1. INTRODUCTION
If we accept that ‘we are all egalitarians now’, then the central remaining question for theorists of distributive justice is – in what sense should people be equal? Or, as Amartya Sen famously put it in his 1979 Tanner Lecture, if our aim is equality, then ‘equality of what’? Sen's answer was that we should be equal in our capabilities, or substantive opportunities, as opposed to other theories that advocated equality of resources or welfare. G.A. Cohen has advanced an influential critique of the answer that Sen, and other capability theorists, have provided to this question. Cohen (Reference Cohen, Nussbaum and Sen1993, Reference Cohen1994) suggests that capability theorists are wrong to focus on freedom or capability to function, and ignore the value of functionings achieved without the exercise of freedom.
The crux of Cohen's argument is, simply, that capability theorists take individuals’ well-being to consist in the capabilities available to them. He contends that by focusing on capabilities – and so the freedom to control one's life – capability theorists ignore passively achieved benefits, and provide an excessively ‘athletic’ account of well-being. He maintains that, on the contrary, our well-being does not depend on the world conforming to our will because it is our will, or as a consequence of our choices or actions: ‘[t]here are many benefits I get which I do not literally succeed in getting’ (Cohen Reference Cohen1994: 23). Given this, Cohen insists that theorists of distributive justice should be concerned with how good individuals’ lives actually are, as well as whether they have the opportunity to lead a good life: ‘[s]urely what matters, normatively, is whether individuals are living well’ not just whether they can (Richardson Reference Richardson2000: 318 (my emphasis)).Footnote 1 Cohen suggests that capability theorists can only incorporate a concern for these passively achieved benefits if possessing a capability is no longer taken to entail that we are free to achieve a benefit for ourselves.
Cohen, therefore, contends that capability theorists must choose between an implausibly athletic definition of well-being (wherein we are only benefited by outcomes we achieve ourselves), and an implausibly weak or expansive reading of freedom (wherein freedom does not require that we control outcomes). Previous responses to Cohen have defended the capability approach by grasping the latter horn, and insisting that an unathletic interpretation of freedom in the capability approach is not implausible.Footnote 2 Yet, I argue, capability theorists need not face this dichotomy: they can maintain both that freedom requires control, and that passively-achieved benefits may improve an individual's well-being. This is because capabilities are not presented simply as components of individual well-being, but as the appropriate goal of just distributive policies. Freedom as control can, therefore, be defended as a policy goal, without implying that such freedom is constitutive of well-being.
Thus, Cohen's critique is not, as he implies, an internal one, arguing that capability theorists have failed to meet their own goal of properly characterizing individual well-being. Rather, it amounts to the suggestion that capability theorists have erred in providing an answer to the ‘equality of what?’ question that does not incorporate a comprehensive and accurate account of well-being, and is, instead, concerned to protect and promote individual autonomy and agency. I will not take a view on whether capability theorists should become (completely or partially) welfarists. I will, however, defend capabilities as a coherent metric of distributive justice, and one likely to be appealing to those concerned to devise policy that prioritizes individual autonomy and avoids paternalism, rather than maximizing well-being by any means. We may, as Cohen (Reference Cohen1994: 124) claims, ‘unambivalently welcome’ the world coincidentally conforming with our will, without similarly welcoming government action coincidentally ensuring such conformity, without reference to our will.
I will, therefore, suggest that capability theorists should not respond to Cohen by abandoning an athletic understanding of freedom (as requiring control) (§2), and will defend prioritizing such freedom in public policy (§3). Thus, I will argue that, understood as a guide to distributive policy, the capability approach's focus on individual freedom and control is justified: in the public domain it is important not just that individuals receive ‘benefits’ but that they participate in their achievement. The ‘athleticism’ of which Cohen accuses the capability approach is not an element capability theorists should aim to eliminate, but one they should celebrate.
2. GOALS OF THE CAPABILITY APPROACH: PUBLIC POLICY AND NON-INSTRUMENTAL FREEDOM
Sen developed the capability approach in response to perceived problems in development policy and so, in part, to problems in measurements of well-being and quality of life on which these policies were based. The capability approach, therefore, has been, and continues to be, used as a ‘standard of individual advantage’ or well-being (Olsaretti Reference Olsaretti2005: 90). However, my focus here is on the capability approach as a response to the ‘equality of what?’ or, as Olsaretti (Reference Olsaretti2005: 89) more accurately terms it, the ‘distribution of what?’ question. In this role, the capability approach functions as a guide to policy: to the distributive outcome that governments should promote.Footnote 3
Understood as a theory of distributive justice, concerned with guiding policy, and delineating the legitimate basis of government action, the approach will not provide a comprehensive account of all factors that contribute to our well-being. As Nussbaum (Reference Nussbaum2011: 32–3) puts it, its purpose is to establish the ‘task of government . . . [under] a decent political order’, and, if we reject welfarism, the ‘task of government’ is not merely the promotion of well-being. Thus, like Carter (Reference Carter2014: 76) ‘[t]he object of my investigation is the capability approach understood as a normative political theory . . . not merely as a non-political (or not-necessarily-political) theory of the quality of life . . . [but] as a theory that contains or entails certain political prescriptions’. It is as political prescriptions that capabilities, requiring freedom as control, are promoted, and in the political context that the focus on control is justified.
