Constitutional pluralism and loyal opposition

Constitutional pluralism has long been controversial, but has recently come under renewed attack. Critics allege that by justifying departure by national courts from the Court of Justice of the European Union’s (CJEU) orthodoxy on the primacy of EU law, constitutional pluralism is (at best) susceptible to abuse by autocratic member state governments, or is (at worst) a favored tool of Europe’s new authoritarians. This article is a defense of heterodoxy in European constitutional thought, and of constitutional pluralism in particular. It uses the concept of “loyal opposition” as a framing discourse, allowing us to see that heterodox approaches to EU constitutionalism and opposition to the received and dominant interpretation of the primacy of EU law are not necessarily any less “loyal” to the principles and values of European integration than agreement with the CJEU. A “legitimacy test” is proposed, by which we can determine whether a given instance of national judicial disagreement with the CJEU is loyal, principled opposition, or disloyal, abusive opposition.


Introduction
Constitutional pluralism (CP) is the umbrella term for a range of theories which all share a core descriptive and normative thesis: that the legal systems of the European Union and of its member states are best regarded as being arranged in a conceptual heterarchy, where no one system is normatively superior to the other, rather than in a hierarchy, where one system-European or national, depending on one's preferences-must be supreme.
This has always been controversial, but recent developments have sharpened the tone of the debate. Whereas critics of CP have long sought to rebut either or both of the idea's twin claims to descriptive accuracy and normative desirability, and have gone so far as to describe it as a threat to the rule of law, 1 the ongoing slide of Hungary 2 and Poland 3 into what GM Tamás described long ago as "post-fascism," 4 along with the potential for Italy, France, and many others to follow suit, has lent new urgency to the critique. Now, CP is castigated as a beloved tool of autocrats, 5 and critics argue that the actions of Fidesz in Hungary and Law and Justice (PiS) in Poland, and of the puppet courts that they have captured or created in their countries, prove the longheld suspicions of the sceptics that CP is an intolerable deviation from the Court of Justice of the European Union's (CJEU) account of the primacy of EU law. 6 As Daniel Kelemen puts it, [CP] should be abandoned by all those who value the survival of the EU legal order and of the European Union itself. Scholars should embrace the CJEU's straightforward and compelling approach to the question of supremacy: For those states that voluntarily choose to join and voluntarily choose to remain members of the Union, EU law, and the Court of Justice as the ultimate guardian of that law, must enjoy unconditional supremacy. 7 This analysis presents us with a clear and attractive binary: the absolute "supremacy" 8 of EU law or barbarism.
In this article I argue that there are strong reasons why those of us who, as Kelemen puts it, "value the survival of the EU legal order and of the European Union itself " should not abandon CP, and should instead defend the idea from the charge that it is a dangerous enabler of autocracy within EU member states. I do so by proposing that the notion of loyal opposition might usefully be transposed from the domain of politics to that of law and legal scholarship, and adopted as an analytical framework through which we can conceptualize different kinds of "opposition" to the absolutist conception of the primacy of EU law. 1 explain the critics' wholesale rejection of CP, and that an embrace of the concept of loyal opposition would have important and beneficial knock-on effects on EU law and legal scholarship, as we shall now see.

The structure of the argument
Just as the lack of a loyal opposition in the European political sphere leads to the catastrophization of disagreement into existential questions of "pro-Europeanism" or "Euroscepticism," intolerance of reasonable national court objection to the application of the primacy principle in particular cases catastrophizes legitimate legal dissent into constitutional crisis. In Section 2, I interrogate this case law, and propose a framework, drawn in part from the jurisprudence of the CJEU, whereby we can learn to distinguish legitimate, "loyal" opposition in European constitutional relations from abusive, 15 "disloyal" opportunism. In doing so, I analyze five particular cases of national rejection of CJEU jurisprudence: two cases of loyal opposition; two cases of disloyal opposition; and one that is less straightforward. Close analysis of this case law reveals the high degree of asymmetry between different instances of judicial opposition in EU law, and thus the poverty of any discourse that explicitly or implicitly bunches these disparate histories together as instances of some ill-defined "neo-sovereigntism," 16 or uses the latter cases as a reason to abandon the entire discourse of CP.
In Section 3, I will argue that far from being a threat to the integrity of EU law or to the existence of the Union, legal and constitutional conflict between the European Union and its member states can be a good, useful, and justifiable thing in particular political and legal contexts, serving as a kind of animating tension, analogous to the distribution of normative power across different institutions-executive, legislative, and judicial-at state level. Moreover, preserving the legitimate possibility of such an animating tension may yet become painfully necessary, if the widespread belief in the European Union as a bulwark against authoritarianism proves to be mistaken, as I fear it yet might. In the actually existing context of imperfect member states in an imperfect Union, scholars-like judges, politicians, and citizens-should be hesitant to accept the claims of one constitutional order over another uncritically, and equally hesitant to accept the equation of heterodox thinking with disloyalty to the claimed values of the European project.

From opposition to EU law to opposition in EU law
The reduction of European politics to the "pros" and the "antis" is mirrored in EU law, whereby any deviation from CJEU orthodoxy by national actors gets dramatized and cast as a fundamental threat to the unity and integrity of the EU legal order. It is of course true that a politician or citizen arguing against Union policy is not the same 15 David Landau, Abusive Constitutionalism, 47 u.c. davis L. rEv. 189 (2013). 16 Fabbrini & Sajó, supra note 6, at 458. thing as a court refusing to follow the CJEU's interpretation of the requirements of EU law. But the kinds of instances of legal opposition that the CJEU has faced need not be catastrophized in this way. As we shall see, they are sometimes rather ordinary legal disputes that the overconstitutionalized 17 nature of EU law transforms into constitutional crises. The CJEU's orthodoxy generally gives hierarchical superiority in cases of conflict to all EU law, including EU secondary law, over all national law, including national constitutional law. 18 The result is that not only are the EU Treaties themselves given constitutional status in their entirety, but in the member states, any piece of EU law, from the most fundamental to the most pettifogging, must be treated as if it were the bedrock of the legal system, and as normatively superior to the deepest statement of the national polity's self-image. Certainly the CJEU's longstanding arguments for its conception of primacy are good ones: legal certainty, uniformity across the European Union, and the principle of sincere cooperation 19 are important, and member states cannot be allowed simply to ignore any aspect of EU law they happen to dislike. But however strong, these arguments are not insuperable in every set of circumstances, and it is for precisely this reason that in hard cases many national courts have been rather more heterodox in their interpretation and application of the primacy principle.
In this section, I will examine five examples of oppositional practice by national courts, emphasizing their diversity as regards origins, impact, and resolution, but also as regards their political orientation: what were these national courts seeking to achieve by their actions; what vision of the constitution and of the polity does this reveal; and were the national courts engaging in legitimate loyal opposition, or abusive opportunistic illegality? In this respect, the analysis looks not merely at legal text but at political subtext. In order to answer these questions, we must first set out some preliminary criteria against which we can measure the courts' behavior.

