Hierarchy,equality,or democracy: Examining the tradeoffs in the governance of economic organizations

Abstract

This paper investigates the Standard Hierarchical View – the perspective that the only way to satisfactorily organize business entities in the modern world is through a hierarchical governance structure. Although this view is dominant across sociology, it is in need of serious revision. Whereas the Standard Hierarchical View assumes that there are only two fundamental possibilities for organizing economic enterprises – firms can be efficient-and-hierarchical or inefficient-and-egalitarian – we show that the universe of possibilities for organizing economic firms is better understood as a spectrum characterized by three ideal-types: conventional hierarchical firms, egalitarian collectives, and democratic worker cooperatives. In order to assess the pros and cons of democratizing the economy, the paper engages in a comprehensive evaluation of the performance of these three organizational types in terms of: productivity; survival; ability to grow and spread; innovation; ability to minimize domination; respect and dignity; distributive equality; job security; and job satisfaction. We find that both hierarchical firms and egalitarian collectives have major disadvantages (domination and inefficiency, respectively). Overall, the evidence suggests that the Standard Hierarchical View is wrong: worker cooperatives, though far from perfect, generally perform better than the other two possibilities in terms of avoiding their major pathologies. Democracy, not hierarchy, should be the default form of economic organization.

Keywords

hierarchy democracy equality worker cooperatives egalitarian collectives efficiency domination workplace democracy

Introduction

Should we aim to democratize the economy? This paper returns to a classical issue in sociology – the potentials, constraints, and tradeoffs in organizations that value both democracy and equality, on one hand, as well as complexity and efficiency on the other. Specifically, this paper investigates what we might call the Standard Hierarchical View (SHV). The Standard Hierarchical View is the perspective that the only way to satisfactorily organize business entities in the modern world is through a hierarchical governance structure. Egalitarian firms do occasionally exist, but they are not viable alternatives for two basic reasons: they are either chronically inefficient, which is why we do not see them in large numbers, or they survive by shedding their egalitarian commitments and degenerating into conventional hierarchical firms. We take this to be the dominant view across sociology, political science, and economics, as well as that of most regular citizens (Alchian and Demsetz, 1972; Leavitt, 2005; Michels, [1911] 1962; Webb and Webb, 1907; Williamson, 1973).

However, although this view is dominant, it is in need of serious revision. In contrast to the SHV, we advance two main arguments. First, although the SHV assumes that there are only two fundamental possibilities for organizing economic enterprises – firms can be efficient-and-hierarchical or inefficient-and-egalitarian – we argue that the universe of possibilities for organizing economic firms should actually be understood as a spectrum characterized by three ideal-types. Firms can be organized hierarchically (such as in conventional capitalist businesses), or they can be organized in a strictly egalitarian manner (such as in egalitarian collectives) – in which case there are indeed stark tradeoffs in terms of complexity versus equality. However, there is also a third option which is that firms can be organized via representative democracy (such as in worker cooperatives), which allows for complexity, efficiency, and a high degree of equality (though by no means complete equality of power or pay).

Second, the paper aims to provide a rigorous answer to the question of “should we democratize the economy?” by way of a careful review of the empirical evidence. We evaluate the three organizational types on the basis of widely desired criteria: productivity; survival; ability to grow and spread; innovation; ability to minimize domination; respect and dignity; distributive equality; job security; and job satisfaction. The evidence reveals that each firm type has specific pros and cons. In particular, both hierarchical firms and egalitarian collectives have major disadvantages (domination and inefficiency, respectively). Overall, the evidence suggests that the Standard Hierarchical View is wrong: worker cooperatives, though far from perfect, generally perform better than the other two possibilities in terms of avoiding their major pathologies. All things considered they appear to be superior forms of social organization (for much, though perhaps not for every sector of the economy). Democracy, not hierarchy, should be the default form of economic organization.

Hence this paper offers two main contributions: a clearer conceptualization of the possibilities for organizing economic life, and a rigorous evaluation of their pros and cons. Though most of the extant literature comparing worker co-operatives (henceforth simply “co-ops”) with conventional firms focuses on their relative economic performance, this paper attempts to synthesize a broader range of evidence – the economic plus the social, in order to provide a scorecard that is more comprehensive. Given the huge literatures that have now developed surrounding the performance of co-ops, we have had to bracket certain issues (such as community impact, and impacts on health). Nevertheless, within these parameters our aim is to provide the most comprehensive overview of the evidence that exists to date.¹

Part one describes the three ideal types of firms; part two carefully weighs the empirical evidence; and part three analyzes the results.

Conceptualizing possibilities

In the world today there is a wide diversity of governance structures of firms, from large hierarchical corporations to tiny anarchist collectives. There are unionized firms, B Corps, ESOPs, codetermined firms, and more. According to the SHV, although many different types of firms, including egalitarian ones, are technically possible, hierarchical governance is the only feasible possibility for large, efficient firms in the contemporary world. The implicit assumption is that there are really only two fundamental possibilities: hierarchical (and efficient) or egalitarian (and inefficient).

In contrast, we demonstrate that what the SHV sees as “egalitarian firms” is actually a conflation of two very different types of organization: egalitarian collectives and democratic cooperatives. Though both of these have egalitarian attributes, co-ops are what we might call “pragmatic egalitarian”; they operate very differently from egalitarian collectives, which we might label “strict egalitarian.” In other words, we propose to categorize the messy diversity of real-world firms on the basis of three ideal types.

The first ideal type is the “conventional” or “hierarchical” firm. Hierarchical firms are those where decision-making power is organized in an unaccountable top-down manner (in the sense that those at the bottom have no formal power or accountability over the decisions made at the top). In addition to this governance structure, these firms usually, but not always, possess a cluster of correlated features: substantial inequality of status and respect, substantial inequality of remuneration, a central goal of productive efficiency and profit maximization, and a core value of authoritarian right (i.e., that those at the top have legitimate authority and those below owe deference and obedience). Major examples include fascist workplaces, state-communist workplaces, and most importantly for our purposes here, conventional capitalist workplaces.

Within the umbrella of hierarchical capitalist firms there is a very wide diversity, for instance, maquiladoras and sweatshops, corporate multinationals, small businesses, as well as unionized workplaces.² It is also possible for hierarchical capitalist firms to incorporate elements of worker participation within the overarching hierarchical governing structure.³

At the other end of the spectrum, the second ideal type is that of egalitarian collectives. Egalitarian collectives are defined by a governance structure whereby decisions are made collectively through direct participatory democracy, aiming at a more-or-less equal distribution of decision-making power. This structure is undergirded by collective ownership of the business assets. Usually, but not always, this structure has a cluster of correlated features: equal status and respect between members, equal (typically exactly equal) remuneration, a central goal of humane work, and a core value of strict egalitarianism. Examples include the traditional kibbutzim in Israel, the wave of counter-culture co-ops in the 1960s and 1970s in the US, some of the more radical worker-recovered enterprises (Empresas Recuperadas por sus Trabajadores, or ERTs) in Argentina, as well as a subset of the world’s worker co-ops (usually small, highly ideological ones).

A third ideal type is that of democratic co-operatives. Democratic cooperatives are worker cooperatives where the highest decision-making power is accountable to all members on an egalitarian basis. Typically, the organization is pyramidal, meaning power is unequally distributed; managers have more power than shopfloor workers, but they are also accountable to all members (usually via one-person one-vote elections of top officials in a representative democratic manner, in addition to having the ability to vote on a range of issues in regular general assemblies). Worker co-ops are owned collectively by the member-owners who constitute the majority of employees of the firm. Usually, but not always, this structure has a cluster of correlated features: equal status and respect between members, a relatively small amount of wage inequality, a central goal of balancing profitability with humane work (especially in terms of providing genuine democratic voice and job security for workers), and a core value of pragmatic egalitarianism, meaning that equality is a core value but not an absolute value or the only one; pragmatic compromises are seen as necessary in order to achieve efficiency and profitability. Major examples include the Mondragon co-ops in Spain (at least from the 1950s to 1990s), as well as the relatively vibrant Italian, French, and Uruguayan cooperative sectors.

We stress that these three types are idealized. They are meant to highlight important family resemblances. In the real world, of course, things are messy, and there is a certain amount of overlap (see Figure 1). To take one example, Germany’s codetermined firms (which allow workers to elect a portion of the supervisory board of directors) are partly hierarchical and partly democratic. In practice, many firms combine hierarchical and democratic elements; moreover, firms often change over time, sometimes becoming more hierarchical, sometimes less. Nevertheless, despite the inevitable marginal blurriness, the three types of firms are relatively clear and crisp categories. As we will see in the next section, the three types do behave quite distinctly. Despite the inevitable blurriness, their essential differences are, we claim, important, stable, and robust.

Figure 1.

The spectrum of governance structures of firms.

Before moving on, an important terminological note. In common discourse the word “hierarchy” usually refers to any and all kinds of organizations that are pyramidal, involving managers giving orders. This is problematic because it is far too broad. Instead, we will use the word “hierarchy” more narrowly and precisely to refer only to governance structures that are unaccountable. In other words, we insist on distinguishing between two fundamentally different types of pyramidal organizations: those that provide democratic accountability of the top to the bottom through regular election involving all the members that are subject to the decisions will be called “democracies” (even if they have tiers of management), whereas those that do not provide this accountability will be called “hierarchies” in the strict sense of the word.

