Burned at the Stake: Stakeholder Theory and Shareholder Interests Don’t Line Up
Persons have a natural right to go into business together, and they may do so either as a partnership or a corporation. In the latter type of organization, the persons setting it up limit their civil liability (I won’t be concerned in this article with the vast and complex issues raised by this). The people who set up a corporation are the owners, and they hold stock in it. Modern corporations normally consist also of managers, including the CEO, and a board of directors, but all of these are agents of the shareholders and act in their interest. I have spoken so far of people’s moral rights, but until recently, at least, the legal system in the Anglo-American world corresponded in general to the scheme I have set out. A school of thought called “stakeholder theory” challenges this, claiming variously that the managers and board of directors should not be restricted to acting as the agents of the owners but should take into account also others who have a “stake” in the corporation. In this week’s article, I’d like to consider some objections to this view and also to hold up to ridicule some of the weak arguments advanced in its support.
An obvious criticism of stakeholder theory is that the key notion of the theory is undefined. Who counts as a “stakeholder”? If the term means anyone whose activities are affected by the corporation, this includes a vast array of people. Competitors of the corporation will be adversely affected by its success; do they have a “stake” in preventing this? Are labor unions that sponsor crippling strikes against it also stakeholders? What about environmental activists who aim to put the company out of business?
In her book Corporate Governance (Institute of Economic Affairs, 1998), Elaine Sternberg expresses this criticism with great clarity:
Given the divergent interests of the different stakeholder groups, that which benefits one group will often harm another. Higher wages for employees can mean higher prices for customers and/or lower returns for shareholders, cleaner emissions into the environment may mean harder work for employees and loss of market share for traditional suppliers. What weight is to be given to these conflicting interests? … Stakeholder theory does not indicate which of these benefits is to be preferred, or how conflicting interests are to be balanced. Are stakeholder interests all strictly equal? Are some more important than others? If so, which are they? And when, and by how much, and why? Stakeholder theory gives no clue as to how to rank or reconcile the normally conflicting interests of stakeholders. (p. 97)
R. Edward Freeman, one of the pioneers in the development of stakeholder theory, Robert Phillips, and Rajendra Sisodia have an answer to this criticism that you aren’t going to believe, even after you read it; but I assure you that the following quotation is accurate. In their journal article “Tensions in Stakeholder Theory” (Business and Society, 2018), Freeman, Phillips, and Sisodia say:
The real issue is not shareholder versus stakeholder but a narrow/reductionist versus broad/holistic perspective on business. It is the difference between a value chain (linear and singularly focused on financial value) and a value network (which includes the importance of shared purpose and values). A value chain has one end point and one desired outcome for one stakeholder, the shareholder; all other players in the system are a means to that ultimate end. In an interconnected and interdependent system, each stakeholder must be a means and an end. Each contributes to collective flourishing and each must also benefit for the system to continue flourishing. Every business is a system, embedded within a set of larger systems. (p. 217)
How can they possibly think that this piffle addresses the problem? Doesn’t their claim of harmony just restate the outcome he needs to prove? I am afraid matters get worse. Freeman, Phillips, and Sisodia reject exact definition of terms; you must instead grope toward a solution in practice in particular cases. So the pragmatists have taught us. “Stakeholder theory is about ‘knowing how’ to engage stakeholders and create value for them, rather than the technical ‘knowing that’ such and such is the case for all times for all problems for all configurations of stakeholders” (p. 97). But in what way is this a solution to the problem? Why should we think that the “knowing how” to which Freeman appeals results in harmony, unless he defines harmony as whatever arrangement results in each case from the jostling of the various interest groups? Why should we assume that all the stakeholders aim at harmony? Does being part of the “network” assure this?
If you ask questions like this, Freeman, Phillips, and Sisodia think you are adopting the discredited notion of “essences.”
Is “stakeholder theory” really a theory or is it merely a perspective? If it is a theory, what are its essential arguments? If it is not a theory, then what is it and which problems does it/can it address? This question has plagued stakeholder theory almost from its inception. But the tension has always been more apparent than real. The framing suggests that being a “theory” creates some privileged position … and that arguments from such a position have immutable “essences.” (p. 218)
But maybe stakeholder theory has a better answer to our question than Freeman’s oozing verbiage. In asking how the interests of the stockholders can be reconciled with the conflicting interests of the stakeholders who don’t own stock, we are assuming that these interests conflict and have to be reconciled. What if that assumption is false? What if promoting the interests of the shareholders requires promoting the interests of the stakeholders and vice versa? Then our question is solved, or better, dissolved.
Tom Donaldson, another leading supporter of stakeholder theory, adopts just this approach, and he supports it through a peculiar argument. Alexei Marcoux in his important essay “Triadic Stakeholder Theory Revisited” explains Donaldson’s argument in this way. (He proceeds next to reject it, though with arguments different from the one I’m going to give.)
According to Donaldson, this is an implication of the principle ought implies can: if one ought to do as stakeholder theory’s normative thesis demands and one ought to satisfy one’s fiduciary duty to shareholders, then it must be the case that one can do both. One can do both only if managerial action satisfying stakeholder theory’s normative thesis is (or does not conflict with) managerial action satisfying fiduciary obligations to shareholders. Therefore, concludes Donaldson, there exists a relationship of mutual support between the normative and instrumental theses where it counts—in the psychology of the manager. (p. 276)
This is a gross abuse of the ought-implies-can principle. That principle says, I can’t be obligated to do something I’m unable to do. I can’t have an obligation, for example, to run a mile in four minutes (or in thirty minutes, for that matter) because I can’t do it. The principle does not license inferences of the form “If I ought to do something, and my doing so requires that the consequences of my doing so turn out a certain way, then it’s possible that they do turn out that way.” Philosophers have argued over whether you can derive an “ought” from an “is.” Donaldson is trying to deduce how it’s possible for the world to be from an obligation. He is saying that because stakeholder theory says that you ought to satisfy the interests of the shareholders and also that you ought to satisfy the interests of the stakeholders, then it’s possible that you can do both. He is arguing from “ought” to “possibly is,” and that is not legitimate.