The legitimacy test
Usefully, the CJEU has already done some of the work for us. In the case of LM, 20 the CJEU held that a national court may be entitled to refuse to execute a European Arrest Warrant if, "on the basis of material that is objective, reliable, specific, and properly updated concerning the operation of the system of justice in the issuing Member State," 21 there is a "real risk," connected with a lack of judicial independence, of the right to a fair trial being breached.
Drawing on this, I propose that if there is good evidence that an apex court in a given member state should not be regarded as a properly independent judicial body, this is strongly persuasive evidence that its invocation of constitutional identity, or any other "interface norm" 22 -such as fundamental rights or the principle of conferral-to justify departure from EU law should be regarded as illegitimate and abusive-an instance of "disloyal opposition." Such evidence could be drawn, for example, from the work of the Venice Commission, the Helsinki Committees, or other respected bodies with expertise on judicial independence, the rule of law, democracy, and human rights, along with the work of scholars. Similarly, but in reverse, a verifiably independent judicial body governed by-and in turn applying-the rule of law is entitled at the very least to be given a fair hearing when it holds that certain factors, whether these be issues of fundamental rights; constitutional identity; the question of Kompetenz-Kompetenz; or any other interface norm, require it to disapply or modify any norm of EU law insofar as the national constitutional order is concerned.
But this is only the first-institutional-element of the test. The second elementwhich is substantive or analytical-concerns the quality of the reasoning employed. By this I do not only mean justifiability and cogency, but what Miguel Poiares Maduro long ago called "horizontal and vertical coherence." 23 Relevant factors here include, but are not limited to, whether the national decision is grounded in persuasive engagement with common European and international standards; serious engagement with EU law and with the jurisprudence of the CJEU; good-faith attempts to enter into formal dialogue with that court through the procedure under Article 267 of the Treaty on the Functioning of the European Union (TFEU); 24 and whether or not the decision evinces an image of the national polity as one member state among many-co-equal to all the others, and deserving of neither less nor more concern and respect than any other-consistent with the equality of the member states under the EU Treaties. 25 This two-legged "legitimacy test" is by no means exhaustive, but it allows us at least to begin developing a more sophisticated approach to judicial pushback against the unqualified "supremacy" of EU law than outright rejection of all such action on the mere grounds that this is how the pet courts of autocratic governments behave. Much like a stopped clock being right twice a day, even the most obviously captured kangaroo court might sometimes hand down a justifiable and cogent judgment, and thus fail the first element of the test, but pass the second; and even a court of the most impeccable constitutionalist and democratic credentials might deliver a judgment that is poorly reasoned or solipsistic, and thus pass the first element of the test, but fail the second. With this legitimacy test in mind, let us turn to the case law. 22 See nico krisch, bEyond constitutionaLisM: thE pLuraList structurE of postnationaL LaW 285-6 (2010); toM fLynn, thE trianguLar constitution: constitutionaL pLuraLisM in irELand, thE

Legitimate, loyal opposition a) The Taricco saga
The Corte Costituzionale (Italian Constitutional Court, ICC) has long maintained that it has the right to prioritize the Italian Constitution over EU law where it believes EU law to be in breach of fundamental constitutional principles. 26 However, this claimed right to depart from EU law, including CJEU judgments, is tempered by a generally accommodating attitude. 27 The Taricco dialogue 28 of the CJEU and the ICC demonstrates exactly the kind of legitimate, "loyal" oppositional practice that should be regarded as an inherent and desirable component of the interplay between Union law and national law, and not as some pathological outlier, or as succor for authoritarians. The first element of our legitimacy test can be dealt with briefly: the ICC is an independent body, against which no serious allegation of executive capture can be maintained. The second element requires close examination of the ICC's reasoning.
The saga began with a preliminary reference to the CJEU from an Italian court dealing with a criminal case of value-added tax (VAT) fraud. 29 The referring court asked whether the national period of limitation for proceedings such as these was so short as to amount to de facto impunity, contrary to EU law. 30 The CJEU held that if, "in a considerable number of case[s]," the time limit would result in "the commission of serious fraud [escaping] criminal punishment," 31 the EU Treaties would require the disapplication of the relevant national statute of limitations. 32 However, the application of the CJEU's Taricco rule gave rise to constitutional difficulties in subsequent cases. For the ICC, the Taricco conditions were insufficiently precise to satisfy the Italian conception of the constitutional principle of legality, and, in any event, required the judiciary to involve itself in matters of criminal policy, contrary to Article 101(2) of the Constitution's requirement that "[j]udges are subject only to law." Accordingly, [I]t is necessary to ask whether the [CJEU] took the view that the national courts should apply the rule even where it conflicts with a supreme principle of the Italian legal system. This Court thinks that it did not, but considers that it is in any case appropriate to bring the doubt to the attention of the [ Here we see a loyal, reputable national court suggesting an answer to its question that would satisfy both the domestic and European legal orders. Elsewhere in the Order, the ICC makes clear its cooperative-but watchful-approach to EU law: The recognition of the primacy of EU law is an established fact within the case law of this Court pursuant to Article 11 of the Constitution; moreover, according to such settled case law, compliance with the supreme principles of the Italian constitutional order and inalienable human rights is a prerequisite for the applicability of EU law in Italy. 34 In this way, the ICC politely suggested that the Luxembourg court may wish to "clarify" (that is, "correct") the rule in Taricco I.
The CJEU obliged in Taricco II. 35 In answering the ICC's questions, the CJEU essentially recited the reasoning in Taricco I 36 before backtracking: now, rather than automatically requiring judicial action, "[i]t is primarily for the national legislature to lay down rules on limitation that enable compliance with the obligations under [the Taricco rule]." 37 With just a few words, the CJEU entirely altered the thrust of its previous judgment, and seemingly admitted that it had asked too much of the Italian courts-and, by extension, all national courts. I argue that the CJEU was aided in coming to this conclusion by the ICC's prior statement that even if it were to have held the Taricco rule unconstitutional, this would not "alter the liability of the Republic of Italy for having failed to provide an effective remedy against serious tax fraud affecting the financial interests of the Union." 38 The ICC, having recognized that something had to be done to bring Italian law into line with the CJEU's requirements, was merely pointing out to the CJEU the intense constitutional difficulty in Italy of the judiciary doing the heavy lifting. This surely made it less difficult for the CJEU to shift the burden from the judges to the legislators: the ICC's acknowledgment of potential Italian state liability cast the ICC not as a recalcitrant external actor, but as a loyal Union court opposing the CJEU's interpretation of the Treaties for justifiable reasons of legal and constitutional integrity.
In resolving the specific problem in casu, the CJEU spoke in a harmonious mode of how there is nothing specifically Italian about the principle of legality in criminal matters, which forms part of the well-worn "constitutional traditions common to the Member States." 39 It went on to note that the principle's requirements of foreseeability, precision, and non-retroactivity would preclude Union law from requiring the disapplication of the statute of limitations in cases where the alleged offence occurred before the judgment in Taricco I was handed down. This being the case with the offences at issue in Taricco II, the CJEU seems to have hoped that the issue of constitutionality would no longer arise. 34 Id. at 3.  The final instalment in the saga came in judgment 115/2018. 40 One might be forgiven for imagining that the CJEU's backtracking would be the end of matters. 41 However, the ICC went much further in its response than simply acknowledging that the CJEU had held the Taricco rule inapplicable to events that predate it, and went on to hold that it could never be applied in Italy. Here, we must turn to the second element of the legitimacy test. The court's reasoning is principled and compelling: to hold that the alleged crimes were time-barred because they took place before Taricco was decided would still be "applying" the Taricco rule, no less than if it were held that the crimes were not time-barred because they post-dated the rule's creation. 42 The overall thrust of the ICC's judgment is that the Taricco rule is so vague that even if it were to "take on a less hazy outline" over time, 43 this would not compensate for its initial underspecification. The rule, being incompatible with a fundamental constitutional principle, therefore never entered the Italian legal system through the domestic provisions providing for the effectiveness of EU law, which only import EU law to the extent that it is compatible with the constitution. It is for the Italian legislator, the ICC held, to ensure that Italy honors its obligations to the European Union by making the law compliant with the EU Treaties: such changes to the statute of limitations must be legislative in nature, of exclusively prospective effect, and clear enough to satisfy the principle of legality.
Chiara Amalfitano and Oreste Pollicino argue that the ICC and CJEU are speaking different "languages" here; 44 but it is entirely unsurprising that two differently situated courts in a legal heterarchy might seek to achieve similar (or at least compatible) results by different means. The point is one of institutional orientation and epistemological viewpoint. It is in the nature of the CJEU, as a supranational body, to seek to emphasize the commonalities of the different legal systems the coordination of which is part of its jurisdiction and the infiltration of which (I use the word literally and non-pejoratively) is part of its mission. It is in the nature of the ICC, as a national apex court, to seek to preserve the sphere of its own jurisdiction and to preserve the integrity of the national constitutional order. These are valid approaches. Moreover, they are not mutually exclusive. As the ICC made clear, the primacy of EU law really is part of the Italian constitutional order: it is just not its only or main principle. Similarly, respect for national constitutional identity-whether scholars like it or not, on which more below-really is part of the Union legal order: it is just not its only or main principle. It is not that the Italian court is trying to dodge EU law, it is that the court-a reputable, independent judicial body-had specific, principled, and justifiable objections to the incompatibility with EU law being remedied in the manner proposed-or rather  43 Id. at 6. 44 Amalfitano & Pollicino, supra note 41. demanded-by the CJEU. The correct approach, for the ICC, is legislative. If this leaves Italy in breach of the Treaties pending such legislative change, then so be it: such is the price of the preservation of the integrity of the constitutional order.
Such conflict or disagreement-such loyal opposition-is a justifiable and desirable part of a composite legal order like that of the Union and its Member States, and this is all the more so in the context of an overconstitutionalized and insufficiently politicized order. The CJEU-and scholarship generally-does not currently recognize behavior such as the ICC's in this context as loyal opposition. It should.

