Liberty to Request Exemption as Right to Conscientious Objection

Abstract

There is a regulatory option for conscientious objection in health care that has yet to be systematically examined by ethicists and policymakers: granting a liberty to request exemption from prescribed work tasks without a companion guarantee that the request is accommodated. For the right-holder, the liberty’s value lies in the ability to seek exemption without duty-violation and a tangible prospect of reassignment. Arguing that such a liberty is too unreliable to qualify as a right to conscientious objection leads to the problem of consistently distinguishing its effects from those of a right to conscientious objection that is made conditional on an individual assessment of the objector’s motivation. These properties require that we distinguish the liberty to request exemption from more restrictive policy choices, and that we subject it to greater scrutiny in the wider moral discourse as a possible variant of a right to conscientious objection.

KEYWORDS:

• conscientious objection
• Hohfeldian liberties
• claim-rights
• conscience absolutism
• incompatibility thesis

1. Introduction

Rival views on the permissibility of conscientious objection (CO) in health care – the choice of healthcare workers, on the basis of conscience, not to provide medical services or goods that are considered safe, legal, and within the scope of their professional responsibilities (Minerva 2017, p. 109) – can be distinguished by how they balance concern for healthcare workers’ conscience with enforcement of their professional duties, as well as by the procedural and material conditions that are thought necessary for exercising a right to conscientiously object.¹ This paper seeks to mediate discourse amongst ethicists and policymakers by examining how our answers to the question of permissibility depend not only on moral preferences and practical considerations but also on fundamental preconceptions about what it means to have a right. To lay these preconceptions bare, we must unpack the legal components that make up a right to CO and map the distinct ways in which these components can be assembled. This exercise reveals four main policy groupings: (1) to grant a liberty to opt out of providing treatment, coupled with a claim that the objector is accommodated, and without a mandatory referral; (2) to grant a liberty to opt out of providing treatment coupled with a claim to be accommodated, but combined with mandatory referral; (3) to grant a liberty to request exemption from prescribed work tasks without an accompanying claim that the request is accommodated; and (4) to grant no liberty to request exemption or claim to be accommodated.

Of these regulatory options, the third clearly stands out. Within the vast literature on CO it is a virtual unknown. While regulatory options (1), (2), and (4) can all be paired with developed moral theories, there is no such available theory for option (3). Nor is there an accepted terminology. I have opted here to refer to the regulatory option as that of granting a bare liberty, reflecting how healthcare workers under this policy lack protection in the form of a legal guarantee to exemption from prescribed work tasks. In the few instances where ethicists have considered this policy, it has been classed as the equivalent of regulatory option (4) – as the absence of a right to CO. We find a particularly clear example of this approach in Christian Munthe’s (2017) assessment of policies in Sweden. A Swedish healthcare worker may request exemption from some of the tasks assigned to them, including for reasons of conscience. This triggers a procedural duty for their employer to assess the request, but no substantive duty of accommodation. Decisions are made on a case-by-case basis, with legal guidelines stipulating that employers must consider factors including the demands of service provision and efficiency, the work environment, and the interests of the employee. If the employee refuses to provide treatment without obtaining accommodation, they must accept penalties, be discharged, or resign, an eventuality that has led prolife groups to repeatedly challenge the Swedish model related to legal abortion before the European Court of Human Rights (Domenici 2021). In short, while Swedish healthcare workers have a liberty to request exemption, accommodation is optional. There is no liberty to refuse to provide treatment without first being granted accommodation. According to Munthe (2017, p. 257), this approach ʻis, simply put, to provide no legal right to conscientious refusal of individual health professionals'.

