1 Introduction

The role of courts in democratic states is often characterised as a counter-majoritarian safeguard to protect individual rights. Yet, courts as such Waldron (2006) and international courts in particular Bork (2010) and Mounk (2018) have often been as criticised as undemocratic. While the current backlash against international courts is in part driven by populist governments (see Voeten 2020), there may also be a genuine concern for the collective decision-making in democratic states (Bellamy 2014).

International human rights courts are in an especially difficult situation as they aim to protect human rights against the very states that have to comply with and implement their judgements.Footnote 1 As a result, several international courts have developed mechanisms of deference to states. One of these is the European Court of Human Rights’ (ECtHR) margin of appreciation doctrine (Benvenisti 1998; Shany 2005; Letsas 2006). The ECtHR grants a wider or narrower margin in which it follows the state’s authority to determine whether the European Convention on Human Rights (ECHR) has been violated in a particular case. However, when is international courts’ deference to states problematic—undermining their very mandate—and when is it a case of courts showing due respect for the states’ appropriate scope of discretion?

In order to understand this issue, it is of no use insisting that states are sovereign, since sovereignty is only valuable insofar as it protects popular self-determination and human rights. Therefore, I propose to distinguish between personal and political autonomy. Personal autonomy is an individual’s capacity to choose among different options according to her reasons, thereby shaping her future and being the author of her own life. Political autonomy refers to the status of individuals as equal norm-givers within their political community and its institutions. Following these ideals, international human rights courts should ensure the conditions of personal autonomy by protecting human rights but respect the democratic decisions of states, insofar as they realise political autonomy.

This paper has two aims: First, it seeks to show that the margin of appreciation can be reconstructed in these terms. The focus on political autonomy brings a new perspective that enables us to understand some of the controversies around the ECtHR’s practice in new terms. Second, once this framework has been established, the paper will use it to critically evaluate the Court’s practice on this basis. In doing so, it highlights the implications of the normative account of personal and political autonomy in terms of the scope and forms of deference. Furthermore, the paper contributes to a more nuanced understanding of personal and political autonomy by drawing on how the ECtHR handles conflicts between them in practice.

To do so, the paper analyses the arguments and normative premises regarding personal and collective autonomy that the European Court of Human Rights brings to bear on three salient cases:

  1. a)

    SAS v. France: concerning the French ban on face covering—which the Court found did not violate the European Convention on Human Rights’ (ECHR) provisions on the right to privacy or freedom of religion.

  2. b)

    Dahlab v. Switzerland: concerning a teacher wearing an Islamic headscarf in class; in violation of alleged community values regarding teachers not imposing “identification” on their pupils—where, again, such claims and trade-offs are contested both within and across European states.

  3. c)

    Strand Lobben v. Norway: concerning child custody dilemmas. Here, the question arises as to how the ECtHR should address issues that are alleged to be based on community-wide values. It concerns trade-offs regarding the interests of the child vs. the interests of “the family”—when those trade-offs are contested and differ across Europe.

These cases have been selected because they refer in different ways to national values and democratic decision-making (subsidiarity), as well as their relation to rights and trade-offs among them. While it is not possible to draw a generalised conclusion about the Courts practice overall, they do provide case studies for the theoretical argument that the paper makes. In analysing the cases, the paper pays attention to the normative claims to have such expressions of cultural values and collective decision-making respected by international courts, and the somewhat conflicting normative claims that an international human rights court should be authorised to review and possibly rule against some of these expressions. The analysis attends to the relevant arguments by the parties in the particular cases, as well as normative assessments of such arguments against the background of personal and political autonomy. Furthermore, the cases involve different domestic institutions (national courts, parliaments, and administrative bodies) and speak to questions of “procedural turn” in the margin of appreciation doctrine (Brems 2017) and the use of the European consensus (Dzehtsiarou 2015; Kleinlein 2017).

The paper proceeds as follows: First, it outlines challenges to the legitimacy of international human rights courts and the potential role of the margin of appreciation to resolve them. In the second part, it illustrates the use of the margin of appreciation by analysing the reasoning of the three cases mentioned above and shows how it can be critically assessed from the perspective of personal and political autonomy. The third part discusses appeals to the role that European consensus plays as part of the margin of appreciation and “the procedural turn” in how the ECtHR carries out its review with regard to the margin of appreciation. The paper reviews these developments in the ECtHR’s case law and practice and assesses whether and how they can be justified in the light of personal and political autonomy.

2 Legitimacy Challenges for International Human Rights Courts

One aim of international courts is to protect personal freedom and minority rights, in particular through human rights adjudication. Yet, human rights courts have increasingly come under fire for judicial activism and interfering with the sovereignty of legitimate states (e.g. Madsen et al. 2018; Pauwelyn and Hamilton 2018). An interesting example is the legitimacy critiques that have been voiced against the European Court of Human Rights (ECtHR) claiming that it oversteps its purpose by interfering with democratically generated and acceptable interpretations of human rights.Footnote 2

How should we evaluate such legitimacy contestations? From the perspective of sociological legitimacy—the acceptance of a specific court by those subjected to it—human rights courts have a difficult stand because, given the lack of an inter- or supranational enforcement mechanism, cases are usually brought against the very states which need to implement their rulings. So, when states do not see them as legitimate, the very functioning of the courts is at stake. Human rights courts therefore have to walk a thin line between fulfilling their mandate and not antagonising the states that are parties to their human rights system. When we focus on normative legitimacy—the justifiability of these courts’ authority—different questions arise. One perspective from which we can analyse legitimacy challenges to international courts is that the states’ claims to sovereignty against human rights court’s judgements may be normatively justified because the state protects certain values, such as national culture or democracy, better than the court.

