There have been significant debates around the Labour Party’s increasing sympathy for universal basic income (UBI). Though UBI’s alternatives have also been discussed by both Jeremy Corbyn and John McDonnell, UBI itself has been the starting point for policy responses to increased precariousness of employment and the possible effects of greater automation, with McDonnell setting up a working group to look into the idea in February last year.

The concept of UBI has been a mainstay for left and green parties in recent decades, and has rapidly become mainstream as it gains increasing support from both centrist parties and individual capitalists, becoming at least an inspiration for further policy development across the political spectrum. However, as this concept has become more popular even among mainstream commentators who seem increasingly ready to entertain at least one form of radical change, it has remained controversial among Marxists.

Authors in the “post work” tradition such as Nick Srnicek and Alex Williams see a UBI as a central feature of a highly automated post-capitalist society. Within capitalism, it is put forward as a transformational reform, smoothing the path towards post-capitalist society by challenging the ideology of work that makes one’s job a signifier of social purpose and worth.

In contrast, the “workerist” tradition has argued that it would undermine the key aim of post-capitalism. A recent critique in Jacobin magazine is exemplary. UBI proponents are charged with abandoning the idea of worker’s emancipation in the workplace through control of the production process. UBI, it is said, offers workers only a subordinated and alienating position within a market economy.

So should a UBI be part of a socialist vision for the future? Does a UBI even make sense as a transformative reform within capitalism? Debates on UBI as a practical policy are well-rehearsed, but this article seeks to answer these questions through a critical review of the Marxist perspectives on the concept.

Freedom in work

“Workerism” is the often derogatory term for Marxist analysis that dogmatically privileges manual production workers as creators of economic value and agents of history, calling for workplace organisation to the exclusion of all else. Though this is a fair criticism of specific Marxist movements, much of the Marxist tradition is “workerist”—with a small “w”—in emphasizing wage labour as the origin of economic value. As workers, we are said to create value and continually remake the world, yet at someone else’s direction. Seeing ourselves as passive “employees” rather than active creators is crucial to the concept of alienation, which underpins capitalists’ ability to expropriate value from production, yet herein is said to lie our latent revolutionary potential.

Accordingly, emancipation cannot be seen as freedom from work, which would appear to be an abdication of agency. Instead, class struggle centres on the workplace, on the fate of the value we create, and on the control of the work process itself. The goal is freedom-in-work through the socialisation and democratisation of the workplace. The free association of producers replaces wage labour.

This position can be caricatured—and historically often has been—as ignoring those who are outside the realm of capitalist production. However, it maintains that unpaid work (such as domestic labour) can, like all labour, create use value, which is something to celebrate. Greater recognition of this fact gave rise to ideas such as “wages for housework”. Although proposed in opposition to narrow workerism, these ideas still embody a certain workerist logic: the response to unpaid labour is to demand that it should be recognised as labour proper by linking it to a wage, and that it ultimately should be socialised, in order to emancipate it. In this vision, the realm of work expands outwards, but work is to become self-directed, and no longer alienating.

In this context, it is not hard to understand the workerist opposition to UBI. To the extent that UBI means we can drop out of work, we gain “free time”, but we lose the historical agency we have as workers. As UBI-funded consumers, we are seen as passive, alienated, taking as given a world shaped by others. We may do work, such as housework, that creates use value outside of the market, but there is no reason to believe that this work must be “free”: it may fall victim to other structures of domination such as patriarchy. Might not UBI create an enormous pressure to “volunteer” in some approved manner, to symbolically earn one’s UBI cheque?

From the beginning, Marxism has pointed out that wage labour offers an escape from “pre-capitalist” structures, even as it imposes a new form of subjugation. Workerism embodies the classical Marxist teleology that sees capitalism as a step on a liberatory path that is only completed through the socialisation of labour. From this perspective, UBI seems to miss a crucial step in the process towards post-capitalism.

This critique recognises the fact that UBI is usually proposed as a specifically social democratic reform within capitalism. Despite all the talk of “free time”, such a formulation of UBI can only be intended to leave the system of wage labour intact, and this is how left critiques of UBI tend to understand it. As such, the liberation involved appears limited: if we still have to perform wage labour to supplement our UBI cheques, then we still cannot escape the alienation of working as someone else’s tool.

