The workplace, writes Kathi Weeks, exhibits ‘the most immediate, unambiguous, and tangible relations of power that most of us will encounter on a daily basis’ (2011, p. 16). As a result, any political project committed to minimising or eliminating domination should pay very close attention indeed to the organisation of work. One potential solution to the problem of domination in the workplace, now gaining traction in both the academy and the wider political world, is universal basic income, or UBI. Many of its proponents expect it to have genuinely radical, transformative effects on the workplace and on the power dynamics between workers and employers. It is also conceptually very neat—one fairly simple redistributive policy would have knock-on effects that go well beyond the realms of more equally parcelling out the social surplus. In the language of self-help books and pop-up adverts everywhere, it can seem like ‘one weird trick’ to address the domination of workers by employers, promising quick and counter-intuitive returns on a comparatively small investment. Approaches like this, because they expect solving the problem of domination at work to require little, if anything, other than fighting for the introduction of such a policy and its predictable consequences, can be called ‘UBI-first’.

In this paper I want to argue against such approaches. The benefits of an anti-domination UBI would be real, and the effects of an appropriate UBI on the workplace would be transformative. But I argue that a UBI oriented towards solving the problem of domination at work—rather than, say, oriented towards reducing poverty, or compensating domestic labour—faces some particular obstacles concerning both the appropriate level of the income and the method by which that income would have to be introduced and maintained. On the basis of some minimal assumptions about political change, a UBI of this kind will be very difficult to implement and sustain. UBI-first strategies are at best a risky bet, staking an approach to domination at work on the unlikely introduction of a single policy. At worst, they are a distraction, and when considered sufficient or near-sufficient for the task of minimising domination at work may divert attention from more piecemeal but more direct approaches to the same ends.

This paper has three parts. In the first I explore and defend the idea that a UBI of a certain kind could minimise the domination of workers by their employers, and I give an account of some recent criticisms of that proposal. In the second part I explore the criteria of adequacy for such a UBI, and describe two particular hurdles for a UBI designed for this purpose. The level of the income will have to be especially demanding if it is to provide workers with the power of radical exit envisioned by UBI-first proponents. And the maintenance of the UBI will need to be similarly demanding: the existence and level of the income will have to be protected from electoral and political whims, both to avoid merely exchanging employer domination for state domination, and in order to give workers the credible threat of exit in the first place. These are obstacles, neither of which are faced by UBIs suitable for other purposes, and they combine to make a UBI-first approach an unstable and uncertain strategy. The third and final part explores the place of UBI within alternative approaches to the problem of domination at work, ones that do not require us to wait for the unlikely jackpot of a truly liveable UBI, securely provided by a truly non-dominating state.

This paper is chiefly a criticism of a set of strategies which consider the pursuit of a liveable UBI as a sufficient or near-sufficient response to domination at work, and an argument that those opposed to domination in the workplace (and republicans in particular) should consider universal incomes of various levels to be just so many tools in the republican toolkit. To be clear at the start, then, I do not intend to argue that a less-than-liveable UBI would not reduce domination at work, perhaps significantly. In the final part, I will argue that we have good reason to suppose that it would, and that it would reduce domination in other areas of life as well. As such, my argument should be amenable to many supporters of UBI, those who support what I call ‘UBI-too’ as opposed to UBI-first approaches.

The Case for a UBI-First Approach

UBI and the Problem with Work

A UBI is a ‘periodic cash payment unconditionally delivered to all on an individual basis, without means-test or work requirement’. This is the definition of the Basic Income Earth Network (BIEN n.d.), and will do for the general formula.Footnote 1 The income level of the UBI will vary from proposal to proposal, and the value of two UBI proposals of a similar income level will vary as well, depending on whether it is intended to supplement or replace existing public spending.

An intuitive way of sorting proposals is according to whether they provide a liveable or non-liveable income. The distinction is obvious enough—is the proposed UBI supposed to supplement income from other sources, or is it supposed to be high enough to replace income from other sources? Theorists are understandably reluctant to use too many numbers when talking about these levels of income, and in any case it is the value of the income that matters for our purposes, rather than the income itself.Footnote 2 Erik Olin Wright characterises a liveable income as one that ‘…would enable a person to live at a respectable, no frills level. That is, the level of the grant is sufficiently high that withdrawing from the capitalist labor market is a meaningful option’ (Wright 2006, p. 8). Philip Pettit, on the other hand, acknowledges that there exist various intuitive criteria of adequacy, himself favouring Amartya Sen’s basic capabilities approach to determine a basic poverty line below which recipients would not fall (Pettit 2007, p. 2, n.1). In this paper I am only interested in those theorists who propose a liveable UBI, and what the consequences of that proposal may be—so for the remainder of this paper, consider ‘UBI’ to be shorthand for ‘a liveable UBI’.Footnote 3

Onto the issue of the workplace. Roughly speaking, I am dominated to the extent that I am in a relationship with someone who has power over me, and whose power I do not control. This account of domination is not uncontroversial—it goes against, for instance, the depersonalised use of the term inYoung (1990) and Weber (2013)—but it covers the most mainstream modern conceptions of domination: those used by republicans, sometimes called ‘neo-’ or ‘neo-Roman’ republicans.Footnote 4 On this account, wage labour is worrying. People who do not own significant amounts of capital—the vast majority—must work for someone else in order to live. And once they do work for someone, they will in the overwhelming majority of cases enter a firm that is structured not as a democracy, but instead as a fairly rigid hierarchy of command and control.Footnote 5 Managers have authority over decisions that will affect workers in all sorts of ways, from everyday decisions about tasks, pacing, and conduct, to larger and more individually consequential decisions concerning issues such as the location of the firm, or its long term priorities.Footnote 6 Employment contracts provide a measure of restriction to this authority, but they are incomplete, usually providing only sketches of a worker’s duties. For theorists such as Nien-hê Hsieh, this is part and parcel of the very idea of a successful company—‘many of the decisions that need to be made within an economic organisation can only be specified during the course of conducting economic activity and not at the outset’ (Hsieh 2005, p. 122).