Evaluation

We are now ready to evaluate the three types of firms. Methodologically, each section reviews the evidence, relying as much as possible on studies that compare similar kinds of firms (similar size, industry, location, etc.) This is important so that we don’t bias the evaluation by comparing, say, the worst performing conventional firms with the best performing co-ops, or vice versa. The best comparative studies are those that derive from large, nationally representative samples of firms, covering many industries, and where close attention is paid to issues of endogeneity, reverse causality, and selection bias, such as Burdín and Dean (2009) and Burdín (2014) in Uruguay, Pérotin (2006), Fakhfakh et al. (2012), and Young-Hyman et al. (2023) in France, Podivinsky and Stewart (2007, 2009) in the UK, Pencavel et al. (2006) in Italy, and Arando et al. (2012) in Spain.

At the end of each section, we will also attempt to assess the confidence that the reader should have in the results. “High confidence” will mean that the evidence largely points in the same direction, and that the studies are rigorous (large-n, sensitive to reverse causality, etc.) “Medium confidence” means that the evidence mostly points in the same direction, but the studies are less rigorous. This might be the case with quantitative studies that are small-n; or it might be that the only evidence available is that of qualitative case-studies. Of course, qualitative studies can be highly informative, provided they are cautious, attentive to tradeoffs and so on, but it is harder to know how generalizable their results are. “Low confidence” means either that the evidence is contradictory, or the underlying studies are questionable (for instance, case studies can be questionable if they are ideologically dogmatic or non-transparent regarding opposing evidence).

Productivity

The first major issue is that of efficiency. “Efficiency” of course is a broad and contentious term.⁴ We focus first on several key aspects of efficiency in the narrow economistic sense of the word: productivity, survival rate, a firm’s ability to grow, and innovation.

First we will look at productivity. In terms of theoretical predictions there is much dispute. Advocates of conventional firms offer a number of reasons why one might expect hierarchy to be productivity-enhancing: control by relatively homogenous shareholders allows for unified goal-setting and the clear prioritization of profitability above all else; hierarchical control allows for quicker decision-making and more expert decision-making; hierarchy provides powerful incentives for workers – large pay differentials can motivate workers to try to climb the ladder, and workers can be more motivated due to increased fear of sanction, leading to less shirking or absenteeism – as well as powerful incentives for residual claimants (since the more effort workers exert, the more rewards owners receive) (Alchian and Demsetz, 1972; Hansmann, 1996; Williamson, 1973).

On the other hand, critics point to a number of reasons why co-ops might have productivity advantages. Co-ops allow for more motivation, commitment, and loyalty since members get a share of the profits; cooperative workplaces may well have less conflict, anger, or resentment; co-ops are better able to benefit from the ideas and wisdom of shopfloor workers; and less money needs to be spent on monitoring and supervising (Cornforth et al., 1988; Dow, 2018; Kruse et al., 2010).

What does the empirical evidence say?

The most robust studies are those of Craig and Pencavel (1995), Fakhfakh et al. (2012), Monteiro and Straume (2018), and Young-Hyman et al. (2023).

Fakhfakh et al. (2012) use a large, representative data set (n = 7000 firms, 500 co-ops) and find no significant difference in productivity (in only one industry – “Paper and Wood” – is there a significant difference, which is in favour of worker co-ops). They also find no significant difference in the size of firms or in the level of capital investment. Their conclusion is that “Overall . . . [worker co-ops] are as productive as conventional firms or more productive, and use their inputs better” (p. 18).

Monteiro and Straume (2018) in Portugal (n = 685 of all kinds of co-ops, not just worker co-ops) find co-ops to be less productive in one estimation, but not in their “preferred empirical strategy.”

Young-Hyman et al. (2023) use French data on workers’ co-ops that are matched with similar conventional firms. Their data allows for a difference-in-difference design, examining changes in productivity before and after the conversion to workplace democracy. They find “consistent and robust evidence that worker cooperatives are more productive in knowledge intensive settings.” Firms in a knowledge-intensive industry that switched governance structures from conventional to a worker co-op enjoyed an average productivity increase of 8.9% (pp. 1354, 1365).

Craig and Pencavel (1995) gather longitudinal data on plywood firms in the northwestern US (n = 7 co-op mills, 19 unionized, 8 non-unionized), all of which are very similar in terms of production methods and technologies. They find that co-ops have a productivity advantage of 6%–14% over conventional firms (they find no difference between unionized and non-unionized firms).

In addition to these leading studies, most other studies have similarly found co-ops to be just as productive as conventional firms, if not more so. This is the case in Brazil (Silva and Morello, 2021), in Italy (Bartlett et al., 1992; Maietta and Sena, 2008), in Spain (Bayo-Moriones et al., 2003; Thomas and Logan, 1982), in France (Defourny, 1992), in Denmark (Mygind, 1987), and in the US (Berman, 1967). However, the literature is not unanimous (Abebe and Gebreeyesus, 2023; Jones, 2007; Jones and Backus, 1977).⁵

The conclusion is that, on average, worker co-ops are just as productive as comparable conventional firms, and this is well supported by the evidence, at high confidence.

Survival

A distinct but important way to think about the issue of efficiency is to ask about a firm’s survival rate.

Here too there is wide consensus in the literature that co-ops have just as good survival rates as conventional firms (most studies find an advantage for co-ops, especially for young firms) (Burdín, 2014; Fakhfakh et al., 2023; Mirabel and Lomuscio, 2025; Olsen, 2013; Pérotin, 2006; Vieta et al., 2017).

The state-of-the-art study is from Burdín (2014), who uses data on the entire population of Uruguayan worker co-ops (n = 223), comparing them to conventional firms (controlling for size and industry type). He finds that the likelihood of dissolution is 29% lower for co-ops than conventional firms.

The conclusion that co-ops survive just as well (if not somewhat better) than conventional firms is strongly supported, at high confidence.

The ability to grow and spread

Since states first granted corporations legal rights in the 17th century, and in particular the extended right of limited liability in the 1800s (allowing large numbers of people to invest some capital, with the potential for enormous rewards but only limited risks), corporations have shown a stunning ability to grow. Although the vast majority of hierarchical firms are small, the largest are truly immense. Walmart currently employs about 2.1 million people, McDonalds employs 2 million (McDonalds, 2024; Walmart, 2025).

Can co-ops grow? The answer appears to be yes. Most famously, Mondragon started in 1956 with five workers and basically no assets. By the late 1980s, it composed 166 co-ops, with 20,000 worker-members, and $1.6 billion in sales.⁶ Indeed, the first 27 years saw the start-up of 103 new co-ops with only one three failures (Whyte & Whyte, 1988, p. 172). This is stunningly successful growth, especially considering that for conventional firms the failure rate is 50% in the first 5 years (BLS, 2016). Mondragon is structured as a parliamentary system, where each of its worker co-ops manages itself democratically, and also elects delegates to attend the Cooperative Congress, where decisions about the structure as a whole are made (Malleson, 2013). Though unusual, this remains a powerful and inspiring model for what large-scale cooperativism might look like. Another example is the Consorzio Cooperative Costruzioni (CCC), a consortium of worker co-ops, and one of the largest general contractors in Italy with a turnover (in 2010) of 3.5 billion euros (Zamagni, 2015).

Of course these are only a couple of examples. Luckily there is also more comprehensive evidence. Most studies find that co-ops have similar (or slightly superior) productivity vis-à-vis conventional firms (as shown above), and importantly, they also have similar investment levels (Bartlett et al., 1992; Estrin and Jones, 1998; Fakhfakh et al., 2012). Given this, it is no surprise that most studies find that co-ops grow just as fast as conventional firms. Fakhfakh et al. (2012) find that “in all industries, the growth of capital in SCOPs [i.e., worker co-ops] is the same as or higher than in conventional firms” (pp. 863–864). Likewise, Arando et al. (2010) found that Mondragon grew faster and had higher investment rates than conventional firms, as did Moye (1993) for an earlier period.

Can co-ops spread? There is a real puzzle here since it is clear that in most contexts co-ops are rare. In the US, there are only 465 co-ops out of 6,100,000 employing firms, that is, only 0.0076% of enterprises (Prushinskaya, 2020; SBECouncil, 2018). Co-ops are somewhat more prevalent in Europe, though still quite rare overall: there are 30,000 worker co-ops representing 654,000 worker-members, which constitutes roughly 0.13% of European firms (Terrasi and Hyungsik, 2017: 15).

Why are they so rare?⁷ The data make it clear that the answer is not that co-ops are rare because they are inherently inefficient. As we have seen, they are not. Rather the problem is that they are very rarely set up in the first place. In Dow’s (2003) words, “The general conclusion . . . is that LMFs [labor-managed firms] are not rare because they fail disproportionately often. Once created, they appear robust. Rather, they are rare because in absolute numbers they are created much less often than KMFs [capital-managed firms]” (p. 227). In other words, co-ops can have difficulty in initial “formation,” but they do not have difficulty in “performance” (Dow, 2025). A similar conclusion is reached in the investigation by Olsen (2013).

