Foreign aid's double-edged sword effect on carbon emissions: A machine learning approach

Qinglong Xiong, Yanxia Zheng, Hai Long,
Foreign aid's double-edged sword effect on carbon emissions: A machine learning approach,
Journal of Cleaner Production,
Volume 530,
2025,
146890,
ISSN 0959-6526,
https://doi.org/10.1016/j.jclepro.2025.146890.
(https://www.sciencedirect.com/science/article/pii/S0959652625022462)
Abstract: Although previous studies have empirically demonstrated that foreign aid (FA) may decrease carbon emissions in recipient countries, the causal relationship remains unclear. This study uses machine learning methods to investigate this dynamic relationship and seek ways of improving FA's effectiveness. Examining World Bank data from 152 recipient countries, this study demonstrates that FA has a double-edged sword effect, exhibiting a nonlinear relationship. FA reduces carbon intensity within a “Golden Zone.” Beyond this range, however, the attenuation law of aid effectiveness applies, meaning that in manufacturing-dominated economies FA may even increase carbon emissions—a dynamic we call the “decarbonization bottleneck.” This dilemma persists until a service-oriented economy becomes dominant. These findings extend the environmental Kuznets curve hypothesis, suggesting that advanced economic structures help overcome the carbon reduction bottleneck and support sustainable FA, while preventing the “too-much-of-a-good thing” effect of FA on environmental performance. Practically, the recipient countries should pay attention to the potential negative effect of FA.
Keywords: Foreign aid; Carbon emissions; Decarbonization bottleneck; Machine learning; Environmental Kuznets curve