But however necessary this open-endedness might be for a successful business, it is also inevitable. No contract can fully legislate for what happens day to day, minute to minute, on the job. And to the extent that the residual authority to interpret those gaps in the contract rests in the hands of employers—whether they choose to wield that power harshly, or not at all—workers are dominated by those employers.Footnote 7

Radical Exit

How could we expect a UBI to minimise this domination? The most basic mechanism is a genuine ability to exit. Currently, workers have a formal right to leave any job they like, subject to notice periods and so on. But it is a largely empty right for those without independent resources, significant savings, or another job lined up. Not finding another job means throwing myself on the patchy and worn safety net of the welfare state, a demeaning and uncertain way of meeting my basic needs. As such, leaving a job is normally very costly indeed.Footnote 8 This inability to be independent from work restricts my ability to be independent at work. It will be very costly for me to leave, and my employer knows this as well as I do. So the threat of exit is barely one worth making, especially for those with low prospects of finding another job quickly. The vast majority of individual workers need their employer more than their employer needs them, so an individual worker’s bargaining power in any decision at work is low. Even lower is the chance that an individual worker might contest the overall decision-making structure of a firm, rather than any particular decision.

A UBI could turn this merely formal right to exit into a meaningful ability to exit, and various forms of UBI allow various forms of exit. Simon Birnbaum and Jurgen De Wispelaere (2016) distinguish between weak, strong, and radical forms. The weak form is the ability to drop one’s hours, creating a shorter working day or week, the money from the basic income picking up the slack. The level of the grant in this case would not need to be liveable, in the sense described in the last section. The second option, strong exit, gives workers the resources to sustain a longer search for a new job once they have left the old one, making their threat of exit more credible and thereby reducing power imbalances in the workplace. Alternatively, those workers could simply use this power of strong exit and find a different, less dominating job. Again, the level of the grant required to provide strong exit would not have to be liveable, since it does not rely upon enabling independence from all work, just independence from one’s particular job.

From the perspective of reducing domination, strong exit suffers from some problems. Say that I use my power of strong exit, secure in the knowledge that my finances can now withstand a relatively prolonged hunt for a new job. Even if I manage to find another job within a week, this will be of little comfort if the problem with my old job was that I resented managerial authority. There are, and under current conditions perhaps only can be, few opportunities to work for a cooperative or be meaningfully self-employed.Footnote 9 My new boss might be nicer or more permissive, but absent a significantly different power structure this means little from the perspective of radically reducing domination—I do not control their interference in my life, so the basic relationship is still a dominating one. There may be many ways to act upon the background conditions that make default employment contracts relationships of domination in this sense, but on the face of it the power of strong exit—the effective power to move between workplaces—is not one of them.Footnote 10 Even when it comes to redressing power imbalances within the workplace, Birnbaum and De Wispelaere show that strong exit can produce some unwelcome consequences. While some highly skilled and valuable employees might be able to effectively increase their bargaining power, the most dominated workers, those who might be easily replaced, will still be at the mercy of their employers. Their employers ‘have very little incentive to accede to any of her demands, for the simple reason that replacing that worker with another one may be the cheaper option’ (Birnbaum & De Wispelaere 2016, p. 67). In those cases where employers decide to keep on vulnerable workers, and change in accordance with their demands, they are not being forced to do so. It amounts, they write, ‘to a privilege being granted by the employer out of the kindness of her heart. This is a far cry from the change in the power relations between employee and employer’ (Birnbaum & De Wispelaere 2016, p. 67).

Radical exit, the third option, is the one most relevant to my argument and to the arguments of most of the theorists that adopt UBI-first positions. This power of exit provided by a truly liveable UBI would, in principle, work in a number of different ways to effective fight domination in the workplace. In the most basic case, a dominated worker can simply choose to exit the labour market entirely, falling back on the adequate income provided by the state. If you are dominated by your employer, and all alternative jobs are similarly dominating, then a UBI allows you to escape. But we do not need to go too far into the economics of a basic income to recognise that this cannot be the answer for every dominated worker (at least, not at the same time) if a UBI is to be sustainable. A sufficient number of people need to work and—we might assume, for the purposes of this argument—want to work, just not in jobs that are dominating and hierarchical, demeaning, or otherwise unpleasant and unrewarding. In any case, the more interesting and transformative effects of a UBI are indirect, using newfound independence from work to produce independence at work.

