Workplace Democracy as the Liberal Republican Default

Abstract

Liberal republican justice requires adequate, reasonably equal protection against domination while giving everybody a reasonably equal opportunity to live their lives as they see fit. This article investigates which central policy follows from it for workplace non-domination, under realistic assumptions about the nature of our working lives and some structural characteristics of advanced economies (especially the presence of large firms). It argues for a legal default of workplace democracy, understood as any arrangement that gives workers equal individual control rights over the management of their firm, and, as a body, more control rights than any other groups of enfranchised stakeholders, taken together. The argument has two steps. The first demonstrates where the most promising alternative liberal strategy – that of enhancing all individuals’ power to exit from their job through granting unconditional resources – fails. The second shows how this failure leads to the specific conception of default workplace democracy outlined.

Introduction

Liberal republican conceptions of social justice have attained prominence in recent decades. According to them, justice requires that people be protected adequately and equally, or not too unequally, against domination by others in relevant social interactions and relations (republican) while giving everybody a substantively equal, or not too unequal, opportunity to live their life according to their own conception of the good (liberal). Within such conceptions, non-domination figures as the sole requirement of social justice, or at least as one enjoying a certain pre-eminence, or priority, over others. ¹

Applied to the workplace as a site of justice, republican views direct attention to the features and quality of individuals’ social relations at work, and not, or not just, to their distributive implications (in terms of pay, or the general division of society’s cooperative surplus). This is welcome, given that extensive and often rigid hierarchies offering copious opportunities for abuses of power by employers and managers characterise the working life of many in even the most affluent countries, and affect not only the poorest, or the most precariously employed (though they affect them most). The last two decades have seen a striking revival of debates around the just structuring of labour markets and employment relations (for example, Blanc and Al-Amoudi, 2013; Estlund, 2018; Ferreras, 2017; Gheaus and Herzog, 2016; Herzog, 2022; Landemore and Ferreras, 2016; Tsuruda, 2020), and republicanism plays a prominent role in that revival (Anderson, 2015, 2017; Cabrelli and Zahn, 2018; Cicerchia, 2022; González Ricoy, 2014; Gourevitch, 2015; Taylor, 2017; Thomas, 2017).

This article investigates which central policy, if any, follows from liberal republican principles for achieving workplace non-domination, under some suitably realistic assumptions about the nature of our working lives as well as about structural features of advanced economies (such as the existence and prominence of large firms). It argues that this central policy is a legal default of workplace democracy for firms, understood as any arrangement in which workers, as a body, have more control rights over the management of their firm than all other groups of enfranchised stakeholders, if any, taken together (at least ‘50 + 1’), and workers’ rights are individually equal to each other’s. Its question is not that of which policies might reduce domination at work compared to the status quo, or deliver a minimum degree of protection. Many, and very different, policies would do so. But only default workplace democracy can claim to be indispensable for realising the liberal republican ideal of adequate and sufficiently equal protection against domination at work – and to be sufficient, or very close to it, if coupled with other policies already commanding broad agreement, such as standard liberal freedom of occupation and full employment policies. By showing that default workplace democracy should be the central plank of a liberal republican ‘realistic utopia’ for the workplace, the article advances debates both about the central implications of liberal republicanism as a family of theories of justice, and about the nature and desirability of different conceptions of workplace democracy. It makes an argument for workplace democracy that is based on non-domination alone, negatively understood, and does not depend on any conception of the good of work, and of workers, that goes beyond standard liberal premises. Thus, while different proponents of workplace democracy should be able to agree with the general thrust of the argument made in this article, its specific contribution is to show that workplace democracy can be arrived at from weaker premises than endorsed by many of them, and that these lead to a specific conception of default workplace democracy.

This argument proceeds in two steps. The first demonstrates that the seemingly most promising, because avowedly most liberal, alternative strategy – that of enhancing individuals’ freedom and bargaining power over their work conditions by enhancing their effective opportunity to exit their jobs through unconditional resources for all (henceforth: ‘exit power’) – is insufficient. Under some modest realistic assumptions, it can only offer a modicum of workplace non-domination, not an avenue for realising the ideal. Economies with large firms need default templates for firm organisation in which firms retain stable authority over management decisions and a level of standardisation of other working conditions, neither of which are replaceable by individual negotiations without risking excessive conflict with requirements of productive efficiency. So, the second step argues, from a transitional point of view, abolishing current hierarchical defaults requires setting new ones, and, from an ideal-theoretical point of view, defaults and templates will continue to be needed, even under (sufficiently realistic) ideal conditions. Given a commitment to equal non-domination, adequate defaults need to be workplace democratic, and specifically liberal republican defaults take the form outlined above, exceeding existing workplace co-determination regimes, but differing from mandatory workplace cooperativism.

The article is structured as follows. The section ‘Liberal Republicanism and Work’ sets the stage by delineating the liberal republican core principle of equal non-domination, its central demands for individuals’ working lives and a set of surrounding labour market policies and commitments, about which liberal republicans interested in workplace non-domination agree. The section ‘Why Exit Power Falls Short’ outlines the exit power strategy and its supposedly unique attractions, before raising the objection based on the general need for significantly standardised workplace regimes, both within (reasonably large) firms and across firms. The section ‘Liberal Republican Default Workplace Democracy’ shows how this objection provides a direct and strong basis for a corresponding case for default workplace democracy as a central part of the appropriate standardisations, and develops the proposal for a specifically liberal republican workplace democratic default. The final section concludes by demonstrating the virtues of this proposal both in realistic ideal theory and for purposes of transition to more comprehensively non-dominating economic arrangements, also beyond workplace non-domination, and responds to liberal objections to it.