UBI is also unlikely to be seen as a transformative reform from a workerist perspective. Organising for a (higher) UBI, to lessen the impact of low pay and precarious conditions, appears as an alternative to workplace organisation. And, from this perspective, organising to plead for state protection from capital looks like a dead end. Indeed, to the extent that UBI is really a subsidy of low pay and precarious conditions, it is not clear who the state would be protecting from whom.

Freedom from work

The post-work tradition has its roots in the post-operaismo of Italian Marxists such as Carlo Vercellone, Antonio Negri, and other “post-work” thinkers such as André Gorz. Its key contention is that work and value-production increasingly fail to conform to the model we see in Marx, one that is better suited to early industrial times. As production becomes more and more immaterial, less bounded by the conventional workplace, and less defined by physical time or output, labour’s role in production becomes harder to measure and define. In other words, the sphere of value creation becomes diffuse. It comes to include not just domestic and reproductive labour, but all activity that is generative of a functioning society and the knowledge that it possesses.

As a result, much less emphasis is placed on the idea of emancipation within work. Instead, the receipt of a basic income often takes centre stage as the emancipatory moment, releasing us from the compulsion of performing alienated activity for the market, or indeed for the state. Work is not transformed so much as progressively abandoned in favour of UBI-funded, self-directed, productive activity.

Technological advance plays a critical role in this approach. The growth of labour productivity through automation is seen, in the tradition of Gorz, as outpacing our demand for the products of labour. This would appear to make productive full-time work for everyone an impossibility, both as a source of income and as a route to emancipation. A related critique is David Graeber’s view that “bullshit jobs” have become the norm in capitalism—jobs which may have some tactical rational for employers yet cannot be seen as meaningfully creating value for society. For Graeber, such jobs should not be socialised; they should be abolished.

So workerist and post-work traditions actually share a vision of production through free association in a society that has moved beyond wage labour, but they foresee very different paths to get there: the one through the transformation of work, the other through escaping it.

The moral agenda is also different. Paying workers “the full fruits of their labour” is a caricature of the workerist position, but not entirely without basis. By contrast, proponents of UBI see it as a way for society to reward all useful activity performed beyond the market. It is often promoted as a “dividend” of past efforts to develop knowledge, technology and “social capital”.

Lastly, there is the question of the transformative potential of a UBI on capitalism itself. Inasmuch as it creates an alternative to wage labour, a UBI is often seen by left proponents as a tool to undermine the bargaining power of capital over labour, forcing improvements in pay and working conditions, or even challenging capitalism itself. Even the implementation of a modest “partial” UBI is seen as a transformative reform, undermining the ideology of work that makes one’s job a signifier of social purpose and worth, and that stands politically in the way of the transition to a post-work society.

Some sources of confusion

One conclusion we can draw from this controversy is that it is not always helpful to read the labour theory of value (LTV) as a normative theory, despite the undoubted rhetorical power of doing so. With LTV treated as a moral theory of theft, the UBI debate becomes one of competing moral claims to the social product, based upon competing claims about value-creation.

But in what sense should value-creation in capitalism be a moral yardstick when we are talking about transformation to post-capitalist society? Marx famously proposed “to each according to his (sic.) needs” as a better principle; a post-capitalist society has to transcend the logic of reward-for-contribution to realise that vision. LTV, on the other hand, can be seen as a descriptive theory of the determination of prices of production under specific circumstances. This is why Marx only considers wage labour in that theory—unpaid work is not considered to have any effect on prices.

There is also a lot of confusion about sequencing in this debate. Are we talking about before or after the revolution? To attack UBI, it is tempting to counterpose the promise of after-the-revolution socialism to the dystopia of a capitalism shored-up by a lumpenproletariat-pacifying UBI. At the same time, it is tempting for the post-work side to counterpose the self-directed, unpaid activity and free association of hypothetical UBI-receivers to the horrors of actually-existing wage labour.

But neither of these are entirely fair comparisons to make. Both compare capitalist realities with states of affairs that are really only possible post-capitalism. We need to consider what role UBI might play in a post-capitalist/socialist/communist economy. In parallel, we need to consider exactly what kind of UBI is possible in a capitalist economy, and the limits thereof.

