The making of landmark rulings in the European Union: the case of national judicial independence

ABSTRACT We study the conditions under which the European Court of Justice (ECJ) expanded the European Union's reach into issues of national judicial independence. The ECJ's 2018 ruling in a case known as Portuguese Judges contained a far-reaching constitutional interpretation that had a transformative impact on the European Union's rule of law crisis. We conduct a theoretically guided investigation into both the domestic origins and the judicial outcome in the Portuguese Judges case. We show how the ECJ, with implicit support from the majority of the Member States, strategically exploited suitable characteristics of an inconspicuous case to produce a landmark ruling that enabled unprecedented enforcement action against democratic backsliding in Poland and Hungary.


Introduction
On 27 February 2018, the European Court of Justice (ECJ) quietly handed down what would turn out to be one of its most important rulings in recent decades. In a case that came to be known as Portuguese Judges or ASJP, the Court radically expanded the scope of its own jurisdiction to comprise the protection of national judicial independence. According to Pech and Kochenov (2021), 'this judgment can be viewed as belonging to the Pantheon of ECJ rulings, on a par with Van Gend en Loos and Costa'.
In essence, the Court interpreted Article 19(1) of the Treaty on European Union (TEU) as requiring Member States to ensure that national courts comply with the principle of judicial independence, regardless of whether they happen to be applying EU law in a specific case. This obligation may appear trivial to the casual observer, but it represents one of the most farreaching judge-led developments in EU law in the recent past, especially as the Court needed to contend with limitations on the scope of the Charter of Fundamental Rights of the EU (CFR). By creating a new EU-wide obligation, the ECJ also empowered the European Commission to bring infringement procedures for its violations. Combined with the preliminary reference system under which national judges submit cases to the Luxembourg Court, the ECJ gave national judges a new tool to protest against government interference with the judicial branch. The expansion of the ECJ's competences into the area of national judicial independence produced a backlash in democratically backsliding EU countries, including a direct challenge to the primacy of EU law in Poland.
While the constitutional and normative implications of the judgment have not escaped the attention of lawyers (Bonelli & Claes, 2018;Elsuwege & Gremmelprez, 2020;Pech & Platon, 2018), this paper situates the Portuguese Judges case in literature on domestic legal mobilization and the ECJ's constrained ability to pursue preferred outcomes. We draw on documentary evidence, interviews and quantitative data to study both the domestic conditions affecting the referral of the case to the ECJ in the first place, as well as the EU-level factors shaping the eventual ruling. This integrated, mixed-method approach to studying a key event in European integration enables us to understand the domestic origins behind the preliminary reference which the Court uses to advance European integration in a decisive manner.
In the next section, we discuss the theoretical conditions under which we expect EU law to be mobilized at the national level and cases referred to the ECJ (supply side), and the conditions under which the ECJ is likely to expand the reach of EU law vis-à-vis the Member States (demand side). The following two sections document, respectively, our case study into the domestic mobilization effort in Portugal and the ECJ's ruling in the litigants' case. The final section discusses the findings in the context of our theoretical expectations, highlighting the broader significance of our case study.

Theoretical conditions
We begin by surveying extant theories of not only the ECJ's judicial behavior but also of legal mobilization in the context of the preliminary ruling system. Although the ECJ has rendered important decisions through other procedures as well, the preliminary ruling procedure is most commonly associated with ground-breaking constitutional moments, including judgments in the foundational Costa/ENEL and Van Gend en Loos. In contrast, the fact that we associate the infringement procedure in particular with few landmark rulings 1 is likely endogenous to its functioning. The infringement procedure is about compliance with existing law and the Commission prefers bringing cases where it has a high chance to winboth these aspects are antithetical to making up novel legal claims that could foster a breakthrough constitutional ruling by the ECJ. Moreover, the bottom-up dynamic of preliminary references provides them with a layer of legitimacy that is unavailable for purely 'Brussels-driven' cases.
We divide both the theoretical and empirical part of this paper along two strands: the supply side consists of the factors that create the case and lead to its referral to the ECJ; the demand side of constitutional change comprises the ECJ's preferences and its constrained ability to pursue them in the increasingly politicized EU system.

Supply side: legal mobilization
There are no landmark preliminary rulings without successful mobilization of EU law at the national level. 2 Although preliminary questions are referred by national judges and litigants do not have a right to have a question referred to the ECJ, preliminary references are frequently driven by litigants. It is not uncommon for preliminary questions to be drafted by litigants rather than judges (Pavone, 2019, p. 867). Conant et al. (2018) distinguish three levels of factors influencing legal mobilization of European law. At the macro level, there are the substantive EU rules that can be mobilized by national actors, as well as procedural limitations attached to the EU legal system. The meso-level comprises the national opportunity structure for legal mobilization. National legal systems tend to vary when it comes to procedural rules governing litigation before national courts, including rules on who can bring cases, before which courts, within what time limit and for how much money. Finally, the micro-level factors cover litigant characteristics such as their identity, resources and influence.
While helpful as a systematic overview of variables potentially impacting mobilization, the various levels might be interlinked in practice. Many successful examples of mobilizing EU law are tied to the activity of so-called 'Euro-lawyers' (Pavone, 2019). The variable ingenuity of 'Euro-lawyers' to find an EU law angle to the case at hand is a micro-level factor that actually shapes the seemingly fixed macro-level availability of EU rules. A good 'Eurolawyer'the hiring of whom may, in turn, depend on litigant networks and resourcesmight be able to craft a convincing EU law argument where others do not see one. The importance of the lawyers' capacity to 'reimagine' the macro-level opportunity structure is more important when constitutional issues are at stake where sector-specific legislation does not act as an anchor around which mobilization takes place.
Extant literature gives ample reason to consider also the possibility of references being driven by entrepreneurial national judges, possibly seeking empowerment vis-à-vis other courts or branches of power in the process (Alter, 2001) in which instance the onus of formulating the EU law dimension of a case might not fall on the litigants. However, the increasingly more important role of higher courts in the preliminary ruling system (Dyevre et al., 2020), as well as knowledge and resource constraints faced by lower courts make it overall less likely that constitutional breakthroughs would nowadays originate in the activism of lower courts.

