Investigating Lessons for the EU’s Fundamental Rights Policies

Elise Muir lets us know early on that her book ‘is less an essay on equality law than an enquiry into the way a supranational organization, such as the EU, develops a fundamental rights policy’.1 Equality law is chosen as the ‘case study’2 from which wider lessons will be drawn.3 Muir puts forward two main arguments in this book which can be summarised as follows: first, creating a fundamental rights policy such as that on equal treatment through legislation based on a distinct legal basis understood as giving specific expression to a fundamental right ‘creates a significant risk of over-constitutionalisation’.4 That means the political debate on the definition of fundamental rights is limited5 because such definition is too closely tied to primary law (the constitutional version of the right). Second, Muir deals with what she calls the ‘governance’ of EU equality law in domestic spheres. She maps the governance ‘infrastructure’ as set up by secondary legislation

(and as influenced by judicial intervention) to support a fundamental rights culture (on equality) at the domestic level. For her, this is where the real added value of equality law may lie today. 6 This is the second argument permeating the book, namely that 'legal innovation' at the domestic level is preferable to a top-down prescriptive approach, and equality law, with its own set of distinct tools and instruments, could teach us lessons here. In this review, I shall focus on the first argument of the book, concerning the risk of 'over-constitutionalisation', which takes up a larger share of the book. Before doing so, however, I shall first reflect on the notion of 'fundamental rights policy' as used in this book and on the delimitation of this study.
T '   :  ''     '  ' In examining the concept of a 'fundamental rights policy' for the EU, this book revisits an old theme. That is valuable as the literature in this field remains limited.
This theme emerged in the late 1990s when Alston and Weiler 7 famously urged the EU to adopt a human rights policy, comprising 'positive' measures as opposed toand in order to complementnegative 'legal prohibition on violations'. 8 Alston and Weiler's human rights policy also had to be comprehensive. They highlighted gender equality as a single and laudable, even if 'far from perfect', initiative 9 but this only illustrated that other areas were lagging behind. They also used a thick 10 understanding of the term 'human rights policy', meaning going beyond legislative initiatives. 11 Relatedly, Alston and Weiler put forward a forceful call for institutional reform in order to create policy structures and an administration tasked with policy formulation and implementation. Post-Lisbon, such reforms were undertaken. New bodies, structures and with them a series of mechanisms were introduced to meet the fundamental rights promotion duty per Article 51(1) CFR. 6 Ibid., p. 145.  Alston and Weiler, ibid., p. 666. The terms 'human rights' and 'fundamental rights' are often used interchangeably in the English-speaking literature. Muir does the same. Therefore, the same approach is maintained in this piece. 9 Ibid. 10 This choice of term is this author's. 11 E.g. administrative and budgetary prioritisations, working through civil society, funding human rights initiatives in external relations.
It is notable that Muir's use of the term 'fundamental rights policy' in this book is narrower. She is mainly concerned with a thin (as opposed to a thick) understanding of a specific (as opposed to a general and comprehensive) 'fundamental rights policy'. That means, in examining EU equality law as giving expression to the fundamental right to equal treatment, Muir's main focus is on the concretisation and realisation of this specific fundamental right through secondary EU legislation. Within that, special value is ascribed to legislation with an explicit legal basis, solely concerned with this fundamental right.

R      
Both choices made in this book: (1) to focus on the thin dimension of the term 'fundamental rights policy'; while (2) investigating equality law, warrant further consideration.
Focusing on the thin dimension is not immediately obvious at first sight because a broader understanding is conventionally used for the term 'policy'and not just in practice. Policy studies literature 12 reveals that the definition of the term 'public policy' is not straightforward. Still, a certain core can be agreed on. The simplest and often quoted definition is a very broad one: 'anything that a government chooses to do or not to do'. 13 Policy means to achieve policy objectives are wide-ranging and encompass much more than legislation. Muir, however, wants to put the emphasis on the legislation, presumably to emphasise the distinctiveness of equal treatment in this regard.
This leads to the second choice of investigating this specific rather than (aspects of ) the EU's more general fundamental rights policy.