2.1 Cohen's Critique
Cohen approves of much of what the capability approach sets out to achieve, and commends Sen for his identification of a space between resources and utility in answer to the ‘equality of what?’ question. Cohen accepts capability theorists’ contention that resourcist approaches fail to take account of individuals’ different resource needs and conversion capacities (ability to ‘convert’ resources into achieved functionings). Consequently, he agrees that we should not consider what goods or resources individuals possess, but what goods do for people. He also agrees with capability theorists that the welfarist assessment of goods, in terms of the utility they generate, takes too narrow a view of what goods do for people. Given this, Cohen agrees that we should not adopt either of these standard approaches wholesale, and should focus instead on the (valuable) activities or states of being that goods enable us to achieve: the functionings they give us the capability to perform.
Cohen disagrees, however, when Sen calls this space, between resources and utility, ‘capabilities’, since this implies that what matters is what goods enable us to do for ourselves. Cohen (Reference Cohen, Nussbaum and Sen1993: 19) contends that the space Sen has identified is broader than the categories of either capabilities or functionings imply: ‘[w]hat goods do to people is identical neither with what people are able to do with them [their capabilities] nor with what they actually do with them [their achieved functionings]’. Thus, Cohen argues that we should not be concerned only with individuals’ capabilities.
Cohen points out that goods may also provide us with passively achieved benefits. For example, babies do not maintain themselves through the exercise of capabilities, but nonetheless get more from goods than just utility: they also experience (without participating in the achievement of) valuable functionings, such as nourishment or health. Moreover, it is not for the sake of their utility alone that we think we ought to provide them with these goods. Adults, too, can get benefits from goods without the exercise of capabilities: health from being in a malaria-free environment or from the rays of the sun, for example, or nourishment from a nutrient drip.
Cohen (Reference Cohen, Nussbaum and Sen1993: 18) calls the broader category of ‘what goods do to (or for) human beings’, with or without their participation, ‘midfare’. On this view, it is not always our capability to achieve functionings that matters, but sometimes merely their achievement. Midfare, then, incorporates three valuable categories:
(a) the substantive opportunity to perform a valuable functioning (capabilities);
(b) a (valuable) activity or state of being (functioning), achieved:
(i) through an individual's activity (exercised freedom to function); or
(ii) without their participation (passively achieved).
Cohen argues that this final category – passively achieved benefits – contributes to individual well-being, but that Sen ignores it, being concerned only with our freedom and its exercise, and that, as such, his account of well-being is excessively ‘athletic’.Footnote 4 Cohen does not dispute the value of freedom and, indeed, includes it as an important element of midfare.Footnote 5 His concern is that capability theorists wrongly insist that for a functioning to contribute to our well-being it is necessary that we are free to perform it, and that we achieve this functioning through an exercise of our freedom.
Cohen focuses his critique on Sen's elaboration of the capability approach, and in presenting his account I have maintained this focus. However, the same points could be raised against any approach that considers individuals entitled to capabilities, as substantive opportunities to perform functionings. I will not, therefore, defend a particular version of the capability approach – such as Sen's or Nussbaum's – nor discuss the nuances of these approaches. I will understand the capability approach, in broad terms, to propose that the justice of a distribution is measured in terms of individuals’ access to capabilities, and that redistribution should aim to ensure such access.Footnote 6 I aim to demonstrate that protecting and promoting the athletic freedom entailed by capabilities may be a plausible guide to distributive policy, and that Cohen's critique should not lead us to rule out capabilities as a metric of distributive justice (distinct from midfare).
2.2 Freedom and choice in the capability approach
I follow Cohen (Reference Cohen1995: 102) in accepting that ‘real freedom’, or freedom ‘worthy of the name’, is ‘the circumstance of genuine control over one's life’ (my emphasis). I therefore agree that cases where the world coincidentally conforms to one's will – Sen's (Reference Sen1992: 64–5) notion of ‘effective freedom’, or what Cohen (Reference Cohen1994: 120–5) calls a ‘weak’ reading of freedom – are not instances of freedom. I will take ‘freedom’ to mean that our choices (or preferences) must exert control in the world or determine an outcome: my choice will be satisfied because it was my choice.Footnote 7 Thus, in contrast to other responses – such as Sen's and Olsaretti’s, discussed below – my response to Cohen does not depend on insisting that he overemphasizes the ‘athleticism’ of the freedom that capability theorists promote. I agree, then, that freedom in the capability approach requires control. I disagree that this gives us reason to reject this approach to distributive justice.