Is it possible for co-ops to flourish and spread widely? The evidence seems to point to a contingent yes. Yes, because there are concrete examples where they have clearly done so, such as Mondragon and La Lega (the National League of Cooperatives) in northern Italy. For example, in the region of Emilia Romagna, there are 1,199 worker co-ops, employing 98,689 workers (roughly 6% of all employees), generating 3 billion euros of value-added (Lomuscio et al., 2023: 138). Co-ops of all types constitute 12.75% of GDP and dominate in many industries, including construction, agriculture, food processing, wine making, transport, retail, and machine production. In the town of Imola (population 100,000) 50% of the adult population are members of the area’s 115 co-ops (of all kinds, not just worker co-ops), and over 60% of the town’s GDP comes from the co-ops (Ammirato, 1996: 2; Restakis, 2010: 56–57).

But contingent in that co-ops clearly flourish best when they operate within a broader supportive eco-system (not just any market system but a cooperative market system). This includes legal supports (such as a clear and well-designed legal structure and tax advantages), financial support (such as credit from co-op development banks), managerial and business support (such as co-op consultants, and federations and consortia for scaling up and sharing risk with other co-ops), educational support (such as cooperative universities and management training programs), and political support (via established networks with political parties, unions, and various government agencies). Indeed, a large part of Mondragon and La Lega’s success is due to the fact that they have been able to enmesh themselves in many of these supports. Inside such generative cooperative eco-systems, co-ops have indeed been able to multiply impressively (Adeler, 2014; Ammirato, 1996; Corcoran and Wilson, 2010; Hemingway and Pek, 2024; Smith, 2001).

It’s important to realize that precisely the same is true for conventional hierarchical firms: their ability to grow and succeed is similarly dependent on a supportive background infrastructure (an effective legal edifice, universities training students in corporate management, a massive system of banks and stock markets to provide financing, and so on). The difference is that this vital background infrastructure is often taken for granted or assumed as “natural,” in the case of conventional firms in a way that it is not for co-ops (Gunn, 2006; Malleson, 2013, 2014).⁸

Can co-ops spread to every sector of the economy? One area of obvious concern is that of capital intensity. Theoretically it seems that co-ops should find it hard to move into these areas, since relatively poor workers would have to buy expensive firms (which is far less likely than large numbers of relatively rich shareholders each buying a small number of shares) (Roemer, 2013).

Empirically, the evidence is mixed. Using a large data set (n = 9041 conventional firms and 110 co-ops) Podivinsky and Stewart (2009) (and similarly in Podivinsky and Stewart (2007)) find that co-ops enter capital intensive industries less commonly than conventional firms. Mirabel (2024; n = 84 co-ops) finds that the likelihood of co-ops emerging is higher when the sunk cost of human capital is dominant and the sunk cost of physical capital is negligible. On the other hand, Pencavel et al. (2006; n = ~200 co-ops) do not find any significant difference in average capital/labour ratios, nor do Fakhfakh et al. (2012). So the jury is still out on this question. Nevertheless, a sensibly cautious interpretation is that co-ops may well have less ability than conventional firms to spread into highly capital-intensive industries (yet given the mixed evidence, our confidence in this result is low).

There is one additional area where there is good reason to expect co-ops to perform significantly worse than conventional firms: in international competition involving large multinationals. Whereas conventional corporations will happily fire a highly paid workforce in, say, Spain, in order to hire low-paid workers in, say, China, worker co-ops will be reluctant to do the same because worker-owners will obviously not want to fire themselves.⁹

Mondragon is a clear example of this. Reluctance to sacrifice secure, well-paying jobs at home (in Spain) has made it hard for them to compete against capitalist firms who have no qualms in doing so (Basterretxea et al., 2019b; Errasti et al., 2017). Exchanging high-quality, well-paid, secure, democratic jobs for low-quality, poorly-paid, precarious, undemocratic ones clearly can increase profit, but whether it should therefore be called “efficient” in the broad sense of the word – whether it’s better for people and planet overall – is, of course, an entirely different question.

In sum, it appears that co-ops can grow just as well as conventional firms (at least to medium size) – we can have high confidence in this fact. The ability to spread appears moderate (though the evidence for this is less certain, giving us medium confidence in this regard).

Innovation

It remains an open question as to whether co-ops are more or less innovative than conventional firms. The few existing empirical studies find that co-ops are typically more innovative, but the difference is small (Basterretxea and Martínez, 2012; Basterretxea et al., 2019a; Santos et al., 2024). The most rigorous study to date is from Basterretxea and Martínez (2012) (n = 861 firms; 44 co-ops) which finds that co-ops perform slightly better in terms of management capabilities, innovation capabilities, actual competitiveness, and future competitiveness (p. 268). They also find that co-ops invest a higher percentage of their turnover in Research and Development than conventional firms. So a provisional conclusion is that co-ops appear to be just as innovative, however, given the paucity of the evidence, confidence in this result is relatively low.

A brief aside on degeneration

Since the writing of Marx ([1894] 2010, Chap 27) and Webb and Webb (1907) many have believed that co-ops are doomed to degenerate under market pressure. Indeed, this view is now so common that it constitutes an important part of what we are calling the Standard Hierarchical View. One prediction is that co-ops will inevitably replace members with (exploited) non-members; another is that they will degenerate through under-investment (the so-called “horizon problem” (Furubotn and Pejovich, 1970)). In the real world, however, neither of these appear to be common problems, likely due to institutional rules (common in Spain, Italy, France, and elsewhere) mandating that co-ops maintain majority member-ownership and build up their indivisible reserves (Fakhfakh et al., 2023; Pérotin, 2014).¹⁰

A third worry is a variant of Michels’s ([1911] 1962) “iron law of oligarchy”: the prediction is that as democratic organizations become large and complex their governance structure will inevitably degenerate into oligarchy. The evidence for this kind of degeneration is stronger than the other two. Practically all medium to large co-ops do indeed face serious pressures of this sort (Cheney, 1999; Hernandez, 2006; Holmstrom, 1989; Schoening, 2010). However, there is nothing like an “iron law” here as there are frequently countervailing pressures too, such as workers pushing for more democracy. The historical record shows that some co-ops do indeed see their democracy degenerate, but many do not (even after decades). Others degenerate but then, interestingly, regenerate (Bretos et al., 2019b; Cornforth et al., 1988; Holmstrom, 1989; Storey et al., 2014; Stryjan, 1994).

It is also worth pointing out that in practice many firms are often slowly shifting how they operate, sometimes becoming more democratic, sometimes less. In other words, degeneration and regeneration are often partial processes, rather than binary black-or-white transformations. The ideal-type classification between “hierarchical firms” and “democratic co-ops” is a relatively clear and crisp distinction based on governing structure (either a firm allows its member-workers to elect the governing board, or it does not). But in the real world, the extent of democratic accountability in a firm is not simply due to the presence (or absence) of formal governing structures. Other things matter too, such as the degree of democratic culture within the firm (vibrancy of participation, transparency of information, norms of collaboration and inclusivity, etc.), the extent to which all employees are included as co-op members, the presence (or absence) of other kinds of mechanisms of worker voice or ownership (such as unionization, works councils, or shared stock ownership), the degree of meaningful worker involvement in shopfloor decision-making, and so on. For example, Mondragon’s degeneration was largely due to a massive expansion of employees in firms outside of Spain who are not co-ops members (even though the “home co-ops” are still fairly robustly democratic). Hence Mondragon today is a kind of democratic-hierarchical hybrid (Errasti, 2015). It’s also possible for conventional capitalist firms to be structured hierarchically but to nevertheless operate day-to-day in ways that are quite democratic (where managers routinely involve workers in egalitarian decision-making, work happens in co-managed work teams, inclusive consultation is the norm, and so on (Blasi and Kruse, 2006; Frega, 2021)). Since the boss could at any time unilaterally revoke these democratic practices, a firm like this is still formally a hierarchy, but here too it is one with complicated hybrid features. To reiterate, while the ideal-type classification is useful for clarifying the big picture, the real world is messy and complex, and democratization and hierarchicalization are often partial, ongoing, and cross-cutting processes.

Egalitarian collectives

Unfortunately, there is little empirical work directly comparing egalitarian collectives with other types of firm; this is in part because so few of them exist, and in part because they are rarely recognized as their own distinct organizational form different from co-ops (with which they are usually conflated). The relative lack of evidence, and the fact that the studies that we do have are qualitative rather than large-n comparative studies, mean that we must tread cautiously.

Whereas co-ops often grow to medium size (Pérotin, 2016), egalitarian collectives are practically always small (usually less than 15 people) (Rothschild and Whitt, 1986; Schoening, 2010). Do they operate productively at this scale? The evidence is mixed.

On the one hand, it is clear that a radically egalitarian governance structure does limit the complexity of the business. Since egalitarian collectives try to avoid having experts who have more power than others, they rotate tasks so that everyone does everything. This is likely to significantly reduce productivity because it limits specialization. To see this, imagine a simple restaurant with three workers: a chef, a waiter, and a bookkeeper. Each have specialized skills that took them years to acquire (the chef spent years in culinary school, the waiter has keen social skills from years of experience, and the bookkeeper has an accounting degree). It is clear that if they rotated positions, the restaurant’s productivity would suffer: the chef is not going to be a good bookkeeper, nor the waiter a good chef. They can learn, of course, but in the meantime the restaurant will perform badly. The obvious explanation is that people typically have skills only in certain specific areas, so the more time they spend working in areas where they are skilled, the more productive they will be (and similarly, the more time they spend working in areas where they are unskilled, the less productive they will be overall). The only case in which rotation won’t matter is when every job is simple, requiring no specialized technology or advanced skills.

