Exit, voice and consensus – A legal and political analysis of the emergency brake in EU criminal policy
The evolution of EU criminal justice policy in the last 30 years has been remarkable. From being on the periphery of EU integration at the beginning of the 1990s, Treaty amendments in Maastricht, prolific legislative activity and a strong political commitment by the European Council in Tampere, Hague and Stockholm have helped push criminal justice to the centre stage of EU policy-making. This evolution culminated with the ratification of the Lisbon Treaty where the former pillar system was abolished and the making of EU criminal law was integrated into the traditional ‘Community’ decision-making structure. It is, therefore, widely held that the Lisbon Treaty has transformed EU criminal law from an ‘intergovernmental’ to an essentially ‘supranationalised’ policy field (Peers, 2011).
The most powerful objection to the ‘communitarisation’ of decision-making is the introduction of an emergency brake in Articles 82(3) and 83(3) TFEU. Those provisions read in identical wording as follows: ‘Where a member of the Council considers that a draft directive … would affect fundamental aspects of its criminal justice system, it may request that the draft directive be referred to the European Council. In that case, the ordinary legislative procedure shall be suspended …’ (emphasis added).
The implications of the emergency brake are potentially significant. If Member States threaten to employ the emergency brake, EU action in this area may be seriously compromised (House of Lords, 2008). Notwithstanding the significance of this provision for the direction of EU criminal policy, it has not been subject to comprehensive scholarly scrutiny (Rosin and Kärner, 2018). This blog post intends to make such a contribution by offering a critical analysis of the emergency brake. The first part explores the legal dimension of the emergency brake with a particular focus on how the concept of ‘fundamental aspects of a criminal justice system’ should be construed and enforced. The second part offers a political examination of the emergency brake including a comprehensive qualitative analysis of EU legislation adopted so far in the area of criminal policy.
The legal dimension of the emergency brake
At the outset, it is important to reflect on the reasons for the introduction of the emergency brake into the Lisbon Treaty. Working Group X, which was responsible for the drafting of the relevant provisions in the Constitutional Treaty, argued strongly in favour of extending qualified majority voting (QMV) to the area of criminal law (Final Report, Working Group X). Nonetheless, a number of Member State delegations, which were not involved in the drafting of the Constitutional Treaty, considered that the general move to QMV in criminal law was premature and subsequently proposed the insertion of an emergency brake in the Treaties (Intergovernmental Conference, Naples, 2003). The emergency brake could thus be conceived of as a ‘voice’ mechanism in the absence of a proper ‘exit’ option (Hirschmann, 1970). Since ‘unanimity’ as an option for Member States and other venues for ‘selective exit’ (Weiler, 1991) was ‘closed’ after Lisbon, the introduction of the emergency brake must be viewed in the light of the Member States’ appeal for an ‘enhanced voice’ in important outcomes to which they would have to adhere.
Next, the concept of ‘fundamental aspects of a criminal justice system’ is considered. It is argued that one distinguishing feature of a rule or principle belonging to ‘fundamental aspects’ is, that it must be ascribed a certain value within the legal order of a specific Member State. This suggests that criminal rules of constitutional value stand in a privileged position when being invoked by Member States under the emergency brake procedure (Wittgenstein).
The contested Melloni judgment from the Court of Justice – concerned with national in absentia rules and a surrender procedure under the European Arrest Warrant (EAW Framework Decision) – exemplifies a national criminal law principle which, due to its importance in the national legal system, would qualify as a ‘fundamental aspect’. In Melloni, the Court of Justice rejected the possibility of conferring powers on the executing judicial authority to apply the national constitutional standard of protection of fundamental rights, as this would compromise the primacy and effectiveness of EU law (Melloni, paras 55-63). Notwithstanding the Court’s conclusions, it is claimed that the contested Spanish rules on in absentia trials constituted a ‘fundamental aspect’ of the Spanish criminal justice system. The fact that the provision at issue was derived from the Spanish constitution and the fact that no exceptions from the right to challenge in absentia trials was allowed pursuant to the Spanish Constitutional Court’s case law supports this conclusion (Bachmaier Winter, 2016).