Liberal Republicanism and Work

Domination is the asymmetrical capacity of one agent to interfere in an arbitrary fashion in the choices of another agent (Pettit, 1999: 52), or to interfere in them in a matter that is not subject to the control of the interferee (Pettit, 2012). Subjection to arbitrary power is bad, and a failure to protect against it adequately is unjust, for several reasons. Being at the mercy of superiors regularly creates sizable uncertainty for the subjected (Lovett, 2010; Pettit, 1999), and puts them under pressure to ingratiate themselves with (Pettit, 2012: 64), and display servility towards, the powerful. It slots them into a position of social inferiority (Kolodny, 2019; Schemmel, 2021: 67), and it can undermine their self-respect as an equal, if they internalise, and accept and rationalise, this inferior social standing.

Liberal republicanism aims to account for these concerns while staying faithful to the aims of liberalism, by conceiving of non-domination primarily as a constraint on the structure of social relations, not as a mandatory goal for individuals’ personal lives. ² Within these, individuals should enjoy the greatest possible leeway for developing, and pursuing their own conception of the good, within the bounds set by justice (Rawls, 1996, 1999). Liberal republicans thus seek to avoid perfectionism, aiming at a suitable conception of neutrality between different conceptions of the good (Pettit, 1999: 291; Taylor, 2014; Thomas, 2017: 260).

There is much debate about the exact nature and content of individual choices that need to be protected against domination (Pettit, 2012: 92–104; Schemmel, 2021: 71–78; Thomas, 2017: 15–17), as well as of the question of whether republicans who regard non-domination as the primary aim of justice can really lay claim to any aspirations of liberal neutrality. ³ These important theoretical debates need not concern us here, as long as the fundamentals of individuals’ working lives are within the scope of choices and interests to be protected against domination, whatever else is, or is not. Clearly, they are. Work covers a lot, and for some most, of our waking lives. Total time spent at work, effort demanded, duration and timing of breaks granted, planning of work shifts, distribution of tasks and setting of deadlines, commuting and other work-related travel requirements, compatibility with family life and caring obligations, whether one’s job is secure or not and so on – these are important factors for how well almost everybody’s life is going (Tsuruda, 2023: 154, 162). If these matters are left to unilateral determination by managers who are not effectively and robustly accountable to workers for their decisions – through direct, institutionalised accountability mechanisms, or because workers can easily turn their backs on them and leave for a better employer – workers are dominated. Liberal republicans hold that the evil of domination is primarily interagential (Lovett, 2010; McCammon, 2015; Pettit, 1999, 2012), consisting of subjection to other agents (whether individual persons, or collective agents), not structures as such. However, it is important to highlight that agential workplace domination rests on two interlocking structural levels. The first is that it is not somehow due to the individual prowess or charisma of managers, but inscribed into typical work contracts and dominant types of firm organisation, mandated or at least heavily favoured by existing laws. The second is that the latter, in turn, tend to be upheld by superior power of those who have an interest that managers be accountable to them, but not to workers – that is, capital owners – at the level of political decision-making. ⁴

Unaccountable managers dominate workers whether or not the latter have an active, intrinsic interest in self-determination at work which extends to directly participating in management decisions, such as decisions about what should get produced, how much of it, price- and wage-setting and firm investment strategy. For liberals, that is not a necessary, or privileged, worker interest, but merely one on a par with many others they will want to accommodate to whichever degree all of them can be, conjunctively. However, who takes these decisions, and how they are constrained and controlled in the process, and how much uncertainty they expose workers to in the aforementioned respects, matters because, as noted, some important interests of all workers are at stake, whether or not an intrinsic interest in taking management decisions is among these. According to liberal republicanism, domination over these interests is unjust for the reasons outlined at the start of this section, not merely because, or when, management decisions reliably affect them negatively. This will often, but need not always, be the case (we can imagine unconstrained, but particularly benevolent, managers).

There is agreement among liberal republicans interested in workplace non-domination that it requires at least three sets of policies operating in the background – but that these are not, taken together, sufficient to secure it adequately. The first is stringent protection of freedom of occupation as a basic, or quasi-basic, liberty. ⁵ With freedom of occupation in place, the mechanism for slotting (would-be) workers into jobs is that of a labour market where they compete for jobs, and firms for workers, operating alongside the product markets on which firms compete for selling their output.

The second point of agreement is that competition for jobs on labour markets ought to take place against the background of generous, largely unconditional, egalitarian provision of essential social services and goods, such as education, health care and infrastructure. Most think that that provision should take place through a Scandinavian-style universal welfare state (Taylor, 2013, 2017: 49 – advocating a ‘Nordic model’; Schemmel, 2015; Thomas, 2017: 199–205).

The third point of agreement is that some degree of external workplace regulation, settling and policing essential work conditions (such as general health and safety, and non-discrimination) needs to be in place, on whose basis workers enjoy effective rights to contest relevant managerial decisions in fora above firm level (Hsieh, 2005; Taylor, 2017: 58; Thomas, 2017: 270). These fora can be located at state level (mainly labour courts) or at subsidiary sector levels (e.g. unions discharging some official functions), or a combination of both. However, liberal republicans also agree that such workplace regulation can never be adequately complete, and that it would even be undesirable for it to aim to be. That is mainly because labour contracts and their external regulation can never fully codify justified expectations on both sides: it is in the nature of work that it is characterised by the need for flexibility at firm level. ⁶ Firms need to be able to react to changing economic and business circumstances flexibly and quickly, and need an internal allocation of discretionary directive powers to deal with these. This is an integral part of being able to take effective management decisions.