Rothschild and Whitt (1986) provide the example of a small alternative newspaper firm, The Community News, with a staff of 12–18 people. All the major decisions were made collectively by consensus at weekly meetings, and the various jobs of the paper (reporter, photographer, proof-reader, salesperson, etc.) were rotated, so no one was allowed to continue working in only one area. Rotation enabled some to learn new skills, but it also severely limited the development of specialization and expertise which are clearly vital for high levels of productivity. The authors conclude that The Community News, as well as the other egalitarian collectives they studied, suffered significant losses in productivity due to these factors (p. 112). Other illuminating case studies show much the same (Jaumier, 2017; Meyers, 2022; Schoening, 2010; Vieta, 2020).

On the other hand, egalitarian collectives frequently benefit from countervailing forces, in particular, a high level of motivation and commitment from members.

Vieta shows that of the 410 ERTs (i.e., firms that had failed as conventional firms and then been taken over by their workers) in Argentina, as of 2016 only 43 (i.e., 10%) had failed (p. 116). Given that 100% of these firms had previously failed as capitalist firms, this is powerful, though of course not conclusive evidence that they are reasonably productive. It is not known how many of these firms are collectives as opposed to cooperatives, but Vieta estimates that roughly 30% could be classified as collectives.¹¹ However, Vieta also points out that the ERTs which are closer to the egalitarian collective model are mostly found in less competitive sectors (he finds that where market competition is fiercer, ERTs are more likely to be structured as worker co-ops rather than as egalitarian collectives) (pp. 487–489).

So the evidence seems to indicate that egalitarian collectives will be limited in their ability to spread to areas of the market where simple technology and low levels of specialization suffice. But within this niche they can operate productively (at least compared to other firms of similarly small scale). This conclusion has medium confidence (since it’s based on high-quality case studies, but not comparative statistical data).

We do not have data about their survival rates of their rates of innovation. But there is good reason to doubt that these firms will be particularly innovative since they are confined to areas of low specialization and minimal technology.

The major limiting factor for egalitarian collectives is their inability to grow. One tight constraint flows from the fact that as firms grow and become more complex, they necessarily require increased coordination between the various parts, which in turn requires managers to oversee the whole. An egalitarian collective that does not want managers, will likewise not want to grow. Second, when a firm becomes too large (much more than 20 people) it becomes increasingly impossible – far too time consuming and exhausting – for all important decisions to be made in a participatory fashion. This problem is exacerbated if the collective insists on consensus decision-making (which requires even more deliberation and time than majority-based voting). Another tension is that large, complex organizations require specialized skills, and individuals with specialized skills often (though not always) desire more pay than others, and can threaten to leave if not paid more than others, which conflicts with the norm in egalitarian collectives of equal pay. A final constraint is that collective decision-making (especially consensus decision-making) is strongly aided by cultural homogeneity, but homogeneity is hard to maintain beyond small handfuls of people. For all these reasons, egalitarian collectives are under strong pressure to remain small. And indeed the empirical evidence bears this out (Dar and Getz, 2020; Horrox, 2009; Meyers, 2022; Rothschild and Whitt, 1986; Schoening, 2010; Vieta, 2020). In practice, most collectives are tiny, fifteen members or less.

The case of the Burley cooperative is both typical and instructive. In the early years it was highly egalitarian, with nine members who all shared a similar counter-cultural ideology. “Everybody did everything. Small tasks too. Everybody was on the Board [of Directors]. We all made all the decisions. At the time, it worked because we were just on the verge of the explosion. It was great. We would all eat lunch together. It was one of those things like, as soon as you came on, you were equal to everybody in every way. There was no seniority. There was no pay differential” (Schoening, 2010: 316). After a decade, the business took off, growing to 26 members. However, as the business grew, so too did the number of issues and their complexity. Participatory decision-making involving all members required longer and longer meetings, eventually becoming untenable. Finally, a breaking point was reached and Burley was faced with a stark choice: either to shrink back to a smaller business in order to maintain itself as an egalitarian collective, or to grow and shift structures into a conventional worker co-op with representative democracy replacing direct democracy. Burley chose the latter. Members began to elect a board of directors responsible for long-term planning of the co-op. Internal work was divided into “task teams,” with each team electing a representative to a management team that would oversee the day-to-day business operations. Finally, a general manager was hired to lead the management team. The restructure worked, and the co-op successfully grew to sixty members. Schoening reports that “the new democratic structure of the cooperative seemed to have resolved the tensions of growth . . . spreading the benefits of cooperative work to an ever-larger workforce” (pp. 326–327).

Although the Burley co-op did not survive forever, what is particularly instructive is the clear tension between egalitarian collectives and growth. Egalitarian collectives may work at a small scale, but they generally cannot grow. They thus have to choose between three basic options: to remain flat and small, to altogether abandon their values and become a conventional hierarchical firm, or to balance their values through a pyramidal but representative structure that can grow as a democratic cooperative (this conclusion has high confidence, given the degree of consensus among the authors of the various case studies).

Minimizing domination

Domination is being subject to another’s will, being used as a tool for another’s purpose, and being denied the ability to self-direct, self-govern, or self-determine – in other words, being made into a servant. In what follows, we use the term to refer to starkly unequal decision-making power between order-givers and order-takers, enforced by a threat of substantial harm (Pettit, 2012).

In theory, co-ops should be less dominating for two reasons: first, the democratic structure gives workers an indirect say in governance via representation, which reduces their vulnerability (workers are less likely to be fired easily or arbitrarily). The second reason is less commonly appreciated in the literature, which is that democratic firms may well be more amenable to providing opportunities for at least occasional direct participation in decision-making on the shopfloor via self-managed work teams (Frega (2021) is right to point out that worker involvement in shopfloor decision-making is a neglected but important pillar of workplace democracy).¹²

In general, the empirical evidence bears this out. The comparative evidence on subjective evaluations finds that even though co-ops can still have bureaucracies, pyramidal decision-making, and managers directing workflows, they also tend to be more inclusive, participatory, and egalitarian than conventional firms (Bayo-Moriones et al., 2003; Tannenbaum et al., 1974). For instance, Bradley and Gelb (1981) found that 21% of co-op members felt there was a large division between management and workers, compared to 62% in regular firms; 13% felt they participated in important decisions compared to 4% in capitalist firms; and 30% felt they had no participation, compared to 80% in capitalist firms.

Co-ops also tend to have more open and freer communication (Bradley and Gelb, 1981; Hoffmann, 2001), less domineering management (Atzeni and Ghigliani, 2007), and significantly fewer supervisors overall (Bartlett et al., 1992; Fakhfakh et al., 2012; Greenberg, 1986). Greenberg (1986) describes how after a co-op transformed into a conventional hierarchical firm, the first thing management did was to quadruple the number of supervisors. In the words of the General Manager who used to be a manager in the co-op, “We need more foremen because, in the old days, the shareholders supervised themselves . . . They cared for the machinery, kept their areas picked-up, helped break up production bottle-necks all by themselves. That’s not true anymore . . . we’ve got to pretty much keep on them all of the time” (p. 44).

However, a couple of caveats are in order. First, having a formal democratic structure (with regular elections) is necessary but not sufficient for an organization to be meaningfully democratic. There must also be a democratic culture (so that there are clear norms and expectations regarding self-management, transparency about decision-making, open communication, willingness of managers to collaborate and include others, a desire for at least some participation by some workers, and so on) (Cornforth et al., 1988; Hernandez, 2006). It’s worth noting that Mondragon’s famous co-op bank, the Caja Laboral, which successfully launched numerous new co-ops, preferred focussing on start-ups rather than conversions precisely because of the difficulties it encountered in building a cooperative culture in previously undemocratic firms (Whyte and Whyte, 1988).

A second caveat is that it’s clear that many co-ops fail to provide meaningful shopfloor participation. At Mondragon, for example, some senior managers argue that assembly line technology makes shopfloor voice impossible (Basterretxea et al., 2019b); Kasmir (1996) reported Mondragon workers complaining that they were treated like “chickens in a coop” (p. 187). Clearly co-op members can feel dominated, even if they formally have voice and a vote. Overall, it’s probably fair to say that, ceteris paribus, co-ops are more interested in providing opportunities for shopfloor participation than conventional firms, but there is wide disparity here.

In sum, co-ops appear to be moderately successful in reducing domination (a conclusion of which we can have medium confidence given the quality of the data).¹³

It’s also worth noting that over the last several decades there have been many attempts to implement a degree of “worker participation” within the structure of conventional hierarchical firms, such as quality circles (later integrated into broader kaizen and lean production systems), worker surveys, open-book management, referring to the company as a “family,” holacracy, profit sharing, shadow boards, and so on. Sometimes these processes result in genuine and meaningful (if only partial) democratization and the reduction of domination (Frega, 2021); yet sometimes they are low impact or even propagandistic (aimed at convincing workers that they are being listened to so that they will care more about the firm and work harder, without actually giving away any power or profit) (Blasi and Kruse, 2006; Wilkinson et al., 2020). How do we know which is which? Bernstein provides a useful measuring stick composed of six elements of genuine democratization: the level of control that workers have over decision-making; the degree of information sharing; the extent of individual rights; the existence of an independent mechanism for adjudicating disputes; the extent of a participatory-democratic culture; the extent of profit sharing (Azzellini and Vieta, 2026: 56; Bernstein, 1976). In differentiating genuine democratization from candy-wrapped hierarchy, the heart of the matter is whether there is real sharing of decision-making power and profit with the bulk of employees, or not.