Muir sets out the 'fundamental rights policy'-nature of EU equality law with reference to two characteristics. First, its 'transformative mandate' or 'transformative function'which for her appears to be a key qualifier of a fundamental rights policy. She derives this idea from an article by Von Bogdandy (but in a changed, paraphrased way), 14 that EU law scholars will remember as a prominent and critical response to Alston and Weiler's calls for an EU human rights policy. 15  EU equality seeks 'to achieve 'inter-personal equality per se' 16 and to 'change democratic societies'. 17 This she contrasts with 'just protecting individuals against threats to the said fundamental right or limiting European intervention that may affect such right'. 18 At times, the term 'transformative function' is used here interchangeably with the term 'private function', i.e. the application of EU equality law in the private sphere, though the former notion seems to be broader in nature than the latter. In any case, it stands next to EU equality law's so-called 'infrastructural function', namely its application in the public sphere.
The second, for Muir, key distinctive characteristic of EU equality law as a fundamental rights policy relates to its current mandate in the Treaties. We are reminded that, since the Treaty of Amsterdam, equal treatment profits from a specific legal basis: Article 19 TFEU (for combatting discrimination based on sex, racial or ethnic origin, religion or belief, disability, age or sexual orientation), which importantly stands next to and complements the older Article 157(3) TFEU (for ensuring the application of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation, including the principle of equal pay for equal work or work of equal value), that includes both economic and social elements. This is special according to Muir because it gives the EU a specific competence on a fundamental right and the fundamental rights dimension has been acknowledged early on by the Court of Justice (the Court) and has been enhanced over time. There are, of course, also other fields where the EU has a more specific fundamental rights competence, and Muir acknowledges this fact. She mentions in particular two well-known examples: asylum law and data protection, but distinguishes them. Asylum law is said to be different because it is 'more concerned with consistency of EU intervention'. 19 Data protection, on the other hand, resembles equality law 20 post-Lisbon because it now has a distinct legal basis (Article 16 TFEU), but is for her still different 21 because this legal basis maintains a reference to the free movement of data and is thus reminiscent of the internal market (Article 16(2) TFEU); this is said to be also mirrored in the relevant legislation. Muir thus places the distinctiveness of equality law in the fact that action in this realm 16 Muir, supra n. 1, p. 15, citing F. Scharpf, 'Perpetual Momentum: Directed and Unconstrained?', 19 JEPP (2012) p. 127 at p. 132-133. 17 Muir, supra n. 1. 18 Ibid. 19 Ibid. 20 Ibid., p. 178: 'EU data protection law to some extent constitutes a fundamental rights policy' (emphasis added). 21 Ibid., Chapter 4, section C 'EU Equality and Data Protection Law: Parallel Constitutional Designs and Challenges', at p. 136 ff.
is based on a legal basis that is explicitly concerned with enhancing a fundamental right and includes no other objective. She thus distinguishes 'instruments impacting on fundamental rights protection while achieving another central policy aim' 22 as distinct from EU equality law. Because of the absence of other co-existing objectives, she qualifies this policy as an 'autonomous' fundamental rights policy, which is the 'first' such policy.
Given the distinction Muir makes for data protection, logically this must then be for her the 'only' such policy thus far. 23 Based on this, the 'lessons to be learned' appear to be directed at 'emerging' 24 fundamental rights policies. Given the emphasis on the nature of the legal basis, however, there is no room for other fundamental rights policies (in Muir's use of the term) to emerge in the absence of Treaty amendment. Here the reader may wonder what these lessons to be learned are then directed at? It appears throughout the book that in present times the main candidate is data protection law, even if there are limits to that.
Let us now turn to the 'over-constitutionalisation' argument. Muir first lays down the risks associated with over-constitutionalisation in theory (Chapter 2) before illustrating those based on her case study (Chapter 3). She then seeks to draw wider lessons for other fundamental rights policies (Chapter 4).
