Section I: the concept of legitimacy

Despite extensively discussing accounts of legitimacy, bioethicists have offered little analysis of the concept of legitimacy. This section partially clarifies the concept of legitimacy used in bioethics by extrapolation from the political legitimacy literature. Two reasons make such extrapolation appropriate: the political literature offers the most developed discussion of legitimacy, and bioethicists routinely draw on it to defend their views.

Accounts of political legitimacy describe two different types of concepts.^iv Accounts of descriptive legitimacy, frequently indebted to Weber,21 construe legitimacy as a construct of citizens’ attitudes about (or dispositions to support) governments, policies, laws or other governmental actions. Such a concept is important in various respects, such as in assessing citizens’ tolerance of regimes or policies.

In contrast, accounts of normative legitimacy describe a normative property, usually understood as a property of states, regimes, laws, policies or government actions. There are several types of such accounts, which explicate different concepts of normative legitimacy, but two are most prominent. On accounts of authoritative legitimacy, the fact that a government (or the like) is a legitimate authority entails a right of that government to create duties. It might be a claim right of governments over its subjects that they comply with its legitimate directives22 or a power of governments, corresponding to a liability of those subject to it, to create rights or duties for the subjects.23 24 In contrast, on accounts of coercive legitimacy, the fact that government (or the like) is a legitimate coercer entails a (more limited) privilege to exercise coercive force to enforce (some of) its laws.25 ^v,vi

Generally, bioethicists are interested in a normative concept. For instance, Daniels believes that moral controversy about resource allocation creates ‘a legitimacy problem’, which he glosses as a question: ‘Under what conditions do decision makers have the moral authority to set the limits they impose?’ (Daniels, p. 103, his emphasis).26 Such emphasis on moral authority suggests that Daniels is employing a normative concept of legitimacy, as does the fact that his account of legitimacy appeals to properties of the decision-making procedure, rather than features of stakeholders’ psychology after an outcome.

Which normative concept? It seems to be one roughly like that of authoritative legitimacy. First, such a reading makes Daniels’s question both interesting and relevant. While all people are welcome to reflect on the morality of resource allocation, not everyone has the power to impose duties on others based on their decisions. If some entity possesses authoritative legitimacy, it has such a power.

Second, reading the question as about coercive legitimacy will prove problematic to much bioethical usage of the term legitimacy. Bioethicists generally think that not only governments, but a variety of entities might hold legitimacy and make legitimate policy, including hospitals and insurance companies. Yet, generally, these non-governmental institutions do not have a privilege to coerce their clients. If they have claims meriting coercion against their clients, they must appeal to the state to exercise its coercive privilege—for instance, through the court system via contract law.

How shall we understand authoritative legitimacy in bioethics so that such entities might be legitimate? I suggest it refers to a property of policies, such that if a policy is legitimate, the institution associated with it has a right to impose pro tanto duties on its subjects to comply with the verdicts of its decision-making process. More formally, if policy P of institution I is legitimate,^vii then I can impose duties on those subject to its policies to comply with P. While this suggestion offers only a sufficient condition, rather than a full analysis, it illuminates questions like Daniels’s and will be sufficient to address the orthodox view. A full analysis is important but would take us further afield.

Section II: Orthodoxy

With this clarification of the concept of legitimacy, we can consider the orthodox view of legitimacy. To understand the account, consider two of its most prominent representatives: Daniels and Sabin. They believe that moral theories are too coarse-grained to determine just allocation policies—especially in the face of disagreement. Instead, they think an actual procedure must settle the question of what policy is legitimate. On their view, a procedure will produce legitimate policies only if ‘decisions … and their rationales [are] publicly accessible’, and these rationales are reasonable. ‘Rationale[s] … [are] reasonable if [and only if they] appeal[] to evidence, reasons, and principles that are accepted as relevant by fair-minded people’.^viii ‘Fair-minded’ people are those ‘who in principle seek to cooperate with others on terms they can justify to each other’ (Daniels, p. 118–119, Daniels, p. 44–45).18 26 Hence, ‘fair-mindedness’ is roughly Rawlsian ‘reasonableness’ (Rawls, p. 49).27 ^ix

Daniels and Sabin offer two further conditions, such that each of their four conditions are necessary and they are together jointly sufficient. The Revision and Appeals Condition requires mechanisms for ‘challenge and dispute resolution … of policies’, and the Regulative Condition requires some process to ensure that the other three conditions are met (Daniels, p. 118–119, 131–133; Daniels, p. 45–46, 58–65).18 26 ^x However, I will make a simplifying assumption that these conditions are orthogonal to my concerns about the orthodox view. I will defend this claim at the end of section IV.

