1. THE CORE COMMITMENTS OF SHAREHOLDER PRIMACY

2. NORMATIVE MANAGERS AND PRINCIPALS

A manager has know-how, competencies, and expertise that make her particularly well-suited to administer the business; principals are seldom as qualified as she is (Kriegstein Reference Kriegstein2016, 453).^{Footnote 7} The same is true about other types of professionals that act as fiduciaries of principals. Doctors and lawyers have expertise that patients and clients typically lack; it is because of this expertise that patients and clients need to hire doctors and lawyers. When doctors and lawyers make decisions on behalf of a principal, they are required to act in the principal’s interest. Decisions (especially the most important ones) are usually discussed with the patient or client. He is typically offered information about the benefits and risks of the different potential alternatives available to him, information that is supposed to help him make an informed decision about what he considers best. When it is not possible to get direct input from him, doctors and lawyers have to make the decisions on his behalf, trying to put themselves in the shoes of the patient/client so that they act as he would have chosen them to act. The agent, in sum, should conduct the affairs of the principal as the principal would have wanted her to.

Because the manager is institutionally separated from shareholders and seldom has direct contact with them, she is in the position faced by a doctor/lawyer who lacks access to the direct input of her patient/client. This suggests the following guiding managerial directive: “Is this what shareholders would want the manager to do? Is this what shareholders would do if they were directly administrating the business and had the information, know-how, expertise, and competencies that the manager has?” (cf. Hart and Zingales Reference Hart and Zingales2017a, 263).

When articulating the moral obligations of managers at the most general level of analysis, this directive is keyed in a fully normative register. From a fully normative perspective, the idea that managers are required to do what shareholders would want the manager to do should not be construed as a question of what actual shareholders would effectively want managers to do. Scholars have tended to miss this (Bower and Paine Reference Bower and Paine2017, 55, 59; Hansmann and Kraakman Reference Hansmann, Kraakman, Rasheed and Yoshikawa2012, 36; Hasnas Reference Hasnas1998, 22; Kriegstein Reference Kriegstein2016, 457; McMahon Reference McMahon1981, 258–9; Hart and Zingales Reference Hart and Zingales2017a, 248; Rodin Reference Rodin2005, 171, 176). If one fails to recognize this, one ends up committed to the view that if shareholders are indolent and want the manager to pursue unfair business practices, the manager is supposedly obligated to follow suit.

This is mistaken. The morally inflected question that we are addressing—how should the manager conduct the business she is entrusted to administrate?—is a full-blown normative question. As such, it is not an empirical question about how shareholders, as a matter of fact, want the manager to conduct the business (Heath Reference Heath2014, 32–3). It is a normative question that involves thinking of shareholders in normative terms; that is to say, assuming that shareholders are moral and, therefore, that what they want passes moral muster. Thus, articulating a normative theory of business that fleshes out the responsibilities that a manager has towards shareholders requires us to think of “shareholder” and “manager” as embodying normative standards. In particular, it requires thinking that they are moral agents who abide by moral norms.^{Footnote 8} In what follows I will assume that shareholders and managers are moral agents that live up to what morality requires of them.

This leads to what one may call the normative directive that guides the manager’s decisions:

A corollary follows: the fiduciary obligations that an agent has to the principals do not give the agent moral immunity to pursue immoral activities; fiduciary obligations do not override the principal’s moral obligations or license behavior by the agent that would be prohibited to the principal (Goodpaster Reference Goodpaster1991; Kriegstein Reference Kriegstein2016). It is natural to infer from this that agents inherit all the obligations of principals. A number of scholars in business ethics have actually made that inference (Goodpaster Reference Goodpaster1991; Hsieh Reference Hsieh2004; Kriegstein Reference Kriegstein2016; Mansell Reference Mansell2013; Ohreen and Petry Reference Ohreen and Petry2012). As we will see, this inference is mistaken because there are discretionary obligations that do not carry over from principals to their agent. To see this, however, we need to first clarify the nature of nonelective and discretionary obligations.