If workers can simply leave and subsist on an adequate level of basic income, then they can very credibly threaten to do so rather than suffer under the dominating power structures of their workplaces. More than this, employers will know that their workers have this option. In response to some gross overstep of managerial power this might take the form of an actual threat, but in normal circumstances it would function as an implicit threat on the part of the worker. The threat to strike works in the same way: where the right to strike is protected, that threat functions quietly and implicitly, and needs only rarely become explicit. The genuine ability to exit would become more than a tool to contest, after the fact, managerial decisions; it would be an ever-present possibility, raising the bargaining power of individual workers and reshaping their relationship to their employers. Rather than just the ability to exit, an adequate UBI gives workers various abilities—by the reckoning of one supporter, the powers to ‘enter, undominatedly stay, exit, and restart all kinds of social relations, starting with work relations’ (Casassas 2016, p. 9). In this sense the power structures of individual workplaces would be transformed, with managers no longer able to monopolise the residual authority described in the previous section.

On a wider scale, proponents of UBI argue that this individual bargaining chip would have a collective effect on the labour market as a whole. A whole class of very dominating jobs will become almost impossible to fill. ‘On the plausible assumption that the great majority value their enjoyment of non-domination highly’, writes Frank Lovett, ‘few will [accept dominating work] except where necessary to meet their basic needs’ (2010, p. 197). Outside the relationship of any individual employer to any individual worker, employers as a class would be forced to change the jobs they offer. Workers would have the advantages of a tight labour market magnified beyond recognition. And while some dominating jobs would be rendered attractive with higher pay and better conditions, it is plausible to assume that directly reducing domination—offering contracts with less room for managerial authority, and reshaping power structures to give workers control over their jobs—would be a winning and necessary strategy for attracting and retaining employees.Footnote 11

On a more speculative and indirect level, a UBI may well have other transformative effects on work. With the security and bargaining power that it provides, we can expect many people to work in formal, paid employment less, and some people to not work at all. This newly freed-up time allows for activities that we might loosely group together as ‘experimentation’. The direct, decommodified provision of labour would be a real possibility to all. The ability to rely upon a basic income would breathe life into cooperative and non-hierarchical working arrangements, taking much of the personal risk out of what is otherwise a very risky endeavour. And time is freed up too for political action, on whichever level—engaging more with the state, or organising workplaces and communities autonomously. Although harder to predict, these knock-on effects of free time and economic independence may be the most transformative of all: challenging other sources of domination in society, or engaging in the non-dominating provision of goods and services outside the strictures of a firm.

A UBI of this genuinely liveable kind would eliminate a basic inequality between owners of the means of production and workers—the ability of the former, but not the latter, to subsist without selling their labour—and as such the claims made on behalf of this kind of UBI as a solution to domination in the workplace do not seem unreasonable. We could expect a labour market with a liveable UBI to provide an escape route for very dominated workers, to exhibit significantly less overall domination than it currently does, and—crucially—to encourage longer-term movement away from dominating employment relations. And, although there may be many routes to the same goal, UBI is a particularly tidy way of getting there. By guaranteeing material independence, and with only the assumption that people will not accept domination if they have other acceptable options, the decisions of countless individual workers about which jobs (if any) to accept will exert pressure on the system as a whole. Without state coercion, significant collective action, or the need for drastic invigilation of firms, dominating jobs would wither away.

So is UBI enough on its own? The answer differs according to who you ask. For certain republicans, like Frank Lovett, a free market and a UBI are sufficient. In fact, Lovett thinks that workplace regulation could be repealed, saving on the costs of inspection, tribunals, and enforcement (Lovett 2010, p. 200). This is a minority view. More common are approaches like Philip Pettit’s, in which a UBI of this kind would be a not-quite-sufficient condition for reforming the workplace, doing the heavy lifting alongside a raft of smaller, tinkering proposals: some increased workplace regulation, the prosecution of offending firms in a public court, the removal of certain legal barriers to collective bargaining, and the prohibition of some practices like termination-at-will and non-compete agreements (Pettit 2007, 2012, p. 113–116).

Existing Criticisms of UBI

Before moving on, I want to discuss some existing criticisms of an anti-domination UBI, and criticisms of UBI in general that would affect its ability to reduce domination at work in this radical sense. I disagree with some and agree with others, but it is valuable in any case to get clear on the state of the field.

Taking the broad sweep of academic and non-academic work, criticisms of UBI work across a number of levels of ideality, from ‘a UBI would strip work of dignity and therefore be morally wrong’, to ‘a liveable UBI would not sufficiently reduce the domination of workers’, to ‘a UBI of any kind is not politically possible given the current electoral arithmetic’. These are very different kinds of critique. Some cut to the very heart of the idea of a UBI, and others merely raise doubts about its implementation. UBI might violate some important principle of reciprocity, or might constitute the exploitation of productive citizens by less productive citizens, for instance.Footnote 12 Two criticisms are especially relevant for my purposes here, and both are well represented in the work of Alex Gourevitch (2013, 2016) and Lucas Stanczyk (Gourevitch and Stanczyk 2018). Respectively, they claim that an anti-domination UBI would be ineffective, and that even a modest UBI would be impractical.

To claim that an anti-domination UBI would be ineffective is to deny the mechanisms I describe in the last section, to deny that a UBI could redress power imbalances between workers and employers, transform the labour market, and so on. So for this first criticism, the question of how difficult it would be to win a UBI of this sort is irrelevant—assume that we could click our fingers and implement such a policy. Gourevitch makes two arguments against the transformative potential of that UBI. The first argument denies that a UBI, of whatever level of income, could sufficiently meet the costs of leaving a job, and would therefore fail to give workers the power to credibly threaten to do so. He distinguishes between instrumental costs, which include ‘work-related social standing…the needs of family, ties to community, [and the] value of workplace relationships’ (Gourevitch 2016, p. 24), and intrinsic costs of exit, where ‘many [workers] will still validly feel an obligation to work and/or they will seek a creative outlet for their capacities’ (Gourevitch 2016, p. 24). These are things that money cannot buy. You can make leaving a job less costly, but these non-material costs establish a hard floor to the project of giving workers a credible threat of exit.