Demand side: ECJ as constitutional entrepreneur
The ECJ is seen as an actor that tries to push European integration forward when possible, be it through expanding the reach of EU law or strengthening the EU's supranational dimension (including its own role) (Burley & Mattli, 1993). These pro-federal inclinations have been shown to be tempered by Member States' preferences (Carrubba et al., 2012;Larsson & Naurin, 2016;Ovádek, 2021). While it would be wrong to conclude that Member States have a 'final say' on judicial outcomes (Larsson, 2021), the ECJ appears consistently more likely to nudge European integration along when its preferred policy choice is supported or at least not opposed by the Member States. Compared to the early period of European integration characterized by landmark rulings such as Costa/ENEL and van Gend en Loos, the increased politicization of the EU could force the ECJ to be even more mindful of Member States' positions (Alter, 2000).
Historically, the propensity of the ECJ to hand down landmark rulings advancing European integration was linked to political gridlock (Lenaerts, 1992), partly thanks to law's ability to function as mask and shield (Burley & Mattli, 1993). Blauberger and Martinsen (2020) nuance the 'mask and shield' proposition by pointing to the constraining role of public discourse in addition to Member State preferences. It has been shown that public discourse around politically charged subjectssuch as migration and social benefitsis liable to affect the Court (Blauberger et al., 2018). But public debates only exist around issues that are salient and many ECJ judgments garner little public attention. Nonetheless, even in the absence of public salience, an issue can be salient for the Member State government, which is particularly likely in the case of preliminary references challenging domestic laws. In this way, supply-side factors influencing the salience of the domestic case feed into the ECJ's decision-making calculus, as governments are more likely to hold stronger preferences over more salient issues. For example, although the public salience of the Costa/ENEL dispute was relatively low, the case was important to the Italian government (Arena, 2019). Similarly, the constitutional issues raised in the landmark van Gend en Loos case attracted the attention of the Dutch, Belgian and German governments.
In advancing integration or achieving some other policy goal through the preliminary ruling procedure, the ECJ is likely to be opposed by the Member States whose laws are most immediately affected. As a result, the ECJ faces incentives to make side payments to governments to dampen their opposition. The canonical side payment in the preliminary ruling context is the ECJ limiting the impact of a new doctrine in the case at hand. For example, while establishing the principle of supremacy of EU law in Costa/ ENEL, the ECJ largely upheld the contested statute of the Italian electricity monopoly (Arena, 2019(Arena, , p. 1030. A notable feature of the preliminary reference system is the ECJ's ability to reformulate the question posed by the national court. Such reformulations enable the Court to put a preferred twist on the question or highlight a specific aspect of it (Šadl and Wallerman, 2019). This type of discretion sets the preliminary ruling procedure apart from other procedures and harbors the potential to facilitate jurisprudential breakthroughs that might be otherwise hampered by the original wording of the question. The power to reformulate also affects the supply-side perspective on court outcomes: the margin to amend a preliminary question means the ECJ is not strictly dependent on what cases arrive from the national level (Šadl and Wallerman, 2019). At the same time, the reformulated question must bear resemblance to the original question lest the rewording be considered illegitimate by other stakeholders.

Supply side: judicial mobilization in Portugal
This section traces the roots of the ASJP case in Portugal. We draw on documentary evidence, including court submissions, and semi-structured interviews with four persons closely involved in the mobilization effort to understand the domestic origin of the case, as well as the motivations and strategy of the litigants. All interviews took place in June 2021 via individual video calls.

Background to the dispute
Portugal was among the countries most severely impacted by the Eurozone sovereign debt crisis. At the brink of bankruptcy, the Portuguese government negotiated in May 2011 a 78 billion EUR bailout package with the European Commission, the European Central Bank and the International Monetary Fund. In exchange for the emergency funding, Portugal had to adopt austerity measures in line with conditions detailed in three memoranda of understanding with a view to reducing the deficit in public finances. On this basis, the government expanded public sector pay cuts each year until 2014.
The salary cuts impacted virtually all state employees, including judges. According to data reported to the European Commission for the Efficiency of Justice (CEPEJ), the average monthly gross salaries of Portuguese judges in 2010 ranged from approximately 2891 EUR at the entry-level to 6950 EUR at the peak courts. 3 In 2014, these figures dropped to 2391 EUR and 5517 EUR respectively. In addition to this base salary, judges received a special allowance of approximately 700 EUR which the government intended to permanently reduce as of 2014, whereas the pay cuts were temporary for the duration of the economic crisis.

