The paradox-ridden relationship between the constitution and security

Modern constitutions have an ambivalent relation to security. Security considerations have often – in a Hobbesian way, we might say – weighed heavily among the objectives of a new political formation and its constitution; that is, behind the constitutive functions of the constitution. That was the case with the US Constitution, for instance. Security objectives, such as insuring domestic tranquillity and providing for common defence, occupy a prominent place in the Preamble:

Yet constitutions fulfil not only constitutive but restrictive functions too: state constitutions are expected to discipline the use of public power, primarily through fundamental rights, alongside democratic and judicial control. The restrictive function, closely related to the ideas of a democratic Rechtsstaat (a constitutional democracy), focuses on executive bodies and activities related to internal and external security, i.e., police and armed forces. For the restrictive function, security as embodied in such institutions and activities is not something to be brought about by the constitution but something pre-existing it; something that must be subordinated to constitutional forms, procedures and limitations. However, constitutionalization has not been an easy task, due to the specific nature of security activities: they belong to the executive domain, consist of factual rather than normative measures and seem to escape the reach of the legislative function. We shall note these features resisting constitutionalization and juridification in general in the security dimension of the EU too.

Some constitutions put more emphasis on the constitutive and others on the restrictive function, but as a rule, state constitutions are expected to accomplish both of them. Even if the accent is on the constitutive function, constitutions also impose restrictions on security agencies and activities. Restrictions are necessary for managing what might be called the first paradox of a security constitution: if unchecked, security agencies are prone to turn themselves into security threats for citizens.

A further paradox inheres in the restrictive function. A constitution is supposed to curtail security activities and bring them under democratic and judicial supervision. However, security objectives are allowed to justify curbing the effect of those very restrictions and monitoring mechanisms, even suspending them. This dialectic between fundamental rights and security considerations can be seen in both limitations on and derogations from fundamental rights; a distinction adopted, for instance, in the European Convention on Human Rights (ECHR). Limitations are allowed even under conditions of ‘normalcy’, whereas derogations are justifiable only in a state of emergency. Arts. 5–11 of the Convention include clauses on the grounds warranting limitations to rights, and some clauses explicitly invoke security grounds, such as national security or public safety.Footnote ¹ In turn, derogation, i.e., temporary suspension of rights, ‘in time of war or other public emergency threatening the life of the nation’ is provided for by Art. 15 of the Convention.Footnote ²

In standard cases, states of emergency are justified by security exigencies. In addition to suspending the protection offered by fundamental rights, a state of emergency may also entail empowering the executive at the expense of the legislature, and adjourning democratic and judicial control of security organs. In sum, states of emergency – or states of exception, as they also are called – imply setting aside the fundamental constitutional principles defining a democratic Rechtsstaat. A security constitution centred around the doctrine of state of emergency has traits of an anti-constitution: another paradox in the relations between security and the constitution!

For both the constitutive and restrictive function, security is an extra-constitutional objective, a common good, which public authorities are expected to produce under constitutional restrictions and supervisory mechanisms but which also justifies limiting or even suspending these constitutional guarantees. However, in recent decades, a tendency to transform security into a fundamental right, equal in rank to traditional liberty rights, has made itself manifest. Such a transformation would shake the doctrinal foundations of a liberal Rechtsstaat: a previously extra-constitutional collective good, the main target of the restrictive function of modern constitutions, would be turned into an inherent element of the restrictive system of fundamental rights. Evidently, the list of paradoxes attached to a putative security constitution must again be amended!

In total, a basic ambiguity permeates the relationship between security and the constitution. Security seems to be unable to find a stable location with respect to the constitution and to establish itself either within or outside the constitution. But if, in spite of these paradox-ridden relations, we still engage in a search for a security constitution, how should we proceed? The natural way would appear to be to point to individual constitutional provisions relating to security and to label their assemblage the security constitution, as, indeed, has been done in post-9/11 debates, especially in the United States. In these debates, the term ‘security constitution’ or ‘national security constitution’ has been employed at the state level, although rather hesitantly.Footnote ³ Candidates for inclusion in such a ‘surface-level’ constitution comprise, for instance, provisions concerning decisions on war and peace; on armed forces; and on states of emergency and emergency powers. In the European Union, ever since Maastricht the Treaties have involved distinct Titles or Chapters on Common Foreign and Security Policy (CFSP) as well as issues of internal security too. So arguably at present the European security constitution consists of Title V of Chapter III TFEU (Area of Freedom, Security and Justice), complemented, perhaps, by security-related provisions in Title I TEU. But if ‘security constitution’ is understood, as I think it should be, in more substantive terms than just as an additive or enumerative category, then a minimum degree of coherence uniting individual constitutional provisions should also be required. In effect, some observers have claimed that, constitutionally speaking, the AFSJ does not amount to more than a fragmented collection of policy provisions. No justification exists – so the argument continues – for elevating the AFSJ above other EU policy domains and privileging it with the status of a distinct constitutional dimension, comparable to, say, the European (micro)economic constitution. A crucial testing-stone in the search for a European security constitution is whether a unifying background concept(ion) of security exists which injects coherence into the amalgamation of security-related Treaty provisions.

We have already noticed how contentious is the underlying constitutional theory in the various dimensions of the European constitution. The security dimension is no exception. In this dimension, the main formal criteria of constitutionalization pertain, on the one hand, to legal and political transnationalization at the expense of international law and institutional intergovernmentalism, and, on the other hand, to individualization: subjectivization of the individual in the two senses of the term. Both these criteria hint at substantive controversies concerning the foundations of the putative security constitution.

In the security dimension, the process of detaching from international law has been particularly arduous. The reason is obvious: security issues touch on the heart of the traditional understanding of state sovereignty, which includes, as Max Weber famously put it in the very definition of ‘state’, state monopoly over the use of legitimate violence. Lingering features of international law arrangements and enhanced institutional intergovernmentalism testify to Member State sovereigntist concerns. As these concerns are more accentuated in some Member States than others, they have led to legal and institutional fragmentation: a complicated system of opt-outs and opt-ins. Potential tensions between the transnational and the national characterize European constitutionalization in all its dimensions, but they are especially pertinent in the security dimension.

The dialectic between the constitutive and restrictive function of the constitution is reflected in the dual position of European individuals. A tension between security considerations and fundamental rights permeates the security constitution, implying two different perceptions of the individual: the individual is treated as both a security risk and as a citizen endowed with fundamental rights.

In line with other constitutional dimensions, the security constitution fulfils a positioning function and, again in line with other constitutional dimensions, with regard both to Member States and the rest of the world. Ever since Maastricht, the Treaty on the European Union has included a Title on Common Foreign and Security Policy (CFSP). Below, following the focus of preceding chapters, my focus will be on internal Union security and the relations between the European and Member State level. This choice is also warranted by the low degree of juridification of the external security dimension; to a great extent, CFSP is the province of ‘high politics’ in the European Council and the Council of Foreign Ministers. As Art. 31(1) TEU explicitly sets out, adoption of legislative acts is excluded. In turn, Art. 275 TFEU precludes ECJ jurisdiction with respect to CFSP provisions or acts adopted on that basis.Footnote ⁴

From Trevi to Maastricht

From the very beginning, since Paris 1951, European integration possessed a prominent security aspect. One of the major objectives – if not the major objective – of integration was to guarantee peace among former enemies through reinforcing their economic ties. Accordingly, in the Preamble to the Treaty of Rome the Signatory States defined pooling their resources as a means ‘to preserve and strengthen peace and liberty’. But peace was a second-order, extra-Treaty objective, comparable to Hobbesian security as the rationale behind polity-building and constitution-making. Achieving this objective would be the work of the economic constitution and not a separate security constitution.

Security considerations entered the domain of what would develop into the European microeconomic constitution through the justificatory regime of free movement law. The provisions on economic freedom explicitly allow for Member State derogations on security-related grounds. Free movement of goods does not ‘preclude prohibitions or restrictions on imports, exports or goods in transit justified on grounds of public morality, public policy [in French, ordre public] or public security’ (now Art. 36 TFEU). And, correspondingly, free movement of workers is subject to ‘limitations justified on grounds of public policy (ordre public), public security or public health’ (now Art. 45(3) TFEU).Footnote ⁵

Such exemption clauses did not give rise to a distinct security dimension but remained an integral part of economic constitutional law. Furthermore, they unfolded their effects at the Member State rather than the European level; indeed, they aimed at preserving Member State sovereignty in security-related questions. However, free movement of workers has played an intermediary role between the economic and security dimensions, as it has in the interrelations between the economic and social constitutions too. Particularly relevant is the clause permitting Member States to restrict free movement of workers on grounds of ordre public and public security. Member States perceived worker mobility and opening of internal borders as a potential threat to security, which required vigilance over border-crossing movements of persons potentially jeopardizing ordre public or public security. Samuli Hurri argues that this led to the birth of a European individual whose legal status was defined outside the dichotomy of the law of nationals and the law of aliens.Footnote ⁶ What is significant for our present discussion is that this European individual was born as a problem: as a security threat.

Hurri locates the birth of the European individual in the 1970s, when the ECJ issued its pivotal rulings on interpretation of the derogation clause in the provision on free movement of workers. This was the decade when terrorism appeared as a major security risk in Europe: the decade of intensified activity of the IRA in Northern Ireland and elsewhere in the UK; ETA in Spain and France; Black September and the Rote Armee in Germany; and Brigade Rosse and extreme-right terrorism in Italy. The upsurge of terrorism was also a central backdrop to the Trevi framework, one of the ‘laboratories’ where border-crossing security cooperation was experimented with in advance of any constitutional mandate.Footnote ⁷ The Trevi framework anticipated many pertinent traits of the EU security dimension: not only the initial experimental nature of cooperation ‘in the shadow of the Treaties’ but also the perseverance of intergovernmentalism and the weight of state-sovereigntist obstacles to transnationalization; the importance of cooperation and networking among Member State security officials; and exchange of information as a central operational modus.

The Trevi framework was founded by Member State Ministers of Justice and Home Affairs in 1975. The objective was to combat terrorism through exchange of information about training, equipment and organization of police action. In the mid-1980s the fight against football hooliganism and serious organized crime, such as drug trafficking, arms smuggling and large-scale bank robberies, was added to the tasks of specialist Working Groups, composed of security professionals from the participating states. Ministers of Justice and Home Affairs were responsible for policy development and overall review of cooperation. Trevi was intergovernmental cooperation with emphasis on technical issues and information sharing among Member State security officials, and subject only to international law. The problems to which Trevi initially responded, such as terrorism or football hooliganism, were not directly related to the establishment or functioning of the common market. In its last phase the Trevi network was associated with the Schengen process, where the link to economic constitutionalization was evident.