In Munthe’s account, the status of a bare liberty as a non-right is posited rather than argued. The same pattern is found throughout the literature on CO (Cowley 2016, Fiala and Arthur 2017, Savulescu and Schuklenk 2017). I aim to test this view, and the premises that underpin it, in two ways. I will first adopt the perspective of legal and moral theorists who argue that bare liberties are rights whose value is lessened, but not erased, by the absence of protective claims. I will then examine the practical effects of granting healthcare workers a bare liberty to request exemption, and argue that there is no clear point of transition from this regulatory policy to regulatory options (1) and (2) when exemption from duties is conditional on an individual assessment of genuineness or reasonableness. Note that this examination does not amount to a normative assessment of policies like those adopted in Sweden. It is meant, instead, to clarify the status of bare liberties in the landscape of rights and to encourage more careful scrutiny of regulatory option (3) in the wider moral discourse. For the sake of clarity, my examination is limited to policies in the public healthcare sector that closely resemble the Swedish model, illustrated throughout with examples from the most widespread and disputed area of application: abortion. Its conclusions, however, are generalizable to CO more broadly, including the potential future applications that rapid technological and societal development may prompt (Oderberg 2019, pp. 217–218).

The suggestion that a bare liberty can qualify as a (severely limited) right to CO is bound to be highly contentious. If the argument succeeds, we are left with a new minimal conception of the right to CO and, in turn, a need to reframe accounts of its permissibility. Julian Savulescu and Udo Schuklenk (2018), for instance, have argued that granting a right to CO is morally impermissible and that attempted exemptions should be handled on the basis of labour law. Yet the policies they advocate grant a liberty to request exemption that may be accommodated even when made on the basis of conscience. If this bare liberty should qualify as a right to CO, then what they are proposing is not the absence of a right to CO but rather a more limited version of it. Even if the argument in favour of treating bare liberties as a right to CO is ultimately rejected, there is value in articulating the status of such liberties more clearly. Identifying nuances within the right to CO (or its absence) is a valuable tool for identifying sources of moral disagreement and presenting policymakers with a full range of regulatory tools (Nieminen et al. 2015). In particular, the exercise reveals crucial differences between regulatory options (3) and (4) that are currently underexplored.

The paper proceeds as follows. Section 2 sets out the possible components of a right to CO. Section 3 then maps the ways in which these components can be assembled, pairing each resulting policy with a companion moral theory and regulatory examples. Section 4 assesses the formal status and inherent benefits of bare liberties. In light of these benefits, excluding bare liberties from the scope of a right to CO requires an argument that they only qualify as generic rights and do not grant a sufficient degree of protection to count as a right to CO in the proper or strict sense. This objection leads, in turn, to section 5 and the problem of distinguishing the effects of bare liberties from those of a protected right to CO that is conditional on an individual assessment of the objector’s motivations.

2. Components of a policy response to conscientious objection

When someone has a right, they occupy a position that is apt to be advantageous relative to someone else (MacCormick 1977, Mackie 1978, Kramer et al. 1998). The right to conscientiously object is usually shorthand for a complex right, meaning a composite of multiple, distinct advantages. The most widespread and successful tool for identifying the specific content of such complex rights, developed by the American jurist Wesley Hohfeld, has long served as the standard model for rights theorists and is now the subject of renewed interest from legal practitioners and courts as well as moral and political philosophers (Hohfeld 1964, Thomson 1990, Wise 2001, Wenar 2005, Barker 2018). That it has not yet seen widespread use in the field of bioethics is, I believe, a missed opportunity.

Hohfeld’s key observation was that the same term – right – is used indiscriminately to denote normative incidents that are functionally distinct, leading to potential inaccuracy and confusion. His revised typology replaces the single all-encompassing right with four subtypes of rights. Of these four, there are two which are particularly salient in the context of CO. Claims, or claim-rights, are rights requiring someone other than the right-holder (i.e. the duty-bearer) to do or refrain from doing some act, while liberties are rights allowing the right-holder to do or refrain from doing some act.² In the present context, the starting point is that public healthcare workers have a general duty to provide any legal medical service or good, or treatment for short, that falls within their area of responsibility and competence. To have a moral or legal right to CO is, at its heart, to no longer be wholly bound by this duty. This absence of duty can be expressed as a liberty to seek exemption. The liberty can be coupled with a claim to be accommodated through reassignment and referral of patients. Note the analytical distinction at work here: whereas the liberty-right is what gives healthcare workers a choice to exercise – to provide treatment or seek exemption – the claim-right is what requires their employer, a review board, or other authorities to accommodate that choice. Hence the liberty-component of a right to CO is exercised through the right-holder’s own acts or omissions, and the claim-component through the acts or omissions of others.

