Many theorists believe that a just society should be committed to freedom of occupational choice. Moreover, most theorists also believe that that commitment entails both (a) that states cannot use their coercive powers to force people into working in particular professions, and (b) that individuals are not under an obligation of justice to choose particular professions. However, the vast majority are also committed to the view that justice requires, at the very least, that people have access to the goods and services needed to meet their basic needs. But what if not enough of those goods and services are produced when individuals choose their occupations freely?

Following Lucas Stanczyk, let’s call this the “problem of productive justice.”1 The problem of productive justice will be a difficult problem for any comprehensive political theory to solve. But Stanczyk argues that liberal egalitarianism is particularly ill-placed to address it, since the central commitments of liberal egalitarianism appear to give rise to an incoherent position. On the one hand, liberal egalitarians’ strong commitments to (at least) a social minimum mean that they must say that the failure to provide these goods and services is a grave injustice. On the other hand, they must say that no injustice has been done in failing to provide these goods and services, since their equally strong commitment to freedom of occupation means that they must take individuals to be both morally and legally entitled to choose not to provide them.

The purpose of this paper is to rescue liberal egalitarianism from this charge. One proposal might be to argue that liberal egalitarianism can and should reject (b), that individuals are not under an obligation of justice to volunteer for particular professions.2 In this paper I put that route aside. Instead, I argue that liberal egalitarianism can and should reject (a), that states cannot use their coercive powers to force people into working in particular professions. I will argue that, when properly understood, Rawls’s theory of justice (and “high liberal” approaches to justice more broadly) can allow for certain, moderate kinds of mandatory civilian service, and consequently that it contains the resources for an attractive solution to the problem of productive justice.3

My argument rests on the high liberal interpretation of the priority of liberty that Rawls describes most clearly in Lecture VIII of Political Liberalism.4 In that discussion, Rawls makes clear that once a list of basic liberties has been settled upon, a series of difficult open questions remain about how those basic liberties are to be specified—that is, questions regarding precisely what is and what is not incompatible with those basic liberties. On the high liberal approach, that process of specification should take place in the light of an underlying conception of citizens as free and equal persons. Significantly, this means that claims about the implications of high liberal commitments to certain basic liberties are importantly incomplete without further specification of those liberties in the light of that conception.

In this paper I argue that when it is specified in this high liberal way, freedom of occupation can be seen to be compatible with some forms of moderate mandatory service. On the Rawlsian view, the basic right to personal property should be specified such that it is consistent with significant restrictions on rights of acquisition and bequest, and the basic liberty of free speech should be specified such that it is consistent with significant restrictions on the right to advertise. Similarly, I argue that for Rawlsians, freedom of occupational choice should be specified such as to allow for moderate mandatory service. Moreover, I argue that this general framework offers a useful and plausible way of understanding the demands of productive justice in general. Pace Stanczyk, therefore, Rawlsianism and liberal egalitarianism are far from nonstarters when it comes to productive justice. On the contrary, they are a very promising place to begin in looking for guidance in thinking through difficult questions pertaining to how justice applies to the realm of production.

The problem of productive justice can be expressed by the following three jointly inconsistent claims:

As Stanczyk argues, it can’t be that each of these three statements is true, since it is possible that people cannot be enticed or persuaded to do whatever work is necessary to supply the goods that justice requires us to provide (beyond ensuring liberties). Suppose, for example, that (i) justice requires the provision of basic healthcare to all members of the relevant society. But suppose further that an insufficient number of people can be either persuaded or incentivised to enter the medical profession and work in areas where their services are needed to provide that basic healthcare. If the failure to provide basic healthcare is genuinely unjust, then either individuals act unjustly when they decide not to provide the needed service (in which case (1) is false) or the state acts unjustly in failing to force people to provide the service through some sort of conscription scheme (in which case (2) is false). But this result generalizes, since we can always imagine a plausible alternative version of (i) for some good or other, just so long as (3) is true. So (1), (2), and (3) are jointly inconsistent. In Stanczyk’s words, “If occupational choices cannot be unjust, and if justice forbids forcibly assigning welfare-improving jobs, then there can be no injustice in any failure to bring about any state of affairs in which people fare well, whenever able people will not do the necessary work without being forced.”6

Of course, forcibly assigning jobs is not the only thing a state can do in response to shortfalls in the production of basic goods. Incentivising people to enter areas that are affected by shortages and seeking more creative ways to provide services will certainly go some—and perhaps even all—of the distance towards addressing such shortfalls. Just how far is a complex empirical question, and unfortunately reliable, relevant evidence is thin on the ground.7 Nevertheless, some states at least seem to think that policies that appear to involve forcibly assigning jobs are necessary, or at least justified—states such as Ghana, for example, where all graduates of tertiary education are required to spend a year in civilian service.8 And, though the issue is certainly controversial, at least some political philosophers have agreed that in certain circumstances (from “brain drain” scenarios to labour shortages in rural areas in relatively well-off nations) compulsory service requirements are likely enough to help reduce shortages that they can be justified on those grounds.9 Moreover, as Stanczyk notes, even where incentive policies could address shortages, it may be the case that such policies can only do so on pain of threatening other important goals of a just state. Using incentives to persuade already very well-remunerated medical professionals to relocate to deprived areas, for example, will tend to involve introducing considerable material inequalities into those communities, and so introduce serious injustices that we should wish to avoid.10