b) The Ajos case
A comparable case of principled, justifiable, loyal opposition to the CJEU's interpretation of the requirements of EU law was in Denmark, where the Højesteret (Danish Supreme Court, DSC)-a body no less independent and reputable than the ICC-explicitly refused in Ajos 45 to disapply a national legal provision which conflicted with the general principle of EU law prohibiting age discrimination.
Danish law provided that long-serving salaried employees would receive a severance allowance on dismissal. There were exceptions to this rule, which in a previous case, Ole Andersen, 46 the CJEU found to constitute direct discrimination on grounds of age, contrary to the Employment Equality Framework Directive. 47 Having been dismissed by a state body, Andersen could rely directly on the Directive, and the Danish court was required to disapply the discriminatory exceptions and order that the severance allowance be paid.
Ajos was a case was between private parties. A lower Danish court, relying on Ole Andersen, had ordered Ajos to pay the former employee the severance allowance, disapplying the exceptions not in favor of the Directive itself, but in light of the horizontally directly effective general principle of EU law prohibiting age discrimination, of which the Directive is only a concrete expression. 48 Ajos appealed to the DSC, which referred two questions to the CJEU. 49 The first question was on the substantive issue of whether the general principle precluded the discriminatory exceptions. The second question asked if, in determining whether to disapply the Danish law, the DSC would be entitled to weigh the general principle prohibiting age discrimination against the competing principles of legal certainty and legitimate expectations, and whether it  was relevant that the employee would be able to claim compensation from the state for the loss caused by the incompatibility of Danish law with EU law. The Supreme Court's preferred solution to the problem in the case was clear from its second question. Ajos was not even tangentially responsible for Denmark's failure properly to implement the Directive (unlike the public-sector employer in Ole Andersen), and had acted in good faith and in accordance with Danish law as it stood at the time. The principle of legal certainty therefore required that it not be ordered to make a payment that it had a legitimate expectation that it would not have to make. The principle that national law must be interpreted consistently with EU law does not go so far as to allow for an interpretation contra legem. 50 Consistent interpretation being impossible, and disapplication being contrary to the principles of legal certainty and legitimate expectations, the case must be decided in Ajos's favor. Of course, none of this would solve the problem that Denmark (and not Ajos) had breached EU law to the detriment of the dismissed employee. The solution, therefore, was for the employee to claim Francovich damages from the state for its inaction. 51 In this way, Ajos is not punished for an error that was not its fault; the employee is not denied, for discriminatory reasons, money to which they were entitled; and the member state is punished for failing in its obligations under EU law. EU law is upheld; the rights of both parties to the case are upheld; and all this in a manner consistent with both EU law and Danish law.
But the CJEU did not take the DSC up on its offer, instead holding fast to its Mangold 52 and Kücükdeveci 53 case law on the horizontal direct effect of the general principle prohibiting discrimination on grounds of age. The Court dealt with the issue of legitimate expectations tersely, 54 and made no substantive observations on the principle of legal certainty at all. The possibility of Francovich damages being claimed against Denmark for its breach of EU law was also not enough, in the Court's eyes, to absolve the DSC of its obligation either to interpret the national legislation compatibly with EU law, or, if this is impossible, to disapply it, though this is a conclusion the court reached without reasoning. 55 The instruction from Luxembourg was therefore as clear as it was peremptory and badly reasoned: whether by interpretation or disapplication, the DSC must order Ajos to pay the severance allowance.
The DSC declined to do so, by a majority of 8:1. Its reasoning was based closely on the specific manner in which Danish law incorporates and gives effect to EU law, and on the DSC's conception of the relationship between the two systems, which closely mirrors that of the ICC in the Taricco cases: The [CJEU] has jurisdiction to rule on questions concerning the interpretation of EU law . . . . It is therefore for the [CJEU] to rule on whether a rule of EU law has direct effect and takes precedence over a conflicting national provision, including in disputes between individuals.
The question whether a rule of EU law can be given direct effect in Danish law, as required under EU law, turns first and foremost on the Law on accession by which Denmark acceded to the European Union. 56 The Danish Accession Act allows the EU institutions to exercise the Danish state's constitutional powers in accordance with the EU Treaties. 57 However, the DSC noted that the CJEU located the source of the general principle prohibiting discrimination on grounds of age specifically not in the Treaties, but in "various international instruments" and in the constitutional traditions common to the member states. 58 For the DSC, A situation such as this, in which a principle at treaty level under EU law is to have direct effect (thereby creating obligations) and be allowed to take precedence over conflicting Danish law in a dispute between individuals, without the principle having any basis in a specific treaty provision, is not foreseen in the Law on accession. 59 The point is not whether the horizontally directly effective general principle prohibiting age discrimination is a good thing or a bad thing: rather, as with the Taricco case law, the issue is one of institutional orientation and epistemological viewpoint. The DSC held that just as it is the right and obligation of the CJEU to determine the content of EU law, it is the right and obligation of the Folketing (Danish Parliament) to determine how EU law is made effective within Denmark. The question of the interpretation of the Danish Accession Act, 60 and of whether, to what extent, and in what way this particular general principle of EU law flowed into Danish law, was therefore a threshold question exclusively within the jurisdiction of the Danish court. 61 In this particular case, the Court detected an irreconcilable tension between its duties as a national court and as an EU court, and felt itself unable, as a Danish court, to act in such a way as to go beyond the extent to which the Folketing had imported EU law into the Danish legal system. The Italian and Danish cases illustrate situations where reputable judicial institutions engaged in reasoned and principled-loyal-oppositional practice within the constitutional sphere. They did so for reasons with which it is perfectly possible to disagree: it is legitimate to argue that both the ICC and DSC were wrong, and should have followed the reasoning of the CJEU, whatever its flaws. But the decisions of the ICC and DSC do not only make sense within the internal perspectives of Italian and Danish constitutional law: they are also justifiable from within the epistemology of a European Union of member states united in diversity. They pass both elements of the legitimacy test, and must be distinguished from the cases to which we now turn.

