Preserving the Rule of Law Through Transnational Soft Law: The Cooperation and Verification Mechanism

Abstract

This contribution reflects on the role of soft law instruments to address the rule of law crisis, a topic of high relevance in the context of this special issue. Indeed, the EU-15 enlargement towards the ‘periphery’ proceeded also through soft law instruments, perceived to allow greater flexibility in monitoring. Soft law – or rules of conduct that have no legally binding force but may have legal and practical effects (Snyder 1993, 64) have been considered inappropriate by the literature in dealing with the regulation of values in EU law. However, a closer look at the career of soft law issued for Romania under the Cooperation and Verification Mechanism appears to suggest a more nuanced picture. Whilst time is not yet ripe for a full assessment of the effectiveness of the tools, there are signs that show that the reports issued by the Commission within the Cooperation and Verification Mechanism process have had an influence in fostering change. They did so by catalysing litigation and, as a result, informing and fuelling an institutional dialogue on the rule of law at the transnational level.

1 Introduction

The EU rule of law crisis, or backsliding, first documented by President Barroso,¹ has acquired gargantuan proportions in recent times. It is characterised by a reduction of judicial independence in some Member States and political capture of some supreme courts, because of rampant corruption, or indeed, propagating corruption at various levels. Given the complete ineffectiveness of the procedure provided in Article 7 TEU, certain concrete actions have been implemented, such as the Justice Scoreboard first published in 2013 by the Commission,² the Rule of Law Initiative and the Rule of Law Dialogue launched by the Commission³ and the Council,⁴ respectively, in 2014, and the Rule of Law Mechanism, first introduced in 2014 and overhauled in 2020. The latter is conceived as an annual dialogue between the Commission, the Council, and the European Parliament, together with Member States as well as national parliaments, civil society and other stakeholders on the rule of law.

All these mechanisms and instruments are characteristic of the “new” governance – as opposed to government, whereby elected representatives pass legislation, which is in turn implemented by the executive, with courts ensuring judicial scrutiny (Treib et al 2007). New governance methods can rely on institutional infrastructures falling entirely outside the realm of traditional branches of ‘constitutional’ government (Iancu and Tănăsescu 2018). They generally consist of information sharing through reporting and discussion supporting progressive learning and innovation through non-binding rules such as recommendations, opinions or reports, also known as ‘soft law’ (Korver and Oeij 2005; Sabel and Zeitlin 2008). Even though tagged as ‘new’ governance, such tools have been a feature of EU law since the 1960s but have become increasingly common over the years. For instance, the EU response to the COVID-19 pandemic was initially conveyed mostly through soft law showing that, in times of crisis, soft law can act as a quick-fix (Ștefan 2020). But soft law fills in also a different function, which is to stimulate conversation between national and European authorities and courts to encourage the harmonious application of EU law in the Member States, as well as an experimentalist way to find common solutions to problems on which there are differences of opinion (Jacobsson 2004; Sabel and Zeitlin 2008).

This article argues that both crisis-solving, as well as the dialogical function of soft law are key in rule of law regulation. One of the manifestations of the rule of law crisis is the rejection of rule of law values à l’Européenne by powerful apex courts of Central and Eastern Europe. This is generally done under the pretext of national constitutional identity and the respect of national sovereignty. The literature concurs, with Krastev and Holmes (2020) blaming the current crisis of the rule of law in Europe on differences between East and West over the interpretation of fundamental principles. They argue that liberal values were imposed by the West on Eastern Europe in return for a promise of welfare that did not materialise. However, although differences on the interpretation of certain principles exist between all EU countries, Weiler (2021) is right to warn that a governance of a majority without the constraints of the rule of law and human rights is nothing short of the tyranny of the majority, ready to fundamentally change a people whenever a new government is elected. In the same context, Perju (2020 at 1016–1019) advises caution in persisting in classical EU theories advocating for constitutional dualism at the EU level, and he militates for a European hierarchical structure which should allow for critical dialogues (at 1021–1022). Dialogue between the different legal orders and their courts is key also for Kochenov (2024). He believes there is still hope for constitutional pluralism to triumph in protecting fundamental values, on the condition that both Luxembourg and its national interlocutors are genuinely invested in exchange and cooperation.

Can the regulatory framework help foster this dialogue? At least in theory, a hybrid framework relying on binding law and non-binding, experimental, dialogical rules should be ideal for this. This could follow established models identified in the literature, according to which new governance (or soft law) can serve as compliance/enforcement mechanism for traditional hard law and vice versa (Trubek and Trubek 2007; de Búrca and Scott 2006) or indeed could help connect the various layers transnational governance (Ștefan 2014). But the literature criticizes at length the use of soft law as a tool for solving the rule of law crisis. Kochenov (2017) invites for a new approach to the enforcement of EU values, whilst showing that EU enlargement law it itself could be considered soft law rather than binding acquis. Pech (2021, 322) considers that non-binding tools “neither help contain backsliding nor reverse the damage done to the rule of law.” EU institutional belief in dialogue appear thus naïve, confronted to the systematic undermining of democracy in some Member States. Grogan and Van Dam (2023, 402) undertake an evaluation of the tools, reaching similar results: even in its most formal manifestations relying on strong steering pressures, such instruments cannot foster dialogue. Indeed, given that, proverbially, it takes two to tango, soft law would not lead to change in countries which are impervious to recommendations and do not engage in any form of exchange with the European institutions.