The second, related argument for the ineffectiveness of a UBI denies that the threat of exit could ever sufficiently increase the day-to-day bargaining power of individual workers, even if exit costs were radically reduced. For the purposes of contesting discrete and exceptional managerial decisions, the threat of exit might work well; if your boss threatens to cut your pay in half, you can threaten to leave. But Gourevitch writes that ‘threatening to leave a job is a kind of nuclear option that is simply not credible in many low-level disagreements’ (Gourevitch 2013, p. 608). These low-level disagreements are the norm. You might disagree with your boss on how best to perform some task, in which order to perform a series of tasks, who might best perform that task, or a range of pettier issues: what constitutes appropriate clothing or speech in the workplace, how to keep your workstation or desk, and so on. That each disagreement might be small—even insignificant—means little. The series of disagreements, and the regularity with which the manager can enforce their will in those disagreements, characterise the ongoing relationship with a manager as one of living under the power of another. That series might also be the only obvious expression of that domination, if no individually very consequential decisions need to be taken. But in such cases, reacting to any of those decisions with the threat of exit is not particularly credible. Are you really going to leave your job because your boss told you to organise those files alphabetically, rather than by date? For Gourevitch, this coarse-grained power is not a useful corrective to the fine-grained reality of workplace authority.

Although a response to this first set of critiques is not the object of this paper, I suspect that Gourevitch somewhat overstates the case against the effectiveness of a UBI here. He is right to say that there are many non-material costs of leaving a job that could not be compensated for, in kind, by any basic income. But in order for the structural effects of a UBI to kick in, leaving one’s job does not need to be costless, but merely low enough and bearable, for that threat to be used as an effective bargaining chip. In order to make those non-material costs bearable, the level of a basic income may well need to be higher than strictly required to compensate for the material costs of exit. But the brute fact that some kinds of costs cannot be eliminated by a basic income does not establish that a basic income would fail to make most workers’ threat of exit credible.Footnote 13 Birnbaum and De Wispelaere, for example, include a similar list of ‘non-monetary, social values [of employment] that cannot be compensated for by basic income in any immediate way (Birnbaum and De Wispelaere 2020, p. 9)’ as a reason to doubt that partial basic incomes could give workers a credible threat of exit. This is compatible with my suggestion that a suitably generous income may nonetheless make exit a realistic option for many workers. Similarly, Gourevitch’s worry about the unwieldy ‘nuclear option’ of the threat of exit is a real one—employers may well reasonably doubt that workers will quit their jobs over some individually inconsequential managerial decision. But this worry may be overstated for the same reason as it is a worry in the first place—a series of such decisions, and the relationship that results from it, is consequential enough and harmful enough that the threat of exit is a perfectly credible response. ‘I’ll leave unless you reverse your decision to make me sort these files inefficiently’ is not credible; ‘I’ll leave unless you stop enforcing petty decisions, the latest of which is the decision to make me sort these files inefficiently’ is much more credible, and reflects the problem with the failure of the first example. In any case, proponents of a work-orientedFootnote 14 basic income do not see the threat of exit as primarily an explicit threat, a mechanism to discipline the boss triggered in response to some decision. Instead, they see it as an implicit threat, one attaching risks to the continued exercise of power formally available to managers. This is enough, I think, to establish that we need more than Gourevitch provides in order to show that any basic income will fail to seriously alter the bargaining positions of workers and employers.

The second kind of criticism is that even a modest UBI would be politically impractical—Gourevitch and Stanczyk argue that a ‘bare subsistence’ (2018, p. 5) basic income of $15,000 a year would be almost impossible to achieve. Specifically with reference to the US, the sheer level of taxation and appropriation required to finance a UBI of this level, and the predictable opposition of the classes of owners who stand to lose from it, make even this low income a far-off dream. Their conclusion is that viewing UBI as a relatively easy route to a much stronger working class—a ‘capitalist road to communism’ (2018, p. 2, after Van der Veen & van Parijs 1986)—gets it the wrong way round. Overcoming these obstacles and winning a UBI would require the strong and unified working class that UBI promises, class power that must be achieved through other, more traditional means. They make the case, we could say, that we would need a communist road to a capitalist road to communism.

Responding to this criticism is much easier: although it does not entail it, my argument below is perfectly compatible with their doubt that even a modest, bare subsistence UBI would be practical or affordable. My goal is different—I want to argue that the kind of UBI required for transforming work will demand more than bare subsistence, and be more demanding than UBIs suitable for other purposes. For those convinced by Gourevitch and Stanczyk’s argument, this paper offers more bad news for defenders of a work-oriented UBI. But it is possible to reject their conclusions and accept mine: to think that a bare, subsistence UBI would be practical and achievable, but that an adequate UBI for the sort of UBI-first strategies I am concerned with would likely face significant additional barriers.