Illegitimate, disloyal opposition
In this section, we will see that not all opposition to EU law is justifiable or loyal. The first case to be examined is one of merely misguided and poorly reasoned opposition by an independent, reputable court. It can rightly be criticized, but thankfully has had little negative impact. The second is far more disturbing, and shows a disreputable, captured institution engaging in oppositional practice for reasons contrary to the EU's values, and ultimately incompatible with continued membership in good standing of the European Union.

a) The Holubec case
The decision of the Ústavní soud (Czech Constitutional Court, CCC) in Holubec 62 was the first time a national court declared a judgment of the CJEU (in Landtová 63 ) to be ultra vires. 64 The dispute arose in the politically, jurisdictionally, and historically contingent context of the breakup of the Czecho-Slovak federation; of a treaty between the two successor states regarding the pension rights of citizens of one state who had worked for an employer domiciled in the other; and against the background of a long-running institutional turf war between the CCC and the Czech Supreme Administrative Court (CSAC) on the matter, which developed throughout the 1990s and 2000s, and became highly personalized between the members of the two courts. Czech pensions being generally more valuable than Slovak ones, the CCC held that the Czech state must top up the smaller Slovak pensions of Czech (but not Slovak) citizens living in Czechia. This requirement was contested by the CSAC, which referred the matter to Luxembourg: the question was whether the top-up discriminated on grounds of nationality contrary to EU law. Quite rightly, in Landtová, the CJEU found that it did.
In declaring the Landtová judgment ultra vires, the CCC did not ground its analysis in any specific Czech or EU constitutional principle or right that the CJEU had failed to respect. Indeed, the CCC engaged in no analysis of the CJEU's reasoning at all: the mere fact that Landtová had contradicted the CCC on the issue (and had agreed with the CSAC) was enough for the CCC to declare that the CJEU had exceeded the scope of the powers transferred to the Union under the Czech constitution. 65 The one actual deficiency that the CCC claimed to identify in Landtová was an alleged infringement of the right to a fair trial. 66 Almost comically, the "victim" of this infringement was the CCC itself, which was not a party to the proceedings in Landtová, and had attempted to write a letter to the CJEU explaining and justifying its case law, which the CJEU rightly ignored. 67 Zdeněk Kühn uses Holubec to argue that constitutional pluralism is "a flawed theory because it celebrates very dangerous outcomes for everyday cases." 68 But there is nothing for anyone, constitutional pluralist or otherwise, to celebrate about Holubec. The CCC's declaration that the CJEU had acted ultra vires was both inapt and inept, yet we can recognize this without abandoning the fundamentally sound reasons why such a declaration should be possible from the domestic perspective: the issue of Kompetenz-Kompetenz; the fact that the European Union is not a vertically integrated, complete system of legal and constitutional authority; and the institutional obligation of national courts both to enforce Union law and uphold the national constitutional order simultaneously. The simple fact is that the CCC failed in both of its obligations in Holubec: it failed to enforce EU law; and it did so for reasons that had nothing to do with upholding the national constitution, protecting fundamental rights, or policing the limited competences of the Union. The CCC's decision in Holubec is therefore not the sort of loyal opposition that I have been conceptualizing in this section. This does not mean the case is without its uses: it serves as a helpful sort of negative precedent, a foil that illustrates the importance of loyalty in opposition, and in this case, loyalty not only to the Union or the national constitution, but to the reasoned justification that courts are supposed to provide. Holubec therefore passes the first element of the legitimacy test, but fails the second.

b) Decision 22/2016
Similarly useful as a negative precedent, but far more serious, is Decision 22/2016 69 of the Alkotmánybíróság (Hungarian Constitutional Court, HCC). The HCC is a captured institution, operating in a context of democratic and constitutional backsliding. As Halmai notes, the case was cherry-picked and confected from a previously abandoned application by the Commissioner for Fundamental Rights, a similarly captured constitutional officer, in the aftermath of the Fidesz government's anti-migrant referendum of October 2016. The government having failed to secure a valid result in the referendum due to low turnout, its judicial servants in the HCC were only too happy to offer their masters a way forward. 70 The Commissioner asked two questions of the HCC. The first related to a Council Decision providing for the relocation of refugees. 71 In a display of rank intellectual perversion, the Commissioner asked whether the decision, because it proposed the 67 Id. at 14.  The Commissioner's second question was more abstract. It asked, in a general sense, whether Hungary is entitled a) to review norms of EU law for conformity with the national constitutional conception of fundamental rights or b) to examine whether they go beyond the competences of the European Union; and, if Hungary can do this, which institution of the state may carry out the review. 74 The HCC declared itself competent to perform such review. In doing so, it made reference to a wide range of national court judgments in this area, though this was done unsystematically and without real engagement with the reasoning or constitutional context of each different legal system. 75 The HCC also noted the CJEU's practice of taking account of the "constitutional demands" of the member states. 76 The main judgment is therefore quite innocent and anodyne in appearance. The HCC attempts to position itself firmly in the mainstream of national apex court practice, acknowledging the constitutionally authorized nature of national membership of the EU; the special and original nature of the EU as a new legal order; the primacy principle; and the place of fundamental rights in the EU. It also makes exactly the kind of arguments about Kompetenz-Kompetenz and the duty of national courts to police the boundaries of the powers of the Union that I have presented, in this article, as being so compelling and justifiable. In the HCC's words, [T]he Constitutional Court . . ., in compelling cases and as a resort of ultima ratio . . . can examine whether exercising competences on the basis of [EU Membership] results in the violation of human dignity, the essential content of any other fundamental right or the sovereignty . . . and the constitutional self-identity of Hungary. 77 By "constitutional self-identity," the HCC declared that: [It] interprets the concept of constitutional identity as Hungary's self-identity and it unfolds the content of this concept from case to case, on the basis of the whole Fundamental Law and certain provisions thereof, in accordance with the National Avowal and the achievements of our historical constitution . . . . 78 But context is everything. The Fundamental Law was enacted by Fidesz using its twothirds parliamentary majority in 2011, though the possibility of a new constitution had formed no part of the 2010 election debates or campaign. 79 74 Id. ¶ ¶ 12-14. 75 Id. ¶ ¶ 33-44. 76 Id. ¶ 44. 77 Id. ¶ 46 (footnotes omitted). 78 Id. ¶ 64. 79 On the Hungarian Constitution generally, see constitution for a disunitEd nation: on hungary's 2011 fundaMEntaL LaW (Gábor Attila Tóth ed., 2012). the constitution was enacted "without any elementary political, professional, scientific or social debates." 80 There was, prior to the enactment, a "national consultation" involving the sending out of questionnaires to every voter, but the results of this exercise have never been made public. 81 The "National Avowal" to which the HCC refers is the Fundamental Law's Preamble. It is phrased, in many respects, in the kind of aspirational, solemn, and self-reverential language that one finds in such preambles across Europe and around the world, though it lays it on rather more thickly than most. It speaks of Hungary being part of "Christian Europe" for a thousand years; of Hungary having "defended Europe" "over the centuries" (an obvious reference to Hungary's historical place on the front-line between the Ottoman Empire and the empires of Europe); and of "honour[ing] the achievements of our historical constitution." 82 The HCC's attempt to dress its judgment in the borrowed robes of legitimate constitutional adjudication elsewhere in Europe is therefore complicated by the context in which it operates. It is a clever attempt, and one which might fool the unwary. But it is in the concurring and dissenting judgments that the mask slips. The concurring opinion of Judge Pokol adds that "the provisions, and in particular the National Avowal, of the Fundamental Law of Hungary-contrary to many other European constitutionshave raised above all the national independence of Hungary and the safeguarding of State sovereignty in several declarations and value-statements." 83 Quite what this means is not clear, and the dig at "many other European constitutions" is unexplained. What are explained very clearly are the ideological underpinnings of Judge Pokol's opinion, of the Decision itself, and the politics behind the historical claims of the National Avowal: We cannot expect any future decrease of the flow of millions of migrants targeting Europe, and the deep conflicts in the parallel societies of the big Western cities warn us what we have to protect Hungary from. Considering the tensions we see today within the European Union, the need to protect Hungary from the proliferation of tensions from the Western part of the Union can result in the future primary importance of the [HCC]'s procedure of sovereignty and constitutional identity control, the foundations of which have been laid down in this decision. 84 It is in this baleful context, of racist conspiracy theories about "migrants targeting Europe" and the "parallel societies" of "the big Western cities," that the Decision must be read. This is not loyal opposition: it is paranoia. 85 A body such as the HCC will take its opportunities where it finds them, and its abuse of the legitimate jurisprudence of other national courts discredits only itself. It certainly does not provide grounds for the rejection of a legally justifiable and ethically sound concept such as CP.

