I argue that, in a sufficiently just society, all adults who are able to work have an obligation to do so on the basis of the value of reciprocity.¹⁰ We can start by thinking of a just society as a “cooperative venture for mutual advantage,” which involves people cooperating to produce the goods and services that constitute the material and social means for people to live flourishing lives.¹¹ A person who benefits from this cooperative activity by consuming goods and services without doing anything to produce them when they are able to do so is taking unfair advantage of the productive members of society, not so unlike the idle rich do now.¹²  

The principle as formulated does not say what amount of productive activity people must perform.¹³ That said, the fairness argument implies that society should strive to make sure productive burdens are shouldered as fairly as possible. To the extent that some people take on substantially greater burdens — because, say, they work jobs that are more dangerous, less esteemed, or more unpleasant, or they work more hours than others — they should receive compensation for doing so.

The duty to work, then, is grounded in obligations of fairness to fellow members of society’s cooperative scheme.¹⁴ Call this the Fairness Argument.

Why is the qualification “in a sufficiently just society” necessary? Some philosophers have argued that victims of significant social injustice do not have an obligation to work. The basic idea is that these people are not enjoying the fruits of social cooperation to a degree that could ground obligations of fairness. Tommie Shelby has argued that, even if something like the Fairness Argument is sound, the black urban poor in the United States do not have obligations to work because they are victims of egregious injustice: vastly inferior economic opportunities, lack of wealth, and racial discrimination in various domains. Michael Cholbi has argued similarly that most US workers today do not receive sufficient net benefit from society’s cooperative scheme to have such duties. And socialists should be skeptical about a general duty to work under capitalism on the grounds that this work is generally exploitative and harmful to workers, subjecting laborers to domination and meager compensation while generating unearned value for employers.¹⁵

These objections bring out that a duty to work obtains only if certain background conditions of justice are satisfied. While those conditions are probably not satisfied in the contemporary United States, our question is one of “ideal theory” in political philosophy: a question about the norms that apply in a society in which people and institutions generally conform to principles of justice. For socialists, that means a society in which the value of political equality is realized, class exploitation has been abolished through collective ownership, and distributive justice is at least roughly achieved.

Some writers, however, have raised in principle objections to the Fairness Argument, even understood as an argument in ideal theory. I discuss two significant ones here. 

One worry is that fair contributions to the cooperative scheme need not take the form of working.¹⁶ Social cooperation involves other burdens, such as respecting the rule of law and private property, that all citizens are supposed to bear; why aren’t those restrictions of liberty sacrifice enough to discharge fairness-based duties?

This line of objection is not compelling. Social cooperation does require everyone to bear a variety of costs, like those involved in complying with the law. Yet labor involves a distinctive burden over and above these: using one’s finite time and exercising one’s productive capacities for the benefit of one’s fellow humans. It is no mystery why fairness demands that all who are able should take on an equal share of this burden — a demand captured, in overly punitive form, in the New Testament dictum that “He who does not work shall not eat,” endorsed even by Vladimir Lenin in revolutionary Russia.¹⁷

A second worry: it is not clear why everyone should have a duty to work when everyone’s labor may not be required to produce sufficient goods and services.¹⁸ Perhaps voluntary labor contributions can ensure sufficient production, without our needing to appeal to a universal duty to work. Yet even if it were true that sufficient labor would be performed in the absence of this obligation, a situation in which some work and others do not is manifestly unfair. Ensuring a fair distribution of this burden is the point of the Fairness Argument.¹⁹

To say that people have a duty to work is not yet to say whether and how we might enforce that duty. There are three possible mechanisms.

