Semantic Clustering of Material ESG Issues: A Comparative Study on Indian Private and Public Sector Banks

Abstract

The role of banks in emerging countries such as India is indubitable, and with sustainable development being recognised as the path forward, this role has now added on the facets of sustainability. Similar to companies in other sectors, banks too report their sustainability practices and performance through sustainability reports, within which the concept of materiality assessments is an important aspect. Materiality assessments provide stakeholders with information on environment, social and governance (ESG) issues that companies deem to be material. While materiality assessments and issues have been studied for a myriad of industries, works comparing the material issues of banks in India remain scarce. This study makes use of multiple Python libraries to semantically cluster material ESG issues of Indian banks and compares the frequencies and proportions of clusters across the categories of private and public sector banks. Since discrepancies were found in material topics reported by banks across various reporting frameworks in a given year, the mandatory Business Responsibility and Sustainability Report (BRSR) was given precedence. Semantic clustering identified eight material ESG issues, with “Managing Climate Change Impacts” being the most frequent issue and “Regulatory Compliance” being the least frequent. Comparing the aggregate frequency of ESG using Wilcoxon-Mann-Whitney Test reveal that private banks report ESG issues more frequently than their public sector counterparts. However, testing differences in the proportions of frequencies across private and public sector exhibit insignificant differences, highlighting that across ownership categories, Indian banks are not qualitatively different in how they view ESG issues they deem material.