Worker-managed firms can curb inequality and protect free markets

Rising income inequality in Western high-income countries threatens democratic governance and social cohesion, and AI risks making it worse. Drawing on a forthcoming book, this column argues for worker-managed firms as a durable remedy. With no external owners taking profits, worker cooperatives protect against the rising bargaining power of employers and provide an enhanced sense of self-worth and self-determination for employees. Furthermore, the data suggest that well-designed worker cooperatives compete with conventional firms in free markets with no evidence of a productivity gap. Because they are underprovided by the market, it calls for government action to support worker cooperatives.

Inequality in Western high-income countries has reached levels not seen since the early 20th century (Piketty 2014). In the US, median earnings of full-time employees remained essentially unchanged between 1979 and 2019 despite a near-doubling of national income per capita (Bureau of Labor Statistics 2024). Even the most optimistic measures of market income growth for households in the bottom half of the income distribution show huge gaps opening with the top 10% of the distribution (Auten and Splinter 2024), and even more so with the top 1% and top 0.01%. European countries exhibit similar patterns, with top-decile income growth systematically outpacing bottom-half growth by factors of 2-to-3.5 across the UK, Germany, France, and Italy (Blanchet et al. 2022). This relative stagnation of living standards for the majority, combined with extreme concentration of income and wealth at the top, is widely seen as a systemic failure of current economic arrangements, threatening social cohesion and the very foundations of democratic governance. The meteoric rise of artificial intelligence (AI) threatens to worsen both these inequality trends, and the resulting erosion of democratic norms.

Two main approaches to addressing rising inequality have received extensive attention from leading economists and policymakers. The first is aggressive redistribution through progressive taxation and expanded social transfers (e.g. Piketty 2014, Saez and Zucman 2019, and Atkinson 2015). The second focuses on regulatory reforms designed to strengthen worker and consumer bargaining power. This includes raising minimum wages, supporting the unionisation movement, strengthening antitrust enforcement to break up concentrated industries, and imposing restrictions on anti-competitive labour market practices (e.g. Dube and Lindner 2024, Rosenfeld 2014, Philippon 2019, Krueger and Posner 2018).

However, it is increasingly clear that these conventional remedies face a fundamental challenge: they are unlikely to be sustained over time due to the disproportionate political influence of wealthy individuals and corporations, who have both incentives and means to systematically erode policies designed to constrain their economic dominance. Extensive evidence across the social sciences shows that this influence operates through multiple channels, including campaign contributions, lobbying expenditures, media ownership, and the ‘revolving door’ between government and private sector. Perhaps more fundamentally, the scale of the redistribution required raises troubling philosophical, social, and psychological questions. If most people end up living on a ‘basic-income’ cheque cut by the AI companies, which is really a form of charity, what will that do to their self-esteem, sense of purpose in life, and self-determination? Not to mention: what will they do with their time – other than consuming AI output?

My forthcoming book Cooperatives in the Twenty-First Century uses economic theory and evidence to propose a remedy against extreme inequality which is more robust against backlash from wealthy interest groups; more protective of the immemorial link between one’s work and one’s livelihood; more liable to expand, rather than restrict, the sphere of individual self-determination; more capable of preserving collective democratic norms; and perhaps more likely to influence the direction of the development of AI in a way that complements work rather than substitutes for it – as recently advocated by Acemoglu et al. (2026). 

It is intuitive (and amply confirmed by the data) that firms democratically managed by their own workers will feature less pay inequality – especially as driven by the pay packages of CEOs, which have now crossed the trillion-dollar line in the corporate sector. But that is not where the greatest inequality-reducing benefits of worker democracy come from. The main benefit is the absence of external owners to take a share of earnings in the form of profits – which are increasingly inflated by the rising bargaining power of employers. Furthermore, because worker cooperatives do not issue tradable ownership shares, they make it very difficult to build disproportionate individual fortunes and positions of overwhelming political influence. These benefits are complemented by the enhanced sense of self-worth and self-determination coming with co-responsibility in the management of firms. In addition, when the businesses that are the customers of the AI companies are run by the workers, it seems more likely that the AI firms will skew their training of large language models (LLMs) and other forms of generative AI towards a ‘pro-worker’ type of AI future.