The Schengen Agreement, concluded in 1985, anticipated completion of the internal market project, scheduled for the end of 1992. Removal of internal border controls was seen to require compensatory security-related measures, which were included in the Convention on the Implementation of the Schengen Agreement, signed in 1990.Footnote ⁸ The Convention was an agreement under international law, manifesting the typical pioneer function of international law instruments in the security dimension. Schengen introduced another distinctive feature of the security dimension, which also has its backdrop in state-sovereigntist concerns: territorial fragmentation or so-called variable geometry. Recourse to an international law agreement instead of European legislation was motivated not only by issues of competence but also by the doubts of some Member States in respect of abolishing internal border controls. In effect, the Schengen Agreement and Convention were originally signed by only five Member States. The Schengen acquis was integrated into the framework of EU law through a Protocol attached to the Amsterdam Treaty (1998), but this was accompanied by a complicated system of opt-outs and opt-ins, so that by no means did it spell the end of variable geometry. The Schengen Convention included provisions on external border controls, visas and residence permissions, and asylum applications;Footnote ⁹ on police cooperation, including cross-border police operations; mutual assistance in criminal matters; application of the ne bis in idem principle; extradition; narcotics; as well as on firearms and ammunition. In addition, the Convention established the Schengen Information System, which reflects the importance of data processing and information exchange for the security dimension.

Neither Trevi nor Schengen had any anchorage in Treaty law. Only in Maastricht did the security dimension receive a place in the ‘constitutional charter’ within the new three-pillar structure of the Union. A particular Title on Cooperation in the Fields of Justice and Home Affairs – the third pillar of the Union – acknowledged internal security as a separate policy field, irreducible to the economic constitution. Justice and Home Affairs were, though, still defined as derivative from free movement of persons, one of the cornerstones of the economic constitution. However, Maastricht enlarged the personal scope of free movement to European citizens, and by the same token, enhanced its potential security consequences.

The Justice and Home Affairs Title listed nine policy areas as matters of common interest and subjects of cooperation.Footnote ¹⁰ Still, in light of the transnationalization criterion, constitutionalization made but modest progress. Third-pillar provisions were impregnated by intergovernmentalism and recourse to international law instruments. An explicit bow in the direction of state sovereigntism was made by declaring that the provisions ‘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’. Administrative information exchange and consultation were to be the main operational modus in areas of common interest (Art. K.2). Decision-making in the third pillar rejected the Community method, based on the agenda-making privilege of the Commission and the supervisory role of the ECJ and also presupposing the involvement of the European Parliament. The main third-pillar institution was the Council, and the requirement of unanimity, with only few exceptions, enhanced the prevalent intergovernmentalism (Art. K.3).Footnote ¹¹ The Treaty granted the Commission only limited initiative, and in the areas where it was competent, it shared the initiative with Member States (Art. K.4). The main legal instruments at the disposal of the Council were international law conventions which the Council could recommend to Member States for adoption in accordance with the national constitution.Footnote ¹² Somewhat paradoxically, conventions under international law could induce steps in the transnational direction: if the conventions did not provide otherwise, implementing measures could be adopted by a majority vote among the contracting parties.Footnote ¹³ As a rule, the ECJ did not possess jurisdiction within the third pillar (Art. K.3(2)).Footnote ¹⁴

Maastricht did not immediately affect the Schengen Convention, which remained outside the umbrella of Treaty law and was not applicable in all Member States. However, in visa policy replacing Schengen by EU legislation was made possible.Footnote ¹⁵ In addition, the Justice and Home Affairs Title sanctified the variable geometry of the security dimension, already manifested by Schengen. Closer cooperation between two or more Member States was possible ‘in so far as such cooperation does not conflict with, or impede, that provided for in this Title’ (Art. K.7).

The significance of Amsterdam

The Treaty of Amsterdam signified a new boost to the constitutionalization of the security dimension. Internal security was explicitly included among EU objectives, although in the Preamble to the TEU and in Art. 2 TEU it was still subordinated to the primary objective of free movement of persons. In the amended Preamble, the Member States declared their resolution to ‘facilitate the free movement of persons, while ensuring the safety and security of their peoples, by establishing an area of freedom, security and justice’. A corresponding objective was added to Art. 2 TEU: ‘to maintain and develop the Union as an area of freedom, security and justice, in which the free movement of persons is assured in conjunction with appropriate measures with respect to external border controls, asylum, immigration and the prevention and combating of crime’.

Amsterdam changed the designation of the internal security dimension into the Area of Freedom, Security and Justice (AFSJ) and split the provisions concerning this area into two. Provisions on police and judicial cooperation remained in the Treaty on the EU as the third pillar of the Treaty architecture. A free-standing security objective opened the new Title: ‘Without prejudice to the powers of the European Community, the Union's objective shall be to provide citizens with a high level of safety within an area of freedom, security and justice by developing common action among the Member States in the fields of police and judicial cooperation in criminal matters and by preventing and combating racism and xenophobia.’ Cooperation was extended to crimes with no inherent cross-border aspect, which constitutionalized developments within the Trevi framework and its successor, Europol.Footnote ¹⁶ The envisaged high level of safety was to ‘be achieved by preventing and combating crime, organized or otherwise, in particular terrorism, trafficking in persons and offences against children, illicit drug trafficking and illicit arms trafficking, corruption and fraud’ (Art. 29 TEU).

In Amsterdam, matters directly related to free movement of persons, such as external and internal borders, visas, asylum and immigration, were transferred to the Community pillar. Still, this did not deprive them of all their state-sovereigntist and intergovernmental features. Firstly, the new Title of the Treaty on the EC reconfirmed the assurance that Member State responsibilities in maintaining law and order and safeguarding internal security would not be affected (Art. 64(1) TEC). Secondly, AFSJ issues of the Community pillar were subject to a transitional regime for five years. During this period, the Council acted unanimously on a proposal from the Commission or on the initiative of a Member State and after consulting the European Parliament. After the transitional period, the Council could act only on the initiative of the Commission; the Commission, however, was obliged to examine any request by a Member State that it submit a proposal to the Council. Thus, a step towards the Community method was taken, but not a full step. Yet the Title also opened the possibility of further integration. The Council could, through a unanimous decision, transfer AFSJ policy fields to the co-decision procedure, which entailed qualified majority voting in the Council and the involvement of the European Parliament, as well as jurisdiction of the ECJ (Art. 67 TEC). In 2004, this passerelle clause was applied to asylum and immigration. The typical fragmentation of the security dimension was manifested by the opt-outs granted to the Member States most worried about their national sovereignty: the UK, Ireland and Denmark. These countries were also accorded special treatment in the Protocol integrating the Schengen acquis into the EU framework. The Amsterdam Treaty retained the possibility of closer cooperation between Member States too, which was now regulated by general TEU provisions.

Intergovernmental features were even more pronounced in the fields which remained in the third pillar and under the Treaty on the EU. While the AFSJ Title of the TEC introduced intergovernmental elements into the Community pillar, the stripped-down Title VI TEU brought some transnational traits into the third pillar. The Council was granted the power to decide that police cooperation and judicial cooperation in criminal matters fell under the procedures laid down in the AFSJ Title of the TEC. Provisions on the legal instruments at the disposal of the Council also took a step in a transnational direction. International law conventions were still provided for, but they found an alternative in framework decisions, a juridical novelty of the Amsterdam Treaty. Framework decisions could be adopted for the purpose of approximation of the laws and regulations of the Member States. They were similar to directives in the Community pillar in binding Member States as to the result to be achieved but leaving to national authorities the choice of form and methods. However, an explicit provision denied them any direct effect; here the drafters had had an eye on ECJ case law on directives. A similar disclaimer applied to decisions, which the Council could adopt for other purposes consistent with the objectives of the Title than approximation of Member State laws and regulations.Footnote ¹⁷ Diverging from conventions, framework decisions, although requiring a unanimous Council, did not fall under international law. In practice, they superseded conventions as the primary legal instrument of the third pillar.Footnote ¹⁸ Their importance was still enhanced by the Pupino ruling, which accorded a framework decision a legal effect similar to the indirect effect of Community directives.Footnote ¹⁹ All in all, in the pre-Lisbon era juridical constitutionalization took interesting sidesteps in the security dimension.

A further sign of transnationalization, related to juridical constitutionalization, was the extension of ECJ jurisdiction into the third pillar. The Court was granted jurisdiction to issue preliminary rulings on the validity and interpretation of framework decisions and decisions, as well as on the interpretation of conventions and on the validity and interpretation of measures implementing them. Yet accepting the Court's jurisdiction was optional for Member States, which could restrict the competence to request a preliminary ruling to the highest courts or extend it to all courts and tribunals: again a token of the power of state sovereigntism, contributing to the constitutional mishmash of the AFSJ. The Court was explicitly denied jurisdiction to review the validity or proportionality of operations carried out by the police or other law enforcement services of a Member State or the exercise of Member State responsibilities in law and order maintenance and in safeguarding internal security (Art. 35 TEU).

That these responsibilities were incumbent on Member States was reconfirmed in Amsterdam (Art. 33 TEU). However, police cooperation was no longer confined merely to mutual consultation, data processing and information exchange, but reached out to operational cooperation between law enforcement services, thus constitutionalizing the inroads into national sovereignty already accepted by Schengen states. State sovereigntism, though, retained an impact: authorities of a Member State could only operate in the territory of another Member State in liaison and in agreement with the authorities of that State (Art. 32 TEU). Another important opening of the Amsterdam Treaty pertained to harmonization in criminal and criminal procedural law; a traditional core area of state legislative sovereignty. Common action in judicial cooperation included ensuring such compatibility in national legislation as was necessary to improve cooperation in procedural and enforcement matters and facilitate extradition between Member States. It also included ‘progressively adopting measures establishing minimum rules relating to the constituent elements of criminal acts and to penalties in the fields of organized crime, terrorism and illicit drug trafficking’ (Art. 31 TEU).

The Amsterdam Treaty created a decisive impetus to activities in the security dimension, which rapidly gained a high ranking on the legislative and policy agenda of the EU. New agencies and networks were established, while administrative cooperation, which had been exercised ever since the establishment of Trevi, was intensified.Footnote ²⁰ In the eyes of many observers, the AFSJ became the most dynamic field of the EU in the post-Amsterdam era. The new guiding role of the European Council, whose position had been enshrined in the Maastricht Treaty, was a token of the political significance accorded to the AFSJ. In October 1999 in Tampere, Finland, the European Council adopted the first comprehensive programme for activities in this area, followed by the Hague Programme in 2004 and the Stockholm Programme in 2010.Footnote ²¹ The Tampere Conclusions introduced in the security dimension the principle of mutual recognition which subsequently found application in, for instance, the Framework Decision on the European Arrest Warrant (EAW).Footnote ²² Just as in free movement law within the economic constitution, mutual recognition balances between transnationalism and intergovernmentalism. Unlike in harmonization, the Member States retain sovereignty over national legislation, but they are obliged to acknowledge the legal effects of the legislation and decisions of other Member States. The programmes adopted by the European Council have been complemented by Commission action plans, manifesting the further intrusion of the Commission into the (former) citadel of intergovernmentalism.

