December 2020

What's Wrong With 'Social Responsibility'?

This publication is a part of:

Collection: Hamilton’s Corner


There is a lot of talk these days about "stakeholder capitalism," especially as applied to publicly owned companies, including banks. The idea is that companies that seek only to maximize their value, subject to the constraint of abiding by the law, are missing the mark. They should instead target what is good for society, not just what creates value for their shareholders.

In the case of banking, in the interest of social responsibility, some banks sought to deny credit and other financial services to whole legal sectors of the economy that their managers or "stakeholders" did not approve of politically, such as gun makers, fossil fuel companies, or private prisons. The Office of the Comptroller of the Currency issued a proposed fair-access rule in November 2020 clarifying its longstanding guidance against such sectoral redlining.

This example raises a broader question that applies to many other corporate practices and many other firms. Is it beneficial for our society to pressure Chief Executive Officers (CEO) and asset managers to consider objectives other than value maximization and legal compliance when deciding how to run their companies (if they are CEOs) or what stocks to invest in (if they are asset portfolio managers)? In this essay, I summarize some economists' views that value maximization still provides the right overarching objective to guide corporate managers, from the standpoint of economic efficiency and from the standpoint of achieving bona fide public policy goals determined by our democratically elected representatives.

I argue that the alternative "stakeholder" perspective does not even offer a coherent alternative to value maximization. Those who advocate social responsibility as an alternative objective ask that companies do good things, but they cannot offer an alternative standard for determining what exactly they should do. I also explain why stakeholder capitalism is gaining popularity. Precisely because it does not offer a clear standard for measuring achievement; it can be used by CEOs, asset managers, and politicians to avoid accountability for their failures. From that perspective, it should not be surprising that so many politicians, CEOs, and asset managers have embraced the new orthodoxy. Indeed, they have already succeeded—to their own gain, but to our loss—in making it much harder for shareholders, fund investors, and voters to hold all three groups accountable for their failures.

In explaining this view, I have to spend some time discussing how fuzzy compliance with so-called environmental, social, and governance (ESG) objectives work, or more accurately, why they don't work, in contrast to compliance with the law, and with the discipline that comes from shareholder value maximization. Then I discuss why pursuing ESG objectives leads to inferior outcomes from the standpoint of economic efficiency and the achievement of bona fide regulatory goals.

Much of my analysis draws on a recent PhD thesis by Ruoke Yang a year ago.1 Ruoke was a student of Patrick Bolton, Xavier Giroud, Harry Mamaysky, and myself, and he currently is at Imperial College Business School in London.

I want to begin by noting that Milton Friedman's classic analysis a half century ago of this question is often misunderstood.2 Friedman did not call for unbridled pursuit of current profitability, as is often portrayed, but rather for value maximization subject to full compliance with laws and regulations. In the world Friedman considered, democratically elected politicians decide the constraints under which firms operate. That is, we decide together, as a group.

Specifically, our elected officials and the regulators they appoint set pollution standards to hold firms accountable for polluting the environment. They also set labor standards to protect workers (which are enforced by the various federal agencies). And they set rules on corporate governance practices—enforced in the United States mainly by the U.S. Securities and Exchange Commission (SEC)—to protect investors. Firms, according to Friedman, should abide by all these rules, but subject to the constraint of abiding by the law, they should maximize the value of their shares for their stockholders.

Notice two things about this formulation: First, Friedman does not say that firms should skirt the laws to create value for their stockholders, but rather, that they should comply with all laws fully. In this view, it is not up to firms to determine society's regulations, but rather it is up to our elected officials, and they are accountable to voters.

I would add that in this view, activists who lose their battles in our legislative houses are not able to impose those rules on firms by other means. In a democratic system of government, we believe (or at least, we hope) that democracy will achieve the best collectively determined outcome about environmental, labor, and corporate governance standards.

Second, Friedman focused on long-run profits, and hence value maximization, as the criterion managers should employ, not just current profit maximization. If a firm's CEO found a way to increase current profits while engaging in legal practices that nevertheless harmed its consumers or workers in a way that generally was viewed as egregious, that firm likely would find it harder to sell its product or hire its workers, and that would affect the firm's stock price adversely. Value maximization, therefore, already includes a dimension of self-interested ESG compliance beyond the law: employers who do things contrary to the interests of their workers and consumers that many people consider unfair (but that has not yet been codified as a regulatory violation) will not succeed in maximizing the value of their enterprises.