The Case Against

The Level of the Income

As discussed above, quite what it would mean for a UBI to be ‘liveable’ is unclear. Erik Olin Wright opts for a no-frills level of income, whereas Pettit, nodding to Sen’s capabilities approach, nevertheless accepts any ‘intuitive criterion of adequacy’ (Pettit 2007, p. 2). In this section I want to argue that several nevertheless intuitive criteria of adequacy are the wrong ones to aim at in order to produce the structural transformation of the labour market that UBI’s more optimistic proponents rely upon: an anti-domination UBI of this kind would have to provide a subjectively acceptable standard of living, and do so as obviously as possible. As a result, it will likely have to be especially generous compared to those UBIs suitable for other purposes.

For the purposes of reducing poverty something like a bare subsistence level, or indeed a capabilities approach to determining the level, is perfectly reasonable. For the purposes of reforming a humiliating and conditional welfare regime, the level of an adequate UBI would not have to exceed the level currently offered conditionally, with the mechanism of delivery doing the work. If we wanted to make sure that workers freely chose to work for their employers, such that they were not forced to do so, we could aim for an income that provides a ‘reasonable or acceptable alternative’ (Cohen 1983, p. 4) to work. In this case, what constitutes a reasonable alternative will be a matter of fact, either an entirely objective account of a minimum sufficient level of welfare, say, or an option that a reasonable worker would choose to live upon. For the purposes of transforming work, none of these standards will do.

The mechanisms discussed in the last section depend upon workers being able to credibly threaten to leave their jobs, and the more structural mechanisms depend upon many workers actually leaving their jobs, and many people without jobs refusing to take dominating jobs for long enough to put pressure on recalcitrant firms. This relies upon the alternative—living on the income provided by a UBI, and nothing else—to be subjectively acceptable to as many workers as possible, and as obviously as possible. Characterising this level of income is fairly easy: as much as it takes for people to leave their jobs in sufficient numbers (and be able to credibly threaten to do so), and for people to remain unemployed in sufficient numbers and for sufficiently long. Would people leave their jobs if their UBI did not allow them to save enough to replace a broken boiler, or run a car? If they could not go on holiday, or buy Christmas presents that do not make their children feel hard done by? If they could not have a full social and cultural life? Some workers, and certainly those workers currently unable to afford those things anyway, would choose to subsist on such an income. A UBI of this level would have much to recommend it—it would reduce poverty and reduce the domination of some workers. But if everyone knows that life on a UBI is dreary, thrifty, but just about tolerable, then the relative bargaining power of most employees is not dramatically increased, and the structural effects of a UBI on the workplace would likely fail to kick in. Quite how demanding that level of income would be is a complicated empirical question, and will in any case differ according to how much one is expected to buy on the market (in some countries the value of the income would have to cover health insurance, in others education for one’s children).

In order to start answering that complicated empirical question, it may be helpful to think about some of the less obvious costs of leaving a job, for which an appropriate UBI of this kind would have to compensate in order for exit to be subjectively acceptable to a large enough number of workers. I have already discussed some in the last section—the social ties to one’s workplace, colleagues, community (Gourevitch 2016, p. 24); the status and value of working in a society in which unemployment is stigmatised (Birnbaum and De Wispelaere 2020, p. 9). Birnbaum and De Wispelaere offer some other hidden costs of exit, including the loss of employment-related social protection (pension contributions, for instance), and ‘scarring effects’ of unemployment that might make it harder to find a job in future, such as poor references or gaps in a CV (Birnbaum and De Wispelaere 2020, pp. 8–9). As in the last section, these costs may well be able to be compensated for satisfactorily by a high enough level of income, such that workers really could, and would, leave the workplace in great enough numbers. But that level of income will be unusually demanding, more demanding than would be required for a UBI suitable for other purposes, if it is to have its more radical intended effects.

Many of the obvious practical obstacles to achieving a UBI grow larger along with the proposed level of the income. As Gourevitch and Stanczyk (2018) discuss, funding even a modest UBI would require very high taxation or expropriation of wealth, both proposals certain to generate a great deal of opposition from the targets of that tax or expropriation. It seems uncontroversial to assume that the higher the UBI, the greater that opposition. Many intuitive and good-faith objections to the unconditional nature of a UBI also scale with the level of the income: if it is unfair to give everyone a $15,000 a year for nothing, then it is even more unfair to give them $25,000. If the lower income generates an unmet reciprocal duty on the part of the recipient, then the higher income generates a more demanding and equally unmet duty. And if a lower income will attract freeloaders, a higher income will surely attract more of them. Even if defeasible, these intuitive objections will naturally be exploited by those with a material interest in opposing that UBI—the wealthy and the owners of productive assets—and exploited more and more easily the higher the level of the income. If a work-oriented UBI must be much more generous than other potential UBIs, it will as a result be much more difficult to achieve.

The Stability of the Income: The Problem of State Domination

Another difficulty for proponents of a work-oriented UBI lies in the required stability of that income. Again, a comparison with UBIs suitable for other purposes is helpful—assume now that the level of the income remains the same between examples. A UBI that reduces poverty will do so for as long as it is provided, and will stop doing so as soon as it is withdrawn or reduced to below the poverty line. A UBI that reforms a degrading, conditional welfare regime will do so for as long as it is provided, and will stop doing so as soon as it is made conditional. The problem of an unstable UBI—one that could be reduced or eliminated at any moment—is indeed a problem for these purposes, but it is a future-oriented problem. For as long as that unstable UBI is in place, it has its intended effects, but it would be better if we could ensure that those effects would continue for as long as possible. The issue of stability for a domination-oriented UBI is more complicated. Republican proponents already appreciate this, for distinctively republican reasons. In this section and the next I want to briefly highlight quite how demanding that republican stability criterion is, and then give another, more general reason—one that non-republicans must accept as well—to doubt that an unstable UBI is adequate for the purposes of transforming work.