First, the duty to work may be imposed through what we might call interpersonal moral sanctions. That is, people might feel and express resentment, indignation, or contempt toward fellow citizens who shirk the duty to work and take advantage of others; we might resort to scolding, ostracizing, or otherwise socially penalizing able-bodied persons who simply refuse to work while praising and expressing esteem for those who do contribute.²⁰ This sort of minimal “enforcement” can help inculcate an internal sense of duty to do one’s fair share.²¹

Second, the duty might be imposed through direct, coercive legal requirements, as the “parasite laws” established in the Soviet Union and other state socialist regimes in the twentieth century did. Such laws might punish nonworkers with financial penalties or jail time.²²

Third, we can impose labor obligations through what Marx calls “the silent compulsion of economic relations”: encouraging people to work by making some or all of their income dependent on working.²³ This would be the case if society makes no welfare provision for people outside the market at all. Or working might be made a condition of receiving welfare benefits, as is now standard in the United States. Finally, a society might offer some minimum level of unconditional welfare but make additional income above that minimum dependent on working.²⁴

The second key principle of productive justice I argue for is that society should respect every worker’s right to freedom of occupational choice. People should be able to pursue and apply for any jobs they would like; they should not be forced to take certain jobs by the state, nor barred from certain jobs or occupations by law or by arbitrary discrimination (due to, e.g., sex or race) by employers.

My argument here takes inspiration from John Rawls. For Rawls, freedom of occupational choice belongs to a set of “basic liberties” that includes civil liberties like freedom of speech, freedom of religion, and freedom of movement as well as fundamental political rights. Protecting these liberties is of paramount importance for social justice because they are essential to safeguarding individuals’ status as equal citizens and allowing them to live autonomous lives. For that reason, they cannot be restricted or sacrificed merely for the sake of improvements in welfare or to promote greater economic equality. Freedom of occupational choice, like other basic liberties, takes priority over these aims.²⁵

The basis of Rawls’s position, which socialists ought to endorse, is that the value of autonomous agency is of overriding importance.²⁶ He sees autonomy as centrally involving a capacity to form and revise one’s own “conception of the good”: an understanding of what is fundamentally important or valuable to one and a “plan of life” that reflects those values. The civil liberties, like freedom of speech and freedom of conscience, as well as freedom of the person — including freedom of occupational choice — are essential to giving people the reflective space to formulate their own life plans.²⁷ Without knowing that I am free to choose my own occupation, my ability to reflect on what matters most to me and how to arrange my life so as to pursue it is severely compromised. Loss of freedom in how I exercise my productive capacity does not merely constrain my practical options for pursuing a plan of life; it also forecloses the possibility of even formulating my own plan of life, since this sort of deliberation has been deprived of much of its point.

The right to freedom of choice of occupation (like other civil liberties and freedoms of the person) thus protects individuals against deformation of our autonomous agency. It is for this reason that it and the other basic liberties take priority over promoting overall welfare, equality of opportunity for jobs, and equality in other social and economic resources.²⁸ This view may seem implausibly extreme: Shouldn’t people suffering from severe material deprivation be willing to sacrifice freedom of occupational choice to ensure everyone has a roof over their heads and enough to eat? 

Here we need to introduce a crucial caveat. For Rawls, the priority of the basic liberties is established only when society is at a high enough level of social and economic development to ensure people can effectively exercise those liberties. Robert S. Taylor, elaborating Rawls’s argument, claims that this involves society meeting two thresholds: “a level of wealth sufficient for it to maintain a legal system with courts, police, etc., that can define and protect the basic liberties of citizens within the bounds of the rule of law,” and “a level of wealth sufficient for it to allow its citizens to engage in the meaningful formation of life plans. For example, citizens must have access to media, public forums, and schools and must have sufficient leisure time to make use of these resources and reflect on their plans.”²⁹

Once these material conditions are in place, citizens’ autonomous agency, as protected by the basic liberties, cannot be compromised for the sake of greater welfare or expanded economic opportunities. A lack of material resources, of course, might limit an individual’s options: I may be unable to afford to commute to my preferred job, for instance.³⁰ This may represent an injustice that society should seek to address, but it is not an injustice of the same order as an interference with one’s free choice of occupation. For while lack of money might interfere with my ability to take a certain job, the lack of even a formal right to choose my own occupation interferes with my autonomy at a deeper level: it constrains my ability to rationally deliberate about certain options in formulating my conception of the good. It is for that reason that the freedom of occupational choice, like other basic liberties, takes priority over greater economic welfare.