Court action
When asked about the motives of the ASJP action, all interviewed actors were quick to point to the defense of judicial independence as the overarching objective. However, there was also admission of the salary cuts being economically significant for many judges (Interviews 2 and 3). The membership expected the association to take some action against the wage reductions and the idea of a general strike was mooted at one point as well. At the same time, the judges were aware that the whole country was suffering economically and that with the exception of the special allowance, the pay cuts applied to all state employees. Still, the feeling of being 'disrespected' by the government came up in the interviews, largely because the government did not consult judges prior to the pay cuts (Interviews 1 and 2).
While rejecting the idea of a strike, the ASJP leadership actively contested the austerity measures as from early 2013 (Interview 1). In both 2013 and 2014, ASJP asked the Prosecutor General and the Ombudsmanboth of whom have privileged access to the Constitutional Courtto challenge the constitutionality of the measures (Interviews 1 and 2). This approach failed, not least because both offices were aligned with the government (Interview 1). In addition, in 2013 ASJP asked the Portuguese government and parliament to revise the salary cuts in a way that would give weight to the principle of financial independence of courts and judges (Interview 1). It was only following this failure to persuade other actors to change the course that the ASJP initiated disputes before ordinary courts that eventually led to a referral to the ECJ.
The failed attempt to reach the Constitutional Court is particularly significant. Although access rules ultimately foreclosed this possibility, the litigants' initial attempt at overturning the salary cuts and bolstering judicial independence were not directed to Luxembourg. Had ASJP managed to get its case before the Constitutional Court, it is likely that, building on its increasingly anti-austerity rulings, the Constitutional Court would have granted ASJP's application at least insofar as the special allowance was concerned. At the same time, a preliminary reference to the ECJ would be highly unlikelythe Portuguese Constitutional Court referred its first question only in 2021. In the end, it would be the Constitutional Court developing the principle of judicial independence, enshrined in Article 203 of the Portuguese Constitution, instead of the ECJ. This counterfactual appears all the more probable given that in Italy, the Constitutional Court ruled in such a manner in a similar dispute in 2012, without any involvement of the ECJ.
ASJP's attention thus turned to other courts. Three actions targeting the lawfulness of the pay cuts were initiated on the grounds of violating the principle of judicial independence. The three actions targeted different jurisdictions but the first and only to gain traction within a reasonable time was a lawsuit before the Supreme Administrative Court (SAC), formally brought against the Court of Auditors but taking aim at the government's austerity policy.

Judicial independence and EU law
Although the court submission was prepared by a law firm representing ASJP since 2012 (Interview 4), interviewees were unanimous in identifying the then ASJP president, judge José António Mouraz Lopes, as the driving force behind the action and its principal claim that salary cuts for judges are problematic from the perspective of judicial independence. In an interview, he emphasized the role of Magistrats européens pour la démocratie et les libertés (MEDEL), a network of European judges where he was the Portuguese representative for over 10 years (Interview 1). Membership of MEDEL gave the ASJP president an opportunity to not only maintain international contacts but also keep abreast of developments in other countries to draw inspiration from, such as the Constitutional Court challenge of salary cuts in Italy. Other persons close to the ASJP case were similarly embedded in European networks (Interviews 2 and 3).
Yet it would be a mistake to think the ASJP case was brought with Hungary and Poland in mind (Interviews 1 and 2). In late 2014 when the court submissions began to be prepared, the most notorious attack on judicial independence in the EUin Polandwas still yet to come. The European-wide implications emerged more clearly only later, as the rule of law situation in Poland gradually deteriorated from 2015 onward, and especially when the ECJ handed down its judgment.
Nor was the ASJP case just an artificial vehicle for the ECJ to develop its doctrine. Interviews revealed no contact between case initiators and Luxembourg insiders. Moreover, the case was not built exclusively or even primarily around EU law. Legal documents show that the initial submission by ASJP from 20 January 2015 relied on four legal claims, only one of which was supported by EU law arguments. In paragraph 58 of the submission, the plaintiff makes the claim that ends up revolutionizing EU law: In addition, the principle of judicial independence also derives from Articles 19 (1) [TEU] and 47 of the Charter of Fundamental Rights, and therefore, the salary reduction in question, effected by Portuguese executive and legislative powers, the European Union law will also be put at risk. The submission continues by arguing that the origin of the litigated salary cuts rests in European Union law by virtue of the finding of an excessive deficit in Portugal and the conditions attached to the emergency funding in 2011. The purpose of this part of the argument is to establish that Portugal was 'implementing Union law' within the meaning of Article 51(1) CFRa condition for the application of the Charter to a Member State's actionswhen it imposed the pay cuts on judges.
Crucially, despite the tenuous connection between pay cuts and EU measures, the plaintiff did not propose a radical reinterpretation of the scope of EU principles and jurisdiction. In contrast to the ECJ's ruling, ASJP did not claim that the scope of EU law provided for by the term 'fields covered by Union law' in Article 19 (1) TEU is substantively different from the scope of the Charter limited by the term 'implementing Union law' in Article 51 (1) CFR. Unlike the ECJ, ASJP tried to fit the case of Portuguese judges' salaries within the existing confines of the case law on the applicability of the Charter, which required establishing a link between EU law and the disputed national measures.
Nonetheless, invoking Article 19 (1) TEU and the principle of judicial independence explicitly was innovative. This and other EU law arguments in the submission originate in an amicus curiae brief drafted by professors Alessandra Silveira and Pedro Madeira Froufe of the University of Minho in Braga. Facilitated through personal contacts (Interview 2), ASJP asked the professors to elaborate an opinion supporting the EU law dimension of their claims. Interviewees identified the opinion as key to convincing the SAC to refer a preliminary question to the ECJ (Interviews 2 and 3). Comparing the text of the opinion with arguments raised in the ASJP submission shows that the professors were de facto in charge of the EU law dimension of the plaintiff's claims.
On 5 February 2016, The SAC referred the question to Luxembourg almost verbatim as prepared by ASJP. As one person familiar with the case put it, 'we made everything for the [SAC]' (Interview 2). Still, ASJP harbored doubts about the chances of the case reaching Luxembourg. Interviewees felt Portuguese courts were at the time relatively reluctant to refer questions to the ECJ due to lack of familiarity with EU law, the delay generated as a result (Interview 2) or feeling of own importance in the case of peak courts (Interview 3).