But, irrespective of where the empirical chips fall, the problem still presents a powerful theoretical challenge for liberal egalitarianism, and for Rawls’s theory of justice in particular. For, whether or not we are currently in circumstances in which forcibly assigning jobs would be necessary for providing for the basic needs of all, such circumstances are clearly possible. If liberal egalitarianism issues implausible or inconsistent judgments in those cases, that’s a significant philosophical problem, whether or not it also represents a practical one. Moreover, it’s a philosophical problem with important broader ramifications for our assessment of Rawlsianism and liberal political morality. The worry that liberalism overvalues negative liberty at the expense of properly valuing access to the material resources required to meet basic needs is an old one, and versions of that criticism have been and continue to be levelled at modern iterations of liberal political theory (not least Rawlsianism), sometimes with explicit reference to the problem of productive justice.11

In the case of Rawls in particular, the problem is especially powerful because he seems to unambiguously endorse each of (1), (2), and (3). On (1), he is clear that on his theory “what kind of work people do, and how hard they do it, is up to them to decide in light of the various incentives society offers.”12 On (2), he is equally clear that “the priority of liberty means that we cannot be forced to engage in work that is highly productive in terms of material goods.”13 And on (3), he is equally firm that his theory requires the provision of a guaranteed social minimum—indeed, such a guarantee should be included as a constitutional essential in any just society.14

Even more troublingly, each of those endorsements seems to follow from foundational commitments of Rawls’s theory, and so for all liberal egalitarian theories that share those basic commitments. Rawls’s commitment to (1) follows from his commitment to restricting the scope of justice to the basic structure, which can in turn be seen as grounded in basic commitments to, e.g., individual freedom and publicity.15 The commitment to (2) follows from Rawls’s commitment to the priority of liberty.16 Since protecting the basic liberties (and not their worth) is lexically prior to other goals of social justice (such as fair equality of opportunity and the difference principle), it seems that restricting people’s freedom of occupation by forcing them to work in ways that will improve the conditions of their fellow citizens cannot be justified by reference to those improvements.17 And the commitment to (3) follows, as Rawls himself argues, from the underlying conception of citizens as free and equal that undergirds and guides the entirety of his theory. A robust guarantee of an adequate social minimum “is what is required to give due weight to the idea of society as a fair system of cooperation between free and equal citizens, and not to regard it, in practice if not in speech, as so much rhetoric.”18

The problem of productive justice thus seems to place both Rawlsianism and liberal egalitarianism in an extremely uncomfortable position. It’s not just that Rawls and other liberal egalitarians appear to be committed to three incompatible claims. It’s that those commitments seem to follow from some of the deepest and most foundational aspects of their approach to political morality. If this is right, then the problem of productive justice seems to present a profound challenge to liberalism, threatening the coherence of its very foundations.19

In the last section we set out the problem of productive justice, and we saw how it represents a deep challenge to Rawlsianism. In this section I set out the groundwork for a Rawlsian response to this challenge. The heart of that response is that when Rawls’s commitment to freedom of occupation as a basic liberty is properly understood in the high liberal terms that Rawls intends it, it can be seen to be compatible with moderate mandatory service schemes. It is therefore false that Rawlsianism (and liberal egalitarianism more broadly) cannot allow that it is just to forcibly assign jobs when doing so is necessary to secure a basic minimum for others. The next section develops this response with respect to the specific case of freedom of occupation. This section explains Rawls’s distinctive high liberal approach to the basic liberties in the general case.

A useful way to begin with this explanation is to note that identifying a particular freedom as a basic liberty only goes so far in determining one’s views about the implications of that identification for particular policy debates. That’s because, in a range of cases, it’s simply not clear just what counts as an infringement of a particular freedom, nor what should be done in cases where different basic liberties seem to clash with one another. Suppose, for example, that I think that freedom of speech is a basic liberty. What that means with respect to my views about, for example, the right to advertise is difficult to say without a more detailed specification of just what freedom of speech requires, and thus of just what counts as an infringement of it.20

