Signals of Legitimacy or Tools of Strategy? Investigating ESG Score Drivers in the Age of Sustainability Accounting Standardization, an Analysis of Canadian Firms

Abstract

Environmental, Social, and Governance (ESG) scores have become a central metric for assessing corporate sustainability performance. Yet significant concerns remain regarding their reliability, comparability, and sensitivity to firm-level disclosure practices, including the choice of sustainability reporting framework. This study examines whether the ESG reporting system chosen by firms, specifically, the Global Reporting Initiative (GRI) or the Sustainability Accounting Standards Board (SASB), affects the ESG scores assigned to Canadian publicly listed companies. Using a panel dataset of 345 TSX-listed firms from 2014 to 2023, the analysis incorporates firm-level ESG scores from LSEG Refinitiv, sustainability reporting data, and financial and industry characteristics sourced from Compustat. The study adopts a quantitative research design rooted in legitimacy theory, employing multiple regression methods including pooled OLS, fixed and random effects models, and more advanced nonlinear approaches such as Generalized Additive Mixed Models (GAMMs) regressions. Results consistently show that the choice of reporting framework is not a statistically significant determinant of ESG scores. Instead, firm characteristics such as revenue turnover, market capitalization, industry affiliation, and reporting year are stronger predictors. These results suggest that ESG rating systems may either normalize across reporting frameworks or overweight internal financial and governance metrics, minimizing the role of voluntary reporting format. This finding supports the comparability of ESG scores across frameworks but also raises concerns about the limited influence of narrative disclosures on divergent rating outcomes. The study concludes that the impending global shift toward IFRS S1 and S2 may have less impact on ESG scores than anticipated, calling into question some of the assumptions underpinning current regulatory convergence efforts. Recommendations for researchers, regulators, and firms are provided, along with directions for future study as ESG disclosure standards and rating systems evolve.