1. INTRODUCTION
Some egalitarians believe that the well-off are under a duty of justice to transfer material resources to the worse-off. Of those who hold that view, some also claim that there are good reasons to turn that moral duty into a legal duty via coercive taxation. By contrast, very few egalitarians also believe that the talented are under a moral duty of justice – let alone a legally enforceable duty – to provide services to the worst-off in the form of socially useful occupations. On the whole, egalitarians thus differentiate between coercive taxation and forced labour. And yet, consider Nozick's well-known statement to the effect that:
Taxation of earnings from labour is on a par with forced labour. Some persons find this claim obviously true: taking the earnings of an hours [sic] labour is like taking n hours from the person; it is like forcing the person to work n hours for another's purpose. Others find the claim absurd. But even these, if they object to forced labour, would oppose forcing unemployed hippies to work for the benefit of the needy.Footnote 1
Nozick's challenge to coercive taxation as on a par with forced labour has not been taken seriously by egalitarian liberals, with the notable exception of G. A. Cohen. In Self-Ownership, Freedom and Equality, Cohen argues that one can object to forced labour without thereby committing oneself to rejecting taxation (at pp. 230ff.). In his recent Rescuing Justice and Equality, he mounts a sustained critique of coerced labour, against the background of a radical egalitarian conception of distributive justice (ch. 5).Footnote 2
In this article, I examine Cohen's views on money and labour. I defend the following two claims:
(1) Cohenian egalitarians are committed to holding the talented under a moral duty to choose socially useful work for the sake of the less fortunate. (section 2)
(2) Cohen's arguments against coerced labour fail, particularly in the light of his commitment to coercive taxation. (section 3)
In the course of defending those claims, I shall argue that Cohen's remarks on freedom of occupational choice and taxation exhibit partiality towards the interests of the better-off to the detriment of the less fortunate – a partiality which is in tension with his commitment to equality.Footnote 3 Before I begin, however, let me set out briefly the key features of Cohen's just society:Footnote 4
(a) Individuals are not worse off through no fault of their own.
(b) The appropriate egalitarian metric is access to desirable conditions of life (or ‘advantage’), and thus combines resources and welfare.
(c) The better-off are permitted to confer greater weight on their own goals, projects and attachments, to the detriment of the worse-off, at least within limits.Footnote 5
(d) The requirements of justice apply not merely to society's legal institutions, but also to agents’ personal choices in their daily lives.
Cohen accepts that there are values other than justice which might conflict with the latter and which, all things considered, might be deemed to outweigh it (RJE, 7). According to Cohen, Pareto-efficiency is one such value. So (to give another example) is democracy. A society might thus be democratic, or organized, distribution-wise, along Pareto-efficient lines, and yet be unjust. Contrariwise, it might also be undemocratic, Pareto-inefficient, and yet just. My claim that, if one accepts (a)–(d) above, the talented are under a moral duty to provide labour to the less fortunate is a claim about what a just society is – not about what the talented may or should do, all things considered, for the sake of the less fortunate. Likewise, when rejecting Cohen's objections to the legal enforcement of that moral duty, I shall argue that he has not shown what he seeminglyFootnote 6 wants to show, namely that a society in which the talented are legally coerced to act justly is unjust. Whether or not values other than justice support the claims that those agents ought not to be made to do what justice requires of them is a separate question which I do not address here.
One final remark before I begin. Cohen is often referred to as a radical egalitarian – on the grounds that his theory of justice is particularly demanding on the better-off. And yet, he too balks at some of the implications of his theory for freedom in general, and freedom of occupational choice in particular. That tells us something about the enduring appeal of freedom, and the task which any conception of justice faces – particularly welfarist conceptions, whether or not those conceptions are egalitarian, prioritarian, or sufficientist. For on those conceptions, the poor, or the worse-off, are not owed material resources so much as whatever is required to further their well-being. As we shall see below, such duty can be discharged via the provision, by the talented, of socially useful work. Accordingly, my objections to Cohen's views on labour and money are relevant to any theory of justice which, on the one hand, claims that the better-off are under an enforceable duty to transfer material resources to the poor or worse-off, and, on the other hand, maintains that the talented are not under an enforceable duty of justice to labour for the sake of the latter.