Demand side: judicial independence as a pillar of European integration
While it is broadly true that the ECJ on average prefers outcomes that favor European integration, the Court has good reasons to be specifically invested in the defense of the principle of judicial independence at the national level. National courts are key conduits for the enforcement of EU law, including ECJ rulings. Decreasing their independence is likely to hamper their ability to participate in the preliminary ruling procedure and enforce domestically unpopular EU commitments (Mayoral & Wind, 2021). Moreover, ECJ judges, many of whom have domestic experience, might feel professional solidarity towards their national colleagues and view court packing or court curbing measures as negatively impacting their careers and life.
There is little doubt that the ECJ cared about judicial independence prior to the ASJP saga. In the Wilson case, the Courtunder the rapporteurship of later President Koen Lenaertsemphasized the importance of judicial independence where EU law required the Member States to make available remedies before 'a court or tribunal'. And while judicial independence formed part of a 'general principle of Community law stemming from the constitutional traditions common to the Member States and enshrined in Articles 6 and 13 of the European Convention for the Protection of Human Rights' (Wilson, para 46), the applicability of this principle was still subject to finding a material connection between EU and national law in any given case. It was not until ASJP that the ECJ took the radical and consequential step of interpreting Article 19(1) TEU as a general obligation that all national courts be independent regardless of the substance of the case. This section traces how the ECJ's attitude towards austerity cases shifted from near-complete disinterest to finding the perfect ground for a jurisprudential breakthrough.

Austerity case law
The ECJ was initially uninterested in preliminary references concerning the impact of austerity measuresin particular salary cutson fundamental rights. It declined jurisdiction in at least six cases from Portugal and Romania between 2011 and 2014. 4 The Court argued in all its dismissals that the Member States were not 'implementing Union law' in the meaning of Article 51(1) of the Charter which prevented it from ruling on these cases.
As the effects of the economic crisis recededand to some extent coinciding with the appointment of Koen Lenaerts as President of the Court in October 2015the ECJ began to open the door to some austerity related cases. In September 2016, the ECJ ruled that the Commission and the European Central Bank could be sued for damages for their involvement in the implementation of bailouts under the European Stability Mechanism. 5 However, the applicability of the Charter to Member States' action remained untouched in Ledra, not least because the Mechanism was outside the EU legal order.
That changed in Florescu. 6 Romania was found to be implementing EU law for the purposes of Article 51(1) of the Charter when it prohibited combining the receipt of pension with income from other activities for public sector retirees. Curiously, like the ASJP case, the Florescu action was brought by judges. They complained not only about the lost pension resulting from their university teaching income, but also about the fact that the measure exempted holders of legislative, executive and high judicial offices from its impact.
While on the merits the ECJ ruled against the Romanian judges, the Court found itself competent to interpret an MoU between the EU and Romania and the national austerity measures in question to constitute an implementation of that MoU, triggering the applicability of the Charter. As others have observed (Dermine & Markakis, 2018), this was a notable departure from the previously dismissive ECJ case law on national austerity measures.
The similarities between the Florescu case and the ASJP case are numerous. Both cases were initiated by judges whose income was affected by austerity laws. Both actions concerned national measures taken at least partly in response to conditions attached to EU emergency financial assistance which were at least partly expressed in MoUs. The question of the applicability of the Charter loomed over both cases. In both cases, the ECJ took the case law forward in a pro-integration direction.
The key difference between Florescu and ASJP lay in the invocation of Article 19 TEU and the judicial independence framing in the latter. This is not to say that the ECJ could not have added this perspective to the case in Florescu, but it would do so at the risk of drawing considerable criticism and appearing overly activist. Instead, the Court opted for an incremental expansion of the applicability of the Charter which, albeit potentially useful, was no silver bullet in the fight against rule of law backsliding. Importantly, when Florescu was being decided by the ECJ, the Court had already received the ASJP preliminary reference and heard the AG opinion. The ECJ was therefore aware that a case spontaneously raising the issue of judicial independence was on the horizon. A radical constitutional interpretation was bound to be received with less opposition when it built on arguments developed by national judges and after Florescu provided a steppingstone towards ECJ jurisdiction over austerity cases.

