Oliver Hart and Luigi Zingales have proposed a revision to the dominant model of the objective of the firm, most famously defended by Milton Friedman, arguing that executives’ obligation is not to maximize shareholder value, but shareholder welfare. Their proposal hopes to be more ethically attuned than Friedman’s. As it stands, however, it falls short from an ethical perspective, argues Fordham professor Santiago Mejia. 


Oliver Hart and Luigi Zingales, two of the most influential contemporary economists, have proposed a revision to the dominant model of the objective of the firm, most famously defended by Milton Friedman. Their motivation for this proposed revision was their conviction that Friedman’s model is not sufficiently responsive to ethical and social values. 

Hart and Zingales propose that executives’ obligation is not to maximize shareholder value (i.e., profits), as Friedman proposed, but shareholder welfare. While their proposal is promising, it falls short from an ethical perspective. We need to recognize that executives should only satisfy the preferences of shareholders when they are ethically permissible 

By saying that a preference is ethically permissible, I mean to say that it conforms with the ethical norms that prescribe how we should behave and interact with others. For instance, ethical norms prohibit profiting from selling drugs to children or promoting lies that cause disinformation, polarization, and violence. Some may think that these norms are the product of social agreements, others may think that they are Platonic realities that we ought to discover. The important point is that ethical norms establish what is prohibited and permissible, are meant to apply to everyone, and are supposed to be justifiable with sound arguments. 

Amending the Hart-Zingales Model 

The amendment I am proposing—namely, that executives should only maximize the ethically permissible preferences of shareholders—would not be necessary if shareholders’ preferences and values were ethically permissible. However, it is mistaken to assume that this will always be the case. Some shareholders may be immoral or their self-interest may cloud their ethical judgments. 

Allow me to illustrate with an example by Hart and Zingales themselves. In their 2017 paper “Companies Should Maximize Shareholder Welfare Not Market Value,” they use a hypothetical example to illustrate that the executive of a tobacco company should decide whether to sell tobacco to children by polling investors on whether they want to profit by selling tobacco to children. This example is ethically problematic. 

We allow adults to smoke, despite the health hazards involved, because we consider them equipped to make decisions concerning the harms and benefits of tobacco. Children, however, are not adults. We don’t consider them to be well-positioned to assess the risks and benefits of smoking. Smoking also causes more developmental damage to children, and its effects are often more lasting than in adults. These arguments explain why it is ethically objectionable to sell tobacco to children. 

It follows from this that if shareholders were to sell tobacco to children directly, they would be breaching an ethical prohibition. Shareholders who want the executive to do it are asking her to do on their behalf what ethics prohibits them from doing directly. But this is wrongheaded: an agent is not allowed to behave in ways that would be forbidden to the principal. 

It is frequently emphasized by economists that the executive’s fiduciary obligation is to pursue the interest of shareholders. However, the executive’s fiduciary obligations also require her to discharge the principals’ ethical obligations. And that entails that the executive should not satisfy preferences of shareholders that are not ethically permissible. 

Two related objections may arise at this point. First, if most people recognize that this activity is wrong, won’t shareholders, a subset of these people, recognize it? Not necessarily. Some shareholders may not care much about ethics. Others may claim to care, but persuade themselves, through self-serving and specious arguments, that selling tobacco to children is legitimate. 

One may also object: if most people people recognize that this activity is wrong, they will be less likely to do business with a company promoting it, forcing the company to change its behavior. In a few cases, such as the 1965-1970 Delano grape strike, stakeholder pressure may force a company to improve its ethical behavior. But to think that the market will always serve as an ethical check is naive. First, it is very difficult for stakeholders to be fully informed about the ethical or unethical activities of a company. Second, savvy companies are often able to manipulate the reputational risk, either by conducting aggressive PR campaigns or spinning off the company’s risky divisions. Third, if a stakeholder values something that the company offers (say, if a consumer likes a particular cigarette brand) their ethical qualms may not be enough to deter them from doing business with the company. 

“Satisfying the legally permissible preferences of shareholders is not the same as satisfying their ethically permissible preferences.”

Law and Ethics 

It may be suggested that the concerns I raised above could be addressed by requiring executives to satisfy only the preferences of shareholders that are legal. According to this line of argument, ensuring that the preferences of shareholders are legal would ensure that managers would not sell tobacco to children. 

However, this proposal will not do. Satisfying the legally permissible preferences of shareholders is not the same as satisfying their ethically permissible preferences. 

As Hart and Zingales recognize, legal norms are the product of messy political processes that sometimes lead to imperfect laws that don’t track what ethics would permit. In addition, the law “is a blunt instrument” that is ill suited to regulate the entirety of our moral world. For instance, ethical norms prohibit most forms of deliberate deception, yet democratic countries do not penalize many deliberate lies because that would squash free speech

Think of a social media company whose major voting shareholders push executives to increase “user engagement” (i.e., the monetizable time that a user spends in the platform) through the use of algorithms. Higher levels of engagement are usually best promoted by false posts that trigger strong negative emotions. The systematic promotion of such posts has increased the spread of misinformation, worsened political polarization, and has played a key instrumental role in massacres and forced migrations. 

Ethical norms prohibit individuals from profiting from the use of such socially toxic algorithms. Yet this prohibition does not disappear when individuals nominate an agent to act on their behalf—it carries over from individuals (shareholders) to their agents (executives). 

However, while ethical norms prohibit managers from profiting from the use of these toxic algorithms, the law allows them to do so. The two reasons I mentioned explain it: first, the company’s lobbying power may have kept reforms at bay; second, there are laws (such as Section 230 of the Communications Decency Act) which, for the sake of protecting free speech online, don’t impose legal penalties on platforms for how they distribute users’ content. 

