Exploring university sustainability performance through the Environmental, Social, and Governance (ESG) framework using Bayesian belief networks

Abstract Sustainability is increasingly critical for universities as institutions influence environmental stewardship, social equity, and the management of economic resources. The Environmental, Social, and Governance (ESG) framework is a widely used tool to assess university sustainability efforts, providing insight into institutional policies, practices, and external contributions, such as alumni research and societal engagement. This study analyzes data from 1,397 universities using the 2024 QS (Quacquarelli Symonds) sustainability rankings, which evaluate environmental impact, social impact, and governance indicators. A Bayesian Belief Network (BBN) model is applied to uncover the complex interdependencies among governance, social impact, and environmental impact, revealing how these factors jointly influence overall sustainability performance. The findings demonstrate that governance plays a pivotal role, with strong governance enhancing both social and environmental outcomes. By highlighting these interdependencies, the study provides valuable insights for university leaders and policymakers seeking to optimize ESG-based strategies to improve institutional sustainability. The main contribution of this research lies in applying a network-based probabilistic model to ESG indicators, offering a more nuanced understanding of how governance and other ESG factors interact to shape sustainability performance in higher education.