More than twenty framework decisions were adopted within the third pillar before the entry into effect of the Lisbon Treaty. In addition to the European Arrest Warrant, framework decisions pertained, for instance, to terrorism, environmental crimes, sexual exploitation and child pornography, and drug trafficking.Footnote ²³ At the institutional level, the constitutional position of Europol was strengthened through explicit provisions in Title VI TEU. Europol is the institutional site for police cooperation, data processing and mutual exchange of information and experience. Eurojust was established as a corresponding site for cooperation among public prosecutors, and in Nice a Treaty basis was created for this agency too (Art. 31(2) TEU). New post-Amsterdam agencies within the AFSJ include Frontex, operating in the field of external border guarding; the European Asylum Support Office (EASO); the European Anti-Fraud Office (OLAF); the European Police College (CEPOL); and the European Monitoring Centre for Drugs and Drug Addiction (EMCDDA). The agencies are supplemented by networks of security professionals, a characteristic feature of the EU security dimension.Footnote ²⁴ Another important aspect in reinforcement of the AFSJ is extension of information systems, a development which was launched even before Maastricht by the Schengen Information System. The end result is a hardly controllable, fragmented whole, with obvious overlaps among its constituent parts.Footnote ²⁵

The 9/11 attacks fell in the post-Amsterdam period. The EU reacted rapidly. The Justice and Home Affairs Council met on 20 September 2001 and agreed on a package in the fight against terrorism, which was endorsed the next day by the European Council. The conclusions of the European Council included the need for a European definition of terrorism, which was provided on 13 June 2002 by the Framework Decision on Combating Terrorism. The ‘jewel in the crown of the EU's response to the terrorist attacks’Footnote ²⁶ was the EAW Framework Decision, adopted by the Council simultaneously with the terrorism decision. Anti-terrorism measures were taken within other pillars too. Implementation of UN Security Council Resolutions 1373 (28 September 2001) and 1390 (16 January 2002) fell under the first pillar of Common Foreign and Security Policy and the second, Community pillar. On 27 December 2001, the Council adopted two common positions under the first pillar, which focused on the fight against financing terrorism, such as freezing funds of individuals and organizations suspected of terrorist contacts and included in lists annexed to the common positions. The Community pillar was involved through Council regulations on implementation of the CFSP common positions.Footnote ²⁷ EU reactions to 9/11 and the bomb explosions in Madrid (2004) and London (2005) also included the European Security Strategy and the European Union Counter-Terrorism Strategy, which the European Council adopted on 12 December 2003 and 30 November 2005, respectively. Moreover, a particular EU Counter-Terrorism Coordinator was appointed and such networks of security professionals as the Counter-Terrorism Group and the Police Working Group on Terrorism were created.Footnote ²⁸ On the external front anti-terrorist cooperation with the United States was greatly intensified in the aftermath of 9/11.

Although 9/11 had a weighty impact on development of the security dimension, it can hardly be considered a decisive turning point. Anti-terrorism had stood high on the agenda ever since the launch of Trevi, and some of the major legislative measures adopted in the aftermath of 9/11 had been drafted before the World Trade Center towers fell in New York; this applies, for instance, to the Framework Decision on the European Arrest Warrant. The attacks of 9/11 or the subsequent attacks in Madrid and London did not leave any immediate trace in Treaty law. However, the Solidarity Clause in Art. 222 TFEU can be traced back to the attacks. Furthermore, Art. 75 TFEU was expected to clarify competences in the fight against the financing of terrorist activities.Footnote ²⁹ Despite the potential importance of the Solidarity Clause, the indirect constitutional implications of anti-terrorism measures are probably more important. They concern the pillar structure and the dividing line between transnationalism and intergovernmentalism. Combating terrorism was seen to require complementary measures in diverse pillars, as is shown by the simultaneous adoption of first-pillar common positions and Community-pillar regulations on financing terrorism. Anti-terrorist policy also questioned the separation of external and internal security, the former confined within the first pillar, and the latter divided between the Community pillar and the rump third pillar. Furthermore, legal measures approved in the wake of 9/11 gave rise to constitutionally important case law in both the EU and Member States. In their decisions on implementation of the EAW Framework Decision, several Member State constitutional courts addressed fundamental issues concerning the relationship between EU law and national constitutional law.Footnote ³⁰ On the side of the EU, Kadi reaffirmed the autonomy of EU law from international law, clarified the role of fundamental rights in security measures, and even pronounced on relations between the pillars and the competence basis of Community regulations on financing terrorism.

Lisbon and after

For the purposes of this chapter, the (hi)story of EU security-related constitutional law can be brought to an end with a brief summary of the security-related provisions of the Lisbon Treaty. I will start with the Solidarity Clause (Art. 222 TFEU).

The Solidarity Clause not only possesses deep symbolic significance but extends solidarity obligations to military resources which Member States are expected to put at the disposal of the Union. The Clause calls for joint action in a spirit of solidarity by the Union and its Member States in preventing and responding to terrorist attacks and in assisting a Member State hit by a natural or man-made disaster. The Union should mobilize all the instruments at its disposal, including military resources which Member States have made available. Assistance following a terrorist attack or natural or man-made disaster presupposes a request by the political authorities of the Member State concerned. The Council decides on the measures on a joint proposal by the Commission and the High Representative of the Union for Foreign Affairs and Security Policy. A decision that has defence implications requires unanimity. Thus, even when acting in a spirit of solidarity, the Union is expected to heed state-sovereigntist concerns!Footnote ³¹

As the Solidarity Clause demonstrates, the Lisbon Treaty pushed further the piecemeal constitutionalization of the security dimension. Yet it did not eradicate all remnants of intergovernmentalism, nor did it cut the ties linking the security dimension to free movement. In the Preamble to the TEU and in the Treaty itself, security remains subordinated to the primary objective of facilitating free movement of persons. Still, significantly enough, offering citizens an area of freedom, security and justice has been upgraded in the list of objectives and now precedes establishment of the internal market (Art. 3(2–3) TEU). This reflects the new prominence of security in the post-Amsterdam era, as does the explicit enshrinement in the Treaty of the strategy role which the European Council had already exercised in Tampere in 1999 and in The Hague in 2004.Footnote ³²

The Lisbon Treaty abolished the pillar structure, integrated the AFSJ into the institutional structure of the former Community pillar and extended the use of Community legislative instruments and procedures to fields which the Amsterdam Treaty had still left to the truncated third pillar of police and judicial cooperation. The jurisdiction of the Court was accordingly amended. In consequence, regulations and directives can now be adopted in all policy fields of the AFSJ. As a rule, they are approved in the ordinary legislative procedure, involving a Commission initiative, a co-decision of the European Parliament and qualified majority voting in the Council. These are important developments for transnationalization and constitutionalization of the security dimension. However, the TEU still emphasizes the primary responsibility (and sovereignty) of Member States in the security field. Art. 4(2) lays down that the Union ‘shall respect their essential State functions, including ensuring the territorial integrity of the State, maintaining law and order and safeguarding national security’ and that ‘in particular, national security remains the sole responsibility of each Member State’. Moreover, exclusion of Member State security measures from the Court's jurisdiction is reconfirmed in Art. 276 TFEU.

The TFEU Title on the AFSJ also makes several concessions to state sovereigntism and intergovernmentalism. AFSJ measures are supposed to respect not only fundamental rights but the different legal systems and traditions of the Member States as well. The specific objectives listed in the Title accord primacy to Member State action but call for mutual trust and solidarity too. Common policy on asylum, immigration and external border controls is to be based on solidarity between Member States,Footnote ³³ while in criminal matters and access to justice emphasis is on the principle of mutual recognition; harmonization of criminal laws can be resorted to only ‘if necessary’.Footnote ³⁴ The Title also re-invokes the reference to ‘the responsibilities incumbent upon Member States with regard to the maintenance of law and order and the safeguarding of internal security’ which has accompanied constitutionalization of the security dimension from its Maastricht beginnings (Art. 67 TFEU).

Provisions on policy competences too defer to state sovereignty as well. Art. 67(4) TFEU declares that Member State power concerning the geographical demarcation of borders will be left intact. Accordingly, Art. 79(5) proclaims that the Member State right to ‘determine volumes of admission of third-country nationals coming from third countries to their territory in order to seek work, whether employed or self-employed’, will not be affected. In the legislative procedure, Member States have retained a residue of their power of initiative: legislative acts can be adopted within the AFSJ, not only on a proposal by the Commission, but also on the initiative of a quarter of the Member States.

Finally, the powers of EU agencies and authorities of other Member States in police and judicial cooperation are accompanied by substantive and procedural qualifications which also heed state sovereigntism. Operational action by Europol or authorities of other Member States must be carried out in liaison and in agreement with the authorities of the Member State or States whose territory is concerned. The main function of Eurojust and Europol remains to provide a site for coordination and cooperation between Member State authorities and professionals. Prosecutorial cooperation under the auspices of Eurojust does not cover ‘formal acts of judicial procedure’, which are to be carried out by competent national officials (Art. 85(2) TFEU).Footnote ³⁵ The Treaty, though, enables developing Eurojust into a European Public Prosecutor's Office, responsible for ‘investigating, prosecuting and bringing to judgment, where appropriate in liaison with Europol, the perpetrators of, and accomplices in, offences against the Union's financial interests’. Such an agency would also exercise ‘the functions of prosecutor in the competent courts of the Member States in relation to such offences’. However, such a move in a transnational direction requires surmounting an intergovernmental threshold. It requires a special legislative procedure involving unanimity in the Council and consent by the European Parliament. The procedural hurdle is still higher for extending the mandate of the European Public Prosecutor's Office to serious crime having a cross-border dimension. In addition to a unanimous decision by the European Council, the consent of the European Parliament and consultation with the Commission is needed (Art. 86 TFEU).

A special legislative procedure, including a unanimous Council decision and consultation with or consent from the European Parliament, must be followed even in other matters where action by EU agencies or authorities of other Member States has been considered to be particularly sensitive for national sovereignty. These include judicial cooperation in family law questions with cross-border implications (Art. 81(3));Footnote ³⁶ extension of the list of serious crimes with a cross-border dimension where minimum rules on the definition of criminal offences and sanctions may be established through directives adopted in the ordinary legislative procedure (Art. 83(1–2)); and operational cooperation among Member State police, customs and other specialized law enforcement services (Art. 87(3)). As a counterbalance, some of the provisions facilitate a minimum of nine Member States proceeding with enhanced cooperation; a token of variable geometry, characteristic of the security dimension.