Respect and dignity

There is a close link between domination and disrespect. Starkly unequal decision-making power makes those at the bottom of the hierarchy into servants and tools for the use of others; and there is, arguably, something inherently demeaning about being made into a tool for purposes over which you have no say or influence (Anderson, 2017). Moreover, the existence of profound inequality in power and authority tends naturally to bleed into inequality of status. To be disempowered is almost always to be disrespected.

There is substantial evidence showing that disrespect and degradation of workers by managers is common in hierarchical firms. One dramatic example is Tyson (the second largest meat processing firm in the world), where American workers were prevented from using the bathroom, forcing some to urinate and defecate on themselves while their supervisors mocked them; others felt compelled to wear diapers (Oxfam, 2016).

A particularly vivid example of the inherent disrespect generated by undemocratic firms is the fact that tens of thousands of front-line workers around the world died from exposure to COVID-19, in large part because they had so little say about their working conditions (and so were forced to continue to work and interact with the public without proper safety gear even after the dangers were well known). For instance, in California, food and agriculture workers (who frequently have little power in their workplace) faced a 39% increase in excess mortality from the pandemic (compared to a 6% increase for White Californians in general) (Chen et al., 2021). It is hard to believe that such workers would have been forced to work without safety gear had they been co-owners of their own worker co-ops. These are thus deaths of disrespect.

More broadly, among restaurant workers, an astounding 80% report experiencing sexual harassment (Jayaraman, 2016: 38). Among production workers, only 29% strongly agree that they are treated with respect (similarly for only 32% of food preparation workers and 34% of sales staff); among workers in general, only 24% say they’re expected to be creative at their job, and only 20% strongly agree with the statement that “my opinions seem to count” (Zanini, 2022). Disrespect and disregard appear to be ubiquitous in the conventional capitalist workplace, particularly at the bottom of the hierarchy.

There is little comparative data directly comparing conventional firms to co-ops in this regard, but there is a fair amount of indirect data. Workers in co-ops are allowed to participate more in decision-making (see above); they also typically have higher job satisfaction (see the section below on “Job Satisfaction”); co-ops commonly invest more money into worker training than do conventional firms (Arando et al., 2010; Bartlett et al., 1992; Basterretxea and Martínez, 2012; Morrison, 1997); and they usually have less antagonism with management (Greenberg, 1986; Hoffmann, 2001), for instance, co-op managers frequently embrace little signs and symbols of equality, such as eating with the workers and (in Italy) using the informal “tu” mode of address (Holmstrom, 1989). In her study of a British coal mine, Hoffmann (2001) describes how under the hierarchical British Coal, the miners echoed the disrespect the men felt from the managers with their own disrespect toward the managers, resulting in mutual resentment, often leading to formal grievances and work stoppages. After the conversion to a worker co-op, the managers were in more frequent and informal contact with workers showing increased respect and openness.¹⁴

All in all, such evidence strongly suggests that workers in co-ops are generally, but of course not universally, treated with more respect and dignity than in hierarchical firms. Indeed, given the increased power that workers have in co-ops – as well as their formal status as equal partners – it would be strange if they didn’t also experience heightened respect. (This conclusion has medium confidence because most of the evidence points in the same direction, but we lack large comparative studies on the question).

Minimizing inequality and exploitation

One of the most empirically robust facts about co-ops is that they have less distributive inequality than conventional firms: in general, the lowest earners are paid more while the top managers make less. There is broad consensus on this point, and the quality of the evidence allows for high confidence in the result (Arando et al., 2011; Bartlett et al., 1992; Burdín, 2016; Craig and Pencavel, 1995; Magne, 2017). Given this, plus the fact that workers are also co-owners receiving the residual profits (there are no outside shareholders), it follows that there is also less exploitation (in the technical sense of value produced by some being appropriated by others).

However, the extent of the differences in inequalities between the two types of firms is often underestimated. The comparative studies typically match co-ops against similar size conventional firms. This makes good sense for many purposes, but the problem in this case is that since most co-ops are small or medium size, such comparisons inevitably ignore the largest firms – but that is precisely where inequality is starkest. For example, Craig and Pencavel (1995) find wage inequality between the highest and lowest paid in capitalist unionized firms to be a ratio of 2.5:1. Bartlett et al. (1992) find a ratio of 2.86:1. These are incredibly low levels of inequality for capitalist firms, and so give a very distorted view of the overall economy. We get a much more accurate picture by looking at the dominant players in the economy. Among the top 350 American firms in 2020, the average CEO was paid roughly $24.2 million (including wages and various stock benefits). Hence the CEO-to-typical-worker pay ratio is a staggering 351:1 (Mishel and Kandra, 2021). In other words, the data from comparative studies underestimates the level of inequality in large capitalist firms by a factor of 100. And even this enormous inequality still underestimates the true extent of inequality in the contemporary US. For instance, just before he stepped down as CEO, Jeff Bezos had a wealth of approximately $150 billion from ownership of Amazon shares. At a standard 5% return, this translates into an annual income of $7.5 billion; in contrast, the lowest paid at Amazon make $15 per hour; this is a wage ratio of 240,000:1. In the largest co-ops, on the other hand, the difference is typically less than 3:1, and only very rarely exceeds 10:1 (Dow, 2003). As far as we’re aware, even after its degeneration, the maximum inequality that any of Mondragon’s co-ops ever experienced was 9:1 (Reuten, 2022, p. 347). It is inconceivable that co-ops would ever allow inequality anywhere near the levels seen in the largest hierarchical firms. An economy dominated by co-ops would have few multi-millionaires, and certainly no billionaires.

This matters because the inequality that exists within firms ripples out across society and causes all kinds of significant harms. The evidence is now robust showing that increased inequality is associated with increased poverty, reduced opportunity, reduced democracy, more xenophobia and racism, worse mental health, and increased crime of various sorts (Malleson, 2023).

Job security

Another widely agreed upon fact is that co-ops provide better job security than conventional firms (Bartlett et al., 1992; Burdín and Dean, 2009; Cristini et al., 2023; Fakhfakh et al., 2012; Pencavel and Craig, 1994; Pencavel et al., 2006). Here too the quality of the evidence allows for high confidence in the result.

Whereas conventional firms lay off workers whenever it is valuable to do so, co-ops do not see their members as disposable forms of human capital, as merely means to the end of increased profit, and so lay off their members only as a very last resort. Indeed, the difference between firm types in this regard is striking: in hierarchies, human workers are completely disposable. If it were possible to replace every single human worker with cheaper and more productive machines, the hierarchical firm would instantly do so. Co-ops on the other hand would not, since maintaining good work for the people involved is their essential purpose; their mission is fundamentally different.¹⁵ This difference is particularly evident in times of recession (Bretos et al., 2019b; Burdín and Dean, 2009; Cristini et al., 2023).

Job satisfaction

Measuring job satisfaction is notoriously difficult because it is always highly contingent on expectations. The most useful data comes from research which asks people not just about their happiness in general, but instead focuses on people who have direct experience with both types of firms in order to evaluate their satisfaction in comparative terms (e.g., “having experienced both kinds of workplace, which is better?”). Cheney (1999), Moye (1993) and Bradley and Gelb (1981) all find that the majority of Mondragon workers would prefer to stay in their co-op than transfer to a capitalist firm (even at higher pay). Harnecker (2007) finds the same in Venezuela. In Italy, Holmstrom (1989) finds that 86% of co-op workers prefer to work in a co-op than a conventional firm because of the more relaxed discipline, lack of petty restrictions, and the feeling of being trusted and consulted. Likewise, Cornforth et al. (1988) report that even though co-op workers are frequently disappointed with various aspects of their co-op, nevertheless “people did not want to give back control over their working lives once they had experienced it” (p. 111). Even Kasmir (1996), who is highly critical of the Mondragon co-ops, acknowledges that the majority of members do not want to see their firms converted into conventional hierarchical firms (p. 166).

This evidence is reinforced by other studies showing that co-ops typically have less worker conflict and antagonism with managers, as well as lower quit rates, than conventional firms (Bartlett et al., 1992; Greenberg, 1986; Hoffmann, 2001).

So a reasonable conclusion is that co-ops generally provide higher levels of job satisfaction than conventional firms. But given the lack of large-n studies here, as well as the high-degree of inherent ambiguity concerning people’s self-evaluations of “happiness,” we should probably rank the confidence level here as low to medium.

Egalitarian collectives

In egalitarian collectives, the case studies find that there is typically little domination relative to conventional firms since there are no bosses with the power to fire, nor servants forced to obey (Horrox, 2009; Rothschild and Whitt, 1986).