UBI would be contingently necessary in a post-capitalist economy

UBI has emerged as a policy response to the problem of persistent labour oversupply and the precarity and poverty that goes with it. This oversupply, however, is a necessary feature of capitalism. In the workerist vision of a socialist economy, the problem is solved by guaranteed access to employment. In the post-work vision, UBI solves the problem that employment can no longer be guaranteed. My aim in this section is to suggest how this chasm might be bridged.

First of all, I believe that full employment—understood in a certain way—should be seen as a basic requirement in a post-capitalist economy. The opportunity to work as much as we would like is not just a nice-to-have; it is needed to underpin the strength of labour and ensure the democratic character of the production process. I am tempted to say that worker power through effective labour scarcity is even more fundamental than the de jure control of the productive apparatus. There is no state enterprise or worker coop that could not potentially become the site of alienation if you had to beg to be allowed to join, i.e. if surplus labour put some people in a position of de facto power over others.

But there is no reason why we should all want to work forty hours a week. In fact, “how much we want to work”, on aggregate, is not a matter of immutable individual preferences but something to be consciously and collectively decided. It should be an object of collective economic management, and this is where UBI may play a role post capitalism.

How much we should want to work, moreover, depends on how much we want to produce. In capitalism, growth is just the accidental product of uncoordinated profit maximization. Associated production, by contrast, implies a collective decision about how big the economy should be at any one time. Even putting aside the issue of ecological limits, the balance between well-being now and in the future, and thus between current consumption and investment, is fundamentally an ethical and political one. To let the current and future scale of output be determined accidentally, through uncoordinated activity, is to abdicate responsibility. A socialist society has to target a certain scale of output at any date, thus aiming for a certain trajectory of growth or de-growth.

A target level of output then implies a target for the input of labour time, given the state of production technology at any one moment. Looking forward, this is unlikely to mean everyone working all week; not in the context of automation, ecological constraints and the satiation of basic needs. So there can be no question of simply trying to stimulate whatever output it takes, Keynesian style, to absorb however much labour we might happen to want to supply. Instead, full employment would require that we adjust the amount of labour we want to supply to the desired level of output.

Now this is not a problem where there is no market. For example, if I agree, in a self-sufficient commune, to a plan which foresees everyone working 20 hours a week, then by definition I decide and want to work the required 20 hours a week. The supply and demand sides are brought together in a conscious decision. The same applies to voluntary activity and any work done for its own sake, rather than for payment.

However, there are few visions of a post-capitalist economy that do not foresee at least a consumer market and payment for labour (in money, “labour tokens”, etc.), at least transitionally. And where work is performed for payment, the problem arises of setting this reward at the right level to incentivize the required supply of labour. Rationing out scarce paid work is possible but suboptimal, leaving nearly everyone dissatisfied.

This is where UBI comes in. Clearly, by raising or lowering a UBI, we could collectively influence aggregate labour supply, whilst leaving each of us free to decide how much we each want to work for payment. In fact, we should speak here not just of UBI but of the general balance between labour and non-labour income. The latter includes goods and services provided in kind, like public healthcare, and money transfers such as pensions, as well as UBI. Here then, UBI becomes an intelligent tool for planning aggregate labour supply in the context of prior defined goals for aggregate production and consumption.

Consequently, instead of seeing “workerist” and “post-work” visions of post-capitalist society as opposing each other, we can see them as contingently necessary solutions to different historical circumstances. We can see UBI as a contingently necessary tool, rather than as something mandated by a particular view of the LTV. In a low-productivity, early industrial context, with a strong desire to increase output to meet basic needs, we face the problem of how to motivate and compensate the performance of a large quantity of arduous work. In this context, there is limited scope, macroeconomically, for extensive non-labour income.

By contrast, with high labour productivity, high living standards and ecological constraints, we face the problem of how to allocate output without giving rise to unquenchable demands for paid employment. As a result, a shift towards non-labour income becomes macroeconomically necessary. Meeting particular needs comes first, in my view, but something like a UBI becomes increasingly desirable, and even hard to avoid. So the post-work vision of UBI in a highly automated communist economy, in which little work needs to be paid for, also makes sense. Vitally, both visions are ones of full employment, in the sense that we each get as much paid work as we want.