To this sovereignty argument, one may, of course, object that human rights conventions and their associated courts are based on the states’ consent through the ratification (and by not withdrawing). In other words, the authority of courts is derived from the sovereignty of states and does not stand in contradiction to it. However, normatively speaking, there is a question of how meaningful state consent is, due to issues of representativeness and voluntariness (e.g. Buchanan 2002; Christiano 2011). Leaving this fundamental question aside, it is still important to ask how far the authorisation through this consent reaches. From a legal theory perspective, international human rights courts, and, in fact, any court, must have a certain degree of discretion to interpret rights. While the Vienna Convention on the Law of Treaties of 1969 (VCLT) provides certain specification on treaty interpretation in Articles 31–33, which include a textual approach favouring a narrower understanding of state consent and a teleological approach that gives more leeway for interpretation following the purpose of the treaty, the VCLT does not indicate how to choose between or balance them. In the case of the ECtHR, the convention, similar to a constitution, has been interpreted in an evolutive manner, as a living instrument (e.g. Letsas 2013). Yet, what the limits of this interpretative freedom are is hard to establish.

Another approach to assess the legitimacy of an authority is the so-called service conception (Raz 1986). The constraint on states’ power that international human rights courts provide is assisting them in protecting human rights in a way that states cannot achieve on their own. If this is the case, sovereignty claims cannot immunise states from the requirement to comply with human rights courts. Of course, there might be other reasons why compliance with a particular court is not warranted, for example, efficiency or procedural issues. Yet, states might acknowledge the authority of human rights courts generally speaking, but reject it in certain cases, claiming that these courts overstep the domain of human rights protection and thereby interfere with their sovereignty. This is particularly the case if we think that there is a domain for which it is better for states to decide according to their own judgement. This idea is captured by Raz’s independence condition, which limits legitimate authority to decisions that are not better taken by oneself (Raz 2006). One way to understand this idea is that only other-regarding reasons, not self-regarding ones, can ground legitimate authority (Scherz 2022). However, based on the service conception, it is not clear where the domain of other-regarding reasons ends and that of self-regarding reasons begins. Therefore, there exists ambiguity around the question of when states should be able to decide for themselves and when an international court has legitimate authority to decide for them.

We can either see this as an issue of value pluralism that the court has to resolve (Raz 2017), or as balancing sovereignty concerns with concerns of human rights protection. One approach the ECtHR has been taking to address this issue is the so-called margin of appreciation doctrine (hereafter margin).Footnote 3 It enables international courts to exercise restraint and flexibility in reviewing the decisions of national authorities, granting them some measure of discretion. However, the margin does not by itself resolve the tension between the two poles. Rather, it provides the Court with a practice to do so. In his seminal paper, “Two Concepts of the Margin of Appreciation”, Letsas (2006) describes what he calls the substantive view of the margin as a balancing between individual rights and public interest. He also makes a similar point, saying that the margin of appreciation itself does not provide a systematic account of when the Court should defer to states but that the margin is itself in need of a normative theory. Letsas suggests that different theories of rights, in particular rule-utilitarian, interest-based, and reason-blocking theories, can fulfil this function. The structural concept of the margin holds that, in certain cases, the power of the Court should be restrained if its nature as an international court means that national authorities are better placed to decide on human rights issues. Letsas argues that “the Court’s case law on consensus and public morals under the structural concept of the doctrine is in clear violation of anti-utilitarian liberal principles under both interest-based and, even more so, reason-blocking theories of rights” (2006, p. 731). While the structural concept may still be useful under interest-based models, in his view, these models come with rather severe disadvantages. Therefore, Letsas argues that the structural concept of the margin should be abandoned. In contrast, I suggest that shifting the focus to political theory, centring on questions of legitimacy and democratic decision-making, instead of a purely moral theory, can make sense of both concepts of the margin, while still remaining critical. However, from the political approaches discussed above, neither the idea of sovereignty nor the service conception can establish how the margin of appreciation doctrine should be used. To do so, I am proposing a different normative ideal, namely that of autonomy. In this sense, we can understand the tension between human rights protection and national sovereignty as one between personal and political autonomy. To be clear, this does not mean that this approach rejects a moral view of rights, but rather that it also attributes value to collective political decisions insofar as they realise the right to participation, i.e. if they are sufficiently democratic.

Both of these forms of autonomy require institutional realisation through political institutions. Personal autonomy is an individual’s capacity to choose among different options according to her reasons, thereby shaping her future and being the author of her own life. In order to do so, a sufficient set of valuable options is necessary to act autonomously (Raz 1986, 369), and importantly these options have to be robustly guarantied (Pettit 2012, Chapter 1). Such a guarantee is best provided in the form of institutional protection of human rights not only by states but also through an additional safeguard by international human rights treaties and courts. Political autonomy refers to the status of individuals as equal norm-givers within their political community and its institutions (Habermas 1996, Chapter 3; Forst 2012, 125–37). This requires reciprocal recognition as equals: not only as subjects of the law, but as co-authors of it. Such control as a norm-giver over the rules to which one is subjected is institutionalised through equal participation rights in democratic decision-making. Collective decisions are normatively valuable in so far as they are expressions of political autonomy, establishing a political order, based on the participants’ own agency, values, and convictions. Political autonomy is a group right that grounds claim to sovereignty in order to protect this freedom to collective self-determination.

How does the margin balance between personal and political autonomy? The margin can be understood as a means to take into account reasons based on political autonomy in the adjudication of the court. Many regard the margin of appreciation as a limitation of evolutive interpretation through the consensus amongst contracting states (Letsas 2013). Importantly, however, the margin of appreciation is never unlimited and therefore does not afford total deference to states. Rather, international courts retain the authority to review the reasonableness of national decisions. According to Yuval Shany (2005), the margin of appreciation is specifically appropriate for intrinsically uncertain international law norms. Furthermore, it can help to combine comparative institutional advantages of the two levels: benefiting from national courts’ competences in fact-finding and fact-assessing as well as democratic accountability while ensuring norm-interpretation at the international level. In the following, I will propose that the margin of appreciation can and should be seen as a mechanism to balance democratic political and personal autonomy.