Political gridlock in the rule of law crisis
In parallel with the legal fallout from the austerity era, the EU was becoming gradually more preoccupied with a deepening rule of law crisis in Hungary and Poland. The European Parliament's 2013 report on the situation of fundamental rights in Hungary was the first institutional document highlighting the democratic backsliding taking place under Viktor Orbán's government. Out of a myriad of issues, only the firing of the data protection supervisor (C-288/12) and lowering the compulsory retirement of judges, prosecutors and notaries (C-286/12) were followed up by the Commission through infringement proceedings. Although the Hungarian government was transparently upsetting the domestic balance of power by targeting independent institutions, the Commission relied on age discrimination legislation in the latter case to bring Hungary before the ECJ. Because in most areas of backslidingunlike data protectionthere was no relevant EU legislation, most developments received no institutional pushback from the Commission and the Council. Interestingly, at this time the ECJ signaled no interest in making a landmark ruling on national judicial independence by not assigning case C-286/12 to its Grand Chamber, an expanded court formation of 15 judges where interpretive advancements usually take place.
Instead of pursuing more infringement actions or cutting EU funding to the offending Member States, the Commission's next move was to propose the creation of monitoring instruments such as the Justice Scoreboard and a 'rule of law framework', a kind of pre-Article 7 TEU procedure. For its part, the Council introduced an annual dialogue on the rule of law. Accompanied by a 'rhetoric of inaction' (Emmons & Pavone, 2021), the EU converged on an 'authoritarian equilibrium' (Kelemen, 2020) with its political institutionsmost of all the intergovernmental Councilproving distinctly unwilling to sanction Hungary or Poland (Closa, 2020).
It was not until December 2017 in the case of Poland and September 2018 in the case of Hungary that Article 7 TEU was triggered. Nonetheless, it soon became clear that even the initiation of a procedure previously known as the 'nuclear option' would do little to resolve the institutional gridlock over the rule of law crisis. By 2017 undermining judicial independence contributed to a 42 and 31 per cent fall in Hungary's and Poland's liberal democracy score according to data by the Varieties of Democracy project compared to pre-Fidesz and pre-PiS levels. Despite endangering the EU's identity as a 'community based on the rule of law', European integration was 'failing to fail forward' in this area (Emmons & Pavone, 2021).