Some socialists might be skeptical of the emphasis this argument places on individual liberty or its Rawlsian lineage. But the value of autonomous agency is central to the socialist ideal, and to Marx’s ethical vision in particular. Marx looked forward to an ideal society in which “the free development of each is the condition for the free development of all.”³¹ Such a society would, crucially, allow people to consciously direct their creative energies toward their own freely chosen ends.³²

The Marx of The German Ideology seems to associate this overcoming of alienation with the elimination of the division of labor. There he writes, “In communist society . . .  [it will be] possible for me to do one thing today and another tomorrow, to hunt in the morning, fish in the afternoon, rear cattle in the evening, criticise after dinner, just as I have a mind, without ever becoming hunter, fisherman, shepherd or critic.”³³ Though it seems doubtful we could construct a desirable society that does away with the division of labor like this, it’s striking that the vision of human flourishing gestured at here emphasizes pursuing activities “just as I have a mind” — deciding for myself how I want to deploy my productive energies. We should hold on to this kind of autonomy as an ideal, one that freedom of occupational choice protects.

The simplest answer to the labor question is to deny socialist societies face any such problem. This is essentially what postwork socialists argue: they claim socialist societies can and should do away with work. Instead of living off wages, postwork theorists argue, everyone should receive a generous universal basic income (UBI).³⁴ Any work that people chose to do would then be fully voluntary. In fact, postwork socialists tend to denigrate the idea that there is any kind of necessity or duty to work as a piece of false consciousness.

Alex Gourevitch has convincingly criticized the postwork perspective.³⁵ The core of his criticism is that, by proposing everyone live off a generous UBI, postwork socialists tacitly presuppose that work is being done. After all, someone must be making goods and services for people to buy with their basic incomes. Yet despite presupposing the continued existence of work, the postwork view says nothing about why or how people will be motivated to do this work. Gourevitch also makes a persuasive case that the problem cannot be solved by appealing to automation.³⁶

Still, suppose that some people feel called to carry out nonautomated work by nonmoral but still nonmaterial motives — a love for particular jobs or activities, say.³⁷ It seems doubtful such motives would be sufficiently strong and widespread to produce the collective wealth needed to sustain a generous UBI. But even if they were, the resulting society would involve significant injustice, for the burden of work would simply fall on whoever felt called to do it, while everyone else would enjoy the fruits of their labor.

Gourevitch rightly rejects postwork socialism, advocating instead “a democratically planned economy supported by a socialist work ethic.”³⁸ He doesn’t lay out what this planning would look like but is clear that, in such a society, people would collectively decide what work needs to be done and then be motivated to do that work by a sense of duty. Moreover, he suggests that this sense of duty could be sufficient to motivate people to do the needed work:

In a society where everybody is ready to do their share of the necessary labor, nobody is forced to work. That is because, if someone were to stop doing what has to be done, others would step in and do it, meaning all the necessary work would still get done. . . .  Sharing labor is the path to avoiding having to force people to work and securing free time for everyone.³⁹

This view is consonant with my claims that people have a moral duty to work and the idea that it may be enforced, in a minimal sense, through interpersonal moral sanctions. But Gourevitch does not say anything about whether or how people might otherwise be compelled to work and indeed suggests that stronger compulsion may be unnecessary. If he truly means to say this duty would be a sufficient source of motivation to work under socialism, then his account is plainly inadequate.

While we shouldn’t underestimate the importance of moral commitments as a motivation, it is unlikely that this motive (and other “internal motivation[s]”) will be sufficiently powerful to get workers across an entire national (or global) economy to perform all of society’s necessary labor — as opposed to the work of, say, a family business or a small religious commune. People will find much of that necessary work unpleasant or worse, and the social and emotional ties and sense of moral obligation connecting people to their fellow citizens are unlikely to be strong enough to motivate people to make the significant personal sacrifice of taking the required jobs, when they could instead pursue unnecessary work of their own choosing or simply refuse to work at all. The historical experience of attempts at socialism offers evidence for this claim: although periods of revolutionary upheaval often see workers voluntarily making great sacrifices for the sake of moral ideals, this sort of motivation has not been sustainable over longer periods.⁴⁰

