Selling the Last Unencumbered Asset: Coase, AI, and a Property Right in Labor-Market Access
In 2002 the mathematician and economist Eric Weinstein published a proposal in the International Labour Review constructing what he called a canonical Coasean market solution to economic immigration — recognizing native workers as the holders of a right of preferential access to their own labor market, and requiring employers who wish to hire outside it to license that right for a fair price rather than receive it by administrative grant. Native workers keep working at a lower wage and receive licensing income on top of it; on Weinstein's account their blended income rises. He has since extended the proposal to artificial intelligence, treating AI as one more supplier of labor that workers license, and framing it as the free-market alternative to universal basic income. The proposal is genuinely Coasean and it is more rigorous than most of what occupies that space. This essay takes it seriously on those terms and then subjects it to the framework's own apparatus. Fekete's test finds no exchange of income for wealth on the worker's side: the worker does not sell the capacity to work, and the human capital stays where it was. What the proposal does is quieter. It splits the wage in two — a return to skill, which remains human capital, and a premium for access to the market, which becomes a licensing payment routed through government and worker representatives, confined to one jurisdiction and one sector. Human capital is unconfiscatable, not fixed in nominal terms, and portable; the licensing claim has none of those properties. In the immigration case the claim is a supplement. In the AI case it is the growing share, and at full displacement it is the whole. The proposal does not capitalize labor. It replaces it, by degrees, with a claim.