From a republican perspective, Pettit argues that any UBI must be ‘non-nullifiable’ (Pettit 2007, p. 2). That is, it must be something to which citizens are entitled as a right, rather than merely as a kindness—it must be ‘suitably entrenched’ (Pettit 2007, p. 3) against change or dilution. This reflects two distinctively republican fears: the domination of citizens by the state, and the domination of recipients of charity by their benefactors. Welfare payments in general ‘must not have the aspect of a gift that may be withdrawn…not even at the whim of an electoral or parliamentary majority’, Pettit writes, ‘unless the income support comes under conditions that guard against such whim, the claimant is denied freedom as non-domination in the very act whereby that freedom was meant to be promoted’ (1997, p. 162). From the perspective of avoiding domination overall, how a state provides a UBI is just as important as how much that state provides—getting it wrong risks reducing domination by employers while increasing the domination by the state.

Although these are worries raised more in the republican tradition than elsewhere, they are salient ones for anyone opposed to domination in general, rather than just with the domination of workers by their employers. If the argument is convincing, then introducing a UBI is certainly risky. Introducing a UBI that successfully transforms the workplace is even riskier, since it would involve introducing an income that most people would be willing to leave their jobs for, hoping that large numbers of workers or unemployed people would move towards relying entirely on the state for their income. Pettit’s solution is to insulate the level and the existence of the income from short-term electoral contestation, perhaps constitutionalising the UBI. Quite apart from the practical difficulties of guaranteeing the stability of even a constitutionalised basic income, this magnifies the predictable obstacles faced by a UBI in the same manner as in the previous section. If a UBI hurts the rich and powerful, it will stand to hurt them more the harder it is to water down or repeal. If it is unfair, it is even more unfair when it cannot be electorally overturned like most other policies. In this sense too, then, a work-oriented UBI is more demanding than a UBI required for other purposes.

There might be plausible optimistic responses to this republican demand for stability. Assume that we introduced a UBI tomorrow, of an adequate level, but one that could be withdrawn by a simple executive decision. UBI recipients would no longer be dominated by their employers (for the sake of argument), but they would instead be more dominated by the state. Is not this an improvement? There is no obvious reason to suggest that any individual will be more dominated than they were previously, merely that their domination has been shuffled about. And, at the very least, this consolidates the problem. Instead of limiting the power of a million little tyrants in the form of employers, the route to non-domination for those in receipt of a UBI is clearer and easier to pursue—press to constitutionalise that UBI, or press for the further reduction of state domination in other ways. An insecure UBI solves the problem of domination in the workplace, and the additional state domination to which its recipients are subjected is in some senses more tractable. As such, the demand for stability does not make it impractical to radically reduce the domination of workers by their employers—that can be achieved even with an unstable basic income—and UBI-first approaches to the workplace are still a realistic strategy.

The Stability of the Income: The Problem of Effectiveness

A quite different and more direct case for a stable basic income can be made. The republican argument is compatible with an unstable UBI effectively and radically reducing the domination of workers by their employers, albeit increasing the domination of those workers as citizens in the process. Instead, I want to argue that an unstable basic income will likely be less effective even for the purpose of transforming the workplace.

As discussed above, the more transformative effects of a UBI on the workplace only kick in if most workers would be willing to leave their jobs and subsist upon it. The ability to credibly threaten to do so increases the bargaining power of workers, and the actual withdrawal of many workers and potential workers from the labour market would put pressure on firms to reform in order to attract and retain workers. Both mechanisms require a UBI attractive enough to compensate for the costs of leaving a job, and an uncertain UBI, one that might change on the political winds, is an unattractive UBI. For those living on it, anxiety and insecurity are real psychological harms, ones that we tend to recognise as some of the worst features of insecure employment. Even supposing that the level and existence of a UBI could only be changed at elections, for recipients this would amount to a series of four-year contracts, with no guarantee of an extension at the end of each. The more unstable the UBI, the more precarious its recipients would become—if it could be repealed by simple executive decision, every day might be the last day that recipients could meet their basic needs. Some longer-term insecurity might just be the cost of democracy (and removing that insecurity would, therefore, not obviously increase non-domination overall). But for relatively unstable UBIs, leaving one’s job would be choosing a life of walking on eggshells—it seems much less likely that people would do so, or be able to credibly threaten to do so, in sufficient numbers to act as a sufficient or near-sufficient solution to workplace domination.

A UBI whose future is uncertain in the short- or medium-term is one that will not provide a particularly credible threat of exit. Who would leave their jobs in an election year? Which employers would be worried by the threat to do so? And an uncertain UBI is also not one that can provide the longer-term withdrawal of labour required to provide the benefits of a radically tight labour market for workers. Contrast this again with a UBI for the purpose of, say, poverty reduction. An unstable poverty-reduction UBI is worse than a stable one because the former, but not the latter, might stop reducing poverty in the near future. An unstable domination-oriented UBI is worse than a stable one because it is less effective even while it exists. The future stability of a basic income directly affects its current value, and an unstable basic income is, in effect, an unliveable basic income.