Thus, even when these three sets of policies are as fully implemented as it is reasonable to wish, there is a regulation and accountability deficit left that gives rise to worries of workplace domination, and must be filled in some way. The question is who has the necessary discretionary power, and how it has to be controlled, within the firm. It could be handed to workers directly, or their representatives, through legally requiring a form of workplace democracy, minimally defined by González Ricoy (2014: 233) as ‘a form of managerial organization where workers have control rights over the management of the firm’. This definition needs two clarifications. First, to qualify as workplace democracy, not just any arrangement where workers have some say, it must require that workers have greater control rights over firm-internal decisions, at least pertaining to electing representatives, ⁷ than all other stakeholders in the firm, if any, so that their vote can be decisive on a range of policies. Second, their control rights must be equal to each other’s (no firm-internal ‘worker aristocracies’). Many republicans argue for a form of workplace democracy, such as worker-owned cooperatives (e.g. Gourevitch, 2015) ⁸ , though some do explicitly based on a conception of active worker self-determination (Breen, 2015, 2017) that goes beyond liberal republican non-domination as outlined above (absence of arbitrary power over working lives).

Against this, an important strand in liberal republicanism aims at empowering workers to resist potential workplace domination, and hold whoever takes management decisions duly accountable, in a more indirect fashion. It comes down against any form of legally mandatory workplace democracy, arguing for ramping up workers’ exit power instead (Taylor, 2017; Thomas, 2017 – although, for Thomas, this is just one element of a more demanding conception of societal background fairness). ⁹ According to this strand, exit power promises the best balance of liberal concern with workers’ freedom to choose from, or set up, an array of differently structured workplaces, and non-domination; mandating any form of workplace organisation is not necessary, and subject to independent liberal objections. ¹⁰ So, it is to the promise of this position that we must now turn, and to the main objection to it, as this objection points towards a specifically liberal republican case for workplace democracy.

Why Exit Power Falls Short

The Promise of Exit Power

Adherents of exit power hold that liberal non-domination in labour markets requires enhancement of workers’ capacity for exit specifically through equalisation of internal and external assets, thus increasing the bargaining power of all. This is to be done by granting everybody, on top of unconditional, high-quality social benefits, unconditional resources in the form of a basic income, stakeholder grants, shares in social wealth funds or any combination thereof. As already noted, they do not argue against all workplace regulation, but for a general, in instances defeasible, priority of shoring up exit. As long as roughly equally empowered individuals compete for jobs, and employers for workers, on labour markets approaching perfect competition, regulation should be merely a last resort (Taylor, 2013: 559, 2017: 58).

This clarifies also how exit power aims at more than merely an opportunity to leave bad workplaces; it is not a sufficientarian but an egalitarian view, aiming at more control over, and substantive opportunities for qualitative improvement of, workplaces for all. It does so by explicitly drawing on Hirschman’s (1970) key insight that, within many social practices and institutions, it is often (only) the credible threat of exit that enables proper voice (Taylor, 2017: 14). Such a credible threat gives clout to worker voice that is not present where, for example, voice merely takes the – now popular – form of informal worker ‘consultation’ by managers. Adherents of exit certainly want to empower workers to choose and form democratic workplaces, too, such as cooperatives, if they wish, and some go as far as maintaining that, given their current rarity, the state ought to take at least some measures favouring their formation, such as providing some tax relief (Thomas, 2017: 279). ¹¹ However, exit power also stresses the fact that real-world democratic procedures are always imperfect, allowing for various sources of domination. Avowedly democratic procedures can fail to control and hold to account officials sufficiently, ¹² or, for example, there could be cases of divided firms leading to a ‘tyranny of the majority’, where a majority of ‘skilled’ workers reliably advances their particular interests at the expense of a minority of ‘unskilled’ ones. What is more, even where well-functioning democratic procedures succeed in giving individual workers roughly equal shares of effective control, at least in the context of larger firms, these individual shares will inevitably be small. ¹³

Therefore, exit power promises more control for individuals, and where workers cannot attain any reasonably full measure of control over their work environment, this, under equal exit power on reasonably perfectly competitive labour markets, will only be because of the pressure that markets generate on work conditions and the shape of existing job opportunities. Under such conditions, that pressure stems from a myriad of non-intentionally coordinated individual actions and choices of firms and individuals. Subjection to impersonal forces, against an equal background, is not domination, at least not in the particularly morally salient and urgent sense that neo-republicans mainly want to capture: the specific evil of domination, on this view, is subjection to other agents. ¹⁴ At the same time, at state level, exit power only requires impersonal administration of a universal and unconditional resource scheme for all, which should minimise domination by officials and elected representatives at that level. ¹⁵

The Need for Stable Workplace Authority, and Other Standardisations

Nobody denies that enhanced exit power can reduce or mitigate workplace domination: that, other things being equal, more of it is a good thing for individual workers. As a central strategy for achieving, or approaching, a liberal republican ideal of adequate and reasonably equal workplace non-domination, however, it is subject to a serious, general objection.

This objection is not the one, familiar from the recent, and burgeoning, literature on workplace justice that enhancing non-domination through exit power facilitated by unconditional resources is bound to work unequally well for different categories of workers – less well for vulnerable workers. This category includes workers whose skills are not highly sought-after; workers for whom exit might create particularly high non-monetary costs (Herzog, 2022: 431), including in their personal lives, and even workers who are in fact highly skilled, but so specialised that they will struggle to find an alternative workplace catering to these skills (the ‘too-much-invested-to-quit effect’, González Ricoy, 2014: 240). Some go as far as arguing that facilitating exit through implementing a (modest) unconditional basic income might worsen the situation of low-paid workers, by increasing precariousness of employment (Birnbaum and De Wispelaere, 2016, 2021). However, while it stands to reason that exit power will not protect all categories of workers exactly equally well, it is not clear how far-reaching an objection this is. On the one hand, it requires specifying liberal republican principles of equal and adequate workplace non-domination further, with regard to how stringently egalitarian opportunities for workplace satisfaction and individual protection against domination for all workers must be, independently of personal differences. Different liberal republicans will disagree about this. On the other hand, for this egalitarian objection to turn into an argument for workplace democracy, we would need to specify a conception of the latter that would demonstrably do better for those categories of workers less well protected by exit. This is difficult to show with any precision. So, issues of interpersonal (in)equality between workers are not the right place to look for the most general and serious objection to exit power. ¹⁶

