Abstract
This paper examines the causal relationship between Environmental, Social, and Governance (ESG) factors and regional mental health in Italy from 2004 to 2023. Although the single ESG factors have been linked to sustainability and public health, their combined and individual contributions to mental well-being remain under-researched, particularly at the sub-national territorial scale. By adopting a decomposed ESG framework, the paper aims to estimate three single Instrumental Variable (IV) Panel Data specifications—representing the environmental, social, and governance pillars—using both Fixed Effects (IV–FE) and Random Effects (IV–RE) specifications. Based on the ISTAT-BES dataset, this approach eliminates endogeneity and regional heterogeneity, thereby enabling robustness-based causal inference. The empirical results indicate that environmental dissatisfaction and ecological decay have a negative association with the Mental Health Index (MHI) score, whereas favorable climate aspects and sustainable material use have a positive association with the MHI score. Poverty risk affects mental health negatively the most, in the poverty pillar, whereas lower work intensity and welfare protection have a positive effect. Institutional quality is a determining factor that emerges as such: lengthy judicial processes, perceived insecurity, and territorial decay have been correlated with low mental health, whereas institutional belief and public protection have cooperating effects on psychological well-being enhancement. High intra-class associations indicate resilient structural gaps between the regions. The results show that mental health gaps are highly diffuse along environmental, social, and institutional directions. Moving further into the exploration of these gaps, the paper applies informal clustering methods based on the K-Means algorithm, tuned using the Silhouette coefficient, to recognize stable socio-territorial patterns within each individual ESG pillar. The clustering outcomes indicate the persistent presence of a north–south bipolarity, as the northern regions, as well as the center, consistently appear within high-performing clusters, characterized by better environmental quality, stronger social protection, and more efficient governance. Southern regions, as the islands, are still more often found in clusters characterized by ecological degradation, labor market weakness, poverty risk, and institutional inefficiencies. The clustering analysis presented here suggests IV estimations, providing a typology of territorial determinants in mental health, as well as highlighting a structural coincidence between ESG weakness and low MHI scores. The integration of ESG principles into public policies should have priority, then, as a means of promoting sustainability as well as mental robustness. The article presents a transferable framework for analyzing regional well-being by assessing psychological health as a multidimensional product of sustainable governance and socio-environmental conditions.