As Rawls is keen to emphasise, specific judgements will rely in various ways on the particular social circumstances of individual societies.21 But it is possible to give general guidelines for how such difficult questions of specification should be approached in a just society, and Rawls tries to do just that.22 Key to his discussion is the notion that each basic liberty has a “central range of application.”23 The central range of application of a basic liberty is identified by reference to the liberty’s role in securing the conditions for the “adequate development and full exercise of the two moral powers of citizens as free and equal persons.”24 As Rawls also puts it elsewhere, different particular liberties can be seen to have different degrees of “significance” where “a liberty is more or less significant depending on whether it is more or less essentially involved in, or is a more or less necessary institutional means to protect, the full and informed and effective exercise of the moral powers in one (or both) of the two fundamental cases.”25 The task of specifying the basic liberties can be thought of as the task of identifying the central range of application of each basic liberty—the range within which those liberties have the most “significance”—and then ensuring that the overall package of basic liberties is best designed to protect those central ranges of application.26 In another Rawlsian formulation, while infringements of basic liberties that affect the central range of application restrict liberty, infringements that are intended, “so far as possible, to preserve intact the central range of application of each basic liberty” do not restrict liberty, but rather regulate it.27 And while there is “a general presumption against imposing legal and other restrictions on conduct without sufficient reason” outside of the central range, regulating a liberty in the interests of preserving its central range (or the central range of a different liberty) provides just such a reason.28 In this way, an overall package of basic liberties is to be specified in detail under the guidance of the overarching thought that “these liberties and their priority are to guarantee equally for all citizens the social conditions essential for the adequate development and the full and informed exercise of [the two] moral powers in the two fundamental cases.”29

Rawls illustrates this general approach with two instructive examples—the right to personal property and the right to free speech. In the first case, Rawls is clear that a right to personal property would be protected as a constitutional guarantee in a just society. But in specifying the content of this constitutional guarantee in further detail, we should be guided by the thought that the right has a central range of application, determined by reference to its role in providing the conditions for the development and exercise of the moral powers. That role, in the case of the right to personal property, is “allowing a sufficient material basis for a sense of personal independence and self-respect.”30 As such, the central range of application of the right to personal property does not include wide-ranging rights of acquisition and bequest, nor a right to own means of production and natural resources.31 Though in a broad sense such rights could be taken to fall under the right to personal property, given that they are not required to allow a sufficient material basis for a sense of personal independence and self-respect, they fall outside of the liberty’s central range. As such, broad rights of acquisition and bequest, and rights to own means of production and natural resources, are not protected by lexical priority. So long as there is sufficient reason to do so, limiting those broader freedoms is quite compatible with respecting both liberty in general and the basic liberties in particular.

In the second case, Rawls is again clear that while freedom of speech is protected as a constitutional essential in a just society, the weight of particular claims to freedom of speech should be judged by reference to the liberty’s central range—that is, by reference to the liberty’s particular role in providing for the conditions necessary for the development and exercise of the moral powers.32 On the one hand, there are various ways in which free speech must be regulated in order to allow for its proper exercise. Rules of order (such as regulations regarding time, place, and manner, and the use of public space) are not to be seen as restrictions on freedom of speech. They rather regulate freedom in order to better protect its central range (especially the public use of our reason, as an exercise of the first moral power, the capacity for a sense of justice).33 On the other hand, certain uses of speech which do little to advance (and sometimes much to hinder) the development and exercise of the moral powers lie clearly outside of the liberty’s central range and so may be restricted in those cases where there is sufficient reason to do so—such as cases of libel and defamation, and advertising.34 Thus, though advertising is a kind of speech, the right to advertise is not a basic liberty, and so it should not be protected as an exercise of the basic liberty of free speech in a just society.35

The central upshot of this treatment is that identifying a right as a basic liberty should not be conceived as definitively answering the question of what Rawlsians (or other liberal egalitarians) think about particular cases in which such rights seem to come into play. Rather, such identification is the first step in a longer process of specifying the content of that right and adjusting it to the other members of the set of basic liberties. Moreover, that process must take place, for Rawls, under the guidance of an underlying conception of citizens as free and equal, and the accompanying underlying conception of social cooperation on terms of fairness and mutual respect.36 It is in this sense that Rawls offers a distinctly high liberal interpretation of the basic liberties and their priority. Unlike accounts of the basic liberties which attempt to ground them on fundamental natural rights, for example, the Rawlsian account anchors them in a particular liberal understanding of citizenship and social cooperation.37

So, for Rawlsians the basic liberties should be specified in the light of the underlying conception of citizens as free and equal, and in this way we can come to a sharper understanding of the content of those liberties, and thus of precisely what they do and don’t rule out. What does this mean for the specific case of freedom of occupation?38

The starting point for the process of specifying freedom of occupation is to think about its role in supporting the development and exercise of the two moral powers in the fundamental cases. First, and most obviously, freedom of occupational choice seems to be importantly connected with the second moral power—the capacity to formulate, revise, and pursue a conception of the good. Choosing a particular kind of career will often be an important part of pursuing a particular conception of the good; think of embarking on a career as a carpenter or a social worker or a philosopher, for example. As Rawls himself puts it while discussing a different issue: “When an individual decides what to be, what occupation or profession to enter, say, he adopts a particular plan of life.”39

Secondly, freedom of occupation (like the right to personal property) also contributes in important ways to self-respect, which is itself a significant part of developing and exercising both moral powers.40 This is partly a matter of something like integrity—being able to live one’s life in accordance with one’s deeply held commitments. Being forced to work in ways that are inconsistent with deeply held values can present serious threats to self-respect. But it is also a consequence of the relationship between self-respect and a sense of self-authorship. As we’ve already noted, choosing one’s occupation is a basic part of what it is to exercise choice over the sort of person that one is and will become. As such, it is an important component of the capacity to think of oneself as the director or author of one’s life, to think of one’s life as one’s own in a significant way. And that capacity in turn plays an important role in developing a robust sense of self-respect.