A mixed bag: The Weiss saga
We have now seen two examples each of loyal opposition and disloyal opposition in the European constitutional context. The final case to be examined, the PSPP judgment of the Bundesverfassungsgericht (German Federal Constitutional Court, GFCC), is somewhat more difficult to classify. In March 2015, the European Central Bank (ECB) launched the Public Sector Asset Purchase Programme (PSPP), 86 whereby it would purchase public bonds of Eurozone member states in an attempt to get inflation rates-then very low-back to the ECB's target of being below, but close to, 2%. In 2017, the program was subject to a constitutional challenge in Germany, the applicants arguing that the PSPP contravenes the EU Treaties' prohibition of lending to or otherwise financing the member states; 87 the principle of conferral; 88 and the principle of democracy under the Grundgesetz (Basic Law, GG), an element of which is that the Bundestag must "take all essential decisions on revenues and expenditure." 89 The GFCC referred the case to the CJEU, which held that the PSPP is within the competences of the Union. This was unsurprising, given the CJEU's generally expansive approach to the delineation of Union competences, and the particularly generous approach it has taken in the context of the Eurozone crisis, as demonstrated by the cases of Pringle 90 and Gauweiler, 91 to which the CJEU in Weiss makes constant reference. As regards the prohibition of monetary financing, the Court followed Gauweiler in declaring that the safeguards built into the PSPP were enough to ensure that the PSPP would not reduce the impetus for Eurozone States to conduct a "sound" budgetary policy. 92 The core of the CJEU's judgment, for present purposes, concerns the principle of conferral. Here, the Treaty distinction between monetary policy (an exclusive EU competence 93 ) and economic policy (the province of the member states) was crucial. For the Court, one must look principally at the objectives of a measure in order to determine whether it falls within the area of monetary policy. 94 The Court acknowledged as undisputed the potential for the PSPP to have an impact on member state finances and   on the real economy that could be brought about by economic policy measures, but noted that this does not necessarily stop the PSPP being a monetary measure, as long as these effects are indirect. 95 Given that the PSPP's objective-to raise interest rates to the ECB's target-can be "attached" 96 to the EU Treaty objective of price stability; 97 and that its means-the buying of marketable instruments-is specifically provided for in the Statute of the European System of Central Banks (ESCB) and the ECB, 98 the Court found the PSPP to fall within the sphere of monetary policy, and thus within the conferred powers of the Union.
The Court had no difficulty in declaring the Union action proportionate. The PSPP was a suitable means of achieving the ECB's goals given the general monetary and financial climate, and the similar activities of other central banks. 99 As regards necessity, the CJEU held that, given the risk of deflation and the virtual impossibility of further cuts to interest rates, the PSPP's stringent eligibility requirements, its limited volume and duration, and other factors, the program did not manifestly go beyond what was necessary to achieve the objective sought. 100 What emerges from the CJEU's reasoning is the formal, procedural nature of the proportionality review it is conducting, its willingness to take the ECB at its word, and the relatively low level of scrutiny that results. While this may theoretically be justifiable in light of the ECB's independence and the broad discretion it is granted by the Treaties, it was not to be enough for the GFCC.
Upon the case's return to Karlsruhe, the GFCC agreed that the proportionality of the ECB's program was of central importance, but was distinctly unimpressed with the CJEU's analysis, describing it as being "not comprehensible and thus objectively arbitrary." 101 Crucial was the distinction between a procedural and a substantive approach to proportionality review of the acts of the Union's institutions. For the GFCC, what is important is not the ECB's objective in coming up with the PSPP, nor the instruments that were employed, but rather the PSPP's actual effects, monetary and economic. Such effects include, for the GFCC, the risks that "necessary consolidation and reform measures" in certain member states will not be implemented; 102 that member states will be dissuaded from pursuing "sound budgetary policy"; 103 and that some unspecified "impact" will fall on "shareholders, tenants, real estate owners, savers, or insurance policy holders." 104 The CJEU's alleged "complete disregard" 105 for such risks results in a "structurally significant shift in the order of competences to the 95 Id. ¶ ¶ 59-70. 96 Id. ¶ 57. 97 TFEU, supra note 24, arts. 127 Under the first leg of the legitimacy test I have sketched in this article, the GFCC passes institutional muster. The German Rechtsstaat has not been subject to attack in the manner of Poland or Hungary, and whatever disagreements one may have with its decisions, the GFCC is an independent body worthy of respect. However, it is in the second element of the test that PSPP fails to convince. Though based on sound theoretical fundamentals regarding the logical impossibility of a Union of limited competences being the sole determinator of the limits of those competences, certain aspects of the GFCC's approach are regrettable, and serve as a stark illustration of the importance of communication, self-restraint, and self-awareness in the context of European constitutional conflict. It beggars belief that the GFCC was surprised by the CJEU's reasoning, which hewed so closely to that of Pringle and of Gauweiler. If the German Court was so intent on the CJEU conducting a probing, substantive proportionality assessment of the ECB's actions (as is emphatically not its wont), it should have made such a desire clear in its Order for reference, which it did not. Moreover, if unsatisfied with the CJEU's response to its questions, it could have acted as the ICC did in the Taricco saga, and referred further questions in light of the CJEU's findings. Though it was noted above that the ICC did not relent even after the return of its second set of questions, its conduct throughout the Taricco litigation was beyond reproach (at least up until its decision that the Taricco rule could never be applied, a finding on which views will differ). The GFCC, by contrast, comes out of the PSPP dialogue looking like a court spoiling for a fight, rather than one seeking to engage in good-faith efforts at the coordination of overlapping national and supranational constitutional systems.
But the problems with PSPP go deeper: the judgment also suffers from a lack of vertical and horizontal coherence. Vertically, the GFCC's approach to proportionality review-specifically the third step, proportionality stricto sensu-differs from the way in which the principle is applied and assessed by the CJEU, at least in the context of Article 5(4) TEU. 107 The GFCC's failure to take into account that proportionality is a principle with different concrete expressions across the Union shows a regrettable lack of self-awareness and indeed of humility, and is arguably contrary to the GFCC's own stated practice, invoked in PSPP itself, whereby ultra vires review of EU norms is to be coordinated with the CJEU and conducted "cooperatively in accordance with the European integration idea and relaxed through mutual consideration," 108 and where the CJEU "has a right to tolerance of error." 109 Horizontally, the GFCC in PSPP seems to be operating in a curiously bipolar constitutional universe, consisting of the German constitutional order, the EU constitutional order, and no others. PSPP is not a decision of a court that has taken into account its place in a polyarchic constitutional system. A crucial part of the analysis here is attention not merely to legal text but to political subtext. It does not breach the principle of the equality of the member states to recognize the political and economic reality that a dispute about the statute of limitations in Italy, severance payments in Denmark, or pensions in Czechia is of a rather different order of magnitude than the question of continued German participation in quantitative easing programs on which the economic fortunes of a continent depend. This being the case, citizens right across the European Union are entitled to feel a certain indignation at the cavalier behavior of the most powerful court in the most powerful member state. Relatedly, it is a mistake to read the GFCC's decision in Weiss in a vacuum, incognizant of the political orientation of the applicants in the case. 110 As in Gauweiler, the applicants in PSPP were a range of right-wing figures seeking, by means of judicial review, to place legal limits on the independence and autonomy of the ECB because they simply happened to disagree with the manner in which the ECB has exercised this independence in its "whatever it takes" approach to quantitative easing. Underlying all this is the deeply tiresome conception, unmoored from reality, of the euro as some kind of German charity project. The verdict in PSPP is replete with references to the risk of the PSPP reducing the incentive for "certain" member states 111 to pursue "sound" budgetary policies, 112 as if budgetary "soundness" were an objective legal standard, and not an entirely contingent notion that shifts according to the given economic landscape and the particular conception of political economy one employs in the analysis.
But we cannot necessarily blame the GFCC for its repetition of the concept of budgetary "soundness," given that such language is to be found in the EU Treaties themselves, 113 and it is here that we come to the heart of the issue: PSPP is simply the necessary result of the unsustainable and unworkable distinction at the core of the political and legal constitution of the Economic and Monetary Union (EMU). The EU Treaties' allocation of monetary policy to the EU and economic policy to the member states has had disastrous consequences, including leaving the European Union unable to respond to the Eurozone crisis without taking actions, such as the PSPP, which are at the very least of questionable legality under the scheme of the EU Treaties as they currently stand. Moreover, this division has no basis in economic, monetary, or constitutional theory, but is rather the result of the compromise whereby the European Union was to have a shared currency but not a shared budget and shared liabilities.
In the aftermath of PSPP, there have been calls for national ultra vires review to be abandoned, 114 or for the EU Treaties to be reformed with a view to establishing an institutional mechanism for the resolution of Kompetenz-Kompetenz questions by a special formation of the CJEU, to include national court judges. 115 For reasons which will become clear in the next section, such proposals, though well-intentioned, are misguided. This does not mean that the Treaties are not in need of reform: only that the reform must focus on the root cause of the problem (the unworkable nature of the economic/monetary distinction) and not its mere symptom (disagreement between EU and national courts on where to draw the line). There is much to criticize in PSPP, but it is ultimately an instance of constitutional conflict caused by the structural tensions at the very heart of the EMU, and not a threat to the claimed values of the Union itself or to the rule of law. It therefore serves as a useful interstitial marker, a midway point on the second leg of the legitimacy test, that illustrates both the dangers and the advantages of the incorporation of loyal opposition into EU legal practice, and-in its political effects-the importance of constitutional conflict as an animating tension.

