Workplace Democracy Democratized: The Case for Participative and Elected Management

1 Introduction: The Blind Spot of Work Organization in Worker Cooperatives

This discussion of the methods of choosing leaders under a democratic industrial system may seem to be somewhat dull and detailed; but it is one of the fundamental problems of Guild Socialism. For the most frequent argument urged against industrial democracy is that it is incompatible with workshop discipline and productive efficiency. (Cole 1920, 55)

Workplace democracy defines “systems for the exercise of power by workers or their representatives” (Frega, Herzog, and Neuhäuser 2019, 1). Worker cooperatives – organizations owned and governed by their workers or their representatives – are often viewed as the most ‘complete’ embodiment of workplace democratization.^[1] Unlike co-determination or bicameral models (Ferreras 2017), workers there hold the majority of voting power, allowing them to both own and manage their enterprise. While traditional firms operate on the capitalist principle of political power as ‘one share, one vote’, worker cooperatives adhere to the democratic principle of ‘one person, one vote’. During general assemblies, which typically take place once a year, workers exercise several key rights. First, they elect the general manager (in larger cooperatives they elect the board of directors, which in turn, elect the general manager). This is what allows worker cooperatives to represent an alternative to the traditional issue of domination: because the general manager is elected (and can be removed at any time), workers are less likely to be subjected to unjust budget cuts, unjustified relocations, or downsizing strategies. In this way, they have a voice, albeit indirectly, in the company’s general policies, the horizon line the ship is following. Second, they approve the budget, vote on profit distribution, and make ‘major’ economic decisions, such as relocating the company or merging with another firm. As co-owners of the firm, workers indeed decide what to do with the profits each year (mainly whether to redistribute them as compensation or reinvest them). This is what allows worker cooperatives to avoid the traditional issue of exploitation: by being their own employers, workers sidestep the exploitation system rooted in the division between owners and workers.

Yet, since their emergence in the 19th century, worker cooperatives have maintained a significant blind spot – one now shared by their advocates of workplace democracy: while they focus on democratizing governance, they often overlook a crucial locus of power – the daily organization of work itself. Indeed, workers have no say over the management team. Its members are typically appointed by the elected general manager or by the elected board of directors without their input or involvement – a bit like how the government is appointed by the elected president, without the citizens having a say in it. I propose to call this neglected problem the hourglass paradox. On the strategic governance side, the structure is democratic: employees are members and participate in major strategic decisions. In contrast, on the management side, the executive team directs and implements actions, with decisions cascading down to middle and line management and, ultimately, to the workers. By default, the power structure of a worker cooperative thus takes the form of an hourglass.^[2]

In this paper, I will argue that between the decisions made by a cooperative’s general manager and the workers themselves lies a broad spectrum of intermediate decision-making, carried out by middle and line management – such as team leaders, supervisors, superintendents, foremen. I will argue that this intermediary layer plays a crucial role in shaping the daily organization of work and, as we will see, in ruling on ethical issues. Yet, it remains an often-overlooked site of power within worker cooperatives, seldom the focus of empirical studies, which tend to focus instead on democratic tensions at the level of governance.^[3] What we do know about the everyday practices of worker cooperatives managers is that they tend to be significantly more tolerant of minor infractions than their counterparts in traditional firms (e.g. Bellas 1972; Greenberg 1986; Sobering 2022, 79). This tolerance is partly facilitated by stronger peer surveillance mechanisms among workers at the same hierarchical level – a phenomenon Katherine Sobering refers to as ‘lateral management’ (Sobering 2022, chap. 3). However, there remains a lack of information on how managers (or the administrative council, as in the Hotel Bauen case studied by Sobering) make key decisions regarding work organization, particularly in areas such as working hours, task allocation, recruitment, and layoffs. While it appears that these decisions are often made by default without consulting employees, it would be valuable to investigate whether such ‘default management’ in cooperatives is indeed as top-down as it seems.

The fact that middle and line management’s decisions are an often-overlooked site of power in worker cooperatives is problematic, as the most pressing injustices for workers stem not from the owners’ overarching control but from the daily organization and evaluation of their work. What is the point of lengthy daily discussions with their relatives at diner, what keeps workers awake at night is less the fact of being exploited than being subjected to relentless work rhythms, having their opinions on how to perform their tasks properly disregarded or mocked, being treated unfairly compared to their colleagues, being sidelined, subjected to everyday harassment, and moreover forced to suppress their anger. Of course, exploitation is a problem that must be addressed on its own. But the way managers can exercise their power unjustly is a matter of primary importance. Worker cooperatives are promising models for avoiding exploitation, domination and other unjust social dynamics. But how can they represent also a model that goes against the objectionable power exercised on a daily basis by those in charge on organizing every-day work, namely managers?

This hourglass paradox presents a challenge to the attractiveness of the political program of workplace democracy. Strategically speaking, its neglect has led, over the past 50 years, to this program being overshadowed on its ‘right wing’ by participative management – a form of worker empowerment in their tasks, implemented in capitalist firms such as quality circles. In participative management, however, this autonomy is quite limited: while workers may have the freedom to organize themselves as they see fit, it is still within the framework of production objectives set for them in advance, over which they have little room for maneuver (Rothschild 2000, 197–202). In other words, participative management democratizes management, but not governance; whereas worker cooperatives democratize governance, but not management. In this paper, in line with Deranty and Renault’s (2020) critique of workplace republicanism, I will argue that a comprehensive and complete theory of workplace democracy should aim to democratize both spheres of power: governance as well as management.