By making value maximization (within the confines of the law) the sole measure of CEO performance, we can hold CEOs to account. They can compete with each other and we potential stock investors have a clear metric for gauging their performance and for deciding when to fire them.

Under this rule, we also have a clear way to measure the performance of mutual or pension fund managers. If the shares under their management underperform the market (using observed stock returns, on a risk-adjusted basis, as our metric) then we will move our funds to some other fund.

And we have a clear way of evaluating our politicians' performance, too. If the environment is very dirty, or if workers are being badly abused, or if companies are able to use tricks to skirt their accountability, we know who to blame: our politicians, who pass legislation and appoint civil servants to write regulations. And we have a remedy: vote them out of office and elect people who will do a better job.

If you believe in accountability of agents (CEOs, fund managers, and politicians), and if one believes in democracy as the best means of determining laws, there is really no room for improvement on this model. I am not saying that our system—which combines democracy, the legal duty of management to maximize the interests of shareholders, and market competition among firms and among asset managers—works perfectly to achieve our objectives as a society. I am saying that there is no way I can see to improve upon it. Everyone is assigned clear objectives, which means it is easy to detect failure, and we the investors and voters have ultimate power—weighted equally as voters and weighted according to our stockholdings as investors—to make all our agents accountable.

Let me emphasize what I am not saying. I am not saying that you as a voter or investor must invest in a firm, or vote for a politician, just because your compatriots support them, if you don't like them. In particular, if you believe a firm is being unfair, you are free to dump your shares as a means of expressing your views, and as I already noted, if enough people agree with you, then firms following the Friedman standard of value maximization will have an incentive to take that into account.

And you are free to protest and raise the consciousness of others to get them to agree with you about unfair practices by a firm. But if others don't agree with you, then under the Friedman approach, you will have little power, and firms that are constrained by law to maximize shareholder value will not be able to respond to your wishes just because you are a powerful person. I would say that is a good thing.

Contrast that with stakeholder capitalism as it has evolved—that is, the view that corporations should not only maximize their value subject to complying with all laws, but that they should also be willing to do many things that reduce their value in pursuit of some other set of objectives related to good citizenship as it is defined by some group, some pension fund, some politician running for office, some newspaper, or some ESG rating agency.

If one is an ESG enthusiast, one should be able to answer three questions: First, exactly whose preferences should the stakeholder-maximizing firm consider, recognizing that the preferences of people are bound to conflict?

Second, what weights should the firm attach to each of the possible stakeholder objectives other than value maximization?

Third, without a clear delineation of which people, which specific preferences, and which weights are being pursued, how do we hold the CEO or the investment manager accountable for failure to achieve desirable outcomes?

There is a lot of hubbub about stakeholder capitalism, but I have not heard advocates even try to address these three fundamental questions. I believe the reason for that is that they don't have good answers to any of them. That is because it's not possible to come up with good answers to them.

With respect to the first question, once the CEO or fund manager has given up on democracy to determine the constraints on value maximization, there is no clear way to decide which objectives "society" is demanding of him or her.

Second, unless one is very self-confident about his or her ability as a central planner, and also willing to impose his or her own views on others, no one has any way of knowing what weights to attach to those many amorphous objectives. Should the firm sacrifice value mainly to make its workers happier (perhaps by paying them above-market wages) or should it sacrifice value by avoiding the use of fossil fuels?

Third, given that investors have no clear way of weighting or even measuring these actions credibly, it is impossible to define a standard that can be used to hold CEOs or ESG-minded asset managers accountable for achieving ESG objectives.

So, what do we do? We can reject this approach to governing firms and investing in stocks as incoherent. But some might object, hoping that perhaps we can rely on some wise sage who can answer these three difficult questions to advise us on how to measure ESG compliance so we can decide whether we are willing to tolerate the value-creating shortcomings of the CEO or asset manager because of an overall ESG achievement that more than compensates for it.