Constitutional pluralism, constitutional identity, and loyal opposition in EU legal scholarship
We have now seen examples of legitimate and illegitimate opposition to the CJEU's conception of the primacy of EU law, and one case that lies somewhere in-between. Some scholars have concluded that, however well-intentioned the scholarship on CP may have been, the development of the new European post-fascism, and its adoption of constitutional pluralist language in order to justify deviations from European norms, means the theory is now so dangerous that it must be abandoned. As Daniel Kelemen and Laurent Pech put it, "[c]onstitutional pluralism is a theory designed for polite society, but we live in brutal times." 116 On this analysis, constitutional pluralists must recognize that "their creation is having destructive consequences that threaten the entire EU legal order." 117 There is little room for disagreement here: you are either with the EU or you are against it. However, this new critique is by turns misbegotten, inattentive, and unwise: first, in its grounding on a fundamentally mistaken reading of the nature of the primacy of EU law; second, in its failure to distinguish between legitimate and abusive constitutional identity claims; and, third, in the underlying political, legal, and constitutional assumptions that lead to its willingness to place so many eggs in the single basket of a hierarchical conception of EU law.

The primacy of EU law
Kelemen is explicit that his critique of CP is based on a worldview where Union law has "supremacy" over national law. 118 However, the concepts of "supremacy" and "primacy" are conceptually distinct, and it is to "primacy" that the CJEU refers in its own jurisprudence. "Supremacy" implies and requires an erga omnes hierarchy between norms or institutions within a single, integrated system or order. "Primacy" is concerned with the in casu preference given to one norm or institution over another in the context of interacting but distinct systems or orders. 119 In seeking to bolster this monist view of the relationship between EU law and national law, Kelemen refers to the member states' affirmation of the CJEU's primacy jurisprudence in Declaration 17 annexed to the TFEU, whereby they: "recall[] that, in accordance with well settled case law of the [CJEU], the Treaties and the law adopted by the Union . . . have primacy over the law of Member States, under the conditions laid down by the said case law." But Kelemen makes no mention of the normative status of this Declaration, nor of its legislative history. As a Declaration annexed to the Treaties, and not as part of the Treaties themselves, it is not legally binding. Though it expresses the agreed political position of the member state governments at the time of its declaration (and is thus worthy of respect), it cannot override the specific means by which EU law is constitutionally incorporated in each member state. Moreover, this non-binding status arises precisely because the explicit and legally binding "primacy clause" of the failed Constitutional Treaty (CT) 120 was removed in the process of turning that Treaty into the Lisbon Treaty. Article I-6 of the CT would have stated that "[t]he Constitution and law adopted by the institutions of the Union in exercising competences conferred on it shall have primacy over the law of the Member States." We cannot simply pretend that this statement entered into law. It did not. 121 In cases of conflict, Kelemen writes that national apex courts can "remedy the situation by compelling their government either to amend their constitution, to seek to change the EU legal norm involved by working through the EU political process, or, if necessary, to withdraw from the Union altogether." 122 However, these three options do not describe practical or theoretical reality, and though Kelemen does acknowledge that withdrawal would be a "drastic" step, he does not balance the (always) extreme difficulties and heavy consequences of such a move against the (variable) weight of the particular legal issues that may be at stake in a case of constitutional conflict between a member state and the EU. It is correct that withdrawal is the ultimate safety valve when it comes to constitutional conflict in Europe, and marks the EU as being significantly different from other (quasi-)federal systems. However, exit from the European Union is a momentous step, with enormous consequences for the withdrawing state (and for the continuing European Union), and, more importantly, with potentially lifechanging implications for the citizens of the withdrawing state and the remaining states, particularly those who have exercised their freedom of movement under the 119  EU Treaties. The "speedy-exit" approach of the critics is far too quick to show member states the door in the event of disagreement. Is it not rather glib to answer the alwayspresent possibility of constitutional conflict between the member states and the EU by saying "well, you can just leave"? 123 Consider the Taricco context: should Italy now withdraw from the European Union because of the (still unresolved) incompatibility with EU law? After all, neither Italian law nor the Italian constitution have been amended. The CJEU did not reverse its jurisprudence. This then leaves us with exit. But though the principle of legality in criminal matters is important, the precise length of the statute of limitations in cases of VAT fraud would be a rather recherché ground on which to base the undoing of sixty years of European integration and the legal unmooring of the lives of millions of people. On the other hand, the (vital) importance of Italy's European obligations, and the (unquestioned, but not unquestioning) loyalty it owes to the CJEU, are not sufficient to prejudice the rights of criminal defendants. Put simply, in the Taricco saga, the ICC was right and the CJEU was wrong-a state of affairs demonstrated by the cogency of the ICC's reasoning and the poverty of the CJEU's reasoning; and one that is not captured by or reflected in Kelemen's three possibilities, or in the general "supremacist" view of the relationship between EU law and national law.