It is worth emphasizing that given that my concern is with the capability approach to distributive justice, I need not make the strong claim that freedom per se must involve control. Rather, I suggest that freedom as it is understood and promoted (by capability theorists) as a political value, should involve control. Thus, in response to Cohen's (Reference Cohen1994: 124) contention that:
[t]here are two values associated with the successful exercise of freedom. One is that the world conforms to my will and the other is that it is I who achieve that result. Sometimes the second value does not matter much . . .
I would suggest that the second value matters a great deal in the political domain. Most liberal approaches aim to avoid the imposition of paternalist policies, where paternalism constitutes interference with an agent's autonomous choices, motivated by a distrust of their ability to make decisions in their best interests and an assumption that the intervening agent's judgement is superior.Footnote 8 Rather, liberals aim to respect individuals as agents capable of formulating goals (‘originators of ends’ (Taylor in Carter Reference Carter2014: 82)), and to allow them the freedom to pursue these ends, without the imposition of a perfectionist conception of the good. The capability approach is one such liberal theory, and though promoting capabilities may increase well-being, its goal is not to maximize well-being but to respect individuals as autonomous agents.Footnote 9
I will not provide an independent argument for promoting anti-paternalist public policy, or try to convince those unmoved by the value of individual autonomy, and the importance of a state allowing its free exercise.Footnote 10 Rather, I will show how the liberal values and anti-paternalist commitments in the capability approach can be plausibly realized. I believe that this requires that individuals have freedom as control over their lives (Cohen's ‘second value’, above). As Sen (Reference Sen1992: 65) notes, ‘it is often very hard, if not impossible, to have a system that gives each person all the levers of control over her own life’, however it is access to these ‘levers’ we should promote. I will expand on my understanding of freedom as control below (§2.3), before arguing that capability theorists’ focus on such ‘athletic’ freedom is justified, when devising distributive policy (§3). First, though, I will consider the problems with responding to Cohen by claiming that capability theorists’ focus on freedom is justified because freedom need not be understood athletically.
In his response to Cohen, Sen (Reference Sen, Nussbaum and Sen1993: 43) insists that ‘athleticism was never intended’, and Cohen was simply ‘misled’ by words such as ‘capability’ and ‘achieving’. Indeed, Sen's (Reference Sen, Nussbaum and Sen1993: 44) position as he presents it here seems very close to Cohen’s:
an active exercise of freedom might well be valuable for a person's quality of life and achieved well-being . . . [but] freedom has many aspects . . . and it would be a mistake to think of achievements only in terms of active choice by oneself.
Active choice is ‘an important component of living freely’ (Sen Reference Sen, Nussbaum and Sen1993: 44), then, but it is only one component, and freedom does not always need to involve such activity. In this way, Sen may avoid Cohen's charges of athleticism, but Cohen seems right to point out that it is rather counterintuitive to say that someone's freedom ‘is enhanced . . . when something he or she values occurs . . . even when the person had nothing to do with the occurrence’ (Sen in Crocker and Robeyns Reference Crocker, Robeyns and Morris2010: 77).
Rejecting Sen's interpretation in favour of freedom as control will require biting the bullet on a point that is considered a reductio ad absurdum of using athletic freedom in the capability approach: that we should have control both over which functionings we exercise, and the circumstances in which we choose.Footnote 11 Capabilities are, essentially, substantive opportunities, ‘created by a combination of personal abilities, and the political, social, and economic environment’ (Nussbaum Reference Nussbaum2011: 20 (my emphasis)). Sen (Reference Sen1999: xi–xii; Reference Sen2001: 54–6), too, emphasizes the contribution of external circumstances, policies and decisions to the provision of these opportunities. However, though Sen acknowledges the instrumental value of background conditions, he insists that having a capability does not depend on our ability to control these conditions (and so determine whether an opportunity is available to us). For example, Sen (Reference Sen2001: 54) does not distinguish between a disabled individual who has the capability ‘to go out of her house whenever she wants and to move around freely’ because she ‘is always helped by volunteers with goodwill’, and one who is enabled to do so by servants who ‘have to obey . . . her command’.Footnote 12
The freedom (and well-being) of the individual dependent on the goodwill of volunteers may indeed be increased, but this is not the freedom that a liberal, respect-based, anti-paternalist capability approach should promote. Individuals cannot freely form and pursue a conception of the good if they cannot be certain that they will continue to have access to central capabilities (the volunteers’ goodwill may cease), and are not respected as agents if they must court this goodwill to maintain such access. Thus, non-contingent, secure freedom should be the goal of liberal public policy. I believe that this requires both the first-order freedom to control whether we function, and the second-order freedom to control which functionings are available to us (§2.3). Opportunities alone do contribute to individual freedom, then, but the opportunity to determine the nature of these opportunities also contributes, and governments have reason to promote both (as §3.2 will argue).