Relying on and extrapolating from a Romanian case study, this article advocates for a more nuanced view. Whereas soft law cannot be a panacea to address rule of law concerns,⁵ it has the capacity, in the long run, to stimulate change through a dialogue involving the judiciary but also the executive or indeed the legislative branch of traditional government. In line with the theme of the special issue,⁶ the article traces the career of a particular type of EU soft law, the reports issued by the EU Commission under the Cooperation and Verification Mechanism (CVM). The CVM is a hybrid, soft/hard law instrument, which was used in the context of the EU-25 enlargement. The article shows how the soft law issued under the CVM has stimulated an important dialogue at various levels of the civil society, the national and transnational institutions, eliciting changes in national hard law. Whilst not engaging with an evaluation of these changes, the ambition of this article is to show the multidimensional character of the dialogue generated by soft law and its clear potential for incremental change. After an initial complete refusal to enter a dialogue from the Constitutional Court of Romania (CCR), transnational soft law acted as a catalyst for the civil society, as well as for private litigants, and, little by little, began to permeate the official discourse at the national level. Since the preservation of values is an enterprise exceeding strict legal frameworks and requiring culture change or maintaining a certain standard, soft law – and hybrid soft-hard regulatory frameworks – are unavoidable. This finding is important as it illustrates the potential of other soft law instruments in the area of the rule of law to follow a similar path. As shown by Dumbravă (2021, 442–443), the similarities between the CVM and the rule of law mechanism mean that the Romanian case law on the rule of law is not CVM-specific, and can be generalised to the reports produced under the rule of law mechanism. Furthermore, the Commission has recently embarked in a structured rule of law dialogue based on soft law in order to address issues with the Spanish Council of the Judiciary,⁷ showing that such techniques are constantly resorted to.

The article proceeds as follows. In its second part, it outlines the various soft law instruments at play in the Romanian case, with a primary focus on the Commission reports in the framework of the Cooperation and Verification Mechanism. Part three unpacks the judicial dialogue and conflict generated by these instruments. Part four illustrates how soft law has served as a catalyst for legislative change in Romania by mobilising the stakeholders (and even the civil society)⁸ as well as the courts, whilst fuelling transnational soft law dialogue.

2 The Case Study: the Cooperation and Verification Mechanism as a New Governance Tool Based on Soft Law

The Cooperation and Verification Mechanism is a control tool employed for Romania and Bulgaria after their accession to the European Union. Indeed, following the closure of negotiations, outstanding issues remained. Therefore, the Treaty of Accession included a supplementary safeguard to those employed for the previous eastward enlargement. On the basis of this Treaty (Articles 37 and 38), a Cooperation and Verification Mechanism Decision (hereinafter CVM Decision)⁹ was issued by the Commission with effect from accession day, and touching on issues in the reform of the judiciary and the fight against corruption. The procedure was as follows: Romania needed to submit regular reports to the Commission on the progress towards the benchmarks, while the Commission was supposed to submit its comments and conclusions in a Report to the European Parliament and the Council. The latter reports also contained recommendations and are thus akin to non-binding instruments mentioned in Article 288(4) TFEU. Even though reports are not mentioned in the Treaty, they are widely used in European law. For instance, the Commission publishes every year a report on competition law, which often contains, in addition to statistics and descriptions of enforcement activity, rules of conduct.¹⁰ Moreover, the Open Method of Coordination is also based on multiple Commission reports. Such instruments belong to the soft law category, defined as “non-legally binding instruments, but which can produce practical and legal effects.” (Snyder 1993, 64). The CVM regulatory framework is therefore hybrid (Trubek and Trubek 2007, 539), with the CVM Decision, binding hard law, being operationalised by the non-biding reports of the Commission.

As with any hybrid mechanisms, the CVM is exposed to the various critiques concerning soft law. The first one is legitimacy, with soft law being criticized often for lack of appropriate safeguards and involvement of stakeholders, or indeed, the transparency of proceedings (Ștefan 2020). In the case of the CVM, reports up to 2019 are accompanied by technical notes regarding their drafting, the involvement of the stakeholders, as well as the evidence on which the Commission based its conclusions. Regrettably, this was not always the case, with recent reports lacking such detail –this default covers only the last two years. Furthermore, the procedure for issuing such reports was not always followed to the letter. The CVM decision required, under Article 2 (a clear rule of hard EU law), reports to be published at least every six months, yet, since accession, the Commission did not always publish two reports per year. During the pandemic, there was no report for Romania, leaving the period October 2019-June 2021 unreported under the CVM Decision.¹¹ However, a Country Chapter Romania was issued in September 2020 under the distinct procedures of the Rule of Law Mechanism.¹²