This objection focuses instead on general constraints on job negotiation that any reasonable regime for labour markets and workplaces has to abide by, if it wants to safeguard their functionality and efficiency – at least in a context of developed economies with large firms. It shows that this – rightly – puts significant limits on individual bargaining power over the nature of one’s working life, and not only for disadvantaged workers. Firms need to offer workers some standardised work conditions and rely on a model of firm organisation that standardly includes retaining authority over necessarily discretionary management decisions. This objection is familiar in outline, too, but it needs development, showing where which form of standardisation is needed, and how it grounds the argument for a corresponding form of default workplace democracy that is fully available from the thinner liberal republican premises outlined (‘Liberal Republicanism and Work’ section). Let us first work out the argument for standardisations and see how exit power responds to it, and then show in which respects this response fails to address the underlying problem.

Among the basic interests of all participants in an economy is an interest in a sufficient level of productivity. ¹⁷ In developed economies, firms – often large ones – are the main agents organising production; we have seen that their interest in being able to organise production grounds the need for flexibility, for possessing the discretionary power needed to react to changing market circumstances reasonably quickly. Firms cannot respond to any significant change in such circumstances by re-negotiating individual work contracts. In fact, they do not tend to bargain with workers, at least certainly not with individual workers, in cases of strategy changes. It is a well-known observation that firms are, internally, not an ensemble of markets, but bureaucracies relying on command chains (Coase, 1937: 388; Williamson, 1973: 320ff). Subjection of workers to the authority of their firm in a relatively open-ended fashion (Breen, 2017; González Ricoy, 2022; Gourevitch, 2016: 18) is an integral part of this.

States respond to this need for standardised firm authority by providing a number of templates for firm organisation in corporate law, so that those setting up a firm have off-the-shelf solutions firmly and stably constituting firm authority ready to hand. These help ensure that firms can coordinate large numbers of workers while being able to give outside actors, such as customers and investors relying on their performance, credible assurance that they will stably perform well (Anderson, 2015: 60). Now, according to one possible argument in economic theory, the simplest standardised solution would also be the best, at least from the point of view of investors, consumers and other business partners. This would be to give maximally unconstrained authority to management, always, with minimum accountability to workers, in the name of maximum flexibility, concerning the allocation of tasks, laying off and hiring workers, and work conditions and hours – perhaps even allowing worker remuneration to fluctuate significantly. Firms could thus react and tailor performance to changing market circumstances at greatest speed.

However, this solution is morally unacceptable because it entirely overlooks the interests of workers and amounts to their near complete domination. ¹⁸ Moreover, under any arrangement meeting the conditions outlined in the section ‘Liberal Republicanism and Work’ – freedom of occupation on a competitive labour market against the background of good, universal benefit provision – firms so structured will struggle to attract self-respecting workers with any skills to offer.

In fact, states provide not only templates for firm organisation, but also some standardised framework for work contracts and conditions. This joint framework of corporate and employment law expresses political decisions balancing workers’ interest in protection with firms’ interest in authority and flexibility, significantly narrowing the range of possible negotiations between the parties. Based on what we have said so far, states are, in principle, right to so step in even if they do not get the balance of interests exactly right, because having to negotiate job packages from scratch would create enormous transaction costs (Anderson, 2015: 50, 60), and sizable firm-internal conflicts. To help firms avoid these transaction costs, institutional solutions are needed (North, 1990); templates in corporate and employment law are these solutions (Anderson, 2015: 63). As noted, all participants in the economy, not just firms, have an interest in reasonably efficient production, and thus in avoiding excessive transaction costs in the productive sphere.

Based on this, in all advanced economies, some standard forms of employment contracts work in tandem with standard templates for firm organisation. ¹⁹ These express different ways of balancing the underlying interests in accordance with the overall imperative to lower transaction costs. ‘Coordinated’ political economies (Hall and Soskice, 2001), with strong trade union presence, provide less freedom to exclude worker interests from firm governance, and tend to specify some form of mandatory presence of representatives of workers on corporate boards of firms of a certain size, such as in the German co-determination system. ‘Liberal’ political economies (Hall and Soskice, 2001), such as the US and the UK, provide more such freedom of exclusion, and are in fact more dominated by companies which have much more pronounced accountability to shareholders than to workers (if any). Laws on firm organisation nowhere mandates non-democratic organisation, but it is, in fact, heavily dominant everywhere. In no OECD economy is it currently mandatory that workers command some form of majority.

At the firm level, legal templates for firm organisation and work contracts leave some freedom for individual arrangements, where provisions are derogable or simply leave some matters uncovered. However, at least reasonably large firms tend to rely on firm-specific templates for work contracts, constraining room for individual bargaining further, once more with a view to lowering transaction costs, and ensuring uniformity of treatment of workers to avoid firm-internal conflicts (leading to often quite large ‘Human Resources’ departments). Therefore, some standardisation operates at that level, too – though it has come under pressure. Especially in countries with little compulsory representation of worker interests (but also elsewhere), many firms have, over the last decades, retreated from monolithic corporate organisation and highly standardised contracts by ‘fissuring the workplace’ (Weil, 2014), slimming down the main company, outsourcing and creating smaller, dependent, supply firms with worse work conditions, often operating across jurisdictional boundaries. However, they have evidently mainly not done so in order to accommodate any interest of workers in more diverse workplaces, but because they hold that this strategy enables them to withstand increased global competitive pressure, or increase short-term profitability for shareholders, or both. ²⁰