2. FROM MONEY TO LABOUR: ON THE DUTY TO CHOOSE SOCIALLY USEFUL OCCUPATIONS
In chapter 5 of RJE, Cohen examines what some might regard as an insurmountable trilemma between three values: equality, Pareto-optimality and freedom of occupational choice. Consider a person, A, who can be a doctor or a gardener. Her preference-ordering is as follows (in descending order): doctoring at £50,000 p.a.; gardening at £20,000 p.a.; doctoring at £20,000 p.a. As doctoring is of particular social value to the badly off, her community's preference ordering somewhat differs from hers: they would much rather that she be a doctor at £20,000 p.a., followed by doctoring at £50,000 p.a., and, finally, gardening at £20,000 p.a.Footnote 7
According to the trilemma thesis, we cannot have all three of income equality, freedom of occupational choice and Pareto-optimality. For if we decide to preserve equality and to respect A's freedom, we have to freeze salaries at £20,000 and to let her choose her occupation. However, in so doing, we violate Pareto (since both A and her fellow community members are worse off than they could otherwise be). If, on the other hand, we seek to satisfy both income equality and Pareto, we have to freeze salaries at £20,000 and force A to work as a doctor, thereby violating her freedom of occupational choice. Finally, if we choose freedom and Pareto, we have to allow A to choose doctoring at £50,000 p.a., in violation of income equality.
Let us accept that we must be able to satisfy both income equality (within the constraints of the personal prerogative) and Pareto. The question is whether or not to sacrifice freedom of occupational choice, understood as not being coerced, by the state or anyone else, into particular jobs. Some egalitarians, Cohen concedes, would make that sacrifice and have the state force A to be a doctor – a ‘Stalinist solution’ which he rejects as illiberal (RJE, 186–7). Fortunately for egalitarians, however, there is another solution – the ethical solution – which was first proposed by Joseph Carens, and which, Cohen tells us, meets all three desiderata.Footnote 8 Let us suppose that A freely chooses to be a doctor for £20,000 p.a., out of principled commitment to equality as well as concern for other human beings. In so far as A chooses to do that which she is not legally constrained to do, the ethical solution preserves her freedom of occupational choice. Moreover, the ethical solution satisfies Pareto, since it makes both the less fortunate and A better off than they could otherwise be. Finally, the ethical solution does not require of A that she should choose a socially useful occupation which would be so burdensome as either to make her worse off than others or to violate her legitimate personal prerogative. In fact, her willingness to be a doctor (for a price) suggests that doctoring is not unacceptably repugnant to her. Accordingly, the ethical solution preserves equality. Crucially, what applies to A's case applies to the talented in general, since it is standardly true of the talented, Cohen avers, that the jobs which they normally do are rewarding to them, even if they are not their first choices (e.g. RJE, 55–6, 103). Thanks to the ethical solution, then, we have no trilemma, provided that we are able to overcome a number of organizational difficulties, such as epistemic constraints which make it very difficult to know what counts as socially useful occupations and how many agents would have to be directed to them.Footnote 9
I shall not assess here whether the ethical solution solves the trilemma, and, in particular, preserves freedom.Footnote 10 Instead, I shall argue that even if the ethical solution succeeds where the Stalinist proposal fails, Cohen's dismissal of the latter is too quick. As we shall see in section 3, his objections to forced labour are not convincing, particularly in the light of his commitment to coercive taxation. As a preliminary step towards that conclusion, I shall now argue that A is under a duty of justice, at the bar of Cohenian egalitarianism, to choose doctoring over gardening. Cohen himself briefly makes that point in the course of defending his account of the ethos of justice (RJE, 370–1): in what follows, I seek to provide a fuller argument to that effect.
Recall A's choices: doctoring at £50,000 p.a.; gardening at £20,000 p.a.; doctoring at £20,000 p.a. Her decision, thus, has two distinct components: a reward decision to extract a given salary (£50,000 v. £20,000 p.a.), and an occupational decision to embark on a particular career rather than another (doctoring v. gardening). According to Cohen, equality governs A's reward decision (£20,000 p.a., which would ensure that A is not much better off than most others); but the commendable feature of A's occupational decision is something else altogether, to wit, the Pareto-satisfying fact that A's fellow community members are better off as a result of A's choice for doctoring, since their preference influences what gets produced (RJE, 185 and 191).
However, A's occupational decision, like her reward decision, can also be seen in a rather different light – to wit, as the fulfilment of a moral duty of justice to choose socially useful professions.Footnote 11 In fact, so to regard A's decision is entirely consonant with Cohen's egalitarianism. For consider. Cohen's account of egalitarian justice stipulates that individuals should have equal access to desirable conditions of life. This in turn imposes on others an obligation to help secure such access, by refusing to live a life that is more rewarding, all things considered, than the lives of other individuals, within the constraints of a legitimate personal prerogative. But there are several ways in which duty-bearers can discharge their egalitarian obligations to the less fortunate. Health – a desirable condition of life if there is one – perfectly illustrates the point. Our doctor-gardener can help the sick either by agreeing to distribute, via taxation, a share of her income, so as to help fund the National Health Service, or by agreeing to become a doctor, or both. It is a key principle of Cohen's theory of justice that individuals are under an egalitarian duty to divest themselves of part of their income for the sake of the worst-off. On his view, A is under a duty to forego £50,000 p.a. and be content with £20,000 p.a. She is thus faced with a choice between doctoring at £20,000 and gardening at £20,000. Given that (ex hypothesi) doctoring would neither make her worse off than is permitted by equality nor violate her personal prerogative, justice requires of her that she forego gardening and opt for doctoring instead.
