Cooperative Economic Performance: Calculation and Self-Management
Position
Two linked empirical claims underpin most objections to cooperative economics: that non-capitalist systems cannot perform economic calculation, and that workers cannot manage complex enterprises without bosses. Both fail against the evidence — the calculation claim outright, the management claim against a literature strong enough that its absence from mainstream economic discourse is itself the thing needing explanation.
The Calculation Problem Dissolves
Mises’s economic calculation argument (1920) holds that without market prices for capital goods, rational resource allocation is impossible. This argument was formulated against centralized state ownership of all production. It does not apply to cooperative market economies, which use market prices for exactly the same allocation functions as conventional firms. Worker cooperatives buy inputs at market prices, sell outputs at market prices, and use internal accounting to allocate resources — the calculation problem simply does not arise.
More fundamentally, every large corporation is a centrally planned economy internally. General Electric, Amazon, and Walmart allocate resources across thousands of divisions using internal transfer prices, managerial directives, and bureaucratic planning — not market mechanisms. The internal economy of a Fortune 500 company is larger than most national economies that Mises worried about. If central planning is impossible, these firms should not exist. They exist because the calculation problem is about information and incentives, not about the metaphysical impossibility of non-market allocation.
Stock markets — the supposed engine of capitalist price discovery — finance approximately 0.5% of net capital formation. The vast majority of investment is funded through retained earnings, bank loans, and internal corporate allocation. The price-signal argument for capitalism vastly overstates the role that equity markets play in actual resource allocation.
The Self-Management Evidence
The strongest claims the literature actually supports, stated at full confidence: worker-owned and worker-managed enterprises match or outperform conventional firms on productivity and on survival — including survival through downturns — and outperform on pay equity and employment stability. The safety and job-satisfaction findings run favorable with more mixed results across studies. That is the precise version, and it is sufficient: the boss’s efficiency justification does not survive parity, let alone advantage.
Mondragon (Basque Country): 80,000+ worker-owners across manufacturing, finance, retail, and education. The cooperative complex has operated for seven decades, weathered multiple recessions with near-zero layoffs (using internal redeployment and temporary pay cuts voted on democratically), and maintains a maximum pay ratio of roughly 6:1 compared to 300:1+ in conventional corporations.
The record is not unblemished, and pretending otherwise is unnecessary. Mondragon’s Fagor appliance division collapsed in 2013 after over-expansion — the federation declined to bail it out, and workers lost jobs. Mondragon’s international subsidiaries employ non-member workers under conventional labor relations. These are real. They are also design data, not disproof: Fagor faced market discipline without the too-big-to-fail bailout guarantee that conventional firms receive, and the subsidiary labor structure reflects the gravitational pull of operating within capitalism, not evidence that cooperatives require exploitation. When a conventional firm collapses, no one concludes conventional ownership is structurally impossible. Apply the same standard. → hist/mondragon.md
Emilia-Romagna (Italy): A cooperative-dense regional economy producing approximately 40% of regional GDP. The region consistently ranks among Italy’s wealthiest and most innovative, with cooperative enterprises competing successfully in advanced manufacturing, construction, agriculture, and services.
Systematic evidence: Pérotin’s surveys of econometric studies, the CECOP reports on European cooperatives, and Dow’s analysis of labor-managed firms converge on the core finding — democratic workplaces perform at least as well as hierarchical ones on conventional economic metrics, and better on most human-welfare metrics studied.
The Capital Channel: Why Co-ops Are Rare If They’re Good
The critic’s sharpest move is revealed preference: “Co-ops are legal everywhere and have been for centuries. Workers keep choosing conventional firms over the republic next door anyone is free to found. If your model needs a mandate to outcompete a status quo that already permits it, the mandate is doing confession.”
The answer is not “the terrain is tilted” in the abstract. It is three specific, literature-grade mechanisms — all running through capital, none through preference:
Credit rationing. Banks lend against collateral and concentrated control. Worker-owners pledging the same firm that employs them present lenders a correlated risk no conventional borrower presents, and democratic governance gives creditors no controlling counterparty to discipline. Bowles and Gintis formalized the result: cooperatives are systematically credit-starved at formation, independent of their operating performance.
The equity asymmetry. A conventional startup sells ownership stakes to raise capital. A cooperative cannot sell control without ceasing to be a cooperative — its defining feature is also a structural lockout from the equity channel that funds everything else in the economy. This is not a market verdict on co-op quality; it is a constraint built into the intersection of co-op form and capital-market design.
The demutualization ratchet. When cooperatives succeed and accumulate value, members face a standing offer: convert to conventional ownership and capture the windfall. Conversions run overwhelmingly one direction — successful mutuals and co-ops get demutualized for capture reasons orthogonal to efficiency. The population of co-ops is continuously harvested from the top, then the survivors’ scarcity is cited as evidence of weakness.
The natural experiments close the case. Where the financing constraint was specifically fixed, formation rates jump: Mondragon’s growth ran through Caja Laboral, a cooperative bank built for exactly this; Italy’s Marcora law (workers’ right to capitalize unemployment benefits into buyouts) produced hundreds of durable worker takeovers; the UK’s Employee Ownership Trust framework triggered a formation boom within years of enactment. Same workers, same legal permission as before — different capital channel, different outcome. That converts “workers don’t choose it” into “the capital channel is the binding constraint”: a testable claim, partly verified, and the policy implication is a level channel, not a mandate.
