Ronald Coase

In the [YOU] on AI Field Guide

The cycle that began with [YOU] on AI documents a twenty-fold productivity multiplier at one hundred dollars per person per month—a backend engineer building user interfaces, a designer writing complete features, each individual absorbing transactions that previously occurred between specialists within a firm. Coase's framework translates this immediately: each eliminated transaction is a transaction that previously justified organizing those specialists in an organizational structure. The specification meeting, the handoff, the design review, the revision cycle—each was a transaction cost, real and substantial, whose management justified the firm's existence. When the AI-augmented individual can perform both functions, the coordination that justified the structure is no longer necessary. The Coasian boundary of the firm has shifted inward not because the firm became less efficient at what it does, but because the individual became vastly more capable.

The cycle's documentation of the Software Death Cross—a trillion dollars of market value leaving software companies in the first weeks of 2026—is, in Coasian terms, the market repricing firms whose primary value resided in the production layer. When code can be produced by individuals at negligible cost, the firm that existed because writing code was expensive has lost its transaction-cost justification. What survives, Coase's framework predicts, is the firm whose boundary-justification was never primarily production: the data, the integrations, the customer relationships, the institutional trust that AI cannot replicate. The market has identified the direction correctly; it has not yet calibrated the magnitude for individual firms.

The Coasian lens in the cycle sits alongside but distinct from Ronald Dworkin's, which asks what is morally at stake in the institutions the boundary reorganization is touching. Coase explains the economic logic of why the boundary is moving; Dworkin explains why its movement in courts, agencies, and schools demands more than economic analysis. The two framings do not compete. They address different levels of the same transformation, and both are necessary to see it whole.

Origin

Coase was born in Willesden, England in 1910, educated at the London School of Economics, and spent the defining months of his intellectual formation at American factories asking why they organized as they did. The paper that resulted, published in Economica in 1937, introduced the concept of transaction costs—though the phrase itself came later from Oliver Williamson, who systematized Coase's insight into a full theory of organizational boundaries. The profession initially received the paper with benign indifference. It was too simple, too obvious, too concerned with the real world rather than the elegant formalisms then in fashion. It took decades and Williamson's extension into asset specificity and vertical integration before the profession understood what Coase had done: he had given economics the tool it needed to understand institutions, not just prices.

His second great contribution was the Coase Theorem—the proposition that when property rights are clear and transaction costs are zero, private bargaining produces efficient outcomes regardless of the initial allocation of rights. The theorem is famous for the qualifier: it was designed to show how important transaction costs are by showing what happens in their absence. “I never liked the Coase Theorem,” Coase himself told an interviewer. “It's a proposition about a system in which there were no transaction costs. It's a system which couldn't exist.” The theorem's purpose was diagnostic: it showed that the real world was everywhere defined by the transaction costs the theorem assumed away.

Coase received the Nobel Prize in Economics in 1991, at eighty-one, and remained intellectually active until his death in 2013 at 102. He spent his final decades at the University of Chicago, characteristically unhappy with the profession's tendency to operate on blackboards rather than in real markets. His last book, How China Became Capitalist, co-written with Ning Wang, examined institutional transition using the framework he had spent eighty years refining. Throughout, his method never changed: look at what is actually happening, rather than what the theory predicts should happen, and build the framework from the observation up.

Key Ideas

Transaction Costs as the Foundation. Firms exist because using the price mechanism is costly. Search costs, bargaining costs, specification costs, enforcement costs—these are the real resources consumed by the process of exchange itself, as distinct from the resources consumed by what is being exchanged. When those costs are high enough, internal coordination under managerial direction is cheaper than contracting on the market. The transaction cost taxonomy is the analytical foundation of everything Coase built.

The Moving Boundary. The boundary of the firm is not fixed; it is a frontier determined by the continuous comparison of internal coordination costs against market transaction costs. Any change in the underlying cost structure moves the boundary. AI has moved it by collapsing the most expensive categories of knowledge-work transaction costs: search costs approach zero when the AI has been trained on the relevant documentation; bargaining costs disappear when the person who conceives a product can also produce it; translation costs—the enormous expense of converting intention into a form a specialist in another domain can act on—are eliminated when the machine speaks the language of intention. The boundary has moved inward for production, and the institutions on the wrong side of it face the Coasian question they cannot avoid: what do we still do that justifies organizational coordination?

The Individual as Firm. The solo AI-augmented builder—producing in a year what previously required a team of five—represents the limiting case of the Coasian firm: a firm of one, in which coordination costs are zero because there is no one to coordinate with. The AI is not an employee; it does not shirk, renegotiate, or require management. It is a production technology that has absorbed the functions of coordination, allowing the individual to direct it as a manager directs an employee while incurring none of the costs of management. The individual as firm is real, measurable, and proliferating.

What Markets Cannot Provide. The Coasean Singularity—the hypothetical endpoint where AI reduces transaction costs to zero and the economic rationale for firms disappears entirely—will not arrive, because some functions that sustain the firm are not reducible to transaction costs: trust, the transmission of tacit knowledge, the maintenance of professional standards, and the provision of belonging and professional identity. Markets have never been able to provide these things efficiently, and AI does not change that. The firm persists, but as a social institution—a relationship technology whose production function has been largely externalized to AI-augmented individuals but whose coordination, quality-assurance, and community functions remain essential. These residual functions are what the firm of the future will be built around.

Institutional Lag. The boundary has already moved; the institutions on both sides of it have not yet reorganized to reflect its new position. This gap—between the speed of technological change and the speed of institutional response—is the defining structural problem of the current moment. Legal frameworks for liability, property rights in AI output, regulatory structures for algorithmic accountability: none of these has caught up to the boundary shift. The institutional lag is where the transition is most dangerous, because the costs of misalignment between the moved boundary and the unmoved institutions fall on the people—workers, consumers, citizens—who had no say in moving it.