In response to that last point, some people might be tempted by the following move. Namely, although one may hold agents under a moral duty to help others by way of income transfers, we may not ask them to discharge that duty by spending the majority of their waking hours doing a particular job. The difference between money and labour, those critics might insist, is that the former is not, but the latter is, intimately connected to the person we are. As we shall see below, that thought is what drives one of Cohen's objections to forced labour. It is also a motif which we find in some of his articles on freedom. In contrasting the worker's lack of freedom (forced as he is, Marxists allege, to sell his labour) with the capitalist's lack of freedom (forced as he is, some libertarians allege, to invest his capital), he writes:
For the worker is more closely connected with his labour power than the capitalist is with his capital. When I sell my labour power, I put myself at the disposal of another, and that is not true when I invest my capital. I come with my labour power. I am part of the deal.Footnote 12
Correspondingly, when we hold the better off under a moral duty to forego part of their income for the sake of helping the less fortunate, we do not thereby hold them under a duty to put themselves at the latter's disposal; yet, we do precisely that by holding A under a moral duty to become a doctor.
However, that line of argument would not be available to Cohenian egalitarians – at least, not as an argument about what justice requires. For ex hypothesi, that which A is asked to do neither makes her worse off than the less fortunate nor violates her personal prerogative – any more than giving a share of her material resources to the less fortunate would do. Even if there is a sense in which she would thereby be at other people's disposal, her life would not, thereby, be blighted. By contrast, her decision to become a gardener would be highly costly to the less fortunate, deprived as they would be of her doctoring services. Under those circumstances, it is not clear why we should confer greater weight on the satisfaction of A's first occupational preference than on helping the worse-off; in fact, to do so seems at odds with the kind of commitment to equality which Cohenian justice otherwise displays. Cohen is absolutely right, therefore, when he writes: ‘Egalitarian justice requires people to have some regard to equality not only when negotiating for rewards but also when making career choices’ (RJE, 370).
3. FROM THE MORAL TO THE LEGAL DUTY TO HELP: COHEN'S REJECTION OF COERCION
I have argued that, at the bar of Cohenian egalitarianism, the talented are under a moral duty of justice to choose equality-friendly occupations. Should those moral duties be legally enforced? Prima facie, if the moral duty of justice to help the worse-off by way of financial transfers ought to be turned into a legal duty, so ought the duty to provide labour – unless one can point to features of the confiscation of labour which (a) tell against it and (b) are not also features of taxation. In Rescuing Justice and Equality, Cohen, who endorses coercive taxation as just, claims to have identified such features.
In order to justify conferring on an agent the legal right to act wrongly, one must show that granting the right is preferable to enforcing the prohibition on acting wrongly. Typically, granting the right is thought preferable to legal enforcement for three kinds of reasons: enforcement is practically unfeasible; it violates the very value (here, the value of justice) which informs the moral requirement at issue (and is therefore, in the case at hand, unjust); it violates some other value whose importance is deemed to override the value which informs the moral requirement (and which, in the case at hand, overrides justice). Now, whilst some of the considerations which Cohen adduces against coercion are phrased as pragmatic concerns, they do in fact raise importance issues to do with justice (or so I shall show). Moreover, it is not always clear whether Cohen's explicitly normative objections to coerced labour, at RJE, 218ff., are deployed on the front of justice, of values other, and more important, than justice, or both. However, in so far as his main concern, at least in Rescuing Justice and Equality, is to delineate what a just society is, I shall interpret his non-pragmatic objections to enforcement as stemming from justice.Footnote 13
Cohen deploys four objections to the legal enforcement of the moral requirements of justice with respect to labour. The deterrence objection claims that compelling individuals to work in socially useful occupations might deter them from acquiring the relevant skills. The informational constraints objection notes that the state simply could not acquire the information necessary to direct the right kind of individuals into the right kind of socially useful occupations. The motivational objection holds that even if those two problems could be solved, it is preferable that agents do the right thing (working in socially useful occupations) for the right reasons (out of commitment to equality). Finally, the Kantian objection rejects compulsion for unacceptably treating agents as means. I shall argue that, on grounds of justice, none of his arguments are particularly convincing against the Stalinist proposal for coerced labour.