Serena Olsaretti also responds to Cohen by insisting that freedom in the capability approach is not athletic. Her argument runs, briefly, as follows: the endorsement of valuable functionings is necessary for well-being; endorsement is best secured when people choose freely which functionings to achieve; and people are better placed to choose freely when they also have the freedom to forgo functionings (Olsaretti Reference Olsaretti2005: 98–100). I believe Olsaretti's response moves in the wrong direction in two ways. First, capabilities, on her view, have merely instrumental (and so contingent) value as a means to ensuring individuals endorse valuable functionings. Second, Olsaretti insists that providing these capabilities does not involve an athletic (or, in Olsaretti's (Reference Olsaretti2005: 100) words, ‘hyperactive’) notion of freedom. Endorsement is not understood in a strong sense, involving active choice or control, but is merely taken to mean that ‘the functioning is not forced on me’ (Olsaretti Reference Olsaretti2005: 100). This thin definition of endorsement may include instances when a person's will is bypassed, such as cases of brainwashing, indoctrination or hypnosis (Olsaretti Reference Olsaretti2005: 103–4).
Both these elements are incompatible with the central concerns of the capability approach to protect individual freedom, and avoid paternalism. For many capability theorists, including both Sen and Nussbaum, freedom has intrinsic, not merely instrumental, value.Footnote 13 Capabilities are not promoted as the best means to ensure all individuals perform valuable functionings, but because ‘a focus on dignity will dictate policy choices that protect and support agency’ (Nussbaum Reference Nussbaum2011: 30). Olsaretti's instrumental approach also means the justification for promoting capabilities depends on there being a stable causal connection between providing individuals with capabilities, and their endorsing valuable functionings. Given Olsaretti's weak definition of endorsement, it seems likely that this causal connection often will not hold. This makes the toleration of individuals pursuing various conceptions of the good, and a variety of functionings, rather tenuous. If a superior means were found to make people endorse the ‘right’ choice, and perform ‘valuable’ functionings (more effective brainwashing techniques?), such toleration would no longer be necessary.Footnote 14
Thus, Olsaretti's abandonment of an athletic understanding of freedom, in favour of a more minimal interpretation (as ‘not being forced’), and instrumental construal of the value of freedom, means individuals’ autonomous choices are only contingently, and so not securely, protected. Similarly, Sen's lack of concern with how opportunities are established does not securely protect capabilities, or individuals’ ability to form and pursue a conception of the good, free from dependency on the goodwill of others. In defending an approach, it is preferable to uphold its central motivating concerns, rather than abandoning them, and an unathletic conception of freedom is a poor base from which to pursue the liberal and anti-paternalist goals endorsed by many capability theorists. I will argue that it is possible to defend a version of the capability approach that considers athletic freedom (as control) as of intrinsic importance, and which can, therefore, avoid paternalism, and respect individuals’ autonomy and agency.
2.3 (Political) freedom as control
Yet two challenges remain. First, more needs to be said about what it means to have freedom as control. Second, it should be explained why what seems attractive in Cohen's criticism – that sometimes passively achieved benefits are worth pursuing, regardless of the affected individuals’ preferences – is not, in fact, compelling. I will begin my response to the first challenge here, outlining what it means for individuals to have control. The following section will continue this response, considering the sense in which individuals can be said to have control in ‘group’ cases (§3.2), as well as taking up the second challenge, arguing for the plausibility of the focus on such athletic freedom in the political domain (§3.1; §3.3).
Simply put, I understand freedom as control to require that our preferences or choices, made in a procedurally appropriate way, should determine outcomes independently of their content and context. Each of these three elements – appropriate procedure, content and context – require some elucidation. First, then, I suggest that our choices should be made under procedural conditions, such that ‘endorsement’ means something more substantive than Olsaretti's use of the term, according to which endorsement is mere assent in the absence of force. Instead, I mean something more akin to Nussbaum's practical reason, with its emphasis on being able to plan one's life, and ‘engage in critical reflection’ about that plan (Nussbaum Reference Nussbaum2000: 79). Thus, we should require that our choices meet something like Feinberg's (Reference Feinberg1986: 104–6) conditions of voluntary choice: they should be formed without coercion, with knowledge of the relevant empirical facts, in a clear emotional state, not based on mistaken reasoning, and be carefully considered.