If a conscientious objector has been granted a liberty not to provide treatment and an accompanying claim to be accommodated, and should those rights have been exercised, the objector has two further courses of action available. One is to make a referral or facilitate referral to other, willing providers. The other option is to simply refrain from providing treatment and do nothing more. In order for the latter course of action to be permissible, an objector must possess a further, negative liberty, namely the liberty to refrain from referral. The absence of such a liberty amounts to a duty of referral or mandatory referral.

These basic components are all that are required to construct the four groupings of policy options available in response to conscientious objectors. In the next section, each resulting assembly will be paired with a companion moral theory justifying the regulatory policy in question, an exercise that doubles as a means of communicating the content of these theories in a way that is immediately recognizable to legal professionals. There will be one exception, of course. The lack of systematic engagement with regulatory option (3) means that there is no applicable, developed moral theory to draw on.

3. Regulatory options and companion moral theories

Regulatory option (1) is the strongest possible version of a right to CO. It combines a liberty to opt out of providing treatment with a claim to be accommodated, as well as a liberty to refrain from ensuring or facilitating patient referral. Examples of this regulatory option can be found, for instance, at state level in Louisiana, Arkansas and Illinois (Wicclair 2011, p. 212). Doctors in these states cannot be compelled to perform an abortion, nor to facilitate its provision. Their right is in line with a moral theory that rules out any constraints on the exercise of conscience by healthcare professionals, a position termed conscience absolutism (Wicclair 2011). Such a theory is commonly supported by a form of moral integrity argument, whereby being forced to act against one’s conscience is thought to lead to a loss of identity, feelings of guilt, remorse, and shame (Gold 2010, p. 139, Wicclair 2011, pp. 25–26). Conscience absolutism is also situated as the potential endpoint of an argument from complicity, the idea being that making someone facilitate a treatment they object to makes them bear some of the moral responsibility for its provision (Cowley 2017, Trigg 2017).

Regulatory option (2) is similar in nearly all respects to the first. At its heart is still a liberty to opt out of providing treatment and a claim to accommodation. The one difference is that the liberty not to refer the patient has been replaced with mandatory referral. This combination is found in the majority of jurisdictions where CO is allowed in cases of abortion, including France and Italy (Chavkin et al. 2013). Since it accommodates healthcare workers to an extent, while also attempting to avoid excessive impediment to patients’ access to treatment, it is commonly referred to as the compromise approach (Brock 2008, Cowley 2016, Harris et al. 2018). Yet calling it the compromise approach makes it seem like there is only one regulatory option between conscience absolutism and an outright ban on CO, thus ignoring regulatory option (3).

Regulatory option (3) is to grant a liberty to request exemption from prescribed tasks without a companion claim that the request must be accommodated. Note that the liberty I am concerned with here goes beyond the mere right to voice objections that healthcare workers can be said to enjoy by virtue of their freedom of speech. It refers, instead, to a formally recognized and institutionally anchored ability to seek exemption. The Swedish model continues to serve as a helpful illustration. Here, healthcare workers have access to a formal mechanism for requesting exemption that includes a claim to an individual assessment. Unlike policies in line with conscience absolutism or the compromise approach, however, this claim is procedural rather than substantive. It guarantees an assessment, and not the outcome of that assessment. Thus regulatory option (3) clearly offers less protection than the previous two: without a guarantee of accommodation healthcare workers cannot be certain of the outcome of their request. This unpredictability is twinned with flexibility on the part of the employer, or any other authority tasked with reviewing requests, who may opt to prioritize interests other than conscience. To ensure fair treatment, this flexibility will likely be constrained by legal guidelines mandating that certain interests must be weighed, as it is in Sweden. Crucially, neither the absence of a legal guarantee nor the presence of assessment guidelines will preclude any occurrence of an exemption granted on the basis of conscience (Savulescu and Schuklenk 2018, p. 475). Furthermore, and despite its unfamiliarity, there is nothing legally or conceptually incoherent about the components that make up regulatory option (3). It is this tension, between weakened protection and the legibility of its components, which drives the question of whether a bare liberty can qualify as a right to CO.

