Print
E-MailEhsan Azarmsa is an Assistant Professor at the University of Illinois Chicago (UIC) College of Business Administration and Joel Shapiro is a Professor at the University of Oxford Saïd Business School. This post is based on their recent article, forthcoming in the Journal of Finance.
ESG ratings have become an important input into investment decisions. With the advances of sustainable investment, a large industry has developed to collect, analyze, and sell information about firms’ environmental, social, and governance performance. At the same time, ESG ratings have attracted considerable criticism. Ratings from different providers often disagree, raising concerns among investors, regulators, and academics about their accuracy and usefulness.
One common response is to compare ESG ratings with credit ratings, but credit ratings ultimately assess one dimension: credit risk. ESG ratings cover a wide range of potentially unrelated categories. Environmental performance can include climate change, biodiversity loss, and pollution; social performance can include human capital, product liability, and stakeholder relations; and each of these categories contains further subcategories. An ESG rating provider therefore faces a basic choice about where to devote its resources. It can specialize in a narrower set of categories, or it can generalize and spread its effort across many of them.
In our paper, The Market for ESG Ratings (Journal of Finance, 2026) we study how competition among ESG rating providers affects this choice. Our main finding is that competition can lead rating providers to become generalists even when specialization would produce more information. This distortion is most likely to arise when investors care very strongly about ESG performance.
Specialization versus generalization
Consider two rating providers assessing two ESG categories. If one specializes in one category and the other specializes in the second, their information complements each other and gives investors a relatively complete picture. If instead both providers cover both categories, they duplicate some of their effort and produce less precise information overall. Specialization therefore produces more information.
Competition, however, does not always lead to specialization. Suppose an investor will invest only after receiving sufficiently positive information about both environmental and social performance. A provider specializing only in environmental performance then has relatively little value on its own: its information becomes useful mainly when combined with another provider’s information about social performance. By covering both categories instead, the provider can make its own rating more useful to the investor, even though doing so reduces the total amount of information produced by the two providers.
Thus, when investors place substantial weight on performance across several ESG categories, competing rating providers may both choose to generalize. They cover the same ground rather than specializing in different areas. Thus, stronger investor demand for ESG information can lead to less informative ESG ratings.
Rating disagreement may not be bad
Our results also change how one should interpret disagreement among ESG ratings.
Low correlations among ratings are often treated as evidence that the ESG ratings market is not functioning well. Prior research has documented substantial divergence among major providers, and the lack of agreement has motivated calls for greater consistency and standardization.
Our analysis shows that disagreement can have a very different interpretation. When two rating providers specialize in different categories, their ratings will naturally disagree more because they are measuring different things. Indeed, in our framework, disagreement is greatest when the two providers specialize in different categories—the same outcome that maximizes the information available to investors.
This distinction is empirically relevant. Berg, Kölbel, and Rigobon (2022) attribute 38% of the discrepancy in category ratings among major ESG rating agencies to differences in the subcategories that providers examine—what they call “scope divergence.” In our framework, this type of divergence could reflect specialization and can be beneficial rather than harmful.
The implication is not that all ESG rating disagreement is desirable. Ratings may also disagree because providers use different data or measurement procedures, or simply because measurement is noisy. Rather, disagreement by itself is a poor measure of the quality of the ESG information market. Understanding why ratings disagree is more informative than observing whether they disagree.
Greenwashing can change the structure of the ratings market
Greenwashing provides another reason why the distinction between specialization and generalization matters.
ESG rating providers rely substantially on information disclosed by firms. If firms can manipulate that information—for example, by making environmental commitments whose implementation is difficult to verify—investors rationally become more skeptical about favorable ESG ratings.
We show that greenwashing can reduce the quality of ESG information in two ways. The first is direct: manipulated corporate disclosures make ratings less informative. The second operates through the organization of the ratings industry. As investors become more skeptical, they may require favorable information across several categories before investing. This reduces the value of a specialized rating on its own and gives raters a stronger incentive to generalize, producing less information.
Implications for regulation
These results suggest several considerations for the regulation of ESG ratings.
First, reducing disagreement among rating providers should not itself be a regulatory objective. Rules that push providers toward common scopes, methodologies, or data sources may raise correlations among ratings while reducing the diversity of information supplied to investors. Transparency about what a rating measures and how it is constructed is valuable. Standardization that goes beyond transparency, however, can discourage useful specialization.
Second, improving corporate disclosure can affect the ESG ratings market through more than one channel. Better disclosure directly gives rating providers more reliable inputs. It can also reduce investor concerns about greenwashing and thereby strengthen providers’ incentives to specialize and provide more information.
The broader lesson is that the quality of the ESG ratings market cannot be assessed simply by asking how accurate an individual rating is or how closely different ratings agree. ESG information is inherently multidimensional. The organization of the market determines which dimensions rating providers investigate and how much information they produce about each one.