Indeed, decisions regarding the content and organization of work involve ethical, political, and cultural considerations, and, as such, workers should have a say in them. Therefore, I will argue that the primary responsibilities of managers should include facilitating and fostering democratic deliberation on work-related issues, supporting coordination efforts, and providing assistance to workers. Finally, I will contend that managers themselves should be elected and held accountable by the workers, rather than being appointed by the cooperative’s general manager or board of directors.

2 What’s the Point of Management in a Democratic Firm?

The first reason worker cooperatives may require management hierarchies is the same as in any other organization: coordination. As a firm’s operations grow in scale and complexity, the amount of information to process exceeds what any individual can manage alone, making coordination through managerial roles necessary. According to the neo-classical theory of the firm, management exists to ensure that employees fulfill their contractual obligations in alignment with the company’s broader objectives. Unlike the price mechanism that governs market exchanges, firms require internal coordination among agents to produce goods or services (Coase 1937). Hierarchies of authority emerge to define individual roles clearly and ensure compliance.

But one important difference between democratic and non-democratic firms is that, in the latter, management is primarily designed to meet ambitious targets and sustain high work rhythms – most often driven by the goal of maximizing output while minimizing costs. This is even the case in public organizations, often undermined by public management practices. In worker cooperatives, by contrast, it is common for workers to agree that the only legitimate pace of work is the one that ensures everyone receives a fair wage. Thus, empirical studies show that workers are in general more engaged within democratic organizations, which could explain why they tend to be more sustainable in the long run (Pérotin 2016). Therefore, “statistics for firms before and after democratization generally show declines in purposefully poor work, sabotage, absenteeism, wildcat strikes, and other disciplinary problems (Potvin 1958; Blumberg 1968, chap. 5)” (Bernstein 1976, 86).

Some worker cooperatives have clearly adopted a self-management model, where they simply try to do away with middle and line management (Rajan and Bino 2021; Rothschild and Whitt 1986). Participatory meetings are almost daily, and no one is specifically in charge of organizing them or ensuring that decisions are properly followed through. However, it is possible that some workers may prefer a form of workplace democracy that demands less time commitment. Additionally, it seems that, beyond a certain company size, there are clear advantages to incorporating management into the organizational structure. Thus, the primary function of management^[4] is that of coordination, aimed at achieving a goal established either collectively or by the general manager, who serves as the representative of the workers.

Actually, most political philosophers who have addressed the topic of the enterprise agree on the legitimacy of establishing hierarchies within the organization of work, specifically for coordination purposes. However, they also acknowledge that this issue presents a significant challenge for political philosophy. For instance, Elizabeth Anderson asks: “Can egalitarians offer a systematic analysis of the proper content and scope of command relations in the domain of work (including the administrative branches of government), acknowledging that, given the scale and complexity of modern production, some form of command hierarchy is necessary?” (Anderson 2015, 150) as for the representatives of French critical theory of work, they assert:

The management issue has always been taboo within the left. But if one takes seriously the fact that work has to be managed, no program of social transformation, be it reformist or radical, could be consistent or plausible without tackling the management issue. (Dejours et al. 2018, 165, my emphasis)

In their various publications, Christophe Dejours (a psychoanalyst specializing on work), and the philosophers Emmanuel Renault and Jean-Philippe Deranty highlight that even in Marx’s writings, there is the recognition that hierarchical coordination is essential to ensure the smooth functioning of collective labor. He famously likens the manager to a conductor, whose role is to prevent one musician’s performance from interfering with another’s: “A single violin player is his own conductor: an orchestra requires a separate one” (Marx 2004, 1, 448–9 cited by Dejours et al. 2018, 219). As we see it, it appears to be a broad consensus that hierarchical authority is both necessary and legitimate. Yet philosophers express concerns about the potentially harmful effects these hierarchies can produce.

3 The Degeneration of Workplace Hierarchies: Domination and Social Hierarchy

The problem emerges when the inherent authority hierarchies within the workplace morph into oppressive social hierarchies. Continuing a long-standing tradition of republican thought that critiques the autocratic tendencies of workplace organization, Anderson emphasizes the extensive privileges and powers granted to executives and business owners (Anderson 2008, 2015, 2017). As she explains, the relationship between a manager and their subordinate – legally defined as one of subordination – creates a profound vulnerability that labor law often fails to address adequately. She illustrates this point with a striking example: a CEO who allegedly required his secretaries to shape their hands into ashtrays so he could drop the ashes of his cigar as he passed by (Anderson 2008, 146). While such behavior might qualify as workplace harassment under the legal frameworks of certain countries, it is undeniable that many forms of managerial abuse, however immoral, remain perfectly legal.

This power imbalance leaves workers effectively ‘at the mercy’ of their employers’ goodwill. In the absence of countervailing powers – such as robust labor laws or strong union representation – there is a genuine risk that ‘hierarchies of command’ will degenerate into ‘hierarchies of standing’, where, as Anderson puts it, “those at the top are entitled to make claims on others in their own right, and to enjoy rights and privileges, while those below are denied rights or granted an inferior set of rights and privileges, and denied voice to make claims on their own, or given an inferior forum in which to make their claims” (Anderson 2015, 144).

But hierarchies of command can also degenerate into another problematic kind of hierarchies: the one that Anderson calls ‘hierarchies of esteem’.^[5] As Néron explains, authority hierarchies in companies rely on self-legitimating discourses that have a self-devaluing effect on workers (Néron 2015a, 2015b, 2024). By arguing that – in a Foucauldian vein – managerial authority “typically goes hand in hand with the expression of a wide range of attitudes and forms of treatment incompatible with the ideal of social equality” (Néron 2015b, 325), Néron answer to Anderson’s worry in two ways: 1. Yes, command hierarchies do frequently transform into hierarchies of standing and esteem: 2. Command hierarchies in the workplace obviously enhance workers’ feelings of their lower value.