Regulatory option (4) is to have no legally recognized liberty to request exemption or claim to be accommodated. If abortion is part of a healthcare worker’s portfolio of responsibilities, the treatment must be provided without exception. This is the position occupied, for instance, by public healthcare workers in Finland, and by auxiliary staff in the United Kingdom (Chavkin et al. 2013). In effect, this policy treats the healthcare worker as having given up the prospect of exemption when assuming the responsibilities of the role. Hence the labelling of the companion moral theory as the incompatibility thesis: to opt out of providing treatment on the basis of conscience is considered incompatible with the professional duties of healthcare workers (Wicclair 2011, Stahl and Emanuel 2017).

4. The status of bare liberties

4.1. The bare liberty as a right

I turn now to the conceptual status of bare liberties as illustrated by a liberty to request exemption from a duty to provide abortion; the next section examines the liberty’s practical effects. A useful starting point is the authoritative account provided by the legal theorist Herbert Hart (1982, pp. 166, 172), in which he critiqued the failure of contemporary scholars to recognize the significance of liberties unprotected by claims, without which, so he argued, there could be no clear understanding of the legal order. Since then, legal and moral theorists have generally assumed that the standing of bare liberties is reasonably clear: a rule that permits something is a kind of right (Thomson 1990, Kramer 2019). But not all. There is still a vocal minority which views bare liberties as something devoid of significance (Hurd and Moore 2018). To explain the lack of attention to bare liberties in debate on CO, it is helpful to briefly examine the factors that have caused such lasting disagreement elsewhere.

Two properties in particular fuel the classification of bare liberties as non-rights. First, they are constituted by an absence: the absence of duty. My liberty to sit on a park bench, for instance, is evinced only by the fact that sitting on the bench is not a violation of duty. Contrast this with the tangible events required to fulfil claims-rights, be it the payment of an outstanding debt or a patient transfer. It is easier to note that which occurs than that which does not, and so liberties are easily overlooked. Second, the exercise of a liberty is vulnerable to disruption. Imagine that I have been granted a formal liberty to participate in political assemblies, but that I am forced to stay home because of a threat of violence or arrest. Without companion claims to police protection or against arbitrary arrest it would seem that my right to free assembly has been hollowed out to the point of non-existence. Similarly, if health authorities have the option to refuse requested exemptions from work tasks, then we might legitimately worry that the liberty to make such requests is too weak to do morally substantive work. It is on these grounds it has been argued that the concept of a bare liberty is too weak to qualify as a right in the real sense – that it equates to a legal and moral ʻnothing'. On this line of argument, a true right to CO must be both permissive, in the sense of allowing the healthcare worker to choose, and inviolable, in the sense of protecting that choice with a legal guarantee (Barker 2018).

The sceptical position has great intuitive appeal. Joining a liberty with a protective claim clearly makes for a more effective right. Why, then, do theorists insist that unprotected liberties should be treated as independent and significant rights? Part of the answer lies in the doctrinal observation that all legal orders seem to in fact contain bare liberties (Thomson 1990, Schlag 2015). Another reason is conceptual. The argument against treating bare liberties as rights is founded, in part, on their lack of efficacy. Yet setting efficacy as an existence criterion for rights leads to serious boundary problems. Besides removing the guarantee of accommodation, a right to CO can also be weakened by the absence of channels for raising objections; by the lack of administrative mechanisms, judicial enforcement, or sanctions for employers who do not adequately accommodate conscientious objectors; by the absence of immunity against the reimposition of duties due to staff shortages; and so on. The upshot is that demanding efficacy can lead to a convoluted and unwieldy conception of what it means to have a right (Hart 1982).