T   '-'   Muir rehearses here familiar arguments for judicial vs. legislative approaches to fundamental rights protection. She relies mainly on well-established accounts of Waldron, who 'emphasizes the role of legislation in placing disagreement on the nature and shape of fundamental rights protection at the core of political activity', 25 and on Somek holding 'constitutions that require public intervention to elaborate rights and political institutions ought to be given leeway to choose how to flesh out the positive dimension of a fundamental right'. 26 She also draws on Von Bogdandy, 27 to highlight risks that come with framing such legislation in fundamental rights terms or as being part of a fundamental rights policy. Those risks are presumably amplified in the EU context. I understand Muir to connect the points made by Waldron, Somek and Von Bogdandy in the following way: settling disagreement on fundamental rights at the legislative level is desirable. What is less desirable is the couching 22 Ibid., p. 15. 23 Ibid., p. 14. 24 Ibid., p. 14. 25 Ibid., p. 44 discussing J. Waldron, Law and Disagreement (Oxford University Press 1999). 26 Ibid., discussing A. Somek, The Cosmopolitan Constitution (Oxford University Press 2014). 27 Von Bogdandy, supra n. 14, p. 1315, discussed by Muir at p. 37. of legislation in fundamental rights terms because that may lead to unwanted 'interferences between two layers of norms'. 28 The Court may, under these circumstances, find it easier to interpret secondary legislation based on the constitutional (primary law) version of a right, and thus take the matter away from the political processes. In the EU context, this is even more tricky than it would be in national contexts, because the EU legal order is said to be already heavily constitutionalised (in particular its economic law) and 'a great emphasis on fundamental rights discourses would only enhance the appeal to intensify' it even more. 29 This would, in turn, defeat the idea of placing special emphasis on political avenues for resolving conflicts between competing interests in the EU. 30 So, Muir's concern of 'over-constitutionalisation' seems to relate to the specific idea of judicial reliance on the constitutional version of a fundamental right in order to constrain the legislature in shaping the same right via the political process (where there is an explicitand 'sole aim'fundamental rights competence to do so). And this is presumably made easier for Courts when legislation is couched in fundamental rights terms.
The reader may relate the concern Muir identifies to Grimm's 'over-constitutionalisation'-problem in the EU, 31 in the sense that both are concerned with the ordinary legislature not being too constrained by constitutional provisions. But the two are not quite the same. Grimm states that in the EU, 'the treaties are not confined to those provisions that reflect the functions of a constitution. They are full of provisions that would be ordinary law in the Member States'. 32 This became problematic the moment the treaties were no longer treated as traditional international law but became 'constitutionalised' through the doctrines of direct effect and supremacy. 33 Grimm's important point is that many provisions that at the moment do exist as primary Treaty law should not be there.
So, while Grimm is concerned with what should and should not count as primary ('constitutional') Treaty law, Muir's concern is how much room (she argues for much more than was hitherto the case) should be left to the political process for articulating a fundamental right (equal treatment) whose existence in primary ('constitutional') law is not questioned.
Be that as it may, Muir also notes the added complication resulting from the impact of EU law on national law (the 'supranationalisation of fundamental rights 28 Muir, supra n. 1, p. 39. 29 Ibid.

30
Per Von Bogdandy, supra n. 14, we should insist in doing so. See arguments in support at p. 1328-1329, not as such presented by Muir.  32 Ibid., p. 470. 33 Ibid. discourse' 34 ) with reference to arguments by Walker 35 for a fundamental rights policy on the grounds of enhancing the Union's legitimacy; and arguments against it with reference to Kennedy, 36 Ruzza 37 and others on the basis of its potential divisive effects.