3.1. The deterrence objection
Cohen's first objection to coerced labour – the deterrence objection – claims that compelling individuals to work in socially useful occupations might deter them from acquiring the relevant skills. Although Cohen describes deterrence issues as a pragmatic concern, the objection can be construed as a justice-based argument against the enforcement of a duty of justice, as follows: ‘if A knows that she will be forced to doctor, she will not acquire doctoring skills, and will thus make sure that she can work as a gardener (her preferred occupation). As a result, the less fortunate will be deprived of the doctoring services which they are owed as a matter of justice. If, however, A is not coerced, she might develop doctoring skills, and deploy them to help the less fortunate. From the point of view of the less fortunate, a policy of coercion fails to bring about justice.’
At first sight, deterrence problems do seem insuperable, though it pays to note that they arise, not simply with agents whose own choices might be subject to coercion, but with their parents as well who, knowing that their children would be coerced into socially useful jobs which they do not want to do, would discourage them from acquiring the relevant skills. However, the objection is vulnerable to the following two replies. First, the claim that the talented would not acquire socially useful skills under a policy of coerced labour and that the less fortunate would therefore not get what is owed to them as a matter of justice stands only if the talented are permitted to hold out for extra rewards (in other words, if justice-based coercive taxation, which Cohen endorses, is not in place). Recall our doctor-gardener, A, whose first occupational preference is gardening, but who would rather, all things considered, doctor for £50,000 p.a., followed by gardening for £20,000 p.a. and doctoring for £20,000 p.a. According to the deterrence objection, forcing A to doctor would deter A from acquiring doctoring skills. But if A's society is one in which earnings are taxed as required by justice, A will not be able to get the income for which she would be willing to doctor. In that case, she would have no reason to acquire doctoring skills anyway, and every reason to satisfy her first occupational preference by working as a gardener. Cohen might rejoin, at this juncture, that there might be good reasons (for example, as pertain to incentives) for not taxing A at the level required by justice. Under those circumstances, A would have a material incentive to acquire doctoring skills, but would be deterred from doing so if she knew that she would be forced to do so. However, even if that concession were to rescue the deterrence objection to coerced labour, it would come at a very high cost, that of inegalitarian taxation. Whether Cohen would be willing to pay that cost for improving the chance that the talented might develop the required skills remains to be seen.
Second, suppose that agents could be educated, from an early age, into developing the principled conviction that they – and, one generation down, their children – ought to choose equality-friendly occupations and develop the appropriate skills. Under those circumstances, coercion would not have deterrent effects on agents’ training choices. Cohen might counter that, under those circumstances, where agents are motivated to fulfil their duty anyway, it is wholly unclear why coercion would be necessary. And yet, it is possible that agents may come to have the aforementioned principled conviction and, at the same time, be too weak-willed to act on it, when the time comes to embark on the occupation which they are under a moral duty to choose. As he himself admits, principled conviction and weakness of the will are wholly compatible (RJE, 171; If, 155–6). Coercion would be applied to force agents to take the jobs for which they trained (as required by their moral duty) but which they would not have the willpower to choose if not coerced to do so.
Cohen would not, I think, be persuaded by that particular move – not least because he believes that a truly just society is one where people voluntarily and willingly do that which they are required to do, without threat of coercion (RJE, 174). He might well be right that such a society is not truly just. Be that as it may, that move does meet the deterrence objection. And whilst it might not get us to a truly just society, it might get us to a society that is just enough. For here are the alternatives: (a) a truly just society, without coercion, whose agents are strong-willed enough to act justly; (b) an unjust society whose agents are weak-willed and where there is no coercion; (c) a semi-just society whose agents are too weak-willed to act justly absent coercion, but strong-willed enough to set up a justice-promoting coercive structure. On the – not unlikely – assumption that we cannot get (a), we might want to go for (c) rather than (b), given that (c) is, on the face of it, less unjust than (b). Or is it? Not according to Cohen, who claims that even if we could ‘efficiently force just the right doctor-gardeners to doctor’ (RJE, 219), we should still oppose coercion, on grounds of motivations and respect for the aforementioned Kantian principle.
3.2. The informational constraints objection
I shall discuss those grounds (as found in his third and fourth objections to coerced labour) below. Meanwhile, let us turn to the informational constraints objection. It claims that, as the state cannot know with precision what counts as burdensome labour for individual agents, it would risk conscripting into socially useful occupations agents who ought not to be forced to take those up. As Cohen points out, ‘we . . . can't tell how much the doctor-gardener dislikes doctoring, or not without an enormously invasive apparatus’ (RJE, 219). As we shall see in section 3.4, Cohen rules out the use of such an apparatus on the grounds that its use would unacceptably breach the privacy of the talented. Absent such apparatus, then, we do face informational constraints, which, for two reasons, pose a normative difficulty for the Stalinist proposal:Footnote 14 either our conscript hates doctoring so much that doing it would make her worse off than most others; or our conscript dislikes doctoring enough that it would violate her personal prerogative (without making her worse off than others). In either case, informational constraints render legal enforcement unacceptable from the point of view of justice.