Harmonizing criminal and criminal procedural law through minimum rules falls under the ordinary legislative procedure. But here too emergency brake clauses address state-sovereigntist sensitivities. If 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 be referred to the European Council. If that happens, the ordinary legislative procedure will be suspended and can only be resumed if the European Council reaches a consensus. In case of disagreement, nine Member States can proceed with enhanced cooperation with the draft directive (Arts. 82(3) and 83(3) TFEU).

In total, even after Lisbon the AFSJ possesses conspicuous intergovernmental features and defers to state-sovereigntist concerns. Yet, since Maastricht, progress in the transnational direction has been considerable, and inroads have been made even into the hard core of traditional national sovereignty in policing, and criminal and criminal procedural law. State-sovereigntist sensibilities account for the complex system of opt-outs and opt-ins with which the AFSJ seems not to be able to dispense and which are provided for in Protocols annexed to the Lisbon Treaty. Coherence within the AFSJ has increased due to abolition of the pillar structure and the ‘communitarization’ of the former third pillar. But this increased coherence finds its counterweight in the fragmentation brought about by opt-outs and opt-ins, as well as possibilities for enhanced cooperation provided by provisions on special legislative procedure and emergency brakes.

From state of emergency to security

We have been able to observe an advancing constitutionalization of the security dimension at the Treaty level starting in Maastricht from the declaration of a number of security issues as common interests of the Member States, and ending up in Lisbon with the upgrading of security objectives and the ‘communitarization’ of the entire AFSJ field. Moreover, we have followed the institutional build-up, often enough antedating Treaty-level constitutionalization: the establishment of EU agencies, information systems and networks for administrative and professional cooperation. But, as we have also seen, by no means has the development been unequivocal. Persistent state-sovereigntist concerns explain the concessions which even the Lisbon Treaty makes to intergovernmentalism and the lingering juridical peculiarities that constitutionalization in the security dimension shows in comparison with the (micro)economic constitution and the former Community pillar.

Individual Treaty provisions do not yet make up a security constitution. What is also needed is sufficient coherence in security-related constitutional law: a coherence deriving from an underlying conception of security issues. My search for such coherence bears some resemblance to the constructivism of the Copenhagen School of security studies. The central premise of the Copenhagen School is that what is called security or a security problem does not have any objective, independent existence but results from a naming, from a speech act by a securitizing actor: ‘by definition, something is a security problem when the elites declare it to be so’.Footnote ³⁷ This is not an innocent declaration, an innocent speech act, but implies the need of and, by the same token, a justification for extraordinary measures to maintain or restore security: ‘Security is the move that takes politics beyond the established rules of the game and frames the issue either as a special kind of politics or as above politics.’Footnote ³⁸ In legal terms, ‘politics beyond the established rules of the game’ corresponds to a state of exception (state of emergency) and extraordinary emergency measures permitted during such a state.

The emphasis on the explicit declaration of a security problem – or, to use a more dramatic term, an existential threat – and the link from there to the right to take extraordinary measures brings the Copenhagen School close to Carl Schmitt's ideas of sovereignty and state of exception. The Schmitt citation index is probably topped by the opening sentence of Politische Theologie: ‘Souverän ist wer über den Ausnahmezustand entscheidet.’Footnote ³⁹ Post-9/11 developments have launched a new wave of debates on the concept of state of exception and the pertinence of Schmitt for understanding and conceptualizing reactions to terrorism. No agreement has been reached on even the fundamental question whether a state of exception should be, or in general can be, conceived of in constitutional terms or whether it is located beyond the law and the constitution. The debates, where Walter Benjamin's and Jacques Derrida's theorizing on the relationship that law maintains with violence has also been brought in, touch on fundamental paradoxes in the very foundation of law and the constitution, and certainly possess great legal philosophical interest.Footnote ⁴⁰ They also illuminate the uneasy relation that security holds with the constitution and the difficulty in defining a fixed place for the former with regard to the latter.Footnote ⁴¹

This is not the right place to intervene in the legal philosophical discussion on sovereignty, state of exception and the constitution. I would only wish to stress the relevance of the constructivist approach and its account of the typical consequences that a declaration of a state of exception or an emergency is intended to justify. These include an at least partial suspension of fundamental rights, the rule of law and separation of powers; that is, the basic constitutional principles which define a democratic Rechtsstaat. In addition, I would also stress the wider cultural presuppositions of security-related speech acts and their success. These speech acts are not Schmittian Entscheidungen aus NichtsFootnote ⁴² but are dependent on the cultural factors which inform the Vorverständnis of ‘securitizing actors’. Nor do the speech acts succeed in their justificatory intent if they do not find (cultural) resonance among the political and legal elites, as well as the public at large.

The Copenhagen School is aware of the importance of the cultural context of securitizing speech acts. By contrast, what the Copenhagen School can be criticized for is holding – in Schmitt's wake – on a sharp distinction between state of exception and normal times. Here it follows traditional constitutional doctrine which has seen in the state of exception and concomitant emergency powers the main constitutional response to security concerns (in other words, they are viewed as constitutional means to suspend the constitution). But, as many observers have contended, the boundary separating exceptional from normal times has become increasingly blurred,Footnote ⁴³ although by no means have arguments from emergency and constitutional emergency powers exhausted all their relevance. Nor does a strict demarcation correspond to the rationale of the European security constitution. It is true, though, that the Solidarity Clause in Art. 222 TFEU invokes a pending emergency. Yet it does not adhere to the Continental European constitutional tradition, which includes the requirement of an explicit declaration of a state of emergency and the consequent (partial) suspension of such constitutional basics as fundamental rights, the rule of law and separation of powers.Footnote ⁴⁴ Tellingly enough, no emergency provision comparable to Art. 15 of the European Convention on Human Rights or Art. 4 of the UN Covenant on Civil and Political Rights was included in the EU Charter of Fundamental Rights.Footnote ⁴⁵

What, then, has been the argumentative device which has facilitated lowering the wall separating exceptional from normal times, and emergency measures from normal legislation? The concept of security has played a major part in this development. Under justification by security exigencies, normal legislation has in many countries acquired features that bring it close to the characteristic effects of a declaration of emergency or state of exception. Typically, states did not react to 9/11 and the terrorist attacks in Europe with a declaration of emergency and subsequent time-limited emergency measures but with rather far-reaching amendments to normal legislation; in the aftermath of 9/11, packages of anti-terrorist legislation were adopted not only in the United States but in such European countries as, for instance, the UK and Germany too.Footnote ⁴⁶ However, 9/11 only accelerated a long-term post-Second World War development. In the United States, the concept of national security has been invoked to warrant expanding presidential powers ever since the Truman Presidency and the start of the Cold War.Footnote ⁴⁷ Security or national security has become the main justification for a ‘permanent state of exception’. EU constitutional law must be analysed in this general cultural framework where a clear-cut concept of emergency or state of exception has increasingly yielded to a diffuse, expansionist and boundary-breaking concept of security.

‘Security’ is a tricky concept, hard to pinpoint in an unambiguous definition. The most promising way is to approach it through the concept of risk – for many sociological observers a key concept for analysing modern societyFootnote ⁴⁸ – and to define ‘security’ negatively, as connoting combating risks. Emergencies are temporarily and territorially limited and located risks that have already realized, with identifiable appearances and sources. ‘Security’ may also cover risks that have already occurred, but the emphasis is elsewhere: in risk prevention. Security calls for continuous detection and assessment of risks and security threats. Hence its inherent expansionism: ever new risks can be perceived and brought under the umbrella of ‘security’, labelled as security threats and seen as necessitating security measures. Expansionism is conspicuous in the development of the EU security dimension too, say from the modest start of anti-terrorist cooperation in Trevi to the Internal Security Strategy for the European Union, approved by the European Council in March 2010.Footnote ⁴⁹

The Strategy is explicit in its support for a comprehensive concept of internal security which reaches far beyond the traditional core fields of policing, crime prevention and maintenance of public safety. In its introductory section, the Strategy declares that ‘[t]he concept of internal security must be understood as a wide and comprehensive concept which straddles multiple sectors in order to address…major threats…which have a direct impact on the lives, safety and well-being of citizens, including natural and man-made disasters such as forest fires, earthquakes, floods and storms’.Footnote ⁵⁰ The Strategy identifies the following security threats, ‘main challenges for the internal security of the EU’: terrorism; serious and organized crime; cybercrime; cross-border crime; violence itself, such as youth violence or hooligan violence at sports events; as well as natural and man-made disasters, such as forest fires, earthquakes, floods and storms, droughts, energy shortages and major information and communication technology breakdowns.Footnote ⁵¹ A comprehensive concept of security entails that ‘internal security must be seen as encompassing a wide range of measures with both horizontal and vertical dimensions’. In the horizontal dimension, ‘to reach an adequate level of internal security in a complex global environment requires the involvement of law-enforcement and border-management authorities, with the support of judicial cooperation, civil protection agencies and also of the political, economic, financial, social and private sectors, including non-governmental organizations’. In the vertical dimension, the Strategy emphasizes that EU security policies and initiatives must be seen in the context of international cooperation, regional cooperation between Member States and harmonization of Member State national, regional and local policies.Footnote ⁵²

Related to its diffuse and expansionist character, ‘security’ possesses an intrinsic boundary-breaking tendency. It tends to blur not only the distinctions between exceptional and normal times, and extraordinary measures and normal legislation, but other divisions of established constitutional doctrine as well: such as external and internal security, war and crime, and soldiering and policing. ‘Surface-level’ EU constitutional law has still stuck to the dichotomy between external and internal security, the former – the previous first pillar – still regulated by the TEU and still dominated by intergovernmentalism. But as responses to 9/11 clearly proved, the dichotomy is difficult to maintain in practice, and AFSJ strategy documents make incursions into domains falling under Title V TEU. The Internal Security Strategy includes a section on the ‘external dimension of internal security’, premised on the claim that ‘a concept of internal security cannot exist without an external dimension, since internal security increasingly depends to a large extent on external security’.Footnote ⁵³

The Strategy also highlights a further typical feature of ‘security’: an accent on prevention. The Strategy calls for prevention and anticipation, a ‘proactive, intelligence-led approach’. It underlines that ‘a comprehensive approach must be taken that is geared to constant detection and prevention of the threats and risks facing the EU in the various areas of internal security, and the main issues of concern to the public’.Footnote ⁵⁴ The Strategy stresses the need for ‘guidelines for hazard and risk-mapping methods, assessments and analyses’ as well as an ‘overview of the natural and man-made risks that the EU may face in the future’. Such a strategy of prevention and anticipation implies the need for continuous flows of information about potential risks. Accordingly, data processing and exchange of information have formed an essential aspect of the EU security dimension. Europe-wide information systems have abounded, and EU agencies, such as Europol and Eurojust, as well as diverse networks of security professionals, have, to a significant extent, been sites for information sharing. The Hague Programme, adopted by the European Council in 2004, established the principle of availability which was to epitomize ‘an innovative approach to the cross-border exchange of law-enforcement information’. This principle ‘means that, throughout the Union, a law enforcement officer in one Member State who needs information in order to perform his duties can obtain this from another Member State and that the law enforcement agency in the other Member State which holds this information will make it available for the stated purpose, taking into account the requirement of ongoing investigations in that State’ (Para. 2.1). The Internal Security Strategy, in turn, indicates developing a comprehensive model for information exchange as one of its strategic guidelines.Footnote ⁵⁵

The crucial role of information sharing and data processing is not confined to law enforcement, but covers the whole field of security, defined in terms of risk management. Risks are something quantifiable, something which can be calculated and assessed, and calculation and assessment are impossible without information. Here we encounter a distinct feature of the modern mode of governance which Michel Foucault discussed under the label of dispositif de securité and where the main object of power does not consist of individuals, as it does for disciplinary mechanisms, but populations.Footnote ⁵⁶ Individuals may constitute risks, and preventive measures may be taken against them; yet individuals appear as risks not so much in their irreducible individual capacity but as members of a population.