Lastly, the sceptical position can be met head-on, by stressing the inherent value of bare liberties (Schlag 2015, Barker 2018). As Judith Thomson (1990, pp. 46–52) forcefully argues, this value lies in the ability to act, or refrain from acting, in the safety of knowing that one’s conduct does not violate a duty. In a jurisdiction with policies resembling those in Sweden, employees requesting exemption on the basis of conscience would know they are acting within the bounds of their professional duties. It is only if they refuse to provide treatment without obtaining exemption that a violation of duty occurs. Without this liberty, healthcare workers lack protection from negative professional consequences following ad hoc attempts at exemption (nor are they guaranteed an individual assessment). With the liberty, no one is wronged by the attempt. That, as Thomson (1990, p. 52) puts it, ‘is certainly not nothing’. It is for these reasons that scholars have opted to firmly distinguish the question of whether a right exists from the question of how valuable that right is to its holder, with the latter being contingent on a multitude of subjective factors and material and procedural conditions (Mackie 1978, Rawls 2009). Thus vulnerability and unpredictability do not deprive bare liberties of their status as rights.

How are these conceptual points reflected in the literature on CO? Munthe’s (2017) assessment that the Swedish model provides no legal right to CO is offered without further explanation, and so the conceptual premises behind it must be teased out from other accounts. Savulescu and Schuklenk (2017) also cite Sweden as an example of a jurisdiction that provides no right to CO. Their underlying ontology of rights is evident from the acknowledgment that a ‘few conscientious objection accommodation requests, dealt with on a case-by-case basis, may be unavoidable, but they should not be based on a right to be accommodated’ (Savulescu and Schuklenk 2018, p. 475). As long as there is no claim to be accommodated, the policy is regarded as consistent with removing ‘the legal right to object’ (Savulescu and Schuklenk 2018, p. 473). This framing mirrors that of legal theorists who require that real rights must be not only permissive but also guaranteed by claims.

For comparison, a parallel ontology of rights is evident in Christopher Cowley’s (2016) assessment of Dutch legislation regarding euthanasia or physician-assisted suicide (PAS). According to the Termination of Life on Request and Assisted Suicide Act of 2002,³ Dutch patients can request euthanasia provided they fulfil six separate conditions. Yet if a panel of medical reviewers determines that these conditions are fulfilled, the patient still cannot direct their request to any doctor. The procedure is performed only by doctors who volunteer their names to a centralized registry. Cowley (2016, p. 363) describes the resulting status of Dutch patients in the following manner:

Do patients (who meet the six conditions) have a right to euthanasia or PAS in the Netherlands? No. The patient requests PAS, and the (volunteer) doctor offers it.

The point here is not to directly compare the rights of Dutch patients with those of Swedish healthcare workers, but to illustrate how Cowley’s account demonstrates the same strict emphasis on claim-rights. Here, the liberty to request euthanasia is treated as a non-right because the claim to have the request fulfilled cannot be directed at an individual doctor as duty-bearer.

To the formalist, Cowley’s conclusion and Munthe’s claim that a liberty to request exemption or euthanasia is ‘no legal right’ cannot be defended; a legal liberty is a right. I have attempted to outline the conceptual reasons why we may want to adopt the formal stance, and some of those reasons are borne out in Savulescu and Schuklenk’s account. For one, it leaves us with an unanswered question (to which I return below): if labour law, or any other set of legal norms not ostensibly designed to protect conscience, allows healthcare workers to avoid having to perform abortions in practice, how accurate is it to say that the right to conscientiously object is wholly absent? Furthermore, merging regulatory policies (3) and (4) leaves no conceptual room for distinction between the Swedish model and other, more restrictive jurisdictions. After stating that Sweden does not grant a right to CO, Savulescu and Schuklenk (2017, pp. 162–163) state that ʻ[t]he same holds true' for Finland. But there is an important difference between the two jurisdictions. Once an abortion application has been considered, there is no established means for Finnish doctors, nurses, and midwives tasked with performing the treatment to request exemption on the cited basis of conscience (Fiala et al. 2016).⁴ Nor would (informal) attempts at exemption be guaranteed an individual assessment. On these two points, Savulescu and Schuklenk’s account would have benefited from a clearer conception of how bare liberties fit within the landscape of rights. Even if bare liberties ultimately do not qualify as a variant of the right to CO, it is still worthwhile to adopt a conceptual framework that allows clear distinctions between policies, and thus to present policymakers with a full range of regulatory options.