A second problem is soft law overload – this happens at both the EU and transnational level – although, as we will see in Sect. 4.2, this is not necessarily a negative feature. Since 2020, Romania was included, as all other Member States, within the rule of law mechanism, a structured annual process of cooperation between the Commission, Council, Parliament, Member States and stakeholders. Similarly to the CVM, the rule of law mechanism ultimately presupposes reports drafted by the Commission after a process of consultation with various organisations including fact-finding missions in the Member States.¹³ For Romania, two reports were drafted in 2021 and 2022 respectively under the CVM and the Rule of Law Mechanism – two different procedures duplicating, in substance, the assessments, as even a cursory reading of the reports produced would reveal. As shown by Dimitrovs and Kochenov (2021), the CVM procedure had a discriminatory and unnecessary character, having become even redundant after the Court formulated the principle of non-regression in Repubblika¹⁴ (Leloup et al. 2021). Given the large-scale violations of the rule of law in states that have not been subject to this additional conditionality, such as Poland or Hungary, all European states should have been subject to similar monitoring. This is the purpose of the Rule of Law Mechanism, which consists of reporting mechanisms and feedback from the Commission similar with the CVM. It appears logical that, in 2023, the Commission formally¹⁵ closed the CVM for Romania, with only the Rule of Law Mechanism applying. Yet, other transnational soft law has also been at play in the case of Romania. The EU is not the only organisation preoccupied by the rule of law crisis. At the level of the Council of Europe, several mechanisms are in place for many years now. The Venice Commission (or the European Commission for Democracy through Law) is an advisory body on constitutional matters, which issues opinions on a range of rule of law relevant topics. Similarly, the Group of States Against Corruption (GRECO) monitors compliance with anti-corruption standards, and, in so doing, issues various reports and recommendations. Whilst through its numerous reports the Venice Commission was tagged as a “norm entrepreneur”, such instruments are also plagued by various legitimacy/transparency issues (Iancu 2019).

The third problem with soft law is its effectiveness. In the case of the CVM, the binding Decision is only outlining four benchmarks on which progress needs to be reported. It is thus the Commission non-binding report which operationalises these benchmarks. This begs the question if such instrument can produce any effects. Over the years, the CJEU has accepted the legal effects of soft law instruments in particular where they served a purpose of “regulation by publication” (Hofmann 2006, 169–170) or “regulation by information” (Snyder 1994, 199–201). One of the major objectives of soft law rules is to increase transparency in the application of European law by explaining the Treaty, regulations or directives to individuals and authorities alike. The publication of soft law rules increases both legal certainty as well as the potential for uniform application of European law. Consequently, these instruments are often binding on the issuing institution. The mechanism is as follows: by issuing soft law, the enacting EU institution limits its discretion and is obliged to apply that instrument in the name of general principles of law such as equality, respect for legal certainty and legitimate expectations.¹⁶ Things get more complicated at the national level. The Court of Justice stated in Grimaldi¹⁷ that national courts are obliged to take soft law into consideration – a rather vague pronouncement that has not been clarified by the case law (Korkea-aho 2018). Of course, disregarding soft law might have consequences down the line, when such instruments are applied in EU decisions imposing obligations for Member States (such as, for example, to recover aid illegally granted to industry).¹⁸ Be that as it may, national authorities are not, in principle, bound by the non-binding rules issued by the European institutions; these instruments are not covered by the principle of supremacy. Thus, the main function of soft law is to act as an instrument of dialogue between the various levels of EU governance.

Yet, as empirical research shows, soft law is not static: it evolves as a result of a dialogue between various levels of governance, but it also can foster change at the national level ‘socialising’ institutions to EU principles and rules (Eliantonio et al. 2021). The rejection of soft law instruments at the national level may lead to the modification of these instruments at the European level, following a dialogue between the various institutions. For example, in Expédia, the Court clarified that a notice on the de minimis threshold in competition law had no effect for the French competition authority, which was free to decide on the impact of a clause in an agreement between undertakings, even if that agreement covered a very small protion of the market.¹⁹ The direct consequence of the exchange between the national and European courts in this case was the modification of the European de minimis notice, which proved to be ill-adapted to the reality on the ground. Also, in areas such as State aid, some soft law rules are negotiated, and thus become binding on Member States’ authorities.²⁰ The European legislative framework sometimes reinforces dialogue by employing soft law rules, as is the case in the area of financial market supervision. The European Supervisory Authorities’ operating regulations require the competent national authorities to confirm that they intend to comply with the guidelines and recommendations issued by the Supervisory Authorities, or, otherwise, to state their reasons for so doing. Sometimes soft law is used in an opportunistic way, for example where soft law rules are useful to resolve a concrete situation, such as complicated definitions in environmental law.²¹This opportunistic use by courts has proved to have long-lasting effects on some Member States’ administrative institutions, which have become “socialised” with European soft law rules, and have started to apply them frequently.²² It is through these constant exchanges that soft law evolves, and can even become more legitimate, by connecting various levels of governance (Ștefan 2023).

In the next sections, this article shows that the CVM reports follow the scheme described above. First, it is through this adaptable tool that a post-accession rule of law non-regression recommendation was addressed to Romania. Second, these reports were rejected by one of Romania’s apex courts, the Constitutional Court and their status as legal instruments was not recognized. As a result of this rejection and the follow-up preliminary references, the CJEU had the opportunity to clarify its interpretation of soft law at the national level. Third, by persistently issuing reports on the rule of law situation in Romania, whilst, at the same time, drawing on other transnational soft law, the European Commission has catalysed national litigation which, in turn, has led to the building of a legal community attuned to rule of law values and ultimately to legislative change.