3. THE DISTINCTION BETWEEN NONELECTIVE AND DISCRETIONARY OBLIGATIONS

At the heart of the distinction between nonelective and discretionary obligations is the fact that moral obligations are not all mandatory in the same way (Herman Reference Herman1993; Hill Reference Hill1971; Kant Reference Kant and Gregor1998; Rainbolt Reference Rainbolt2000; Schneewind Reference Schneewind1990; Schroeder Reference Schroeder2014; Stohr Reference Stohr2011). There are obligations, such as the obligation not to murder an innocent person, that impose demands that do not afford latitude and require compliance on every occasion. Other obligations, such as the obligation to aid those in need, allow for latitude about when and how to discharge them. We are always required to avoid murdering an innocent, but we are not required to always help a person in need. This latitude has to do with, among other things, the fact that how one decides to fulfill a discretionary obligation can depend on one’s own sensibility and personal inclinations.

I will refer to duties that impose demands that don’t allow for latitude and always require compliance as nonelective obligations. I will refer to obligations that allow for latitude about when and how to discharge them as discretionary obligations. The labels nonelective and discretionary are meant to single out one single criterion at play in the more complex distinction between perfect and imperfect duties.^{Footnote 9} I use these labels instead of the more familiar perfect/imperfect labels to indicate that the distinction at play here is only related to this single aspect of the perfect/imperfect distinction and to signal to readers that the distinction between nonelective and discretionary obligations should be unhinged from the philosophical baggage associated with the distinction between perfect/imperfect duties, in particular from the Kantian theoretical apparatus with which these are typically associated.

There is, in fact, nothing specifically Kantian about the distinction between nonelective and discretionary obligations. The idea that not all obligations allow for the same degree of latitude is a pretheoretic intuition that runs deep within our moral landscape. Denying it amounts to affirming that all obligations are always mandatory or forbidden, a view that imposes too stringent demands on us. The case of beneficence illustrates it. If it were demanded that we comply with the obligation to be beneficent on every occasion, we would be required to help those in need every time we could. Because it is almost always possible to help someone in need, this would require us to devote all of our spare time to pursuing beneficent actions. However, it is natural (and legitimate) to deny this. Arguably, we are not contravening the duty of beneficence by enjoying an afternoon with friends, even if we could be spending this time contributing to worthy humanitarian causes. And this is so, not merely because doing so allows us to replenish our vigor to pursue our charitable work, but because “there is such a thing as having done enough when it comes to beneficence” (Stohr Reference Stohr2011, 52; cf. Hill Reference Hill2002, 201). Part of the force of the intuition that guides the distinction between nonelective and discretionary obligations is that it is permissible to pursue self-interested projects, that even though we should help those in need, we don’t need to always be devoted to serving others or to contributing to society.^{Footnote 10}

One might object that the idea of a discretionary obligation is an oxymoron. An obligation that is discretionary, one might say, is not an obligation. This worry is mistaken. The objection fails to recognize that even though discretionary duties allow for latitude, they are obligatory.^{Footnote 11} That they allow for latitude does not entail that they are optional. We have an obligation to help those in need, this obligation is mandatory, and everyone is bound by it. But while this obligation is obligatory in this way, it does not require us to help every single person who is in need, nor to help those in need every time we have the chance to do so; this obligation allows discretion about how and when to help.^{Footnote 12}

4. MAXIMAL VERSIONS OF SHAREHOLDER PRIMACY GO WRONG WITH DISCRETIONARY OBLIGATIONS

It is natural, but mistaken, to think that all obligations that bind shareholders are inherited by the manager who serves as their agent. Goodpaster provides a concise articulation of this rationale:

As I will now show, this argument, suggested by a number of business ethicists (Arnold Reference Arnold2003; Goodpaster Reference Goodpaster1991; Kriegstein Reference Kriegstein2016; Mansell Reference Mansell2013; Ohreen and Petry Reference Ohreen and Petry2012), is flawed. The fact that all shareholders share a particular obligation, interest, or need does not mean that the manager is obligated to discharge this obligation, serve that interest, or fulfill this need, on their behalf. A fiduciary relationship is a “triadic relation” (Marcoux Reference Marcoux2003, 3). It involves two parties (an agent and a principal) and a project (in this case, the business). Managers are agents of shareholders only in a circumscribed domain: the administration of the business (Heath Reference Heath2014, 32). It follows that the obligations that carry over from shareholders to managers are only those that bind shareholders as part of their participation in the business (Kriegstein Reference Kriegstein2016).