Since 9/11 terrorism has topped EU lists of security threats. However, in 2001 terrorism was not a novel concern: as our historical exposé has shown, anti-terrorism cooperation was already a main impetus to the Trevi framework in the 1970s. Terrorism exemplifies many of the typical features of the concept of security which pervades EU legislation, as well as programmatic documents and practical activities. Terrorism is hard to capture in an unequivocal definition. Within the EU, terrorism has not found a fixed location in relation to criminality. As the Framework Decision on Combating Terrorism makes clear, terrorism has been considered a criminal offence or, rather, a series of criminal offences. But it seems not to be reducible to criminality but goes beyond it. In anticipation of the declaration of the Union value basis in the Lisbon Treaty (Art. 2 TEU), Recital (1) of the Preamble to the Framework Decision proclaims that ‘the European Union is founded on the universal values of human dignity, liberty, equality and solidarity, respect for human rights and fundamental freedoms’, and ‘based on the principle of democracy and the principle of the rule of law’. Recital (2) identifies terrorism as ‘one of the most serious violations of those principles’. In EU legislation and programmes, terrorism is at times assimilated with ‘other’ serious crimes, at times dealt with as a security threat separate from crime. The EU maintained a distance from the ‘War on Terrorism’ rhetoric of the Bush administration, but it has not taken a firm position in favour of treating terrorism as a crime either. It is as if defining terrorism ‘merely’ as a crime would downplay the enormity of the violation of the universalist values and principles invoked in the Recitals of the Framework Decision on Combating Terrorism.

Be that as it may, terrorism proves how conceptual diffuseness and breaking down traditional constitutional distinctions go hand in hand: war and crime, external and internal security or soldiering and policing are hard to keep separate in combating terrorism. The fight against terrorism corroborates the connection between security, prevention and information too. Potential terrorists must be identified and terrorist plots unearthed before their realization. Hence the never-satiated need for registering individuals and intercepting their communications. Terrorism is a de-localized and de-individualized security threat: it does not possess fixed headquarters but operates through networks for which territorial boundaries pose no obstacles. Terrorists, in turn, are seen not as individual persons, but as replaceable cogs in an anonymous machine; the identity of a suicide-bomber is not important, and Al-Qaeda is not in need of even Osama bin Laden. The difference between perceptions of criminal offenders and terrorists is conspicuous.Footnote ⁵⁷

Freedom, security and justice

EU Treaty law puts (internal) security into the context of the Area of Freedom, Security and Justice. An analysis of the conceptual cluster of ‘freedom’, ‘security’ and ‘justice’ rounds up our discussion of the notion of security which informs security-related EU constitutional law. How do ‘freedom’ and ‘justice’ relate to ‘security’?Footnote ⁵⁸

Much has been written about the ‘securitization’ of the AFSJ, the subjection of freedom and justice to security and their definition through their relation to the latter. And, indeed, a tendency to securitization is easy to perceive, although it is not perhaps as unequivocal as some critical discussants are prone to claim. To begin with freedom, its subordination to security is discernible, first, in its depiction as the source of the security problems which the AFSJ is expected to address. In this quality, ‘freedom’ is defined, first of all, as freedom of movement within the EU; abolishing internal border controls calls for complementary, flanking security measures. Art. 2 TEU-Amsterdam introduced the AFSJ objective in the following terms: ‘to maintain and develop the Union as an area of freedom, security and justice, in which the free movement of persons is assured in conjunction with appropriate measures with respect to external border controls, asylum, immigration and the prevention and combating of crime’. Lisbon retained this connection between freedom and security, entailing an implicit reduction of freedom to free movement (Art. 3(2) TEU).Footnote ⁵⁹

Seeing the threats addressed by security arising from freedom is not the only way of perceiving the latter through the lens of the former. In the first two AFSJ programmes, adopted by the European Council in Tampere in 1999 and in The Hague in 2004, the policy measures envisaged were divided along the sub-areas of freedom, security and justice. What fell under ‘freedom’ was common policy in the field of asylum, migration and borders, with clear emphasis on security-related measures. In essence, the Hague Programme section Strengthening freedom was primarily about strengthening security; about building up ‘Fortress Europe’, as critics have put it.Footnote ⁶⁰

Yet, from the very introduction of the AFSJ, an ambiguity in the definition of ‘freedom’ can be observed. The programmatic documents define ‘freedom’ primarily as free movement within the EU, but they intimate that ‘freedom’ can be assigned wider contents too. In the Tampere Conclusions, the challenge of the Amsterdam Treaty was seen in ensuring ‘that freedom, which includes the right to move freely throughout the Union, can be employed in conditions of security and justice accessible to all’ (Para. 2; italics added). A wider notion of freedom was also implicit in the characterization of security measures as responding to ‘the threat serious crime poses to people's freedom and legal rights’ (Para. 6).

Fundamental rights are an obvious point of reference for elaborating a more comprehensive notion of freedom, reaching beyond mere freedom of movement and abandoning the rationale of the economic constitution. All the AFSJ programmes have at least hinted at a link between the freedom dimension of the AFSJ and fundamental rights, although they have been rather indecisive as regards the significance of this link for policy measures. An Annex to the Tampere Conclusions defined the composition and working methods of the Convention which was set up to draft the EU Charter of Fundamental Rights. In turn, the Hague Programme put improving ‘the common capability of the Union and its Member States to guarantee fundamental rights’ high on the list of objectives included in the introductory section. In the section on ‘General orientations’, it called for full respect for ‘fundamental rights, as guaranteed by the European Convention on Human Rights and the Charter of Fundamental Rights in Part II of the Constitutional Treaty and the explanatory notes, as well as the Geneva Convention on Refugees’. It was recalled that ‘incorporating the Charter into the Constitutional Treaty and accession to the European Convention for the protection of human rights and fundamental freedoms will place the Union, including its institutions, under a legal obligation to ensure that in all its areas of activity, fundamental rights are not only respected but also actively promoted’. Furthermore, the Hague Programme reaffirmed the ‘firm commitment to oppose any form of racism, antisemitism and xenophobia’, expressed by the European Council in December 2003, and welcomed ‘the Commission's communication on the extension of the mandate of the European Monitoring Centre on Racism and Xenophobia towards a Human Rights Agency’ (Para. II.2).

At least at a justificatory level, the post-Lisbon Stockholm Programme has taken the most far-reaching steps in the direction of detaching ‘freedom’ from its narrow free movement connotation and defining it by reference to fundamental rights and citizenship. Promoting citizenship and fundamental rights heads the list of political priorities in the introductory section of the Programme: ‘European citizenship must become a tangible reality’, and ‘the area of freedom, security and justice must, above all, be a single area in which fundamental rights and freedoms are protected’. Accordingly, the freedom-related section of the Programme bears the title ‘Promoting citizens’ rights: a Europe of rights’. Top priority is accorded to rapid EU accession to the European Convention on Human Rights, which ‘will reinforce the obligation of the Union, including its institutions, to ensure that in all its areas of activity, fundamental rights and freedoms are actively promoted’. This will facilitate the continuing development in step of the case law of the ECJ and the European Court of Human Rights, ‘reinforcing the creation of a uniform European fundamental and human rights system based on the European Convention and those set out in the Charter of Fundamental Rights of the European Union’. The programme invites Union institutions and Member States to ensure that throughout the legislative process, legal initiatives will be consistent with fundamental rights and freedoms. For this purpose, Union institutions are called on to make full use of the European Union Agency for Fundamental Rights (Para. 2.1). The measures envisaged pertain not only to full exercise of the right to free movement but also to respect for diversity and protection of the most vulnerable; rights of the individual in criminal proceedings; protection of citizens’ rights in the information society; participation in the democratic life of the Union; and entitlement to protection in the information society (Paras. 2.2–7). Migration, asylum and borders are no longer subsumed under the freedom-related section, but external borders and visa policy are dealt with as a sub-section (‘Access to Europe in a globalised world’) in the security section (‘Europe that protects’), while migration and asylum have been dedicated a section of their own (‘A Europe of responsibility, solidarity and partnership in migration and asylum matters’). This rearrangement and renaming testifies to an effort to moderate the security emphasis of previous AFSJ programmes.

The Stockholm Programme as a whole can be seen as a response to criticism of the AFSJ, where main concerns have included ‘securitization’ of the area and lack of democratic and judicial accountability. This goes a long way towards accounting for the attempt to define the freedom aspect through fundamental rights and citizenship, and to sever its umbilical cord to the economic constitution and concomitant cross-border security concerns. Still, the AFSJ Title of the TFEU includes no reference to the EU Charter, nor are there any mutual references between the AFSJ Title and the citizenship provisions of the TFEU.Footnote ⁶¹

To conclude, a high degree of ambiguity permeates the discussion of freedom in the AFSJ programmes. This is not merely manifested by the alternation of a narrower, free movement-related understanding and a wider, fundamental rights and citizen-related notion, but also, and perhaps even more significantly, by the definition of the relationship of freedom to security. Three alternatives are conceivable among which the programmes do not make a clear choice. Firstly, freedom can be seen as a source of security problems, as is the case when ‘freedom’ is primarily defined as free movement of persons. Secondly, it can be understood as an object of protection, as is the case when terrorism and criminality are seen as threatening exercise of citizenship or fundamental rights or basic Union values, at present enshrined in Art. 2 TEU. And, thirdly, freedom in terms of fundamental rights can be defined as a side-constraint which security measures must heed. In Tampere, clear emphasis was on the first alternative: free movement as a source of security problems. The subsequent programmes entailed a shift in the direction of the second and third alternatives without, however, clearly distinguishing between the two.Footnote ⁶² The ambivalence in defining the freedom element and specifying its relation to security may be seen as a token of a basic tension within the constitutional theory of the security constitution: the tension between security and fundamental rights, which will be discussed in more detail in the following section.