And finally, it is also worth noting that freedom of occupation plays an important supporting role with respect to other basic rights and liberties, which in turn bear important ties to the development and exercise of the two moral powers. Restricting freedom of occupation will often involve some restrictions on freedom of movement, for example, which in turn threaten what Rawls calls the “liberty and integrity of the person.”41 As such, protecting freedom of occupation might be a necessary means of protecting other basic liberties that are themselves required for the adequate development and full exercise of the two moral powers.42

With this account of the ways in which freedom of occupation supports the development and exercise of the moral powers in hand, we’re now in a position to assess the extent to which different specifications of that freedom cohere with a Rawlsian approach to the basic liberties. Specifications of freedom of occupation that contribute to the kinds of support for the moral powers just outlined fall within the liberty’s central range and are thus the focus of the protections afforded by constitutional provisions. Those which are less important for providing that kind of support do not fall within that central range and thus lie outside of the focus of that protection.

With all this in mind, let’s now return to the problem of productive justice. Suppose that we live in a reasonably just society, in what Rawls calls “reasonably favourable conditions.”43 Our society’s constitution includes a guarantee of freedom of occupation, which is justified by reference to the idea that freedom of occupation is a basic liberty and so protected by lexical priority. Suppose further that the basic needs of certain members of the population are going unmet because of the ways in which individuals are using their freedom of occupation. In response, legislation is proposed that would introduce a mandatory civilian service, according to which every citizen (with reasonable exemptions for those who cannot serve) must spend a period of one or two years, perhaps distributed across the course of their life, working in socially necessary professions—providing basic services in medical contexts for example, cooking and cleaning in municipal buildings, social work and care for the young and the elderly, and so on.44 Suppose that by enacting a scheme of this sort, the shortfall in socially necessary labour can be met, and the basic needs of all can be satisfied. Call a scheme of this rough form a moderate mandatory service scheme.

In setting out the problem of productive justice as a challenge to Rawlsianism in Section II, we supposed that a scheme of this sort would be incompatible with Rawlsian commitments to the priority of liberty. But we’re now in a position to see that that supposition may yet prove too hasty. Whether such a scheme is compatible with a Rawlsian commitment to freedom of occupation depends on how that freedom is to be specified, where the process of specification should proceed along the high liberal lines described above, i.e., by reference to the ways in which freedom of occupation supports the development and exercise of the moral powers.

So, how much of a threat does a moderate mandatory service scheme pose to the capacity of freedom of occupation to support the two moral powers? Consider first the connection between freedom of occupation and the second moral power. Would a moderate mandatory service scheme threaten people’s capacity to formulate, revise, and pursue a conception of the good by choosing an occupation? There are clearly some restrictions of this kind involved with moderate mandatory service. For a year or two the state would dictate to people what kind of work they would do, and so individuals would have one or two years less in which they could (for example) choose to pursue a particular kind of career that makes a particular contribution to their chosen conception of the good. But it’s important not to overstate the extent of this threat. Moderate mandatory service might delay people from entering a profession of their choosing, and this will create some obstacles to the pursuit of certain conceptions of the good. But citizens would remain free to enter an occupation of their choosing once they have completed their service, and so their capacity to shape their life in a particular direction by choosing the occupation that they will devote the majority of their lives to remains, to a very significant degree, protected under the proposed scheme.45

Something similar can be said for the connection between freedom of occupation and self-respect—indeed in that instance the case for thinking that moderate mandatory service does not affect the central range of freedom of occupation is even greater. For though moderate mandatory service does involve the state deciding occupations for individuals, the time spent working in an occupation of the state’s choosing is limited enough that it does not threaten the grounds of a confident sense that one has authored one’s own life, in the sort of way required for the self-respect needed to develop and exercise the moral powers.46 A well-designed mandatory service scheme could also afford conscripts some scope for significant choice within the context of the scheme, with conscript preferences between different kinds of work taken into account in job assignments wherever possible. And finally, though there will be some restrictions on corollaries of freedom of occupation (such as freedom of movement or association), it is again not clear that those restrictions will threaten the central range of those liberties. From 9–5, five days a week, for a year or two, conscripts will be required to be in a particular place, with particular people, in order to discharge their service obligations. But outside of that time they will be free to move and associate as they please. It’s not clear that this isn’t entirely adequate to protect the requisite measure of freedom of movement and association required for securing the social conditions of the adequate development and full exercise of the moral powers in the fundamental cases.