Does this mean that egalitarians must give up on the question of organizing work? Would it be possible to set limits on hierarchies of command to prevent them from degenerating into privileges and social disdain? While Schuppert’s analysis has the merit of highlighting the hierarchical relationship between managers and workers – as an “almost unavoidable part of contemporary division of labor” (Schuppert 2015, 112) – he does not seem to move beyond the broad call for “democratic procedures in relevant decision-making [that] would give each and every member of a corporation a fair share in control and influence” (2015, 117). What is more, this share of control and influence appears to be limited to “major decisions” and “vast differences in power” (2015, 117).^[6]

As for Anderson and Néron, they propose making it explicit and transparent that the hierarchy of command is rooted in a hierarchy of competence (even though Néron, as Schuppert, clearly argues in favor of workplace democracy). Elizabeth Anderson outlined how this idea of a hierarchy of competence could be institutionalized in her 2008 paper that fell far short of receiving the same level of attention as her later, renowned book, Private Government, even though the argument she made in it was probably more original. In this article, Anderson contends that bureaucracy could be “a vital tool in the egalitarian toolbox” (Anderson 2008, 159). This idea will be explored further in the next section, before showing that it is not without its challenges.

4 The ‘Back to Bureaucracy’ Solution: A More Efficient and Egalitarian Organization of Work?

To begin, bureaucracy has two advantages: through the implementation of procedures, it is driven by “an efficiency principle” (Anderson 2008, 157) and it makes the arbitrary use of power impossible. Anderson draws on the opposition between the ideal types proposed by Max Weber, specifically between bureaucracy and patrimonial domination. In patrimonial domination, legitimized by reference to tradition, the relationship between the one who obeys and the one who commands is one of dependency. The one who commands “pledge their personal loyalty and submission” (Anderson 2008, 151). Furthermore, “cronyism, not competence, determines subordinates’ access to such positions” (Anderson 2008, 151). In contrast, bureaucracy is based on competence and qualifications as the legitimate criteria for access to power and position. You got the job because you are the most competent for it, not because you get along well with the boss. Another defining feature of bureaucracy is that the rules are not decisions taken on the spur of the moment by the powerful, but are permanent, written, accessible and therefore predictable rules. Bureaucratic roles do not confer special treatment or privileges. They ensure in theory equal treatment for all. The limited scope for discretion and the predictability of the rules make arbitrary behavior impossible, thus reducing the prevalence of actions like brown-nosing. Empirical studies support this thesis: sociologists have shown that many of the issues attributed to it actually arise from the obstruction of its established rules and procedures (Gay 2000). And a recent study also suggests that women tend to feel more comfortable in bureaucratic organizations, viewing them as more efficient and egalitarian, while men are more likely to perceive them as overly controlled and regulated (DeHart-Davis 2009). Therefore it is quite likely that hierarchical relationships within a bureaucratic democratic workplace pass Philip Pettit’s famous ‘Eyeball Test’ for identifying interpersonal domination, which involves determining whether one can look another in the eye without fear or deference caused by the potential for interference (Pettit 2012, 84–7).

So, what we would have here is a democratic workplace in which managers make decisions competently, without placing their subordinates in a position of vulnerability. However, Anderson herself acknowledges certain limitations of this model, highlighting two factors that should temper – without entirely extinguishing – the egalitarian enthusiasm for it: democracy and “dignitary and humanistic constraints” (Anderson 2008, 158). The objections she raises actually resonate with the numerous critiques that came from the left against this principle, which have, in fact, led many worker cooperatives to build themselves in radical opposition to the bureaucratic model (Rothschild and Russell 1986; Rothschild 2000).

The first limitation of bureaucracy lies in its potential conflict with the principle of democracy, as Max Weber himself recognized – thereby continuing the tradition that portrays bureaucracies, on the whole, as a negative force in society. In her article, Anderson appears to suggest that bureaucracy may be problematic when the decisions it enforces are not themselves democratically determined – such as in the case of a dictatorship – or when they lead to policies that are democratically questionable.^[7] An interesting objection to the idea of a technical hierarchy in worker cooperatives is that it reinforces information asymmetry among workers, making it easier for managers – who are perceived as experts – to capture political power. This issue is indeed a well-known challenge in worker cooperatives (Bernstein 1976, 73; Meister 1964; Cornforth 1995).

As we explained, bureaucracy is characterized by “the rule of law, tied to a principle of efficiency. The powers of office are not arbitrary but limited by rules rationalized by institutional rather than personal objectives” (Anderson 2008, 155). But the other side of the coin is that the efficiency principle itself can reduce individuals to a lower status of being, treating them impersonally as interchangeable “cogs in a machine”, as Anderson (2008, 157) herself phrases it. She also notices the potentially impoverishing effects bureaucracy can have on the nature of work itself, as it may reduce the richness and variety of tasks to rigid, standardized procedures.^[8] A situation that occurred within the world’s largest cooperative group powerfully illustrates both aspects of this issue.