It is of course possible that people’s psychology will be so different in a future socialist system that moral motives will be sufficient to mobilize all socially required labor. But this possibility seems like a thin reed on which to base a fundamental transformation of society. Our vision of socialism should not require that human beings come to behave in ways radically at odds with our actual experience in the here and now.⁴¹

Attempting to implement the model is also bound to produce significant injustice. Recall Gourevitch’s assertion that “if someone were to stop doing what has to be done, others would step in and do it, meaning all the necessary work would still get done.” But this proposal manifestly generates productive injustice. It means that nonworkers will be able to free ride off the productive efforts of those who work, effectively taking advantage of workers’ greater sense of duty or degree of conscientiousness. It threatens to create, as Gourevitch himself puts when criticizing the postwork view, “an arbitrary class of idlers next to an arbitrary class of workers” — where the former group in effect exploits the latter.

Now, it may be compatible with Gourevitch’s account that a socialist society use force or other material incentives to get people to work if there is a sufficiently widespread breakdown of voluntary compliance. My point is that our conception of a feasible socialism should not assume generalized voluntary compliance solely on the basis of internal motivations and interpersonal moral sanctions. We need to explain how we might supplement the sense of duty with other motivations.

Participatory economics, developed in a series of works by Robin Hahnel and Michael Albert (both jointly and singly), is one of the most prominent attempts in recent decades to formulate a democratic, nonmarket alternative to capitalism. The core idea behind this model is that decisions about what to produce and consume are made collectively through an iterative process of deliberation within and between democratic bodies of various kinds — including worker councils, neighborhood-level consumer councils, and higher-level federations of worker councils and consumer councils. A complete assessment of “parecon,” as it is sometimes called, is beyond the scope of this paper. My focus is on how jobs are assigned and compensated. 

At the beginning of each year, an annual production plan is determined by comparing collective consumption proposals from neighborhood consumer councils and production proposals submitted by worker councils (i.e., democratically worker-
managed firms); these proposals then go through a process of iterative mutual adjustment until consumption and production plans match. The final plan authorizes each worker council to “to employ a certain number of members with particular skills”; this might mean they need to lay off current workers or “hire” new ones.⁴² The planning process is meant to ensure full employment. People may apply to whatever worker councils they want, and worker councils are free to make hiring decisions on whatever basis they see fit.⁴³

Crucially, wage offers play no role in this hiring process. That is because an individual worker’s compensation is supposed to be based on the effort or sacrifice they made at work, as assessed by their coworkers at the end of the production year. Note that this is merely a recommendation; Hahnel and Albert leave it up to each worker council to democratically decide how to allocate compensation.⁴⁴

That said, participatory economics does involve material incentives to work rather than not work. Though the model advocates for taking care of the basic needs of the nonworking population through welfare provision (as well as providing certain goods like health care and education to all free of charge), those who work are provided additional “credits” for their households’ private consumption needs, in proportion to the effort ratings they are given by their coworkers. To prevent workers at a given workplace from colluding to artificially inflate their effort ratings, Hahnel and Albert suggest capping the average of the effort ratings workplaces may assign. 

Consider, first, the idea of a labor levy: a legal requirement that all citizens do jobs of a particular sort for a specified amount of time. Kandiyali has recently defended such a policy.⁵⁵ His argument starts from the observation that many socially necessary jobs impose great burdens on workers while providing them few corresponding benefits — they involve drudgery, or significant health or safety risks, and don’t command much in the way of pay or social esteem; a socialist state might reasonably demand that everyone spend, say, a year or two performing a rotation in one of these jobs. Perhaps everyone should spend a year working as a garbage collector, or a janitor, or a certified nursing assistant, for instance. 

Kandiyali argues that such a policy is justified in order to make sure this sort of work is fairly distributed across the population.⁵⁶ But such a labor levy might also be justified by the aim of inculcating an ethos of public service or civic-mindedness, instilling a sense that “we’re all in this together.”