While EU intervention is said to be particularly intrusive at domestic level and claims to diversity at that level need to be taken seriously, Muir argues that EU law's 'infrastructure' does not make that easy. The EU legal order (unlike the ECHR system) is not subsidiary and she considers the principle of subsidiarity ill-suited to regulate EU legislative intervention in fundamental rights matters. According to Muir, subsidiarity, while to some degree useful, cannot effectively address the concern for maintaining diversity on fundamental rights at the domestic levelfor two reasons: first, because 'Art. 5(3) TEU relies on the assumption that the principle articulates the relationship between the EU and the Member States in a transnational context'. 38 That makes it, for her, an inappropriate tool for those fundamental rights competences which, as is the case with equality law, are 'concerned with regulating relationships within states; they go further in deepening European integration rather than merely regulating relationships among states'. 39 Second, she argues that since subsidiarity consists of a 'two-tier comparative efficiency test' which 'is based on an assessment of the effectiveness of the law to pursue a pre-established objective' 40 such test cannot address what fundamental rights standard-setting is about, namely prioritising and balancing values. She states 'key conflicts on the definition of fundamental rights standards [ : : : ] cannot be solved by comparative efficiency tests'. 41 It is easy to agree with Muir that at present subsidiarity may not be a tool with teeth in ensuring respect for the local level because of the Court's well-known reluctance to seriously review or strike down legislation for violating this principle, 42 34 Muir, supra n. 1, Ch. 2.C. 35 Ibid., p. 45, referring to N. Walker, 'Human Rights in a Post- which inherently asks for considerable political discretion. However, the two arguments noted above which she presents for rejecting subsidiarity's role in this context on conceptual grounds may be contestable on the following grounds: first, the text of Article 5(3) TFEU does not indicate the need for a 'competence with a crossborder component', 43 and it is clear that it applies to all shared competences. Therefore, it is arguable that subsidiarity could assume different meanings when it is applied to legislation that articulates values at EU level, such as is the case with equality law, as opposed to harmonising legislation of national regulatory regimes such as that of the internal market (which may also or incidentally articulate values at EU level). Second, and relatedly, subsidiarity in this context can take a role different from the kind of comparative efficiency applicable to the harmonisation of national regulatory regimes and in any case does not have to address conflicts of values. Rather, in relation to legislation that articulates values at EU level (and therefore also prioritises or mediates between conflicting ones) subsidiarity can question the detail in which matters should be set out at the EU level or left to the national level for local experimentation. So, subsidiarity may not be 'ill-suited' as a tool, but of course there does remain the question of how to operationalise it in practice.

I    '-'
Muir starts her next chapter by mapping the fragmented field of EU equality law to demonstrate the problem of 'over-constitutionalisation'. Subsequently, she 'investigates options for judicial interpretation to be able to accommodate the political dimension of fundamental rights policy-making', 44 in order to draw lessons.
We are led here through seminal cases on EU equality law starting with Defrenne II 45 where the Court famously pronounced the horizontal direct effect of what is today Article 157(1) TFEU. In doing so the Court is said: (1) to have placed this article and itself in a central position on the development of EU equality law; (2) to have 'revealed the transformative mandate of EU equality law' 46 because of the direct regulation of private relationships; and relatedly (3) to have increased the potential for EU induced change at the national level through domestic litigation between private parties based on EU rights. The subsequent codification of legal concepts defined in sex and nationality 43 Ibid. discrimination case law is said to have forestalled political debate on the issue, 47 leading to a 'political entrapment'. While we gain here an impression of this case as a seminal early example of the type of 'over-constitutionalisation' of EU equality law that Muir laments, she seems to somewhat row back from that in the epilogue to this chapter. This occurs in response to an anonymous reviewer of the book proposal arguing that her suggested approach in favour of non-or de-constitutionalised approaches to EU equality law could 'call into question the entire edifice of EU equality jurisprudence'. 48 In negating this, Muir seeks to distinguish more recent equal treatment case law from Defrenne II based on how the Court must have perceived Article 119 EEC with reference to its 'tone' in this ruling. She argues that the article was regarded as a roadmap for further integration enforcing pre-existing political commitments; it was seen as 'an ordinary regulatory tool[] and not a fundamental right'. 