5 When Bureaucratic Efficiency Threatens Workers’ Dignity

In the late 1980s, the Mondragon cooperatives reorganized themselves into the Mondragón Cooperative Corporation (MCC), which was designed to reflect “the need to achieve maximum business efficiency in the new competitive context of Europe”, as anthropologist Sharryn Kasmir studied (Kasmir 1996, 180). From this point onward, several Mondragon cooperatives began implementing new models of work organization. First, “shifts were introduced to amortize investments in new machinery. Two and three shifts were scheduled to keep machines running 24 hours a day, seven days a week […]” (Kasmir 1996, 181). Second, just in-time production was introduced, which “sets work schedules according to incoming orders, requiring workers to stay late some days while they are sent home early other days, without prior notice” (Kasmir 1996, 181). Third, consultants were brought in to redesign production. Kasmir recounts an exchange on this topic with Kris, an employee working in the preassembly of washing machine parts in Fagor Garagartza:

as a result of the consultants’study, she [Kris] was now told how to coordinate the movement involved in her preassembly task, and she was expected to do her work faster. I met her after work one day, and she told me that she was nervous all the time. Her hand shook as she lit her cigarette. (Kasmir 1996, 183)

The most detrimental changes, however, occurred, rather paradoxically, in the same plant that had implemented participative management norms in the 1970s – Fagor Copreci, a plant specializing in the manufacture of fittings for copper tubing – where Methods-Time Measurement (MTM) was enforced. “Jobs were then redefined,” explains Kasmir, “to cut out what were considered to be inefficient movements (it was determined that there were 12 essential movements in production and that all other movements were unnecessary) […]” (Kasmir 1996, 187). With the new standard, output increased by 25 %, but workers complained that the rate was exhausting and stressful. She recounts:

A KT member [a small group of Mondragon’s militant workers] told me that in Copreci: “ten to fourteen have fallen ill with hysteria, depression, nerves, vomiting. They are treating us like chickens in a coop.” (Kasmir 1996, 187)

As we see, this case offers a striking example of the negative outcomes that can arise when the political question of labor is not addressed.^[9] It seems that it would have taken reaching such extremes before this managerial policy was eventually abandoned (though it is unclear how this actually occurred). This – quite dramatic – case demonstrates that relying on the principle of efficiency, which lies at the heart of bureaucracy, is not without risks for a democratic organization. Its supposedly ‘objective’ and impersonal approach to organizing work can, in fact, undermine individual dignity.

Nevertheless, there is a more fundamental limitation to the bureaucratization of hierarchical relations in the workplace, one that Anderson doesn’t seem to address: that, contrary to what she and Max Weber himself thinks, it could be, at the end, inefficient. For understanding why, we need to go deeper in the nature of work, of real work, or, as critical theory calls it ‘living labor’ (i.e. labour-power in action, according to Marx (Capital, t. 1)).

6 How Work Gets Done: An Epistemic Critique of Vertical Management

In the 1980s, participative and horizontal management emerged as the new trend in work organization, following the decline of Taylorism and Fordism. The Human Relations school of thought not only recognized the crucial role of employee motivation but also highlighted the importance of workers’ interpersonal skills. From a strategic perspective, this shift stemmed from a move away from mass production of standardized products toward the creation of a diverse range of highly personalized products designed to meet the changing tastes of the emerging consumer society, as well as the tertiarization of the economy. To enhance quality and avoid excess inventory or breakdowns, Lean Management relied typically on employee responsibility, particularly through quality circles – volunteers from all levels who discuss ongoing or upcoming projects. The reason these ‘participative’ management trends have enabled capitalism to thrive for nearly four decades is that they grasped something essential: effective work is not about rigidly applying the same method over and over again.

Indeed, work is essentially an improvisation aimed at achieving a goal. Ergonomics provides a useful concept for understanding this: the distinction between prescribed work (tasks as formally defined) and real work (how tasks are performed in practice). The prescribed work refers to what is asked of the employee. It represents the expectations, which are formalized through instructions and objectives (requests, procedures, directives, protocols, codes, programs, and other guidelines imposed by the hierarchical superior). Merely following instructions to the letter is not sufficient to meet the expected outcome. In practice, an employee may realize that the situation does not precisely align with the prescribed guidelines. Based on their experience or capabilities, they must make adjustments or modifications in order to effectively execute the tasks at hand. That is why real – the work truly performed by the employee – is far more complex than prescribed work. It encompasses all of the actions undertaken and strategies employed by the employee to complete their tasks. Constraints may be material, technical, organizational, etc. and often require the worker to adapt and adjust their approach. For example, if I am a nurse and need to perform blood draws on two different patients, the task is ostensibly the same, and it should take me the same amount of time. However, patients can be quite different, and the conditions under which I perform the blood draws may not be identical. A contrasting example of this can be seen in the phenomenon of go-slow strikes, where workers follow instructions or regulations to the letter, thereby intentionally slowing down the productivity of the organization.

Now, what holds true for individual activity, as Dejours explains, also applies to collective work (Dejours 2013b, chap. 3). Group collaboration – what he refers to as cooperation – also requires departing from established rules. It often involves improvising according to the circumstances, occasionally bending procedures, and exceeding narrowly defined roles. In group cooperation, if each worker merely follows their prescribed tasks to the letter, true collaboration becomes impossible, as not all situations can be anticipated in advance. For instance, it is common for an employee to cover for a sick colleague to ensure that an urgent task is completed, even if it falls outside their official duties.

However, to adapt to unforeseen circumstances and perform their tasks effectively, workers need to maintain a comprehensive understanding of the broader operations in which their work is embedded. Successful cooperation among individuals requires knowing who is responsible for what – and also recognizing when someone is failing to do its part of the job. In fact, this is exactly what workers do intuitively in their daily activities to perform their tasks in the most effective and appropriate way. As Deranty and Renault point out, this often leads workers to develop a more profound and practical knowledge of their own and their work and their colleagues’ work than their managers do:

Only the people who actually engage in the productive process and who actually produce the outputs know what it takes to deliver them. For many tasks, this critical insight can be minimal, differing only infinitesimally from what engineers, coders, and managers have put in place and think is actually happening in their workplace. For other tasks, this critical insight can be much more substantive. (Deranty and Renault 2020, 152)