3 Justifying the Margin of Appreciation in Terms of Autonomy

The ECtHR’s margin of appreciation doctrine allows the Court to grant states a certain measure of discretion to decide how they comply with their obligations under the Convention, in a specific case. The margin of appreciation doctrine has developed as a practice of the Court: First, it was stated as a requirement of “reasonable relationship of proportionality” in the Belgian Linguistic case as the Convention implies “a just balance between the protection of the general interest of the community and the respect due to fundamental human rights while attaching particular importance to the latter”.Footnote 4 Second, in Handyside v. the United Kingdom, the Court departs from this requirement and evaluates the interference with a right based on the necessity of this interference in a democratic society. Importantly, the Court also emphasises that the Contracting States do not have an unlimited power of appreciation and that the Court is empowered to establish the final ruling. The margin is generally only granted in circumstances where trade-offs arise between Convention rights and other Convention rights that can be limited, but not for non-derogable rights such as Articles 2, 3, and 4. In the three cases analysed in this paper, the margin concerns specifically restrictions of Articles 8 (the right to respect for one’s private and family life) and Article 9 (freedom of thought, conscience, and religion), regarding the relationship between the state, private life, and religions. In SAS v. France, the Government contends that the “Court afforded States a wide margin of appreciation when it came to striking a balance between competing private and public interests (Evans v. the United Kingdom)”.

Yet why is a deferential review valuable? First, it may be used to avoid conflicts in a pluralistic legal system. This in turn should be understood on the basis of a sociological legitimacy requirement for international courts: They need the continued compliance of states. A second value that is often attributed to deferential review is subsidiarity. In fact, the margin of appreciation was historically developed to incorporate subsidiarity considerations in the Court’s decision-making. Yet there is also a risk in deferring too much and thereby undermining the effective protection of human rights (Follesdal 2021).

Subsidiarity is, however, not a value in itself. So, on what values does subsidiarity itself rest? Subsidiarity is often justified on an epistemic basis, namely, that local authorities know the situation better because they have direct access to local facts and needs (Young 2009). Yet there seems to be a broader understanding of subsidiarity at play here. For example, Gerards summarises it as follows: “Domestic authorities are usually better equipped to make such assessments, since they are likely to be more closely acquainted with national problems, (constitutional) traditions, sensitivities and debates” (Gerards 2011, 85). This raises the question of whether this reliance on subsidiarity is a different explanation that conflicts with the reasons for respecting the political autonomy expressed in democratic decisions. In my opinion, the democratic justification is also partly based on this epistocratic basis but adds a self-determination reasoning to it. Therefore, it is not contradictory to subsidiarity but rather complementary by shifting the normative focus. Eva Brems lists three different but not contradictory rationales for the ECtHR to assess the quality of domestic processes: process efficacy, subsidiarity, and process value (2017, 33). In this case, process value, such as fairness or democratic quality, is distinct from subsidiarity.

In Hatton and Others v. the United Kingdom, the Court also acknowledges this democratic element of subsidiarity:

The national authorities have direct democratic legitimation and are, as the Court has held on many occasions, in principle better placed than an international court to evaluate the local needs and conditions (…). In matters of general policy, on which opinions within a democratic society may reasonably differ, the role of the domestic policy maker should be given special weight. §97

The same point is made in SAS v. France to argue for a wide margin of appreciation, noting:

[the] fundamentally subsidiary role of the Convention mechanism. The national authorities have direct democratic legitimation and are, as the Court has held on many occasions, in principle better placed than an international court to evaluate local needs and conditions. In matters of general policy, on which opinions within a democratic society may reasonably differ widely, the role of the domestic policy-maker should be given special weight (see, for example, Maurice v. France [GC], no. 11810/03, § 117, ECHR 2005-IX). §129

This expresses the fact that courts are in general differently democratically legitimised than democratically elected bodies—courts are legitimised as interpreting laws set by such bodies but should not insert their own views for those of the legislature or executive. The setting of international courts that are not directly embedded in a democratic system makes this issue even more pertinent. International treaties are often phrased broadly and therefore require more interpretation that falls on international courts, particularly in the absence of an international or supranational legislature. In many cases, it may not make a difference whether or not we regard these two rationales for subsidiarity as distinct, since they are compatible and often coincide. However, the democratic rationale that the autonomy-based argument supports grants the margin of appreciation for different reasons and can therefore delimit specific conditions under which the Court should do so.

Beyond these general points, the autonomy-based reading of the margin differs from the general understanding of subsidiarity in several ways: first, does not just require minimal respect for domestic decision or, second, substantive deference on the grounds of expertise (Kavanagh 2008), but rather regards the actual democratic decision as a reason in itself. In other words, there is an intrinsic value associated with deciding for oneself: the fact a decision is our own gives it value. For individuals, voluntary informed decisions are generally an expression of autonomy. For political autonomy, things are more complicated since it is not the case that just any group agent, and in turn any collective decision, is normatively valuable. Furthermore, political autonomy attributes normative value specifically to democratically made decisions as they realise the equal autonomy of all citizens. It is important to note that this accounts attributes value to the political decision and not to “public morals” held by the majority, as rightly criticised by Letsas (2006).

