Institutional quality and risk depletion of natural resources: Evidence from explainable artificial intelligence modelling

As global temperatures continue to rise, urgent action from public authorities is required to conserve natural resources more effectively, since widespread environmental degradation and overexploitation of resources already threaten ecosystems and human well-being in many parts of the world. Using data from 167 countries over the period 2007–2023, this study identifies the most effective institutional dimensions for resource preservation. The analysis utilized explainable artificial intelligence modelling to capture nonlinear and complex interactions among institutional and investment environments, business conditions, economic contexts, and infrastructure and market access, thereby generating accurate predictions and interpretable insights into the key drivers of resources conservation. The results show that land, forest, soil and freshwater resources are better protected in countries with strong political accountability and efficient regulatory institutions. From a practical standpoint, on one hand, the potential threat of electoral backlash can motivate politicians to adopt environmentally and ecologically friendly policies. On the other hand, high quality regulatory frameworks strengthen the adaptive capacity of institutions, enabling timely responses to changes in the natural environment. Additionally, strengthening these institutional dimensions should be complemented by public policies that promote affordable electricity, foster innovation, and enhance labour force engagement. Finally, our results provide supporting evidence for neo-institutional theory, emphasizing the environmental and ecological benefits of institutional legitimacy.