Consequently, as Deranty and Renault argue, “workers are always in a critical position in relation to management” (2020, 152). This position is in fact primarily epistemic. It states that managers lack an accurate understanding of the actual work performed by employees. Instead, they often hold a distorted, idealized view of it, which is frequently trivialized or undermined:

From the perspective of these individuals and collectives, the discourse of expertise, which is one of the strongest justifications for hierarchical power, often proves to be seriously out of touch with the reality of production. Workers frequently complain that external models of management are superimposed onto productive processes without consideration of what is really required. In such situations, the experts, technicians, and engineers who tell workers what to do from on high might benefit from organizational power to enforce their decisions, but they lack the actual productive justification for their alleged expertise. (Deranty and Renault 2020, 159)

Bureaucracy, in this light, can be seen as a rather ineffective form of work management, and actually has been criticized on this front since the 1960s. The perspective we propose to adopt here – following Dejours, Renault and Deranty’s insights – brings about a significant shift in the contemporary understanding of work, offering a more realistic and collective view of labor.

7 Fostering Collective Decision-Making and Support: Two Key Functions of Democratic Management

What type of management should then be envisioned? Of course, the relatively autonomous decision-making of the manager may be necessary and desirable for large-scale coordination tasks, particularly when employees lack visibility and do not desire it, such as in the coordination between different teams (Deranty and Renault 2020, 160). But in fact, it seems legitimate that much of the organization of work within work teams should be discussed among the employees themselves. It is indeed certainly both possible and desirable to ensure that guidelines are developed by the employees themselves, or at the very least, presented to them before being implemented, as suggested by Deranty and Renault (2020, 160). In any case, they should remain open to discussion and revision by the employees. “Managers’ task may be more appropriately seen as examining the costs ans benefits of competing otions, which they present to the membership for decisions”, as Cornforth, Thomas, Lewis and Spear briefly suggest (Cornforth et al. 1988, 153). The democratization of management in democratic workplaces would therefore involve the democratization of power beyond the sole mode of governance of institutions; it would extend to “various power relations and hierarchies operating within them” (Deranty and Renault 2020, 161). The democratization of management could also provide an opportunity to revalue group work, countering the trends toward the individualization of labor, and to enrich intellectually unengaging and monotonous tasks by incorporating responsibilities previously reserved for other positions within the company – both of which are advocated by Axel Honneth (2024, 211–4).

Thus, by reusing part of the typology proposed by Dejours et al., it can be argued that the primary function of a manager would be to establish, facilitate, and sustain such spaces for deliberation over time. I would even add that, in a sense, the manager should act as a support for the work collective, a mediator enabling the collective to meet, express itself, sustain the exchange of viewpoints, and ultimately have the courage to collectively arrive at a decision. The manager’s role would be to ensure that the conditions for deliberation within the work collective are met, within the framework of what Dejours et al. call ‘cooperative management’ (2018). These conditions are spatial, temporal, and psychological, as Dejours aptly explains (2013, 81–119; Dejours et al. 2018, 168–70). The spatial and temporal aspects are particularly crucial: the manager must ensure the preservation of both formal deliberative spaces (such as meetings) and informal ones (such as cafeterias, lockers, kitchens, restrooms, and social areas). Psychological conditions are equally crucial: deliberation among workers entails some risks for each individual worker, as Dejours points out. What risks? By explaining their work to others, a worker faces several risks: having their ‘tricks’ stolen, exposing their imperfections and mistakes, and, most importantly, being criticized for failing to comply with official rules. Such transparency places the worker in a vulnerable position. This explains why many team meetings are characterized by an almost palpable silence – the fear that speaking up might backfire on workers in the end is pervasive. Therefore, the manager must ensure that those risks are mitigated: neither the workers nor the superiors will use this information against them (Dejours et al. 2018, 170).

When workers reach an agreement on the procedures they wish to implement, this results in the establishment of ‘work rules’, Dejours explains. When several such rules are formulated, they are referred to as ‘rules of the trade’. This collective process of rule-making is known as “deontic activity” (Dejours et al. 2018, 169–70; Dejours 2013b, chap 3). Creating spaces for deliberation where each worker can articulate and defend their perspectives on their methods of working – their modus operandi – as well as their interpretations of objectives and directives, is crucial. Workers should have the opportunity to reflect on and debate what is acceptable, what is more or less effective, and what should or should not be prohibited.

As we will see, and this is an element we will return to shortly, the viewpoints expressed are not limited to the most ‘technical’ way of performing a task. In fact, any theorization of work involves normative judgments – judgments of values and aesthetics, as Dejours states – on what constitutes “well-done” work, and what is a “beautiful” and “good” work (Dejours 2013a, 44–6; 2006). There is no ‘one best way’ at work. The definition of what constitutes ‘good work’ is fundamentally tied to values beyond mere efficiency (Dejours et al. 2018, 168). Thus, one essential function of a good manager, as I propose in this context, is to analyze the ongoing debates and, through conceptualization and reflection, to highlight for the workers that the decisions being made in these deliberations involve ethical, political, and cultural choices. This is an argument I insist upon, as I believe it is crucial to recognize and attribute this role to the manager, who plays a key part in guiding the collective decision-making process by bringing these broader dimensions into focus. The manager must remain attentive to this and assist the collective in considering the different options available for what they truly are: choices that engage these ethical, political, and cultural dimensions. Furthermore, it is quite likely that the personal interests or preferences of managers may disrupt the proper course of deliberation. In her study of an urban crisis center (Helpline), which was run in a self-managed manner, Mansbridge demonstrated that during deliberation periods, the ‘representatives’ designated by each of the six service groups often struggled to faithfully advocate for the perspectives they were meant to represent:^[10] “I recorded several instances in which the representatives, with the best will in the world, did not adequately communicate both sides of an argument to the service groups or report all the dissents and positions in the service group back to the CPC [Community policy meetings] or DPW [Daily, Personnel and Weekly policy meetings].” (Mansbridge 1983, 213) There is clearly room for discretion in how different options are presented and in whether or not relevant information is shared to facilitate an informed discussion. This presents a challenge for workplace democracy, one that requires vigilance and, perhaps, the implementation of counterbalances – such as hiring an external consultant or even reinvigorate union support in this context to assist in certain decision-making process.

