Are Corporate Governance Ratings Biased? Evidence from Institutional Shareholder Services

We examine whether the corporate governance (CG) ratings issued by Institutional Shareholder Services (ISS) are biased in favour of potential consulting clients. Using novel ISS data (2013–2021), including underlying governance inputs and model points, we test whether large firms, more likely to be clients, receive systematically higher discretionary scores or benefit from model adjustments. Controlling for observable governance characteristics, we find no evidence of upward bias, which holds around changes in ISS’s ownership that plausibly altered its incentives. Our findings contrast with prior evidence of bias in ISS’s proxy voting advice and suggest that differences in product design and verifiability constrain the scope for bias in CG ratings. We contribute to the literature on information intermediaries and CG ratings by identifying conditions under which conflicts of interest do not lead to biased outputs. When CG ratings involve limited analyst discretion and high verifiability, conflicts may not translate into distortions.