Is “stakeholder capitalism” real?
In 2019, 181 CEOs declared shareholder primacy over. Five years later, the question is whether anything behind the press release was ever real.
The watershed that wasn’t (yet?)
“Each of our stakeholders is essential. We commit to deliver value to all of them, for the future success of our companies, our communities and our country.”
— Business Roundtable, Statement on the Purpose of a Corporation, signed by 181 CEOs, August 19, 2019
Jamie Dimon, Jeff Bezos, Tim Cook, and 178 other chief executives put their names to it. The press called it the end of the Friedman era — the moment American capitalism stopped serving shareholders alone. Klaus Schwab built the 2020 Davos meeting around “stakeholder capitalism”; Larry Fink’s annual BlackRock letters turned ESG into a boardroom expectation. The watershed was declared. The question this whole walkthrough asks: was it real?
Underneath the announcement sits an old argument about what a corporation is for, and there are two clean answers. The residual-claimant view: a corporation is property. Shareholders own it, they bear the residual risk after everyone else is paid by contract, so managers owe their duty to them. Maximize shareholder value and you have done your job. The nexus-of-stakeholders view: a corporation is not a thing anyone owns but a coalition — customers, employees, suppliers, communities, and shareholders all contribute and all have a claim. On this reading, serving shareholders alone is a category error.
That is the genuine dispute the 2019 statement waded into, and naming the two readings is all you need here — the next two stages argue each one at full strength. But hold one distinction in reserve, because it will decide everything: a statement of purpose is not the same as a mechanism of accountability. A corporation can declare any purpose it likes. Whether that purpose binds anyone depends on who can hold a manager to account for breaching it — and that is a different question entirely.
“While each of our individual companies serves its own corporate purpose, we share a fundamental commitment to all of our stakeholders.”
— Business Roundtable Statement on the Purpose of a Corporation, August 2019
“Corporate purpose officially changed in 2019.”
The watershed claim: 181 CEOs abandoned shareholder primacy, and the era of “the business of business is business” is over. Did the signatures change what corporations do, or only what they say?
A turning point — or a talking point?
“This new statement better reflects the way corporations can and should operate today. It affirms the essential role corporations can play in improving our society when CEOs are truly committed to meeting the needs of all stakeholders.”
— Alex Gorsky, then-chairman & CEO of Johnson & Johnson, on the 2019 Business Roundtable statement
The proponent reads the moment as overdue honesty. Corporations have always affected more than their shareholders — the workers whose lives they shape, the towns built around their plants, the air their factories use. Saying so out loud, at the highest level, is the first step toward governing as if it were true. The signatures are a commitment that can be invoked against the firms that signed.
“It’s largely a rebranding… corporate leaders are genuinely worried about the legitimacy of big business. This statement, by accountability standards, is mostly a PR move.”
— the early-skeptic reaction, voiced by governance scholars within weeks of the 2019 statement
The skeptic asks the only question that matters in August 2019: what, concretely, will be different on Monday? No board resolution, no changed metric, no new party empowered to sue if the promise is broken. If a commitment costs nothing to make and nothing to break, the most economical hypothesis is that it was made for how it reads, not for what it binds. The full case waits for Stage 3 — but the doubt is already here.
Where this leaves us
A press release is not a constitution. One hundred eighty-one CEOs signed a statement — but a statement of purpose is only as real as the accountability mechanism behind it, and at this stage we have seen no mechanism. Before we decide whether they meant it, though, we owe the proponents the question they were answering: is pure shareholder primacy actually wrong? Because if it is, the stakeholder turn is a correction the economics demands — and if it isn’t, it is just talk. The next two stages take those two questions in order.
Start with the strongest version of the case those CEOs were making. Suppose you take the stakeholder argument completely seriously — not as PR, but as economics. Does it hold up?
The case for stakeholder capitalism
“The corporation is best understood not as a bundle of assets that belongs to shareholders, but as a team of people who bring together complementary, often firm-specific, investments in a joint enterprise.”