The default mode of decision-making in this deontic activity is consensus. Indeed, Dejours acknowledges having drawn on Habermas’s theory of communicative action to develop his model (Dejours 2013b, chap. 3). However, there may be times when consensus cannot be reached. We can reasonably expect, and hope, that these situations will be infrequent. But in these cases, it is the manager’s role to “arbitrate”, Dejours et al. argue (Dejours et al. 2018, 170). This point, in my view, deserves more attention than it has received. First, one might ask whether we should also allow for procedures to achieve proportional outcomes – such as bargaining, side payments, taking turns (‘A wins today, B tomorrow and C the next days’) or the division of distributable goods (see Mansbridge 1983, 265 sqq.) – as well as voting, where the majority decide. After all, those procedures are traditional in democratic decision-making process, even though they each have their flaws. Why would it be more legitimate for a manager, who, unlike a director, is not elected, to make the final decision? However, one potential issue with voting, even in secret ballots, is that it may breed animosity among workers – something that may be less likely when the final decision is left to a third party, such as the manager.

More fundamentally, one might question the default reliance on consensus. For example, in her study of a self-managed center, Mansbridge argues that its rigid commitment to consensus was a mistake, as it notably led some individuals to refrain from expressing their true preferences or, worse, to perceive them shamefully as obstacles to the common good. Acknowledging the merits of the agonistic tradition and, in particular, the value of proportional outcomes, she concludes that such worker collectives should “shift from unitary to adversary procedures depending on the goals they wish to pursue and the extent to which their members actually have interests in common […]” (Mansbridge 1983, 292). We lack the space to fully explore this discussion here, but further reflection is needed on the desirability of relying on consensus in workplace democracy – particularly since the workplace represents a unique context for its application, especially given the repeated interactions among the same individuals.

For now, let’s then assume that arbitration by the manager is the least problematic method when consensus breaks down. How exactly does this arbitration work, and why is it considered legitimate?

8 Rethinking Managerial Issues Beyond Technical Expertise

Arbitration must be preceded by attentive listening on the part of the manager and, to be effective, must rest on the legitimacy granted to them. It is at this point that Dejours et al. introduce a concept that is both complex and central to any theory of democracy: authority. This authority should be “founded from the bottom, meaning that it is conferred on the leader by the subordinates” (Dejours et al. 2018, 170). On what criteria, though?

This authority proceeds first from the technical competence of the manager, specifically her ability to provide advice and assistance to subordinates when they fail to tackle the obstacles presented by the real by themselves. This authority is thus based on technical competence (…). But the authority of the manager also depends on her ability to listen and make decisions, that is, to take responsibility for the consequences of the decisions. (Dejours et al. 2018, 170–1)

Ideally, a technically proficient manager should possess the ability to perform any technical task assigned to their subordinates, if necessary. I believe this idea is crucial and should be central to any theory of management in democratic workplaces (in fact, it could be recognized as one of the official functions of a manager, as Dejours et al. argue). Many workers feel, often unjustly, that their supervisors lack the competence to understand their work and, even more so, to help them. It is clear that assigning the role of assisting subordinates as part of the manager’s function is a valuable and important idea, even though this often requires a significant cultural shift. Kasmir, for example, recounts how managers struggled to accept such a change when Mondragon attempted to implement participative management before it lost momentum in 1983:

The study recognized that it was difficult for the first-line supervisor to adjust to these changes. […] Instead of concentrating on the details of operations and giving orders and imposing discipline, the supervisor will be expected to guide workers to build self-managing work groups, to coordinate activities, and to offer advice but not intervene in work operations unless help is needed. It is easy to describe in general the nature of this role but difficult to carry it out effectively. (Kasmir 1996, 127)

However, what concerns us here is whether this technical knowledge of the work and “her ability to listen and make decisions” grant the manager legitimacy to make decisions in the case of conflicts. There is something intuitively problematic about this idea, given that, as the authors have noted, the issues the manager must address are often not just technical, but also involve other values, particularly moral values (for instance, Dejours et al. mention “respect for principles, loyalty, thoughtfulness toward colleagues, and so forth” (2018, 168)).

To gain a clearer understanding of this crucial aspect of the issue, I believe it is essential to move beyond the abstract nature of discussions on the functions of management and instead focus on identifying the specific decisions employees are likely to delegate to managers. With this aim in mind, the following list could be proposed:

1. Working conditions:

   1. Offices allocation and equipment (furnitures, etc.)

   2. Workplace discipline (general work location and working hours)

   3. Setting work standards, work procedures and pace

   4. Safety rules and practices

   5. Job assignments and tasks allocations

   6. Evaluation procedures

2. Employment conditions:

   1. Recruitment of new employees and employee layoffs

   2. Setting wages and salaries

   3. Fringe benefits

   4. Distribution of training opportunities

   5. Employee promotions (advancement) and job transfers

   6. Collective welfare income (e.g., medical, housing)

3. Job content:

   1. Choice of products, markets, and pricing policies

   2. Research and development

As previously mentioned, within the framework of a participative management theory – which we now consider to be the defining characteristic of work organization in a democratic workplace – the fundamental principle is decision-making through consensus. By default, the areas listed below are those over which employees in a well-functioning democratic enterprise should have sovereign authority. However, in cases of unresolved disagreement, it falls to the manager to make a final decision.