ASJP case before the ECJ
The ASJP case reached the ECJ on 5 February 2016. Written observations were lodged by ASJP's lawyerrestating the arguments presented before the SAC the Portuguese government and the Commission. The latter two also attended the oral hearing on 13 February 2017. The Advocate General delivered his opinion on 18 May 2017 and the Court ruled on 27 February 2018. The timeline can be considered standard in comparison with preliminary references lodged and decided around the same time.
It is customary for the Commission and the Member State from which the preliminary reference originates to submit observations to the ECJ. Other Member States tend to intervene more often in cases that are salient and important. In this instance, the low salience of the case combined with the wording of the preliminary question successfully obscured the intention of the ECJ to hand down a landmark ruling, resulting in no additional Member State observations being submitted (see Figure 1).
It is not only the Member States who misjudged the importance of the ASJP preliminary reference. Despite being already at the time embroiled in an escalating rule of law conflict with Poland, the Commission argued for the case to be declared inadmissible. Instead of pleading for a robust judicial independence doctrine and an expansion of the EU's rule of law competences which would be consistent with the Commission's overall pro-supranational preferences in court disputes (Ovádek, 2021) the Commission missed the strategic opportunity represented by the ASJP's action and treated the case as just another austerity preliminary reference that should be declared inadmissible.
It is clear from the summary of arguments prepared by the Advocate General that the plea of inadmissibility centered on the question whether the Portuguese government was implementing EU law within the meaning of Article 51 CFR. Both Portugal and the Commission considered the link between the austerity measures and EU law to be insufficient to satisfy this requirement with the consequence that the ECJ would not be empowered to rule on the case. Like ASJP, neither of the opposing third parties thought about distinguishing the scope of applicability of Article 47 CFR and Article 19(1) TEU.
In contrast, Advocate General Saugmandsgaard Øe emphasized the distinction between the two provisions. He concluded that the wording of Article 19(1) TEU warranted consideration of the case separate from the question of the applicability of the Charter, although he was also of the opinion that the conditions for the applicability of the Charter were satisfied, so the preliminary reference would be admissible under both headings. However, when it comes to the substance of Article 19(1) TEU, the Advocate General did not think the provision, and in particular the phrase 'effective judicial protection', extended to the principle of judicial independence. That principle was to be found in Article 47 CFR but not in Article 19 (1) TEU. Either way, the Advocate General considered the austerity measures to not violate judicial independence and recommended the ECJ rules against ASJP.
While arriving to the same conclusion, the ECJ adopted a radically different, and globally more strategic, interpretation of Article 19(1) TEU. The Court used the different wording of Article 19(1) TEU ('fields covered by Union law') to make the point that it applies regardless of whether the Member States are 'implementing Union law' within the meaning of Article 51(1) CFR. The ECJ proceeded to connect Article 19(1) TEU to the more general value of the rule of law, enshrined in Article 2 TEU, as well as Article 47 CFR which governs specifically the conditions attached to the 'right to an effective remedy and to a fair trial'. Regarding the latter, it is notable that the ECJ spent no time on justifying why the expression 'effective legal protection' appearing in Article 19(1) TEU is equivalent to the principle of effective judicial protection, a general principle of EU law that has been codified in Article 47 CFR. It is equally notable that, consistent with theoretical expectations, the ECJ reformulated the question referred by the SAC by dropping mention of Article 47 CFR and keeping the focus solely on Article 19(1) TEU (see Appendix for full wording).
Nonetheless, the most important consequence of the ASJP ruling resides in the extension of the EU law obligation that courts must be independentas part of the principle of effective judicial protectionto all national courts by virtue of the fact that they are also EU courts. This logic is different from that traditionally governing the use of general principles of EU law or indeed the applicability of the Charter, which were always premised on the existence of a substantive connection between EU and national law in a specific case. This novel interpretation would enable the ECJand the EU more generallyto address violations of judicial independence in the Member States directly without the need to artificially connect the case to a substantive provision of EU law. 7 ASJP revolutionized Article 19(1) TEU. A provision that was scarcely mentioned before became the centerpiece of a judicial agenda aimed at safeguarding national judicial independence. Figure 2 illustrates this point quantitatively. Prior to ASJP, 31 rulings or AG opinions mentioned Article 19(1) TEU and typically only in one paragraph. In less than three years following the judgment, the provision appeared in 60 texts, frequently in more than 10 paragraphs, indicating a more substantive deployment. Following the ASJP ruling, judgments using Article 19(1) TEU referred to the provision in over nine more paragraphs where the case concerned judicial independence (see Appendix for regression results). The ECJ went from barely discussing judicial independence to weighing in on politically sensitive questions of judicial organization in Malta, Romania, Germany and, above all, Poland.  (1) TEU. Each data point in the plot represents the number of paragraphs mentioning the second paragraph of Article 19(1) TEU in one ECJ case. We also distinguish cases by whether the term 'judicial independence' is mentioned to ensure the effect is not driven by irrelevant cases. The fitted lines are drawn from bivariate Poisson regression.
The new doctrine found immediate use in cases concerning rule of law backsliding in Poland (Pech & Kochenov, 2021) and it was not long until the Commission came around to recognizing its importance for triggering infringement procedures. Instead of searching for EU laws that might tangentially apply to backsliding practices, the Commission could finally tackle domestic interference with judicial independence directly and with more realistic implications than in the context of Article 7 TEU. The new interpretation given to Article 19(1) TEU by the ECJ drew the EU much more closely into the rule of law crisis in Poland.
Three and a half years and more than 20 preliminary references and infringement actions later, the ASJP revolution enabled the ECJ to fine Poland 1 million EUR per day for non-compliance with previous decisions concerning judicial independence (C-204/21).
The rise in the quantity and prominence of Article 19(1) TEU case law was so stark that even the Court's insiders worried where it might lead. Advocate General Bobek in his opinion on judicial reforms in Romania warned of 'gates which are too open'. Yet, the ECJ has been arguably aware of the need to set boundaries for Article 19(1) TEU. It refused to be drawn into a dispute between an Austrian judge and his colleagues regarding case allocation (C-256/19). It also declared inadmissible two preliminary references from Poland on account of being disconnected from EU law, an argument that could well have applied in ASJP (C-558/18 and C-563/18). And while it reaffirmed the importance of Article 19(1) TEU in a case about the Maltese judiciary (C-896/19), the ECJ decided not to question the decisive role of the Maltese prime minister in the appointment of judges. The very choice of focusing interpretation on Article 19(1) TEU, which is limited to court matters, as opposed to relaxing constraints on the applicability of the Charter, can be considered conservative. As previous studies have shown, the Court is wary of overly antagonizing its most important stakeholdersthe Member States.

