Abstract
In the three decades since it was established in the context of a secretive and technocratic intergovernmental organization, Europol has evolved into a European Union agency with some, albeit limited, supranational capacities. This article overviews the gradual legal development of Europol’s powers and discusses the obstacles to creating a European federal police force. Limits to powers and accountability continue to frame discussions on EU’s operational criminal justice powers. While many EU agencies can lay claim to embryonic supranational enforcement agency, the EU Member States have closely guarded operational and prosecutorial enforcement powers. This guardedness still shows, especially in EU criminal justice agencies’ reliance on intergovernmental structures, such as colleges of national members, and mutually recognized but ultimately national decisions and judgments. Through the lens of its history, this article examines how and what kind of balance has been struck between accountability and competences in the current state of evolution toward Europol’s potential supranational authority.
Introduction
Europol has, in the last three decades, developed from a technocratic intergovernmental organization into a de novo body with autonomous capacities. The legal framework within which it operates – that which frames both its powers and accountability – has evolved so as to demonstrate fledgling supranational features. Still, states remain resistant to granting Europol full autonomy, leaving the agency legally constrained in varying areas of law enforcement.
In what follows, this article overviews the legal and political turning points in Europol’s development. Focusing on the shifting divisions of Member State and Union competences in law enforcement, and the changing forms of accountability which have accompanied them, the article showcases the ways in which the Union legal framework has been worked so as to both enable and restrain Europol’s agency. The most recent case in point is the process of reform that followed the Lisbon Treaty. Focusing on the post-Lisbon changes to Europol’s legal framework, the last section of the article discusses the tensions that are at play – those tensions being much the same ones as the ones arising throughout Europol’s history.
On the development of Europol from TREVI to Amsterdam
Europol transformed out of the TREVI group, a network of national officials known for its secrecy and, as experts in cross-border policing often describe it, its opacity. 1 Especially in its early years, Europol’s legislative framework shared some of these qualities. 2 Despite misgivings about Europol’s transparency and legitimacy, by the time Europol came into being it was considered, if not a remarkable improvement, at least a notable change in the mode of legally establishing international police cooperation. 3 In contrast to, for instance, Interpol and TREVI, Europol benefited from a clearer legal mandate. Perhaps for this reason, it represented a more coherent approach to law enforcement cooperation. 4 Europol’s opaque background, in turn, prompted ‘concerns for democracy’ that would increasingly come to be reflected in its legal framework. Occhipinti, for one, has argued that the concerns for democracy arising from the TREVI group’s secrecy pushed Europol toward supranationalism, since, to make sure that Europol achieved a higher standard than TREVI, legal safeguards were necessary to ensure increased transparency and accountability. 5
The 1993 Maastricht Treaty established the grounds for police cooperation among Member States to combat terrorism, drug trafficking and other organized crime. 6 The Treaty, however, also set significant limits to the Union’s competences in these areas. Police cooperation was, to the dismay of many scholars, 7 relegated to the Third Pillar introduced by the Maastricht Treaty. The impact of the pillar structure in the case of Europol was significant, especially because the Third Pillar was insulated from judicial scrutiny by the Court of Justice. Even though commentators were adamant that the existing European human rights framework was established and harmonized enough to set tangible limits to police cooperation within the Union, 8 in practice, there was to be no judicial control within the Union over the Third Pillar. The Court had no ability to review the legality, or the validity, proportionality or human rights concomitance of policing actions and measures in this area.