Abstracting from this representative account, we can formulate Orthodoxy.

Orthodoxy: Policy P is legitimate if and only if Institution I produces P via actual democratic deliberation that employs only public reasons, which I must publicise.^xi

Two terms require clarification. First, to publicise reasons is roughly to promulgate them to stakeholders--whether or not they were part of the procedure in question. Second, public reasons, influentially described by Rawls and his students, are reasons that would be acceptable (in some sense) to (possibly idealised) reasonable persons in some (possibly idealised) deliberative scenario.27 Yet, unlike Rawls, bioethicists are interested in actual (rather than hypothetical) deliberation and actual (rather than hypothetical) persons.^xii,xiii

This implies that orthodox theorists owe us some understanding of public reasons that does not rely on idealization. I will follow Daniels and Sabin, who understand such reasons as those "accepted as relevant by [reasonable] people," but I believe that similar problems to those below will arise on all understandings open to orthodox theorists (Daniels, p. 118, Daniels and Sabin, p. 45).18 26

The relationship between Rawls’s view and Orthodoxy is important. Proponents of the orthodox view (henceforth, orthodox theorists) acknowledge debt to and, frequently, invoke Rawlsian views of legitimacy. Sometimes they suggest that their view is an application of a Rawlsian view of legitimacy to the case of healthcare.16 However, their views are susceptible to the challenges below in ways that Rawls’s are not because of these critical—though perhaps often unnoticed—differences.

Rawls’s view that legitimacy requires appeals to public reasons emerges in his liberal principle of legitimacy. Roughly, it holds that political power is legitimate only when ’it is exercised in accordance with a constitution the essentials of which… citizens may reasonably be expected to endorse.' He further endorses a related principle that legislative questions concerning ‘constitutional essentials, or basic questions of justice, should be settled, so far as possible, by principles and ideals that can be similarly endorsed’^xiv (Rawls, p. 137).27 To expound on this, he offers a framework of deliberative democracy in his theory of public reason.

While orthodox theorists claim support from this view, Rawls takes these principles as ones of ideal theory while orthodox theorists are concerned with non-ideal theory.16 For Rawls, ideal theory proceeds within a context of a well-ordered society under conditions of full compliance.28 29 His goal is not to apply the theory of legitimacy immediately to contemporary societies in our actual world, but rather to argue that his theory of justice could stably govern a well-ordered society despite reasonable pluralism about the good.27 Yet, bioethicists generally aim to offer non-ideal theories in light of their utility for affecting our actual world, and so, the counterexamples I will offer focus on those conditions.^xv

Section III: challenging Orthodoxy

We may now address the problem that cases like the ACA and HSA raise for Orthodoxy. As a first approximation of the problem, intuitively, the ACA is legitimate in the sense we have specified in section I. So too, I think, Clinton’s HSA would have been. Yet, Orthodoxy would seem to deem them illegitimate. More generally, Orthodoxy is incompatible with the legitimacy of any reform proposal made in ways like these—no matter what virtues it may have.

To clarify the problem, we should abstract away from contingent issues about healthcare reform in the contemporary USA that might raise questions about whether certain events were publicised or about the likely effects of particular reform proposals. Hence, consider Defective Reform Environment.

Defective Reform Environment: Ration-topia is like the (pre-Affordable Care Act) USA in regard to affordability and access to health as well as economic conditions. Its citizens are averse to ‘rationing,’ defunding other important goals and increasing taxes, but nevertheless generally avow a right to healthcare.