Cohen seems to assume that we need to know with some detail how satisfying people find certain kinds of work as opposed to others, in order to bring about equality – a level of detail which (he says in the passage under scrutiny here) we cannot obtain. He is certainly right, as a matter of fact, that absent such knowledge of agents’ likes and dislikes, we will not know for sure whether equality obtains. The question, however, is whether our inability to secure that kind of information is good enough a justification not to coerce the talented into socially useful jobs. I believe that a Cohenian egalitarian should answer that question in the negative. To begin with, and as Cohen himself acknowledges, we can and do make rough-and-ready judgements about what agents enjoy doing and what we can and cannot legitimately ask them, in their daily life (RJE, 353, 370). Moreover, we reach such judgements about relatively young individuals, in no small measure through the educational system. In particular, we can spot, at least sometimes, and at various junctures in individuals’ trajectory, whether someone will turn to out to have talents in some (broadly defined) areas. Given that the occupations in which the talented usually end up are, on the whole, satisfying in absolute terms, and more satisfying than those of the untalented (or so Cohen tells us at RJE, 55–6, 103), knowing roughly what skills individuals are able to develop, and what broad categories of jobs they enjoy, might well be enough to permit us to coerce them.
To be sure, we will sometimes get it wrong, and I suspect that this is a risk which many, including Cohen, will not be willing to take. Rough knowledge – they might insist – is not enough to justify coercing agents. In criminal cases, for example, we do not believe that rough knowledge of who is guilty or innocent suffices to coerce: we want to know, beyond a reasonable doubt, whether defendant D has committed the wrongdoing of which he is accused.Footnote 15 Likewise – those critics might say – we want to know, beyond a reasonable doubt, that coercing A into doctoring would neither make her worse off than most nor violate her personal prerogative. And in so far as we simply cannot know that beyond a reasonable doubt, we ought not to coerce her.
Clearly, rough knowledge of D's guilt is not enough to warrant punishing him. However, punishment differs from other kinds of coercion. Consider taxation, for example. If knowledge beyond a reasonable doubt of the relevant facts is a necessary condition for coercing agents in general, then we would need to know, beyond a reasonable doubt, whether depriving someone of a share of her income makes her worse off or violates her personal prerogative, in order permissibly to do so by force (in other words, to tax her). A proponent of resource egalitarianism might be able to rejoin that we can acquire such knowledge in the case of income, though not in the case of job satisfaction. Cohen, however, cannot do so, precisely because his egalitarian metric combines resources and welfare. Just as we cannot hope to know, other than in a rough way, whether for A to become a doctor would make her worse off than most others, surely we cannot hope to know, beyond a reasonable doubt, whether loosing n-amount of income would make the well-off worse off than most others with respect to access to advantage. As it happens, Cohen does allow rough knowledge, in the case of taxation: as he puts it,
Coercive progressive taxation can be justified on the egalitarian welfarist ground that on average welfare is higher the more wealth a person has: we only need confidence in the averages, we need not invade individual psyches, to tax on a welfarist basis. (RJE, 222)
As I argued two paragraphs ago, however, and as Cohen himself admits, we do not need such detailed knowledge about individual psyches in the case of labour. Accordingly, if rough knowledge of people's personalities and tastes is enough to tell us whether or not doing a particular job is unacceptably repugnant to them, and if it is no bar to coercive taxation, then it cannot be a bar to coerced labour.
