Unveiling the drivers behind carbon emissions disclosure: an ESG perspective

Purpose This study aims to explore the key drivers of voluntary carbon disclosures among French firms, highlighting environmental, social and governance (ESG) components and sustainable investments as crucial factors. Design/methodology/approach This study uses H2O AutoML, an advanced machine-learning framework, to examine CO2 emissions disclosure among 77 French non-financial companies listed on the SBF 120 index from 2017–2021. The research rigorously assesses CO2 disclosures using the Carbon Disclosure Project Index criteria, enhancing the precision of the findings. This approach paves the way for innovative advancements in ESG research by integrating cutting-edge computational tools with conventional environmental reporting metrics. Findings Social performance and sustainable investments are the most significant predictors of CO2 disclosure, outweighing traditional financial metrics such as the Book-to-Market ratio and direct environmental factors. Variable importance analysis and heatmaps underscore the critical role of social factors in shaping corporate transparency regarding carbon emissions. Practical implications This research offers valuable insights for companies and policymakers aiming to enhance environmental transparency and reporting, underscoring the significance of integrating ESG factors. Carbon emissions disclosure plays a critical role in promoting sustainability, ensuring regulatory compliance, attracting investors and strengthening risk management practices. Originality/value This research presents an innovative methodology for analyzing CO2 emissions disclosure using advanced machine learning within the H2O AutoML framework. It emphasizes the importance of model diversity and variable consideration to comprehensively understand environmental disclosure, potentially guiding policymakers and businesses in improving transparency and sustainability practices.