The level of mutual respect is usually quite high, with members typically viewing each other as comrades if not also close friends (Palgi, 2006; Vieta, 2020).

The standard practice is for wages and profits to be distributed precisely equally (Horrox, 2009; Rothschild and Whitt, 1986; Vieta, 2020).

There is little hard data on job security, but given the deep commitment to equality and solidarity, it seems clear that the general membership will typically fire a colleague only in extreme circumstances. So, in general, egalitarian collectives are likely to have even stronger job security than co-ops.

Members typically have quite high levels of job satisfaction (and do not want to return to hierarchical work), though these are sometimes offset by the very high hopes of what an egalitarian workplace should be like (Rothschild and Whitt, 1986; Tannenbaum et al., 1974; Vieta, 2020: 502).

The confidence in the evidence of these results is medium; there are no large-n analytics, but the case studies are generally rigorous and all point in the same direction.

Summary and analysis

Table 1 summarizes our results.

Table 1.

Evaluating conventional firms, egalitarian collectives, and democratic co-ops.

	Conventional firms	Egalitarian collectives	Democratic co-ops
Productivity	High	Moderate (but only at small scale)**	High***
Survival	High	Unknown	High***
Ability to grow	High	Very low***	High***
Ability to spread	High	Very low**	Moderate to high (depending on institutional supports and with caveats about areas of capital intensity and global competition)**
Innovation	High	Very low**	High*
Minimizing domination	Low	High**	Moderate**
Respect	Low to moderate	Very high**	Moderate to high**
Distributive equality (lack of exploitation)	Low	Very high**	High***
Job security	Low	Very high*	Very high***
Job satisfaction	Low to moderate (but highly dependent on expectations)	High (but highly dependent on expectations)**	Moderate (but highly dependent on expectations)*

The results are colour-coded for easy viewing. High is green, moderate to high is light green, moderate is yellow, low to moderate is orange, low is red. These are, of course, very broad generalizations. It goes without saying that there will always be counterexamples.

Confidence level of the result is denoted by asterisks: high confidence (***), medium (**), and low (*).

The main results of this study are as follows.

In terms of the fundamental categories used to make sense of the diversity of governance structures in firms, it is incorrect to think that the only real possibilities are hierarchical firms governed top-down, or egalitarian firms governed horizontally. There is an important third possibility: democratic worker co-ops governed via representative democracy. These three types are (generally speaking) different and distinct. Although scholars have not commonly distinguished between co-ops and collectives, they should. Empirically, the differences between such firms are robust and important, and so they should not be conflated.

The definition and basic conceptual understanding of what constitutes a co-op should be modified. Co-ops are standardly defined in terms of their democratic governance structure. For example, the International Cooperative Alliance defines a co-op as “an autonomous association of persons united voluntarily to meet their common economic, social and cultural needs and aspirations through a jointly-owned and democratically-controlled enterprise” (ICA, 2018). What is missing from such a definition is that co-ops are fundamentally hybrid institutions; they are institutional centaurs. Our empirical overview has shown that they are defined by a commitment to balance in both structure and mission. In many ways the defining feature of a co-op is that it is a clever institutional design meant to balance efficient production with significant equality. Mission balance, therefore, is just as much a defining feature of co-ops as is the governance structure of “one person, one vote.”

The Standard Hierarchical View which dominates the social sciences is partially correct but partially wrong.

On the one hand, we have seen that it is true that there is a strong correlation between the size and the complexity of an organization; it is unlikely for an organization to be highly complex and also small because complex production typically requires layers of discrete specialization. It is also true that there is a hard tension between equal power and efficiency; it is functionally impossible for very large or complex organizations to allow equal decision-making power because complexity requires managers and planners with the power to give commands (as well as requiring that others obey such commands) in order for the various parts of the organization to work together effectively. This means that if one wants an organization to be complex and efficient, it must also involve pyramidal authority relations. It is also true that there is a tension (though a more malleable one) between equal pay and efficiency; it is very unusual, though not impossible, for complex organizations to provide all members with precisely equal pay since complexity requires specialization, and specialized workers, with more skill than others, often demand special treatment, especially if they can threaten to leave for better pay elsewhere.¹⁶ In sum, for an organization to be highly productive it must usually be complex, and for an organization to be complex it must also be large, governed via unequal decision-making power, and usually involve at least some degree of unequal pay.

On the other hand, it is completely false to conclude that the only way for firms to be large, complex, and efficient is for them to be hierarchical (with no democracy, starkly unequal wages, and rampant disrespect). In fact, democratic co-ops manage to pragmatically combine efficiency with substantial egalitarianism. Most co-ops are not governed through strictly equal power (medium and large ones are almost always pyramidal in structure), nevertheless, they also provide accountability on an egalitarian basis of one-person one-vote. They usually do not provide exactly equal pay, but often come quite close to it, with highly compressed pay scales, and very low levels of inequality compared to large capitalist firms. Co-ops also typically provide high levels of respect and job security.

Hence the Standard Hierarchical View that there is an inherent and intractable tradeoff between efficiency and equality is overblown. There is indeed a hard tradeoff between efficiency and one aspect of equality – equal power – but much less so in terms of remuneration or respect. Large efficient organizations cannot be based on equal power. But they can be based on accountable power, minimal differences in income, and equal respect. All in all, the Standard Hierarchical View is a lie wrapped in a truth. In conflating democratic co-ops with egalitarian collectives, the SHV unfairly tars the former with the limitations of the latter.

Both hierarchical firms and egalitarian collectives are deeply flawed institutions. Egalitarian collectives which prize equality, and hierarchical firms which prize efficiency, are nearly mirror opposites in terms of their virtues and vices. This is a visually striking result from Table 1. The reader should note that practically every strength of hierarchical firms is a weakness of egalitarian collectives, and vice-versa. An important conclusion, therefore, is that both of these types of “pure” firm have fundamental disadvantages. Conventional firms are deeply flawed, particularly in terms of domination, disrespect, inequality, and job insecurity. On the other side of the spectrum, egalitarian firms are deeply limited in their ability to grow.

All things considered, co-ops perform best. The most important finding of this study is that, as Table 1 makes clear, co-ops perform significantly better overall than the other two alternatives. Co-ops have more strengths and less weaknesses than the other types. This is mainly because they do a better job of balancing the fundamental goals of contemporary social life – efficiency and egalitarianism. On the basis of these results, it appears that co-ops, not hierarchical firms, should be the default form for organizing economic life (except in areas of high capital intensity where some form of codetermination may well be preferable (e.g., Ferreras et al., 2024).

One of the limitations of this paper is that since the bulk of the data comes from the Global North, one must be cautious about generalizing its conclusions to other parts of the world, particularly where institutions (such as the legal structuring of co-op firms) may be very different. Another limitation is that some of the data, especially in terms of the less-quantifiable features of working life – such as the degree of domination and disrespect – derive from small-n or qualitative studies.

Conclusion

The Standard Hierarchical View is the perspective that the only way to satisfactorily organize business entities in the modern world is through a hierarchical (i.e., undemocratic) governance structure because egalitarian alternatives are likely to be woefully inefficient or to degenerate. However, this paper has shown that the SHV is in need of serious revision. Conceptually, the SHV is incorrect, or at least overly general. There are not two possibilities (efficient-and-hierarchical or inefficient-and-egalitarian); a third possibility exists, that of democratic co-ops, based on representative democracy.

The main goal of this paper has been to carefully evaluate the real-world performance of these three types of economic organization. We find that both hierarchical firms and egalitarian collectives have severe drawbacks (domination for the former, severe difficulties in growing for the latter). The most important finding is that, contra the SHV, democratic co-ops, though far from perfect, generally do a better job of avoiding the pathologies of the other two forms. All things considered, they appear to be superior forms of social organization (at least for many parts of the economy). Democracy, not hierarchy, should become the default form of economic organization. Co-ops will not, however, spread automatically. An economy dominated by worker co-ops requires a conducive market environment, which must be built carefully and conscientiously. That will require significant costs; but overall, it appears that the costs are well worth it.

Footnotes

Acknowledgements

Many thanks to Fred Block, Gregory Dow, Mark Kaswan, Simon Pek, Marcelo Vieta, and the anonymous reviewers for their invaluable feedback on earlier drafts.

ORCID iDs

Tee Malleson

Yang-Yang Cheng

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Notes

References

Abebe

Gebreeyesus

(2023) How well do worker cooperatives perform in manufacturing? Evidence from a large low-income country. Economic Development and Cultural Change 72(4): 1993–2029.

Adeler

(2014) Enabling policy environments for co-operative development: a comparative experience. Canadian Public Policy 40(Suppl. 1): S50–S59.

Alchian

Demsetz

(1972) Production, information costs, and economic organization. The American Economic Review 62(5): 777–795.

Ammirato

(1996) La Lega: The Making of a Successful Cooperative Network. Dartmouth.

Anderson

(2017) Private Government. Princeton University Press.

Arando

Freundlich

Gago

, et al. (2010) Assessing Mondragon: stability and managed change in the face of globalization. William Davidson Institute Working Paper No. 1003, pp.1–47.

Arando

Gago

Jones

, et al. (2011) Efficiency in employee-owned enterprises: an econometric case study of Mondragon. IZA Discussion Paper No. 5711.