4.2. The effects of bare liberties

I have argued so far that bare Hohfeldian liberties to request exemption are rights with inherent significance, and that the failure to recognize them as such obscures key distinctions between policy responses to CO. Formalism, however, has its limits. If we are to have a meaningful philosophical and policy discussion about CO, the terms of the debate should also reflect intuition and extant usage of terms. It is therefore necessary to assess how the conceptual account of bare liberties matches the effects of adopting regulatory option (3). Imagine for these purposes a large hospital operating under the Swedish model. Its employees have been granted the liberty to request exemption from having to provide abortions, but their employer may refuse such requests in order to safeguard other interests. After these rules come into effect, nine doctors request reassignment on the basis that performing an abortion goes against their conscience. Their employer opts to accommodate them all. Finally a tenth doctor, call her Greta, requests reassignment. This time the employer, citing staff shortages, opts to deny the request and requires that Greta perform the procedure herself. Note that such a ratio is not far-fetched. In Norway, for instance, there is a common practice of doctors referring patients who ask for contraceptives to colleagues without specifying the reasons for the referral. Although health authorities have the power to intervene against such referrals at any time, this power is very rarely used.⁵

The question is, as before, whether employees at this fictional hospital enjoy a right to CO. Three answers are possible. The first is to argue that the first nine doctors enjoyed such a right but Greta did not. This answer is unsatisfactory because the presence or absence of a right would be contingent on factual circumstances. There would be no way to make an a priori determination of whether the policy amounts to a right to CO. The second answer is to argue that the first nine objectors only appeared to have a right to CO, and that Greta’s case revealed the right to be non-existent. This is the answer that is implicit in Munthe, Savulescu, and Schuklenk’s commentaries. Yet it is a fact, and no mere matter of appearance, that the first nine doctors were relieved of prescribed work tasks. Their liberty to request exemption has translated into a tangible good. The third answer is to bow to formalism and acknowledge that all ten doctors enjoyed the exact same right, unchanged by circumstances but subject to the same vulnerability. On this view, Greta’s experience did not reveal the right to CO as non-existent; it just highlighted its limits.

The scenario also serves to demonstrate how classifying Greta’s liberty to request exemption as a right to CO strains our intuitions. How can she be said to have a right to CO if she did not get her desired outcome? Does a fruitless attempt at exemption not equate to a legal and moral ‘nothing’? Even this intuition must be properly qualified, however. Greta was not faced with the inevitable prospect of having to act against her conscience. She could take her chances and request a reassignment – safe in the knowledge that there was no duty not to – and hope that her employer would allow it. Under a policy resembling the Swedish model she would also be guaranteed an individual assessment. Although studies and evidence on this issue are lacking, it is possible that such an established mechanism provides a greater chance of exemption than ad hoc attempts to avoid prescribed work tasks. Lastly, a recognized liberty to request exemption could be considered morally significant because it allows her to formally signal protest against an action she sees as unconscionable (thereby providing employers and policymakers with a gauge on prevailing attitudes). For these reasons, one inherent and the others contingent on policy choices and practice, having a liberty to request exemption without a guaranteed outcome is still preferable to the situation facing auxiliary staff in the United Kingdom or Finnish doctors. This line of reasoning could ultimately be employed as part of a normative justification for adopting regulatory option (3). Since the focus here is solely on the conceptual status of bare liberties, I will not pursue it further.