Member States' preferences
Based on existing research, we have hypothesized that the Court would be less likely to adopt far-reaching interpretations in the presence of significant Member State opposition and vice versa. Nonetheless, the covert strategy by which the ECJ effectuated doctrinal change in ASJP created an information deficit, as the Member States did not submit observations on the case. While the covert strategyif successfulhas the obvious benefit of avoiding open confrontation with the antagonistic Member States when the landmark ruling is in the making, the downside is that the Court receives less information about other Member States' preferences. In theory (Larsson & Naurin, 2016, p. 378), such an information deficit increases uncertainty and makes government backlash more likely. However, if the ECJ could still gauge Member State reactions to its ruling and conclude the majority would not oppose the intended interpretation, the trade-off in the covert strategy would be worthwhile.
We argue that even in the absence of written observations in the ASJP case, the Court could credibly anticipate that the majority of the Member States would not oppose its landmark ruling. In December 2017, just a few months before the ASJP judgment, the French and German leaders publicly and unambiguously supported the Commission triggering (for the first time ever) Article 7 TEU against Poland. 8 Such open backing for EU enforcement of the rule of law by the two most important Member States was evidence that patience with the backsliding Member States in the Council was running out. As shown by Closa (2019), the very act of triggering Article 7 TEU by the Commission was predicated on Member State support which was gradually rallying since 2015 as the Council quietly dropped opposition to the Commission's rule of law framework. At the same time, one of the most vocal opponents of EU action, the UK, was in the process of exiting the Union. Furthermore, the Court could see that the majority of MEPs, many of whom were members of government parties, were since 2012 in favor of more EU action against rule of law backsliding in Hungary and Poland (see Appendix). The Court was likely to have additional information through informal networks between judges and political elites but even on the basis of publicly available statements, the ECJ could have credibly anticipated the majority of Member States to not oppose a landmark ruling on judicial independence.
Member State interventions in cases following ASJP largely vindicated the Court's boldness. Five Member States -Belgium, the Netherlands, Denmark, Sweden and Finlandhave openly supported the new interpretation when it was applied by the Commission in an infringement procedure against Poland (case C-791/19). Similarly, Latvia endorsed the use of the new doctrine to find the controversial Polish disciplinary chamber as lacking in independence (C-585/18, C-624/18, C-625/18 and C-558/18, C-563/18). Implicit support can be deduced from observations submitted by Ireland and Spain, respectively, in two cases (C-216/18 PPU and C-354/20 PPU) questioning the viability of European arrest warrants issued by Polish courts in light of the rule of law crisis. Furthermore, when Germany and Malta were faced with preliminary references questioning aspects of their judicial systems in view of Article 19(1) TEU (C-272/19 and C-896/19), they defended the status quo in their respective countries without attacking the ECJ's jurisdiction to rule on the matter. Romania found itself in a similar situation in a series of cases known as Asociaţia Forumul Judecătorilor din România, but with a twist. As noted by Advocate General Bobek, Romania's position on Article 19(1) TEU changed during the case, as the old socialist government which was responsible for violations of judicial independence gave way to a center-right coalition in November 2019. This means that in the immediate aftermath of the ASJP ruling, Romania should be considered to have sided with Poland and Hungary, the two main resisters and prime targetsespecially in the case of Polandof the new doctrine.
We thus find six Member States repeatedly supporting the ASJP breakthrough and two opposing it. The remaining Member States are best classed as belonging to a non-opposition group: they are either happy for the ECJ to have advanced EU protection of judicial independence without their personal investmentwhich would be compatible with Macron's and Merkel's joint statement supporting the idea of EU enforcement in December 2017or not sufficiently aggrieved by the change to speak out against it. It is particularly notable given the high number of cases concerning Poland that it was unable to muster support from any country other than Hungary.
In fact, the adoption of Regulation 2020/2092 on a general regime of conditionality for the protection of the Union budget in December 2020 further suggests that opposition to the new Article 19(1) TEU interpretation was remarkably low. Although using less direct language, recital 12 of the Regulation partly codified the new doctrine by linking Article 19 TEU and judicial independence and explicitly referring to the ASJP judgment in footnotes. The Regulation was adopted in accordance with the ordinary legislative procedure which requires the Council to decide by a qualified majority vote. All Member States apart from Hungary and Poland voted in favor of the Regulation. Thus, the paradoxical outcome of the ECJ's covert strategy was democratic legitimization of strengthened EU rule of law oversight.
In sum, the ECJ was correct to anticipate any backlash to its ASJP doctrine to be limited to only a few Member States. Post-ASJP position-taking by the Member States confirms that the Court's ruling enjoyed broad support and ran little risk of override. Taken together with the absence of public contestation, Blauberger and Martinsen (2020, p. 387) call this situation 'permissive consensus'. Instead of rejecting greater EU involvement in matters of national judicial independence, the majority of Member States embraced and entrenched (through Regulation 2020/2092) the doctrinal change. There are four possible, mutually compatible explanations for that: first, EU-level protection of judicial independence may help preserve the ability of national courts to enforce, including domestically unpopular, EU commitments. Second, a Member State might be ideologically opposed to the degradation of liberal democracy through interference with the independence of the judiciary. Third, a Member State may believe that the new doctrine would entail no negative implications for itself while enabling it to reap a transactional or reputational reward from non-opposition. Fourth, a Member State may perceive the reputational cost of challenging the ECJ as outweighing any domestic benefit of backlash politics. We leave it to future research to explain which factors motivate Member States' support and non-opposition to ECJdriven constitutional change.