Arando

Gago

Podivinsky

, et al. (2012) Do labour-managed firms benefit from agglomeration? Journal of Economic Behavior & Organization 84(1): 193–200.

Atzeni

Ghigliani

(2007) Labour process and decision-making in factories under workers’ self-management: empirical evidence from Argentina. Work, Employment and Society 21(4): 653–671.

10.

Azzellini

Vieta

(2026) Commoning Labour and Democracy at Work: When Workers Take Over. Routledge.

11.

Bartlett

Cable

Estrin

, et al. (1992) Labor-managed cooperatives and private firms in North Central Italy: an empirical comparison. Industrial and Labor Relations Review 46(1): 103–118.

12.

Basterretxea

Martínez

(2012) Impact of management and innovation capabilities on performance: are cooperatives different? Annals of Public and Cooperative Economics 83(3): 357–381.

13.

Basterretxea

Charterina

Landeta

(2019a) Coopetition and innovation. Lessons from worker cooperatives in the Spanish machine tool industry. Journal of Business & Industrial Marketing 34(6): 1223–1235.

14.

Basterretxea

Heras-Saizarbitoria

Lertxundi

(2019b) Can employee ownership and human resource management policies clash in worker cooperatives? Lessons from a defunct cooperative. Human Resource Management 58(6): 585–601.

15.

Battilana

Yen

Ferreras

, et al. (2022) Democratizing work: redistributing power in organizations for a democratic and sustainable future. Organization Theory 3(1): 1–17.

16.

Bayo-Moriones

Galilea-Salvatierra

De Cerio

JMD

(2003) Participation, cooperatives and performance: an analysis of Spanish manufacturing firms. Advances in the Economic Analysis of Participatory and Labor-Managed Firms 7: 31–56.

17.

Berman

(1967) Worker-Owned Plywood Companies: An Economic Analysis. Washington State University Press.

18.

Bernstein

(1976) Workplace Democratization: Its Internal Dynamics. Kent State University Press.

19.

Blasi

Kruse

(2006) U.S. high-performance work practices at century’s end. Industrial Relations: A Journal of Economy and Society 45(4): 547–578.

20.

BLS (2016) Chart 3. Survival rates of establishments, by year started and number of years since starting, 1994–2015, in percent. Available at: https://www.bls.gov/bdm/entrepreneurship/bdm_chart3.htm (accessed 12 February 2026).

21.

Bradley

Gelb

(1981) Motivation and control in the Mondragon experiment. British Journal of Industrial Relations 19(2): 211–231.

22.

Bretos

Errasti

Marcuello

(2019a) Multinational expansion of worker cooperatives and their employment practices: markets, institutions, and politics in Mondragon. ILR Review 72(3): 580–605.

23.

Bretos

Errasti

Marcuello

(2019b) Is there life after degeneration? The organizational life cycle of cooperatives under a ‘grow-or-die’ dichotomy. Annals of Public and Cooperative Economics 91(3): 435–458.

24.

Burdín

(2014) Are worker-managed firms more likely to fail than conventional enterprises? Evidence from Uruguay. ILR Review 67(1): 202–238.

25.

Burdín

(2016) Equality under threat by the talented: evidence from worker-managed firms. The Economic Journal 126(594): 1372–1403.

26.

Burdín

Dean

(2009) New evidence on wages and employment in worker cooperatives compared with capitalist firms. Journal of Comparative Economics 37(4): 517–533.

27.

Carens

(1981) Equality, Moral Incentives, and the Market: An Essay in Utopian Politico-Economic Theory. University of Chicago Press.

28.

Chen

Glymour

Riley

, et al. (2021) Excess mortality associated with the COVID-19 pandemic among Californians 18–65 years of age, by occupational sector and occupation: March through November 2020. PLoS One 16(6): e0252454.

29.

Cheney

(1999) Values at Work: Employee Participation Meets Market Pressure at Mondragón. ILR/Cornell University Press.

30.

Corcoran

Wilson

(2010) The Worker Co-Operative Movements in Italy, Mondragon and France: Context, Success Factors and Lessons. Canadian Worker Cooperative Federation.

31.

Cornforth

Thomas

Lewis

, et al. (1988) Developing Successful Worker Co-operatives. Sage.

32.

Craig

Pencavel

(1995) Participation and productivity: a comparison of worker cooperatives and conventional firms in the plywood industry. Brookings Papers on Economic Activity. Microeconomics 1: 121–174.

33.

Cristini

Grasseni

Signori

(2023) Can worker cooperatives stabilize employment and remain financially sound during a prolonged recession? Applied Economics 55(33): 3863–3883.

34.

Dar

Getz

(2020) Changes in organizational behaviour – from traditional to renewed Kibbutzim. In: Ben-Rafael

Shemer

(eds) The Metamorphosis of the Kibbutz. Brill, pp.43–65.

35.

Defourny

(1992) Comparative measures of technical efficiency for five hundred French workers’ cooperatives. Advances in the Economics of Participatory and Labor-Managed Firms 4: 27–62.

36.

Dow

(2003) Governing the Firm: Workers’ Control in Theory and Practice. Cambridge University Press.

37.

Dow

(2018) The Labor-Managed Firm. Cambridge University Press.

38.

Dow

(2025) The formation and performance of labor-managed firms – an economic perspective. In: Warren

Biggiero

Hübner

, et al. (eds) The Routledge Handbook of Cooperative Economics and Management. Routledge, pp.34–45.

39.

Errasti

(2015) Mondragon’s Chinese subsidiaries: coopitalist multinationals in practice. Economic and Industrial Democracy 36(3): 479–499.

40.

Errasti

Bretos

Nunez

(2017) The viability of cooperatives: the fall of the Mondragon cooperative Fagor. Review of Radical Political Economics 49(2): 181–197.

41.

Estrin

Jones

(1998) The determinants of investment in employee-owned firms: evidence from France. Economic Analysis 1(1): 17–28.

42.

Fakhfakh

Magne

Mirabel

, et al. (2023) Employee-owned firms in France. Journal of Participation and Employee Ownership 6(2): 101–127.

43.

Fakhfakh

Pérotin

Gago

(2012) Productivity, capital, and labor in labor-managed and conventional firms: an investigation on French data. ILR Review 65(4): 847–879.

44.

Ferreras

Malleson

Rogers

(eds) (2024) Democratizing the Corporation: The Bicameral Firm and Beyond. Verso.

45.

Frega

(2021) Employee involvement and workplace democracy. Business Ethics Quarterly 31(3): 360–385.

46.

Furubotn

Pejovich

(1970) Property rights and the behavior of the firm in a socialist state: the example of Yugoslavia. Journal of Economics 30(3): 431–454.

47.

Greenberg

(1986) Workplace Democracy: The Political Effects of Participation. Cornell University Press.

48.

Gunn

(2006) Cooperatives and market failure: workers’ cooperatives and system mismatch. Review of Radical Political Economics 38(3): 345–354.

49.

Hansmann

(1996) The Ownership of Enterprise. The Belknap Press of Harvard University Press.

50.

Harnecker

(2007) Workplace democracy and collective consciousness. Monthly Review 59(6): 27–40.

51.

Heath

(2025) Are cooperatives more virtuous than corporations? Politics, Philosophy & Economics. Epub ahead of print.

52.

Hemingway

Pek

(2024) Expanding Democratic Employee Ownership in Canada: Policy Options. Canadian Centre for Policy Alternatives, UVicSpace.

53.

Hernandez

(2006) Striving for control: democracy and oligarchy at a Mexican cooperative. Economic and Industrial Democracy 27(1): 105–135.

54.

Hoffmann

(2001) Confrontations and compromise: dispute resolution at a worker cooperative coal mine. Law & Social Inquiry 26(3): 555–596.

55.

Holmstrom

(1989) Industrial Democracy in Italy. Avebury.

56.

Horrox

(2009) A Living Revolution: Anarchism in the Kibbutz Movement. AK Press.

57.

ICA (2018) Cooperative identity, values & principles. Available at: https://www.ica.coop/en/cooperatives/cooperative-identity (accessed 12 February 2026).

58.

Jäger

Noy

Schoefer

(2021) What Does Codetermination Do? Springer.

59.

Jaumier

(2017) Preventing chiefs from being chiefs: an ethnography of a co-operative sheet-metal factory. Organization 24(2): 218–239.

60.

Jayaraman

(2016) Forked: A New Standard for American Dining. Oxford University Press.

61.

Jones

(2007) The productive efficiency of Italian producer cooperatives: evidence from conventional and cooperative firms. Advances in the Economic Analysis of Participatory and Labor-Managed Firms 10: 3–28.

62.

Jones

Backus

(1977) British producer cooperatives in the footware industry: an empirical evaluation of the theory of financing. The Economic Journal 87(347): 488–510.

63.

Kasmir

(1996) The Myth of Mondragon: Cooperatives, Politics, and Working-Class Life in a Basque Town. State University of New York Press.

64.

Kaswan

(2019) Happiness theory and worker cooperatives: a critique of the alignment thesis. Journal of Labor and Society 22(3): 637–660.

65.

Kruse

Freeman

Blasi

(eds) (2010) Shared Capitalism at Work: Employee Ownership, Profit and Gain Sharing, and Broad-Based Stock Options. The University of Chicago Press.