So, the threats presented to the central range of application of freedom of occupation seem limited. But in addition to this, specifying freedom of occupation in such a way as to allow for moderate mandatory service also seems to help produce an overall package of basic liberties that is better suited to supporting the development and exercise of the moral powers. The most obvious way in which this is the case is that, in allowing for moderate mandatory service, such a specification would allow for a policy that would ensure that the basic needs of all can be satisfied. Insofar as those basic needs being met is a necessary condition for citizens to understand and exercise their basic liberties, securing those basic needs is obviously an important part of securing the social conditions necessary for the development and exercise of the moral powers.47

But it is also worth noting that there is some reason to believe that allowing for moderate mandatory service could help support the exercise and development of the moral powers in some more subtle ways too. Though this is necessarily speculative, there is some reason to suppose that (well-designed) mandatory service could support the development of a social ethos that is more conducive to the development and exercise of the moral powers. One reason to think this is that mandatory service schemes can serve to undermine pernicious status inequalities that are bound up in various ways with attitudes towards certain kinds of work. For example, casting care work as a universal duty might encourage a broader appreciation of the significance (and burdensomeness) of care work, and so undermine attitudes that underpin social norms that govern the gendered division of labour.48 Similarly, mandatory service could help change attitudes towards certain kinds of “low status” work, attitudes which can threaten to undermine the self-respect of those who perform such work for a living.49 It’s worth emphasising just how great an obstacle such social norms can constitute with respect to the second moral power in particular. Living one’s life in the shadow of powerful social norms to the effect that one is only good for “low-grade” work can drastically reduce one’s capacity to formulate, revise, and pursue a diverse range of conceptions of the good. Another reason to think that mandatory service schemes could help support the development and exercise of the moral powers is through encouraging greater social equality, by forcing all citizens, from various different walks of life, to work together on equal terms for a common goal.50 Such social equality can, in turn, play a key role in protecting the self-respect of citizens, which in turn strengthens their capacity to develop and exercise the two moral powers.

Bringing these threads together, we’re in a position to see that a moderate mandatory service scheme presents a limited threat to the capacity of freedom of occupation to support the development and exercise of the moral powers. In other words, it puts minimal, if any, pressure on the central range of freedom of occupation. Moreover, specifying freedom of occupation in such a way as to allow for moderate mandatory service better allows for that freedom to form part of a package of basic liberties that better protects the capacity of citizens to exercise those basic liberties within each of their central ranges of application. Given this, it seems that a high liberal specification of the content of freedom of occupation may well entail that moderate mandatory service is compatible with that freedom, at least in circumstances sufficiently similar to those supposed above.

It is important to note that the claim here is not that freedom of occupation would be restricted by such a scheme, but that we needn’t worry about this because those restrictions would be adequately compensated by gains in terms of meeting basic needs and undermining pernicious status inequalities and unjust social norms. That claim would fly in the face of the lexical priority rule, given the distinction between the basic liberties and their worth. Rather, the argument is that (in the right circumstances) a society may well have good reason to rule that moderate mandatory service simply does not constitute a restriction on the basic liberty of freedom of occupation at all. When that freedom is properly specified, then (adequately designed and implemented) moderate mandatory service may regulate the basic liberty of freedom of occupation, but it would not restrict it—in the same way that advertising regulations and time, place, and manner restrictions regulate (and don’t restrict) freedom of speech, and limits on rights of acquisition and bequest regulate (and don’t restrict) the right to personal property.

Equally, the argument is not that moderate mandatory service wouldn’t restrict liberty, in the general sense, at all. It clearly would, and so it would face the general presumption against restricting liberty without sufficient reason that is presupposed in Rawls’s theory.51 Rather, the claim is that appropriately designed moderate mandatory service would not threaten the central range of application of freedom of occupation, and so does not threaten the specially protected basic liberty of freedom of occupation, even if it does affect other liberties that aren’t afforded the special protections of lexical priority.

If the main argument of the previous section is correct, then the Rawlsian has a compelling response to the challenge presented by the problem of productive justice. For in that case, commitment to protecting freedom of occupation as a basic liberty is in fact compatible with the sorts of moderate mandatory service schemes that could help address shortfalls with respect to the productive activity necessary to meet basic needs. In the relevant sense, Rawlsians need not be committed to claim (2) from the inconsistent triad set out in Section II, which claims that justice forbids forcibly assigning jobs (except where other basic liberties are at risk). Rawlsians can reject that claim because some ways of forcibly assigning jobs (such as moderate mandatory service schemes) are not forbidden by plausible specifications of freedom of occupation, as they do not sufficiently threaten the freedom’s central range of application. Just as justice does not forbid extensive regulation of rights of acquisition and bequest, it does not forbid moderate mandatory service.