— after Margaret Blair & Lynn Stout, “A Team Production Theory of Corporate Law”, Virginia Law Review, 1999
This is the case at its strongest — not a slogan but a claim grounded in the economics of externalities and incomplete contracts. Take it seriously: if the corporation really is a coalition of people who each put something at risk in it, then serving only one of them is not just unfair. It is bad economics.
Here is the argument the proponents are entitled to, at full strength. Pure shareholder primacy is genuinely first-best efficient — but only under two conditions: every contract is complete, and every externality is priced. Neither holds in the real world, and once they fail, three things go wrong.
1. Unpriced externalities. A firm maximizing shareholder value will produce exactly the pollution, the systemic risk, and the social harm it is not charged for. That is not managerial malice; it is the objective function working as designed. Where externalities are unpriced, “maximize shareholder value” and “maximize social value” come apart.
2. Under-protected stakeholder-specific investments. Contracts are incomplete, so people who sink firm-specific investments bear hold-up risk a contract can’t fully cover. A worker who trains in skills useful only at this plant, a town whose tax base is one factory, a supplier who re-tools for a single buyer — each has put real capital at risk in the enterprise, and each can be expropriated when the firm later changes terms. A governance rule that answers only to shareholders treats those investments as if they don’t exist.
3. Myopia and long-term value. When capital markets are short-sighted, the things that build durable value — R&D, employee development, brand and social-license capital — get under-supplied, because their payoff lands past the horizon the share price rewards. The team-production view ties the three together: the firm is a coalition of specific-asset contributors, and a rule serving one contributor under-protects the others’ stakes. None of this is heterodox. Incomplete-contracts theory and externality theory are mainstream microeconomics, and they say plainly that pure shareholder primacy is not first-best.
This reading has a long pedigree. The institutional tradition — from Berle and Means’ 1932 diagnosis of the separation of ownership and control, through the mid-century managerialists, to Blair and Stout’s team-production theory — has always read the corporation as a structure of power and coalition rather than a simple object of ownership. That lineage is the proponent’s intellectual home; see History of Economic Thought Ch.15 §15.3 (the institutional theory of the firm).
“The purpose of business is to produce profitable solutions to the problems of people and planet, and not to profit from producing problems.”
— Colin Mayer, Prosperity: Better Business Makes the Greater Good, 2018
“Serving all stakeholders is long-term value creation.”
The strong proponent claim: the firm is a coalition of stakeholders, and looking after all of them is not charity that costs shareholders — it is how durable value gets built. Is the diagnosis right?
A real coalition — or an unenforceable promise?
“Public companies are not simply the private property of shareholders. They are entities that draw on a wide range of stakeholders, and shareholder value alone is a dangerously narrow measure of the value a company creates.”
— the team-production / institutional case, after Blair & Stout and Colin Mayer’s Prosperity
The proponent stands on the institutional tradition that runs from Berle and Means through the managerialists to team-production theory — the long argument that the corporation is a coalition of specific-asset contributors, not a shareholder’s possession (the lineage lives in History of Economic Thought Ch.15 §15.3). The point is not sentimental. It is that a firm which protects its workers’, suppliers’, and communities’ investments builds the trust and specificity that make the enterprise valuable in the first place.
“The diagnosis is right and the prescription is unenforceable. Show me the mechanism — the metric, the principal, the remedy — or you have described a problem, not built a solution.”
— the measured skeptic, who grants the diagnosis and asks for the mechanism
This is not a rout — the skeptic concedes the externalities, the hold-up, the myopia. The objection is narrower and harder: a governance rule needs a who and a how. Whom does a manager answer to when the duty to “all stakeholders” is breached, and by what measure was it breached? Until that is answered, the coalition has a real grievance and no enforceable claim. That is the bridge into Stage 3.