Consider, for instance, the distribution of tasks. Imagine that a temporary increase in workload becomes necessary – due to an unusually large client order or because supply chain disruptions have caused delays that now require urgent recovery. Suppose the work is typically divided fairly between two teams, but neither team is willing to take on this additional burden, and both reject the proposed compromises. While such situations may be rare – being usually resolved through mutual agreement – how should the manager decide when consensus cannot be reached? In such cases, the manager is compelled to draw on a range of values – not only practical concerns like efficiency – for example, assigning the task to the team that can complete it most quickly – but also moral values such as fairness – even if the first team is more efficient, it may be unjust to assign them the extra work if they have already borne a disproportionate burden in the past. In fact, ethical and social justice considerations arise at every level of decision-making. For instance: Regarding office allocation in an open-plan setting, should managers and the elected director be given private offices, even if doing so risks reinforcing the perception that their work is more valuable than that of others? As for working hours, should meetings be scheduled only during regular hours out of respect for employees with childcare responsibilities? Should employees have the option to work longer hours, or should the company adopt a slower pace – for instance, a four-day workweek? When it comes to recruitment, should affirmative action policies be implemented in hiring processes? In choosing a health insurance plan, should the company opt for a cheaper option, even if it fails to adequately cover those with the greatest medical needs? And, of course, wages: What should the pay scale be within the company? Should individual bonuses be allowed, even if they risk fostering competition rather than cooperation among employees?

These examples underscore that managerial decisions in a democratic workplace are never purely technical – they inevitably engage with ethical, political, and social concerns and what constitutes a legitimate organization of work. Given this, is it not problematic to assume that such questions can ultimately be decided by a manager on the basis of their ‘technical’ competence alone?

9 Management by Election

In fact, this question had already been considered at the very inception of the first worker cooperatives in early 19th-century France by their principal founder, Philippe Buchez. While Buchez’s model diverges from participatory management in that it reserves co-decision among employees for matters ‘of some importance’, his views on the elected committee in charge of management in small co-ops suggest an interesting avenue: selecting managers based on their moral convictions. Indeed, Buchez argued that members of the elected administrative committee should be a “gathering of intelligent and upright workers” who are not driven by personal interests. This is particularly significant because, as Buchez notes, “its action would be twofold; it would direct the work and ensure that production aligns, as far as possible, with consumption; it would be moral and charitable, acting for the benefit of all and each, by maintaining a healthy balance in the distribution of tasks according to each worker’s capacities” (Buchez 1832, 52). As we can see, and interestingly, the committee must demonstrate moral virtue because it is not only responsible for managing the workers but also for addressing the needs and demands of external consumers.^[11]

How, then, can we ensure that managers uphold the moral values that are important to the workers? I propose here to revisit the idea of electing or randomly selecting managers as a potential solution to this problem. The idea would be to either choose or randomly select one or more of the most competent workers to listen to and advise other workers, as well as to organize deliberative spaces. I will focus here on the election option, setting aside the possibility of random selection.

In the case of an election – the classic democratic selection mechanism already used in the case of the general manager – the criteria for selection would certainly include a candidate’s ability to listen, advise, and organize deliberative spaces, as we said. However, it would also include the alignment of the candidate’s platform with the personal moral convictions of the workers. It would require the manager to recognize the political and moral choices underlying decisions, be able to articulate them, explain them, and justify the benefits of their moral positions in some way. In the end, this would essentially be the same process that applies to directors and board members in democratic firms. Elections could be held every four years or less, with the possibility of limiting the number of terms.

One significant critique that could be leveled against the model proposed here is that it offers limited prospects for cooperatives to remain competitive in a capitalist market.^[12] A well-known objection against cooperatives was put forth long ago by the economist Henry Hansmann, who attributes their rarity to high transaction costs, arguing that the heterogeneity of worker preferences tends to increase the cost of decision-making. Since the 1990s, empirical research on this issue has made progress, although its findings have struggled to gain traction among economists and the public sphere. As Gregory Dow summarizes in his seminal work, which synthesizes three decades of research on the subject, the rarity of worker cooperatives cannot be explained by an abnormally high failure rate: “Once created, they appear robust. Rather, they are rare because, in absolute numbers, they are created much less often” than conventional firms (Dow 2003, 227).^[13] Contrary to persistent misconceptions,^[14] worker cooperatives survive at least as well as other firms – even in capital-intensive industries – are at least as productive, preserve jobs more effectively during recessions, and demonstrate greater long-term stability than conventional enterprises (Pérotin 2016).^[15]

That being said, could the model defended here lead to an increase in transaction costs within cooperatives, which, under normal circumstances, tend to be less participatory than what this framework envisions? And if so, might the benefits – both in terms of the epistemic quality of decision-making and the enhanced motivation of workers – offset the additional time spent in meetings? I am inclined to believe so. If not, it would be necessary to devise solutions to mitigate this potential shortcoming, such as reinforcing state support policies envisioned by Honneth (2024, 190) and reinforcing “specialized public credit institutions” or “publicly supported cooperative development intitiative” envisioned more specifically by Wright (2019, 78–9).

The election of managers was a solution already considered by G.D.H. Cole in his industrial democracy model. Notably, he also put forward an idea that appears particularly relevant to the viability of our specific model of participatory management: the possibility of their removal.^[16] To illustrate this point, I will present a case study that highlights the – underestimated – potential for power-avoidance management within worker cooperatives.