Finally, we have the concept of justice. How is ‘justice’ defined and what is its relation to ‘security’? Justice has not been discussed as a substantive value or principle but, rather, as a procedural and institutional question. In all three AFSJ programmes, it has quite consistently been understood in terms of judicial cooperation in cases with a cross-border dimension, under the continuing existence of national judicial systems with their territorially delineated jurisdictions. The Tampere Conclusions proclaimed that in a ‘genuine European Area of Justice individuals and businesses should not be prevented or discouraged from exercising their rights by the incompatibility or complexity of legal and administrative systems in the Member States’ (Para. 28). Mutual recognition, already introduced in Tampere, is defined in the Stockholm Programme as the cornerstone of judicial cooperation in both civil and criminal matters. Legislative harmonization is invoked as a means to achieve a European Area of Justice too, but secondary to the principle of mutual recognition.

Especially in judicial cooperation in criminal matters, the policy field of justice maintains a close link with security. Measures which according to Art. 67(3) TFEU aim to ensure a high level of security include ‘coordination and cooperation between police and judicial authorities and other competent authorities, as well as…the mutual recognition of judgments in criminal matters and, if necessary, the approximation of criminal laws’. The Tampere Conclusions made explicit the connection between justice and the security concerns raised by freedom:

Statements like this lend credence to the claim that ‘security’ is the uniting element in the conceptual cluster of freedom, security and justice.Footnote ⁶³ Yet conceding this, as well as the primacy accorded to the fight against terrorism and other security concerns particularly after 9/11, we should register the – at least programmatic – efforts to conceive of freedom with reference to fundamental rights and citizenship and to acknowledge the importance of respecting fundamental rights in security measures. Moreover, ‘justice’ includes a private law aspect, which is not reducible to security but must be linked to access to justice and, hence, to fundamental rights. Indeed, considerations of access to justice balance the security emphasis, as is conspicuous in the above citation from the Tampere Conclusions.

Security and fundamental rights

The intricate process of constitutionalization, the ever-resurfacing state-sovereigntist concerns, as well as consequent intergovernmental features, lingering reliance on international law instruments and a complicated system of opt-outs and opt-ins, reflect the exceptionally high tension between the transnational, European and the national, Member State level within the security dimension. Let us now turn to the other major tension labelling the security dimension: the tension between security and fundamental rights.

Security possesses both an individual and a social dimension: risks and security threats can be approached from the perspective of both individuals and collectives. The individual and social dimensions do not, of course, exist independently from each other: the security of an individual is socially conditioned, and collectives are composed of individuals (although not reducible to them). Yet, especially in constitutional doctrine, the distinction between individual and collective or social security is important. In constitutional analysis, security can be treated as either an individual right or as a collective good. Rights must be protected, collective goods produced through public policies. If security is guaranteed as a fundamental right, it is an individual right, in line with other fundamental liberty rights. By contrast, in its collective aspect security relates to a factual state of society, brought about through policy measures. The liberal Rechtsstaat tradition, in particular, with its emphasis on the restrictive constitutional function, places collective security – public order and safety (Allgemeine Ordnung und Sicherheit), as the term goes – outside the normative sphere of fundamental rights. Security is an extra-constitutional collective good, and one of the central tasks of the constitution is to discipline the exercise of power by police and other security authorities expected to procure the collective good of security. Fundamental rights are supposed to function as side-constraints to security measures and to protect individuals against the security threat posed by uncontrolled security power.

The doctrine of balancing, one of the great global successes of twentieth-century constitutional theory and practice, already signifies a retreat from the liberal Rechtsstaat view. As we know, much of the balancing talk has been about liberty and security; about justifying restrictions to liberty in the name of guaranteeing security. Here we encounter one of the paradoxes of constitutional doctrine, evoked at the outset of this chapter: extra-constitutional collective security interests, which individual fundamental rights are supposed to curtail, (re-)enter constitutional doctrine as grounds for limiting those very rights. As Jeremy Waldron, among others, has pointed out, many problems afflict the idea of balancing, both in general and in the particular context of freedom and security.Footnote ⁶⁴ Yet balancing belongs to the standard package of contemporary constitutional doctrine and is implicit in the limitation clauses of Arts. 5–11 ECHR, for instance. What is important for our present discussion is that the idea of balancing does not necessarily entail effacing the distinction between individual fundamental rights and security as a collective good. It still makes sense to discuss within the balancing paradigm, for instance, whether the personal freedom or physical integrity of a suspect or detainee may be restricted in order to enhance the collective security of society.

In recent decades, a tendency to blur in a problematic manner the distinction between individual fundamental rights and collective goods, such as security, has made itself noticeable. The Finnish example is instructive of a more general trend. In Finnish constitutional argumentation at the turn of this century, security as a collective good was no longer treated as an extra-constitutional societal interest which could justify restrictions on fundamental rights; collective security itself was transformed into a fundamental right. This reading of the Constitution was achieved through three argumentative moves.

Firstly, the right to security, explicitly enshrined in the Constitution, was detached from rights to personal liberty and integrity, confirmed in the same provision, and assigned independent legal status. This already points to elevating security as a collective good to a fundamental right; it is difficult to see what independent legal meaning, not included in the rights to liberty and integrity, could be attributed to an individual right to personal security. The second step in the argument consisted of an appeal to the doctrine of horizontal effect (Drittwirkung) of fundamental rights; fundamental rights pertain not only to vertical relations between public authorities and private individuals but also to horizontal relations among the latter. Finally, reference was made to the constitutional provision which establishes a general obligation for the state to ensure the realization of fundamental and human rights; this corresponds to the German Schutzpflicht (duty of protection) doctrine. Taken together, these arguments – the independent nature of the right to security, the horizontal effect of fundamental rights and the Schutzpflicht of the state – entail the obligation of the state to guarantee realization of the right to security in horizontal relations between private individuals. Security as a collective good has been converted into a fundamental right with a rank equal to that of individual liberty rights. By the same token, balancing between liberty and security is redefined as balancing between diverse fundamental rights. This move obliterates from constitutional doctrine the Dworkinian distinction between policies aiming at collective goods and principles focusing on individual rights, with the concomitant prima facie primacy of the latter.Footnote ⁶⁵

Such an understanding of security allows for fundamental rights justification for, say, ever-new police powers: these powers can always be argued to be necessary for fulfilling the obligation of the state to guarantee the fundamental right to security. If and when these powers encroach on individual fundamental rights, such as rights to personal integrity, privacy and confidentiality of correspondence, their constitutional acceptability is claimed to depend on weighing fundamental rights principles of equal rank. What according to the liberal Rechtsstaat view was to be restricted through the system of fundamental rights has been absorbed into the system as one of its constituent elements. The influence of German doctrine on Finnish development was conspicuous. Indeed, a similar development was already discernible in Germany in the 1980s, when the German Constitutional Court, supported by such academics as Robert Alexy,Footnote ⁶⁶ elaborated its doctrines of fundamental rights as general legal principles, the Dritt-wirkung of these rights and principles and the Schutzpflicht of the state. Critics, such as Ingeborg Maus and Oliver Lepsius, have pointed to the ensuing danger of weakening the protection of individual liberty rights. Lepsius has drawn attention to the tendency to understand even liberty as a general state of society, rather than in terms of individual liberty rights. Hence balancing tends to be perceived as a comparative weighing up of two kinds of collective good!Footnote ⁶⁷

Evidently, in EU law and complementary programmatic documents ‘security’ has primarily been employed in its collective sense. In ECJ case law, balancing has been resorted to but without invoking collective security as a fundamental right; balancing has occurred between extra-constitutional security interests and individual fundamental rights. To my knowledge, no attempt has been made either by EU institutions or academic writers to interpret the right to security of person, enshrined in Art. 6 CFREU, in terms of a collective good.Footnote ⁶⁸ As Jörg Monar has noted, the provisions of the AFSJ Title of the TFEU do not establish justiciable rights for citizens but provide for furnishing them public goods; European citizens are not conceived of as (active) rights-holders but as (passive) recipients of public goods. On Monar's reading, the wording of Art. 3(2) TEU is symptomatic enough: ‘The Union shall offer its citizens an area of freedom, security and justice…’Footnote ⁶⁹

In programmatic contexts, both citizenship and rights vocabulary has been employed. The Tampere Conclusions, the first comprehensive AFSJ programme adopted by the European Council, declared that ‘people have the right to expect the Union to address the threat to their freedom and legal rights posed by serious crime’ (Para. 6, italics added),Footnote ⁷⁰ and the Internal Security Strategy of 2007 even uses the term ‘basic right’: ‘People in Europe expect to live in security and to enjoy their freedoms: security is in itself a basic right.’Footnote ⁷¹ AFSJ programmes and strategies even hint at the protective duty of the Union. This message is conspicuous in the Preamble to the Anti-Terrorism Framework Decision, where terrorism is characterized as the most serious violation of the universal values on which the Union is founded and which include respect for human rights and fundamental freedoms.

The expansion and constitutionalization of the security dimension has largely coincided with the development of Union fundamental rights law. Maastricht both introduced the security-related pillars and confirmed the status of fundamental rights as principles of Community law. In turn, post-Amsterdam privileging of security issues overlaps the drafting and acceptance of the EU Charter, and, finally, the acknowledgement of its constitutional rank in Lisbon. Still, until Lisbon the effectiveness of fundamental rights constraints on EU security measures was hampered by the limited competence of the ECJ within the third pillar.

In constitutional adjudication, Union security measures have been weighed against fundamental rights not only by the ECJ but by national constitutional courts as well. The pluralism of fundamental rights law in Europe, discussed above in Chapter 3, creates an interface for the two basic tensions we have identified in the constitutional theory of the security dimension: between, on the one hand, European and national competences and, on the other hand, security and fundamental rights. Member States have been keen to ensure that Union security measures meet the criteria of national fundamental rights law too. Implementation of the EAW Framework Decision raised fundamental rights concerns in several Member States. These concerns were related to extradition of own nationals; the principle of legality; respect for ne bis in idem and other procedural rights of defendants; as well as confidence in fundamental rights protection in other Member States which the principle of mutual recognition presupposes. Indeed, during implementation of the EAW Framework Decision, several constitutional courts profited from the opportunity to reconfirm their reservations on the primacy of EU law. Even in applying national provisions implementing the Framework Decision, national constitutional courts have tried to defend the autonomy of their fundamental rights law.