Where this leaves us
The case for is correct as a diagnosis. Pure shareholder primacy is not first-best — externalities, incomplete contracts, and stakeholder-specific investments are real, and a doctrine that pretends the corporation affects only its shareholders is incomplete economics. But there is a chasm between “pure shareholder primacy is wrong” and “stakeholder capitalism is right.” The first is about a critique. The second is about a system. And a system needs something the diagnosis hasn’t supplied: a mechanism that says who can hold a manager to account, and how. The valid kernel is real. Whether the movement built on it is a different question — one this walkthrough keeps strictly separate from its converse, whether the shareholder status quo is itself broken (forthcoming).
So pure shareholder primacy is wrong. Does that make stakeholder capitalism right? Here is the question the proponents keep not answering: if a manager owes a duty to everyone — customers, workers, suppliers, communities, shareholders — who, exactly, can hold that manager to account when the duty is breached? And what if the answer is: no one?
The skeptic’s case
“In a free-enterprise, private-property system, a corporate executive is an employee of the owners of the business. He has direct responsibility to his employers… What does it mean to say that he has a ‘social responsibility’? Insofar as his actions reduce returns to stockholders, he is spending their money.”
— Milton Friedman, “The Social Responsibility of Business Is to Increase Its Profits”, New York Times Magazine, September 13, 1970
Fifty years before 181 CEOs declared shareholder primacy over, Milton Friedman asked the question that still hasn’t been answered: when a manager spends the company’s money on “society,” whose money is it, and who said they could? And in 2020 a pair of corporate-law scholars went looking for evidence that the 2019 promise had changed anything. What they found is this stage’s hammer.
Friedman’s point is usually read as ideology, but the durable version is structural. Shareholder primacy is the standard solution to the agency problem. Managers have private information and private incentives; left unchecked they pursue their own comfort, empire, and pay. The discipline that holds them in line is a single accountability metric — shareholder value — tied to a single residual-claimant principal who can actually act: shareholders vote, sue, sell, and replace the board. The metric is crude, but it is enforceable.
Now replace it with “balance all stakeholders.” The manager is handed a metric for everything — customers, workers, suppliers, communities, shareholders — and therefore a binding constraint from no one. A duty to all is an enforceable duty to none: diffuse obligations have no metric and no enforcer, and the absence of an enforcement mechanism is not a detail. It is the whole question of whether something is a “system” at all. This structural prediction generates two testable consequences. First, the doctrine would change rhetoric far more than behavior, because nothing forces behavior to move. Second, it would entrench managers: “I was balancing stakeholders” is an all-purpose defense against accountability to anyone — not a cynical aside but a prediction about who benefits.
The evidence agrees with the structure. Lucian Bebchuk and Roberto Tallarita surveyed the 2019 signatories. The companies’ own boards, in the great majority of cases, never approved the statement — the CEOs signed it personally. The corporate governance guidelines were not amended. No fiduciary duties changed. The statement was, on the evidence, a public-relations document, not a governance commitment. The wider ESG literature finds the same pattern at scale: far more relabeling than behavior change. The structural prediction was confirmed.
This skeptic argument has its own lineage. Friedman’s 1970 essay and the Chicago program that followed — Jensen and Meckling’s agency theory, the shareholder-value-maximization doctrine — sit in the counter-revolution that remade economics from the 1970s on; see History of Economic Thought Ch.10 §10.1 (Friedman and the Chicago counter-revolution). No equation decides this one. It is a claim about who can act when a promise is broken, and the answer for “all stakeholders” is: no one in particular.
“Stakeholderism would impose substantial costs… by inducing and reinforcing managerial insulation from shareholder accountability. The Business Roundtable statement was largely a rhetorical public-relations move rather than the harbinger of a major change.”
— Lucian Bebchuk & Roberto Tallarita, “The Illusory Promise of Stakeholder Governance”, Cornell Law Review, 2020
“A duty to all stakeholders is an enforceable duty to none.”
The strong skeptic claim: without a single metric and a single principal who can act, “balance everyone’s interests” is not a governance system — and the BRT signatories proved it by changing their press releases and not their boards.
Whose money, and who enforces?
