Work

Persistent Rating Disagreement, Measurement Robustness, and Information Reliability in Sustainable Finance: Implications for ESG Transparency and Evidence-Based Sustainability Governance in Chinese Listed Companies

Year: 2026
Type: preprint
Abstract: This study investigates cross-provider environmental, social, and governance (ESG) rating disagreement, temporal persistence, and measurement robustness, with particular emphasis on improving the reliability of sustainability information for ESG management and financial decision-making. Using an unbalanced panel of 1,323 firm-year observations representing 377 Chinese listed companies during 2019–2024, the study compares MSCI- and FTSE Russell-attributed assessments. Disagreement is operationalised through percentile-rank, standardised-score, and min-max transformations. The analysis integrates descriptive statistics, consecutive-year Spearman correlation, exploratory panel regressions, firm-clustered inference, multiple-testing adjustments, and reproducible computational verification. The mean absolute percentile-rank disagreement is 0.2493, indicating appreciable differences in relative corporate assessments. A consecutive-year Spearman correlation of approximately 0.493 demonstrates moderate temporal persistence. Alternative harmonisation procedures produce different inferential conclusions concerning exploratory financial associations, particularly firm size, highlighting the importance of measurement-sensitive interpretation. The study integrates disagreement magnitude, temporal persistence, and measurement robustness within a reproducible framework, extending ESG divergence research beyond isolated provider comparisons. Drawing on Institutional Theory and information asymmetry, it positions ESG disagreement as an information-governance challenge rather than direct evidence of inaccurate sustainability performance. The findings support systematic provider comparison, transparent rating harmonisation, and sensitivity assessment in ESG information management. Such practices offer a basis for improving evidence scrutiny and decision transparency, subject to limitations concerning secondary-data provenance, panel composition, and observational inference.
Language: English
Open Access status: green
OpenAlex ID: W7228195999
Cites:
Cited by:
Related to: 10