Most of the scholars who have defended maximal versions of shareholder primacy have tried to infer the responsibilities of managers from the responsibilities that shareholders have as human beings (Goodpaster Reference Goodpaster1991; Mansell Reference Mansell2013; Ohreen and Petry Reference Ohreen and Petry2012). This is mistaken. As Kriegstein (2016, 446) points out, the moral obligations that apply to shareholders need not coincide with the moral duties that bind human beings in nonbusiness contexts. Shareholders’ obligations take place in a competitive environment determined by the ethics of adversarial relationships, and this context will typically impose a more limited set of obligations than those that apply to human beings in general. Only these more limited sets of obligations would potentially carry over from shareholders to managers.

What most defenders of maximal versions of shareholder primacy get right is that shareholders’ participation in a (successful) business enterprise imposes a duty of beneficence on them. Arguably, because shareholders increase their wealth through their investment in the business, we have a moral expectation that they would share some of these gains with those who are less fortunate. This expectation, however, arises not merely from the fact that they are human beings, but from the fact that they are increasing wealth through their joint enterprise.

But the fact that all shareholders are bound by this obligation does not mean that it carries over to the manger. Defenders of maximal versions of shareholder primacy fail to realize that for managers to be required to fulfill shareholders’ obligations on their behalf, these obligations are such that (moral) shareholders want the manager to fulfill them on their behalf. If it is in the interest of (moral) shareholders to discharge certain obligations by themselves, and it is morally permissible for them to do so, the manager should not attempt to discharge these on their behalf.

The above considerations can be summed up in the following two conditions that have to be met for any obligation to roll over from principals to agents.

With these two conditions, as I will proceed to show, managers are not, in general, obligated to fulfill instances of the wide duty of beneficence (to which I will refer to as “charity”) on shareholders’ behalf. Two reasons explain this. The first has to do with the way in which discretionary duties allow us to express our individuality. In allowing for latitude, the duty of charity provides us with a space to exercise what Rainbolt has aptly called “moral freedom” (Rainbolt Reference Rainbolt2000; cf. Lea Reference Lea2004, 210–1). How and who we decide to help provides us with an avenue to express ourselves morally. Rainbolt illustrates this with familial and amical obligations, such as the obligation to play with one’s child or to show one’s wife that one loves her:

From this perspective, how a shareholder might want to share his proceeds with those in need appears to be a personal matter that should be out of the reach of managerial responsibilities.^{Footnote 13} In other words, it seems that an important part of a discretionary obligation is that it allows us to fulfill it in our own personal way.

A second reason why managers should not, in general, be tasked with fulfilling wide duties of charity is that shareholders might have personal differences that lead them to fulfill the duty of charity differently. These differences will be grounded in both their individual sensibilities and their personal circumstances. Some shareholders might be more inclined to contribute to organizations that help certain dispossessed communities, while others might want to contribute to threatened minorities or at-risk populations. There are also differences in their personal circumstances that lead to variations in the proportion that each shareholder should contribute to charity. Some shareholders might be very wealthy and are morally obligated to make substantial donations to charity; other shareholders might be struggling financially and might not be required to make any financial contribution (Lea Reference Lea2004, 214–5). Given these wide divergences, it will be difficult, if not impossible, for the manager to discharge shareholders’ obligation to fulfill wide duties of charity on their behalf. But because the duty of charity is wide, and because it does not require the manager to discharge it in any specific way, it is a duty that can be separated from the money-making activities of the business (Hart and Zingales Reference Hart and Zingales2017a, Reference Hart and Zingales2017b). Because of this, instead of using corporate resources to engage in charity, managers should transfer these resources to shareholders and allow each of them to fulfill their wide duty of charity directly, as is fit for each of them.