Requiring procedural prerequisites for free choice fits with Pettit's (Reference Pettit2007) suggestion that freedom requires that individuals’ choices are decisive independently of their content and context. Content-independence means that an individual's choice should be decisive regardless of what they choose: for example, whether or not an individual chooses to perform a valuable functioning. This fits with my suggestion that choice should not (and in many iterations does not) have a merely instrumental role in the capability approach: capabilities are not merely a means to ensuring individuals make the ‘right’ choices. For an individual to have the capability for nourishment, then, they must be able to choose both to be nourished and to fast; to have the capability for health they must be able to choose to be healthy or unhealthy.
Context, favour or permit-independence requires that the decisiveness of our preferences, or our ability to control an outcome, should not depend on the ‘gratuitous favour’ of a third party.Footnote 15 Thus, we are not free if our preferences are decisive only insofar as we retain the favour of some other(s). This would rule out our possessing capabilities under a dictatorship, however benevolent. Pettit (Reference Pettit2007: 13–15) gives an example of a benevolent potentate who uses his wealth to improve the health-care and education systems of his country. As long as our access to these benefits depends on his favour, we do not possess the capability for health or education. Even if, for as long as his favour lasts, we can achieve these functionings, we lack these capabilities since we lack effective control over whether we are able to function.Footnote 16 The significance of permit-independence is particularly clear in collective action cases, as §3.2 will discuss.
I suggest that possessing a capability requires that these three conditions be met. This is certainly an athletic form of freedom, in Cohen's sense, but not, I will argue, implausibly so. It is worth reiterating that I am not claiming that this interpretation is true of every version of the capability approach (to distributive justice), though I do believe that it captures the motivating concerns of many prominent versions. Instead, I propose that for those concerned to avoid paternalism in the implementation of public policy, rather than the achievement of individual well-being by any means, this view will have some appeal.Footnote 17 Promoting athletic capabilities is a coherent goal, and a defensible response to the ‘equality of what?’ question, and does not merely signal a failure to acknowledge an important component of well-being.
3. IN DEFENCE OF ATHLETIC FREEDOM
Why is it so important that the government ensures we have control over our environment, rather than simply promoting our well-being, even if our own role in its achievement is a passive one? To answer, I will focus on the passively achieved benefits that most concern Cohen: ‘freedoms-from’ environmental obstacles, in particular the freedom from diseases such as malaria. Freedom from malaria seems like a prime example of something the government ought to promote regardless of our choices or preferences, and even if we do not play an active role in achieving this benefit. Yet, Cohen objects, capability theorists cannot acknowledge the value of such passively achieved benefits. Moreover, the ‘freedom-from’ such a harm or burden removes an available option or limits our capabilities (to be subject to that harm), which may appear contrary to the aims of the capability approach.
I will argue that valuing the freedom-from some environmental obstacles is compatible with the capability approach, but that they should not be unilaterally promoted by governments (§3.1). Instead, in cases where a group is necessarily affected by a policy, the group should decide whether the policy – such as the eradication of malaria – will be pursued (§3.2). In such collective action cases, individuals simply cannot have direct control, since this would undermine others’ freedom. Thus, having freedom as control may merely require participation in a democratic system: the possession of ‘collective capabilities’. These collective capabilities ensure that individual agency is respected, and paternalism is avoided, in cases where individual capabilities cannot be provided (§3.3). I will, therefore, demonstrate that Cohen is wrong to suggest that capability theorists must choose between a plausible reading of freedom (as control), and properly conceptualizing well-being.
3.1 Freedoms-from and capabilities
To undermine the plausibility of capability theorists’ concentration on freedom, Cohen considers cases where benefits are achieved passively. In particular, when we are benefited by being made free from a risk or obstacle ‘that impede[s our] . . . achievement of valuable functionings in choice-insensitive ways’ (Olsaretti Reference Olsaretti2005: 94). For example:
a malaria-ridden environment . . . render[s] the functioning of being . . . disease-free highly difficult or impossible, and . . . [does] so, typically, in choice-insensitive ways, in the sense that someone exposed to these factors is likely to come to lack the relevant functioning through no choice of her own. (Olsaretti Reference Olsaretti2005: 94)
Cohen insists that since capability theorists aim to provide individuals with freedom as control they cannot promote such freedoms-from obstacles (since this would provide functionings rather than the capability to function). Whilst they may enhance our freedom, they are not freedoms in themselves, and so not the direct concern of the capability approach. As Olsaretti (Reference Olsaretti2005: 93) presents Cohen's view:
‘Freedom from malaria’ only counts as a capability on an unduly expansive sense of freedom, one on which someone's freedom is enhanced when something happens to her or her environment, even though she has not chosen that thing and has no control over whether that thing will be chosen.
Whilst Olsaretti's (Reference Olsaretti2005: 95–6) (and Sen's Reference Sen, Nussbaum and Sen1993, Reference Sen2001) response is to defend this ‘expansive’ and unathletic view of freedom, I follow Cohen in agreeing that freedom does require control, and that ‘freedom from malaria’ cannot plausibly be understood as a capability.Footnote 18 Nonetheless, the response capability theorists can make initially seems remarkably straightforward: they need not consider freedoms-from to be capabilities in order to promote them.