3 The Soft Law Side of the Romanian Rule of Law Cases

Following EU accession negotiations, a series of reforms were implemented in 2004 with regards to the fight against corruption and judicial independence. These were reversed in 2017–2019 through a series of initiatives²³ regarding, inter alia, the liability of magistrates, and a clever engineering of the reorganisation of courts and status of limitation with the aim of ensuring impunity for some corruption offences. The democratic legitimacy of the amendments to the Justice Laws is questionable, with the various ordinances issued to amend the initial justice reforms in Romania marked by significant controversy: accelerated procedures, limited debates in Parliament, and massive public anticorruption protests. Most of these amendments do not seem to correspond to needs identified by independent and serious impact studies.²⁴ In the CVM Reports, these initiatives were noted, since 2018, as a setback. In turn, CVM reports were taken into consideration by several Romanian courts of various levels, with doubts being raised with regards to the legislative amendments from a constitutionality and EU law compliance perspective. This was followed by a series of preliminary rulings from the CJEU and decisions of the Constitutional Court of Romania (CCR) expressing diametrically opposed positions on the status of CVM reports, the principle of sincere cooperation under Article 4(3), and even the principle of supremacy of European law.

A first set of decisions on the CVM reports came from the CCR. Since 2018, the CCR considers that the recommendations made in the CVM reports are not covered by the supremacy of European law under Article 148 (2) of the Romanian Constitution, as they are not binding.²⁵ This change in case law is curious, especially since, in 2012, the CCR seemed to lean towards giving constitutional status to this mechanism.²⁶ In 2019, the CCR refused to submit a question for preliminary ruling to the CJEU on the status of CVM reports. The case concerned the constitutionality of Government Emergency Ordinance (GEO) 90/2018, relating to the Section for the Investigation of Offences Committed within the Judiciary (SICOJ). The setting up of this controversial new body²⁷ to investigate magistrates within the High Court of Cassation and Justice’s Prosecutor Office was directly cited in the November 2018 CVM Report. The recommendation was to “[s]uspend immediately the implementation of the Justice laws and subsequent Emergency Ordinances”, as well as to “[r]evise the Justice laws taking fully into account the recommendations under the CVM”.²⁸ The CCR rejected the request to submit a question for preliminary ruling on the nature of this report. This was for the procedural reason that the proposed questions concerned the establishment of the SICOJ, and had no bearing on the case—the constitutionality review concerned GEO 90/2018. The CCR took the decision to rule itself on the status of the CVM reports. It applied self-referential criteria to determine the applicability of the principle of supremacy of European law over constitutional norms, namely (i) whether the norm is sufficiently clear, precise and unequivocal in itself or its meaning has been clearly, precisely and unequivocally established by the European Court of Justice and (ii) whether the norm can be circumscribed to a certain level of constitutional relevance. The Court concluded that the acts that may have an impact upon constitutionality review in Romania are: Treaties, regulations, directives, and decisions. Since the CVM Decision did not contain, in CCR’s opinion, concrete obligations, the reports issued within the framework of the CVM did not have a constitutional character either.²⁹

The CJEU had the opportunity to rule on the status of the CVM reports following various requests for preliminary rulings made by Romanian courts. One of these references, AFJR I, is a landmark case which set the basis for a principle of progression, according to which Romania had a duty to improve the rule of law situation at the national level (Moraru and Bercea 2022). For the purposes of this article, the crucial bit of AFJR I is that the referring courts asked specifically about the legal effects of CVM reports, as well as about connected transnational soft law: Venice Commission opinions and GRECO reports. The CJEU noted that the benchmarks set out in the Annex to the CVM Decision were, like the CVM Decision itself, binding on Romania. Although the CVM reports were not addressed to Romania, but to the European Parliament and the Council, they were in fact recommendations to the Member State, as they formulated requirements for achieving the benchmarks. Based on the principle of sincere cooperation enshrined in Article 4(3) TEU, the Court concluded that all national authorities “must take due account of the requirements and recommendations formulated in the reports drawn up by the Commission under that decision.”³⁰ This is significant, for two reasons.

First, AFJR I is one of the clearest confirmations that soft law can produce effects through the constitutional principle of Article 4(3) TEU. This obligation to cooperate enshrines an obligation of dialogue between the various European institutions with the ultimate aim of transforming the legislative framework and advancing European integration. Before AFJR I, the case law had recognized binding effect to soft law as a result of special duties to cooperate, such as Article 108 TFEU, requiring the Commission to cooperate with the Member States in the area of State aid.³¹ Following the same line of argument, the Court could have recognized effects to the reports by relying just on the lex specialis, the CVM Decision, given that this was, par excellence, an instrument governing institutional cooperation. Instead, by extending the reasoning to Article 4(3), the Court inferred potentially more important legal effects for CVM reports, which became connected to the constitutional structure of the European Union, and also binding evidence for national courts. Compliance with the CVM Decision could thus only be established through these reports, reflecting the hybrid nature of the legislative framework.

Second, AFJR I sheds more light on the mysterious duty of national courts to take soft law into account, which was laid out three decades ago in Grimaldi.³² In the context of the CVM reports, this duty means that, on the one hand, the Member State may not adopt measures which may compromise the outcome of the Commission’s recommendations. I do not interpret this to mean that the Member State must respect to the letter the report (this would be a binding effect), but States cannot fundamentally diverge from it, and, as a baseline, they should at least engage with the recommendations. On the other hand, if the Commission finds an incompatibility between a new national measure and the benchmarks, Romania is obliged to work with the Commission towards “overcoming the difficulties encountered”³³ – thus, an invitation to dialogue.³⁴