This short sketch of how the need for standardisation arises both across firms and at firm level (for reasonably large firms) indicates the principled limits of exit power, understood as an ideal for achieving reasonably equal and sufficiently high protection against workplace domination by enhancing individual control (not merely as a partial remedy for its worst forms; see the previous subsection). As an ideal, exit power advocates for the abolition of any bias towards hierarchical organisation enshrined in law as well as for effectively increasing de facto worker power over the workplace, by enabling voice through (threat of) exit. Giving workers the freedom to say ‘no’ to job offers gives them the opportunity to look for democratically organised workplaces, if they wish, and, even if they do not, they can seek to negotiate bespoke agreements that keep firm authority over them within bounds they are individually comfortable with. So, on the best exit power scenario, work relations should tend to reflect negotiated and agreed allocations of authority rather than general ones imposed on workers with little choice. Ideally, this should, therefore, lead to a ‘pluralistic commonwealth’ (Thomas, 2017: 255) of forms of workplace organisation.

However, this overlooks that stable firm authority as well as other standardisations are not merely a result of de facto superior power of management and capital owners over the terms of work, but, as seen, also respond to a genuine concern for efficient productive organisation, including especially keeping transaction costs at bay. Firms will particularly resist any attempts by workers to put individually negotiated limits on firm authority over management decisions. It is easy to see that any arrangement where the boundaries of firm authority fluctuate over individual workers within the same organisation risks becoming literally unmanageable. So firms are bound to frustrate workers’ wishes on this issue, or, if the latter are sufficiently empowered to keep saying ‘no’, this would push up transaction costs very significantly.

Exit power should, against that, make it easier for workers to voice requests for individualised working conditions that do not impinge on core business matters (for example, pertaining to the kind of computer workers use, or to how many days they can work from home). Even here, however, at least reasonably large firms encounter some pressure to standardise work contracts, as outlined – they need to guarantee reasonably stable work conditions as the composition of their workforce changes over time, facilitating diachronic worker integration. New workers’ job conditions cannot vary too radically from those for workers already present, also in order to avoid firm-internal conflicts between different groups of workers.

Furthermore, in individual negotiations about whether to take up, or retain, a job, workers tend to face a structural asymmetry in bargaining position that favours firms: changing workplaces, especially where it necessitates moving residence, often has a large impact on workers’ personal lives – while employers, if they are firms, have no personal lives. ²¹ This casts doubt on how perfect the perfectly competitive labour markets envisaged by adherents of the exit strategy could ever be, in terms of suppliers’ (workers) capacity to switch between different buyers (employers) sufficiently rapidly (Anderson, 2015: 67–68). ²² Exit power is, in fact, demanding on workers, because, for it to have real purchase, workers have to be quite mobile, ready to accept costs for their personal lives – or at least have to strike a credible pose of more or less constant mobility, to keep employers on their toes.

For reasonably large firms, any kind of far-reaching negotiation over essential work and management matters would thus realistically have to take place between firms and collectives of workers and aim at a settlement for an extended period. Workers in non-democratic firms could try to democratise them – or lobby for whatever form of organisation they, as a collective, regard most desirable – by threatening to leave en masse. Or collectives of workers could operate together on the labour market trying to take over firms looking to extend their workforce considerably, or having to replace a large part of it at once (e.g. due to a retirement wave). While unconditional resource schemes would certainly facilitate this, such opportunities would still be quite rare, and the fact that coordination costs for sustaining such worker coalitions would squarely fall on the workers themselves would endanger their stability. Why hold out together with lots of others if somebody makes you an otherwise reasonable individual job offer? Such workers would have to be very committed to their preferred model of firm organisation.

The deepest problem of the exit power strategy is thus not that any current proposals might envisage a level of unconditional resources that is too low to solve problems of workplace domination. The strategy as such is insufficiently responsive to the reasons why there need to be some standardised solutions for stably allocating sufficient firm authority for discretionary action, and for other work conditions, to start with. As seen, these reasons concern especially lowering transaction costs, and are ultimately grounded in requirements of productive efficiency. It is important to highlight that this objection does not presuppose any imperative to maximise efficiency. That could not be an imperative of liberal republican social justice; it is likely to conflict, at some point, with workplace non-domination itself. The issue is, more simply, that exit power is insensitive to the nature of the problem. It might conflict even with any basic notion of productive efficiency that participants in an economy have reason to desire (or else leave workers in a position unable to escape the most stubborn problems of workplace domination).

Therefore, if there were a strategy addressing workplace non-domination that does not conflict head-on with the need for standardisations, it would, based on the argument so far, be the superior solution. Showing that this solution exists – and that is in fact, one of default workplace democracy, not something less – is the task of the next section. Even if this argument succeeds, it does not rule out that exit power could be adapted, especially through coupling it with other policies aimed at ensuring far-reaching background fairness for all workers (and non-workers), to yield a superior solution to workplace domination, all things considered. We will thus return to this question in the final section, double-checking that default workplace democracy is in fact superior, in both transitional and realistic ideal, theory. First, however, we need to show that the step from a general need for templates to a template of default workplace democracy is indeed short, on liberal republican principles.