A new president and congress perceive significant support for reforming healthcare. Yet, given recent history, political sensibility, and polling, the new officials have reason to believe that if they are transparent about their rationale behind healthcare reform proposals, reform will fail. Further, they have reason to believe that this failure will damage the possibilities of reform for the foreseeable future, leading to healthcare rights violations. Hence, they develop reform behind closed doors and publish the results, but not rationales, of their decisions. They frame such results in terms of offering ‘better healthcare,’ offering no other reasons and consciously denying any suggestion that the reform licenses ‘rationing’ or ‘tradeoffs’ in resource allocation.

Given that significant rights violations will continue without reform, the failure to enact such a policy seems vastly unjust; in turn, assuming the policy in question is a good means of securing healthcare rights, intuitively, it is legitimate. The degree of injustice here^xvi is important. The conditions of justice and legitimacy are likely to be distinct, but a law or policy cannot be legitimate if it is vastly unjust.^xvii If all of this is right, Orthodoxy appears to entail that this reform is illegitimate because the administration publicised the results but not the reasons that they employed. Hence, Orthodoxy is false; actual democratic deliberation that employs only public reasons that are publicised is not necessary to the legitimacy of bioethical policies.

Of course, readers may not share this intuition. So, consider the following argument. If this policy is the best way to avoid such violations, then its citizens appear to have a duty to comply with the policy, even if the rationale is not publicised. Given our analysis of bioethical legitimacy from section I, a compelling explanation is that the policy is legitimate.

What other explanations of this duty might be offered? An appeal to consent is implausible because no consent tokens occur in the case. One might claim that the duty follows from general political obligations rather than the legitimacy of the policy. However, we can imagine a similar case in which a private insurance company enacted a new coverage policy aimed at bringing about more just allocations (say, by enabling them to cover more of the worst off at cheaper rates while charging more to the economically advantaged). Intuitively, the verdict would be similar. Of course, in such a case, individuals would usually have contractual obligations with their insurance companies. Hence, one might claim that it is not the legitimacy of these policies, but the contract that explains the duties. Yet, surely, contracts to illegitimate policies are invalid and so could not explain the duties in question unless the policy is legitimate (in which case Orthodoxy is false).

Hence, appeals to legitimacy in such cases explain duties to comply with the relevant policies even when these policies do not result from the procedures mandated by Orthodoxy. In reply to this argument, orthodox theorists might give three responses: (1) that some revision to Orthodoxy salvages their core commitments, (2) that their views of publicity and extra-procedural considerations do not actually have these implications and (3) that their arguments for Orthodoxy are strong enough to support these counterintuitive results. I reply to these in the next three sections, respectively.

Section IV: other orthodox views

Can orthodox theorists salvage their core commitments with revisions to Orthodoxy? They might claim that they are actually concerned with the more general commitment to some important role for actual procedures in which only public reasons are offered in the explanation of legitimacy in the context of healthcare resource allocation. However, they owe us some alternative condition that would explain this role. In this section, I consider three alternatives, arguing that each fails.^xviii

First, orthodox theorists might accept that the problem is that publicising public reasons cannot be necessary because policy-making sometimes demands secrecy; nevertheless, they might suggest that it is sufficient. Consider Orthodox Sufficiency.

Orthodox Sufficiency: Policy P is legitimate if Institution I produces P via actual democratic deliberation that employs only public reasons, which I must publicise.

Yet, No Mental Health appears to show that Orthodox Sufficiency is problematic.

No Mental Health: Given resource constraints and needs to fund other forms of healthcare, the US Congress decides that Medicaid will no longer cover mental health services. All proponents in Congress want to justify this policy to each other and to those whom it will affect in terms that they can accept. Hence, they discuss the evidence in favour of various mental health treatments, but also evidence that the decision enables other benefits, which other fair-minded people should see as important.

Recall that Daniels and Sabin’s understanding of public reasons is such that they are reasons that are accepted as relevant by some reasonable people in the deliberative democratic scenario. The Congress members are reasonable—in the relevant sense—in No Mental Health and offer reasons that such people would see as relevant even if not decisive. Hence, the policy in No Mental Health satisfies the antecedent of Orthodox Sufficiency.