A reader might be tempted to provide a third justification for why managers should not discharge discretionary obligations on behalf of shareholders, namely, that managers lack the qualifications to do it (Friedman Reference Friedman1970, 122; Rodin Reference Rodin2005, 178). “What qualifies a manager to administrate a business,” this reader might argue, “are things like her understanding of the risks and challenges in the industry, her ability to anticipate the financial needs of the company, and her capacity to lead teams and inspire others. Knowing how to engage in charitable giving to best help those in need, however, is not part of a manager’s job description.” This justification is fallacious; it begs the question. It is predicated on an antecedent conception of the manager’s responsibilities, which is precisely what we are trying to clarify. If my argument showed that managers are obligated to fulfill discretionary obligations, then this would just show that this reader’s understanding of the scope of managerial responsibilities is mistaken and that discharging discretionary obligations should be part of a manager’s job description.

A few caveats are in order. First, if shareholders (or at least a sufficient majority of them) want the manager to discharge a discretionary obligation that binds them in a specific way, then the manager will be typically obligated to do it. There are, for instance, companies whose statutes codify the charitable commitments of shareholders, stipulating that the company should devote a certain proportion of their proceeds to certain charitable causes; Target, for instance, commits 5 percent of its profits for this purpose (Target 2012). Absent this kind of stipulation, however, one would not expect a consensus among shareholders about how to discharge their duty of charity given their wide variety of inclinations and diverse personal circumstances. In these cases, we have a prima facie reason to think that managers should avoid fulfilling these obligations on their behalf.

Second, if making charitable donations is a strategic managerial decision that contributes to increasing the company’s bottom line (and most of the charitable contributions of current companies are of this sort), then the manager is typically obligated to make these donations. Note, however, that her obligation to get involved with charity arises not because she is obligated to discharge discretionary obligations on shareholders’ behalf, but because she is obligated to increase shareholders’ value (cf. Rodin Reference Rodin2005, 175). As such, charity becomes a means to achieve an end: increasing shareholder value. Because it is a means, there is a clear way to determine the charitable deeds that the manager should pursue: those that would make more money for shareholders. Pursuing charity in this way, however, should not be conceptualized as fulfilling a discretionary duty of beneficence.

Third, I have argued that, within shareholder primacy, managers are not required to fulfill certain discretionary obligations such as the wide duty of charity. I haven’t shown, however, that this result generalizes to all discretionary obligations. Although it is beyond the limits of the article to explore this, there may be narrow duties of beneficence, associated with the particular competencies of the company or its specific relationships with certain groups of people, that are discretionary and that managers may be required to fulfill on behalf of shareholders.

5. MINIMAL VERSIONS OF SHAREHOLDER PRIMACY GO WRONG WITH NONELECTIVE OBLIGATIONS

6. ENCOMPASSING VERSIONS OF SHAREHOLDER PRIMACY

I have argued that minimal and maximal versions of shareholder primacy are both untenable. The former do not impose enough obligations on managers; the latter impose too many. Both Heath (Reference Heath2014) and Hsieh (Reference Hsieh2004, Reference Hsieh2009, Reference Hsieh2017b) recognize that the obligations imposed on managers are more encompassing than minimal versions of shareholder primacy recognize, and both of them agree that these obligations are not as encompassing as maximal versions of shareholders primacy suggest (cf. Rodin Reference Rodin2005). I will conclude the article by discussing whether either of their accounts instantiates an encompassing version of shareholder primacy.

Let me start by focusing on Hsieh, who proposes to ground managerial obligations in the “duty not to harm.” While this duty can be shown to ground obligations that are more encompassing than what minimal versions of shareholder theory recommend and not as encompassing as what maximal versions propose, it is nevertheless inadequate to ground an encompassing version of shareholder primacy. The main reason is that in some cases this duty is too demanding and in others it is insufficiently demanding. There are two types of reasons why it is too demanding. First, in failing to differentiate the magnitude of the harm, it is unable to differentiate between major harms that are forbidden and minor harms that may be morally permissible. Second, within shareholder primacy there are plenty of examples where managers are allowed (and even encouraged) to engage in morally permissible harms of significant magnitude. For example, they are allowed to put competitors out of business by developing better products at lower prices, leave a group of employees out of jobs because a plant is relocated to a different location, or leave a group of suppliers in the lurch due to a change in the strategic orientation of the firm. The moral legitimacy of these actions may actually be very strong (for instance, if the new plant is moved to a location that serves a historically disenfranchised group or if the change in strategy is a response to an existential threat to the company).