The other central feature of ‘fundamental aspects’ is to analyse whether EU law adversely affect culturally and historically defined ways of thinking about a criminal justice system (‘legal tradition’). Rules that constitute a central feature of a specific Member State’s legal tradition are particularly sensitive to external legal influence (Summers, 2007). The right of access to a lawyer is an enlightening example of a principle that – because of the legal traditions in the Member States – in many instances satisfies the requirements of being a ‘fundamental aspect’. The negotiations on the EU Access to Lawyer Directive highlighted the very contested nature of access to a lawyer in the pre-trial phase, where Member States with a more ‘inquisitorial legal traditions strongly contested the promotion of active defence to criminal charges’ (Jackson, 2016).
The argument so far suggests that a feature must have firm support in primary legislation, case law or the preparatory works to legislation and constitute a key organising principle for the criminal justice system to reach the threshold of ‘fundamental aspects’. The principle of proportionality in sentencing (von Hirsch and Ashworth, 2010) and free assessment of evidence (Damaska, 1995) are cases in point of such principles. It was also argued that an alternative threshold for satisfying the ‘fundamental aspects’ criterion is that the principles at issue have adverse implications for a specific national legal tradition.
Next, we turn to judicial review of the emergency brake. In order to review the Member States’ use of the emergency brake the Commission or a Member State would have to commence an infringement proceeding against the Member State pulling the brake (Articles 258 and 259 TFEU). The basis for the action would be that the Member State had breached its obligations under the Treaties by illegally pulling the brake, acting in contravention of Article 83(3) TFEU and the principle of sincere cooperation (Article 4(3) TEU). The subjective language of the Treaty, ‘where a member of the Council considers’, nonetheless suggests that it is intrinsically difficult to challenge the Member States’ use of the emergency brake. Drawing on the debate on the identity clause in Article 4(2) TEU, it is proposed that the Court of Justice’s authority only extends to determining the framework of national criminal law rules that may constitute ‘fundamental aspects’ (Besselink, 2010). If the Court were to assess the content of such ‘fundamental aspects’, it would overstep its jurisdictional mandate which is confined to the interpretation of EU law pursuant to Article 19(1) TEU. By extrapolating from the case law on misuse of powers, it is argued that the concrete test for judicial intervention must be founded on a ‘manifestly’ illegitimate use of the emergency brake procedure. This suggests that the Court can only intervene when it has been substantiated that a Member State has illegitimately employed the emergency brake mechanism for predominantly political reasons (Windpark Groothusen; Spain and Italy v Council).
In sum, this section suggests that the emergency brake – whilst not strictly speaking a veto right (which can be used without giving reasons and cannot be contested) – is a powerful legal tool for Member States wishing to safeguard the integrity of their criminal justice systems.
The political dimension of the emergency brake
Articles 82 and 83 TFEU state that decision-making in the Council is to be accomplished by means of qualified majority voting, unless the emergency brake is triggered. It is thus appropriate to consider decision-making in the Council more broadly under the qualified majority rule. A key premise for the discussion is that Council decision-making in most areas of EU policies (including criminal law) is governed by consensus politics. The high proportion of legislative decisions taken by consensus is a remarkable feature of Council decision-making, also in fields where the majority rule applies – and previous findings suggest that only 20–25 per cent of all legislative acts is contested (Van Aken, 2012).
An important account for explaining consensus culture in the Council is based on formal voting rules as ‘drivers’ of decision-making. The implicit threat of voting helps to identify emerging majority views and serves as a deterrent against any opportunistic behaviour of Council members (Novak, 2013; Beaudonnet and Deloche-Gaudez, 2010). It is important here to distinguish between consensus politics in the area of qualified majority and a genuine ‘voice’ in form of a ‘veto’ or an informal agreement, such as the Luxembourg accord amounting to a veto. The Luxembourg accord, which resolved the ‘empty chair’ crisis caused by France in 1965, conferred on every state a de facto right to veto proposed Community legislation when ‘very important interests were at stake’ (Wallace, Winand, Palayret, 2006). In terms of the emergency brake, the conventional understanding among observers, at least pre-Lisbon, was that the very existence of this provision would have a ‘chilling’ effect on majority decision-making and thus work as an ‘informal’ veto compromise (House of Commons Report, 2005, paras 110-120).
