ESG Rating regulation

Let’s speak about the regulation of ESG rating agencies. When talking about it, I often encounter a mix-up between regulation of firm-level disclosure and ESG ratings. But these are, of course, two separate issues.

Let’s start with the regulation of firm-level disclosure. To improve the quality of ESG ratings, we need better input and hence better disclosure practices. This debate should mainly focus on what issues should be disclosed and how to ensure reliability, e.g., impose mandatory auditing. Regarding ESG ratings, regulation is a bit more subtle. As our Aggregate Confusion paper shows, the divergence between ESG ratings can be separated in what is measured, its relative importance, and how it is measured. As every investor is entitled to their preferences, a regulator should not set in stone what is measured and how important the issue is.

Measurement is a different case, however. The measurement divergence hints at measurement noise, i.e., we measure imperfectly. For instance, we do not directly observe how women are treated in a company and thus rely on proxies such as the wage gap. Regarding gender diversity, there is a conflict between measuring perfectly and privacy concerns. If you want to measure how women are treated, in theory, you need to analyze every interaction, including noninteractions between every human in a given organization. This is hardly possible nor welcome as data privacy is one of the concerns of ESG.

However, regulation can favor an improvement of measurement. I suggested in my talks with AMF and IOSCU that we need to put in place regulation that forces ESG raters to be more transparent about their measurement practices and data collection methods. Increased transparency will increase competition because investors can make more informed decisions. And only then academics, NGOs, the media, and rated firms will be able to criticize the practices of ESG raters in a constructive way.