AEA Papers and Proceedings
ISSN 2574-0768 (Print) | ISSN 2574-0776 (Online)
Debiasing Estimates of Global Forest Cover Loss
AEA Papers and Proceedings
(pp. 81–86)
Abstract
Using machine learning predictions as proxies for difficult-to-observe outcome variables can bias empirical estimates when prediction errors correlate with treatment variables. We describe methods for detecting and correcting these biases using a sample of ground truth data. These types of data are often not available in practice, however. We construct a novel dataset on deforestation in Africa using approximately optimal sampling methods and visual interpretation of high-resolution satellite imagery. We use the data to evaluate bias in widely used satellite-derived measures of deforestation. We find that deforestation is systematically under-predicted in areas with higher rates of deforestation.Citation
Gordon, Matthew, Eliana Stone, Megan Ayers, and Luke Sanford. 2026. "Debiasing Estimates of Global Forest Cover Loss." AEA Papers and Proceedings 116: 81–86. DOI: 10.1257/pandp.20261018Additional Materials
JEL Classification
- C45 Neural Networks and Related Topics
- C51 Model Construction and Estimation
- O13 Economic Development: Agriculture; Natural Resources; Energy; Environment; Other Primary Products
- Q23 Renewable Resources and Conservation: Forestry
- Q54 Climate; Natural Disasters and Their Management; Global Warming