While the Treaty marked the formal integration of police cooperation into the pillar structure of the European Community, 9 Europol itself was, nevertheless, established through a convention, placing it outside of the Community legal framework. In other words, Europol was to be an intergovernmental organization, and its legal basis, as such, was much slower to gain validity than its future EU agency counterparts were. 10 The 1995 Convention required ratification by all Member States to come into force. Plausibly because of the lack of consultations with the European Parliament or National Parliaments in the drafting of the Convention, Loader observed that there was now a powerful institutional motor driving the formation of both stronger ties between Europe’s police forces, and an opaque and thinly accountable transnational police elite oriented toward forging common ‘solutions’ to common ‘security’ problems. 11
It ultimately took until 1999 for Europol to commence its activities. 12 Moreover, as Den Boer and Walker contend, policing came to ‘occup[y] a peculiarly hybrid position within the Community legal system, neither merely intergovernmental nor properly supranational…in the core law-enforcement areas of judicial co-operation in criminal matters, customs co-operation, and police co-operation itself, the Commission, as the permanent EC executive, still has no right to initiate policy’. 13 Instead, Europol fell into the mandate of the Council of the European Union. By the time the Convention finally came into force, Member States of the Council were keen to amend it to progressively broaden Europol’s powers. 14
The Council proposed multiple amendments and protocols to the Europol Convention in the early 2000s. Drafting these updates to the Conventions remained in the sole power of ministers, though it was often inspired by professional elites. 15 Although National Parliaments were still sidelined from the drafting processes, they retained their right to ratify or to refrain from ratifying. The unanimity requirement here meant that while Europol/EU competences could not be enlarged without unanimity, neither could they be reduced without all Member States agreeing. 16
Ultimately, Member States acted unanimously to add to Europol’s competences. Through the new protocols and amendments, Europol’s competences extended to participation in Joint Investigation Teams and the right to ask Member States to provide information on ongoing criminal investigations. 17 In the name of the fight against terrorism, Europol was granted access to the Schengen Information System (SIS) and the Visa Information System (VIS) database. To complement this ‘[slow] transformation into a more operational body’ 18 Europol’s accountability, too, increased. Europol officials’ immunity, when supporting JITs, was waived. 19 Instead, they were to be accountable under the ‘national law of the Member State of operation applicable to persons with comparable functions.’ 20
Amsterdam, fragmentation across pillars
The amendments to the Europol Convention all took place under the framework of the Treaty of Amsterdam (1997), which had come into force before the Convention had gained full ratification. With it came renewed aspirations regarding Police and Judicial Cooperation, which was to be achieved through ‘closer cooperation between police forces, customs authorities and other competent authorities in the Member States, both directly and through the European Police Office (Europol) in accordance with the provisions of Articles 30 and 32’. 21 Additionally, the Amsterdam Treaty gave Europol a privileged position in information exchange between national police forces. 22
Along with the renewed commitment to law enforcement under the Amsterdam framework also came reiterated concerns for Europol’s accountability. Under the EU Treaty’s framework, police co-operation remained in the Third Pillar, whereas issues such as visa, asylum and migration were transferred to the First Pillar. To some scholars, these were clear indicia of the Union’s intention to keep Europol ‘at arm’s length from any supranational control by the EU Parliament and the European Court of Justice’. 23
Creating more concerns, even, than questions of accountability, was the growing complexity of the Union’s legislative framework, which brought questions of fragmentation to the fore. 24 Among these complexities were the increased possibilities for opt-ins and other flexible arrangements, making it challenging to ensure consistent, Union-wide parliamentary and judicial control. For instance, Member States could now opt to accept the Court of Justice’s jurisdiction to give preliminary rulings on decisions adopted by the Council regarding police cooperation. 25 Moreover, scholars increasingly criticized the diminished coherency and transparency of the Union’s criminal law framework as a whole, demonstrated, in particular, by the distribution of Union competences regarding fraud and drugs across two of the three pillars. To this end, some scholars suggested that abolition of the Pillar Structure would diminish the ambiguity present during the era. 26 Most of all, however, there was a growing sense that, to ensure the legitimacy and functioning of Europol, the diversity of domestic policing strategies and criminal laws had to be overcome, likely by further harmonization of criminal laws and coherency among institutions at the Union level. 27
In preparation for the EU Constitutional Treaty, the Commission proposed to replace Europol’s convention with a Council Decision. The Council Decision would allow more flexibility in amending Europol’s legal basis, 28 as well as grant it full community status. It would, in other words, turn Europol into an EU agency. Even though the Constitutional Treaty failed, the Council pushed the decision through. The Lisbon Treaty was, at this point, already in sight, and though the European Parliament objected to the Decision, it had not quite yet gained the power to influence the outcome. In other words, The Council’s timing was crucial for avoiding parliamentary oversight. 29
The Council Decision, when transforming Europol from an intergovernmental organization into a Union agency, had implications for Europol’s powers and oversight. Among these implications, among the most important was the European Parliament’s enhanced role in the control of Europol’s budget, which was now funded by the Community. 30 In addition to the indirect control the EP could exercise through its budgetary powers, pursuant to the Council Decision, the Council President and Europol’s Director and Chair of its Management Board were also obliged to appear before the EP at its request. The EP’s increased role, Occhipinti contends, ‘represent[ed] a noteworthy development toward supranationalism in its own right’. 31
This development toward supranationalism was bolstered further by the shift brought about by the Council Decision in regard to future amendments. Whereas the Europol Convention demanded unanimity, with the Council Decision a qualified majority would suffice. Such procedural changes have historically enabled quicker compromises to be found and sealed at the Union level, thus signifying a further step toward integration. 32