10 When Managers Fail to Manage: Lessons from a Cooperative Case Study

I conducted a three-year field study within a small production cooperative specializing in consulting for associations and public entities.^[17] At the time of my arrival, the cooperative had 20 employees. At the outset, the elected general manager and the four group managers had established a plan whereby each team would be assigned a collective revenue target to be communicated and monitored by the group manager. This strategy aimed to avoid the pitfalls associated with individual revenue targets, which risked fostering competition and undermining solidarity among consultants. But the group managers never fully assumed this responsibility – they failed to address the collective revenue targets directly. They didn’t take the collective revenue target head-on. They didn’t take it with the seriousness it required, and failed to take the necessary steps to manage it in a clear and decisive manner. Consequently, there was no clear consensus among the workers regarding whether these targets were to be met collectively or individually, and the subject became somewhat taboo. In parallel to this, some workers privately felt they were not securing enough clients but were hesitant to express their concerns openly. Instead, they withdrew into a form of denial regarding the potential risks this posed to the company. Two years later, the cooperative’s financial situation had deteriorated, leading to the sudden and brutal dismissal of two consultants.

Interviews reveal that the managers were, to a significant extent, in denial about their responsibilities. This attitude was reinforced by their fear of the ethical dilemmas and interpersonal conflicts that might arise from measuring each team’s revenue – even collectively. As one manager confided to me, evaluating the group’s overall performance inevitably required collecting individual figures, which carried the risk of “singling out a colleague”.^[18] Furthermore, it is plausible that the managers feared such evaluations could also reflect poorly on their own performance, exposing the fact that they were not contributing sufficient revenue to the cooperative. Notably, one of the two individuals who were laid off was, in fact, a manager. As a result, two key managerial functions that had been assigned to them remained unfulfilled. Coordination: the managers failed to ensure that workloads were evenly distributed. Support: they did not adequately listen to or advise less experienced employees who would have benefited from their guidance.

Interestingly, this case represents an instance of power-avoidance management, which stands in stark contrast to the example of power-driven management that we found in the Mondragon case. This case suggests that it is likely necessary to establish not only democratic procedures for selecting managers but also mechanisms for oversight and removal in cases of negligence. This is an aspect that G.D.H. Cole had already identified in his Guild Socialism model – even though it does not exhibit the participatory management features that concern us here (Renault 2022, 91). He was careful to specify that “the first grade of industrial supervisors should be chosen directly by the particular body of men with whom they are to work” (Cole 1920, 54). The right to appoint supervisors, previously reserved for the employer, thus became a democratic right, “making the discipline of the shop a matter, no longer of imposition from without, but of self-regulation by the group as a whole” (Cole 1920, 198–9).

In cases where workers are dissatisfied with management, Cole outlined the following procedure:

The workers who chose their manager need not have an unrestricted right to recall him at any moment. Before he could be deposed, he should have the right to appeal to his peers his fellow-managers; and, if they held him in the right, but the workers still desired his dismissal, the case should go for judgment to a higher tribunal of the Guild. But even so I think that after a certain lapse of time the workers under him should have the right to remove him. (Cole 1920, 58–9)

There are, therefore, two stages of ‘appeal’, so to speak: first, before a tribunal composed of other managers, and then before a “higher tribunal of the Guild”, about which Cole provides no further details. Interesting reflections accompanied by practical initiatives have already been undertaken regarding an independent judiciary for workers in worker cooperatives (Bernstein 1976, chap. 8). But it seems that these efforts are primarily directed toward rank-and-file workers rather than managers.

However, in the case I studied, I am not certain that such a procedure would have been effective, as all the managers collectively failed in their responsibilities. Perhaps the matter would have needed to be escalated to the ‘higher tribunal of the Guild’. Another limitation lies in the possibility that, within the appeal tribunal, managers might show solidarity with one another and downplay their colleagues’ shortcomings. Yet, it is also possible that such a tribunal could serve as a space for deliberation among managers on how to perform their duties more effectively – a place for healthy mutual support. Indeed, I would argue that the deontic activity, which enables workers to define what constitutes ‘good work’, should also be present at the managerial level in determining what makes fair and legitimate management.

11 Conclusions

A theory of workplace democracy necessarily depends on premises about how work is carried out and the relationship between individuals and this activity. One might view it as generally unsatisfactory and unattractive to most people, regardless of its organization. On the other hand, one could argue, as Dejours (2013b) suggests, that the individual cannot help but seek to be useful to society and to exercise their creativity in the workplace.^[19] The role of work organization, then, is to prevent discouraging this fundamental desire by implementing an exploitative, alienating, and oppressive work regime. This aspiration to be useful to society, along with the pleasure derived from seeing one’s abilities unfold and evolve, leads individuals to theorize the most suitable and effective ways to perform their tasks and their profession more generally. Too often, the misguided orders from managers reveal to workers that the quality of their epistemic conception is not even recognized. Moreover, their desire to be entitled to equal respect and to have their interests and preferences acknowledged is what underpins their often implicit demands for more democracy in the workplace.^[20]

These premises have an impact on the theory of what constitutes necessary and just management. In this article, I argued that management should support workers in their quest for the best solutions to perform their tasks in ways that align with their own technical, moral, and cultural preferences. To do so, the manager must fulfill several roles: at the individual level, a role of listening, assistance, and guidance to the worker; at the team level, a role in facilitating decision-making – which includes highlighting the ethical aspects of decisions – and in ensuring the follow-up on decisions made; and, at a potentially broader level, a role in coordinating between teams. More specifically, I contended that it is essential for workers to have a say in selecting the managers they work with, a component often overlooked in current theories of workplace democracy. It is also necessary to consider mechanisms for monitoring and, if needed, removing these managers.^[21] I thus hope to have demonstrated that democratizing workplace democracy is not only necessary, but also feasible.