Suppose that Cohen were able to show that we can insist on stringent epistemic requirements in the case of coerced labour without jeopardizing egalitarian taxation. Let us assume, then, that egalitarian taxation is in place, so that A will only take home £20,000 pounds a year, and will thus work as a gardener if left legally free to do so. Even then, a third difficulty would arise with the informational constraints objection. The objection insists on knowing beyond reasonable doubt, or at least with some considerable degree of precision, whether A will be worse off or have her personal prerogative violated if she were to be a doctor. It then posits that we cannot obtain that kind of knowledge, from which it concludes that we ought not to coerce her. But note how biased the objection is towards the agent – A – whose labour is subject to coercion. It is tantamount to saying: ‘We should not run the risk of making A worse off/violate her personal prerogative, even though, by not taking that risk, we run a different risk, that of unjustly depriving the less fortunate of the doctoring services they need.’ But we need an argument in support of the view (implicit in the objection) that we rightly care more about A's risk of being made worse off or having her personal prerogative violated, than about the risk incurred by the less fortunate of being unjustly denied access to doctoring services.Footnote 16
In defence of the objection on that particular point, Cohen might be tempted to say that there are few more important things to us than the way in which we spend most of our waking hours. Accordingly, not only must labour burdens be catered for by any plausible theory of justice (RJE, 369): they must also be given greater weight, as a metric, than other kinds of inequality, when comparing the consequences of coerced labour for the talented and the less fortunate. Now, I agree that labour burdens are particularly burdensome. So are other burdens, however, such as, for example, health-related burdens. Accordingly, although we certainly ought not to risk blighting someone's life by asking him to shoulder considerable labour burdens for the sake of, e.g., ensuring (rough) income equality, it is far from clear, at the bar of equality, that we ought not to take that precise risk for the sake of, e.g., relieving the persistent, chronic, acute and long-term pain of someone else. Moreover, and to reiterate an earlier point, it is also far from clear why running the risk of violating someone's personal prerogative is worse than running the risk of unjustly depriving the less fortunate of the resources they need. Cohen might insist that it is worse: for if a just society is one where equality is constrained by the personal prerogative, then, under circumstances of epistemic uncertainty, we should err on the side of caution, and make sure that the latter is not sacrificed to the former (or so he might claim). And yet, that putative move is worrisome from the point of view of egalitarian justice. For if it is standardly true of the talented that they end up in jobs which are not so repugnant to them as to violate their personal prerogative, it follows that to coerce A into doctoring, absent information on its burdensomeness for her, is less likely than not to violate her personal prerogative. By the same token, however, not coercing her is more likely than not to result in the less fortunate being unjustly deprived of doctoring services. In other words, a policy of coercion is more likely than a policy of no-coercion to bring about justice. To be sure, the former places A at risk of being wronged, whereas the latter poses a risk to the less fortunate. But before concluding that it is better to risk acting unjustly towards the less fortunate than to risk acting unjustly towards A, we must be clear about the precise risks run by the latter. If the risk is that doctoring is so repugnant to her as to make her worse off than most others, then, given the greater risk, at the bar of equality, incurred by the worse-off under a policy of no-coercion, justice dictates that we favour the less fortunate, and thus recommends coercing A to doctor (at least, pending examination of other possible justice-based objections to enforcement). If, by contrast, the risk of coercion, for A, is that doctoring would violate her personal prerogative (rather than make her worse off than most others), then the egalitarian, whose baseline concern, as it were, is for the less fortunate, needs to show why we may never risk harming the personal prerogative, however small that risk. To put the point differently, we may perhaps accept, with Cohen, that equality ought to be constrained by the personal prerogative, all things being equal. Here, however, all things are not equal, since (legal) freedom of occupational choice puts the less fortunate at a greater risk of being treated unjustly than coerced labour puts the talented at a risk of being so treated. Quite why we should confer on the personal prerogative that kind of lexical priority remains unclear, given the egalitarian underpinnings of Cohen's enterprise.Footnote 17
To recapitulate, I have cast doubt on the soundness of the deterrence and informational objections to legally enforcing the moral duty of the talented to choose socially useful occupations. Let us now turn to Cohen's remaining objections to legal enforcement. Even if we could solve deterrence and epistemic difficulties, Cohen argues, we would have two further reasons for rejecting the Stalinist solution – to wit, reasons to do with motivations, and reasons to do with treating people as ends in themselves.
3.3. The motivational objection
According to the motivational objection, we prefer that agents do the right thing for the right reasons. As a policy of compulsion makes it difficult for agents to develop the right motivations, it ought to be rejected. Thus, it is better all way around if agents choose to work in socially useful occupations because they believe in justice, rather than out of fear of legal sanctions – sanctions which, again, would make it difficult for them to foster the right kind of motives (RJE, 219–20).
The objection is of a piece with Cohen's claim that a truly just society is one in which agents are motivated to act justly, without need for legal constraints. Seen in that light, it grants agents the legal right to act wrongly (by withholding that which, at the bar of justice, they are under a moral duty to provide), on the grounds that forcing them to act justly would violate justice itself. Put differently, a state of affairs in which they are forced to act justly is more unjust than one in which they are allowed to act unjustly, since in the latter, there is a greater chance that they will develop the motivations to act justly, and a greater chance, therefore, of fully realizing justice.