But the duty not to harm is also not sufficiently demanding because it cannot ground all obligations that bind shareholders qua shareholders. Arguably, not all instances of nonelective duties of reparation, foresight, fidelity, and justice can be formulated in terms of a duty not to harm. Fulfilling one’s promises and seeking fair arrangements in the workplace are instances of the duty of fidelity and justice. But what grounds these duties is not always the harm that may be inflicted by the failure to discharge them. By breaking a promise or a contract you may wrong (but not harm) the other party (Rodin Reference Rodin2005, 167). To assume that “harm” can be used as a universal token to assess all obligations fails to recognize the multiplicity of ways in which our moral obligations are grounded.

Hsieh connects the idea of grounding managerial duties in the duty not to harm with the idea that “basic principles of ordinary morality . . . provide an adequate basis for specifying the responsibilities of business managers” (Hsieh Reference Hsieh2017b, 293). Our basic principles of morality, however, involve negative and positive duties. Why, one may wonder, should one expect one single negative duty, the duty not to harm, to ground all negative and positive nonelective duties? For instance, there are promising arguments suggesting that there are positive duties of rescue that are nonelective (Merck’s donation of Mectizan comes to mind). A powerful justification for these (nonelective) positive duties is not that the company’s managers have to repair the harm from which shareholders have profited, but that the company has a unique competency to alleviate a significant human need without incurring significant costs to shareholders (Donaldson Reference Donaldson1992; Dunfee Reference Dunfee2006).^{Footnote 15} Thus, the duty not to harm is neither necessary nor sufficient to ground the moral obligations within encompassing versions of shareholder primacy. In certain cases, this duty imposes some obligations that are too demanding, and in other cases, it imposes obligations that are not sufficiently demanding.

Let me now discuss Heath’s account. The fact that he identifies himself as a “tinged” shareholder theorist (Heath Reference Heath2006, 551) may suggest that his account instantiates an encompassing version of shareholder primacy. However, the fact that the duties favored by his account appear to line up with the duties that an encompassing version of shareholder primacy would recommend betrays an important methodological difference. One of the central points defended in this article is that, because of the agency relationship, to identify many of the obligations imposed on managers one needs to start by looking at the obligations that would bind shareholders if they were directly managing the business. However, as Kriegstein (2016, 446) has shown, Heath attempts to infer the managerial obligations directly from the implicit morality of the market, failing to recognize that “the obligations formulated by the market failures approach apply in the first instance to shareholders, rather than to managers, for it is shareholders who are seeking to further their own self-interest via market competition” (446). This suggests that Heath’s approach, as it stands, cannot be identified with an encompassing version of shareholder primacy.

With a few minor tweaks, however, Heath’s account can be modified to instantiate an encompassing version of shareholder primacy. To be an encompassing version of shareholder primacy, his account would need to, first, identify the obligations that bind shareholders qua shareholders. Kriegstein (Reference Kriegstein2016) has argued that it is possible to interpret Heath’s account as applying to shareholders first. For this revised account to be an encompassing (instead of a maximal) version of shareholders primacy, it would also need to examine which of these obligations carry over to managers. In doing so, it would recognize, as Kriegstein does not, that not all discretionary obligations that bind shareholders carry over to the manager. A revised Heathian account that would accommodate these two methodological points would instantiate an encompassing version of shareholder primacy.

This revised Heathian account, however, is but one potential version of an enhanced notion of shareholder primacy. Different business ethics scholars might have reservations with it. Some may argue that shareholders who benefit from investing in a business enterprise should not merely avoid profiting from market failures but also from “justice failures” (Singer Reference Singer2018). Others might argue that the emphasis on efficiency conflicts with other moral duties that shareholders have and which might take priority (McMahon Reference McMahon1981), moral duties that one might simply recognize as “the decent thing to do” (Goodpaster Reference Goodpaster1991, 65). It is also possible to propose encompassing versions of shareholder primacy that do not rely on efficiency or Paretian considerations. For instance, one might try to develop an account of the moral obligations that bind shareholders by relying on, say, Kantian (Bowie Reference Bowie1999) or Aristotelian (Hartman Reference Hartman2013) principles.