Only two days after AFJR I, the CCR issued its Decision 390/2021 on the unconstitutionality objections raised against the amendments to the Justice Laws introducing the SICOJ.³⁵ Claimants argued that noncompliance between those amendments and CVM reports meant noncompliance with Article 148 (4) of the Constitution (the guarantee clause that the State institutions shall respect EU law). Reinterpreting AFJR I, the CCR concluded that the CJEU “does not uphold the binding nature of the reports drawn up by the Commission pursuant to Decision 2006/928, but establishes with regard to the Romanian state authorities the duty to cooperate with an institution of the European Union (the Commission), on the basis of the principle of sincere cooperation provided for in Article 4(3) TEU.”³⁶ This conclusion is not in itself false. The CJEU has never recognised, nor can it recognise, the binding nature of soft law rules. But AFJR I requires that national institutions and courts engage with such instruments, by virtue of the principle of sincere cooperation. In Decision 390/2021, the CCR denied such engagement to courts of all levels: “the national judge cannot be put in the position of deciding to give priority to the application of recommendations to the detriment of national law, since the CVM reports do not regulate and are therefore not likely to conflict with national law.”³⁷ Courts could not use these reports as evidence to determine whether there has been a violation of the rule of law. The CCR added that this is in the interest of legal certainty, another principle of the rule of law, since, if some courts were to “leave unapplied ex officio national provisions which they consider to be contrary to European law, while others apply the same national rules, considering them to be in conformity with European law, the standard of foreseeability of the rule would be seriously undermined, leading to serious legal uncertainty and, by implication, to a violation of the principle of the rule of law.”³⁸ This statement—this time a general one, with no specific reference to CVM reports – appears to go as far as to suggest that the principle of supremacy of European law is contrary to the rule of law.³⁹

Such contestation of the principle of supremacy appears to be premised on constitutional dualism, or the convention that national and EU constitutional orders co-exist in harmony, replacing each other at the level of the highest hierarchy of norms when there is a conflict over the core of the principles and rights (Weiler 1981). Under the argument of national constitutional identity, the CCR considered itself the only court “empowered to ensure the supremacy of the Basic Law on the territory of Romania,”⁴⁰ and therefore impervious to the interpretations of the Commission or those of the Court of Justice. But a legal history analysis of the notion of constitutional identity in Romania shows that the concept is by no means established, and is marked by a hesitation between an ethnocentric identity and a Eurocentric identity. Although it has used the term since 2012, the CCR has never really made a clear choice between these different configurations, not even in Decision 390/2021 (Guțan 2022). This is not to say that the CCR needs to fully adopt the identity dictated by the CJEU – whose missionary approach is also subject to criticism (Kochenov 2024) – but at least to be open to dialogue. Yet, by adopting a formalistic approach, the CCR has emphasized the non-binding character of Commission recommendations to reject any dialogue on the rule of law on this basis.

This comes out very clearly from the text of the CCR judgments engaging with the recommendations but also from extra-judicial statements of leading magistrates. Commenting on earlier constitutional case law on CVM reports, Enache and Titirişcă⁴¹ (2021, 126) rightfully note that the CCR has no jurisdiction to decide on the legal nature of the CVM reports. However, they immediately qualify this statement pointing out that the CCR has the competence to decide whether such instruments are of ‘constitutional relevance’, a notion to be determined exclusively in relation to the case law of the CCR. The authors give no further details as to how is deciding on the ‘constitutional relevance’ of EU recommendations (a CCR prerogative, to be sure) different from deciding on the ‘legal nature’ of the same (a CJEU exclusive competence) leaving this to the readers’ imagination. In a post-Decision 390/2021 statement, the (then) President of the CCR dispelled any doubts: “It has been made very clear what domestic law means: treaties, regulations, decisions. Not reports, not opinions, not recommendations. In two of the judgments of the CJEU, it was clearly stated that they have a guiding value, they give directions for action. … The problem is that those expert committees did not have the role of a European authority. How can you make such reports legally binding? Without getting into a political approach, we said that we could not recognise the legally binding value of these recommendations.”⁴² Such a positivist approach is not unusual for Eastern Europe and may be even reminiscent of the communist era (Avbelj and Vatovec 2021).⁴³ Relying on non-binding reports to hold countries accountable seems in this context to be, as the French would put it, tendre le bâton pour se faire battre. And yet, as mentioned above, soft law is not a static, but a dynamic instrument, and its effects, as we shall see in the next section, are not dependent on one instance of rejection, even by an apex court.

4 The CJEU Multiplying the Voices: Soft Law as Catalyst

As negative reports on the rule of law have multiplied and have been rubber stamped by other transnational soft law instruments (such as Venice Commission opinions or GRECO reports), they started to create a culture shift and a pressure on the system, by catalysing public opinion, militant groups, and indeed, public institutions. In order to illustrate this, this section shows three examples of dialogue: the dialogue between EU and national courts, especially following CCR Decision 390/2021, the transnational soft law dialogue, whereas the CVM Reports, the Venice Commission and GRECO reports keep referring to each other, and finally, the infiltration of EU soft law at the level of national hard law on judicial reform.