The capability approach aims to provide freedom in the sense of substantive opportunities, which include the physical and psychological conditions of making a choice. This seems likely to include the removal of obstacles that prevent people from making choices: we cannot have a secure capability for good health whilst living in a malarial environment, for example. As well as securing new opportunities, freedoms-from may also improve our ability to choose between available options. For example, being free from starvation may allow us to choose to leave a job we hate, if the consequences of doing so are no longer so dire (starvation). Further, the decision to exercise some functionings, once free from the obstacles to performing them, may also increase our ability to choose. For example, being nourished may improve our brain functioning, as well as removing the distraction of perpetual hunger, and so give us a greater capacity for reasoning and decision-making than if we were malnourished.Footnote 19
Thus, capability theorists have many reasons to value and, indeed, promote freedoms-from as a means to promoting capabilities, without having to concede that freedoms-from are, themselves, capabilities. Given that many policies that enhance some capabilities also restrict others, an instrumental argument of this sort might require us to make trade-offs between different capabilities. In some cases, this will only mean the loss of opportunities widely considered disvaluable (for example, contracting malaria) for the sake of valuable, or central, capabilities (the opportunity for good health). For most capability theorists (who rely on a list of specified valuable functionings) this would barely be considered a trade-off at all, since nothing they value is lost. Even where the lost capability is one they consider valuable, giving it up may be justified where another valuable capability can be achieved by doing so. This is a familiar enough phenomenon: people choose to get sterilized (to be ‘free-from fertility’), and thus lose the opportunity to control whether they have children, in order to have other options. The control over our lives that a capability gives us does not require that this control be always maintained.
The capability approach, then, does not proscribe decisions that limit our future option sets, or future opportunities to control our lives. It does, however, proscribe such limitation of options without the consultation of the affected individual. Sterilization is acceptable if chosen by the person to be sterilized, but unacceptable if it were the result of a government policy concerning population control. Responding to Cohen's cases is complicated by the fact that these tend to concern public, non-excludable goods, wherein making us free from an obstacle will limit the future options of many people. I will suggest (in §3.2) that individuals should have control over their shared environment, as well as control over the more specific functionings they perform. Further, that when a group is affected by these environmental changes, it is the affected group who should exercise control (since individual control is impossible).
Before analysing freedom as control in collective cases, however, it is worth considering how capability theorists not committed to providing capabilities for specific valuable functionings should respond to Cohen's critique. On such a view, though we identify specific valuable capabilities (for example, for health), these are interpreted as domains of control, rather than the ability to perform a specific functioning (being healthy). Thus, this view would not distinguish ‘valuable’ functionings – such as controlling our reproduction – from ‘disvaluable’ functionings – such as contracting malaria. We should have control over all aspects of our health. It would, therefore, be problematic if individuals were forcibly denied the opportunity to contract malaria, just as it would be if they were forcibly denied the opportunity to have children. In both cases, we should be able to control both whether we have these opportunities, and the use we make of them.
I believe this approach captures the capability approach's underlying liberal goals: that individuals be enabled to autonomously form and pursue their own conception of the good, and that respect for human dignity requires that we allow individuals to exercise their agency. It also better encapsulates Pettit's suggestion that freedom be content-independent, and provides a response to Carter's claim that an identifiable capability approach cannot provide content-independent freedom, and so avoid paternalism.Footnote 20 However, whether or not this approach seems plausible,Footnote 21 it is worth considering since it is especially vulnerable to Cohen's criticisms, given that it requires that individuals be in control in a greater range of cases (even when all that is lost is an apparently disvaluable opportunity). To defend even this ‘excessive’ athleticism seems the best way to show that Cohen's concerns about athleticism are misplaced.
3.2 Collective capabilities and democracy
I have suggested that the capability approach should provide individuals with freedom, in the sense of the ability to exercise control over central domains of their life, not simply to ensure that we get what we want in these domains (or what some capability theorists consider good for people), but to ensure we are treated in an appropriate (non-paternalistic) way: our agency is respected. Not getting what we most want may be detrimental to our well-being, but it need not be paternalist. Enforced sterilization, then, is not just problematic because some individuals lose a valued opportunity (to have children), which may decrease their well-being. More importantly, overriding individuals’ preferences insults their agency in a characteristically paternalist way, by distrusting their ability to make decisions in their own best interests. To reiterate, then, a capability approach to distributive justice is not designed to ensure that individual well-being is maximized, but to allow them freedom as control over their lives. I will now consider what it means to have such freedom in collective cases.