A comprehensive review of all EU criminal law legislation adopted since Lisbon under Articles 82(2) and 83 TFEU suggests, unsurprisingly, that consensus is the general style of decision-making in this field. There are only five instances where Member States have voted against or abstained from voting for the adoption of a directive (the PIF Directive, the Access to Lawyer Directive, the Money Laundering Directive, Directive on Legal Aid and the Directive on Attack against Information Systems); the remaining Directives have been adopted by consensus. From the review it is also apparent that the emergency brake has not, to date, been used. More importantly, a detailed review of the legislation gives some tentative evidence to support a gradual transformation to ‘supranational’ decision-making in this area. We consider the more controversial pieces of legislation in some detail.
With respect to the Legal Aid Directive, it seems that serious objections made by some Member States were sidelined in the negotiations. Whereas the original comprehensive remit of the directive proposed by the Commission was tempered, the core principles of the right to legal aid was kept intact in the final version (Directive on Legal Aid, Articles 2(1)(C), 2(4) and 4). Poland issued a specific ‘emergency-brake’ type of statement, referring to the Member States’ divergent views on the balance between rights for suspects and the states’ need to effectively combat crime criticising the far-reaching scope of the Directive (Council Doc 12835/16). However, Poland refrained from using the brake and abstained from the formal voting procedure.
In respect of the Presumption of Innocence Directive, in absentia trials were the most contested issue in the negotiations (Council Doc 12955/14, Council Doc 13304/14, Council Doc 11112/15). In this specific instance, it should be conceded that the ‘threat’ of a stalemate in negotiations may have influenced the less ambitious design of the in absentia provision in the final version (Presumption of Innocence Directive, Article 8) compared to the Commission’s proposal (Proposal for a Presumption of Innocence Directive, Article 8). Since the Commission’s design of the provision was not significantly altered in the final version, it nonetheless seems that the potential threat of the ‘emergency brake’ did not have a direct influence on the final outcome.
Another notable example in the analysis is the PIF Directive, which was subject to a majority vote in the Council. The contestation concerned the sensitive issue of VAT where Member States claimed that the inclusion of VAT within the scope of the Directive would have adverse effects on their criminal justice systems (Öberg, 2021). In this instance, the final compromise contained substantial concessions to the dissenting Member States, as the scope of the Directive was confined to serious cases of cross-border VAT fraud (PIF Directive, Article 2(2)). However, the negotiations also offer evidence of a tendency of Member States to turn to voting in this area if there is substantive disagreement, and to abstain if they object to legislation.
The negotiation history of the Access to Lawyer Directive is the strongest testimony that the emergency brake is not employed as an ‘exit’ option to alleviate QMV in the area of criminal law. In this instance, many Member States fiercely objected to the broad scope of the right of access to a lawyer envisioned by the Commission’s proposal (Council Doc 14495/11). Belgium expressly claimed that the proposed directive infringed on certain central principles of its criminal justice system – i.e. the quasi-inquisitorial character of the preliminary investigation procedure – thus satisfying the criteria for pulling the emergency brake (Council Doc 7593/12). The Member States’ objections were partly heeded in the sense that the Commission’s far-reaching proposal was watered down in subsequent negotiations. However, the key principle of the right of access to a lawyer was consecrated throughout the negotiations, suggesting that Belgium’s principled objections to the Directive were not accommodated (Öberg, 2021; Access to Lawyer Directive, Article 3(3)). Despite all this, Belgium refrained from pulling the brake and abstained at the final vote.
There are several implications to be drawn from this analysis. First, the review offers some evidence supporting the idea that the political cost of pulling the emergency brake is greater than relying on the veto right within the context of unanimous decision-making. Member States who oppose a proposal seem to prefer to participate in a blocking minority rather than to take singlehandedly the risk of pulling the emergency brake. Second, there is also some tentative evidence of political ‘peer pressure’ within the Council (Scharpf, 2006) to agree on certain measures in the field of criminal law. The example of mutual recognition is illustrative. Prior to Lisbon, the adverse effects of mutual recognition instruments became visible, and flagship measures such as the European Arrest Warrant attracted strong criticism based on human rights concerns (Mitsilegas, 2006). In light of this, the European Council called for comprehensive procedural rights approximation to mitigate the differences between Member States’ standards (Trauner and Ripoll Servent, 2016). Against this backdrop, it is easy to appreciate why Member States perceive it difficult to obstruct negotiations on a planned procedural rights measure by using the emergency brake. The diligence with which some Member States argue about legal bases (determining whether a measure is adopted by qualified majority or unanimity, Council Doc 7929/17) offers further evidence to support that the political cost of using the emergency brake is significant. The context for its use seems to be exceptional circumstances, where undesirable EU measures that are of fundamental importance to a Member State are proposed, in a situation where the Member State knows that the other Member States can go forward without it via enhanced cooperation (Articles 82(3) and 83(3) TFEU).
