The role of ESG indicators in closing the disaster insurance gap: a machine learning analysis

This study investigates the role of environmental, social, and governance (ESG) indicators in addressing global protection gaps in disaster insurance. Utilizing a comprehensive dataset from EM-DAT and the World Bank, covering the period from 2000 to 2022, the research employs advanced machine learning techniques to analyze the complex relationships between ESG factors and disaster insurance coverage. Among the methodologies applied, CatBoost and Gradient Boosting stand out for their strong predictive performance and reliable generalization capabilities. The findings reveal that governance quality, particularly in terms of stronger control of corruption, is the most significant ESG factor. On the social dimension, improved access to essential infrastructure, emerges as a crucial contributor to disaster insurance coverage. Additionally, environmental conditions related to the economic significance of primary sectors help elucidate variations in coverage. Alongside these ESG variables, disaster-specific factors, particularly total economic damage, remain critical determinants of protection gaps.