But specifically, what objectives does the wise sage use? And how does the wise sage decide whether a firm is achieving those amorphous objectives that he has decided are correct? Is it right to assume that ESG ratings agencies are wise sages who know the right set of objectives and the right weights to attach to them, even though they don't bother telling us what those are? And is it right for us to assume that they have the correct incentives to monitor corporate behavior closely to ensure that companies are adhering to that unspecified weighted objective function?

We now have some research—Dr. Ruoke Yang's recent dissertation, cited above]—that can help us answer those questions. Let me summarize his findings for you (all the facts cited below can be found in his PhD dissertation).

First, Yang collected data from regulatory agencies' actions, legal filings, and newspaper reports to create objective measures of each dimension of E, S, and G. For example, his measure of E using regulatory agencies' actions measures whether a firm is found to have violated an environmental regulation. His measure of lawsuits measures whether it has been sued for an environment regulation, and so on.

Second, Yang asked whether ratings of E, S, and G, or other rating agencies' similar measures, forecast the outcomes for each of these three categories (as measured by regulatory, legal, or news outcomes). He finds that in the case of S and G, ratings have zero forecasting ability for outcomes. Strangely, in the case of E, they have negative forecasting ability. When a firm receives an improvement in its E rating, that is a signal that it will soon be found to be in violation by a regulator, or be sued for polluting, or be the subject of a news scandal about polluting.

How can that be? Yang explains this outcome with a model, and he confirms the implications of that model using econometric analysis. His explanation is actually quite simple: When a firm knows it has committed an environmental violation, it invests more in public relations (known as greenwashing) to improve its E rating so that it will not be held accountable by investors or politicians for the bad news that is about to be revealed.

The distortions in ESG ratings produced by greenwashing are now being studied by the SEC, and Commissioner Hester Peirce recently spoke about the need to think through whether ESG ratings really capture useful information (more about her speech below). The firms that provide ESG ratings fall for these greenwashing public relations stunts because they (like all agents) lack incentives to provide sufficient effort to avoid doing so. Rating agencies are unaccountable for their failings because their ratings are a black box, based on ill-defined criteria, and thus investors have a hard time identifying failures, even when the ratings agencies disagree (as they often do). Ratings agencies, therefore, don't have strong incentives to stay on top of every firm's behavior, which is why greenwashing by firms works so well.

According to Yang, there are now more than $10 trillion dollars chasing amorphous ESG criteria, and he finds that the ratings do affect firm prices. Better ratings seem to cause stock prices to go up. Apparently, ESG-minded investors take the ratings seriously, even though Yang shows that they should not.

Now, you might conclude from this evidence that we should use Yang's objective ex post measures of outcomes for our ESG ratings rather than existing ratings. But it is very hard to move from an ability to show that the ratings agency emperor is naked to devising an alternative E, S, and G ratings scheme.

Yang, or anyone using his data, would not only have to define outcomes that are bad and good (which is quite hard to do), but he would also have to decide what weights to attach to them. Should being sued by someone or having a bad (but unproven) news report about your company be considered a failure by your company? And what weight should be attached to it, compared to greater value creation, or compared to other such "failings"?

As Hester Peirce said in her speech:

"... In our purportedly enlightened era, we pin scarlet letters on allegedly offending corporations without bothering much about facts and circumstances and seemingly without caring about the unwarranted harm such labeling can engender. After all, naming and shaming corporate villains is fun, trendy, and profitable."

"E, S, and G tend to travel in a pack these days, which makes it hard to establish reliable metrics for affixing scarlet letters. ... Not only is it difficult to define what should be included in ESG, but, once you do, it is difficult to figure out how to measure success or failure."

"There is ... a growing group of self-identified ESG experts that produce ESG ratings. ESG scorers come in many varieties, but it is a lucrative business for the successful ones. The business is a good one because the nature of ESG is so amorphous and the demand for metrics is so strong. ESG is broad enough to mean just about anything to anyone. The ambiguity and breadth of ESG allows ESG experts great latitude to impose their own judgments, which may be rooted in nothing at all other than their own preferences. ... Putting aside the analysis that produces the final score, some ESG scores are grounded in inaccurate information."