“We find that the boards of the overwhelming majority of the signatory companies did not even discuss joining the Statement. This indicates that joining was largely a public-relations move rather than a considered governance decision.”
— Lucian Bebchuk & Roberto Tallarita, Cornell Law Review, 2020
This is the empirical hammer, and it lands on the structural prediction. Friedman’s “whose money?” was a question about accountability; the Chicago program that formalized it — Jensen and Meckling’s agency theory — turned shareholder value into the one metric that disciplines managers (the lineage sits in History of Economic Thought Ch.10 §10.1). The prediction was that a diffuse duty would move talk and not behavior. Bebchuk and Tallarita went and checked. The boards never voted. The behavior never moved.
“Accountability mechanisms are being built — stakeholder-metric-linked pay, mandatory climate disclosure, benefit corporations, B-Corp certification. The system is young, not absent. Judge it the way you’d judge any institution still under construction.”
— the proponent’s strongest reply to the accountability charge
The proponent’s reply is not a strawman — these innovations are real, and some have changed real behavior at the margin. The skeptic’s answer is precise rather than dismissive: every one of these works by installing a specific enforceable metric — a tonne of carbon, a chartered duty, a certified standard. That is the opposite of “balance all stakeholders.” It proves the point the skeptic was making: where stakeholder governance becomes real, it does so by ceasing to be diffuse. The kernel survives; the diffuse mandate does not.
Where this leaves us
The skeptics are right about the system exactly where the proponents were right about the diagnosis. A duty to everyone, with no metric and no enforcer, is not a governance system — it is a sentence, and the evidence agrees with the structure: the BRT signatories changed their press releases and not their boards. But notice what the skeptics don’t get to claim — that Friedman was simply right. The accountability argument demolishes stakeholder capitalism as a system; it does nothing to rescue pure shareholder primacy from the externalities and incomplete contracts Stage 2 established. Both sides won something. Neither won everything.
So the critique of pure shareholder primacy is valid, and stakeholder capitalism as a system is mostly not real. Five years after the watershed, the ESG wave is breaking against a political backlash and a wall of evidence. So what was it — a revolution, a fraud, or something in between?
The verdict — mostly rhetoric, with a real kernel
“The end of ESG.” By 2023, the three letters that ran Davos and rewrote BlackRock’s letters had become a culture-war epithet — banned from Texas and Florida pension mandates, dropped from the marketing decks that once flaunted them, and quietly walked back by some of the firms that signed in 2019.
— the 2022–2024 anti-ESG backlash, as it broke across politics and the financial press
The backlash arrived with as much force as the wave — and was just as over-stated. The right turned ESG into a banner of “woke capital”; the same executives who flaunted the label in 2019 stopped saying it. But a label deflating is not an argument winning. The celebration was an over-claim; so is the demolition. The honest answer sits between them, and it is uncomfortable for both wings.
Put the two stages together and the synthesis is clean. The valid kernel from Stage 2 — externalities, incomplete contracts, stakeholder-specific investments — is real and permanent. But look at what would actually fix each piece. Unpriced externalities are fixed by pricing them: carbon taxes, Pigouvian instruments, regulation. Hold-up of specific investments is fixed by contracts and co-determination: enforceable protections, worker board seats, specific remedies. Myopia is fixed by changing capital-market incentives. Each is a concrete, enforceable mechanism — and none of them is “publish a purpose statement.”
That is the whole verdict in one move. A real stakeholder-accountability system would have to install specific, enforceable stakeholder metrics — which is mostly the work of policy and contracts, not of a corporate-purpose declaration. The diffuse “balance everyone” mandate doesn’t address the kernel; the concrete mechanisms do, and where the movement has produced anything real, it has produced concrete mechanisms wearing the stakeholder label. The label was mostly rhetoric. The kernel was always a job for externality pricing, regulation, and stakeholder-specific contracts.
