The case that labor demand compression is real has already been made, and not as a single event but as a pattern of absences: the hiring round that quietly never happens, the department that absorbs a product line without growing, the middle layer of a company that thins out one retirement and one AI subscription at a time. An earlier piece named the mechanism, another dismantled the reassurance that a trade is a safety net, and a third traced the whole idea of wage labor back to a fifteenth-century invention rather than a law of nature. None of them proposed what replaces the wage once it stops doing the job it was invented to do — handing a person access to food, housing, and a reason to get up. This is not another brief for universal basic income; Dauphin, Manitoba already supplied that evidence, and repeating it adds nothing. It is a taxonomy of the institutional shapes that already exist, in some working form, for coordinating a society whose labor market has permanently stopped clearing.

I. What Already Works, and What It Doesn’t Solve

An earlier piece on non-extractive institutions already made the case that Mondragón, Wikipedia, and Elinor Ostrom’s commons share a design pattern — governance by the governed, surplus kept local, exit that is real and cheap — and that the pattern stays marginal because trust has to be earned before it can scale, while extraction can scale first and earn trust later or never. That argument solves a narrower problem than the one this piece is asking about. Mondragón redistributes surplus among people who already have a seat inside the cooperative. Wikipedia’s governance answers to people who already showed up to edit. Those are institutions for the included. The harder question is what happens to the people compression never let in — the un-hired generation, the cohort whose absence from the hiring statistics is the whole story. That is a distribution problem, not a governance-of-existing-surplus problem, and the existing evidence for non-extraction doesn’t by itself answer it.

II. Universal Basic Services

The economist Anna Coote and the policy researcher Andrew Percy make the sharpest version of the argument in The Case for Universal Basic Services (2020): what people lack, in a compressed labor market, is not primarily purchasing power but access to functioning systems — housing, healthcare, education, transport, food, delivered as a right rather than purchased with wages that may not exist. The Nordic welfare states pushed toward this model without ever fully committing to it, funding services generously while still routing most access through employment status. The failure mode is not hypothetical. It’s the same one that afflicts any sufficiently large administrative system: the thing that gets measured — beds filled, cases closed, forms processed — replaces the thing that was supposed to matter, the mechanism an earlier post on Goodhart’s Law already named in a different context. A universal basic service is only as good as its resistance to becoming a bureaucracy that optimizes for its own throughput instead of the person it was built to serve.

III. Platform Cooperatives

The media theorist Trebor Scholz coined “platform cooperativism” in 2014 as the direct inversion of the platform economy Yanis Varoufakis would later call techno-feudal: instead of a platform that extracts a toll from every driver, courier, or freelancer who depends on it, the platform is owned by the people who produce its value. Nathan Schneider’s Everything for Everyone (2018) surveys the working examples — cooperative ride-share apps, freelancer marketplaces owned by their freelancers, data trusts where the people generating the data hold the governance rights. None of it is theoretical; drivers in several cities already dispatch through apps they co-own. What keeps the model from displacing the extractive version at scale is capital and network effects, in that order: a venture-backed platform can burn money for a decade to win a market before it ever turns a profit, and a cooperative raising capital from its own members cannot match that burn rate, while the network effect that made the incumbent platform dominant in the first place actively punishes any smaller challenger for being smaller.

IV. The Guild, Rebuilt

The historian Sheilagh Ogilvie’s The European Guilds (2019) is a useful corrective before reaching for the word: medieval guilds were not romantic craft communities, they were cartels that restricted entry to protect the prices and privileges of existing members, and Ogilvie’s archival work shows the restriction was often more damaging than helpful to the economies that hosted them. The economist Morris Kleiner has spent a career documenting the modern version of the same mechanism — occupational licensing that raises the cost of becoming a florist or a hair braider far beyond any plausible safety justification, gatekeeping dressed as consumer protection. What a rebuilt guild would need to borrow instead is the model the legal scholar Yochai Benkler called commons-based peer production in The Wealth of Networks (2006): the open-source maintainer structure, where standing is earned by contribution rather than purchased through a licensing exam, and where the credentialing body is the community of practice rather than an incumbent association with an interest in scarcity. The failure mode is the one Ogilvie already documented seven centuries ago under a different name: any body that certifies competence can convert that certification into a toll on entry, and the guild becomes indistinguishable from the cartel it was supposed to replace — the same credentialing drift the title-trap thread already tracked inside existing professions.

V. Universal Basic Inheritance

The idea is older than any of its modern advocates. Thomas Paine’s Agrarian Justice (1797) proposed a one-time capital grant paid to every citizen at adulthood, funded by a tax on the land whose value, Paine argued, was never any individual’s private creation to begin with. Bruce Ackerman and Anne Alstott revived the argument two centuries later in The Stakeholder Society (1999): fund the grant instead from a wealth tax, and let people own a stake rather than receive a recurring check, on the theory that ownership changes behavior in ways a subsidy never does. The nearest thing to a running version of the mechanism is the Alaska Permanent Fund, which has paid every Alaska resident an annual dividend from the state’s oil revenue since 1976 — a recurring payment rather than Paine’s one-time grant, but proof that a sovereign wealth fund can distribute a resource windfall directly to citizens for decades without the arrangement collapsing into either dependency or resentment. The failure mode is the one built into any equal-start design: identical capital produces unequal outcomes the moment it meets unequal luck, health, and judgment, which is not a flaw exclusive to this model so much as the thing every equal-start model eventually has to answer for.

VI. The Constraint All Four Share

Line the four up and the same tension that already marked Mondragón and Wikipedia reappears, before any of them are even built at scale. Each requires trust to precede rather than follow growth — a universal basic service has to be trusted not to calcify into bureaucracy, a platform cooperative has to be trusted with capital its members raised themselves, a rebuilt guild has to be trusted not to relicense its way back into a cartel, a basic inheritance has to be trusted to fund itself from a windfall or a tax base that outlives any single government. None of the four can borrow the shortcut extraction gets for free: scale first, prove legitimacy later, or never. That is not a reason to expect any of them to fail. Mondragón, Wikipedia, and the Alaska fund all work, each at a scale that once looked implausible. It is a reason to notice that none of the four, as they exist today, were designed with the un-hired generation as the intended member. They were built by and for people already standing somewhere inside an economy, then adjusted afterward to be more equitable to the people already inside them. Whether any of the four can be rebuilt as the default point of entry for people the labor market never had a place for — rather than a correction bolted onto the system once someone notices they were never hired — is the question compression is going to keep asking whether or not an answer is ready.

Further reading