5. Bare liberties and conditional claims

So far, it seems there is a case to be made that a bare liberty to request exemption on the basis of conscience is strong enough to do morally significant work. At this stage, however, the sceptic may object that this form of argument does not address the right issue. Even if the liberty to attempt referral is a right, so the argument goes, it cannot be properly called a right to CO. Or, while terminally ill Dutch patients may well have a right, theirs is not a right to euthanasia. The alleged difference is between a generic right and a right that grants sufficient protection of the relevant interests to qualify as a right to conscientiously object. Arguably, this objection goes beyond the accounts of Cowley, Munthe, Savulescu, and Schuklenk. Their views can all be interpreted as being founded on a general ontology of rights, whereas this objection pertains to our specific understanding of what it means to have a ‘right to CO’ (or ‘right to euthanasia’).

The new objection also brings new challenges. Presumably, the underlying reasoning for excluding bare liberties is that a proper right to CO provides some form of outcome guarantee. The question is where, using this metric of predictability, a bare liberty to request exemption ends and a conditional right to CO begins. For the outcome under regulatory policies (1) and (2) can also be rendered unpredictable by subjecting objectors to an individual assessment of their motivation. It is this blurring of boundaries that leads Jonathan Hughes (2018, p. 129), in his commentary on Swedish managers and their optional accommodation of requests, to conclude that there may not be much of a difference between a sympathetic Swedish employer and a model that grants a conditional right to CO.

Savulescu and Schuklenk (2017, 2018) favour a model where conscience-based objections are handled in accordance with labour law on a case-by-case basis. Since the few accommodations granted on this approach do not amount to a right to CO, even when they are made on the basis of conscience, the right that occasionally translates into referral must presumably be called something else. Let us call it a ‘labour right’. At first glance, the difference between this labour right and a right to CO may seem obvious. We could claim, for instance, that the right to CO is specifically designed to protect conscience, whereas a labour right is motivated by the more general interests of employers, employees and patients. Yet the intentions of legislators cannot be the only guide to determining whether rights have been granted. Regard must also be had to a policy’s effects.

Procedural requirements, such as providing advance notice or entering a registry of objectors (Chavkin et al. 2013), or substantive conditions, such as demanding that objectors belong to a designated faith, do not cast serious doubt on whether the underlying access to CO qualifies as a right. Instead, boundary issues arise whenever conscientious objectors are subject to an individual assessment of their motivation (Card 2017). A common caveat in various jurisdictions is that objectors must provide evidence that their objection is rooted in beliefs that are sincere, deeply held, and consistent. This is meant to root out conscientious objectors that are motivated by discriminatory beliefs, by the desire to avoid association with stigmatized treatments, to disengage with morally contentious issues, or to alleviate heavy workloads (Chavkin et al. 2013, Harris et al. 2018). Whether such a requirement is fulfilled must, by its nature, be judged individually. Employers, review boards or other authorities thus face the difficult task of gauging inner thoughts and motivations in order to differentiate legitimate and illegitimate cases. Such motivations can never be fully accessed, and so the results are bound to be unpredictable (Fiala and Arthur 2017). Some accommodations may be granted that are not based on legitimate reasons of conscience, and some legitimate objectors may be denied.

Now, we do not doubt that jurisdictions which review referrals to ensure that they are legitimate still grant a right to CO. But in these jurisdictions, too, accommodation is not guaranteed. If there is a difference between this source of unpredictability and the unpredictability faced by those who have a liberty to request exemption without a claim to be accommodated, it seems, as Hughes (2018, p. 129) notes, to be one of degree. It should also be noted that the unpredictability common to conditional rights to CO and bare liberties can never be fully eliminated with formal guidelines. In Sweden, employers are asked to weigh the needs of healthcare provision and the work environment against the interests of healthcare workers, but it is in the nature of such assessments that decisions cannot be wholly determinate. Nor can the outside observer be sure that the decision-maker has strived for objectivity.