Yet, the policy in No Mental Health is not legitimate. Consider attempts to resist compliance with this policy. Suppose Dr Jones is double-certified as an internist–psychiatrist. One of her patients, Robert, has schizophrenia, but no longer has coverage for psychiatric treatment in light of the new policy. Dr Jones slyly undermines the policy by seeing Robert for ‘internal medicine appointments’, performing a routine medical exam each time, during which she asks many questions about his schizophrenic symptoms as if she were doing a psychiatric exam. She also prescribes an atypical antipsychotic (off-label) ‘for sleep’, with which she is treating Robert’s schizophrenia. If the policy in No Mental Health is legitimate, Dr Jones appears to be flouting her duty. Yet, in contrast, she appears to be acting permissibly—perhaps even heroically.^xx

A second alternative is suggested by orthodox theorists’ claims about two scenarios in which the requirement that rationales be publicised does not apply. Perhaps my examples involve such circumstances. First, some allow that, because secrecy is sometimes important for realising public policy, secret mechanisms of forming policies are unproblematic if they are legitimated by higher-order public deliberation about the cases in which secrecy is permissible, deliberation that should accord with Orthodoxy.30 However, this condition is irrelevant to the cases above, in which there is no mention of the public previously endorsing such secrecy.

Second, Gutmann and Thompson distinguish between ‘deliberative disagreement’, in which parties are ‘committed to finding fair terms of cooperation’ and mutually acceptable justifications, and ‘nondeliberative disagreement’, in which parties are not so committed. Perhaps Orthodoxy only applies to the former (Gutmann and Thompson, p. 78–79).30 Yet, cases like No Mental Health appear to show that Orthodoxy produces the wrong verdicts in deliberative disagreement. Moreover, part of the lesson from the ACA is how prevalent nondeliberative mechanisms are. Given that lesson, the theory in question will not apply to much decision-making in contemporary democracies, rendering it much less practically relevant than many bioethicists might have thought.

Third, and finally, we should reconsider whether Daniels and Sabin’s Revision and Appeals Condition and Regulative Condition^xxi help them avoid these problems. To see the irrelevance of the first, let us stipulate that the government meets such conditions in Defective Reform Environment and No Mental Health. It seems that nothing about the verdicts changes. Likewise, if we imagine that, seeing the problem of the policy in No Mental Health, Congress subsequently corrects it, Orthodoxy would still be forced to deem the policy illegitimate prior to its correction. Regarding the second, note that no addition of any regulative mechanism changes the verdicts in these cases.

Section V: publicity and extra-procedural considerations

One key tension for orthodox theorists lies between their appeal to public reasons and the role for extra-procedural moral considerations that cases like the above suggest are critical for a theory of legitimacy. Orthodoxy denies any role for considerations outside of its procedure. Orthodox theorists might suggest two ways of navigating this tension to avoid my arguments. First, they might argue that they are not committed to a view of publicity with the implications that I have suggested because, like other public reasons theorists, they allow for idealised rather than actual persons’ views to determine the content of public reason. Second, they might limit the role for public reasons.

Regarding the idealising response, while bioethicists appear to understand public reason in terms of actual people’s views, perhaps orthodox theorists could appeal to idealised people’s views. Yet, justifying idealisation of the persons to whom legitimacy purportedly requires justification is difficult for orthodox theorists. Similar ideal theories idealise individuals’ beliefs about either normative or descriptive matters.31 Yet, it is unclear how orthodox theorists can do so, especially when focused on non-ideal theory.

An idealisation regarding beliefs about descriptive matters might seem acceptable, yet no reason has been offered to suggest that merely improving deliberators’ non-normative, epistemic resources would guarantee that they would take an absence of mental health coverage—or any healthcare coverage—to undermine the legitimacy of a potential policy. In contrast, normative idealisation that would require that deliberators believe that justice demands coverage appears to lie in tension with the reasons that bioethicists have given for their orthodox views. If moral theorising is inadequate to settle the above policy considerations about healthcare, why would it not be inadequate to settle the question of what beliefs are appropriate to require of reasonable deliberators?^xxii

Therefore, consider the second suggestion—that they limit the role of public reason. The general problems for this strategy are (1) finding a principle that sets this limit while maintaining that deliberation is required after the principle is satisfied and (2) doing so while maintaining orthodox theorists’ claim that moral theory is too coarse-grained to settle questions of policy. In the rest of this section, I will illustrate these challenges by investigating an intuitively promising principle and how the challenge arises for it.