Liberal Republican Default Workplace Democracy

That this step is short is not the impression one receives from current debates about workplace democracy. This is likely because most adherents of workplace democracy tend to argue for particular, often quite radical, conceptions of it, such as worker cooperatives, where workers, jointly and exclusively, own the firm. Critics of mandatory workplace democracy, such as the main adherents of exit power, then, understandably, focus their objections on such conceptions. Anderson (2015, 2017) fully recognises the need for templates, but remains on the fence and endorses no form of workplace democracy, merely pointing to some existing examples of workplace co-determination, and stressing the need for democratic – at state level – experimentation with different forms of workplace organisation. She raises objections against workplace democracy that similarly centre on radical conceptions, arguing that ‘firms run as an egalitarian participatory democracy, where no one has authority over others, and all work decisions are made collaboratively’ (Anderson, 2015: 60) face excessive coordination costs (see also Flanigan, 2022: 166–167). That seems right, for reasonably large firms. However, it leaves open a wide field of candidate conceptions of workplace democracy that do not face this problem, or at least to a much smaller extent. There is no reason to define workplace democracy in such radical terms, disqualifying more complex and indirect arrangements, for example, involving representation and executive elections, on conceptual grounds. We should not withhold the predicate ‘democratic’ from a firm where we would readily grant it to a similarly organised state (González Ricoy, 2019: 679–680).

Frega (2020) recommends directing our focus away from institutionalised decision rules and mechanisms for firms to the question of what broader norms of democratic workplace interaction – that is, a democratic workplace culture – might look like, and how one might encourage it. That is an interesting question, but, whatever the answer, it cannot establish that firm decision rules and mechanisms are not necessary. ²³ Clear and firm decision rules matter greatly particularly in contexts where different participants have significantly conflicting interests, and where they are permissibly strategically motivated. For liberals, this is true of work. Workers need to make good-faith efforts at fulfilling their work obligations and devote a sufficient degree – whatever that is – of energy and discipline to that task. But they do not need to be devoted to their firms’ overall interest, or the interests of other groups connected to it, to the same extent as they are to their own, and vice versa for these other groups. ²⁴ Nor do they need to regard work as the most important element in their lives, in any other respect. On the liberal view, workers need their managers to be accountable to them to safeguard important interests that always tend to be in play, whether or not some workers also have any additional intrinsic interest in actively running their firm (‘Liberal Republicanism and Work’ section).

Let us proceed to demonstrating, then, how default workplace democracy as understood in ‘Liberal Republicanism and Work’ section achieves equal and adequate workplace non-domination, by liberal republican lights, while avoiding the kind of direct conflict with other imperatives applying to the organisation of production outlined in the preceding section. The argument is relatively simple. Unlike any other group of stakeholders in the firm, workers are affected both by the day-to-day running of the firm and by its long-term success, so they have an interest in control rights over the firm that is not fully matched by corresponding interests held by any such other group. Therefore, they need to be enfranchised more than any other group. They should, minimally, be able to block any long-term decisions which they do not regard to be in their interest – such as questions of relocation and mergers and acquisitions, for example – and to be decisive (at least ‘50 + 1’) on matters that affect them in unique ways, such as the structure of the working day. Because the overriding consideration for liberals is that workers can safeguard their interests by controlling decision-makers, not that as many of them as possible can actively participate in governing the firm (‘Liberal Republicanism and Work’ section), this is, in principle, compatible with representative or otherwise mediated democracy at firm level. This could be via elections, or choice by other means, of worker representatives for suitable terms of office who exercise veto and decisive rights. There is no reason to think that representative worker democracy could not keep coordination costs to a tolerable level. This solution (of 50 + 1) is particularly well suited to implementing liberal republican non-domination in the workplace: workers are not underenfranchised, because they do not all need to be able to participate actively in firm governance, and they can block all proposals threatening their interests. They are not overenfranchised, because firms are still subject to market discipline in the interest of overall economic productivity, needing to compete with each other on product markets.

At the same time, if the argument of the previous section was sound, there remains some need for standardisations across firms, beyond default workplace democracy, to keep transaction costs at bay while implementing some universally needed protections of workers. Examples are health and safety, and non-discrimination, or regulation of what counts as a standard working day or week, or setting minimum or sector-wide benchmark wages. Workers overall, across firms, then also need democratic input into setting those. Even if political decision-making at state level were not a site of structural domination where capital owners have disproportionate power, it would still be problematic if, on such matters, which touch on workers’ interest more intimately than on those of other citizens (such as pensioners), workers did not have dedicated channels of influence of their own. That influence can take the form of sector-wide (not company-based) trade union organisation, setting up collective bargaining and negotiation procedures over the relevant matters. ²⁵ However, if the argument up to now was sound, these sectoral regulations can only supplement, not replace, default workplace democracy, precisely because there remains an ineliminable need for discretionary management decisions at firm level. As these cannot be avoided, managers taking them need to be made democratically accountable to workers at that level.

Against that, buyers of goods produced by firms do not need any firm-internal enfranchisement at all as long as the output markets that discipline workers are sufficiently competitive, and they have sufficient assurance that firms will honour their contracts. Investors, on the contrary, do have an interest in enfranchisement insofar as they often have to commit capital to the firm in a long-term fashion; they cannot be expected to commit such capital to a ‘black box’ – to a firm whose future operation is not transparent and not accountable to them, with no possibility of correction. So, for example, one promising proposal for instituting, in law, a conception of liberal republican workplace democracy so understood, is Isabelle Ferreras’ (2017) ‘economic bicameralism’, in which firm policies need to be agreed by combined majorities in two chambers, that of representatives of capital investors and that of representatives of ‘labour investors’ (Ferreras, 2017: 141), aka workers. This blueprint needs some modification to fulfil the conditions for workplace democracy outlined above: as noted, the worker chamber needs to be able to be decisive on some issues which affect workers uniquely, or at least more intimately – such as policies regulating the structure of the working day in that firm, breaks, shifts and so on. An arrangement in which workers, as a block, are always veto players, but nowhere decisive, does already constitute a qualitative shift from existing forms of workplace co-determination, but does not yet qualify as a workplace democracy. It is a limiting case.