49 It was also self-executing (Muir notes here that 'the absence of a legal basis for the adoption of further political guidance in the relevant section of the Treaty was used to support the idea that the said article conveyed a clear legal commitment' 50 ), and the retroactive effect of the judgment was excluded. Additionally, for Muir the stage of European integrationhow far it has progressedis determinative of how cautious one should be towards a 'constitutionalisation' approach. While she does not specify the precise point of integration at which such an approach would become problematic, for her the current stage has reached such a level of sophistication and depth that 'political guidance on EU fundamental rights law is particularly strongly warranted'. 51 And yet at this point the reader may be left wondering what to make of Defrenne II and its legacy. After all, according to Muir's own analysis, Defrenne II and subsequent case law echoes into our times; and she does use the argument of 'political entrapment' to find problems of over-constitutionalisation in contemporary case law where the Court actually refers to legislative content in cases such as Kücükdeveci. 52 Mangold 53 and Kücükdeveci are indeed the next important step in this account. It follows from them that a fundamental right/general principle as fleshed out in a Directive can be invoked in a horizontal situation. Muir notes that the Court 'relies on secondary legislation combined with the General Principle of non- 47 Ibid., p. 82. 48 Ibid., p. 106. 49 Ibid., p. 108. The fact that in cases such as Mangold and Kücükdeveci the Court is actually relying on the politically debated (legislative) content of a Directive when applying the constitutional principle is not considered by her to undo the problem for two reasons. First, with regard to the 'political entrapment' argument mentioned above: that the concepts contained in the equality legislation at play in these cases are partly derived from or influenced by 'constitutional case law' relating to sex equality law. Second, that it will be difficult to amend the legislation due to Article 19 TFEU's unanimity requirement 'and the content of any reform may have to be checked against the content of the old directive, which is informative as to the substance of the constitutional right to equal treatment'. 55 Muir derives this last point from Advocate General Trstenjak's Opinion in Dominguez, where she called thisand Muir borrows the terman 'ossification' of legislative content. 56 It is important to note on this point, however, that after Kücükdeveci, and the publication of this book, the case law has developed so as to de-couple the Charter right (or the general principle) from the relevant directive when applying it in horizontal situations. The first indications of this were to be found in AMS, 57 but the cases Egenberger 58 and later Bauer and Broßon 59 mark the 54 Muir, supra n. 1, p. 84. 55 Ibid., p. 87. changed approach. In these later cases the Court clarified 60 that it is only the Charter right (primary law) that has horizontal direct effect, and not the directive or the Charter right in combination with the directive, provided the Charter right fulfils the conditions that it is mandatory and unconditional in nature. So, the Court 'le[ft] the Directive aside entirely' and found that 'Article 31(2) of the Charter, in and of itself, had the effect of limiting the Member States' discretion to retroactively remove the enjoyment of the right [at issue]'. 61 The function of the directive is considered to merely act as a 'pull factor' 62 bringing the situation within the scope of EU law. Ascribing such function to directives when considering the question of the applicable law leads to doctrinal difficulties. 63 But does a case like Bauer and Broßon lead to overconstitutionalisation in the sense that the legislature's room for further specifying an abstractly formulated constitutional right is overly constrained because the Court relies on and ties it too closely to primary law? The answer is arguably in the negative for three reasons.
First, in this case the Court did not elevate a (detailed) legislative provision to the constitutional levelbut applied only the content of a Charter right that already existed at that level. This avoids the 'ossification of legislative content', because the two sources are not conceptually interchangeable.
Second, there is overlap between the Charter and the Directive at issue (Directive 2003/88 64 ) but that overlap is inherent in the Treaties. The Explanations to the Charter, which the Court duly took into account (per Article 52(7) CFR) state that the right to paid annual leave in Article 31(2) CFR was basednext to the European Social Charter and the Community Charter of the Fundamental Social Rights of Workersalso on Directive 93/104/EC, 65 which was codified in Directive 2003/88 and its provision on paid annual leave at stake here 'reproduced the terms of Article 7 of Directive 93/104 exactly'. 66 Third, the Court applies a Charter right that it is qualified as an essential principle of social law, mandatory in nature and unconditional. It is mandatory because it articulates the right in mandatory terms and not, for example, referring to conditions provided for by Union law and national laws and practices. 67 It is 60 unconditional because it does not need to be given concrete expression by provisions of EU law or national law; those could only specify 'the exact duration of annual leave and, where appropriate, certain conditions for the exercise of that right'. 68 In that sense, the Court applies the core of the right that the legislature cannot deviate from.