Policies of this sort are a means of enforcing the duty to work, and they can also be designed to be compatible with the right to freedom of occupational choice. The policies should presumably only require people to work in specific jobs for a limited time, at specified points in life — for two years when they reach the age of majority, for instance, or for a few months every five years over the course of the person’s working life — and people should be given adequate notice of these work requirements. If those conditions are met, however, the policies would not meaningfully preclude people from choosing their own occupations and developing their own plans of life in line with their values.⁵⁷

Stanczyk offers a different argument for compulsory work requirements.⁵⁸ He claims that the demands of distributive justice sometimes justify the state in imposing compulsory labor schemes. I have assumed, for instance, that everyone is entitled at least to a social minimum. Stanczyk points out that there are situations in which, absent the use of coercive requirements, it will be either a) impossible to get people to produce the goods and services needed to provide the social minimum, or b) possible to entice the necessary labor only with pay incentives that are morally impermissible because they would create objectionable economic inequalities or draw resources away from more urgent needs.

An example of type a): Many people in poor countries do not have access to adequate medical care because not enough medical professionals are willing to work in those countries — most emigrate so that they can earn higher salaries abroad. In some of these countries, the government already offers doctors relatively generous salaries, yet there are still severe labor shortages.

An example of type b): There are also severe shortages of medical personnel in some regions in richer countries. Doctors in these areas make salaries that put them in the top 5 percent of income earners in the country, yet that compensation has proved insufficient to incentivize enough of them to live and work in these regions. It is possible that, if medical professionals were offered even higher pay to work in underserved areas, the labor shortages could be solved. Yet, Stanczyk argues, paying doctors the amount required to entice enough of them to rural work might itself involve creating unjust economic inequalities.

So, he concludes, in cases where it is impossible for the state to meet its obligations to provide certain goods and services to people with labor-market incentives alone, or where the effective use of such incentives would involve impermissible inequalities, the state may justly force people to work certain jobs. Ghana, for instance, forces its nurses to work in the country for five years before they are allowed to practice abroad, withholding their nursing licenses if they refuse to do so. The United States uses “contractual-bonding arrangements,” offering public subsidies to nurses, doctors, and dentists to cover the cost of their education on the condition that they work for a period in an underserved region. 

Some such policies violate freedom of occupational choice, while others do not. Contractual-bonding arrangements, which merely require professionals to do certain kinds of work in certain locations in exchange for the government funding their education, do not interfere with occupational choice at all. But Ghana’s policy of withholding nurses’ licenses unless they work in the country for an extended period — in a context where most of the other jobs nurses could get pay poverty wages — does interfere with this freedom. Stanczyk argues that the Ghanaian state might have to rely on a policy that abridges occupational choice in this way in order to secure adequate provision of health care to its population, and that such a policy is therefore justified.

But this is not really a good argument for denying that people have a right to choose their own occupations, nor for thinking that this right may be overridden for the sake of distributive justice. I have argued that basic liberties like freedom of occupational choice take priority over other moral considerations only in societies that have reached a high enough level of development to ensure the effective exercise of those liberties. Adequate health care looks very much like a precondition for that: without assurance that I can be treated for ordinary injuries or illnesses (which might nonetheless be debilitating or life-threatening if left untreated), my ability to meaningfully formulate a life plan is seriously disrupted. So if Ghana truly cannot provide its population with adequate medical care without compelling nurses to work in the country, this would be a scenario where the developmental threshold for establishing the priority of liberties like freedom of occupational choice has not been reached. In that case, a society is justified in abridging such freedoms if doing so “is necessary to change the quality of civilization so that in due course everyone can enjoy these freedoms.”⁵⁹

If, on the other hand, an adequate health care system (and the other prerequisites to the effective exercise of the basic liberties) has been established, there is not a compelling case for interfering with people’s freedom of choice simply to improve health care provision. If we are talking about contemporary Sweden, for example, where generous health care benefits are provided to the whole population, the state would not be justified in imposing Ghana-style restrictions on its nurses or doctors just to, say, reduce wait times for doctor visits.

The focus of my argument, however, is on socialism in rich countries, where we can assume that social and economic development has progressed to the point of enabling the effective exercise of basic liberties. In this context, it is unjust to restrict individuals’ freedom of occupational choice for the sake of greater economic welfare.