66.

Leavitt

(2005) Top Down: Why Hierarchies Are Here to Stay and How to Manage Them More Effectively. Harvard Business School Press.

67.

Lomuscio

Tortia

Cori

(2023) Worker cooperatives in Italy: legislation, prevalence and recent trends. Journal of Participation and Employee Ownership 6(2): 128–148.

68.

Magne

(2017) Wage inequality in workers’ cooperatives and conventional firms 1. The European Journal of Comparative Economics 14(2): 303–329.

69.

Maietta

Sena

(2008) Is competition really bad news for cooperatives? Some empirical evidence for Italian producers’ cooperatives. Journal of Productivity Analysis 29(3): 221–233.

70.

Malleson

(2013) What does mondragon teach us about workplace democracy? Sharing ownership, profits, and decision-making in the 21st century. Advances in the Economic Analysis of Participatory & Labor-Managed Firms 14: 127–157.

71.

Malleson

(2014) After Occupy: Economic Democracy for the 21st Century. Oxford University Press.

72.

Malleson

(2023) Against Inequality: The Practical and Ethical Case for Abolishing the Superrich. Oxford University Press.

73.

Marx

([1894] 2010) Capital Volume III. Marxists.org.

74.

McDonalds (2024) Impact Report. Available at: https://corporate.mcdonalds.com/content/dam/sites/corp/nfl/pdf/McDonalds_PurposeImpact_ProgressReport_2023_2024.pdf (accessed 12 February 2026).

75.

Meyers

(2022) Working Democracies: Managing Inequality in Worker Cooperatives. ILR Press.

76.

Michels

([1911] 1962) Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy ( Paul

E. A. C.

, trans.). Free Press.

77.

Mirabel

(2021a) Past, present and future of empirical research on employee-owned firms: a structured literature review over 1970–2019. Journal of Participation and Employee Ownership 4(1): 1–25.

78.

Mirabel

(2021b) Survival advantage of worker buyouts over newly created worker-owned firms. Journal of Participation and Employee Ownership 4(3): 212–234.

79.

Mirabel

(2024) Do sunk costs nip worker-owned firms in the bud? Economic and Industrial Democracy. Epub ahead of print.

80.

Mirabel

Lomuscio

(2025) Explaining the rarity gap of worker cooperatives between France and Italy. International Review of Applied Economics 39(2–3): 306–330.

81.

Mishel

Kandra

(2021) CEO pay has skyrocketed 1,322% since 1978. Available at: https://www.epi.org/publication/ceo-pay-in-2020/ (accessed 12 February 2026).

82.

Monteiro

Straume

(2018) Are cooperatives more productive than investor-owned firms? Cross-industry evidence from Portugal. Annals of Public and Cooperative Economics 89(2): 377–414.

83.

Morrison

(1997) We Build the Road as We Travel. Essential Book Publishers.

84.

Moye

(1993) Mondragon: adapting co-operative structures to meet the demands of a changing environment. Economic and Industrial Democracy 14(2): 251–276.

85.

Mygind

(1987) Are self-managed firms efficient? The experience of Danish fully and partly self-managed firms. Advances in the Economics of Participatory and Labor-Managed Firms 2: 243–323.

86.

Nembhard

(2021) Collective Courage: A History of African American Cooperative Economic Thought and Practice. Penn State University Press.

87.

Olsen

(2013) The relative survival of worker cooperatives and barriers to their creation. In: Kruse

(ed.) Sharing ownership, profits, and decision-making in the 21st century. Emerald Group Publishing Limited, pp. 83–107.

88.

Oxfam (2016) No Relief: Denial of Bathroom Breaks in the Poultry Industry. Oxfam.

89.

Palgi

(2006) Pitfalls of self-management in the kibbutz. Revue Internationale de Sociologie [International Review of Sociology] 16(1): 63–77.

90.

Pencavel

Craig

(1994) The empirical performance of orthodox models of the firm: conventional firms and worker cooperatives. Journal of Political Economy 102(4): 718–744.

91.

Pencavel

Pistaferri

Schivardi

(2006) Wages, employment, and capital in capitalist and worker-owned firms. ILR Review 60(1): 23–44.

92.

Pérotin

(2006) Entry, exit, and the business cycle: are cooperatives different? Journal of Comparative Economics 34(2): 295–316.

93.

Pérotin

(2014) Worker cooperatives: good, sustainable jobs in the community. Journal of Entreprenurial and Organizational Diversity 2(2): 34–47.

94.

Pérotin

(2016) What do we really know about workers’ co-operatives? In: Pérotin

(ed.) Mainstreaming Co-operation. Manchester University Press, pp.239–260.

95.

Pettit

(2012) On the People’s Terms: A Republican Theory and Model of Democracy. Cambridge University Press.

96.

Pinto

(2024) Economic democracy against racial capitalism: seeding freedom. In: Ferreras

Malleson

Rogers

(eds) Democratizing the Corporation: The Bicameral Firm and Beyond. Verso.

97.

Podivinsky

Stewart

(2007) Why is labour-managed firm entry so rare?: an analysis of UK manufacturing data. Journal of Economic Behavior & Organization 63(1): 177–192.

98.

Podivinsky

Stewart

(2009) Modeling proportions: random effects models of UK firm entry. The Singapore Economic Review 54(3): 367–377.

99.

Prushinskaya

(2020) Worker Co-ops Show Significant Growth in Latest Survey Data. Available at: https://www.fiftybyfifty.org/2020/02/worker-co-ops-show-significant-growth-in-latest-survey-data/ (accessed 12 February 2026).

100.

Restakis

(2010) Humanizing the Economy: Co-operatives in the Age of Capital. New Society Publishers.

101.

Reuten

(2022) The mondragon worker cooperatives’ employment record 1983–2019. Journal of Labor and Society 26(3): 336–375.

102.

Roemer

(2013) Thoughts on arrangements of property rights in productive assets. Analyse & Kritik 35(1): 55–64.

103.

Rothschild

Whitt

(1986) The Cooperative Workplace. Cambridge University Press.

104.

Santos

Guzmán

Valiente

(2024) Entrepreneurship and innovation in worker cooperatives and conventional firms: the role of external cooperation. Small Business Economics 64(3): 1415–1431.

105.

SBECouncil (2018) Facts & Data on Small Business and Entrepreneurship. Available at: https://sbecouncil.org/about-us/facts-and-data/ (accessed 12 February 2026).

106.

Schoening

(2010) The rise and fall of burley design cooperative. Oregon Historical Quarterly 111(3): 312–341.

107.

Silva

Morello

(2021) Is there a trade-off between efficiency and cooperativism? Evidence from Brazilian worker cooperatives. Journal of Co-operative Organization and Management 9(2): 100136.

108.

Smith

(2001) Blooming Together or Wilting Alone?: Network Externalities and Mondragon and La Lega Co-operative Networks. World Institute for Development Economics Research.

109.

Storey

Basterretxea

Salaman

(2014) Managing and resisting ‘degeneration’ in employee-owned businesses: a comparative study of two large retailers in Spain and the United Kingdom. Organization 21(5): 626–644.

110.

Stryjan

(1994) Understanding cooperatives: the reproduction perspective. Annals of Public and Cooperative Economics 65(1): 59–80.

111.

Tannenbaum

Kavcic

Rosner

, et al. (1974) Hierarchy in Organizations: An International Comparison. Jossey-Bass Publishers.

112.

Terrasi

Hyungsik

(2017) Industrial and Service Cooperatives: Global Report 2015–2016. International Organization of Industrial and Service Cooperatives.

113.

Thomas

Logan

(1982) Mondragon: An Economic Analysis. G. Allen & Unwin.

114.

Vanek

(1970) The General Theory of Labour-Managed Market Economies. Cornell University Press.

115.

Vieta

(2020) Workers’ Self-Management in Argentina. Brill.

116.

Vieta

Depedri

Carrano

, et al. (2017) The Italian road to recuperating enterprises and the Legge Marcora framework. Euricse Research Report No. 15.

117.

Walmart (2025) How many people work at Walmart? Available at: https://corporate.walmart.com/askwalmart/how-many-people-work-at-walmart (accessed 12 February 2026).

118.

Webb

(1907) The History of Trade Unionism. Longmans, Green & Co.

119.

Whyte

(1988) Making Mondragon: The Growth and Dynamics of the Worker Cooperative Complex. ILR Press.

120.

Wilkinson

Donaghey

Dundon

, et al. (2020) Handbook of Research on Employee Voice. Edward Elgar Publishing.

121.

Williamson

(1973) Markets and hierarchies: some elementary considerations. The American Economic Review 63(2): 316–325.

122.

Young-Hyman

Magne

Kruse

(2023) A real utopia under what conditions? The economic and social benefits of workplace democracy in knowledge-intensive industries. Organization Science 34(4): 1353–1382.

123.

Zamagni

(2015) The cooperative movement. In: Jones

Pasquino

(eds) The Oxford Handbook of Italian Politics. Oxford University Press, pp.554–566.

124.

Zanini

(2022) Most people don’t hate their jobs, but they definitely hate how they’re managed. Available at: https://www.michelezanini.com/most-people-dont-hate-their-jobs-but-they-definitely-hate-how-theyre-managed/ (accessed 12 February 2026).