To be sure, there are still constitutionalisation forces at play; they relate to what can be loosely termed the 'constitutionalisation of private law' by extending the application of certain Charter rights to private parties, as the Court has done already with equal treatment in Kücükdeveci and in Mangold (with respect to the general principle). 69 Turning now to Muir's investigation of how judicial interpretation could accommodate the political dimension of fundamental rights law-making, she first discusses cases where room for the political is said to be limited by the judiciary, and subsequently those which demonstrate the Court's ability to resist interpretations that lead to 'over-constitutionalisation'. The aim is to draw lessons from both.
The first cases show two types of clashes between the judicial and the political levels: one that Muir attributes to 'the tone used by the key players' 70 and one to 'the level at which the debate was located'. 71 Starting with the latter, she discusses the Barber 72 saga and the case law on 'positive action'. The early seminal cases Kalanke, 73 Marshall 74 and the later Briheche 75 case are highlighted. With the Barber saga Muir wishes to illustrate the 'reluctance by treaty-makers to reverse the Court's case law framed in constitutional terms'. 76 She explains how the Court provided a different interpretation of Article 157(1) TFEU than that which the legislature had assumed in a Directive pre-dating the case, but what she emphasises is that the member states 'did not [subsequently] use the Maastricht Treaty as an opportunity to overrule 68 Ibid., para. 85. Barber. They limited themselves to expressing their discontent in an explanatory protocol' (by pushing the effects of the judgment 40 years down the line). 77 For Muir this demonstrates the Treaty drafters' reluctance to overrule the Court 'presumably out of fear of unsettling the institutional balance within the EU'. 78 The point on the Treaty makers' reluctance to overrule the Court stands; but as to the potential motives ascribed to it for doing so, it is not so obvious why the institutional balance would be upset given that the role and mandate of the respective actors appears to be intact. After all, the Court's role is to interpret (primary) law and the Treaty drafters' role is to draft or alter the Treaties. 79 The 'positive action' case law is used to illustrate 'the Court's reluctance to respond to relevant changes at the constitutional level'. 80 Muir explains that the starting point for positive action cases in the early days was the Court's strict interpretation of Article 2(4) of Directive 76/207 81 (allowing member states to take measures to promote equal opportunity for men and women, in particular by removing existing inequalities which affect women's opportunities). With the Treaty of Amsterdam, a new Treaty provision was inserted, what is now Article 157(4) TFEU, which Muir states 'unquestionably call[ed] for the Court to accept a wider range of domestic positive action measures'. 82 Muir finds it noteworthy that in the case law after this Treaty change, such as Briheche, the Court 'kept its reasoning on the relevant treaty provision to the strict minimum'. 83 Here one does wonder whether this is sufficient evidence for making the argument that the Court is reluctant to respond to changes made at the constitutional level, especially after Muir notes that a reason for the 'discrete response' of the Court might be that it considers 'its post-Kalanke case law to be in line with the new Article 157(4) TFEU'. 84 As a more general point, the reader will note that the interplay focused on in these two sets of cases is not one between the constitutional version of the right and that set out through secondary legislationthe primary focus of the book. Rather, it is between the constitutional version of a right as manifested in the Court's interpretation of secondary legislation, and the constitutional version of a right as subsequently set out by the Treaty drafters. To the extent that we are looking at that interplay, Muir tells us that both the Court and the 77 Ibid., p. 91. 78 Ibid.  82 Muir, supra n. 1, p. 93. 83 Ibid., p. 95. 84 Ibid., p. 94.
Treaty drafters have an interest in not 'acknowledging the political dimension of fundamental rights law-making'. 85 The former because it 'would amount to a significant loss of influence as the protection of fundamental rights has historically been in its hands'. 86 The latter because 'by intruding more explicitly in the Court's case law ( : : : ) [they] may affect the authority of rights that they are intending to place above ordinary law-making processes'. 87 This last point may be true, but presumably only if the Treaty drafters were to consistently and systematically alter the Court's position on fundamental rights.