I take it that this provides a plausible response to Stanczyk’s triadic formulation of the problem of productive justice. However, it’s worth noting three independent ways in which this response represents the beginnings of a plausible approach to questions of productive justice more broadly, and so the beginnings of an attractive theory of productive justice in general.

First, the high liberal approach to freedom of occupation doesn’t just provide a justification for moderate mandatory service. It also provides us with the resources to explain why more extreme mandatory service schemes would not be compatible with justice, and to understand where the cut off between moderate and immoderate mandatory service lies. For the high liberal approach tells us that freedom of occupation’s central range of application will be threatened when potential incursions upon it do begin to impair the capacity of citizens to develop and exercise the moral powers. Precisely when this occurs will, of course, be a difficult matter to assess, and will most likely vary between different societies. But on either side of a difficult grey area there will be fairly clear-cut cases. Forcibly assigning jobs for twenty years, for example, would certainly seem to represent a serious obstacle to people’s development and exercise of the second moral power as well as to their sense of self-authorship, as would forcing people to do jobs that carry high risks of damage to physical or mental health. Extreme mandatory service schemes thus do clearly represent serious restrictions on basic liberty and would be extremely hard to justify in any but the gravest of circumstances.

A Rawlsian approach to productive justice thus helps us in finding a way to draw a line between moderate mandatory service schemes that do not pose a serious threat to basic liberty and extreme schemes that do. Those distinctions are to be drawn by reference to the underlying conception of citizens as free and equal and the accompanying notion of the two moral powers. In this way, a Rawlsian theory of productive justice seems to offer us a useful framework for making judgments about difficult trade-offs between our productive obligations on the one hand, and the requirement to respect people’s “right to be something other than an engine for the welfare of other people” on the other.52

Secondly, this Rawlsian approach can also helpfully illuminate more complex cases of urgent unfilled jobs that require considerable training and so cannot be filled by means of straightforward moderate mandatory service. Consider, for example, Stanczyk’s main example of medical positions, such as nursing and doctoring jobs, in regions that struggle to attract or retain adequately trained professionals (such as Ghana, and rural areas in the USA, like Mississippi’s Issaquena County).53 May Rawlsians endorse policies designed to force those who already have adequate training to remain in, or relocate to, underserved areas to fill shortages that threaten to leave the basic needs of others’ unmet? The approach developed in this paper allows us to see how they can.

Take Ghana’s former policy of imposing prohibitive fines on nurses who wish to emigrate before providing five years of local service, on pain of withholding educational certification.54 No doubt such schemes place some restrictions on freedoms that are importantly connected to freedom of occupation and freedom of movement. Nevertheless, there is good reason to think that they would not put significant enough pressure on people’s capacity to develop and exercise the moral powers that they fall within the central range of those freedoms and so deserve protection under lexical priority. For a start, such policies do not force trainee nurses to keep on nursing. They retain the right to stop nursing and will in that case face no fines for emigration (as they also retain the right to decline to train as a nurse in the first place).55 It is difficult to see how someone could reasonably complain that the absence of the specific option to train as a nurse without incurring any mandatory service obligations places great strain on their capacity to develop and pursue their own conception of the good life, or their sense of authorship over their lives. More generally, as Stanczyk himself points out, “there is no reason to suppose that [free choice of occupation] requires free choice between unregulated professions.”56 The Rawlsian approach can explain why: free choice between unregulated professions is not necessary for protecting the capacity of citizens to develop and pursue a diverse range of conceptions of the good life, or a healthy sense of self-respect grounded in a sense that their life is self-authored. As such, it falls outside of the central range of freedom of occupation and is not protected by lexical priority.

Note too that this approach can also help us see when and why alternative policies in this area would be ruled out by our commitment to freedom of occupation. For example, a policy that required anyone with sufficient talent in high school to train as a doctor and then to work for ten years in an underserved area (even if they have no interest in doctoring) would clearly have much more significant implications for the capacity of individuals to develop and exercise the moral powers.

Thirdly, a high liberal approach to productive justice also helps explain and justify our intuitions about the specific shape that a just moderate mandatory service scheme would have to take. Because a plausible specification of freedom of occupation must be guided by an underlying conception of citizens as free and equal, we should expect just mandatory service schemes to reflect that conception. Rawls describes in detail how this underlying conception shapes a high liberal specification of freedom of speech, explaining how, for example, it introduces a very high bar for content-based restrictions of speech (as opposed to content-neutral regulations), and how it requires that regulations on speech in the name of securing the fair value of the political liberties must be imposed in an equal fashion.57 But similar arguments will be available in the case of productive justice. For example, mandatory service schemes should respect the freedom and equality of citizens by remaining neutral between competing conceptions of the good. A mandatory service scheme that was introduced in order to inculcate “manly virtues” of “toughness” and “discipline” for example is very unlikely to be compatible with a high liberal understanding of the basic liberties.58