3.1 Margin of Appreciation: Political Autonomy and Collective Values in the ECtHR Case Law

In order to show how an assessment on the basis of personal and political autonomy can be applied to the margin of appreciation, this section provides a short overview of how claims about political autonomy are made in three different cases. These claims and the practice of the ECtHR are then normatively assessed on the autonomy-based account.

  1. a)

    SAS v. France

In SAS v. France, clear reference to specific “national” values was made. In the domestic debate of the law, the claim was that wearing a face veil in public is an infringement of the values of the Republic: liberty (as a symbol of subservience), equality (by violating equal dignity), and fraternity (as a negation of contact with others and a flagrant infringement of the French principle of living together) (§17). In the discussion about the Bill before Parliament, it was stated that: “Public order is not confined to the preservation of tranquility, public health or safety. It also makes it possible to proscribe conduct which directly runs counter to rules that are essential to the Republican social covenant, on which our society is founded” (§24). In addition to the public safety argument, values of the dignity of the person and others and in particular gender equality, which is constitutionally guaranteed, are referred to.

The decision also takes note of the situation in other European states, in particular Belgium and Spain. The discussion about the Belgian law on a similar ban argues for a “societal model where the individual took precedence over his philosophical, cultural or religious ties”. Specifically, three aims are listed: “public safety, gender equality and a certain conception of ‘living together’ in society” (§42). The Judgement of the Spanish Supreme Court argues to the contrary that “No legitimate aims were constituted by the protection of ‘public tranquility’, ‘public safety’ or ‘public order’ since it has not been shown that wearing of the full-face veil was detrimental to those interests” (§47).

The applicant maintains that the ban prevents her from manifesting and living according to her faith and from observing it in public. She argues against the legitimate aim of the provision and that it is not necessary in a democratic society nor for public safety. In particular, she states that the aims can be achieved by less restrictive means and that it remains necessary to weigh up the competing interests: “those of the members of the public who disapproved of the wearing of the veil; and those of the women in question” (§78). It is argued that the ban is paternalistic and can indeed not rely on the protection of gender equality as it punishes “the very women who were supposed to be protected from patriarchal pressure” (§78). By criminalising not only the coercion of others to wear a veil, but also the wearing of the veil itself, disregards the position of women who voluntarily choose to do so. The Government, on the other hand, bases its argument on public safety and on the protection of the rights and freedoms of others. The latter makes reference to three values: (1) the observance of the minimum requirements of life in society; (2) equality between men and women; and (3) respect for human dignity, “since the women who wore such clothing were therefore ‘effaced’ from the public space” (§81). With regard to the first value, it argues that the “face expresses the existence of the individual as a unique person, and reflects one’s shared humanity with the interlocutor, at the same time as one’s otherness” (§81). Therefore, veiling one’s face in public constitutes a refusal of the principle of “living together” (“le vivre ensemble”).

In order to assess whether there is a legitimate aim for the restriction of the rights in question, the Court discusses the aims invoked by the Government, namely, public safety and “respect for the minimum set of values of an open and democratic society” (§114). Here, the three values are understood as respect for equality of men and women, respect for human dignity, and respect for the minimum requirements of life in society. The Court assesses whether they qualify as being linked to the “protection of the rights and freedoms of others”, within the meaning of the second paragraphs of Articles 8 and 9 (§116). First, the Court acknowledges that the respect for equality of men and women prohibits anyone from forcing women to conceal their face, which corresponds to rights and freedoms of others. However, it holds that the State Party cannot invoke gender equality to ban a practice that is defended by women, unless it were to be understood that individuals could be protected on that basis from the exercise of their own fundamental rights (§119). Second, it argues that the respect for human dignity cannot justify a blanket ban. Third, the Court holds, however, that the respect for the minimum requirements of life in society, “living together”, can be linked to the legitimate aim of “protection of the rights and freedoms of others” (§121). When assessing whether the measure is necessary in a democratic society, the Court states that:

In democratic societies, in which several religions coexist within one and the same population, it may be necessary to place limitations on freedom to manifest one’s religion or beliefs in order to reconcile the interests of the various groups and ensure that everyone’s beliefs are respected (see Kokkinakis, cited above, § 33). This follows both from paragraph 2 of Article 9 and from the State’s positive obligations under Article 1 of the Convention to secure to everyone within its jurisdiction the rights and freedoms defined therein. (see Leyla Şahin, cited above, § 106) (§126)

In particular, it emphasises the state’s role as a neutral and impartial organiser of the exercise of various religions. In this role, the state is supposed to ensure mutual tolerance between opposing groups. The Court notes that there is a certain restriction of pluralism involved in the banning of wearing veils in public, but that the state deemed this practice incompatible with social communication and the requirements of “living together” in French society. As a conclusion, the Court holds that France in this way aims to “protect a principle of interaction between individuals, expression not only of pluralism, but also of tolerance and broadmindedness without which there is no democratic society” (§153). Therefore, it characterises the ban on full-face veils in public as a choice of the society. Importantly, the Court describes its role as assessing “a balance that has been struck by means of a democratic process within the society in question” (§154).