There is a further important point to note: as Muir herself acknowledges, both lines of case law (the Barber saga and the positive action case law) are actually 'devoid of references to fundamental rights'. 88 As a result, these cases do also not actually illustrate the argument that the couching of issues in fundamental rights language will be a determinant for diminishing the scope of political deliberation on the content of the relevant rights. Muir considers them, however, still useful to demonstrate that even overturning the Court's position through constitutional Treaty amendment may be difficult to achieve.
Muir's second example of a clash between the judicial and political level is illustrated by Test-Achats. 89 It is a clash between the legislature and the Court and this time both actors did 'express[] themselves in fundamental rights terms'. That, combined with the fact that both had a 'constitutional mandate' (the Court to protect the fundamental right to equal treatment and the legislature to give expression to it) are said to have led 'to a head-on collision', 90 where the Court found Directive 2004/113/EEC 91 'to breach the same higher ranking principle of equal treatment to which it gives expression'. 92 Muir relies here on Davies, to articulate the 'lesson' 93 that the legislature should change its 'tone'. It should emphasise political considerations, rather than principled arguments based on fundamental rights, to increase its chances for a wider margin of manoeuvre granted by the Court. 94 85 Ibid., p. 95. 86 Ibid., p. 95. 87 Ibid., p. 95. 88 Ibid., p. convincing to see this change as being directly linked to the introduction of the Union Citizenship Directive.

L 
Muir distils from the overall discussion indicators for 'what influences the readiness of the EU judiciary to adopt [an approach that gives more room to the legislature]'. 104 These factors are the concrete 'lessons' that she draws. They are four. First, the text of the Treaty provisions at stake is one that 'clearly identified the need for further political guidance'. 105 Second, that 'the secondary legislation ( : : : ) relied upon has a strong organic link with the relevant treaty provisions', which 'may make it easier for the judiciary to shift from one level of analysis [primary law] to the other [secondary law]'. 106 Third, the quality of the legislation, e.g. whether it provides for legal certainty or passes the proportionality test in both its substantive and procedural form. Fourth, 'the tone chosen by the legislature', 107 meaning that the Court will be more deferential to the legislature if the latter makes the policy dimension of its decision clear.
Based on Muir's own observations, the first two factors do not strike as indicators that have a clear explanatory value. Regarding the first factor, that the Treaty provisions clearly identify the need for further political guidance, Muir points out that Articles 19(1) and 157(3) TFEU amount to 'comparable invitations to defer to secondary legislation despite case law going in the opposite direction' (the Mangold-Kücükdeveci case law). As regards the second factor, concerning a strong organic link between the secondary legislation and the relevant Treaty provisions, Muir acknowledges that this can go either way as the case law on age discrimination on the one hand and citizenship on the other demonstrate. 108 The third factor concerning the quality of legislation, e.g. whether it provides for legal certainty or passes the proportionality test in both its substantive and procedural form, persuades as indicative, but this does not seem surprising as these would be criteria that determine the legality of all secondary legislation, irrespective of whether it is concretising fundamental rights. The fourth factor, concerning 'the tone chosen by the legislature', 109 is for Muir 'perhaps the more important' one. 110 As discussed above, this may be a factor that influences the Court's deference (even if, in my opinion, it cannot be derived from the Test-Achats case).
In her conclusion Muir goes from the perceived 'empirical' indicators to the normative. She is concerned with giving more room to the political rather than the judiciary on fundamental rights law-making (to create a more 'healthy balance' 111 ). Thus, 'the constitutional norm itself ought to call explicitly for political guidance'; the legislature should not adopt a principled 'tone' but provide for clear and coherent rules whereas, the judiciary 'may have to refrain from palliating limits or imperfections resulting from the policy-making process'. 112 Apart from the question of the true indicative value of all the factors listed, the question arises to what extent the points raised are particular to EU equality law with all the distinctiveness ascribed to it in earlier chapters or whether they can be transposed to other areas. Muir picks this up in the next chapter.

T ?