But are we really still talking about UBI at this point? The same mechanisms can be discussed with reference to basic service provision or systems of rationing to meet basic needs; making UBI the theme just situates the discussion within debates on automation and the potential outcomes of capitalism’s changing technological base. It is also clear that a post-capitalist “UBI” is functionally completely different from a capitalist one. So does a UBI make sense in a capitalist economy, highly automated or otherwise?

The limits of UBI within capitalism

The question for proponents of UBI as a progressive measure within capitalism is whether it can really be high enough to provide a widespread alternative to wage labour. It seems the answer is “no”.

There are circumstances in which a modest UBI could increase labour market participation, but these are limited. More generally, a rising UBI in a capitalist economy will decrease labour participation and unemployment, with predictable effects. Sooner or later the labour market will grow “tight” and the bargaining power of labour stronger. Inflation is likely to rise as firms try to pass on higher labour costs. Ultimately, without state intervention, the result must be a capital strike: investment will collapse as profit rates fall, regenerating unemployment and returning labour to its former weakness.

This is the point where the state usually intervenes, in the guise of fighting inflation. If the state had just raised the level of UBI, then the obvious inflation-fighting response would be to lower the UBI again, forcing people back onto the jobs market and undercutting wage demands. The possibility of a UBI in a capitalist economy is conditional on it being low enough not to provide a viable alternative to wage labour for most people.

For the same reason, UBI within capitalism is quite likely to mean the redistribution of income from higher- to lower-paid workers, rather than taking capital income and redistributing it to labour. There may be interesting possibilities to tap rental income flows like land rents and speculative capital gains, but the overall scope for sustainably reducing profit rates to pay for a UBI must be limited.

Moreover, a modest UBI that failed to provide a viable alternative to paid employment might actually increase profit rates. This might be the case if a UBI was effectively funded at the expense of higher-paid labour, whilst also facilitating lower pay and more precarious conditions for many workers. In this case, capital would be better off, whereas workers as a whole would be worse off. A UBI could become part of a package to create a “business friendly environment”: business might still be forced to pay some workers well, but some of that excessive pay could be recovered through taxation to subsidize low pay and “flexibility” for others.

What difference does automation make?

In a sense, automation could change all this, but in another sense—and this is where the LTV is crucial—it would make no difference at all. The idea behind the appeal to automation is that labour productivity will become so high that output can simultaneously fund a generous UBI, high wages and plentiful profit. Hence, no one would be compelled to work, enough people would still be willing to sell their time to capitalist firms, and entrepreneurs will still want to invest, all at the same time. The capitalist system of exploitation remains intact, but we all get to live a good life anyway.

There is an extensive debate about how likely this actually is, given recent trends of slowing productivity growth. We must also consider the capacity of capital to find non-utility enhancing ways to absorb surplus and create work through zero-sum competitive games, such as in marketing, or just through the expansion of corporate bureaucracies. However, even assuming the promise of automation is fulfilled, this vision raises certain questions.

Firstly, it is not correct, that UBI will be necessary to create demand for the output of automated production. Competitive consumption by the capitalist class and by states could perform this function perfectly well. We also need to be wary of what are essentially moral appeals to capital, in the form of arguments that say automation will make a UBI necessary in order to preserve some idea of the good (but capitalist) society. This is utopian, unless it can be specified why capital should acquiesce.

More fundamentally, we may ask why this future is not with us already, given that labour productivity has quadrupled in the post war period. The answer is that labour, at least for the first three decades, was powerful enough to maintain its share of output, resulting in rising living standards for workers. After that, increases in productivity were rerouted into supporting the endless accumulation of capital.

The high-tech vision of UBI, the darling of Silicon Valley billionaires, would represent the historic defeat of labour. Working to achieve this vision would mean ensuring that workers’ standards of living do not rise with productivity, in order to make UBI affordable. Indeed, we could also make such a “full” UBI affordable now, hypothetically, just by drastically reducing our standard of living. It would be kind of the same thing.

Can a modest UBI be a transformative reform?

I hope that the preceding sections have shown that UBI under capitalism and UBI in a post-capitalist society can’t be understood as the same thing. Many appealing visions of life with a UBI belong strictly in a post-capitalist world, because they are not compatible with capitalism’s functional requirement for surplus labour and accumulation.