We can motivate this intuitively promising principle by noting that the examples above involve massive healthcare reform to alleviate significant rights violations with vast consequences. Perhaps then Orthodoxy makes sense only in cases which a basic right to healthcare has been satisfied. If so, maybe it can be salvaged with some proviso requiring such care. Consider Basic Coverage Orthodoxy.

Basic Coverage Orthodoxy: Assuming that (most) citizens have (some degree of) basic healthcare provisions, then Policy P is legitimate if and only if Institution I produces P via actual democratic deliberation that employs only public reasons, which I must publicise.

First, while attractive, the principle itself encounters difficulties in that it is only compelling on some interpretations of basic health provisions. Consider an interpretation on which that provision does not require mental health coverage. On such an interpretation, the policy that Congress attempted to develop in No Mental Health would have been legitimate according to Basic Coverage Orthodoxy, but it is clearly illegitimate. Hence, orthodox procedures do not suffice to make the resultant policies legitimate, even if basic coverage is ensured, on such an interpretation of basic coverage.

Moreover, cases analogous to Defective Reform Environment show that they are not necessary as well. Suppose that, facing the possibility of the state of affairs described in No Mental Health, government agents (eg, members of the Centers for Medicare and Medicaid Services (CMS)) found ways to undermine the law and replace it with another policy that provided reasonably good coverage. Suppose they do so without publicising their rationale because publicising it would allow those Congress members that opposed mental health coverage to sabotage their efforts. The resultant policy could be legitimate just as the new administration’s reform plan was in Defective Reform Environment. Whether or not it would be would depend in part on the content of the policy. Hence, the plausibility of Basic Coverage Orthodoxy depends on what provisions are ensured by the relevant services and whether they sufficiently realise, protect or enable the realisation of basic requirements of justice. These include mental health services, but similar cases could be generated to test other requirements that may be necessary (eg, women’s health services).

We might question which particular services would be required (eg, what sorts of mental healthcare). However, in regard to our discussion of No Mental Health,^xxiii the fact that the procedure met the demands of orthodox theorists, but that our intuitions still found the results illegitimate, shows that extra-procedural moral considerations must be invoked to defend them. This point brings us to the second challenge for orthodox theorists: defending such a principle without betraying their claim that moral theory is too coarse-grained to settle policy considerations. To maintain their appeal to this claim in motiving their views without sacrificing their core procedural commitments, they must specify a scope in which extra-procedural considerations play a role but beyond which their arguments for the indeterminacy of moral considerations remain intact.

The chief resource on prominent orthodox theories appears to be an appeal to fair equality of opportunity.16 26 30 32 Orthodox theorists frequently claim that justice requires fair equality of opportunity, which is promoted by fair distributions of health, which is, in turn, promoted by healthcare provision. Yet, they claim that beyond the general requirement for healthcare provision, theories of justice are too coarse-grained.^xxiv In turn, they might argue that fair equality of opportunity requires a specific basic minimum for all citizens.

Yet, what particular appeal to fair equality of opportunity could be invoked? Institutions must make prioritisations between a given type of healthcare benefit and (1) values other than fair equality of opportunity, (2) contributants to fair equality of opportunity other than health,^xxv (3) other ways of promoting health beyond healthcare and (4) other types of healthcare benefits (such as relative priorities of mental health, primary care, women’s health, cancer-related benefits, surgical interventions and so on). Hence, we need a statement of fair equality of opportunity that specifies  both how the tradeoff should be made and to which of these tradeoffs the principle applies.

The most obvious specification of how the tradeoff should be made is one of optimisation on which two contributants should be prioritised relative to the degree that they promote fair equality of opportunity. For instance, perhaps expenditure on social determinants of health should be prioritised relative to healthcare expenditure to the degree that it promotes fair equality of opportunity over healthcare expenditure.