Similarly, there were at least four ways in which the Ajos conflict could have been resolved. First, the legislature could have amended the law in order to bring Denmark into line with its Union obligations. This is in fact what has since happened: the Folketing has removed the discriminatory exceptions, and all is now well. 124 Second, the Commission could have brought an enforcement action against Denmark. Thirdand in the context of the Mangold case law this is rather more unlikely than it perhaps should be-the CJEU might have reconsidered its jurisprudence, as it did (or tried to) in Taricco II. Finally, matters may simply have rested as they were, as they did after the CCC's judgment in Holubec. This may be unsatisfactory, but such a modus vivendi is in fact entirely in keeping with the frequently messy-but nevertheless basically workable-way in which federal systems operate: changes in the composition of the CCC led to improved relations with the CSAC, and a quiet narrowing of the application of Holubec to bring the law into line with Landtová. 125 We must not let the legalistic fetish for perfection and smooth institutional interoperation cloud the observable fact of the difficulties of reconciling the different layers in a composite legal order. Any account of the relationship between EU law and national constitutional law that insists on the total subordination of the latter to the former in the case of conflict, and does so by making dire predictions of chaos if the CJEU's conception of the principle of primacy is breached, must also account for the fact that constitutional law actually exists and operates as a sociological system under non-ideal circumstances of political churn and institutional fallibility, and not as a pure theory. For Kelemen courts to rule EU acts ultra vires "would destroy the legal order." 126 But this is demonstrably untrue: though the CCC, the DSC, the ICC, and the GFCC really did refuse to follow judgments of the CJEU-in 2012, 2017, 2018, and 2020 respectively-eppur si muove. It is certainly the case that such disagreements between legal orders are unfortunate, but the only way to prevent them altogether would be to turn the Union into exactly the kind of vertically integrated legal order-with a specific and explicit EU Treaty clause on the supremacy of EU law-that it is not.
Primacy is an important principle of EU law, to which all member states have assented, subject in each member state to the specific means of national legitimation and incorporation of EU law. It is grounded in a coherent reading of the EU Treaties, and is justified, inter alia, by the duty of sincere co-operation, the requirements of legal certainty, and the equality of the Member States. It is a principle of long-standing and great importance, and is deserving of respect. However, from the national perspective, the principle is neither absolute nor unlimited, and it must be reconciled in hard cases with the continuing-though modified-sovereignty of the Member States; the legally binding nature of their constitutions; their generally greater democratic legitimation than the Union; and much else besides. Each of Kelemen's resolutions (domestic amendment, Union amendment, exit) has its place, but as we have seen, they do not exhaust the range of possibilities in seeking to reconcile EU and national law in cases of conflict.

Constitutional identity claims
A particular focus of the critics' ire is the use of the notion of "constitutional identity" by national courts to justify departure from EU norms. For Federico Fabbrini and András Sajó, "constitutional identity" is indeterminate and arbitrary. 127 It is "drenched with neo-sovereigntist features and is contrary to the rule of law," 128 and "[a]nyone who relies on the language of identity runs the risk of making constitutional adjudication open to nativist populist considerations." 129 A full-blown defense of constitutional identity is beyond the scope of this article, and, in any event, it is not clear that such a defense is desirable or even possible. The concept is indeterminate, and we have already seen its abuse by the captured HCC. If Article 4(2) TEU has a role to play in the policing of boundary disputes by the Union and Member State judiciaries, a rigorous conceptual and analytical framework must be constructed around it. 130 However, this dismissal of all constitutional identity claims as being somehow destructive of the fabric of the Union causes scholars to conflate very different cases. national courts did not formally utter the word[s] constitutional identity, they followed a similar ratio, by refusing to comply with [CJEU] rulings . . . in the name of protecting specific features of the national legal order." But as I have demonstrated, to group the Czech and Danish cases together merely because they both exemplify a heterodox approach to the primacy of EU law is untenable, and shows a remarkable lack of attention to the specifics of the cases. Holubec and Ajos are radically different in terms of subject matter, constitutional and political context, and most importantly as regards the quality and justifiability of the reasoning employed.
The rejection of constitutional identity claims then feeds into the rejection of the entire notion of CP. For Kelemen and Pech, "some ideas are inherently dangerous, and constitutional pluralism is certainly one of them. . . . [C]onstitutional pluralism is an abnormally dangerous product and its manufacturers should be held to a standard of strict liability for the damage it has caused." 131 Fabbrini and Sajó quote this approvingly, but in doing so they simply replace "constitutional pluralism" with "constitutional [identity]," as if the two ideas were identical. 132 Elsewhere, Kelemen and Pech speak of constitutional identity and constitutional pluralism as being "closely related" 133 and "twin concepts." 134 However, the two ideas are analytically separate, and must be disaggregated. Constitutional pluralism is the general descriptor for the whole stable of theories that seek to privilege neither the national over the European nor the European over the national. Constitutional identity is only one of the specific means by which the contingent and non-universalizable question of the superiority of one over the other in a given case might be determined. We could banish constitutional identity from our jurisprudence altogether, as the critics suggest, and it would not make a dent in the lived reality and normative desirability of a pluralist conception of the European legal space.
The approach of the critics is therefore one that seems to imagine that the EU Treaties contain something they do not (an explicit, legally binding statement of the "supremacy" of Union law), and do not contain something they do (an explicit, legally binding guarantee of respect for national constitutional identity). As with so much else in law and in life, identity and pluralism are susceptible to misuse and abuse-as Kelemen and Pech themselves admit before proceeding with their criticisms. 135 The trick is to learn to distinguish between legitimate and abusive invocations of constitutional identity and CP, and not to regard every instance of opposition as encouraging disloyalty to the claimed values of the European project. 131 132 Fabbrini & Sajó, supra note 6, at 472. 133 Kelemen & Pech, supra note 6, at 61. 134 Id. 135 Id.