7. PRACTICAL IMPLICATIONS AND FUTURE DIRECTIONS

Encompassing versions of shareholder primacy provide a framework that is able to accommodate a family of versions of shareholder primacy. These versions recognize that to identify the main responsibilities of managers one should 1) identify the obligations that bind shareholders qua shareholders and 2) examine which of these obligations carry over to their managers. I showed that when we identify managerial obligations in this way, we see that managers are required to fulfill all nonelective obligations that are imposed on shareholders, but they are not always required to fulfill (and are sometimes actually forbidden from fulfilling) discretionary obligations that bind shareholders. From this it follows that both minimal and maximal versions of shareholder primacy are untenable: the former do not impose enough obligations on managers and the latter impose too many. The alternative to these two versions of shareholder primacy, however, is not a single version of shareholder primacy, but a framework that is able to accommodate a family of versions of shareholder primacy.

My analysis in this article has focused on the obligations that are imposed on managers in virtue of the principal-agent relationship. Nothing in what I have said, however, precludes the possibility of providing considerations external to the agent-principal relationship to modify the scope of these managerial obligations. For instance, as an anonymous reviewer put it, there may be a “professional ethos specific to the context of management that is independent of any obligations shareholders may have.” This managerial role may come with an independent set of obligations that are not derived from the fact that managers are agents of shareholders, and these obligations may be more (or less) demanding than the obligations imposed by the principal-agent relationship (cf. Strudler Reference Strudler2017). To illustrate with another example, one may appeal to notions of corporate agency to argue that the managerial obligations are more (or less) encompassing than those entailed by the principal-agent relationship (cf. Hsieh Reference Hsieh, Orts and Smith2017a).

One interesting corollary from my argument pertains to debates about corporate agency. Scholars who attribute agency to firms have tended to think of businesses as agents with the same moral requirements as ordinary human beings. The main thesis I defended in this article—that managers and, by proxy, the businesses they administer, are only constrained by some of the obligations that bind human beings in general—means that if we think of businesses as moral agents, and if we think that the responsibilities of managers to conduct the businesses are those that emerge in virtue of being agents of shareholders, then it follows that businesses do not have the same ethical obligations as ordinary human beings. In other words, my article provides a prima facie argument that businesses and human beings are different kinds of moral agents and, consequently, that the conscience of the corporation is not “a logical and moral extension of the consciences of its principals” (Goodpaster Reference Goodpaster1991, 68; cf. Goodpaster and Matthews Reference Goodpaster and Matthews1982).

As I have demonstrated, because any tenable version of shareholder primacy has to be encompassing, any account that criticizes shareholder primacy as promoting unethical corporate behavior is targeting what is ultimately a straw man. Shareholder primacy, properly understood, does not sanction reckless profit-maximization. It is not guilty of what some stakeholder theorists have called the “separation fallacy” (Freeman et al. Reference Freeman, Harrison, Wicks, Parmar and de Colle2010, 6), i.e., the notion that managers can keep ethics and business separate. The fact that, within shareholder primacy, businesses are required to fulfill nonelective obligations implies that business decisions cannot be divorced from ethical questions. Because managers cannot pursue corporate actions that are forbidden by nonelective obligations, they have the obligation to incorporate an evaluation of this ethical dimension in every one of their business decisions.

Shareholder-centric models have become widespread as shown by changes in the attitudes of managers, recent modifications in corporate governance, and the growing ability of activist investors to influence managers’ actions (Bower and Paine Reference Bower and Paine2017, 55–57). Shareholder primacy has also become pervasive in the business literature and business education, particularly in the areas of economics, finance, and management (Heath, Moriarty, and Norman Reference Heath, Moriarty and Norman2010, 445; Hart and Zingales Reference Hart and Zingales2017a, 248), and has been characterized as “the dominant ideology, or global normative standard” (Hansmann and Kraakman Reference Hansmann and Kraakman2001, 46; cf. Hansmann and Kraakman Reference Hansmann, Kraakman, Rasheed and Yoshikawa2012; Hart and Zingales Reference Hart and Zingales2017a).

