Three-quarters of S&P 500 companies now tie a portion of their CEO's pay to environmental, social and governance (ESG) metrics . They typically include carbon emissions, workforce diversity and worker safety, among others.
The justification is straightforward: if shareholders want corporations to take climate change and social responsibility seriously, firms should pay their leaders for achievements on these dimensions. This practice is encouraged by boards and large institutional investors.
Regulators are now trying to standardize the underlying metrics so investors can compare firms on a common basis. The European Union has gone the furthest by directly regulating ESG rating providers .
In 2024, the Canadian Sustainability Standards Board released its sustainability disclosure standards aligned with two global standards issued in 2023 by the International Sustainability Standards Board . Around 40 jurisdictions have now adopted those standards or taken formal steps toward doing so.
Why the push for standardization?
One persistent problem concerns how ESG performance should be measured in the first place. Today, major rating providers - including MSCI , Sustainalytics , S&P Global and Bloomberg - rate the same firms very differently, even when assessing the same dimension of performance.
One influential paper found that the average correlation between major ESG ratings is around 0.54 - far below the near-perfect agreement between credit rating agencies. The same firm can look like a sustainability leader under one provider's score and a laggard under another's.
This divergence is widely seen as a problem, and the standard prescription is harmonization. The conventional view is that convergence on a common standard is unambiguously desirable. But is that really the case?
In a recent paper , my co-reseacher Nicolas Sahuguet and I set out to answer a simple question: If executives understand how the metrics tied to their pay are calculated, how will they actually respond? The answer points to an unintended consequence of the push for harmonization.
When targets get gamed
Critics, including law professor Lucian Bebchuk at Harvard University and economist Alex Edmans at London Business School, have argued that tying executive compensation to specific ESG metrics invites executives to game the scheme and may end up exacerbating the agency problem of executive pay.
An executive who knows exactly how a carbon-intensity score is calculated does not need to actually reduce their firm's environmental impact to improve that score. They can outsource emissions-heavy production to external suppliers or shift the firm's activities toward those that improve relevant metrics without changing its underlying environmental impact.
None of this requires fraud - only an understanding of how the scoring system works.
What our research shows
To examine this, we built a formal model of the relationship between a manager and a socially responsible board . The model accounts for the manager's ability to anticipate how their decisions will affect their own ESG-based pay, and to game the incentive scheme accordingly.
The picture that emerges is more nuanced than the debate usually allows. ESG-linked pay can be second-best optimal, but it is never without cost.
Because our model showed that the manager games whatever metric is used, ESG bonuses inevitably distort their decisions, diverting resources toward investments that improve the score rather than the underlying outcome.
The board accepts this distortion only when the alternative is worse - that is, when it genuinely wants the firm to do more for the environment or for stakeholders than what would maximize its stock price.
If shareholders already reward social performance through the stock price, as they increasingly do, equity-based pay is already providing adequate incentives. Adding ESG bonuses on top is then counterproductive on two fronts: it distorts investment decisions through gaming and pushes the firm to over-allocate resources to ESG activities beyond what either shareholders or the board actually want.
This helps explain why the sensitivity of CEO pay to ESG metrics is usually small, even in firms that have made serious public commitments to environmental or social goals. That is not necessarily window dressing. Boards keep this sensitivity low precisely to limit the distortion from gaming.
Why disagreement has value
Our model also generated a less obvious finding: the current patchwork of competing rating methodologies may actually be doing useful work. When several raters use different methodologies, gaming becomes harder because what improves one score may not improve another.
An executive who knows that Provider A weights one set of indicators, Provider B weights different ones and Provider C weights different ones still cannot easily satisfy all three without genuinely improving underlying performance.
A single official measure, by contrast, gives every CEO a clear target to optimize against. Once the methodology is public and predictable, the gap between hitting the metric and improving actual performance widens.
For harmonization to be a net improvement, the unified standard would need to be of substantially higher quality than the patchwork it replaces by a factor that scales with the number of providers being consolidated.
The central premise driving the harmonization push - that disagreement among raters is a flaw to be regulated away - deserves more scrutiny than it has received. That is a high bar.
Disagreement among raters has costs, but it also has benefits. By limiting a manager's ability to game the metrics, multiple independent measures allows firms to provide more effective incentives across ESG dimensions. Regulators should consider preserving the discipline that comes from keeping those measures independent.
Pierre Chaigneau is a Research Fellow at the Institute for Sustainable Finance.