Discussion
In lieu of a conclusion, we discuss the ASJP storyfrom mobilization in Portugal to Member States' reactions to the rulingin the context of our theoretical expectations. On the supply side, the empirical investigation showed that what Conant et al. (2018) call the meso level, consisting of the national legal-political context, was particularly significant. The applicants only turned to EU law arguments once their original strategy of having the Portuguese Constitutional Court rule on the compatibility of salary cuts with judicial independence failed. By turning to Europe only once national remedies proved unavailable and political compromise impossible, the ASJP followed a path trodden by many other litigants, like those of the Genoa port in the 1990s (Pavone, 2019). But unlike the effort to liberalize the Genoa port, the ASJP case was not a highly salient cause in search of litigants driven by enthusiastic Euro-lawyers. Rather, from the outset it was a relatively quiet case with welldefined parameters and a clear objectivesecuring the independent standing and material conditions of Portuguese judges at a time of economic crisis and austerity measures.
Apart from domestic obstacles to achieving their policy objectives, the other factors that influenced the turn to EU law were micro-level: the identity, background and networks of those driving the litigation on behalf of ASJP, in particular of the then ASJP president, José António Mouraz Lopes. His and some of his colleagues' involvement in European and international judicial networks made them cognizant of the possibility to re-frame what was originally a purely domestic issue in terms of international law and standards of judicial independence. Still, Mouraz Lopes was not a 'Euro-lawyer' with an instinct to immediately involve the ECJ in their case (Pavone, 2019). He instead hired lawyers, or more specifically legal academicson the personal recommendation of a close collaboratorto work out the international and European framing of the case. In this respect, the ASJP story deviates from established accounts of how lawyers manufacture EU law cases at the national level and instead points to a need for future research to clarify connections between legal academia and EU law litigation.
Importantly, it is worth emphasizing that the ECJ's ASJP ruling and the significance it almost immediately acquired in EU law was completely dissociated from the litigants' objectives. While very much sympathetic to the plight of persecuted Polish and Hungarian judges nowadays, when bringing the case ASJP's leadership was only preoccupied with the predicament of Portuguese judges subjected to significant salary cuts. If the timing of the eventual ECJ ruling seemed too good to be true in light of the unraveling of the Polish rule of law crisis, this was not due to the original litigants' desire to reshape EU law on judicial independence but due to the Court's strategic use of preliminary references. The wording of the referred question, which is fully traceable to litigants' drafting, was sufficiently innocuous for Member States to eventually overlook the intention of the ECJ to hand down a landmark ruling.
Turning to the demand side, we have seen that the ECJ does not jump at the first opportunity to hand down a landmark ruling but instead chooses, in the presence of alternatives, the most suitable case. Assuming the Court's objective was to enhance EU law protection of judicial independence at the national level, this goal was better served by the ASJP case than the Florescu case. Even so, the ECJ still reformulated the question referred in ASJP to better set up its novel reading of Article 19(1) TEU while downplaying limitations related to the applicability of the Charter. This finding to some extent differs from the conclusion reached by Šadl and Wallerman (2019, p. 431), namely that reformulation is unrelated to the Court's efforts to expand the domain of EU law. The fact that the ECJ first used a comparatively inconspicuous case to create the new interpretation of Article 19(1) TEU before applying it to the highly salient context of the Polish rule of law crisis also indicates strategic thinking and the importance of precedent as a legitimizing rhetorical device (Larsson et al., 2017).
The political context in which the ECJ hands down its landmark rulings has shifted significantly since the days of Costa/ENEL and van Gend en Loos. In particular, both European integration and international courts have become more politicized and vulnerable to popular backlash. Concretely, the Court and its perceived federalist bias have been an explicit and prominent target of populist discourse in, for example, the Brexit campaign. Perhaps surprisingly, though consistent with Blauberger and Martinsen (2020), our case study of the groundbreaking ASJP ruling suggests that one way in which the Court responded to the more challenging political context has been to double down on the 'technocratic silence' (Pavone, 2019, p. 852) of its decision-making. Indeed, the expansion of EU regulation of national judicial independence might have been the 'quietest' of the ECJ's 'quiet revolutions' (Weiler, 1994): not only did the ASJP case receive little attention, but the significance with which the Court endowed it was completely unanticipated by the original litigants, national governments and even the Commission.
The covert strategy by which the ECJ enacted the novel constitutional interpretation is only one side of the coin, however. The same fear of backlash that might make the Court double down on 'technocratic silence' also creates incentives to radically advance European integration only where the policy change enjoys majority backing from the Member States, the anticipation of which is rendered more difficult due to the inherent trade-off between covertness and availability of information about Member States' preferences. The Court's strategy in ASJP thus worsens informational hurdles that are already higher for international courts compared to domestic courts (Larsson & Naurin, 2016, p. 383), thereby potentially increasing the risk of override. As a result, covertness is unlikely to become a default decision-making mode for the ECJ. The question is not only whether the Court operates under the enabling conditions of a 'permissive consensus' (Blauberger & Martinsen, 2020, p. 387)which in ASJP we have retrospectively established it has beenbut also how much information the ECJ can be reasonably assumed to possess prior to making a decision. In ASJP, the informational loss was mitigated by political signals expressed outside the judicial procedure, highlighting the need to view proceedings before the Court in the context of the broader EU political system.
Conversely, the Member States and the Commission stand to benefit from the Court's interpretive leap in a situation of political gridlock such as the one arising around the rule of law crisis in Poland and Hungary. From a wider perspective, the Court's ruling in ASJP and subsequent policy developments are therefore a manifestation of the different ways in which the EU tries to manage the tension between technocracy and democratic legitimacy (Rauh, 2016), which we can also look at as an adaptation of the idea of supreme courts as essentially majoritarian institutions, at least insofar as legislative (Member State) majority is concerned.
The Member States embracing ex-post the Court's judicial independence doctrine poses broader questions about the legitimacy of this way of enacting policy. On its own, the Court's expansive interpretation, advancing integration through the backdoor, would be a prime target of normative critiques levied against ECJ-led integration through law (Scharpf, 2012). But when the Member States swiftly codify the Court's new doctrine in legislation, those critiques lose some of their edge. Still, the Member States' willingness to defer to the Court's activism in lieu of tackling recalcitrant governments head-on remains questionable.