In sum, there is a limited role for compulsory service requirements of the kind advocated by Kandiyali, and some of those (like contractual bonding) advanced by Stanczyk. But more extensive regimes of forced labor are ruled out when, and because, they violate the fundamental right to free choice of occupation.

The advocate of labor markets, unlike the postwork theorist, can acknowledge the necessity for people to work under socialism as well as a duty to work. And, unlike both the postwork view and Gourevitch’s shared-labor model, labor markets offer a way of enforcing that duty. They do so by creating a compulsion to work, through opening a gap between the social minimum that society guarantees to all and the income made possible by employment.

To make this vivid, we need to say something about what this minimum would look like. Plausibly, it would take the form of some combination of decommodified services (e.g., free health care and public transportation) and cash income. The more difficult question is how generous the social minimum should be. I have assumed that a just socialist society would give everyone broadly equal access to the necessary means for a flourishing life. This does not entail that everyone must be guaranteed the means to flourish as a matter of right, whether or not they choose to work. But some minimum seems owed to make sure everyone can lead lives of basic dignity, to prevent the emergence of noxious status hierarchies, and to ensure equality of opportunity.

What exactly should count as a dignified social minimum, I think, is a matter for democratic deliberation to decide. But any such deliberation must confront the following dilemma. If the minimum is too generous, such that the material gains most workers expect from working look meager relative to the life they can enjoy living on the social minimum, then people may not be sufficiently motivated to work. That would in turn mean labor shortages or otherwise unsustainable upward pressure on wages, potentially leading to higher prices and lower investment. If the welfare state is too generous in that way, it would undermine the labor market’s ability to solve the labor question. 

On the other hand, it is possible to err in the opposite direction by making the social minimum too stingy: failing to provide the nonworking population with sufficient means for even a modest dignified existence, or encouraging undue stigmatization of nonworkers, or undermining equality of opportunity for those who are out of the labor market for a long time (by, e.g., denying them adequate health care or access to intellectually and culturally enriching experiences).

Therefore, a socialist society ought to strive for a welfare state that achieves a happy medium of providing a dignified minimum to nonworkers without significantly decreasing the motivation to work. Where that happy medium lies is an empirical question, one that would likely have to be found through trial and error.

But there is empirical reason for optimism that a generous social minimum need not have deleterious labor market effects. Lane Kenworthy has found that countries with the highest levels of welfare, the Nordic nations, have among the greatest employment-population ratios and are no worse than other developed economies when it comes to innovation and growth.⁶⁰ In fact, more generous Scandinavian welfare systems have arguably increased citizens’ commitment to working.⁶¹ So it is plausible to suppose we can make use of the motivational benefits of labor markets — compelling compliance with the duty to work — while providing a dignified minimum.

Unlike the postwork and shared-labor views, parecon makes use of the gap between the social minimum and work-based income as a material incentive to work. But labor markets can also make use of supply-and-demand-based pay differentials across jobs and firms to recruit workers to where they are most needed. Hahnel and Albert’s system has no comparable mechanism for directing workers with rare skills or talents to roles where those are most needed, nor for encouraging workers to take more burdensome jobs; the effort-based rating system is totally unable to fulfill this function. 

On this score, labor markets are superior to participatory economics not only on practical grounds but on grounds of justice. When firms offer higher wages to attract workers to less desirable jobs — or even to attract individual workers with important skills to jobs they find less agreeable than their other options — they are in effect offering to pay workers more for taking on greater burdens. This advances the ideal of reciprocity underlying the Fairness Argument, by granting greater compensation to workers who make greater sacrifices.⁶²

The other key virtue of labor markets is that they respect workers’ right to freedom of occupational choice. They share this feature with the postwork, shared-labor, and participatory economics models. But labor markets do this while offering a feasible and otherwise just solution to the problem of allocating and mobilizing labor.