The analysis here may be appositely juxtaposed with the ‘exit’ and ‘voice’ argument made by Weiler with reference to the introduction of qualified majority voting on internal market legislation (Weiler, 1991). As observed by Weiler within the context of Article 114 TFEU, it is suggested that the qualified majority rule in Articles 82 and 83 TFEU has meant that decision-making in the field of criminal law today is conducted under the ‘shadow of the vote’. Reaching consensus under the ‘shadow of the vote’ is, however, very different from achieving it under the ‘shadow of the veto’. The possibility of resorting to majority voting under Articles 82 and 83 TFEU drives negotiators to break deadlocks without actually formally triggering the voting mechanism (House of Commons, 2005, paras 113-114). All this suggests that the new decision rules on QMV in Articles 82 and 83 TFEU have effectively – despite the existence of the emergency brake – transformed the character of decision-making in the field of criminal law from ‘intergovernmental’ to ‘supranational’.
The analysis here leads to a general reflection on the role of the emergency brake within the context of EU criminal justice as a perennial battleground between intergovernmentalism (Moravcsik, 1998) and supranationalism (Sandholz and Stone Sweet, 2004). Scholars of (new) intergovernmentalism (Bickerton, Hodson and Puetter, 2015) have notably challenged the supranational paradigm (‘integration through law’, Cappeletti, Seccombe and Weiler, 1986, Volume 1), delegation of powers to supranational institutions and the ‘Community’ method in the Area of Freedom, Security and Justice (AFSJ). They have pointed to the specific institutional and decision-making features, such as the inclusion of the emergency brake as evidence of retained Member State ‘management’ over AFSJ policies and a strong legacy of ‘intergovernmental’ bias in the integration of Justice and Home Affairs. The latter assumes that consensus politics has become an end in itself in this area since Maastricht, which would suggest a significant undermining of the ‘Community’ method (Wolff, 2015).
Whilst the analysis here acknowledges that the gradual ‘communitarisation’ of EU criminal policy is a delicate compromise along the supranational-intergovernmental axis, it suggests that EU criminal policy post-Lisbon clearly tilts toward ‘supranational’ decision-making dynamics. The analysis here gives further evidence of an emerging development from a rationale of ‘cooperation’ to one of ‘integration’ of national criminal justice systems (Monar, 2013; Öberg, 2021). It is conjectured in this respect that qualified majority voting in the Council is a matter of importance for the general evolution of the area of EU criminal justice. ‘Pooled sovereignty’ in the Council entails not only a devolution of powers from the sovereign state to an interstate body like the EU but also a process whereby the members of the Council begin to create ‘new definitions of the self’ (Gouretvich, 1999). Each Member State would jointly constitute the Council as an entity separate from national governments, with a distinct (European) loyalty and which exercises some measure of genuine autonomous power (Wessels, 1991).
Whereas the veto power – as the ultimate power of state sovereignty – is diluted in criminal matters, there appears to be a stronger direction of supranationalism in the area. Whilst the Member States may still formally be the ‘captain of the ship’ (Harding and Öberg, 2021), the proliferation of EU criminal policy initiatives post-Lisbon under the majority rule suggests a strong ‘integrative’ force (Lindberg, 1963; Haas, 1958), which is bound to reshape our contemporary perspective of EU criminal justice as a Member State-dominated policy field.
This blogpost draws on a longer article by the author: J Öberg’ Exit, Voice and Consensus- A Legal and Political Analysis of the Emergency Brake in EU Criminal Policy’ 46 (4) (2021) European Law Review, 506-530. I would wish to acknowledge the very useful feedback from the editors of the European Law Blog on an earlier version of this blog post.