"Some scorecard producers attempt to get information from the companies directly by submitting surveys to companies, the responses to which are then used to rate the ESG risk of the companies surveyed." 3

In other words, companies can manage their public relations through press releases to substantially influence their ratings. They also can do concrete things to buy an improvement in their ratings—such as when oil companies give many millions of dollars to universities in support of research on energy and the environment.

One might say, "well, in rating companies, can't we at least take into account regulatory actions against firms, if not news stories or lawsuits?" But what weights should the Yang ratings based on regulatory actions attach to each regulatory lapse by a firm? Isn't that sort of weighting precisely what our regulatory agencies are supposed to be doing when they set penalties for violations, and aren't their fines and other penalties already an expression of our collective beliefs about the social costs of those violations and our collective will about how to weight the importance of each kind of violation? If those fines and criminal penalties are a proper expression of our collective will, then we have already altered corporate incentives optimally to comply with the law. If they are not a proper expression of that collective will, then shouldn't we focus our efforts on changing the regulations and the regulatory penalties? Either way, we don't need arbitrarily constructed ESG ratings to guide us.

Yang's research makes plain that the existing status quo of ESG stakeholder investing is much worse than the Friedman value maximization rule from every perspective. Well, almost every perspective. There are some clear winners. CEOs can avoid accountability for shortcomings in value maximization due to their inability or lack of effort by pointing to some other ESG objective that they are pursuing. And then the same CEO can invest in greenwashing to get ratings agencies to rate his or her firm highly, and then point to that rating as evidence of his or her achievement. Asset managers that follow the ratings agencies can also quantify their "success" in terms of the ratings of their portfolio firms and excuse themselves for underperforming the index on that basis. Note that even though the research shows that the purported ESG achievements of the CEO and the asset manager are illusory, or maybe even negative on the E component, Yang finds that stock market investors still are rewarding the CEOs and asset managers for their "good social behavior" with higher stock prices.

So, our society achieves nothing worthwhile on ESG, and perhaps even encourages harm to the environment behavior, but there is a big additional cost. Incompetent or lazy CEOs and asset managers are able to stay in their jobs, leading us to make poor decisions that waste resources.

And that's not all. Our politicians also skirt accountability, substituting speeches about stakeholder capitalism for debating specific legislation to establish appropriate regulatory penalties. And voters may be taken in by this, rewarding politicians for speeches instead of asking them to do their jobs by debating specific legislation. Politicians, after all, face their own agency problem. Voting on legislation setting specific rules and penalties means that they will offend some voters, and that also means that they will lose financial support from some as a consequence. Virtue signaling in speeches is much less risky for an elected official than voting on bills.

We need to pull away the curtain and expose the ESG stakeholder game for what it is: an approach that encourages all the key agents on whom we rely—CEOs, asset managers, rating agencies, and politicians—to avoid accountability in the name of amorphous objectives and virtue signaling. We need to restore accountability by insisting on value maximization under the law as the sole criterion by which we measure the achievements of CEOs and asset managers, and on the virtue of democracy as the appropriate means for aggregating our preferences to ensure that corporations behave responsibly. Failing to do so means that we will increasingly reward economic inefficiency while we also increasingly fail to achieve bona fide social objectives in the E, S, and G areas.

It would help us to get there if we thought more deeply about the risks of accepting the opinions of experts when they are passing judgment on others. (This is a lesson that has broader application than ESG ratings, by the way.) As Commissioner Peirce said:

... A stark [Scarlet] letter A does not always serve to convey the truth. The moral authorities of today, like their puritanical forebears, are motivated by a dream of a better society, but methods matter and so do facts. We ought to be wary of shrill cries from a crowd of self-appointed, self-righteous authorities, even when all they are crying for is a label.

Scarlet letters are produced by someone. They have a purpose. Some gain from their existence. Our society does not.

1 Ruoke Yang, Three Essays on Corporate Social Responsibility, PhD Dissertation, Columbia University, 2019. A revised version of his research is available at

2 Milton Friedman, "The Social Responsibility of Business Is to Increase Its Profits," The New York Times Magazine, 1970.

3 Hester M. Peirce, "Scarlet Letters: Remarks Before the American Enterprise Institute," June 18, 2019. Accessible at


Charles W. Calomiris