A socialist society that eschewed labor markets while recognizing the need to compel people to work would have to resort to heavy-handed coercive measures on a large scale, even more severe than the kind advocated by Stanczyk. Insofar as these measures involved forcing people to work in some jobs rather than others for extended periods under threat of punishment, that would mean extensive violations of people’s freedom of choice of occupation.⁶³

I have explained why a socialist society should regard that as a serious injustice, an affront to the value of autonomy that both liberal and Marxist traditions of thought hold dear. But it is also worth observing that widespread infringement of freedom of occupational choice threatens to be practically counterproductive for a socialist society as well. If society’s demands for individual sacrifices of liberty become too great, people are likely to see their own interests as being at odds with the collective interests articulated by the state, and this may well provoke anti-state backlash and bolster support for anti-egalitarian forms of economic liberalization (e.g., restoring private property rights).

Socialists have a bevy of long-standing objections to existing labor markets. To wit: they often fail to provide workers with the means to live a dignified existence; they are exploitative; they give rise to unacceptable inequalities of income and wealth; they lead workers to accept domination and a variety of physical and psychic harms on the job; and they themselves are affronts to freedom, in that they force people to work on pain of destitution.

These are fair objections to labor markets of the sort we see in most, if not all, capitalist societies today. But they are not intrinsic problems with labor markets. A socialist society can realistically use labor markets to enforce the duty to work without falling afoul of these objections. Explaining how socialist labor markets avoid those traditional problems will allow me to offer a sharper illustration of the institutional framework I am proposing.

Consider first worries having to do, broadly, with distributive justice: the objections that labor markets fail to provide workers with a sufficient income and that they give rise to unacceptable wealth and income inequalities. On the question of sufficiency, I have assumed that all workers are entitled to a social minimum and have explained how this can be made compatible with dynamic labor markets. On the question of inequality, much rests on what we think the ideal of equality requires. Again, I have assumed an ideal of broadly equal access to the means of flourishing, which does not require strict equality in income. What it requires, I believe, is at least a universal dignified social minimum and equal opportunity to achieve a greater income above the minimum. We might think distributive justice requires more than this, though. Rawls famously argued for the difference principle, which holds that economic inequalities are permissible only so far as they maximize the prospects of society’s least well-off.⁶⁴ The difference principle is controversial; we might accept a different principle of distributive justice. My point here is to show how labor markets are compatible with even a very demanding standard of economic equality. 

Economic equality could be achieved through social or collective ownership of the means of production; a generous welfare state to provide the social minimum and supplement the income of the least advantaged workers and the nonworking population; active labor market policies to help workers find employment and move between workplaces; and robust public education, health care, housing, and transportation sectors. Industrial policy and a public banking system could direct investment in line with social priorities, including important values like environmental sustainability.

Accepting the existence of pay differentials does not mean allowing unchecked wage inequality. Workplace democracy, collective bargaining regimes, and minimum wage standards can prevent wage inequalities from rising to objectionable levels.⁶⁵ What sort of wage inequalities are acceptable will be a matter of controversy and debate, and the aim of using pay differentials to incentivize the most productive uses of people’s abilities would have to be balanced against other values like equality and solidarity.

We can now see why the other traditional objections to capitalist labor markets need not apply to socialist labor markets. If firms are collectively owned and profits socialized, then participation in the labor market will not involve workers being exploited by capitalists. Workplace democracy, collective bargaining arrangements, and the “exit options” provided by a generous social minimum and active labor market policies protect against the prospect of people being exploited by their coworkers (or by state managers in the public sector) or tolerating domination or harmful working conditions in the workplace.

Finally, there is the question of freedom. Socialists have long objected to capitalist labor markets on the grounds that they force workers to sell their labor power to capitalists to survive. The framework I am advocating, however, is not vulnerable to this charge. The welfare minimum guarantees that people can survive even if they don’t work; those who do work are not selling their labor power to capitalists but instead joining worker-managed and collectively owned firms or public enterprises and sharing fairly equally in the wealth they produce.

Yet I am arguing that, even under socialism, there is a sense in which people should be forced to work. Labor markets make sure needed work gets done through a kind of structural compulsion.⁶⁶ But this is not in itself an objectionable attack on freedom. There will always be work that needs to be done; people have a duty to share in this collective burden; and we need a means of effectively enforcing it. Perhaps, as Marx thought, a more enlightened society would eventually transcend the need for such material incentives. But in the meantime, labor markets respect these facts better than the alternatives.