The key question in Chapter 4 is 'to what extent could problems similar to those identified [thus far] in the context of EU equality law arise in other contexts?' 113 Muir answers by putting forward 'guidelines' for distinguishing between 'legislation giving expression to fundamental rights' 114 on the one hand and 'ordinary legislation' on the other: if we are concerned with the former (in a field outside equality law) then similar problems may arise. For Muir, those problems do not merely relate to 'the ability to invoke directives in private disputes; it may also affect the conditions for judicial review [of new legislation attempting to change the previous policythe problem of 'ossification' of legislative content discussed above] and more broadly the legitimacy of EU intervention in the field'. 115 Muir's main concern is thus to limit exportation of 'the Kücükdeveci effect', 116 by confining it to 'fundamental rights legislation', which for her is not the same as 'an overlap in subject matter between legislation and primary rights'. 117 But as noted above, the Court's case law has progressed since the publication of this book so that the derived guidelines for limiting 'constitutionalisation' may to a certain extent become perhaps less relevant today. We shall nonetheless note her criteria and how she applies them to distinguish the relevant legislation thereby giving the Court tools for maintaining a restraint approach.
The guidelines distinguishing fundamental rights legislation from ordinary legislation are derived from case law contrasting, in particular, EU equality law (Kücükdeveci, Mangold) with social law (Dominguez, Fenoll, 118 AMS). Here again, Muir moves from the descriptive to the normative. She first searches for distinguishing characteristics in the two sets of case law, then she formulates the same ones into guidelines and finally submits that 'the Court does well in distinguishing ordinary legislation-making from fundamental rights legislation making based on these criteria'. 119 These are the guidelines: (1) the existence of a fundamental right protected by EU constitutional law, meaning it should be protected by primary law; (2) this fundamental right should be substantiated through legislation, i.e. 'there must be a subject matter overlap between the primary and secondary right'; 120 (3) 'this overlap ought : : : to be intended to flesh out the fundamental right at hand'. 121 This last condition she calls sometimes 'a specific relationship' 122 or an 'organic link' between the legislation and the primary right.
Muir first applies them to equal treatment clauses 'which are scattered across EU directives designed to ensure the protection of atypical workers as well as third country nationals'. 123 She concludes here that these 'should be treated as clearly distinct from the primary right to equal treatment'. 124 The conclusion is different when the same criteria are applied to data protection law and more specifically the General Data Protection Regulation (GDPR). 125 The application of the first two conditions is straightforward and can be answered in the affirmative in this case; but the third conditionthe more hazy oneis also said to be met, for three reasons: first, the location of its legal basis (Article 16 TFEU) in the treatiesit is placed under 'provisions of general application' and even before non-discrimination; second, the actual wording of the legal basisthe right to data protection is set out therein, and in the Charter, and the market objective maintained in Article 16 TFEU is not considered an obstacle; third, its ability to produce direct effect. Muir rightly notes that the Kücükdeveci approach is not relevant to this area because we are concerned with a Regulation that already has horizontal direct effect, and the same is likely true for Article 16 TFEU.
Looking at the current data protection regime, Muir does see parallels with equality law when relying on an analysis by Lynskey concerning 'uncertainties in the relationship between the two sets of rights [constitutional versus legislative]'. 126 The reason is that the GDPR can be reviewed based on Article 8 CFR, which in part refers to the earlier Data Protection Directive 127 as its basis and may be defined by reference to that. 128 For Muir this ' : : : may limit the possibility for EU political institutions to redesign a fundamental rights policy through legislation'. This scenario is, however, not the same as the problem of 'ossification' of legislative content identified earlier in the Mangold/Kücükdeveci scenario. In the former it is driven by the Court; in the latter it was the choice of the Charter/Treaty drafters to partly base their understanding of the right to data protection on fundamental principles laid down in the Data Protection Directive.

C
This book does well to revive the theme of EU fundamental rights policy-making and to investigate fundamental rights in legislation. It provides and invites needed contemplation on the interplay between the legislature and the judiciary in the protection of equality law, and fundamental rights more generally, and the identified problem of 'over-constitutionalisation', especially as the case law is developing in this field.
Vasiliki Kosta is Associate Professor, Europa Institute -Leiden Law School, Leiden University, The Netherlands.