Cohen is right, of course, that it is better if people do the right thing for the right reason. However, the motivational objection to coerced labour is implausible. For a start, we often place agents under a legal obligation to do something which we would rather that they do for the right reason and which most of them do for the right reasons anyway. Thus, we certainly want parents to nurture their children for the right reasons, and have good reasons to believe that most of them do precisely that. Yet, we do place them under a legal duty of care. Likewise, we certainly want employers to treat their employees well for the right reasons, and have good reasons to believe that many of them do precisely that. Yet, we do impose on them legal requirements to see to their employees’ health and safety in the workplace, as well as requirements not to discriminate against their employees on arbitrary grounds such as race, gender, disability and sexual orientation.
Moreover, not only do agents often act for the right reasons irrespective of their legal duties: the law, pace Cohen, often helps them develop the right mindset in their dealings with one another. Take laws against sexual harassment in the workplace. While harassment still occurs depressingly often, there are far fewer reported cases of its most blatant expressions (groping, rubbing, sexist name-calling, etc.) than there used to be (at least in the UK and, I surmise, in the US). It is now clearly frowned upon, so that the locus of disagreement is no longer whether or not individuals have a claim to work free of unwanted sexual attention, but what counts as such. It would be naïve to ascribe such a shift to the law alone: widespread mixed education, the concomitant liberalization of the laws on abortion and contraception, together with the en masse arrival of women in the workplace have all contributed to changing attitudes towards working women. Still, there is every reason to believe that the law also played a part, not merely as an instrument for victims to come forward, get redress and gain confidence in the validity of their protest, but also as a factor in the acquisition by agents of the uncoerced motivation to treat one another respectfully in the workplace.
It is worth noting, additionally, that the law can also help instil in agents adaptive preferences which are conducive to equality. While we sometimes resent doing under the threat of coercion that which we would do entirely willingly otherwise, we sometimes come to prefer doing that which we are coerced to do – not because we are coerced, but because, being coerced, we get used to doing it, learn to see the advantages of doing it, and so on.Footnote 18 Relatedly, when pondering whether they are under an obligation to act in certain ways for the sake of others, agents often claim that so to act would be too costly for them, from which they conclude that they are at liberty not to act. Quite often, however, agents who are placed under a legal obligation to do, or refrain from doing, something and who comply for fear of sanctions, come to realize that they can easily bear the attendant costs – a process which in turn softens their resistance to the thought that they are under a moral duty so to act. Consider, for example, attitudinal changes to the appropriateness of drink-driving or of smoking in public places. We have all heard people protest, against the legal prohibition of both practices, that a good dinner cannot really be had without lots of alcohol, that one cannot really enjoy going to the pub if one cannot smoke there, etc. I surmise that those arguments are less often made than they used to be partly because many drinkers and smokers, who initially obeyed the laws largely or solely for fear of being punished, have come to realize that their objections were overstated.
3.4. The Kantian objection
Whether or not Cohen would find any of the above persuasive, he might still want to invoke his fourth objection to coerced labour – the Kantian using-as-means objection. To coerce, or ‘frogmarch’, as he puts it, agents into socially useful professions is to use them as means. Yet, ‘we should not use a person as a means (which is not to say that she should not, as it were, use herself as a means, in the interests of justice)’ (RJE, 220). Interpretative generosity suggests that Cohen here misstates his position – that he does not mean that we should not use others as means, but, rather, that we should not use them as means only. As he himself notes in his discussion of the (alleged) Kantian basis for Nozick's thesis of self-ownership, we constantly do, and in fact may, treat others as means, provided we also treat them as ends in themselves – as agents worthy of concern and respect (SO, 240).Footnote 19 To insist that we never treat them as means would be paralysingly demanding indeed.
Now, Cohen clearly cannot be taken to hold the view that the mere act of forcing people to behave justly violates Kant's requirement: that would be both patently absurd (dictating as it would against, e.g., the legal prohibition on murder) and inconsistent with his stand on coercive taxation – a stand which he affirms in both RJE (221–2) and in SO (240). Nor can he possibly mean that that into which agents are coerced is an instance of wrongful use, since, ex hypothesi, for them to work in socially useful occupations is in breach neither of equality (and, in fact, is demanded by it) nor of their legitimate personal prerogative. It is, rather, the application of coercion to that particular sphere of agents’ lives which constitutes a wrongful use.
Cohen does not provide a direct justification for the use-as-means-only objection, but his remarks on coercive taxation and forced labour give us some indication of what he might have in mind. When confronting the claim that his endorsement of coercive taxation is incompatible with his rejection of coerced labour, he distinguishes between an ideal and a non-ideal society. In an ideal society, agents fully comply with the demands of equality (subject to a legitimate personal prerogative) and neither income nor labour needs to be taxed. In a non-ideal society, by contrast, ‘citizens do not affirm and act upon the correct principles of justice’, though they want (limited) redistribution, and have elected a fully egalitarian government (the only alternative, as it happens, to a laissez-faire regime). In such a society, coercive taxation is allowed, but coerced labour is not, on aforementioned deterrence and informational grounds, as well as for the following two reasons: (a) coerced labour requires collecting such detailed information about people's personalities as to be unacceptably invasive of people's inner economy; (b) using that information to tell people what to do is unacceptably controlling of their behaviour. Neither point applies to taxation, since taxation neither requires the collection of detailed information about people's personalities, nor consists in telling them how to behave (RJE, 221–2).