Seen from the history of the discipline, the 2010s corporate-purpose debate is one front of a broader post-2008 pluralist turn — the same moment that revived institutional critique, put inequality and market power back on the agenda, and reopened questions the counter-revolution had closed; see History of Economic Thought Ch.17 §17.5 (the post-2008 pluralist frontier). The stakeholder revival rode that wave in. The wave was real; the specific vehicle was mostly branding.
“The end of ESG” — the 2022–2024 verdict, declared with the same confidence as the 2019 watershed, and just as over-stated.
— the anti-ESG backlash, as both political slogan and financial-press headline
“Stakeholder capitalism: rhetoric with a real kernel.”
The two-tier verdict: the critique of pure shareholder primacy is valid, but the governance system built in its name is mostly not real. An answer that is uncomfortable for both political wings — which is the tell that it is honest.
A real kernel — or always mostly PR?
“Stakeholder-metric-linked pay, benefit corporations, B-Corp certification, real climate disclosure — these are accountability innovations, and the externality critique is permanent. The label may fade; the substance is being built.”
— the measured proponent, on why there is genuinely something here
This is the strongest version of “there is something here.” The proponent is right that the critique is permanent and that a minority of firms has built real, enforceable stakeholder metrics. The kernel is not a rhetorical concession — it is the part of the verdict that says “mostly,” not “entirely.” Where stakeholder governance is real, it is real because it became specific.
“The deflation was predictable, because the label never bound. What survived is the handful of specific, enforceable instruments — everything diffuse evaporated the moment it was inconvenient. That is what a mostly-PR doctrine looks like under stress.”
— the measured skeptic, situating ESG inside the post-2008 pluralist turn
The skeptic places the whole episode in context: the 2010s corporate-purpose debate was one front of the broader post-2008 pluralist revival that reopened institutional and inequality questions (see History of Economic Thought Ch.17 §17.5). The revival was real and lasting. The specific stakeholder-capitalism vehicle was mostly branding, and the proof is what remained when the branding stopped paying: the enforceable instruments, and nothing diffuse.
The verdict
The critique of pure shareholder primacy is valid — externalities, incomplete contracts, and stakeholder-specific investments are real, and Friedman’s framing under-weights them. But “stakeholder capitalism” as currently practiced is mostly not a real accountability system — diffuse obligations to everyone have no enforcement mechanism, the BRT signatories didn’t change behavior, and it functions largely as PR, at worst as managerial cover. The honest answer: rhetoric with a real kernel. The deflation of the label doesn’t vindicate Friedman, and the validity of the critique doesn’t vindicate the movement — and the actual fix for the valid kernel is concrete mechanisms (externality pricing, regulation, stakeholder-specific contracts), not a purpose statement.
Where this leaves us
We started with 181 CEOs declaring shareholder primacy over and the press calling it the end of the Friedman era. The arc bent twice. First, the case for turned out to be right as a diagnosis: pure shareholder primacy is not first-best, because externalities are unpriced, contracts are incomplete, and stakeholder-specific investments are exposed — mainstream microeconomics, not a slogan. Then the skeptic’s case turned out to be right about the system: a duty to all stakeholders, with no metric and no enforcer, is a duty to none — and Bebchuk and Tallarita found the structural prediction confirmed in the signatories’ own boardrooms, which never voted and never changed. Watershed, valid diagnosis, empty system, two-tier verdict.
So the honest answer refuses to flatten. The critique of pure shareholder primacy is valid; “stakeholder capitalism” as a governance system is mostly not real — rhetoric with a real kernel. Hold the two apart and the verdict is uncomfortable for both political wings, which is the tell that it is honest rather than partisan. And notice the practical consequence: the valid kernel is a job for externality pricing, regulation, and stakeholder-specific contracts — concrete, enforceable mechanisms — not for a purpose declaration that binds no one. This walkthrough asked whether the proposed alternative is real. Its converse half — whether the shareholder status quo is itself a problem (short-termism, externality-dumping, buyback-driven underinvestment at depth) — is a separate walkthrough, “Is shareholder primacy a problem?” (forthcoming). Neither is the whole corporate-purpose debate alone; together they are.