Ferreras’ scheme, suitably modified, is just one example of a suitable arrangement fulfilling the conditions set out above, drawn on for the sake of illustration. For firms that are not capital-intensive, a separate chamber devoted to investors will not be necessary. In general, the conditions leave some space for democratic experimentation with different templates, across different political economies in which firms are embedded (concerning also the exact roles played by unions), and different kinds of firms. However, if the argument so far is sound, this room for experimentation is significantly constrained. It cannot be experimentation between workplace democracy and other arrangements, because the argument identifies workplace democracy of at least 50 + 1 as the lower bound of worker enfranchisement, nor can it be experimentation with templates of workplace democracy straying too much into the other direction. ²⁶ To be sure, there is, in principle, more leeway regarding the upper bound of worker enfranchisement in the default template, beyond 50 + 1; in particular, there is no reason to suppose that, for example, cooperatives are inefficient in a way that runs up against basic productive efficiency. ²⁷ Certainly, workers must be free to form workplace cooperatives if they wish. However, cooperativism should not be enshrined as the default. On the one hand, in cooperatives, worker-owners are committing both their labour and their financial capital to the firm, putting all eggs in one basket (White, 2015: 422): if the firm fails, they lose both their job and their resources, or at least a sizable part of them. It is one thing to be personally free to take such risks, another for the state to force workers to take them, or at least make it sizeably more difficult to escape them, by turning this form of organisation into the default.

On the other hand, given this concentration of risk, workers under cooperativism will be more hesitant to leave for a new employment, and cooperatives more hesitant to take on new workers, so that workplace cooperatives becoming dominant would lead to ‘thinning the labour market’ (Thomas, 2021: 528), making changing jobs too difficult. This is a valid liberal objection to this form of workplace democracy. Because of it, in order to keep exit feasible, liberal republican workplace democracy must not only stop short of enshrining workplace cooperativism as the default; it must, in general, be just as committed to full employment strategies as adherents of exit power are (Taylor, 2017: 52ff; Thomas, 2020, 2021). ²⁸

Of course, default workplace democracy, by itself, would not solve all problems of workplace domination. Within democratic firms, there may still be workers who are subject to domination, perhaps because they find themselves in a permanent minority. If default workplace democracy is organised along representative lines, representatives might turn out insufficiently responsive to their interests. They could be co-opted by investors, ²⁹ or corrupted in other ways. We have seen in the previous section that this was among the worries motivating exit power. Even democratically organised firms need oversight, and some effective opportunities for disadvantaged workers to appeal to firm-external bodies: at state level, through labour courts and ombudsmen, and/or through channel of appeal to sector-wide bodies, in which trade unions can, once more, play an important role. To be sure, problems of corruption and insufficient democratic accountability can arise at these levels, too. Like all republicans, workplace republicans need to devise strategies of dispersing power, and possibilities of appeal and contestation, over several democratic levels.

Furthermore, none of this denies that there must also be a place for ‘fair exit’ (Rouméas, 2023) for workers whose interests are not satisfied within their firms, whether or not any forms of appeal to levels above their firm are possible. As noted, liberal republican workplace democrats, too, need to be committed to full employment, coupled with the array of background strategies outlined in the section ‘Liberal Republicanism and Work’ (freedom of occupation, unconditional goods and services). There could also be more targeted support for particular disadvantaged workers who are, for some reason, currently insufficiently mobile and thus unable to leave a firm that disregards their interests, even if it is sufficiently democratically organised. However, this does not amount to a case for unconditional resources for all.

Therefore, these considerations, important as they are, do not amount to arguments against workplace democracy being necessary. We will turn to such objections – that exit power can be combined with other background policies to yield a superior solution to workplace domination overall, rendering default workplace democracy unnecessary – touching also on issues of economic domination beyond workplace domination, in the next section. Before doing this, and to conclude the argument of this section, it is important to note that it deliberately leaves open the question of the size that firms have to reach for the workplace democratic default to kick in. This is a practical matter beyond the scope of this article. It is also important to note, that, in ideal theory, at least after the dominance of workplace democracy is firmly entrenched, switching away from workplace democratic arrangements must be, in principle, possible, even for larger firms. It should be possible as long as employees not agreeing with the switch have an effective opportunity to move to a democratic workplace, and, generally, those looking for work on labour markets do not have their effective opportunity to choose a democratic workplace impaired. These conditions require that the latter remain the dominant form of organisation. A liberal view cannot demand more.

Conclusion: Objections

One typical liberal objection to arguments for mandatory workplace democracy holds that it jettisons liberal neutrality by enshrining a preference for a particular workplace environment that not all workers share (Taylor, 2014). If the argument so far is sound, then this objection is unsuccessful, because the argument did not rely on any special goods of democratic workplace organisation, such as positive conceptions of self-determination at work, but only on non-domination, negatively understood. This is why it can only lead to a workplace democratic default, not more. However, if there have to be defaults, they inevitably enshrine preferences for some types of workplaces over others that not all workers might share. There is no such thing as a neutral default. What matters is which defaults do better by liberal republican non-domination.

However, it seems possible for adherents for exit power to press on with a variant of the objection. Perhaps what we should do, instead of enshrining a democratic default for firms over a certain size, is to increase background equality of circumstances for all workers to such a high level that all workers have a choice of so many differently organised workplaces that working in a hierarchically organised one is extremely easy to avoid. Even if some still choose this arrangement, the fact that it is so easily escapable means that this does not amount to domination. Or perhaps, no employer would venture to offer such a workplace anymore, anyway, because there would be no demand for it. That would then be the more liberal solution.