My proposal is that just as an individual should not be denied an opportunity without consultation, neither should a group. As discussed, capability theorists do not prohibit individuals from sacrificing opportunities, but they must autonomously choose this sacrifice.Footnote 22 Hence, just as sterilization should not be forced on an individual, so public health policies should not be unilaterally imposed on a group: an individual should not be forcibly made ‘free-from fertility’, and a group should not have their environment forcibly made ‘free-from malaria’. If a concern for losing disvaluable options seems implausible, we can consider cases in which valuable opportunities are sacrificed. For example, the industrialization of a landscape that provides much-needed jobs only at the expense of preventing individuals from ‘being able to live with concern for and in relation to animals, plants, and the world of nature’.Footnote 23 Here, the alteration of the environment, creating the background conditions for some members of a group to have central capabilities, will also lead to (at least some) others losing a central capability.
This loss of capability is legitimate only if the group controls whether the policy is implemented, where ‘control’ will mean all individuals’ preferences are counted equally in the democratic process. (Though protection of individual capabilities will mean such democratic control is not appropriate in every case (as §3.3 will discuss)). Indeed, I would suggest that groups should have control even when the capability lost is disvaluable: there should be democratic control over the introduction of both anti-malarial, and urban planning, programmes. This is because, for capability theorists, the relevant problem with the undemocratic introduction of such policies is not that some people lose a desired (or ‘valuable’) opportunity, but that they are not respected as agents capable of choice.
To elucidate this idea of collective control, it is useful to further consider the role of permit-independence. I have argued (contrary to Sen) that, when used as a guide to policy, the capability approach should incorporate a concern for permit-independent freedom, as Pettit suggests. However, I suggest that it is not enough to focus, as Pettit does, only on first-order, and not second-order, freedom as control. To illustrate this point, imagine two benevolent potentates. The first sets up a social infrastructure, such as a system of health care and education, but once set in motion, he no longer exerts control over its running, and so access to it is not dependent on his whims. Imagine, for example, that he formally relinquishes control over its operation to an independent body over which he has no influence. The second is as Pettit describes: he continues to control who has access to health and education (even if this power is never exercised).
On Pettit's view we are free when we are not being dominated, and are not liable to domination, so Pettit would consider the individuals in the first society free, since the potentate lacks the power to dominate them. Their capabilities for health and education are permit-independent: not reliant on the favour of a third-party. Yet, whilst they do have the individual capabilities for health and education, they lack control over the establishment of these conditions, and so lack freedom in an important sense. Just as it would not be unsatisfactory for a disabled person's capability for mobility to be dependent on the goodwill of others, so would it be unsatisfactory for a population's access to a health service to depend on the benevolence of their dictatorial government. Hence my suggestion that when a group is affected by a policy, it is the group who should control whether this policy is implemented.
Whilst citizens of a dictatorship lack such control, as citizens of a democratic state our preferences can be said to be sufficiently decisive. As Pettit (Reference Pettit2007: 18) notes, ‘[a] democratic government is passively forced to respect what are assumed to be my preferences or the preferences that we in a certain group – perhaps the citizenry as a whole – share’. Thus, even if citizens do not have direct control over policy, their preferences as a group determine the shape policy takes (or should do).Footnote 24 Freedom as control therefore requires that when only an individual is affected, their preference alone should control the outcome; and when a group is affected, the group should have control. This is largely for the simple reason that it would be impossible to allow any other sort of control in group cases: to allow one individual's choices or preferences to be decisive would remove control entirely from other individuals. The ability of groups to exert control over policy implementation will be called collective capabilities.Footnote 25
A natural worry is that such collective freedom to control government policy gives each individual very little ‘real’ control over determining what the outcome will be. However, as discussed, the value of control is not to ensure we get the outcome most conducive to our well-being (or the outcome we would, individually, prefer), but that we are treated appropriately (respected). Remember that an act is paternalist if it is motivated by a distrust of individuals’ choices, and disrespect of their agency. To be in the minority in a democratic vote is not to be subject to a paternalistic insult when the government acts on the majority decision. In contrast, when the government acts on what it considers the good of the majority, without taking account of anyone's decisions, all citizens are insulted, even those who would have supported the policy if they were given the chance.Footnote 26 Avoiding paternalism, and ensuring individuals have freedom as control, does not require that no individuals’ desires are ever frustrated, but that their preferences are respected. I contend that individuals’ preferences are respected when they are all given equal weight and value.Footnote 27
3.3 The limits of collective decision-making
However, such collective decision-making is clearly not appropriate (or appropriately respectful) in every case. I will, therefore, provide a sketch of the kinds of decisions that should be made collectively. Collective decisions establish background environmental conditions, which determine individuals’ option sets – whether a health service, a disease, or a job is available – and individuals retain the individual capability to determine which of the available options they utilize. There will always be a tension, then, between individual and collective capabilities, so it is important to consider the extent of the restrictions a group can impose on its members.