Similarly, a Rawlsian approach gives firm justification to the view that a moderate mandatory service scheme would have to be universal—placing equal burdens on all as much as possible. This would strongly speak against, for example, allowing the wealthy to buy their way out of service by paying someone else to discharge their service obligations.59 If the overall package of basic liberties is to reflect citizens’ status as free and equal, the burdens of mandatory service should apply to rich and poor alike. In each case, useful guidance and plausible judgments can be found by employing the high liberal approach of understanding the basic liberties and their priority in the context of an underlying conception of free and equal citizens engaging in social cooperation on terms of fairness and mutual respect. Once again, the Rawlsian high liberal approach provides us with a fruitful framework to navigate these sorts of complex issues.

Before concluding it is worth considering a pair of objections, the responses to which help to further clarify the advantages of the approach defended in this paper. The first objection alleges that the approach advanced in this paper is importantly incomplete as a response to the problem of productive justice, and the second alleges that it is also unnecessary, given the possibility of a more direct Rawlsian response to the problem, which does not suffer from the alleged incompleteness of my proposed solution.

The incompleteness worry takes its cue from the fact that the argument I have presented establishes only that moderate mandatory service is compatible with affording freedom of occupation the special protections envisaged by the priority of liberty. As noted in the previous section, the envisaged justification would not apply to extreme mandatory service requirements, such as those which would forcibly assign jobs for decades, rather than just years. But, the worry continues, that leaves the central theoretical challenge of the problem of productive justice in place. For what if it were to turn out that moderate mandatory service is not sufficient to fill shortfalls in the provision of basic goods and services? However empirically unlikely this may be, it’s just as much of a theoretical possibility as the possibility that less intrusive policies (such as incentives and subsidies) will not suffice. And so, Stanczyk’s central theoretical challenge to Rawlsianism and liberal egalitarianism might seem to persist, in spite of the arguments offered above.

Moreover, this incompleteness problem may seem to be all the more pertinent when we consider that an alternative Rawlsian response to the problem of productive justice seems available, which might seem to offer a more complete response to Stanczyk’s challenge. The challenge here begins with the later Rawls’s endorsement of Rodney Peffer’s “basic needs principle.” The basic needs principle plays no role in A Theory of Justice, but is endorsed by Rawls as an amendment to his theory in Political Liberalism, in response to a suggestion of Peffer’s in his analysis of Marxist critiques of Rawls in Marxism, Morality, and Social Justice.60 Peffer’s suggestion comes in the context of a discussion of the charge that Rawls’s commitment to the priority of liberty essentially affords absolute priority to negative liberty over all other demands of social justice, and that this extreme privilege for liberty represents an implausible lack of balance.61

After considering various responses available to Rawls, Peffer argues that Rawlsians can most plausibly respond to the criticism by appending an amendment to his theory in the form of an additional principle, requiring “that all individuals must be guaranteed a minimum level of well-being and that this has priority over any and all other considerations of social justice.”62 By appending a basic needs principle of this type to the theory as presented in A Theory of Justice, Peffer argues the Rawlsian view can evade the criticism that its insistence on priority for liberty is implausible, since that insistence looks far more reasonable in the context of a guaranteed social minimum. Most significantly for our purposes here, Rawls himself endorses Peffer’s suggestion in Political Liberalism, writing that “the first principle covering the equal basic rights and liberties may easily be preceded by a lexically prior principle requiring that citizens’ basic needs be met, at least insofar as their being met is necessary for citizens to understand and to be able fruitfully to exercise those rights and liberties,” referring to Peffer’s discussion for a fuller statement of the principle.63 If, however, a lexically prior basic needs principle is added to the two original principles of Rawls’s theory, it might seem that Stanczyk’s problem has little force against Rawls’s theory.64 For if ensuring that the basic needs of all citizens are met is assigned lexical priority over even the basic liberties, the needs of the worst off will already trump the freedom of occupation rights of the better off—the problem of productive justice will effectively be defended by abandoning the priority of liberty in circumstances where basic needs go unmet.

Rawls’s explicit endorsement of Peffer’s amendment provides solid exegetical grounds for this response as far as Rawls’s official position goes. Nevertheless, there are good reasons to be dissatisfied with it, and so to prefer the approach provided in this paper. The first reason is that while Rawls might be (perhaps surprisingly) happy to abandon the priority of liberty in the face of basic needs going unmet, one suspects that many other liberal egalitarians will be less keen to do so, especially in the context of the challenge posed by the problem of productive justice. Many people, even in the richest nations in the world, live in poverty, with inadequate access to things like basic education and healthcare. Perhaps their needs should always take absolute priority over the basic liberties of other members of society. But accepting this would appear to be a significant departure from the usual interpretation of the significance of liberty in the liberal tradition. This more direct response will therefore only satisfy those liberal egalitarians who are happy to follow Rawls in endorsing lexical priority for Peffer’s basic needs principle. The approach advanced in the earlier sections of this paper will have a broader appeal.