Post-work critiques of the narrowness of workerist understandings of value creation do not have merit, but escape from alienating work is not a plausible option for most under capitalism—the broadly workerist vision of emancipation in and through work therefore remains vital. Moreover, the idea that a UBI could dramatically alter the bargaining position of workers within capitalism, or could even cause the tables to be turned, is logical but not realistic. If we had the power to kill off capitalism with a UBI, then we would not need a UBI to do so. And until that point, any UBI will be strictly limited by the imperative of profitability.

This leaves us with the question of whether a modest UBI, under capitalist conditions, is a transformative reform that moves us forward, or rather one that causes us to regress. One argument from the post-work side is that even a modest UBI would undermine the “work ethic” that provides ideological support for capitalism, whilst preparing us psychologically for the “new technological revolution”. This work ethic is certainly functional for capitalism, but it is less clear why we should see it as a barrier to change, and it is not clear why a UBI as such would undermine it. A campaign for a UBI might contest this ideology of work (and highlight how labour has ceased to benefit from productivity increases), but so too can campaigns for shorter working times and better life-work balance.

Another tactic is to present a UBI not as a welfare transfer but as a “dividend” from a stock of capital that is collectively created and/or held by society (“knowledge capital”, “social capital”, etc.). But this draws upon and reinforces the idea that income is a matter of right derived from property. As equal shareholders in public capital, it is implied that we are entitled to equal dividends. If this is true, however, then surely the owners of private capital stock have a right to their dividends as well?

The mainstream UBI debate, of course, is really mostly about welfare reform. Here, its usefulness integrates claims about efficiency, the dignity afforded to beneficiaries, and public acceptability. For example, John McDonnell has linked the idea of UBI to his experience with child benefit. As a non-means-tested measure, child benefit was once considered a dubious proposal, but has since proved popular among the public.

But child benefit is not like a UBI. Neither are means-tested, but child benefit is need-determined, not universal. The arbiter of need is the (bureaucratically-determined) presence of children, which is recognized as implying a greater household need for income. There seems to be widespread acceptance for means-tested progressive taxation (from each according to their ability!) and for need-determined benefits like child benefit, pensions or disability benefits (to each according to their need!). Acceptance is most fragile where beneficiaries’ inability to look after themselves is popularly called into question (the unemployed, “benefit scroungers”, etc.).

This is a problem for any UBI scheme. As described above, the scope for funding it from the capital share is limited. This means it is likely to be largely a transfer from the better-paid to the less well-paid, and from those who work to those who do not. Unlike child benefit, its acceptability will depend on persuading people that it is good to work and pay taxes so that other people do not have to. This is likely to be a hard sell.

The acceptability of child benefit rests on the perception that people do not have equal needs, making need-determined distribution more efficient more efficient than universal provision. For the same reason, need-determined distribution provides greater security because people who need a bigger cushion can get one, or because help is targeted at occasions of need, such as the loss of work or the advent of illness. Benefits that are means-tested or conditional on behavior can be bureaucratically wasteful and intrusive, but cases like child benefit show that means need be tested only once (at taxation) and that much needs-testing can be reduced to simple objective criteria. We also need to push the idea that sometimes it is better to give claimants the benefit of the doubt, to create a more humane and efficient welfare system.

Finally, we need to be wary of the assertion that the assessment of need by the state is always suspect, and that equality of benefits is in principle good. From this, it is but a small step to just handing out equal vouchers to spend on private healthcare and education. It is but another small step to the idea that all means-testing is bad and tax should always be flat. The libertarian support for UBI is not for nothing, after all.

Marxist debates can help us understand UBI’s alluring vision. Only by transcending capitalism, however, can we ever reach that vision. Until then, it seems destined to be a disappointment: as a welfare system, it is sub-optimal; ideologically, it may be, on balance, unhelpful; and as a potential subsidy of low-paid and precarious employment, it may even become regressive.

More promising transformative reforms lie elsewhere. We can strengthen and deepen democracy. We can raise the labour share of income at the expense of speculative and rentier incomes through instruments such as land value taxation, reducing the political power of capital as a whole. Lastly, we can begin to foster the growth of public, worker-owned and similar alternative forms of ownership at the gradual expense of the purely capitalist sector. While a UBI scheme is likely to become subordinate to the requirements of the capitalist mode of production, such policy options can help build the foundations for socialism.