This argument for the stability of any UBI up to the task of transforming the workplace is more direct than Pettit’s, and has a broader audience: even those who reject the republican fear of benefactors dominating the recipients of their charity now have a strong reason to require such stability. This leaves open the mechanisms by which that stability would be increased, and experimentation may be required in order to find suitable ones. De Wispelaere and Morales discuss various threats to the political stability of a UBI over time—outright replacement of the scheme by an incoming legislature, but also effective dilution of the level or universality of the income. Brazil’s 2004 experiment with a basic income (‘Lei de Renda Básica de Cidadania’) was quickly rendered ‘little more than scrap paper’ (De Wispelaere and Morales 2016, p. 529) by a clause that subjected it to budgetary constraint, leaving the scheme at the mercy of successive governments’ unwillingness to provide the necessary resources. They explore the possibility of constitutional protection in a bill of rights, ‘protected by a constitutional tribunal or supreme court that has the power to overrule legislative decisions’ (De Wispelaere and Morales 2016, p. 532). Even this, they argue, may not provide the kind of stability against dilution or replacement that UBI-enacting coalitions might desire—judicial review in practice often defers to other branches of government in complex issues of resource allocation, or else comes under extreme pressure of charges of judicial activism. Constitutionalising a UBI is not the only mechanism for producing stability—earmarked sources of funding, as in the case of the Alaska Permanent Fund Dividend, or citizen dividends from sovereign wealth funds, might also provide sources of political resilience. The point is not that a liveable UBI could not be made more stable in the short- or medium-term, but that high stability—however it is achieved—will be a feature of the kind of UBI required for the sort of strong-exit UBI that may have the radical effects on workplace domination discussed above. And it seems safe to say that any stable UBI will be more politically demanding than an unstable UBI.

Taken together, a work-oriented UBI capable of transforming the workplace requires an unusually high level of income, secured beyond normal electoral contestation. Both are tall orders, and are not required for UBIs sufficient for other purposes. Unless and until we have a genuinely adequate UBI, implemented and maintained in a genuinely secure manner by an appropriately constituted state, the policy’s more transformative goals will not be achieved. This hits upon the core problem with relying upon a UBI of this kind as a sufficient or near-sufficient approach to domination at work. The star attraction of a UBI-first strategy is its simplicity. We do not need to directly fight for workplace democracy, for instance—once we have material independence, we can expect an indirect democratisation of the workplace to emerge from the ensuing shifts in the balance of power. But if the road to a sufficient or near-sufficient UBI is this difficult, much more so than for other purposes, then it seems unwise to put all of one’s eggs in this particular basket. This criticism certainly applies to Frank Lovett’s UBI-and-free-markets approach, but also to more diverse approaches in which non-UBI measures are only given real force once a UBI of the right kind is up and running.

Conclusions

Even though a UBI capable of ensuring radical exit would likely be sufficient or near-sufficient for minimising domination at work, I have given reasons to suppose that this kind of UBI would be more demanding than is often supposed by its proponents. To that extent I have also argued that this will make it more difficult to achieve and maintain than is often supposed by its proponents and, as a result, its apparent benefit as a shortcut to a transformed workplace is weakened. UBI is, broadly, presented as a response to three problems: poverty, unemployment (aided by automation or otherwise), and domination at work. The choice of purpose matters: distinguishing between them is vital for assessing the adequacy and feasibility of some proposed basic income. When it is presented as a sufficient or near-sufficient response to domination in the workplace, it is helpful to be clear about which kind of UBI we would need to accomplish that goal, how difficult it will be to win and maintain, and which more direct and collective strategies are being neglected in pursuit of that goal. But abandoning UBI-first strategies—and even abandoning for now the possibility of a liveable and stable UBI in the sense discussed above—does not require us to abandon UBI as part of an anti-domination project. In this section I want to discuss a number of ways in which UBI or similar schemes might fit into a wider strategy for minimising domination in the workplace and elsewhere.

Would a less-than-liveable and less-than-stable UBI reduce domination at work? It seems very likely. It also seems very likely that it would reduce domination in other, non-work areas of life. But the fact that it will be helpful does not mean that it will be sufficient or near-sufficient, the claim that I am arguing against in this paper. As my title suggests, this is not an argument against UBI, but against the false hope of UBI.

Many proponents of a work-oriented, anti-domination UBI do not suggest a UBI-first approach. David Casassas and Jurgen De Wispelaere, for instance, argue for a three-part republican strategy of a UBI, an economic ceiling, and the direct pursuit of workplace democracy (Casassas and De Wispelaere 2016). UBI functions as a means of providing independence from the workplace, an economic ceiling functions to deny the very rich undue political influence, and workplace democracy functions to provide independence within the workplace. The hallmark of these more plural strategies is that they are ‘UBI-too’ instead of UBI-first; if we accept that UBI should be at most one long-term strand of a plural approach, we can in the meantime directly pursue what it promises to secure indirectly. Here I want to briefly discuss five ways in which a UBI that is less-than-liveable or less-than-stable may nevertheless be very effective at reducing the domination of workers by their employers.

