Stakeholder Theory: Definition, the Three Justifications, Criticisms

Stakeholder theory holds that an organisation should be managed in the interests of everyone who can affect or is affected by the achievement of its objectives, rather than in the interests of its shareholders alone. The formulation is R. Edward Freeman's, from Strategic Management: A Stakeholder Approach in 1984, and the definition of a stakeholder as anyone who "can affect or is affected by" the firm is the sentence most often quoted from it.

It is worth noticing at the outset that the definition is doing two jobs at once, and they pull apart. "Can affect" is a claim about power and belongs to a theory of how firms survive. "Is affected by" is a claim about moral standing and belongs to a theory of what firms owe. Most of the disagreement in this literature traces back to which half is being used.

The three strands

Donaldson and Preston's 1995 paper is the standard way of organising the field, and it separates three theories that share a vocabulary.

Descriptive. A claim about how firms actually behave: that managers do in practice attend to employees, customers, suppliers and communities, and that the firm can be accurately described as a set of relationships among these groups. This strand is tested by observation and is largely uncontroversial.

Instrumental. A claim about consequences: that firms which manage stakeholder relationships well perform better on conventional financial measures. This strand is testable, has attracted a great deal of empirical work, and produces results that are positive but modest and sensitive to how the variables are operationalised.

Normative. A claim about obligation: that stakeholders have interests of intrinsic worth, and that the firm owes them consideration regardless of whether doing so improves returns. This is the core of the theory in Donaldson and Preston's reading, and it is the strand that cannot be settled by evidence.

The distinction matters because the three are frequently run together in a way that lets an argument slide. A writer who defends the normative claim and then cites financial performance to support it has changed theories mid-argument: if the justification is that it pays, then the obligation disappears whenever it stops paying.

Who counts as a stakeholder

The broad definition is criticised for including almost everyone, and the literature has responded with narrower schemes.

Primary and secondary. Clarkson distinguishes stakeholders without whose participation the firm cannot survive — shareholders, employees, customers, suppliers, and the public groups supplying infrastructure and legal standing — from those who affect or are affected but are not essential to survival, such as the media and campaign groups.

Normative and derivative. Phillips separates those to whom the firm has a direct moral obligation from those who matter only through their capacity to affect the first group. A competitor is not owed anything, but a competitor's actions may change what the firm owes its employees.

Salience. Mitchell, Agle and Wood classify by three attributes: power to influence, legitimacy of the claim, and urgency. Holding one attribute makes a group latent; two makes it expectant; all three makes it definitive and, in their account, the group management will actually attend to. The scheme's value is that it predicts managerial attention rather than prescribing it, and it explains why a legitimate but powerless and unhurried claim is routinely ignored.

The case against, and how strong it is

Friedman's objection is the one every assignment expects. In the 1970 essay usually cited, the argument is that a manager is an agent of the owners, that spending corporate funds on social purposes is spending other people's money on the manager's own preferences, and that such decisions are properly made through the political process by people accountable for them.

This is stronger than the caricature. It is not a claim that social outcomes do not matter; it is a claim about who has standing to make the trade-off and to whom they answer. Any serious defence of stakeholder theory has to address the accountability question instead of the caricature.

The multiple-principals objection follows from it. A manager accountable to everyone is accountable to no one, because conflicting interests give a decision-maker latitude to justify almost any choice as serving some stakeholder. Jensen's enlightened value maximisation is the attempted answer: use long-run firm value as the single objective, while recognising that it cannot be achieved without attending to stakeholders. Whether that resolves the problem or restates the shareholder position in gentler language is a live argument.

The weighting problem. The theory identifies whose interests count and offers little guidance on how to trade them off when they conflict, which is when guidance is actually needed. Closing a loss-making plant serves shareholders and harms employees and a community; the theory says all three count, and stops.

The measurement problem. Instrumental claims require the stakeholder orientation to be measured, and the available proxies — corporate social responsibility ratings, disclosure indices — measure reporting at least as much as behaviour. A firm that reports well and behaves badly scores well.

What the theory replaced

Understanding why the position gained ground requires knowing what it was arguing against, and it was not simply greed.

The shareholder primacy model has a coherent defence that goes beyond ownership. Shareholders are residual claimants: everyone else — employees, suppliers, lenders, the revenue authority — has a contractual or statutory claim that is paid first, and shareholders receive whatever is left. On that reading, maximising the residual is the same thing as ensuring every prior claim can be met, and the shareholder is simply the party whose interest aligns with the health of the whole.

The stakeholder response is that the premise has weakened. Employees whose skills are firm-specific, suppliers who have invested in dedicated capacity, and communities dependent on a single plant all bear residual risk that no contract covers. If more than one group is exposed to the residual, the argument for giving one of them exclusive claim to the decision loses its foundation.

That exchange is the core of the debate, and it is more productive to write about than the moral framing, because both sides are making a claim about risk and not about virtue.

The instrumental evidence, stated carefully

The empirical literature is large, and summarising it as "stakeholder management pays" overstates what it shows.

Meta-analyses generally find a small positive association between corporate social performance and financial performance. The association is statistically reliable across many studies and economically modest in each.

Three caveats follow. Causation is contested, and the reverse direction is plausible: profitable firms can afford stakeholder investment, so profitability may be producing the social performance and not the other way round. The measures capture reporting quality substantially, so a firm's disclosure practice contaminates the independent variable. And the average conceals wide variation, with the relationship stronger in consumer-facing industries where reputation converts to revenue, and weak or absent in business-to-business commodity sectors.

For an assignment the usable version is that the instrumental case is supported but weak, and that resting the whole argument on it makes the obligation contingent on evidence that could go the other way.

Where it has had practical effect

Whatever its theoretical status, the position has changed practice in three visible ways.

Company law in the United Kingdom carries it in section 172 of the Companies Act 2006, which requires a director to promote the success of the company for the benefit of the members while having regard to a listed set of stakeholder interests. The duty is owed to the company and the stakeholder interests are considerations rather than claims, which is a recognisably stakeholder-flavoured compromise inside a shareholder framework.

Corporate reporting has moved further, with integrated reporting frameworks and, in Europe, mandatory sustainability reporting that requires disclosure of impacts on people and environment alongside financial results.

None of these settles the priority question, and it is worth being explicit about that. Section 172 makes stakeholder interests something a director must have regard to while pursuing a purpose defined by reference to the members; it does not make them claims a stakeholder could enforce. The compromise is procedural rather than substantive, and critics on both sides say so for opposite reasons.

Governance codes now routinely require boards to describe how they have engaged with the workforce and other groups, which converts a normative claim into a procedural obligation without settling the underlying question of priority.

Writing about it well

Three habits distinguish a strong answer.

Say which strand is in play. An argument that a firm should treat suppliers fairly because it reduces supply risk is instrumental; one that says it should regardless is normative. Marking schemes reward the distinction and penalise sliding between them.

Where a named framework is asked for, use salience instead of the bare definition, because it distinguishes groups that will be attended to from groups that merely have a claim.

Take Friedman at his strongest. Answers that dismiss the shareholder position as greed score poorly. The accountability argument is serious and the reply has to engage with it.

Name the conflict the case actually contains, and say how it should be resolved and on what principle. Stakeholder theory is at its weakest on exactly this point, and an answer that shows where the theory runs out demonstrates more than one that applies it smoothly to a case with no real trade-off in it.