4.1 Soft Law as a Catalyst for Litigation

An analysis of the Romanian cases at the CJEU show that CVM reports have continued to play a crucial role in catalysing the actions of private litigants and courts. Indeed, the story did not end with CCR Decision 390/2021. The latter fuelled disciplinary proceedings against those magistrates who gave prevalence to EU law. For instance, investigations for exercising office in bad faith or with gross negligence⁴⁴ were started against the judge of the Pitești Court of Appeal who applied the AFJR I judgment to declare that the existence of the SICOJ was not justified by imperative objectives connected to the administration of justice and did not present sufficient guarantees for the freedom of the judiciary.⁴⁵ The investigations were closed following the dismantling of the SICOJ.⁴⁶ It should be noted that as a result of the disciplinary proceedings, the judge in question changed his view on the supremacy of European law, ruling completely differently in other cases (Selejan-Guțan 2021a). Other disciplinary proceedings resulted in the sanctioning of a judge from another lower court, who referred a question for preliminary ruling in one of the Joined Cases C-291/19 in AFJR I (Selejan-Guțan 2021b). Moreover, as reported by a prominent member of the AFJR, the judicial association instrumental in promoting the Romanian preliminary references interviews for promotion of judges to higher courts contained direct questions on the perspectives on the application of CCR case law (Călin 2022). As a result of these pressures, several courts have accepted the Constitutional Court’s interpretation, refusing to recognise the legal effects of the CVM reports, and arguing, inter alia, that the CVM reports are not legally binding. Continuing the transnational soft law dialogue (of conflict), the European Commission has asked the Romanian Government for clarifications on CCR decision 390/2021,⁴⁷ expressing concern that the decision contravenes fundamental principles of European law and challenges the authority of the CJEU. The Commission underlined that taking disciplinary action against judges who have not complied with the CCR’s decision is likely to intimidate national courts, contravening the principle of judicial independence enshrined in Articles 19(1) TEU and 47 of the Charter of Fundamental Rights.

With some judges still submitting preliminary requests, the CJEU had to settle the issue of the binding nature of CCR decisions for Romanian courts, and the consequent pressure on judges to apply national law even though the principle of supremacy of European law could be interpreted to dictate a contrary solution. In Euro Box,⁴⁸ the CJEU revisited the effects of the CVM reports, reiterating what was decided in the AFJR I. Most significantly, it held that the obligation for national courts to comply with the decisions of the CCR, even when they found such decisions contrary to EU law, prevented national judges from ensuring the full effect of European law. It added that the existence of possible disciplinary proceedings amplified this consequence.⁴⁹ This decision was promptly rejected by the CCR in a press release,⁵⁰ which reiterated the binding character of the CCR decisions.⁵¹ Whilst such press release does not, in itself, have any binding effect either, it added at the time, alongside Decision 390/2021, pressure on national judges, menaced by disciplinary proceedings.

In a later case RS,⁵² the CJEU declared these practices against Article 19(1) TEU, read in conjunction with Article 2 and Article 4(2) and (3) TEU, with Article 267 TFEU and with the principle of the primacy of EU law. One of the issues raised by the national judge in RS was what to do following CCR decision 390/2021, inter alia, the statement that CVM reports were not binding. The questions submitted to the CJEU were directly related, with the national court asking in essence whether the CCR decision was in line with the principle of primacy, as well as the independence of the judiciary. The CJEU answered in the negative. In AFJR II, the CJEU scrutinised a change in the promotion procedure to regional courts and courts of appeal. The claimant argued that this change was against the CVM reports by the Commission, and the national judge asked the CJEU whether these reports were acts which were amenable to interpretation by the Court of Justice, and whether they were binding on Romania. This was answered affirmatively in line with AFJR I, namely that the authorities need to take these into account in order to comply with the principle of sincere cooperation.⁵³

It is important to note, at this point, that soft law is not the only ingredient of the Romanian cases. Naturally, the regulatory frameworks are rarely just ‘soft’ or ‘hard’; they are hybrid. This is reflected also in the approach of the CJEU. Faced with CCR’s refusal to recognize the EU constitutional requirements with regards to the rule of law, the CJEU response was both hierarchical, hard– by resorting to big guns of the autonomy of EU law – and fostering dialogue through soft law. First, as many rule of law judgments, the Romanian judgments make reference not only to the CVM decision but also to Treaty articles and the principle of autonomy of EU law. With regards to the Treaty, in AFJR I, the CJEU had recourse to Article 19 TEU, Article 2 TEU, as well as Article 47 of the Charter to decide on the various issues regarding Romania’s justice laws. This was despite the opinion from the Advocate General that only the CVM Decision should have been the basis of the judgment. The AG advised that the liberal use of such a general article as Article 19 TEU would give rise to abuses, and unjustified interference by the Court of Justice in areas falling squarely within the procedural autonomy of the Member States. Thus, Article 19 TEU should remain “an extraordinary tool for extraordinary cases.”⁵⁴ The European Court’s attitude is, in my opinion, correct. Romania’s obligations regarding respect for the rule of law are clear from the CVM procedure. They are not inferred by the Court through its case law directly from the text of Article 19. However, given that the CVM procedure is based on hybrid hard and soft legal tools, it is important for the Court to remind that these tools have a constitutional grounding in the Treaty. To strengthen its arguments further, the Court also refers, in the Romanian cases, to the autonomy of EU law. Tagged as a legal, de facto, dimension of the vague concept of European sovereignty (Eckes 2020, 319–329), the autonomy of EU law is a controversial concept. While Odermatt (2018, 291, 293, 313) considers the autonomy of European law to be a principle of EU law in its own right, Ziegler and Moreno-Lax (2022, 229) warn, on the contrary, that it is dangerous from the point of view of the rule of law to elevate the autonomy of European law to this status. This “autonomy” would be no more than a descriptive concept, referring to the functional independence of any international organisation—including the European Union. Whilst I agree with the views that the autonomy of EU law is a questionable principle which might need more spelling out and permeability to superior checks and balances,⁵⁵for the purposes of this paper it is fair to note that the frequent reference to the principle in the Romanian judgments on the rule of law shows clearly the hierarchical positioning of the CVM requirements.⁵⁶