First, a less-than-liveable UBI would, to the degree that it replaces income, allow workers to reduce their working hours. If time spent at work is dominating, then spending less time at work will lead to less domination. And the available studies suggest that people would use additional income to work less. There is a large group of people—some studies put it as high as half the workforce—who would choose to work fewer hours than they currently do, even for a proportionately reduced wage.Footnote 15 That is, many people are ‘overemployed’—they would accept lower wages for fewer hours, but that choice is denied to them by their employers and by the opportunities of the labour market as a whole. We should assume that the group of workers who would prefer to work less and retain their current wages would be even higher than this group. As White (2008, pp. 131–133) discusses, there are some barriers to reducing working hours—employers may not offer that flexibility, and coordination problems with fellow overemployed workers may mean that nobody wants to be the first to ask for a part-time role. Assuming that working hours legislation or collective bargaining could make dropping one’s hours a real possibility for workers—and also assuming that managerial authority over non-work time could be adequately protected against—this would be a valuable strategy for reducing domination.

Secondly, a less-than-liveable (or less-than-stable) UBI would likely increase the bargaining power of many workers by allowing them to credibly threaten to leave their current job for another. This is Birnbaum and De Wispelaere’s power of ‘strong exit’: by providing workers with enough money to survive a longer job search (even if not a sustained break from work), or by opening up options for training or education to increase the range of available jobs, individual workers may be able to leave particularly dominating jobs and better contest managerial power within their current jobs with the implicit or explicit threat of exit. The two basic problems mentioned in the previous section do not disappear. For many easily replaceable workers, the threat may still be neither costly enough nor credible enough to increase bargaining power significantly, but it does not need to increase bargaining power significantly to reduce domination in the workplace.Footnote 16 And the paucity of jobs without significant managerial authority (meaningfully cooperative or democratic workplaces, for instance) means that individual strong exit may just mean switching between relatively homogenous authority structures all of which exhibit significant domination. You may find a better boss, you may even find a less dominating or more regulated authority structure, but a UBI that allows individuals to swap one job for another is less likely to have the more radical effects on the labour market as a whole promised by a liveable UBI. But even this last point will be a matter of degree. Although a power of strong exit may not give workers the power to say ‘no’ to the offer of any dominating employment contract (because there may be no non-dominating contracts available), White (2020) suggests that part-time workers and workers taking breaks from employment may contract labour supply enough to have some significant structural effects. Arguing against the notion that a partial UBI may in fact weaken low-paid workers’ bargaining power (see Birnbaum and De Wispelaere 2020 for discussion), Calnitsky’s (2018) study of a guaranteed minimum income scheme in Manitoba lends some empirical support to the idea that a partial UBI may well tighten labour markets, which in the case of the Manitoba experiment led to higher wages and lower hours.

Thirdly, a UBI strong enough to provide a temporary cushion to loss of pay may make collective bargaining more effective. Wright (2006, p. 8) describes a liveable UBI as an ‘inexhaustible strike fund’, allowing striking workers to wait out their employers. A less-than-liveable UBI provides a less-than-inexhaustible strike fund, but that is much better than no strike fund. It would provide some portion of one’s income that could not be docked or withheld by one’s employer. It may also increase—through the ability to drop one’s working hours—the energy and time available for at-work or off-work political organising, and so help to breathe life into other, more direct strategies for reducing domination at work.

The fourth anti-domination benefit of a less-than-liveable UBI is less closely connected to the workplace and the employment relation. Unconditional income of any non-trivial level has many other potential benefits—it may reduce the domination of women by men within the family (See, e.g., Pateman 2004, pp. 98–101); of disabled people (O’Shea 2018); of would-be debtors by exploitative creditors; it would reduce poverty, which makes one vulnerable to dominating relationships of various kinds. Most people who think that domination in the workplace ought to be minimised rest that claim on a more general opposition to domination. Certainly, all republicans do, and so a UBI that is insufficient for minimising domination in the workplace but that would reduce domination there and in other spheres is, of course, still worth taking seriously.

Finally—and, again, drifting away from the workplace—a UBI of even a relatively undemanding level would seriously reduce the domination of those claiming conditional benefits, where those conditions and their method of evaluation are not sufficiently controlled by the claimant themselves. This may sound like an obscure way of putting the point, but it captures two senses in which conditional benefits regimes may dominate claimants: they may give frontline assessors significant personal discretion to decide whether the relevant conditions are met, or they may impose conditions that are not meaningfully controlled by the claimants even democratically (in an extreme case, for instance, if benefit conditionality were determined by a referendum in which existing claimants were disenfranchised; in a less extreme case, if the state in question were to exhibit more general democratic deficits). They may of course exhibit both features (see Lazar 2019 for some discussion of this point). Making those existing benefits unconditional would at the very least entirely remove the dominating power of assessors (see e.g. Eleveld 2020). A less-than-liveable UBI would be entirely adequate for this purpose, given that this reduction in domination has nothing to do with the level of state payments but their lack of conditionality.

A less-than-liveable UBI could be very effective at reducing domination in the workplace and elsewhere, and could—in the sense described above—complement other strategies for reaching the same goal: increasing the effective collective bargaining power of workers, reducing working hours, introducing or making viable democratically organised workplaces, subjecting economic production to more general democratic control, and so on. But the fact that a liveable UBI may be sufficient or near-sufficient for reaching those same goals does not mean that UBI-first strategies are sufficient for those concerned with domination in the workplace. At its very worst, the pursuit of UBI as a panacea might serve to displace these other, more direct ways of fighting domination in the workplace. And for republicans in particular, who have been criticised for a perceived lack of interest in issues of workers’ rights and economic democracy,Footnote 17 placing all their hope in a UBI-first strategy can seem like just another way of kicking the solution to those problems down the road.