From the perspective of personal and political autonomy, the Court argues very much in line with respecting the balance that the society in question has found between personal and political autonomy. On this basis, it is plausible to reject the restriction of the freedom of religion on the basis of public safety or gender equality as the Court has done in this case. However, I share some of the concerns of the Joint Partly Dissenting Opinion of Judges Nussberger and Jäderblom. To draw out these issues and show how the analysis through the lens of autonomy is useful, I will first analyse whether the interference with the right in question can be justified on the basis of personal or political autonomy, balancing to individuals rights or an individual rights and a collective one. First of all, the justification of the interference in terms of personal autonomy is questionable, as it is not clear which rights of others are at stake in the requirements of “living together”. In particular, it is not clear how it falls under rights of others within the meaning of Articles 8 and 9 para 2. (4). “Case law is not clear on what constitutes the rights and freedoms of others outside of the scope of the rights protected by the Convention” (§5). It cannot be a right not to be shocked or provoked by different cultural or religious practices and identities, as the Court has proclaimed this to be an essential part of pluralism, when it comes to freedom of expression (§7). France tries to justify this restriction of pluralism on the basis of toleration itself. In this interpretation, the argument raises the well-known question of how far a tolerant society and state should be towards intolerance. Yet, again, it is not clear how the wearing of a full-face veil is in fact intolerant. The argument seems to capture only those who coerce others to wear a veil, not those who decide to do so themselves. According to Rawls, restricting the freedom of the intolerant is only permitted “when the tolerant sincerely and with reason believe that their own security and that of the institutions of liberty are in danger” (Rawls 1971, 220). As has been seen, public safety is not a sufficient ground for a ban. Therefore, we would need to understand the requirements of “living together” as necessary for the institutions of liberty. In this sense, the argument more plausibly refers to political autonomy. However, the meaning of the value of “living together” is problematically broad and vague. If we analyse the justification of the interference with the right in terms of political instead of personal autonomy, it is true that there was a public debate and a democratic decision prima facie justifying the ban as an expression of political autonomy. Yet, in order to determine whether the political decision is, in fact, a valid expression of political autonomy, attention needs be paid to the democratic quality of the political process and the effects of this on the preconditions of democratic decision. I will discuss this in Section 2.2. To sum up, the margin of appreciation granted to the state in this case can be justified from the perspective of political autonomy, but only if the political process leading to the ban of face veils was sufficiently democratic.

  1. b)

    Dahlab v. Switzerland

In Dahlab v. Switzerland, the Federal Court states that: “In Geneva, section 6 of the cantonal Public Education Act of 6 November 1940 provides: ‘The public education system shall ensure that the political and religious beliefs of pupils and parents are respected.’ It also follows from Articles 164 et seq. of the cantonal Constitution that there is a clear separation between Church and State in the canton, the State being secular” (§3). It argues that the principle of denominational neutrality acquires particular importance in state schools, since basic education is compulsory for everyone. In addition, the Federal Court contends that the wearing of the headscarf is difficult to reconcile with the principle of gender equality established in the Constitution and that “religious harmony ultimately remains fragile in spite of everything, and the appellant’s attitude is likely to provoke reactions, or even conflict, which are to be avoided”. As to the question of whether there is a sufficient basis in the law, the Federal Court holds that the law is sufficiently precise. The legitimate aim is seen as given in the protection of the rights and freedoms of others, public safety, and order. The measures are argued to be proportionate to the legitimate aims since they only prohibit the wearing of a headscarf in the context of teaching activities, as teachers in public schools are also acting in their role as state representatives.

The Court acknowledges that religious beliefs were fully taken into account in the proportionality testing by the Federal Court. “It is also clear that the decision in issue was based on those requirements and not on any objections to the applicant’s religious beliefs”.

The Court’s assessment states:

The Court accepts that it is very difficult to assess the impact that a powerful external symbol such as the wearing of a headscarf may have on the freedom of conscience and religion of young children. The applicant’s pupils were aged between four and eight, a period during which children wonder about many things and are also more easily influenced than older pupils. In those circumstances, it cannot be denied outright that the wearing of a headscarf might have some kind of proselytising effect, seeing that it appears to be imposed on women by a precept which is laid down in the Koran and which, as the Federal Court noted, is hard to square with the principle of gender equality. It therefore appears difficult to reconcile the wearing of an Islamic headscarf with the message of tolerance, respect for others and, above all, equality and non-discrimination that all teachers in a democratic society must convey to their pupils.

What can the autonomy-based understanding contribute to the analysis of the margin in this case? First, it is worth noting that here, the Court relies not only on the rights and freedoms of others, but also on public safety. This can be a problematic argument from the perspective of autonomy because there is no absolute safety and safety claims can always be used to restrict the rights and freedom of individuals. Therefore, safety claims should be used only in exceptional cases to justify the interference with rights. Second, the rights of others that are protected in this case are the rights of the children and may therefore seem like an opposing personal autonomy concern. However, there is also a public choice involved, namely, the neutrality of the school system and the connected autonomy of the political community that chose to teach school pupils in a context of denominational neutrality. In other words, there are individual rights to freedom of religion, both for the teacher and the pupils, the right to education (personal autonomy), and the decision about the way that these are balanced (political autonomy). This seems well justified from the perspective of political autonomy.

Three points remain problematic though: (1) the Court seems to follow the interpretation of the Federal Court that the headscarf is difficult to reconcile with tolerance and equality even if women speak up to defend it as their choice; (2) the alleged conflict and instability of religious harmony were never perceived by the pupils or their parents. This raises the question of whether neutrality is a sufficiently important value of the state to protect in the absence of actual conflict; (3) in comparison to Lautsi v. Italy where crucifixes were not seen as problematic in schools, there seems to be an inconsistency in how the Court judges the influence of religious symbols on children. “There is no evidence before the Court that the display of a religious symbol on classroom walls may have an influence on pupils and so it cannot reasonably be asserted that it does or does not have an effect on young persons whose convictions are still in the process of being formed”. The last point may, however, be explained by the margin attributed to two states who hold different interpretations of the acceptability of religious symbols in schools. This is underscored by the fact that the Swiss Federal Court in fact also prohibited crucifixes in schools. Nevertheless, the readiness with which both the Federal Court and the ECtHR condemn the headscarf as being in conflict with gender equality seems worrying. In exact opposition to the Court, one can argue that seeing a teacher with a headscarf is promoting the message of tolerance, respect for others, equality, and non-discrimination in a democratic society. We can think of autonomy as a critical theory principle that, in cases of conflict, points to the protection of minority groups and the protection of their autonomy, as this is usually what is under pressure.