On a referral by the Belgian Constitutional Court (Cour d'arbitrage), the ECJ examined the EAW Framework Decision from the perspective of Union constitutional law and found no violation of the fundamental rights principles of legality or equality.Footnote ⁷² AFSJ issues have also figured prominently in ECJ post-Lisbon case law invoking the CFREU with its newly acquired constitutional status. In addition to balancing security and fundamental rights, the case law has touched on the respective jurisdiction of EU and national fundamental rights law, and, through the principle of mutual recognition, the horizontal relationship between fundamental rights protection in different Member States. Here the security dimension has contributed to clarifying issues of larger significance within the juridical constitution.

In Melloni, the main legal issue was whether a Member State could refuse to execute a European arrest warrant ‘in order to avoid an adverse effect on the right to a fair trial and the rights of the defence guaranteed by its constitution’. As explained in Chapter 3, the Court invoked the primacy, unity and effectiveness of EU law, and rejected a reading of Art. 53 CFREUFootnote ⁷³ which would authorize Member States ‘to apply the standard of protection of fundamental rights guaranteed by its constitution when that standard is higher than that deriving from the Charter and, where necessary, to give it priority over the application of provisions of EU law’.Footnote ⁷⁴ Melloni might be labelled a special case, due to the fact that the EAW Framework Decision includes explicit provisions on the Member State right to refuse a European arrest warrant and that these were expressly designed to heed fair trial concerns. However, the Court formulated its reading of Art. 53 CFREU in general terms, with the clear intention of articulating a principle with general applicability. Daniel Sarmiento has portrayed reference to the primacy, unity and effectiveness of EU law as an exceptional remedy; a last-resort safety measure for extraordinary situations.Footnote ⁷⁵ Yet the way Melloni defines the primacy of the principle of primacy with regard to the Member State right to enforce higher national standards appears to deprive Art. 53 CFREU of much of its relevance. Situations addressed by this article are by definition situations of overlapping jurisdiction where the principle of primacy is relevant.

If Melloni is important in specifying the relationship between Union and national fundamental rights law from the perspective of EU constitutional law, NS,Footnote ⁷⁶ which deals with asylum law, addresses the horizontal relationship between fundamental rights protection in different Member States. In its judgment, the Court first stated that the discretionary power which the Dublin II Regulation on the Common European Asylum SystemFootnote ⁷⁷ leaves to Member States forms an integral part of this System. When a Member State exercises this power and decides whether or not to examine an application which is not its responsibility under the Regulation, it implements Union law and, consequently, EU fundamental rights law applies (Paras. 65–8).Footnote ⁷⁸ The Court further held that Member States have a duty not to rely on an interpretation of EU secondary legislation which would contradict EU fundamental rights or other general principles of EU law (Para. 77). The Common European Asylum System is built on the principle of mutual confidence. The underlying assumption is that participating States observe fundamental rights and that the Member States can have confidence in each other in this regard (Para. 78).

However, the assumption that asylum seekers are treated in all Member States in compliance with the requirements of the CFREU, the Geneva Convention and the ECHR is rebuttable. The system ‘may, in practice, experience major operational problems in a given Member State, meaning that there is a substantial risk that asylum seekers may, when transferred to that Member State, be treated in a manner incompatible with their fundamental rights’ (Paras. 80–1). Still, not any infringement of a fundamental right by the Member State responsible suffices to prevent the transfer of an asylum seeker. Only ‘if there are substantial grounds for believing that there are systemic flaws in the asylum procedure and reception conditions for asylum applicants in the Member State responsible, resulting in inhuman or degrading treatment, within the meaning of Article 4 of the Charter, of asylum seekers transferred to the territory of that Member State, the transfer would be incompatible with that provision’ (Para. 86). If such grounds exist, EU fundamental rights law obliges a Member State not to transfer the asylum seeker to the Member State responsible.Footnote ⁷⁹

Thus, according to NS, EU fundamental rights law imposes on Member States, including their courts, an obligation to monitor compliance with that law in other Member States. Yet the basic, though rebuttable, assumption guiding Member State action is that other Member States comply with the relevant fundamental rights; hence the term ‘horizontal Solange’, used by Iris Canor.Footnote ⁸⁰ However, it is important to note that EU law is the applicable law with regard both to the Member State considering the transfer of an asylum seeker and to the Member State responsible. We are dealing with the application and observance of EU fundamental rights law in two Member States and not a direct horizontal relation between national fundamental rights laws. The term ‘horizontal Solange’ is justified in the sense that a Member State's authorities are monitoring the fundamental rights record of another Member State. But, actually, these authorities are exercising ‘vertical’ EU fundamental rights review.

It would be audacious to derive from NS a general doctrine of horizontal Solange and a general obligation of Member States to watch over each other's observance of EU fundamental rights. Horizontal Solange would be especially relevant for AFSJ legislation based on mutual recognition and the underlying principle of mutual confidence. In its ruling on the EAW, the German Constitutional Court took a position reminiscent of horizontal Solange. The Court ruled that ‘putting into effect a strict principle of mutual recognition, and the extensive statement of mutual confidence among the states that is connected with it, cannot restrict the constitutional guarantee of fundamental rights’. Accordingly, extradition decisions must always be based on case-by-case weighing.Footnote ⁸¹ Still, here the standard would be national German but not Union fundamental rights law. In its long dialogue with the German Constitutional Court, the ECJ has never accepted the fundamental rights qualification to the primacy of EU law which the German Court has repeatedly asserted. But nor does ECJ case law provide clear support even for a horizontal Solange where the normative yardstick would be provided by EU fundamental rights law. As regards, for instance, the European arrest warrant, the ECJ has not invoked a general doctrine but relied on the explicit provisions of the Framework Decision on mandatory or optional non-execution of such a warrant.Footnote ⁸² However, in the absence of such provisions, a horizontal Solange-type argument might be possible.

Kadi, issued in 2008, remains the most pivotal ECJ ruling on the constraints that EU fundamental rights impose on security measures. Kadi possesses many legally significant aspects, one of them being the relationship of EU law with international law. As explained in Chapter 2, Kadi reasserted the autonomy of EU law: the fact that the Union measures under scrutiny implemented UN Security Council decisions did not affect the jurisdiction of the Court to assess them by Union fundamental rights law. The Court argued that ‘the obligations imposed by an international agreement cannot have the effect of prejudicing the constitutional principles of the EC Treaty, which include the principle that all Community acts must respect fundamental rights, that respect constituting a condition of their lawfulness which it is for the Court to review in the framework of the complete system of legal remedies established by the Treaty’ (Para. 285). According to the Court, the Treaty provision on the validity of pre-Membership international obligations (now Art. 351 TFEU) ‘may in no circumstances permit any challenge to the principles that form part of the very foundations of the Community legal order, one of which is the protection of fundamental rights, including the review by the Community judicature of the lawfulness of Community measures as regards their consistency with those fundamental rights’ (Para. 304). Furthermore, ‘the review by the Court of the validity of any Community measure in the light of fundamental rights must be considered to be the expression, in a community based on the rule of law, of a constitutional guarantee stemming from the EC Treaty as an autonomous legal system which is not to be prejudiced by an international agreement’ (Para. 316).Footnote ⁸³

In Kadi, the Court also took a position on the extent to which security considerations justify restricting EU fundamental rights. It rejected the view that ‘restrictive measures such as those imposed by the contested regulation escape all review by the Community judicature once it has been claimed that the act laying them down concerns national security and terrorism’. With reference to ECtHR case law,Footnote ⁸⁴ the Court held it to be ‘the task of the Community judicature to apply, in the course of the judicial review it carries out, techniques which accommodate, on the one hand, legitimate security concerns about the nature and sources of information taken into account in the adoption of the act concerned and, on the other, the need to accord the individual a sufficient measure of procedural justice’ (Paras. 343–4). Unfortunately, the Court was not more specific as to what the techniques for accommodating security and fundamental rights concerns are.

Kadi belongs to the pre-Lisbon era when the Charter had not yet been acknowledged as enjoying constitutional status and the Court refrained from invoking it. In contrast, Digital Rights Ireland examined the compatibility of the Data Retention DirectiveFootnote ⁸⁵ with the right to respect for private and family life, guaranteed in Art. 7 of the Charter, and the right to protection of personal data, guaranteed in Art. 8, and assessed whether the Directive met the conditions that the Charter lays down for limiting Charter rights and freedoms. In contrast to the European Convention on Human Rights, the Charter includes a general limitation clause. Limitations must be provided for by law and respect the essence of rights and freedoms. They must also adhere to the principle of proportionality, so that they ‘may be made only if they are necessary and genuinely meet objectives of general interest recognised by the Union or the need to protect the rights and freedoms of others’ (Art. 52(1) CFREU).

In Digital Rights Ireland, the Court applied the three-phase reasoning, analogous to that employed when assessing restrictions on free movement (see Chapter 5), which it has adopted in fundamental rights review. In the first phase, the existence of a restriction is considered, in the second phase a position is taken on justification of an identified restriction and in the third phase the restriction is subjected to a proportionality test. The Court found that retention of data relating to a person's private life and communications in itself constituted interference with the Charter right to respect for private and family life and that access of competent national authorities to the data constituted a further interference. Because the Directive provided for processing personal data it also constituted interference with the fundamental right to protection of personal data (Paras. 34–6). The Court noted that the material objective of the Directive is to contribute to the fight against serious crime and, ultimately, to public security, and stated, referring to Kadi and Tsakouridis,Footnote ⁸⁶ that the fight against terrorism and serious crime constitutes an ‘objective of general interest’ for the purposes of the limitation clause in Art. 52(1) CFREU (Paras. 41–2). In the proportionality appraisal, the Court found that retention of data covered by the Directive may be considered to be appropriate for attaining the objective of the Directive. However, the Directive did not confine the restrictions on Charter rights to what is strictly necessary, as required by Art. 52(1). The Court pointed out that the Directive applies to all means of electronic communication and covers all subscribers and registered users, and, consequently, ‘entails an interference with the fundamental rights of practically the entire European population’. No differentiation, limitation or exception is made in the light of the objective of combating serious crime. Application of the Directive does not require any evidence that persons at issue might have a link with serious crime or that data are related to a threat to public security. Furthermore, the Directive does not set out an objective criterion by which limits on access by the authorities to data and their use for preventing, detecting or prosecuting offences could be determined, nor does it contain substantive or procedural conditions for such access and use. The Court specifically stressed that

Digital Rights Ireland (2014) may prove to be a milestone in asserting fundamental rights as a restriction on security-motivated measures; and not only in the EU, but even more generally.

Constitution or anti-constitution?

