Session summary
This GeoField 2026 methods workshop provides a hands-on introduction to using geospatial impact evaluation to assess forest conservation outcomes. The session uses a case study from Madagascar to examine whether biodiversity offsets associated with the Ambatovy nickel and cobalt mine reduced deforestation.
Offsetting Forest Loss from a Major Infrastructure Development
Katie Devenish leads a workshop on evaluating biodiversity offsets created to compensate for forest loss from a major mining project. The analysis estimates what would have happened to forest cover in the offset sites without protection, then compares that counterfactual to observed deforestation.
Participants work through a practical evaluation workflow using R, including data preparation, control sample filtering, pixel-based matching, covariate balance checks, matched-pair extraction, site-year aggregation, parallel pre-trends testing, and difference-in-differences estimation. The workshop also examines how choices about outcome measurement, control regions, matching covariates, spillovers, and units of analysis shape the credibility of the results.
The session closes with robustness checks and a broader discussion of conservation policy. Alternative matching specifications show how sensitive results can be to design choices, while the case study raises important questions about biodiversity offsets, avoided deforestation, and the social costs that conservation restrictions may impose on local communities.