Second, the CJEU also opened a route to dialogue by mentioning that the CVM reports should be taken into consideration because of the duty of sincere cooperation. Such cooperation, based on flexible norms, is crucial. As observed by Tamanaha (2017, 77–98), the rule of law depends on the institutional and societal structure that promotes respect for the law by state bodies and citizens. This includes elements such as a robust legal tradition that produces an integral community of legal professionals. Whilst, in the Romanian cases, the apex Court CCR refused to acknowledge effects of soft law, these instruments were however intensely relied on by litigants and by lower courts, showing perhaps the emergence of a strong legal community. The CVM story discussed in this article shows indeed an abundance of judicial dialogue on the rule of law based on CVM reports from lower courts and the CJEU. As expressed previously in the literature, soft law instruments can make a very good ‘starting point for a litigation strategy, for collective bargaining or for mobilizing support for future measures of hard law’ (Beveridge and Nott 1998, 309). Doroga and Bercea (2024) identify an explosion of preliminary references from the Romanian courts following the modifications of the justice laws of 2018, several of these being brought by associations of magistrates. An analysis of landmark cases such as AFJR I and II, RS, and Euro Box Promotion shows that the CVM reports feature prominently in the arguments of the parties and of the Court. What is more, subsequent CVM reports refer to preliminary references of the CJEU too, reinforcing the crucial role that litigation can have in endowing these instruments with effects. In the 2021 report for Romania, it is clearly stated that “it is important that [changes to the provisions on the civil liability of judges and prosecutors] properly reflect the decision of the EU Court of Justice [in AFJR I] and take due account of relevant European standards.” This is an established technique whereby the European Commission legitimises its own soft law by referring back to the case law of the Court validating its views (Snyder 1994).

At the national level, in the follow-up case to Eurobox, the Romanian High Court of Cassation and Justice disapplied a Constitutional Court decision as contrary to European Union law, in accordance to the advice of the EUCJ. The High Court did not admit direct binding force for MCV reports; quite the contrary, it noted that the rule of law objectives included in the MCV process are binding. As a consequence, ‘Romania has to take measures to comply with these objectives, whilst taking into account the recommendations made in the reports drawn by the Commission.’⁵⁷ Coming from an Apex Court, this is an important endorsement of the hybrid EU regulatory framework. It shows that, when one ‘guardrail’ (Levitsky and Ziblatt 2018, 97 et s) of the rule of law (the CCR) fails, another one is ready to take its place.

The refusal of the CCR to engage with soft law, as well as the compliance, from some lower courts, with CCR decision 390/2021 show that one should not put their entire faith in courts when rule of law matters are at stake. CVM reports have managed to catalyse public opinion within and beyond the judiciary. For instance, in the context of the recent dismantling of the SICOJ, some magistrates opposed certain procedures as suggested in the draft law to make the prosecution of magistrates.⁵⁸ conditional on the approval of the Superior Council of Magistracy, and urged the Parliament to await the opinion of the Venice Commission and GRECO before deciding.⁵⁹

4.2 Transnational Soft Law Dialogue and Cross-Hybridisation

Tracking further the dialogue based on soft law, one notes that the reports of the Commission on the MCV figure also in other transnational soft law instruments, notably in Venice Commission opinions. Most of these opinions include, in the introductory ‘background’ section, a selection of the recommendations made in various CVM reports. In its opinion on the 2018 amendments to the judiciary laws, the Venice Commission assessed the new rules on the appointment of chief prosecutors against the checks and balances provided in a recommendation made in the 2017 CVM report. It concluded that they evidenced a regression in standards.⁶⁰ Yet, in the opinion on the new draft laws from 2022, the Venice Commission shied away from any discussion regarding the compatibility of the new laws with ‘European standards’. It also expressly refused to formulate any views regarding the opportunity of closing the CVM,⁶¹ perhaps because that instrument formally pertained to a distinct organisation.

The dialogue continues in the CVM reports themselves. Already in 2017, the CVM report was referring, explicitly in its recommendations, to the Venice Commission and GRECO.⁶² The 2018 Report noted that the authorities did not follow up on these recommendations. It referred to the opinions of the Venice Commission on Romanian legislative changes and the reversal of good progress made. It again explicitly recommended that Romania should follow the opinions of the Venice Commission and GRECO on criminal and justice laws.⁶³ The 2019 Report found that, in order to foster positive legislative change, ‘the involvement of the Venice Commission remains the best way to find a balance between the involvement of the government and the judicial authorities which can command public confidence.’⁶⁴ The last CVM report, from 2022, referred to the Venice Commission opinion on the new justice laws, and also to the opinion on the dismantling of the SICOJ. It noted in particular that transparency and inclusiveness in drafting the new changes was not always up to standard, as required also by the Venice Commission. The EU Commission however noted that the Government took account of the opinions of the Venice Commission in relation to Benchmark 1 independence of the judiciary and judicial reform.⁶⁵ In the 2023 Report on the Rule of Law for Romania, the Commission mentioned the efforts of the Romanian Government in addressing the concerns of the Venice Commission and the appointment of a high-level expert panel to deal with the questions raised. It recommended Romania to follow up on the Venice Commission opinion and the conclusions of the experts. After extensively citing the Venice Commission opinion, the Commission concluded that ‘significant progress has been made to address the recommendation made in the 2022 Rule of Law Report on the Justice Laws.’⁶⁶ Venice Commission opinions creep in EU hard law as well, with the Commission decision on closing the CVM referring directly, in the preamble to Romania’s commitment to respect these instruments.⁶⁷