Lisbon Treaty – developing and consolidating Europol’s powers
The Lisbon Treaty had a ‘supranationalizing’ impact on the scrutiny and powers of Europol by way of at least three influential changes. First, by ending the pillar structure, the Lisbon Treaty brought policing into the remit of judicial scrutiny. In particular, the Area of Freedom, Security and Justice has become subject to the ordinary jurisdiction of the Court of Justice. 33 Secondly, by extending the ordinary legislative procedure to policing, the Lisbon Treaty empowered the supranational Commission to initiate amendments to the Europol decision, and, essentially, to propose modifications to Europol’s mandate and powers, both of which already had been reinforced in the Treaty. Clear Treaty limits apply to prevent, without Treaty revision, the development of certain direct operational powers for Europol: Article 88 TFEU stipulates that the ‘application of coercive measures shall be the exclusive responsibility of the competent national authorities.’. Similarly, Article 72 TFEU in the general provisions applicable to the AFSJ states that ‘This Title shall not affect the exercise of the responsibilities incumbent upon Member States with regard to the maintenance of law and order and the safeguarding of internal security.’ The introduction of the ordinary legislative procedure also altered the European Parliaments role. Although the EP had not been able to influence the design of the Council Decision on Europol, through the Lisbon Treaty it gained co-legislative rights to adopt, with the Council, regulations that ‘determine Europol’s structure, operation, field of action and tasks’. These regulations were also to comprehend ‘the procedures for scrutiny of Europol’s activities by the European Parliament, together with national Parliaments.’ 34 The Parliament’s oversight was to be further increased by its new privileged access to Europol Classified Information whence the co-legislators initiated new legislation. Abazi has, however, argued that, despite these changes, Europol’s institutional design would prevent the EP from actually exercising any notably different role in regard to classified information. 35 Third of all, the Charter of Fundamental Rights fortified fundamental rights’ protections, which, under the Lisbon framework, can be enforced through the CJEU’s broadened mandate, thus promising increased democratic accountability – at least superficially. 36
The Commission’s 2013 proposal to recast Europol under the new post-Lisbon legal basis 37 centred on governance and accountability. As one might expect, many of the themes which surface in contemporary data service regulation also appear as issues to be considered when reforming Europol. Key stated aims 38 were to align Europol with the new Treaty requirements not only in the form of a regulation but by introducing a control mechanism by the European Parliament and national parliaments and therefore enhancing democratic legitimacy and accountability. Transparency was to be improved by providing a robust data protection regime for Europol, in particular one which guaranteed full independence, effectiveness and sufficient powers of intervention for the Europol data protection supervisor.
Governance of Europol would also be improved by seeking increased efficiency and aligning Europol with the principles laid down in the Common Approach on EU decentralized agencies. 39 The Stockholm Programme goals envisioned developing Europol as a ‘a hub for information exchange between the law enforcement authorities of the Member States’ 40 including training and exchange schemes for Member State law enforcement personnel. Europol was expected to be granted new responsibilities following an envisaged merger with the European Union Agency for Law Enforcement Training. In addition to providing law enforcement training Europol would develop EU centres of specialized expertise for combating certain types of crime falling under Europol’s objectives.
According to the Impact Assessment published with the draft proposal, several shortcomings prevented Europol from becoming the hub for information exchange between Member State law enforcement officers as envisaged in the Stockholm Programme. First, Member States were not considered to provide Europol with all of the information necessary to fight serious cross-border crime. The Europol Decision required Member States to supply Europol with data within its mandate. 41 However, consultations revealed that the data was not always timely, and in many cases necessary information was not provided at all. 42 Member States preferred bilateral sharing through the Secure Information Exchange Network Application (SIENA) without copying this to Europol. 43 This was also reflected in a set of 2012 Council conclusions. 44 Part of the problem was legal: Articles 4 and 8 of the Europol Decision were somewhat unclear on the existence and extent of Member State obligations to provide data. 45 According to the Commission, the problem was also political: ‘[t]he reason for non-recognition of an obligation is mainly politics and perception of sovereignty’. 46 Additionally, there was a low level of awareness and knowledge about Europol attributed to insufficient training and a cautious police culture towards information sharing. Multilateral Europol information collection was seen as a victim of the success of bilateral information exchange schemes. 47 Heterogeneity of organizational drivers as to their role and status in national systems was also seen as an issue, because ‘attitudes and good will of individual law enforcement officials’ was important in developing successful information exchange. 48
A further issue was the data architecture, which made it difficult for Europol to combine relevant data for analysis purposes. 49 Data provided to Europol populated a number of different databases, originally to ensure that data protection purpose limitations would be respected. This created ‘information silos’ and prevented combinations which would be useful for legitimate analytical purposes. When analysts sought to circumvent this problem by asking data providers for permission to include the data detected in another database in the database to which the analyst had access, this took weeks or months. Cross-matching between databases was possible only where there were two identical sets of data in both databases. As the impact assessment explains, if in one database persons A and B are linked to X and in another A and B are linked to Y, it is possible that X and Y know each other or are the same person. This could not be detected by the analyst as she could only cross-check A and B. 50 The technical separation of the data also meant that data had to be entered separately in multiple databases. In practice this led to data loss in situations where the data needed to be transferred or copied. Solutions proposed included financial incentives for Member States to share such data 51 and furthering data sharing between EU criminal justice agencies by offering Eurojust and OLAF access to the data.