Why the Evidence Is Ignored
The core performance findings — survival and productivity at parity or better — are well-established in the specialist literature; the genuine disputes live in the margins (which metrics, which sectors, which causal mechanism). What needs explaining is the absence of even the established core from mainstream economic discourse, and the explanation is that it contradicts the foundational assumption that hierarchy is necessary for complex economic organization. If workers can manage enterprises as well as or better than bosses — and the evidence says they can — then the boss’s share of the surplus has no efficiency justification. It is pure extraction, maintained by institutional inertia, legal structures that favor conventional incorporation, and differential access to capital.
Objection Handling
| Move | Response | Concession |
|---|---|---|
| ”Mises proved socialist calculation is impossible” | Mises argued against centralized state ownership without market prices. Cooperative market economies use market prices — the argument does not apply. Meanwhile, every large corporation plans internally without markets. Amazon’s internal resource allocation is larger than most national economies. If calculation requires markets, firms above 50 employees should not exist. | Fact — concedes that Mises identified a real problem with fully centralized planning — accepts the importance of price information while showing it exists in cooperative market systems |
| ”Workers lack the expertise to run complex businesses” | Workers already run complex businesses — they do the actual work. “Management expertise” is real but is a form of labor, not a property of ownership. Cooperatives hire and elect managers; they just don’t cede ownership to them. The claim that workers “can’t” manage is belied by every self-managed enterprise that exists. The question is whether management should be accountable to workers or to shareholders. | Fact — accepts that management functions are real and require skill — concedes the need for coordination and expertise while disputing that these justify ownership hierarchy |
| ”Cooperatives don’t scale — show me a cooperative Amazon” | Mondragon: 80,000+ worker-owners, seven decades, diversified across manufacturing, finance, retail, and education. Emilia-Romagna: cooperatives producing 40% of regional GDP. The question is not whether cooperatives can scale but why the legal and financial environment makes scaling harder — conventional incorporation has centuries of institutional advantage, preferential access to capital markets, and legal frameworks designed for it. | Fact — concedes the critic notices cooperatives are rarer at the largest scales — then forces the question of why: the legal and financial environment was designed for conventional firms over centuries. The ‘scaling challenge’ is an institutional barrier, not an organizational one, and the critic must now defend the barrier or concede the point |
| ”Show me the evidence they actually outperform” | Pérotin’s meta-analyses, CECOP European cooperative reports, Dow’s systematic reviews — all find cooperatives match or exceed conventional firms on productivity and consistently outperform on safety, retention, pay equity, and recession survival. The evidence is not ambiguous; it is ignored. Name the metric and there is a study. | Contested — accepts the demand for evidence as legitimate; what the Pérotin-class literature firmly supports is productivity and survival comparable-or-better plus greater employment stability, while across-the-board outperformance remains the framework’s wager rather than the field’s consensus — stronger selection-corrected studies in either direction would move it |
| ”Survivorship bias — we only see the good cooperatives” | Cooperatives have higher survival rates than conventional firms across multiple studies and countries. The bias runs the other direction: conventional firms fail more often, but the failures are attributed to individual management rather than the ownership structure. When a cooperative fails, it is treated as evidence against the model; when a corporation fails, it is treated as evidence of market discipline working. | Fact — concedes that survivorship bias is a real methodological concern — accepts the analytical point while showing the data actually favors cooperatives on survival rates |
| ”Even if cooperatives work, you can’t force everyone to adopt them” | No one is proposing force. The argument is that the current institutional environment — corporate law, capital market access, tax structures, bankruptcy rules — is designed for conventional firms and systematically disadvantages cooperatives. Leveling the institutional playing field is not coercion; it is removing the thumb from the scale. If cooperatives can compete on equal terms, let them. | Frame — accepts that institutional adoption should be voluntary — concedes the anti-coercion principle while redirecting to the structural advantages conventional firms currently enjoy |
| ”Co-ops are legal everywhere — workers keep choosing conventional firms. Revealed preference.” | Legality is not channel access. Three mechanisms — credit rationing, the equity asymmetry (co-ops can’t sell control without ceasing to be co-ops), and the demutualization ratchet harvesting successes into conventional form — constrain formation independent of preference or performance. Where the capital channel was fixed (Caja Laboral, Marcora law, UK EOTs), formation jumped with no change in legal permission. The “preference” tracks the financing, not the form. | Contested — concedes the comparison should be made under equal conditions; whether the capital channel or the governance costs of collective choice is the binding constraint on formation remains a live dispute in the literature — the framework reads the natural experiments as favoring the capital channel, and formation booms that fizzle under level financing would move it |
| ”Workers rationally refuse to stake savings and wages on the same firm — diversification is physics, not law” | The risk-concentration point is real, and co-op design already prices it: members typically capitalize through wage deferral and collective reserves, not lump savings, and external non-voting capital instruments exist precisely to import diversification without importing control. Meanwhile absentee capital’s diversification advantage exists because the legal architecture lets ownership separate fully from work — a design choice, not physics. Fix the instruments and the “physics” turns out to be plumbing. | Fact — concedes the obstacle is risk allocation, not worker capability or co-op performance — an instrument-design problem with existing instruments |