These close links between the various transnational soft law instruments are very important. Venice Commission opinions, CVM Reports, and GRECO opinions validate each other, mutually offering extra layers of legitimation. In turn, this can only increase their effectiveness, showing that soft law from multiple sources can act together for the achievement of goals. At the same time, such instruments can evolve following their interaction, as well as the interactions with the stakeholders, and improve over time. This would lead to a robust inter-institutional dialogue also in soft-law generation.

4.3 The End Result: Soft Law Making its Way into National Legislation

National legislation amending the Justice Laws in 2022 refers to the Venice Commission, GRECO, and EU Commission soft law as a basis (or ‘source’) for reform, as per their respective statements of reasons. For instance, the statement of reasons for the Law on the Statute of Judges and Prosecutors lists a whole host of transnational soft law which are meant to be ‘implemented in national law’⁶⁸ through that act. This is extraordinary, given that the term ‘implementation’ is generally associated with hard law. One may even go as far as wondering whether this legislative (and hence democratic) endorsement of soft law does not qualify and perhaps relativise the CCR decision 390/2021, discussed above in Sect. 3. More significantly, the question is whether such clear confirmation of the national implementation of EU soft law does not also automatically trigger the application of the Charter of Fundamental Rights. An analogy might be made with the ANI judgment, where the CJEU confirmed that the Charter applied to situations arising under the Romanian Law on the Integrity of the Civil Service because this law was issued as part of the obligations under the CVM decision.⁶⁹ Since the Statement of Reasons for the Law on the Statute of Judges and Prosecutors refers to this act as implementing Commission recommendations, an argument could well be developed that the Charter should also apply to any situations arising under this Law. This would open the possibility to challenge, under EU law, a vast host of scenarios exceeding those directly related to Article 19 TEU.

The statement of reasons for the Law on the Organisation of the Judiciary and the Law on the Superior Council of Magistracy contain more discrete references to these soft law instruments such as in relation with the revoking powers of the head prosecutors,⁷⁰ or to the Judicial Inspectorate.⁷¹ Quite naturally perhaps, the statement of reasons for the Law on the Dismantling of the SICOJ draws at length on CVM, Venice Commission, and GRECO soft law, yet, interestingly, mentions these sources only as secondary arguments to the national consultations and case law.⁷²

As a result of these amendments, the Commission decided to close the Cooperation and Verification Mechanism in September 2023. This is not without controversy: a professional association of Romanian prosecutors, “Initiative for Justice”, sought the annulment of the Decision closing the CVM.⁷³ They argue that the Commission erred in law when concluding that Romania has met its benchmarks under the mechanism, that the Commission did not provide enough reasons to support its positive conclusion on matters regarding the independence of the judiciary as well as the application of the principle of primacy in Romanian law; and finally, that the Commission infringed essential procedural requirements by failing to involve the Council and the Parliament in this decision making process. Regardless of its merits, which will be assessed in due course by the Luxembourg Courts, this claim is interesting for the purposes of this paper. It shows that, at least in the eyes of some stakeholders, the hybrid CVM did not appear completely ineffective and seemed a procedure which was worth investing in a court case for.

One may wonder what happened to another protagonist, the Constitutional Court of Romania. A wind of change seems indeed to have blown at its doors too. In November 2022, the CCR rejected several complaints against the Justice laws.⁷⁴ These complaints targeted, amongst others, the removal of the disciplinary offences which were found by the ECJ to fall foul of judicial independence and primacy of EU law. Whilst the judgments do not really discuss in detail the various soft law referred to in this article, they might hint at an opening to the dialogue with European partners.⁷⁵ As with anything related to soft law, the story never really ends.

5 Conclusion

The Romanian case study discussed above does not, by any means, illustrate the full effectiveness of soft law to address rule of law issues. It does, however, show that these instruments are not static: they are constantly reviewed, changed, referred to in other soft law issued by other transnational bodies. They can, therefore, be influential as time goes by, through complex dialogues (sometimes conflictual) between courts of various levels, whilst catalysing powerful stakeholders. It is through soft law that international organisations can validate each other’s work, such as the confirmation of EU Commission’s views by the Venice Commission or GRECO, and vice versa. Through its flexibility soft law allows the various transnational authorities to react quickly and signal violations. As many studies argue, a hybrid legal framework, combining mandatory and non-mandatory rules, is important for accommodating complexity and the necessity for swift action, or, indeed, diversity of views. The same is valid also for EU frameworks regulating values, such as the rule of law. The Romanian case study has shown that, even if no immediate consequences seem to stem from soft law, with time, change can occur. As pointed out by Cameron (2021, 179) in relation to Venice Commission reports, when it comes to changing mindsets and preserving values, patience is key. Dialogue takes time and transnational institutions are obliged to ‘keep talking’ even if that might seem a futile enterprise in the short run. As any dialogue, this is not a one-way street: both national and European institutions should be open to change and adaptation, which in turn will lead to a better and stronger framing of values through soft and hard law.