However, it is unclear how such an optimisation of fair equality of opportunity would be consistent with Basic Coverage Orthodoxy on any interpretation. First, Basic Coverage Orthodoxy only appears compatible with invoking fair equality of opportunity in the fourth type of tradeoff—that between different types of healthcare benefits--because  it is not clear that invoking it at other tradeoffs would gaurantee the healthcare coverage that Basic Coverage Orthodoxy requires. As Gopal Sreenivasan has reminded us, data suggest that socioeconomic determinants of health may make a larger contribution than healthcare to promotion of health equality.33 For example, every dollar (in circumstances like those of current developed democracies) spent to promote fair equality of opportunity through education—a social determinant of health—may better promote fair equality of health than every dollar spent on healthcare. Hence, if employed at tradeoff type (3), it is not clear that the relevant degree of basic provision would be protected by such a fair equality of opportunity principle. The problem is compounded if we consider applying the principle to tradeoff type (2) because many of these determinants affect fair equality of opportunity directly, rather than simply through affecting health. Hence, even if the contribution from healthcare to health favours at least some spending on healthcare, the overall contribution to fair equality of opportunity is less likely to. If so, this optimizing principle is unlikely to support significant (if any) healthcare expenditure.

Second, even restricting the principle to the fourth type of tradeoff, it is unclear that optimising fair equality of opportunity produces the intuitively correct verdicts. As the points about No Mental Health above revealed, orthodox theorists need to defend, not merely Basic Coverage Orthodoxy, but further an interpretation of it that guarantees certain content in the basic elements covered, such as mental health coverage. In turn, orthodox theorists would owe an argument for how such an optimising principle of fair equality of opportunity—or some other principle of fair equality of opportunity that they prefer—ensures an intuitively compelling interpretation of Basic Coverage Orthodoxy—one that avoids the concerns raised by cases like No Mental Health. (Presumably, there are many such cases as our intuitions appear to support a gaurantee of several other types of coverage; for instance, women’s health and other types of services included in the ACA’s essential benefits appear to be promising candidates.) Absent that argument, it is unclear that fair equality of opportunity explains all of the extra-procedural demands of legitimacy in healthcare.

Of course, an appeal to fair equality of opportunity is not the only type of argument that orthodox theorists might offer to meet the challenge.^xxvi Yet, I think the above makes clear the tension that orthodox theorists face in trying to limit their appeal to public reasons. Requirements of publicising public reasons create the sorts of challenges that I have raised for them in the last three sections. Yet, attempts to limit those requirements by appeal to extra-procedural moral considerations will usher in the very sorts of appeal to moral theory to settle specific policy considerations that were supposed to be problematic. It is intuitive that some procedural mechanisms must satisfy some extra-procedural demands of legitimacy, but we need a specific account of the demands in question and of the scope of the procedures. Cases such as those I have raised suggest that these will look very different from Orthodoxy.

Section VI: respect and compliance

Finally, given the influence of orthodox theorists, one might think that their arguments for these publicity requirements may be so powerful that we should accept the counterintuitive implications. Counterintuitiveness need not be a vice of a theory, so long as, on the whole, the theory’s support is compelling. In this section, I consider two powerful arguments for the publicity demands of Orthodoxy, arguing that both support views without such strong publicity requirements just as well.

First, some argue that citizens must provide mutual justification to each other through public reasons in light of the values of reciprocity and mutual respect. On Gutmann and Thompson’s view, ‘citizens owe one another justifications for the mutually binding laws and public policies they collectively enact’ (Gutmann and Thompson, p. 98, et passim)32—at least in cases in which the other citizens are ‘motivated to find fair terms of cooperation’ (Gutmann and Thompson, p. 78)30—because one must reciprocate mutual respect.

At best, this supports a pro tanto requirement for offering public reasons or a principle that claims, roughly, that resulting from a process that invoked such public reasons makes a pro tanto contribution to the legitimacy of some policy. Supposing that showing respect is necessary for legitimacy, and granting that offering mutual justification is a means of showing respect, offering mutual justification does not seem to be the only way of doing so. One can show respect by treating people in the ways that morality demands, for instance, by treating them as rights-bearers.^xxvii,xxviii The administration that undermines the deliberative process in Defective Reform Environment would appear to do this, as would the CMS officials that might undermine the policy in No Mental Health.

