Overview
Despite the prominence of ESG investing, a fundamental gap persists: the empirical link between a firm’s environmental, social, and governance performance and its financial outcomes remains poorly understood. This talk applies scientometric analysis — citation networks, co-word maps, social network analysis, and text mining — to the ESG literature to map what is known, what is assumed, and what is missing.
Research Objectives
The analysis pursues three goals. First, to trace the history and evolution of ESG research, identifying dominant themes and how they have shifted over time. Second, to examine how the relationship between ESG factors, sustainability, and organisational performance is theorised in the existing literature — and where explanatory gaps remain. Third, to identify missing links and potential new data sources that could bridge those gaps.
Key Findings
Co-citation network analysis reveals that ESG research clusters around five themes: firm value, corporate social responsibility, consumer reaction, integrated reporting, and sustainability transition. These clusters are largely siloed; cross-cluster citation is sparse. Keyword network analysis further shows that the field concentrates on performance frameworks and organisational culture rhetoric, with limited empirical testing of the ESG-financial performance relationship.
Crucially, the analysis finds that organisational culture and innovation are under-examined mediating variables. Integrating these dimensions with ESG factors could substantially strengthen the theoretical links between ESG integration and measurable outcomes.
Implications and Alternative Data Sources
The talk proposes that alternative data sources — exemplified by the MIT Culture 500 Project, which uses AI to measure corporate culture at scale — could open new empirical avenues. If culture is a mediating variable between ESG commitments and performance, then machine-readable cultural signals provide a way to test that relationship rigorously rather than relying on self-reported ESG disclosures.
The conclusion is not that ESG integration is ineffective, but that the field has outrun its evidence base. Closing that gap requires more rigorous empirical work, better data infrastructure, and cross-disciplinary collaboration between finance, organisational theory, and AI research.
