disaster aid programs often use property damage as a criterion for eligibility. A household's ability to insure against shocks may be harder to observe but more important in determining how the disaster affects welfare. We develop a model of demand for reconstruction aid and estimate the model parameters using data from a household survey following the 2015 Nepal earthquake and moments derived from a spatial regression discontinuity design. We use the model to estimate welfare from counterfactual targeting strategies. Conditioning aid on damages increases welfare by 9% relative to a random allocation, but a geographically targeted aid strategy based on earthquake intensity would have increased welfare by 12%. Furthermore, policymakers face trade-offs between targeting strategies aimed at consumption smoothing and those prioritizing poor households. Damage-based targeting performed worse than random by a poverty-focused measure of welfare.

[VoxDev Summary] [SSRN]

Advances in machine learning and the increasing availability of high-dimensional data have led to the proliferation of social science research that uses the predictions of machine learning models as proxies for measures of human activity or environmental outcomes. However, prediction errors from machine learning models can lead to bias in the estimates of regression coefficients. In this paper, we show how this bias can arise, propose a test for detecting bias, and demonstrate the use of an adversarial machine learning algorithm in order to de-bias predictions. These methods are applicable to any setting where machine-learned predictions are the dependent variable in a regression. We conduct simulations and empirical exercises using ground truth and satellite data on forest cover in Africa. Using the predictions from a naive machine learning model leads to biased parameter estimates, while the predictions from the adversarial model recover the true coefficients.

The distributional effects of climate change are at the heart of international climate negotiations. This paper shows how different property rights regimes, ranging from “right to pollute” to “right to no pollution”, rationalize different welfare weights in climate-economic models with heterogeneous regions. Commonly used Negishi weights separate the issues of climate change and global wealth inequality. However, we show that the separation of these issues does not yield a unique Pareto efficient allocation since climate change and climate policies have distributional consequences of their own. As a result, different property rights characterize a set of efficient allocations. In addition to Negishi weights, which implicitly reflect mixed property rights, we define beneficiary pays and polluter pays weights, derived from transfer rules consistent with right to pollute and right to no pollution property rights, respectively. These weights correspond to distinct Pareto efficient allocations that differ only in the distribution of the cost burden of climate damages and abatement, and we show how nations’ characteristics shape their preferences for different property rights regimes. Unlike the Negishi solution, the other efficient allocations involve international transfers for abatement and climate damages, providing theoretically grounded definitions for climate mitigation finance and Loss and Damage payments—both widely discussed in international negotiations. We use calibrated simulations to illustrate the distributional implications of different property rights regimes.

What are the impacts of the global trade in plastic waste? Since a 1991 World Bank memo argued, “the economic logic behind dumping a load of toxic waste in the lowest wage country is impeccable” cross country flows of plastic waste have increased seven-fold. Yet the consequences of this trade are poorly understood, partially due to the lack of basic descriptive data about what happens to exported waste at its many destinations. Using satellite data and a machine-learning algorithm trained on crowd-sourced data, we study the spread of informal dumps before and after a major shock to the global trade in waste. Our preliminary findings show that the dumps create severe externalities, while the benefits to jobs and wages appear modest.

Several observational studies from locations around the globe have documented a positive correlation between air pollution and the severity of COVID-19 disease. Observational studies cannot identify the causal link between air quality and the severity of COVID-19 outcomes, and these studies face three key identification challenges: 1) air pollution is not randomly distributed across geographies; 2) air-quality monitoring networks are sparse spatially; and 3) defensive behaviors to mediate exposure to air pollution and COVID-19 are not equally available to all, leading to large measurement error bias when using rate-based COVID-19 outcome measures (e.g., incidence rate or mortality rate). Using a quasi-experimental design, we explore whether traffic-related air pollutants cause people with COVID-19 to suffer more extreme health outcomes in New York City (NYC). When we address the previously overlooked challenges to identification, we do not detect causal impacts of increased chronic concentrations of traffic-related air pollutants on COVID-19 death or hospitalization counts in NYC census tracts.

If the material intensive enterprises in an urban area of several million people shared physical resources that might otherwise be wasted, what environmental and public benefits would result? This study develops an algorithm based on lifecycle assessment tools for determining a city's industrial symbiosis potential—that is, the sum of the wastes and byproducts from a city's industrial enterprises that could reasonably serve as resource inputs to other local industrial processes. Rather than report, as do many previous papers, on private benefits to firms, this investigation focuses on public benefits to cities by converting the maximum quantity of resources recoverable by local enterprises into an estimate of the capacity of municipal infrastructure conserved in terms of landfill space and water demand. The results here test this novel approach for the district of Mysuru (Mysore), India. We find that the industrial symbiosis potential calculated based on analysis of the inputs and outputs of ∼1000 urban enterprises, translates into 84 000 tons of industrial waste, greater than 74 000 tons of CO2e, and 22 million liters per day of wastewater. The method introduced here demonstrates how industrial symbiosis links private production and public infrastructure to improve the resource efficiency of a city by creating an opportunity to extend the capacity of public infrastructure and generate public health co-benefits.

For decades Americans have grown up learning fundamental waste management principles: reduce, reuse, recycle. Yet, over half of what we generate from homes and businesses—known as municipal solid waste—ends up being discarded in landfills. While recycling has increased, so has our understanding of its limits. New policies and strategies can refocus waste management so that it is less costly for residents and more beneficial for the environment. In particular, twenty-first century waste and materials policy should seek to preserve, rather than throw away, the materials and energy that go into the products we use. This priority requires a new focus on material diversion, rather than waste treatment and disposal.