To sum up, the autonomy perspective is critical with regard to the use of safety claims to justify the interference with individual rights. However, political autonomy does allow for a collective decision on how to specifically balance the rights in question. Here, again, the democratic quality of the decision would have to be analysed. Finally, the reasoning against the headscarf are questionable from the perspective of autonomy as a critical principle.

  1. c)

    Strand Lobben v. Norway

Strand Lobben brings to light a range of different and interesting issues regarding the Court’s use of the margin of appreciation. It assesses the violation of Article 8 of the Convention to private and family life, in a case in which Norway had restricted the parental rights of the applicant and given her Child X up for adoption. Therefore, the case does not at first glance fit the private–public conflict that characterises the other two cases. In a prior judgement, the Chamber had found legitimate aims for the restriction of the right to private and family life to be the “the protection of health or morals” and the “rights and freedoms” of X in accordance with Article 8 §2. The margin of appreciation is, however, relevant to determine the extent to which a state is permitted to interfere with family life in the interest of the well-being of a child. The applicant argued that the margin:

Accorded to the competent national authorities would vary in the light of the nature of the issues and the seriousness of the interests at stake. It was well established that in cases relating to placement of children in public care and adoption, the domestic authorities enjoyed a wide margin of appreciation. However, the Court tended to hide behind the margin of appreciation concept in a way which could to some extent undermine its control and functions. (§162)

The Court agreed that the margin accorded varies with regard to the nature of the issues and the seriousness of the interests at stake. While the Court recognised that the “authorities enjoy a wide margin of appreciation in assessing the necessity of taking a child into care”, it held that “stricter scrutiny is called for in respect of any further limitations, such as restrictions placed by the authorities on parental rights of access, and of any legal safeguards designed to secure an effective protection of the right of parents and children to respect for their family life” (§211). It also points to the importance of the decision-making process being “such as to secure that the views and interests of the natural parents are made known to and duly taken into account by the local authority” (§212).

In contrast to SAS v. France, where the argument is clearly geared towards national values and tradition, the Norwegian Government in Strand Lobben seems to argue that they have balanced the interests in the right way, making a universal value claim, not one of political autonomy. Yet, in Strand Lobben, there are different levels of interpretation at play. While Norway seems to argue for the universal value of its interpretation, the Court takes the argument not as a substantive universal one, but instead tries to evaluate whether the balancing by the state has been done properly and whether it constitutes an acceptable “national” interpretation of the best interest of the child and the right to family. The Court contends that “perceptions as to the appropriateness of intervention by public authorities in the care of children vary from one Contracting State to another, depending on such factors as traditions relating to the role of the family and to State intervention in family affairs and the availability of resources for public measures in this particular area” (§210). In this case, following the Court’s reading, we can understand the political autonomy as expressed in the domestic interpretation of the child’s best interest that political institutions have adopted. Connecting this to the autonomy analysis of the other cases, in order to assess whether these decisions are expressions of political autonomy, the democratic quality of the process needs to be assessed. Here, interestingly, the deciding institutions were primarily administrative agencies and national courts. I will discuss the issues of democratic quality and the influence of the deciding institution in the next two sections.

3.2 The Margin Based on Political Autonomy: Fostering Democratic Quality

From the perspective of political autonomy, what should be protected are the democratic decisions in different states. We can say that the expression of political autonomy requires a good faith balancing (proportionality test) of the different positions by taking into account the democratic process. Yet it is not clear what the public deliberation of the law in question was in France. Specifically, the clear, almost unanimous majority opinion in this public deliberation is worrisome.

This raises questions about what kind of a democratic model should be the ideal or at least the acceptable minimum for such a process. Depending on what the democratic process and institutionalisation in a state looks like, there will also be differences in deliberation. For example, there is a broader public deliberation to be expected in Switzerland with a semi-direct democracy than in France, as a semi-presidential system. If we think that there can be not only different democratic outcomes but also different legitimate ways of institutionalising democracy, then variation in public discourse is not a problem as such. Therefore, we would need to know whether the deliberative process was sufficient in comparison to the French level. This means that the deliberation need not be a “public” deliberation but that the one held in the parliament is also relevant. On this basis, the French discussion may pass as a good faith attempt, even noting the worrisome tendency to not take seriously the voices of women who are wearing full-face veils.

More important and problematic is the fact that this very process seems to foster an atmosphere of Islamophobia and intolerance. This undermines the very conditions of political autonomy (and potentially also personal autonomy) to decide collectively in a democratic process among free and equals. This is contrary to the state’s role as a neutral and impartial organiser of the exercise of different religions, which is supposed to ensure mutual tolerance between opposing groups. The Court acknowledges this tension in SAS:

The Court is very concerned by the indications of some of the third-party interveners to the effect that certain Islamophobic remarks marked the debate which preceded the adoption of the Law of 11 October 2010 (see the observations of the Human Rights Centre of Ghent University and of the non-governmental organisations Liberty and Open Society Justice Initiative, paragraphs 98, 100 and 104 above). It is admittedly not for the Court to rule on whether legislation is desirable in such matters. It would, however, “emphasise that a State which enters into a legislative process of this kind takes the risk of contributing to the consolidation of the stereotypes which affect certain categories of the population and of encouraging the expression of intolerance, when it has a duty, on the contrary, to promote tolerance”. (§149) (my emphasis)

The focus on political autonomy therefore allows us to criticise the ban on wearing veils in public and therefore also the respect which that law enjoys in the Court’s margin of appreciation.