As we have seen, ‘security’ is a diffuse, expansionist and border-breaking concept. It is no wonder that it is difficult to mark the exact boundaries of the security constitution. In this chapter I have identified the security dimension with internal security, addressed at the Treaty level, first, by the third-pillar provisions of the Maastricht Treaty and, finally, after Lisbon, by the AFSJ of the TFEU. But one of the boundaries which the concept of security informing European constitutional law and strategy documents tends to break is that separating internal from external security. Hence, my choice to focus on internal security and exclude Common Foreign and Security Policy from the discussion might be questioned. Or, to take another example: should we, at the level of Treaty law, include in the security constitution the whole AFSJ or just those fields where the security emphasis is prominent? I do not deny the relevance of such questions. But, what perhaps is most important is to note the fuzziness of the boundaries of the security constitution, at the poles of both constitutional law and constitutional object. Indeed, related to the diffuse, expansionist and border-breaking character of ‘security’, the security constitution does not possess clear-cut contours either.

Another distinctive feature is the combination of transnationalism and intergovernmentalism which characterizes institutional and procedural aspects of the security constitution. A central criterion of the constitutionalization of the security dimension consists of its advancing institutional and procedural transnationalization. Starting from the pre-Maastricht situation where the security dimension was not addressed at Treaty level at all, and ending up with ‘communitarization’ of the AFSJ in Lisbon, transnationalization has made considerable headway. The powers of the three main transnational institutions – the Commission, the European Parliament and the ECJ – now reach to all AFSJ fields, and the establishment of EU security agencies, such as Europol, Eurojust, Frontex and EASO, has complemented the institutional aspect of the security constitution. But intergovernmentalism has proven to be an ineradicable trait in this constitutional dimension. The Commission has obtained the power of initiative, but a diluted form of Member State initiative still exists too. A quarter of Member States have a power of initiative parallel to that of the Commission in police cooperation and judicial cooperation in criminal matters. In turn, the special legislative procedure called for in several AFSJ provisions includes unanimity in the Council but mere consultation of, or consent by, the European Parliament. Furthermore, EU security agencies do not have operational law-enforcement competences, nor do they wield public power in Member States. Their main function is to support Member State authorities and to provide a site for administrative and professional cooperation, as well as for information sharing and data processing. And, finally, as regards the justice aspect of the AFSJ, mutual recognition as the main option is an intergovernmental alternative to transnational legislative harmonization. All this lends credence to Jörg Monar's point that cooperation rather than integration remains the prevailing principle. This sets the AFSJ aside from domains falling under, say, the microeconomic constitution.Footnote ⁸⁷

The reason for lingering intergovernmentalism and the continuing dominance of the principle of cooperation should be evident by now. The security constitution touches on, and is prone to collide with, traditional core attributes of state sovereignty. These attributes are related to the Weberian characteristics of the modern state as possessor of a monopoly on legitimate violence. Often enough, they are protected by basic constitutional principles which, according to the constitutional doctrine of Member States, pose insurmountable hurdles to additional transfers of sovereignty rights. Transnational legislative and operational competence in criminal law, policing and anti-terrorist measures also touches on individual fundamental rights, and here too collisions are possible with basic constitutional principles of Member States, even constitutional identity as defined by national constitutional actors. In this respect, the saga of the EAW Framework Decision and the constitutional assessment of its legislative implementation in Member States is instructive enough.

Sovereign power is the main attribute in the traditional three-element concept of the state, espoused, for instance, by the German late constitutionalist school of Carl Friedrich von Gerber, Paul Laband and Georg Jellinek, the other two being population and territory. Member States have been quite jealous in guarding their control over the latter two elements too. Transnationalization has, though, progressed further in the domain of migration, asylum and borders than in criminal law and law enforcement. Yet Member States have retained their power to determine volumes of migration from third countries to their territory, as well as the geographical demarcation of their borders.

State-sovereigntist concerns are more pronounced in some Member States than others, and national constitutional doctrines of core constitutional principles defining national constitutional identity and enjoying protection against transnational intrusion vary as well. This is a central backdrop to a further peculiarity of the security constitution: its mix of coherence and fragmentation. The concept of security creates the necessary coherence which in general justifies talking of a European security constitution. Moreover, the Lisbon Treaty has increased unity in Treaty law by assembling AFSJ provisions under the same Title of the TFEU and by harmonizing the legislative instruments and procedures available. Furthermore, the security dimension as a whole displays characteristic operational modes. However, signs of fragmentation are conspicuous too. Ever since its pre-Maastricht origins under the auspices of international law, the security dimension has been the model example of variable geometry. The Schengen system has never – either before or after incorporation of the Schengen acquis into the EU legal framework – covered all Member States, and the opt-outs of some Member States have been complemented by the opt-ins of some non-member Schengen associates. The Amsterdam Treaty introduced the objective of establishing the AFSJ and led to replacing conventions as a legal instrument with framework decisions. But as a counterweight, through its provisions on closer (enhanced) cooperation, it (re-)constitutionalized variable geometry within the AFSJ. And as was proved by the Prüm Convention, signed in 2005 by seven Member States,Footnote ⁸⁸ neither did Amsterdam spell the end of resort to international treaties as a means to push security cooperation further in spite of the sovereigntist hesitations of some Member States. The Lisbon Treaty only reaffirmed and even added to the existing fragmentation with its complicated system of opt-outs and opt-ins, as well as provisions on emergency brakes and enhanced cooperation. If the AFSJ has managed to obtain substantive coherence as a policy area, in geographical terms no unitary area of freedom, security and justice exists.

The emphasis on cooperation instead of integration – to use Monar's terms – points to constitutionally relevant specificities in the modes of operation under the security constitution. Cooperation is the prevalent mode of operation of the executive: of administrations, civil servants and professionals. A major part of AFSJ Treaty provisions deals with such cooperation, and, as we have seen, in many cases Treaty provisions have only added a formal constitutional blessing to developments already achieved. The AFSJ is the promised land of networks connecting security officials and professionals, and even AFSJ agencies, some of them – like Europol and Eurojust – provided for in the Treaty, tend to boil down to such networks. Associated with administrative and professional cooperation, information exchange and data processing, necessary for identifying and assessing security risks, are vital for the operational modus of the AFSJ, raising constitutional concerns especially in respect of data protection and the right to privacy. It remains to be seen what wider significance Digital Rights Ireland will have for the security constitution.

Since its pre-Maastricht beginnings in Trevi and Schengen, the security dimension has been accompanied by allegations of lack of transparency and democratic and judicial accountability. Constitutionalization since Maastricht has not been able to quash this criticism. The post-Lisbon Stockholm Programme bears witness to its continuing pertinence. The European Council refers to ‘the increased role that the European Parliament and National Parliaments will play following the entry into force of the Lisbon Treaty and to the greater opportunity Art. 11 TEU provides for citizens and associations to make known and publicly exchange their views in all areas of Union action’, and expresses its confidence that ‘this will reinforce the open and democratic character of the Union for the benefit of its people’. As regards the AFSJ specifically, reference is made to the co-decision procedure (the ordinary legislative procedure), which grants ‘full involvement’ of the European Parliament, while the role of the National Parliaments in the legislative procedure is invoked as well. The European Council also points to the enlarged jurisdiction of the ECJ (Para. 1).

Yet it is doubtful whether the Lisbon Treaty or the assurances of the Stockholm Programme will be able to efface concerns about transparency, democracy and accountability. The emphasis on the executive instead of the legislative, the confidential nature of data processed in the security dimension and the scarcity of decisions with direct legal relevance and within the reach of legal remedies pose formidable obstacles to realizing these values even at the state level. In the AFSJ of the EU, with its primacy of administrative and professional cooperation at the expense of legislative integration, the obstacles are considerably higher still. Democratic legitimacy and the rule of law, including judicial review of executive action, belong to the major promises made by modern constitutions, characterizing constitutionalism in its thick normative sense. In these respects, the security constitution of the EU does not fare well, although some undeniable progress has been made on the road from pre-Maastricht to post-Lisbon.

Let us wrap up this exploration of the peculiarities of the security constitution with some comments on the position of the individual. A democratic Rechtsstaat, as envisaged by basic modern constitutional principles, such as fundamental rights, the rule of law and democracy, is premised on a conception of individuals as autonomous and responsible persons. European constitutionalization has been a process not only of transnationalization but also of individualization, of connecting transnational institutions directly with European individuals. In this respect, constitutionalization and its expansion to new dimensions have entailed step-by-step building up of European citizenship. The economic constitution, where it all began, treats individuals as economic actors in the marketplace and endows them with market citizenship; juridical constitutionalization, proceeding through such celebrated ECJ rulings as Van Gend en Loos and Costa v Enel, added to citizenship a judicial layer, while political constitutionalization generated political citizenship. Even the embryonic social constitution, the constitutional underdog in the EU context, gives rise to at least a thin social citizenship.

Here the security constitution amounts to an exception. Although both Art. 3(2) TEU and the AFSJ programmes have flirted with citizenship terminology, the security dimension does not furnish European citizens with new rights that would further enhance their citizenship. In the mirror of the security constitution, individuals are passive recipients of collective security goods rather than active citizens or bearers of rights. Individuals enter the focus of security measures primarily as security risks whose characteristics, propensities and actions must be surveyed and recorded. As objects of surveillance and data processing individuals are treated as replaceable members of a group – a population, as Foucault would put it – rather than as autonomous and responsible, both unique and equal persons, as conceiving of them as bearers of rights and citizens of a democratic polity would presuppose. In this sense, the security dimension is a domain of de-individualization, rather than individualization!Footnote ⁸⁹

The security dimension comes up with new divisions and classifications of individuals which do not return to individuals their unique personhood but do counteract the principle of equality. As Samuli Hurri has argued, the ECJ's ordre public rulings in the 1970s gave birth to a ‘European individual’, with a juridical locus in the grey zone between the law of nationals and the law of aliens.Footnote ⁹⁰ With the development of common policies on visas and migration, asylum and borders, new categories of individuals with a legal status different from full-blown citizens and, often enough, objects of diversified systems of information and surveillance have abounded: third-country nationals and illegal immigrants, refugees and asylum seekers, to name some of the most obvious ones. The security constitution appears to have not only de-individualizing but also un-equalizing consequences.

‘Security’ has largely replaced ‘emergency’ and ‘state of exception’ as the principal argumentative device in the domain of the security constitution. European Treaty law hardly mentions emergencies. In institutional practice, though, emergency talk has retained at least part of its justificatory significance, as has recently been proven by the framing and handling of the Eurozone crisis.Footnote ⁹¹ At the national level, the security dimension is perhaps the primary battlefield between constitutive and restrictive constitutional functions. Within the EU too, the security constitution poses grave fundamental rights problems, which are further aggravated by its de-individualizing and un-equalizing tendencies.

Even without the constitutional black or grey hole of a state of exception, the security constitution displays conspicuous features of an anti-constitution. This is, if you wish, the ultimate paradox of the European security constitution.