Implementing the Amendment 

One may think that the amendment I am proposing is all well and good in theory, but problematic in practice: Who in the company should be given the authority to decide which of the preferences of shareholders are ethically permissible? And how should this decision be determined? 

The first question is easy to respond: the board of directors (and, through it, the executive) are the agents directly appointed by shareholders. As their agents, it is their responsibility to promote shareholders’ interests within the bounds of their moral obligations. 

Of course, the preferences of shareholders have a prima facie validity. Moreover, executives and boards should not make decisions about the permissibility of shareholders’ preferences by appealing to their gut feelings or personal values. These decisions should be based on ethical norms that are independently justified and rationally grounded. As in most cases of institutional decision-making, these norms should be accessed through corporate procedures that involve a variety of corporate actors. 

The details may vary across companies and industries, but here are some general suggestions. Boards of directors should keep firmly in mind that it is among their duties to ensure that executives (and their subordinates) fulfill the ethical obligations they inherit from shareholders. This should be reflected in the hiring, promotion, and compensation mechanisms that incentivize decision procedures that assess such obligations. Because employees have privileged access into the internal workings of the company, there should be mechanisms allowing them to raise concerns and red flags. The company should also promote and support independent industry organizations tasked with recognizing the main ethical risks and complexities of the industry and providing general guidance about how companies should deal with them. 

Companies, especially large ones, should also have committees tasked with providing advice to executives and employees about the ethical dimensions of their corporate decisions. These committees (unlike many current ethics committees) should have real voice and power. They should also include a variety of external voices to ensure that they take into account a diversity of perspectives and avoid the self-serving biases and group-thinking that cloud ethical decision-making within homogeneous groups. 

“Accepting the preferences of shareholders as given, without assessing whether they are ethically permissible, will lead to an unethical theory of the firm.”

Intractable and Tractable Ethical Differences 

Some may be skeptical of the success of these proposed procedures, given the high degree of variability in people’s ethical perspectives. For instance, abortion is highly contentious, and reasonable people have legitimate and well-founded disagreements about it. To the extent that different (often opposite) positions are well justified, reasoned debate may help to understand the complexity of the issue, but not necessarily to find common ground to resolve it. 

I agree that there are such cases. And in these cases, executives and their advisers will not be in a position to assess and determine the ethical permissibility of one position over the other. They may have their own views on the matter, of course, but they ought to recognize that they have no authority to make decisions on these issues on behalf of shareholders. In these cases, the executive should defer to shareholders, as Hart and Zingales suggest. However, they should do this, as Hart and Zingales also recognize, only when these contentious issues are inseparable from the money-making activities of the company. 

The variability in people’s ethical views, however, should not be taken too far. 

There is, for instance, almost universal agreement that lying, stealing, and harming others is wrong. It is also uncontroversial that companies should fulfill their promises, respect the rights and dignity of their stakeholders, and abide by laws that reflect legitimate political processes. A variety of widely accepted ethical prohibitions follow from this. Among them, that it is wrong to profit from selling tobacco to children or from promoting lies that cause disinformation, polarization, and violence. 

Of course, to claim that there is wide agreement in large swaths of our ethical landscape is not to deny that ethical decision-making is sometimes complex and messy. Nearly all ethical norms have exceptions, and contexts matter a great deal in identifying the relevant ethical principles that apply. For instance, even though we expect full transparency from auditors and accountants, we don’t expect this level of truthfulness from salespeople. 

How should executives deal with these more messy ethical issues? Through consistent and transparent ethical decision-making that is grounded on reasoned justifications. Making decisions based on consistent and reasoned justifications is a basic constraint on the legitimacy of ethical decision-making, and it helps to establish a clear and relatively independent framework that avoids haphazard or arbitrary decisions. Being transparent about such decisions ensures that all the stakeholders (including current and potential shareholders) know and understand the ethical principles that guide the company. 

Reasoned justifications are often the upshot of reasoned debates that consider the variety of aspects at stake in these messy ethical issues. As I suggested above, such debates are particularly effective when they incorporate a variety of voices. When carried in a spirit of collaboration and openness, they often lead to important and significant areas of consensus. Of course, these decisions will seldom please everybody, but this is as true for ethical decisions as it is for almost any corporate decision that involves uncertainty. 

Concluding Thoughts 

There is nothing unethical about seeking profit or pursuing one’s self-interest. Provided, of course, that one abides by one’s ethical obligations and responsibilities. When studying agency relationships, economists have been particularly worried about the fact that agents may fail to keep their end of the bargain. Part of what I tried to do here is to highlight that principals also fail to keep their end of the bargain when they want executives to engage in activities that are ethically objectionable. When this is the case, the executive is in no obligation to follow through. In fact, she has an obligation to resist. 

Ethical issues are messy and their application not always straightforward. But the difficulties of implementing a policy should not blind us about its desirability. Accepting the preferences of shareholders as given, without assessing whether they are ethically permissible, will lead to an unethical theory of the firm. Governance principles need to recognize that if executives are required to satisfy the preferences of shareholders, they should only satisfy those that are ethically permissible

Author’s note: I want to acknowledge several people who provided valuable input for this blog post: Pietro Bonaldi, Tom Donaldson, Oliver Hart, Joseph Heath, Jose Fernando Jimenez, Ganesh Mejia Ospina, Jeff Moriarty, N. K. Chidambaran, Mark Packard, Asher Schechter, and Luigi Zingales. All mistakes and errors should be attributed to me.