It may be helpful to begin with an example of a capability with both collective and individual elements, such as health. There are numerous health policies that affect all individuals, such as the availability of certain drugs or the establishment of a public health service. In many such cases, collective capabilities are appropriate, and decisions should be made by all individuals, counted as equals. Consequently, not all individuals will get their ideal option set: for example, libertarians may be compelled to contribute to a public health service. However, though individuals are not guaranteed their preferred option, freedom as control requires that individuals are, as far as possible, provided with a range of options.Footnote 28
For example, compare two vaccinations against serious, but rarely life-threatening, diseases; both of which require a high proportion of the population to be immunized to be effective. In the first case, the vaccine has no side-effects, whilst in the second it will cause infertility in a reasonably large number of cases (say, half). A collective decision to adopt a compulsory programme of vaccinations in the first case seems permissible. Although it will deprive individuals of the capability to contract the particular disease, and slightly lessens their control over their health care, they still retain general control over their health. Compulsory immunization is unlikely to be permissible in the second case, however, since this completely removes some individuals’ control over a central domain (their ‘choice in matters of reproduction’; Nussbaum Reference Nussbaum2000: 78).
There will, of course, be many difficult cases: if the disease was life-threatening, could we compel individuals to have the second vaccine?Footnote 29 How high would the risk of becoming infertile have to be before forcing us to have the vaccine is impermissible? Could we expect those with certain religious views, such as Jehovah's Witnesses, to receive even the first vaccine? More generally, then, who constitutes the relevant collective for a collective capability, and what decisions can they make for other members of that collective?
Drawing strict lines here will be difficult, but the intention is that: (a) those who are affected should constitute the relevant group; (b) decisions should be made by a collective only in cases when a group is necessarily affected; and (c) an effort should be made for individuals to retain a space to exercise control over central domains of their lives. These criteria limit the scope of collective capabilities, and prevent a majority decision from unacceptably limiting the freedom of the minority. For example, a policy of forced sterilization would be illegitimate because: individuals outside the affected group would exert control over those within it (violating (a)); it is unlikely that this is a decision that must be placed in the hands of a group (violating (b)); and it clearly and severely limits the capacity to choose of the individuals affected (violating (c)).
Defending the capability approach does not require that all these complex issues be resolved. It is enough to outline how the protection and promotion of ‘athletic’ freedom in public policy may be plausible, and show that this does not commit capability theorists to a similarly ‘athletic’ reading of well-being. I suggest, then, that capability theorists should provide individuals with collective control over the external conditions that affect groups of which they are a member, and individual control over which specific functionings they perform, given these external conditions. For example, a group may decide that a drug is made available, but individuals decide whether they want to take it; a group decides whether a public health service should exist, and an individual decides whether they use it.Footnote 30 This may not always maximize individual well-being, but for capability theorists this is not the goal of distributive policy. It does respect individual agency and autonomy, and allow individuals the athletic freedom to control the environment in which they live, the options they face, and the use they make of them. Such ‘athletic policies’ should not be avoided.Footnote 31
4. CONCLUSION
Cohen insists that the capability approach must choose between an implausibly weak definition of freedom (according to which we are ‘free’ even when we have no capacity for control), or an implausibly athletic account of well-being (wherein something is only good for us if we achieve it for ourselves). This is a false dichotomy. When the capability approach is understood as a guide to just distributive policy, it can adopt a plausible understanding of freedom (as control) without denying that when the world coincidentally conforms to our will this may improve our well-being. Public policy should not aim simply to maximize individual well-being, however, but should aim to protect and promote individual autonomy, treat individuals with appropriate respect, and avoid paternalism. This, anyway, is the goal of the capability approach, and why many find it a convincing answer to the ‘equality of what?’ question. Cohen may object that our answer to this question should give greater prominence to well-being, but he has not shown that the capability approach is internally incoherent. For those who consider the promotion of autonomy a more central concern of justice than always protecting well-being, then, the approach will remain appealing. Thus, to the suggestion that their understanding of freedom is excessively athletic, capability theorists should respond that it is proudly so: government policy should aim to provide people with athletic freedom as control over decisions in central parts of their lives.
ACKNOWLEDGEMENTS
For extremely valuable comments, suggestions and discussion on various drafts of this paper, I would like to thank Chris Bennett, Carl Fox, Martin O’Neill, Mozaffar Qizilbash, Daniel Viehoff, and the referees and editors at Economics and Philosophy (in particular, Ian Carter and Fabienne Peter), as well as audiences at the MANCEPT Workshop on ‘Well-Being and Public Policy’ at the University of Manchester in September 2012, the Pavia Graduate Conference in Political Philosophy at the University of Pavia in September 2012, and the Brave New World Postgraduate Conference at the University of Manchester in June 2012.