The second reason is that the alternative response under consideration has some importantly distinct and, to my mind, much less plausible implications than the approach I have developed here. These implications are easiest to see by considering the second objection that we introduced at the beginning of this section—the incompleteness objection. According to the incompleteness objection, the approach I have developed in this paper does not address the core challenge posed by the problem of productive justice, since it will allow for only moderate mandatory service, and thus will not be able to recommend mandatory service solutions in those extreme cases in which moderate proposals will not suffice to address the relevant labour shortages. By contrast, an approach which held that satisfying basic needs should always be afforded lexical priority over the basic liberties would be able to recommend even extreme mandatory service proposals.

My response to this challenge, however, is to point out that it’s far from clear that an endorsement of extreme mandatory service proposals (even in extreme circumstances) is a virtue in a theory of productive justice. Rather, the “incompleteness” of the approach advanced in this paper strikes me as a feature rather than a bug, and an advantage over the alternative of appealing to a lexically prior basic needs principle. To return to a case briefly considered in the previous section, suppose that we are in a situation where not even moderate mandatory service can address shortfalls in the provision of basic goods such as healthcare. The only way to ensure adequate access to a minimally decent life is to take eleven-year-olds with promising test scores, force them to study medicine, and then force them to spend the majority of their working lives as medics in underserved areas. The appeal to a lexically prior basic needs principle would appear to suggest that such a policy would be obviously justified. But that’s deeply counter-intuitive. The approach advanced in this paper, by contrast, is able to explain why a policy like this most likely would not be justified, even while a more moderate mandatory service proposal most likely would be. As explained in the previous section, specifying freedom of occupation such as to allow for the more extreme proposal puts clear and extreme pressure on the capacity of freedom of occupation to protect the development and exercise of the moral powers in a way that the moderate proposal does not.65

The incompleteness worry is therefore misplaced. The proposal advanced in this paper doesn’t counsel that basic needs should take priority over the basic liberties in all cases. But this is an advantage of the view rather than a disadvantage. Moreover, this shows why the approach is preferable to the alternative Rawlsian response of appealing to a lexically prior basic needs principle. Such a response trades an implausibly inflexible insistence on the priority of basic liberties over the satisfaction of basic needs for an implausible inflexibility in the other direction. The approach defended here, by contrast, provides a framework that can generate reasonable, theoretically grounded judgments about how difficult trade-offs should be struck in scenarios where such trade-offs, tragically, must be made.

It is tempting to think that the answer to the question of whether mandatory service schemes infringe upon freedom of occupation is obvious. Mandatory service forces you to work in ways that you haven’t yourself chosen. Perhaps such restrictions are sometimes justified, but of course they represent restrictions on the liberty itself—how couldn’t they? As such, it is tempting to think that, as Samuel Arnold puts it, mandatory service “plainly violates freedom of occupational choice as liberals are wont to conceive of it.”66 But the arguments of this paper show that that temptation is one that Rawlsians and liberal egalitarians more broadly have good reasons to resist. While certain kinds of libertarian might be keen to embrace such an understanding of liberty, Rawlsians are already committed to a more complex view of what the basic liberties are, and what it takes to restrict them (as opposed to merely regulating them). As such, just as liberal commitments to freedom of speech and rights of personal property do not commit them to opposing the regulation of advertising or rights of acquisition and bequest, liberal commitments to freedom of occupation do not commit them to opposing all forms of mandatory service. The implications of the high liberal method of specifying the content of the basic liberties are already well accepted with regard to speech and personal property. Extending those implications to freedom of occupation reveals that Rawlsians can avail themselves of similarly nuanced and complex positions on productive justice as they can in more familiar debates about speech regulation and taxation.67

Most obviously, this means that Rawlsians and liberal egalitarians are not vulnerable to the problem of productive justice in the way sketched in Section II above. They can simply reject claim (2): that justice forbids forcibly assigning jobs except where liberties are at risk. Since not all forcible assignment of jobs threatens the central range of application of free choice of occupation, not all such assignment restricts that freedom, and so not all mandatory service proposals are subject to the stringent standards imposed by the doctrine of the priority of liberty. Rawlsianism thus has the resources to explain why moderate mandatory service would be called for in the sorts of circumstances presupposed by the problem of productive justice.

But looking forward to future research, the arguments of this paper also give us good reason to think that the underlying approach to productive justice supplied by the Rawlsian model is in fact very well suited to helping us think through issues of productive justice more generally. By maintaining a commitment to the priority of the basic liberties, but combining that commitment with a high liberal understanding of how those basic liberties are to be specified and adjusted to one another, the Rawlsian model offers us a helpful way of thinking about how to weigh our productive obligations against the weight of our rights to live our own lives on our own terms.