Background assumptions of the critique
At the root of all these criticisms is a belief in the Union as a bulwark against authoritarianism. This claim may not be explicit, but it is always a background feature. This idea is not unfounded: it is no accident that the collapse or overthrow of dictatorships in Portugal, Spain, Greece, and the former Eastern Bloc was swiftly followed by EU membership, in part as a way of trying to prevent antidemocratic backsliding.
I share with all the critics their rejection of the new authoritarianism in Europe, but this belief in the European Union as a defense against the worst instincts of Europeans is beginning to look rather forlorn. The European Union's difficulties in ensuring the democratic credentials of the member states are well known, not least to Kelemen, whose excellent work on the issue is required reading. 136 Though the EU may be a Union of ostensibly democratic states, its leadership has shown very little appetite for confrontating national backsliding. Is it really the case that the Union can restrain our worst impulses at national level? If it can, will it? And if it will, is this inherently the case, and guaranteed always to be so? And what happens if and when the boot is on the other foot? Kelemen has noted, as Robert Schütze did before him, 137 the parallels between modern European CP and the dispiriting history of "states' rights" in the United States. There, the tools of nullification and interposition were used by racists in attempts to prevent the Federal imposition of equal rights for African Americans. Just as federal legal hierarchical superiority was the only corrective to white viciousness in the United States, the argument goes, European Union hierarchical superiority is the only corrective to the growth of post-fascism in Europe.
But the recent US controversy regarding the Immigration and Customs Enforcement (ICE) agency of the Department Homeland Security is also instructive: here we see an authoritarian federal power using an oppressive agency to enforce unjust policy, against the wishes of certain states. Suddenly, US liberals generally, and liberal legal scholars in particular, have discovered a new appreciation for states' rights. 138 Certainly the cases are not identical: for one, states of the United States have no right of unilateral withdrawal from the Union, as exists in the European Union. But whether in the United States or in the European Union, arguments for or against absolute federal supremacy are frequently contingent: we simply side with the constitutional order or layer with which we happen to agree, for ethical or political reasons. But with authoritarians sitting on the Council of the European Union, nominating Commissioners and judges of the Court of Justice, and ultimately controlling the votes of large numbers of Members of the European Parliament (MEPs), are we wise to place such faith in EU law as a benevolent influence? It is not beyond the bounds of possibility that, at some point soon, 136  authoritarians will achieve electoral success in more and more member states, and "good" member states may seek to resist "bad" Union law. Joseph Weiler once observed that "a democracy of vile people will be vile." 139 It follows that a Union of vile member states will also be vile. One hopes that if it were the Union delegitimizing opposition and fomenting bigotry in, say, Poland or Hungary, the critics would be turning to their MacCormick in order to see how the member states might legitimately oppose EU law. The rise of post-fascism in Europe is a political problem. It is entirely possible, but only possible, that law, including Union law, and institutions, including EU institutions, may help in combatting this disease. But to place too much faith in these is to let the liberal love of legalism and faith in institutions cloud hard-headed analysis of the threats we face. The CJEU certainly has a role to play: there has already been at least some movement on this front, with the Court-as we have seen-authorizing national judiciaries to conduct their own analysis of the independence of the Polish judiciary 140 and finding the Orbán government's treatment of the Central European University contrary to the EU Treaties and the Charter. 141 However, the CJEU itself is not necessarily immune from capture, or-at least-infiltration. Certainly, it is the EU institution most shielded from partisan interference, and the system of appointments to the CJEU has become more robust since the entry into force of the Lisbon Treaty. Any attempt by an Orbán or a Kaczyński to nominate the kind of opportunist hireling with which they have stuffed their national courts would be resisted. But the CJEU cannot be our only-or even main-hope for navigating the new dispensation in Europe.
It happens to be the case for now that we look to the Union to keep democratic backsliding in the member states in check. But we must remember that in a different context, absolute Union supremacy may look less attractive. The fact that, as things currently stand, this is unlikely is immaterial: it is possible, and that is enough. Just as not all assertions of constitutional identity are wicked, not all assertions of federal supremacy are good. We must become more sophisticated in our analysis, and learn to distinguish between the legitimate and the abusive, the loyal and the disloyal.

Conclusion
There is a certain professional deformation that occurs to scholars of EU law. We feel at home with the subtleties and complexities of a notoriously opaque legal order. We tend to agree with the basic principles of European integration, though we may disagree strongly about its specifics. We are socialized into a cosmopolitan habitus of shared ideas, shared presumptions, and very frequently shared phenotype and social class. It is therefore no wonder that anything that smacks of Euroscepticism, in the overbroad and ideologically incoherent sense we have come to understand it, raises the hackles. But in identifying too strongly with our subject, we impoverish our own perspective. Thus blinkered, it seems natural that, of course, the CJEU is correct about primacy, and has been correct since 1964. Of course EU law is good, cosmopolitan, socially and politically pluralist, just, protective of rights, progressive, democratic etc. Of course any legal approach that seeks to defend or entrench the legitimate goods of national constitutionalism is dismissively and fearfully rejected as "nationalism" or "sovereigntism," by which we mean bad, narrow-minded, parochial, unjust, exclusionary, racist, oppressive, undemocratic, etc. We are therefore inclined to agree with Kelemen that "EU law, and the [CJEU] as the ultimate guardian of that law, must enjoy unconditional supremacy." 142 But aren't we also supposed to be constitutionalists? Constitutionalism teaches us that the unconditional supremacy of anything, emphatically including law, is inherently dangerous.
I argued in the introduction that there is no loyal opposition in Union politics, though there should be. In section two, I asked whether it might be possible for there to be loyal opposition in EU law, whereby loyal state-level deviations on complex questions of the application of EU law within the member states might be domesticated and integrated into the system itself, rather than catastrophized. In section three, I then extended the proposal to EU legal scholarship. The point is not just that constitutional pluralism is both descriptively accurate (which is relatively uncontroversial) and normatively desirable (which is most definitely not): it is that daring to make these arguments should not lead to excommunication. It may well be that I am entirely mistaken. It may be that total acquiescence to the CJEU's view of primacy is what will prevent further decay in European constitutional democracy. The orthodoxy may be correct. But whether or not this proves true, heterodoxy must not only be tolerated in EU constitutionalism: it must be actively engaged with and incorporated into the system, as a form of loyal opposition. 142 Kelemen, supra note 6, at 403.