Of course, if we assume that unconditional resources can be set at such a high level, and stably sustained, that might bring about this ideal state. However, we have seen that, even in ideal theory, at such a high level, transaction costs for slotting workers – who can continue to say ‘no’ to all job offers – into jobs may be so high that this would be a very risky strategy. Aiming at that level might conflict with any reasonable idea of productive efficiency we should hold on to; we cannot know in advance. Default workplace democracy achieves workplace non-domination at much less risk.

Furthermore, and just as importantly, this proposal fares worse than default workplace democracy also from the point of view of transitional theory. The political prospects of implementing such highly ambitious unconditional resource schemes are very low everywhere: being able to bring them about would arguably require fairly complete working-class dominance at the level of political decision-making (Gourevitch and Stanczyk, 2018), where these schemes – if economically sustainable – would then need to be irreversibly entrenched, for example, constitutionally, through supermajorities. ³⁰

Against that, the proposal to change the default for firm organisation to a democratic one would of course also encounter strong initial political resistance, but it requires simpler legislative change, which, in some political economies which already have strong union presence and forms of co-determination, would also be much more continuous with existing arrangements. And if it passes, it has the potential of being a catalytic reform, entrenching worker power in firms, making reversion effectively very difficult, especially if strong sector-wide unions are involved (see the previous section) – as well as offering a platform for combating wider structural sources of economic domination. Such domination can stem from severe inequalities of capital ownership, and/or lack of democratic control of large-scale investment. Default workplace democracy would already significantly constrain investors’ power over workers, but it would not solve this problem entirely. However, it would provide a firm basis for further-reaching measures addressing this problem. For all that has been said, some level of unconditional resources might contribute to solving it, too, by helping to broaden capital ownership, just as it can be useful to combat other scenarios of domination outside the workplace (White, 2020). However, if these transitional considerations are sound, they strengthen the case for default workplace democracy as the policy to focus on for tackling workplace domination, here and now. ³¹

Another proposal adjusting the background so as to combat labour market, and consequent workplace, domination is that of the ‘state as employer of last resort’ (Thomas, 2021), offering a job to all those who want one. Perhaps such a scheme is indeed necessary to make good on the commitment to full employment that all liberal republicans share – especially if one cannot count on continuous sizable economic growth. However, it is hard to see how it could solve the problem of workplace domination, without default workplace democracy also being necessary. It could do so only if workers’ opportunity to resort to the residual state sector somehow brought about the democratisation of (the bulk of) private firms. However, that would seem to depend on the conditions of employment in residual state enterprises. One might think it would if these were internally democratic. However, that would be problematic: because these enterprises are, by definition, freed from any market discipline, it would overenfranchise workers in them. There would be no sufficiently strong mechanism ensuring that anything of sufficient usefulness at sufficient productivity is done. If, on the contrary, these workers are subject to the authority of managers who are, in turn, not accountable to them (but merely, directly or indirectly, to all ordinary citizens), they are dominated.

Similar considerations apply to the proposal to socialise major means of production, bringing them under the comprehensive control of the state. ³² Whether or not, and to which extent, this might be needed to address concerns of structural economic domination, such as problems of democratic control of large-scale investment, it does not itself solve the problem of workplace domination. First, what about workers in the enterprises that are not socialised? Second, without workplace democracy, managers acting on behalf of the state will dominate workers in their enterprises. That state being democratic would reduce domination, compared to managers merely being accountable to unelected party elites, as they were in real socialist state enterprises. However, if the argument from general worker interests and the inevitable need for discretionary management actions made in previous sections is sound, problems of workplace domination that call for democratic accountability within the enterprise remain. ³³

To conclude, we need to address a serious worry for default workplace democracy that had already been mentioned in the argument for the need for standardisations: the trend towards ‘fissuring the workplace’ (Weil, 2014), outsourcing services and parts of the production chain to smaller companies subject to less regulation. It seems that setting up a democratic default for firms above a certain size will not be of much use if firm managements can circumvent it by slimming down the main body of the firm, fracturing production over many smaller firms able to escape this regulation. It might even have the perverse consequence of rendering reasonably good standardised packages for workers even less available.

Now, at the level of one political economy regulating towards workplace democracy, this worry is, in principle, relatively easy to answer: of course moving to workplace democracy needs to be accompanied by regulations making sure that this default is hard to circumvent, for example, by changing tax and accountability rules to erase increased profitability of outsourcing. If there is political will for workplace democracy, this must be part of it.

However, this is not the scenario where the worry is most pressing: this is that of increasingly globalised production chains where companies fissure the workplace over jurisdictional borders, outsourcing jobs into political economies with lower wages and worse worker protection. Here, it is easy to see how any isolated effort at ramping up worker protection by enshrining workplace democracy might well mainly lead to further job loss. This might mean that transition to more workplace democratic arrangements will need to be undertaken by a significant portion of democracies with advanced economies moving in sufficient synchrony – for example, by a coalition involving at least the US and the European Union.

However, while this is a serious worry for the feasibility of workplace democracy, it is important to note that opportunities for fracturing and outsourcing to economies with worse working conditions pose similar problems for all extant proposals to address workplace domination – or if not exactly similar, then still equally serious, problems. For example, how could an unconditional resource scheme, if set at an ambitious level which would really make a difference to workplace domination, be stably sustained if firms are set to react to its implementation by outsourcing those jobs that so empowered workers now no longer want, or for which they would demand much better conditions? Similarly, it seems questionable that states committed to being an employer of last resort could sustain such a scheme at reasonably good conditions for those resorting to it, if its implementation would have the same consequence, with the likely effect of having to enlarge that scheme significantly, exercising downwards pressure on what it can offer to workers.

Therefore, this turns out to be less an objection that applies specifically to workplace democracy, and more of a reminder that, under current global conditions, any effort at securing liberal republican non-domination at work has to give serious thought to its necessary international and transnational dimensions.