As the impact assessment was careful to point out, Europol has neither autonomous investigative capabilities nor coercive powers, so it is very distinct from national police forces or from federal law enforcement agencies such as the FBI. 52 Neither of the main problems to which the Regulation sought solutions proposed either of the hallmarks of a supranational police authority. However, a number of separate challenges were also tackled by the proposal. First, the Lisbon Treaty introduced a Treaty requirement for scrutiny by the European parliament and national parliaments. 53 Second, the data protection case law of the Court of Justice suggested that the data protection officer system was insufficiently robust. 54 The general approach to governance of EU agencies also required reflection on whether Europol governance could be amended to better fit this approach. 55
The reception of Europol’s reform by national parliaments
Europol reform was relatively uncontroversial for national parliaments, especially in the context of a year in which the proposal for the European Public Prosecutor’s Office was published. 56 The Belgian House of Representatives 57 had concerns about the subsidiarity of EU action for several reasons linked to openness and transparency: National authorities would no longer monitor fundamental rights linked to data processing, and EU-level procedures prevented the development of national parliamentary scrutiny and reduced parliamentary freedom of speech in so far as they required the confidentiality of Europol scrutiny. The German Bundesrat was primarily concerned with proposed changes to training and the merger of the European Union Agency for Law Enforcement Training (CEPOL) with Europol. 58 Its principal concern was whether the proposed training system for Europol was within the Scope of EU competence in Article 87(2)(b) TFEU. The Spanish Cortes Generales was also doubtful of the merger of CEPOL functions with Europol and whether this was within the spirit of ‘support’ envisaged in Article 88(2)(b). 59 Concerns about the extent to which mandatory information transmission by Member States would overwhelm the agency with low-quality information and concerns about the possibility that information exchange with third parties would weaken mutual trust in Europol were presented as subsidiarity concerns. At later stages some national parliaments issued political dialogue statements in which they raised concerns about the diminishing or unclear role or structure of national parliamentary scrutiny, 60 the lack of sufficiently precise definition of Europol operational powers, 61 and issues in defining relationships with private parties and ensuring the confidentiality of investigations. 62 In sum, the principal concerns of Member States centred on ensuring that EU-level policing is subject to sufficiently robust and transparent control.
A compromise text was accepted after trilogue negotiations in November 2015 63 and taken up in the Council’s position at first reading of March 2016, 64 which was accepted unamended by the EP in its second reading in May 2016. 65 The trilogue discussions 66 resulted in increased data-protection safeguards and developed detailed rules for parliamentary scrutiny on which the EP had insisted in its first reading position. 67 The European Data Protection Supervisor and national supervisory authorities were to form a ‘Cooperation Board’, while a Joint Parliamentary Scrutiny Group formed by members of national parliaments and the EP was to monitor the impact of Europol activities on the fundamental rights and freedoms of natural persons. 68 Information-exchange agreements with third countries became subject to the Regulation and the Commission was tasked with assessing these agreements. 69 Whilst the original proposal and the Council’s general approach mainly prohibited data exchange with private parties, negotiations resulted in the development of provisions supporting and Internet Referral Unit that allow such cooperation. 70 The new data management architecture forms a lion’s share of the Regulation text, 71 and empowers the Europol Management Board to adopt guidelines for the processing of information after consulting the European Data Protection Supervisor. 72 The trilogue negotiations produced a data protection regime which includes strict restrictions on processing overseen by an internal Data Protection Officer reporting to the European Data Protection Supervisor. 73 The institution is subject to the general access to documents Regulation. 74
Conclusion
Despite its emerging supranational features, Europol is still, in many ways, in the hands of Member States. The most prominent changes to Europol since Lisbon point to the kind of new intergovernmentalism that has been invoked to describe European integration since Maastricht. In particular, the trends that characterize much of the evolution of JHA are present in Europol’s recent history. Seen from the perspective of a possible path towards genuine supranational police powers, it is clear that both formal legal limits and practical issues continue to limit Europol’s potential to develop into an autonomous and operational police power. At the level of legislative competences, the Treaty exclusions will require national authorities’ proactive support in any future operational tasks, given that ‘coercive measures’ remain excluded at the level of the Treaty. It remains to be seen whether Europol’s dependence on national authorities’ data will continue to frame the level of support that the institution can provide in the fight against transnational crime. Changes that improve the accountability of Europol, such as an effective parliamentary monitoring system, should develop trust in the institution and this in turn will facilitate the timely and full sharing of criminal intelligence necessary for those tasks. Europol will, for the foreseeable future, remain an institution predicated on data sharing and the extent to which this is successful depends on trust rather than coercion.
