Research/previous
Green bonds and carbon emissions
Causal machine learning estimates of whether municipal green-bond issuance is followed by lower local carbon emissions.
Municipal green bonds are issued to fund projects with an environmental purpose. Whether issuance is followed by lower county-scale carbon emissions is a harder question, because counties that issue green bonds also differ in income, infrastructure, and commuting patterns.
This paper uses a causal forest in a double machine learning framework on U.S. municipal green bonds issued from 2009 to 2019. Emission changes are uneven across counties and grow over time. The associated paper in Environmental Science & Technology (2025) reports an implied abatement cost of $192 per ton of CO2, with stronger reductions in counties that have more small and medium-sized employers.
People
Papers
- Green Bond Issuance and Carbon Emissions: Can Causal Machine Learning Inform Forward-Looking Policy Decisions?Environmental Science & Technology
Links
- No public repository yet.