In a recent article, Michael Otsuka imagines a system whereby we can spot all, but only those, cases where a hiring decision is informed by sexist prejudices. On his view (which is avowedly critical of Cohen's) surely egalitarians ought not to reject the use of that system, since the practice which it targets flouts equality. By the same token, Otsuka claims, Cohen ought not to reject the use of information-gathering techniques to direct agents towards socially useful jobs coercively.Footnote 20 Yet, Otsuka's example is not apt: simply put, the kind of information which we need in order to unearth sexist hiring decisions falls far short of that which is needed to unearth unjust occupational choices, since it does not pertain to agents’ whole ‘inner economy’. Otsuka's information-gathering system (acceptably) breaches the principle of confidentiality in hiring decisions; but it does not breach agents’ privacy in the sense in which privacy is normally understood and endorsed as an important value. Accordingly, if Cohen is right that we need very detailed information about agents’ inner economy in order to enforce the moral duty to choose socially useful occupations, he can accept the use of intrusive information-gathering in Otsuka's example, without withdrawing his objection to coerced labour.
Note, however, that Cohen's objection admits of two construals, depending on the role of privacy in his theory of justice, to wit: either we must grant the talented the legal right to act wrongly (by not providing socially useful services) on the grounds that enforcement would violate privacy – a value which overrides justice; or, we must grant the talented the legal right to act wrongly, on the grounds that the invasion of privacy which enforcement requires would violate their personal prerogative. Under the first construal, enforcement may well preserve justice but it violates a value which is deemed more important than justice. Under the second construal, enforcement violates justice itself, in so far as agents need a certain degree of privacy in order to fulfil their goals and carry out their projects, and as a policy which would infringe privacy to the corresponding degree would violate the personal prerogative. As I stressed at the outset of this article, I only deal with Cohen's justice-based objections to legal enforcement, and accordingly will focus on the second construal of the objection.
The objection thus understood supposes that we need detailed information about agents’ inner economy in order to enforce their moral duty. However, as we saw in section 3.2 when discussing the informational constraints objection, it is not clear at all that this is true. But if I am wrong on that count, then I believe that Cohen is right to reject coerced labour as unjust, precisely because, ex hypothesi, the personal prerogative constrains equality. However, that is so provided (to repeat) that enforcement requires such detailed information about agents’ inner economy. I maintain that we do not need that kind of information, and thus reject the objection.
In any event, the objection is problematic in one other respect. Suppose, for the sake of argument, that privacy does have the importance which Cohen implicitly confers on it (as part of the personal prerogative), and let us assume that individuals do not object to the collection of detailed information about themselves. Even then, Cohen rejects coerced labour, on the grounds that such information would be used to ‘tell them what to do’. But that is odd, for if they have no privacy-related objection to information-gathering of that kind, then surely we can use the latter coercively with a view to bringing about justice. Absent privacy-related objections, agents are in exactly the same position as someone whose actions (whether they consist in professional choices or not) are visibly unjust and clearly ought to be prohibited on egalitarian grounds. Unless one can point to another value which would be unacceptably jeopardized by enforcement, the justice-based case against coerced labour has not been made.
4. CONCLUSION
To conclude, I have argued that on Cohenian grounds, the talented are under a moral duty of justice to choose occupations which benefit the less fortunate. I have also argued that Cohen's justice-based objections to coerced labour fail. In particular, I have noted that Cohen's friendliness towards agents’ wrongful occupational decisions is at odds with his hostility towards their similarly wrongful reward decisions. More damagingly still, it displays unjustified bias towards the well-off (in respect of talents), to the detriment of the less fortunate.
To be clear: I have not defended the kind of forced labour which has occurred throughout history, and unfortunately continues to occur. Nor have I suggested, explicitly or implicitly, that Cohen is committed to endorsing such practices. My critique of Cohen's views on money and labour has been internal to his overall theory of egalitarian justice, and sensitive to his insistence that any requirement of justice as placed on the talented should be constrained both by equality itself and by agents’ personal prerogative (something which forced labour of the truly Stalinist kind clearly was not). Whether we should reject Cohenian egalitarianism on grounds of its demandingness is another matter, to be settled elsewhere.