These benefits might be worth having even if a structural change to a cooperative economy caused a significant decline in output per worker – or its growth rate. But, in fact, the hundreds of thousands of well-designed worker cooperatives that currently compete head-to-head with conventional firms in free markets show no indication of a productivity gap. If anything, simple means comparisons across large samples often show a productivity advantage for co-ops.

Contrary to a well-known critique (Hansmann 1996), cooperatives with well-designed governance structures, in which managers elected by the membership exert strong executive power, do not spend excessive amounts of time and resources in deliberative activity, nor do they bear any larger costs of disagreement than conventional firms. Part of the reason is that the cooperative form removes the largest reason for disagreement at all: the fundamental conflict of interest between workers and owners. Instead, workers in cooperatives leverage the superior information they have on their own co-workers’ abilities and values to select the best among themselves for management roles, and to monitor their performance – on which their own livelihood depends. This contrasts sharply with corporations, in which a mass of atomistic and dispersed shareholders is essentially in the dark about the true goings-on in the firm and the true value added brought by managers, and gets very little relief from this ignorance from boards of directors that are ever more toothless and conflicted (e.g. Zingales 2017). It is also in contrast with owner-managed firms, where ownership status, not necessarily ability, confers rights of control – a problem especially apparent after the retirement of the founder (Caselli and Gennaioli 2013).

Nor is there any evidence that worker cooperatives are pits of laziness and shirking, as famously predicted by Alchian and Demsetz (1972). If anything, there is suggestive evidence that monitoring costs are higher in conventional firms. The reason is simple. Workers in cooperatives continuously observe each other’s effort and performance as a natural by-product of working side by side. Because they are co-residual claimants in the output of the firm, workers observing shirking behaviour have direct incentives to pass the information on to management, and to expect management to act on it. This incentive is absent in conventional firms, and thus the free information generated by peer monitoring is wasted. (If anything, evidence suggests that peer monitoring in conventional firms is often used to enforce low-effort norms among workers.) As a result, conventional firms must engage to a much greater extent in costly and imperfect supervisory monitoring. To make matters worse, in conventional firms, the benefits of flexible approaches to technology and production methods are often sacrificed to the quest for monitoring-cost-minimising workflows and methods.

Predictions that worker cooperatives would fail at capital accumulation (e.g. Furubotn 1976) have also yet to find any support in the data – at least by comparison with conventional firms. On the other hand, there is some circumstantial evidence of an advantage in human capital accumulation, particularly in the guise of training expenditure (e.g. Pendleton and Robinson 2011). A possible explanation for this advantage is that worker cooperatives largely remove the two-sided hold-up problem in acquiring firm-specific human capital – a problem that plagues conventional firms.

If cooperatives distribute average incomes to their workers that are higher than those paid out by conventional firms, perhaps they expose their members to too much risk – as famously argued by James Meade, among others (Meade 1972)? After all, conventional firms pay fixed salaries which seem to insulate workers from volatility in firm performance. In fact, it is now well understood that whatever insurance benefit workers receive from fixed salaries while employed is dwarfed by the much greater risk of unemployment. Cooperative members, in contrast, enjoy significantly superior employment stability, because cooperatives prefer to temporarily reduce income payments or dip into reserves to protect employment and long-run firm survival (e.g. Pencavel et al. 2006, Burdin and Dean 2009, 2012). 

Worker cooperatives, then, are great for workers, both in theory and in practice. So why are there so few of them? There are two sets of reasons. The first is that the cultural, legal, educational, financial, and professional infrastructure supporting firm creation in high-income countries is almost entirely designed to serve the needs of conventional firms. The second set of reasons is intrinsic to the economics of co-ops. Potential cooperative founders often have stronger private benefits to adopting the conventional form – even though the cooperative form produces larger social benefits. Specifically, the conventional form ensures that the founder will capture a significantly larger share of the surplus created by the firm. In most cases, this results in larger benefits to the founder even though the co-op form would have created a larger overall surplus.

This reasoning reveals that worker cooperatives are essentially public goods, and, as such, underprovided by the market. The textbook prescription to deal with underprovided public goods is government action. Cooperatives in the Twenty-First Century lays out a reform agenda to achieve shared prosperity via workplace democracy.  

Source : VOXeu