With regard to the question whether democratic states are less subjected to international courts’ authority because they are granted a wider margin of appreciation, it is important to emphasise that a margin of appreciation is granted based on whether the state has conducted a proper “proportionality test” in the case in question, not the general democratic credentials of the state (Follesdal 2017). Thus, the international courts’ authority over democratic and non-democratic states is the same. As Wheatley suggests, reasons, such as having epistemic humility about their own decisions, should lead democratic states to accept the authority of international human rights bodies or, in other words, not to take themselves to be superior in understanding the instrumental conception of legitimacy (see Wheatley 2014). Overall, the background duties to establish equal equality provide important reasons for states to comply with human rights courts and while there are certain risks for political autonomy, state consent, and procedures, such as the margin of appreciation doctrine, can reduce them to an acceptable level.

3.3 Democratic Quality and Deciding Institutions

Another question which the evaluation of the margin of appreciation in respect of political autonomy needs to address is whether it is relevant who makes a domestic decision: a court or the legislature. In particular, it is important to know whether this makes any difference for the democratic value of these decisions as expressions of political autonomy. In general, local remedies have to be exhausted, and there needs to be a basis in the law to restrict certain rights. Therefore, in many cases that reach the ECtHR, we can expect a combination of legislative and juridical decisions. Yet there can still be differences in this regard. SAS is about a specific law without a domestic court decision; Dahlab concerns a Federal Court decision and its basis in cantonal law; and Strand Lobben raises more specific questions about the administrative application of relatively broad law and subsequent court decisions.

How should we assess the decisions of these different institutions in terms of political autonomy? National legislation that is set by the respective parliament is clearly an expression of political autonomy. Also, national courts should not be completely disregarded from this perspective, even if they are not directly democratically legitimised, because they are still part of a democratic system. Yet should we still think of this as an expression of democratic political autonomy in the case of administrative decisions? One could argue that there has not been a legislative action to counteract or check this application. This could then be interpreted as a form of tacit consent. However, this does not seem very convincing, since, on the one hand, the normative value of tacit consent has been questioned and, on the other, the absence of intervention could also be due to transparency issues. This means that such administrative decision should, in general, be considered to carry less democratic weight.

Now, some may argue that the Court is itself not a democratic institution and therefore it is problematic to let it interfere with democratic decisions and even decide on their quality. Yet, can we think of the Court’s function as a quality control on the democratic process instead of an interference with or even restriction of it? This is plausible if we understand courts as another forum for discourse that is not independent but connected to the political one. There may be worries that courts on the international level do not fit this description of being connected to the political system. However, if they are sensitive to domestic democratic decision-making as the margin of appreciation suggests, then this worry seems unfounded.

4 The Margin of Appreciation in Relation to European Consensus and Its Procedural Turn

There are two further aspects of the ECtHR’s practice of the margin of appreciation that warrant consideration, namely, its use of a European consensus and the so-called procedural turn in proportionality assessments. These will be discussed here in turn.

4.1 European Consensus

When determining the appropriate margin of appreciation, the ECtHR assesses whether there is a European consensus and grants a wider margin if there is no such consensus among the states on the relevant issue. In this way, the use of a European consensus in the margin of appreciation doctrine can be understood as a mechanism of self-restraint of the Court. Yet its application has often been criticised as inconsistent (e.g. Follesdal 2019). This section seeks to assess this practice from the perspective of personal and political autonomy.

In SAS v. France, the Court argues that there is no European consensus on how religion is regulated. It states that it is:

not possible to discern throughout Europe a uniform conception of the significance of religion in society and that the meaning or impact of the public expression of a religious belief would differ according to time and context. It observed that the rules in this sphere would consequently vary from one country to another according to national traditions and the requirements imposed by the need to protect the rights and freedoms of others and to maintain public order. It concluded from this that the choice of the extent and form of such rules must inevitably be left up to a point to the State concerned, as it would depend on the specific domestic context. (see Leyla Şahin, cited above, § 109) (§130)

However, the joint partly dissenting opinion notes correctly that the Court could have found a European consensus with regard to the narrower issue of banning full-face veils in public: “It is difficult to understand why the majority are not prepared to accept the existence of a European consensus on the question of banning the full-face veil” (§19). This remark in the dissenting opinion on the European consensus is interesting, as they point to the inconsistency in how the Court assesses and uses the existence of a European consensus.

To what extent does the Government in Strand Lobben indicate the role of consensus in the larger argument concerning margin? The government does refer to the fact that there is no European Consensus. The Court, on the one hand, seems to agree with the Government, therefore granting a wide margin of appreciation on the necessity of taking a child into care, yet, on the other hand, it places stricter scrutiny on any further limitations of the right in question. This is interesting as the Courts seem to split the areas with regard to how the margin should apply.

Yet, from the perspective of autonomy, it is not evident why a consensus would have any weight. The idea of a consensus seems to speak to the question of what the limits are and how much the balancing between personal and political autonomy can reasonably differ. It is difficult to see the consensus having importance out of respect for national values. If national democratic decisions and values are important, it should not depend on whether or not other states agree with them.

Could we think of the European consensus as defining areas where there is no reasonable disagreement? Is it for the Court to define the boundaries or reasonableness? If there were a clear process of defining consensus, it would not be the Court’s decision, but there seems to be a rather large amount of discretion involved. Can this be compared to how one defines state practice?

It is important to keep in mind the function of European Consensus here. If there is a consensus against the practice of the state in question, there will be stronger scrutiny, i.e. it is to determine how wide the margin of appreciation should be set. This would also mean that the Court does not treat this as a hard line on reasonableness but rather as an indication of it. In a similar vein, Mattias Kumm (2018) interprets the deferential application of the proportionality test, e.g. through a margin of appreciation, as defining reasonableness in terms of public reason.