As a consequence, and as many in the literature have recognized, it has become almost axiomatic for shareholders, financial advisers, boards of directors, and company executives to conceive of managers as agents of shareholders (Bower and Paine Reference Bower and Paine2017; Ghoshal Reference Ghoshal2005; Hansmann and Kraakman Reference Hansmann and Kraakman2001, Reference Hansmann, Kraakman, Rasheed and Yoshikawa2012; Hart and Zingales Reference Hart and Zingales2017a). Practitioners, however, have tended to limit the focus of their attention on the managerial obligation to maximize profits for shareholders, often short-term profits, failing to recognize the demanding ethical limits that the principal-agent relationship imposes on such maximization. Business ethicists have often addressed this profit-oriented bias by proposing alternative models of corporate governance, a strategy that has had limited success because of practitioners’ deeply ingrained conviction that managers are agents of shareholders. Unlike external critiques of shareholder primacy that propose alternative models of corporate governance or promote prosocial behavior by firms, my argument is more powerful rhetorically in that it offers an internal critique of a model that these business practitioners, lawyers, and scholars embrace, a critique that shows that their own endorsement of shareholder primacy commits them to recognize a more robust set of ethical obligations than they typically acknowledge.

Shareholders, as the principals in whose interest the business is administered, need to recognize that they are not only the beneficiaries of the activities of the business but that they also bear the ultimate responsibility for any moral wrongdoing committed on their behalf. As such, shareholders should be more vocal with managers about the importance of conforming to a more encompassing set of moral restrictions. This requirement might appear to fly in the face of the generally accepted fact that the general public should be allowed to buy shares in companies, despite lacking much expertise on such companies or their industries and without requiring much involvement in their part. What I am suggesting here is not that individual shareholders should be vocal with managers about the specific obligations that they should fulfill on their daily operations. My claim is that shareholders should be vocal about the fact that their financial returns should be constrained by any nonelective obligation imposed on managers via the agency relationship. Of course, given that managers are equipped with the expertise to run the business, expertise that shareholders typically lack, it behooves managers to figure out the specific moral limits of their actions and decisions. Managers should recognize that the moral limits to their external business activities exceed those proposed by minimal versions of shareholder primacy. But they should count on the support of shareholders who would look favorably at lower rates of return when this is the result of fulfilling the nonelective obligations that bind them. Boards of directors should be more conscious that part of their own responsibilities consist of ensuring that managers are complying with these moral obligations. Financial advisers, in turn, should recognize and make more transparent to clients that buying company stock is not simply betting money on a particular company, but owning a share of the residual earnings, earnings which should be limited by a more encompassing set of obligations than previously acknowledged. Finally, lawmakers and regulators should seek institutional reforms on corporate governance that support the view that “managers are accountable to owners not only for the financial performance of the corporation but also for the minimal and maximal moral character of its activities” (Rodin Reference Rodin2005, 178) and that provide clearer avenues for shareholders to communicate their moral demands and preferences to shareholders (Rodin Reference Rodin2005, 179; Hart and Zingales Reference Hart and Zingales2017a, 258, 264).

Some scholars might have the intuition that this is not part of what managers should do, that this should not be part of their job description (Friedman Reference Friedman1970, 122; Rodin Reference Rodin2005, 178). However, a reflection on the principal-agent relationship shows that such intuition is mistaken; as agents of shareholders, managers are required to fulfill an encompassing set of moral obligations on their behalf. If the skeptic pushes back saying that this is not what business managers are trained to do, one can respond in two complementary ways. First, one should insist that this is a mistake, that business managers should be sufficiently conversant with ethics to identify their most important moral responsibilities. Second, one could argue that, as with any other aspect of the company’s administration, managers need not know everything about its management. They can (and should) hire officers who have the competence for dealing with this specific area, officers who are capable of identifying difficult moral issues and offer counsel and recommendations about how to deal with them. Regulated by a more encompassing set of corporate laws, guided by more explicit moral directives in corporate statues and shareholder communications, pressured by boards of directors to comply with these moral requirements, and helped by corporate officers to identify, assess, and take action with respect to encompassing obligations, managers would be well positioned to fulfill the moral demands imposed by the principal-agent relationship.