Second, quoting Frederick Schauer,34 Daniels distinguishes two ways of achieving compliance with decisions: appeals to their ‘source’ and appeals to their ‘right[ness]’. In the latter case, ‘giving reasons becomes a way to bring the subject of the decision into the enterprise’ (his emphasis). He argues that the first option is unavailable to ‘private health plans and public agencies… since their decisions will be respected only if they are seen as right—that is, as fair and legitimate’. Hence, he claims, agencies must give public reasons to their subjects (Daniels, p. 124).26 In turn, giving such reasons to subjects seems to require publicity.

In response, even if giving reasons brings subjects into the enterprise and helps subjects see decisions as legitimate, we must distinguish between decisions being seen as legitimate and their being so.^xxix It is not obvious that the latter presupposes the former. Securing stability of an institutional mechanism over time may be necessary for that institution to be legitimate, and perhaps the public seeing decisions as legitimate (or just or good in some way) is (one of) the best means to being stable. Nevertheless, neither appealing to public reasons nor giving reasons publicly each time is necessary to sustain such a sense. Many public institutions including courts, political offices and various agencies display a mixture of publicity and privacy about their rationales, and many of these institutions have significant public trust. For example, we generally accept that an open courtroom can interfere with the rights of citizens and therefore allow judges to close courtrooms to the press.^xxx

Section VII: after orthodoxy

Bioethical legitimacy involves a power to impose duties on those subject to one’s decisions. Given this, I have argued that processes that publicise public reasons are neither necessary nor sufficient for making legitimate bioethical policy in the way that orthodox theorists claim. Furthermore, various modifications to and arguments for Orthodoxy fail. Finally, attempts to limit the role of public reasons create tension with the general motivations for orthodox theory.

In conclusion, I wish to consider a last, potentially lingering, concern, favouring orthodox theorists. One might be attracted to Orthodoxy if one had doubts about our epistemic abilities to make judgments about the legitimacy of particular policies. After all, this sort of claim appears to lie behind orthodox theorists' view that moral theories are too coarse-grained to determine the legitimacy of bioethical policies. However, we can have meaningful and intelligible disagreement about what healthcare policies are legitimate. Furthermore, we can do so while acknowledging that some of us may have more justification for the beliefs that we express. Such acknowledgement seems to be part of much of our ethical discourse. Finally, the arguments above seem to suggest that we have intuitions about what is and is not legitimate and what is and is not just regarding bioethical policies. The fact that some readers may disagree does not itself mean that the arguments above offer no evidential weight.

Hence, we appear to have epistemic capacities to assess the contributions of extra-procedural considerations to legitimacy. However, the extra-procedural considerations at play will often be complex. Without more argument, it is not clear when they might fail to determine questions of legitimacy as pairwise comparison of cases like No Mental Health shows. Consider Less Mental Health.

Less Mental Health: Given healthcare resource constraints, Congress decides that Medicaid will cover less mental health services than it currently does. All Congress members want to justify this policy to each other and to those whom it will affect in terms that they can accept. Hence, they discuss the evidence in favour of various mental health treatments, but also evidence that the decision enables other benefits, which other fair-minded people should see as important.

Intuitively, the policy in Less Mental Health may be legitimate, even though that in No Mental Health is not. It would, I think, depend on the specifics of the plan and the context, including the alternatives that were possible. One may consider further iterations involving changes in healthcare provisions outside the domain of mental health. As it seems to me, some cases will be comparatively clear ones, while others may be the ones in which procedure-independent considerations genuinely fail to settle questions of legitimacy, making the appeal to a process attractive.

I believe these arguments favour considering alternative accounts of bioethical legitimacy—ones that appeal to extra-procedural considerations and embrace the possibility that we have the epistemic resources to do so. One promising way to build such a theory would be to start by striving to find the most promising interpretation of Basic Coverage Orthodoxy. One could proceed by  considering various potential extra-procedural demands of legitimacy in bioethics and cases with which to test them.^xxxi In short, if my arguments have been right, we must be heterodox about bioethical legitimacy to understand it.