Even superficially clear rules can introduce unpredictability by ways of a safety valve. Mark Wicclair (2011), for instance, defends a version of the compromise approach in which conscientious objectors do not need to provide reasons for their refusal. This policy appears to fall neatly into category 2 – as a protected liberty to conscientiously object combined with mandatory referral – and its status as a right to CO has never been in doubt. Yet Wicclair also sets out an exemption for referrals made on the basis of discrimination, beliefs contrary to the goals of medicine, or demonstrably false clinical beliefs. These exceptions make the right that he advocates subject to individual review, and therefore conditional (Meyers and Woods 2007, Card 2017).

The same need for individual assessment is an inbuilt feature of accounts that make the permissibility of CO conditional on the ʻgenuineness' of the objector’s beliefs, as well as accounts that require objections to be ʻreasonable' (Meyers and Woods 2007). David Oderberg claims that the latter is more objective and allows us to exclude from the scope of permissible CO medical practices that are of such a nature that ʻno reasonable person could object to its performance' (Oderberg 2020, pts. 10, 12). However, a reasonableness criterion at the policy level would lead to the same problems of indeterminacy as any other guidelines. If the goal is to prevent illegitimate objections, standards of reasonableness would also have to be deployed at the level of individual assessment (Card 2017). Some assessments of reasonableness do not create boundary issues; if the request is based on an empirical error, for instance, then accommodation is not warranted. Barring such clear cases, reviewers assessing reasonableness face the same problem of discerning the real motivations of the objector.

6. Conclusion

A bare liberty for healthcare workers to request exemption from prescribed duties is a kind of right, and it grants tangible benefits. In order to persist in excluding such liberties from the scope of a right to CO, one must argue that the absence of a protective claim to be accommodated leaves them overly vulnerable. This objection leads, in turn, to the problem of distinguishing the unpredictability caused by not having a legally guaranteed claim to be accommodated from the unpredictability caused by rights to CO that are conditional on an individual assessment of the objector’s motivation. These delineation issues can be avoided by extending the notion of a right to CO to regulatory frameworks that give healthcare workers bare liberties to request exemption, as in Sweden. This new, minimal conception of what it means to have a right to CO is not inherently disruptive to the wider normative discourse. The moral arguments in favour of and against a right to CO remain unchanged. The shift in conceptual premises should, however, push ethicists to subject policies that grant bare liberties to more rigorous examination.

Acknowledgements

I am grateful to César Palacios-González and Matthias Brinkmann, as well as to members of the Uehiro Centre for Practical Ethics at the University of Oxford and participants of the Practical Philosophy Working Group at the University of Oslo for their valuable comments on earlier drafts.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Stages of manuscript revision were supported by the Norwegian Research Council under the Centres for Excellence grant nr. 223274 ʻPluriCourts – Centre for the Study of the Legitimate Roles of the Judiciary in the Global Order'.

Notes on contributors

Johan Vorland Wibye

Johan Vorland Wibye is a postdoctoral fellow at the University of Oslo, based at the Institute for Public and International Law. He holds a Ph.D. in philosophy of law, an M.St. in Practical Ethics, an MJur, and an M.A. in Law.

Notes

1 Conscience refers here to a set of deep moral beliefs which provide uniquely strong reasons to act or to refrain from acting (Wicclair 2011, pp. 4–5, Nehushtan and Danaher 2018, p. 542).

2 The two remaining subtypes are powers to change normative relations and immunities which protect the right-holder against attempted changes in normative relations (Hohfeld 1964, pp. 50, 60).

3 Termination of Life on Request and Assisted Suicide Act 1 April 2002; The Netherlands: Articles 2b, 2c. Available from https://www.worldrtd.net/dutch-law-termination-life-request-and-assisted-suicide-complete-text.

4 See Finland Law No. 239 of 24 March 1970 On the Interruption of Pregnancy. Available from http://cyber.law.harvard.edu/population/abortion/Finland.abo.htm, section 6.

5 See evidence cited in Norway Supreme Court judgment HR-2018-1958-A: at para. 50. Available from https://lovdata.no/dokument/HRSIV/avgjorelse/hr-2018-1958-a.