To conclude, the Court’s use of assessing whether there is a European consensus within the margin of appreciation is not only often inconsistent and arbitrary; its normative relevance itself is questionable. From the perspective of political autonomy, being bound by other states’ practice without having agreed to it is not normatively valuable even if the other states agree. At best, the existence of a consensus can be taken as an indicator of reasonableness and its absence as an indicator of reasonable disagreement. However, the possibility remains that the consensus itself is unreasonable. Therefore, Letsas’ (2006) assessment of the European consensus as a problematic tool can also be upheld from the more political theory of autonomy.

4.2 The Procedural Turn

A further way in which the ECtHR has sought to defer to the contracting states within certain limits is the so-called procedural turn in its review (Brems 2017). Here, the Court does not assess the substantive issues of the proportionality test, but rather assesses whether the state has carried out a good faith test turning to the quality of the domestic decision-making procedure. How can such a mechanism of deference be justified from an autonomy-based perspective?

To understand this question, the case law of the Court is helpful. For example, the third-party comments in Strand Lobben point to the difference in expertise between the domestic authorities and the Court. They contend on this basis that the Court should not undertake its own assessment of the facts and should not render a substantive decision but rather a procedural review of the authorities’ decision-making:

With respect to subsidiarity, the UK Government pointed to paragraph 28 of the Copenhagen Declaration. In cases such as the present, account should be taken of the relative expertise and involvement of the domestic authorities compared with the Court, the level of participation of the parties affected by the domestic process, and the level of consensus amongst Contracting States. The seriousness of the intervention at issue was also relevant, but a closer scrutiny could not entail a fresh assessment of the facts and particularly not if considerable time had elapsed since the decision under review. The Chamber minority could be understood as seeking to establish that the Court should undertake its own assessment of the underlying facts, rather than reviewing the decisions, particularly by its reference to the need for “a forensic examination of the facts” and by indications that the dissenting judges envisaged that the Court itself should render a “substantive” decision. The Grand Chamber was invited to reject this approach; as had been stated by the Chamber majority, the Court was required to consider whether the domestic authorities had adduced relevant and sufficient reasons for their decisions, but only the domestic authorities were in a position to determine what was in the child’s best interests. (§191)

The Court acknowledges that the national authorities have an epistemic benefit because they are in direct contact with all the persons concerned. Therefore, it understands its task not as substituting its judgement for the domestic one, “but rather to review under the Convention the decisions taken by those authorities in the exercise of their power of appreciation (see, e.g. K. and T. v. Finland, cited above, § 154; and Johansen, cited above, § 64)” (§210). However, it contends that “the domestic authorities did not attempt to perform a genuine balancing exercise between the interests of the child and his biological family (see paragraphs 207 and 208 above), but focused on the child’s interests instead of trying to combine both sets of interests, and moreover did not seriously contemplate any possibility of the child’s reunification with his biological family” (§220). This means that the Court criticises not the substantive decision but the decision-making process as not establishing the required “safeguards that were commensurate with the gravity of the interference and the seriousness of the interests at stake” (§225). Therefore, this is often understood as a procedural review of the proportionality test.

The joint dissenting opinion in Strand Lobben of Judges Kjølbro, Polackova, Koskelo, and Norden on the Merits of the Case is very interesting, as it argues that the Court only claims to apply a procedural requirement of a proportionality test, but in fact is applying a substantive view:

The more profound problem is that by giving priority to its own preferences as to how the competing interests should be weighted and balanced, the Court in effect curtails the margin of appreciation that it is important to preserve, especially in situations where the domestic authorities must consider individual rights and interests that may well be contradictory and where views may differ as to how the relevant values, principles and competing considerations should best be reconciled in the given circumstances. This is all the more so in a context such as the present one, where the domestic authorities are under a duty to fulfil positive obligations toward a vulnerable child.” (§15)

They argue further that: “In the present case, it clearly appears that the manner in which the majority have identified ‘procedural shortcomings’ in fact arises from the substantive view taken, as a result of which the domestic authorities are faulted for ‘focusing on the interests on the child’ instead of his reunification with the biological family” (§16). This is not obliviously true. Of course, it is always a question of how much an interest must be considered in order to count as having been considered. In this sense, a minimal substantive criterion is necessary even on a procedural account of a proportionality test. Therefore, the boundaries between substantive and procedural review are not as sharp as they are sometimes made out to be.

5 Conclusion

The aim of this paper was to show how the margin of appreciation can be understood and critiqued through the concepts of personal and political autonomy. It has argued that the margin of appreciation should be understood as a positive practice of the European Court of Human Rights, as it acknowledges the value of domestic democratic decision. However, in its use, the Court should attend more to the actual democratic quality of such decision-making when granting a wide margin. Second, the paper has shown that the European consensus in the doctrine is not only often inconsistently applied but also not well founded from a normative perspective. Finally, with regard to the procedural turn of the ECtHR’s review, the paper argues in favour of such a practice but also contends that the boundaries between substantive and procedural review are not always as sharp as they are often taken to be. This also means that the Court should not give up its function of protecting personal autonomy by ensuring the effectiveness of the Convention.

As a consequence of the suggestion to use personal and political autonomy to assess the margin of appreciation, the cases in which a large margin is granted to the Contracting States will be more restricted. In this sense, the proposed approach is a critical normative tool to restrict the deference of the Court. Some may, however, wonder whether the Court should really engage in such contentious political questions about democratic quality of national decision. While this may politically not be popular, one of the aims of the Court is to protect the democracy of its Contracting States. If states knew that good faith proportionality procedures involve a democratic dimension, they may be encouraged to heighten the quality of these procedures. Therefore, from a normative perspective and potentially even a political one, the Court should take the autonomy perspective more seriously in its margin of appreciation doctrine.