International institutions and the state monopoly of force
In recent years, a growing literature has argued that Western states have undergone a profound transformation due to the increasing number and strength of international institutions, most notably the EU. This thesis has been supported with many indicators and factors describing and explaining the hypothesized change (e.g. Strange, Reference Strange1996; Held et al., Reference Held, McGrew, Goldblatt and Perraton1999; Goldmann, Reference Goldmann2001; Sørensen, Reference Sørensen2004).Footnote 1 Given the overwhelming importance of the state as an actor, as an analytical concept, and as a normative reference point (Lake, Reference Lake2008), this concern is not surprising. What is surprising, however, is that this literature has largely spared the state monopoly of force.
This oversight is unfortunate because a state exists ‘insofar as its administrative staff successfully upholds the claim to the monopoly of the legitimate use of physical force in the enforcement of its order’ (Weber, Reference Weber1978 [1922]: 54, emphasis in the original). Many authors have followed Weber's famous definition or at least included the monopoly of forceFootnote 2 in their own concepts (e.g. Mann, Reference Mann1987; Poggi, Reference Poggi1990; Levi, Reference Levi2003). More concretely, the state monopoly of force has external and internal dimensions which are represented by the military and the police, respectively. As only the police are instrumental for the enforcement of the state's political order which is so central to Weber's argument, we will focus on it here.
While there is broad evidence that international institutions and the EU, in particular, have become much stronger over the last decades and that the state has changed from a monopolist to a manager of political authority in many areas (Genschel and Zangl, Reference Genschel and Zangl2008), we still lack this evidence for its core function. Although the growing literature on international criminal law, international police cooperation, and most notably on the EU's emerging ‘area of freedom, security, and justice’ (e.g. Andreas and Nadelmann, Reference Andreas and Nadelmann2006; Peers, Reference Peers2007; Cassese, Reference Cassese2009) seem to indicate a strengthening of international institutions in the field of policing parallel to their strengthening in other areas, we cannot simply assume that the state will give up control to international institutions or to the EU in its core area. Instead, as Stanley Hoffmann once remarked, the state might resemble an artichoke: Although international institutions could take away the outer leaves without much resistance, the state would defend its inner core (Hoffmann, Reference Hoffmann1966: 883–884).
Against this backdrop, we can formulate two competing hypotheses:
H1: As it is their core sovereign power, states will protect their monopoly of force from external influence and attempt to retain as much domestic policy autonomy as possible. International police cooperation will be sovereignty-driven and remain largely informal or ad hoc, with weak institutions, and exchanges taking place mainly between police officers. This hypothesis is drawn from the literature on European and international police cooperation (e.g. Bigo, Reference Bigo1996; Busch, Reference Busch1999; Deflem, Reference Deflem2002; Sheptycki, Reference Sheptycki2002; Andreas and Nadelmann, Reference Andreas and Nadelmann2006).
H2: States do not aim to protect a ‘core’ but create institutions on a sectoral basis in order to make credible commitments, resolve disputes, or create policy bias. International police cooperation will be efficiency-driven like any other policy and may lead to strong institutions, the involvement of central executives, or binding law. Both pooling (decision-making by procedures other than unanimity) and even delegation to independent institutions is possible. This hypothesis is drawn from the liberal theory of international relations and from theories of delegation in European and international politics (Moravcsik, Reference Moravcsik1998; Hawkins et al., Reference Hawkins2006; Franchino, Reference Franchino2007).
There are thus different expectations regarding the strength of international institutions in the field of police cooperation. According to the first hypothesis, they will be weak. According to the second, they can also be strong. In order to find out how far international institutions impinge upon the state monopoly of force, we will in the following section measure their strength in the field of policing. Strong institutions give prescriptive status to the norms they ban, increase the cost of non-compliance, and shape the opportunities for state action in the future. A parallel argument in the debate on the legalization of world politics is that higher degrees of legalization are more constraining on states than lower degrees of legalization (Abbott et al., Reference Abbott, Moravcsik, Slaughter and Snidal2000). Strong international institutions are not primarily important because their prescriptions could be used against the resistance of states; rather, institutions shape actor preferences and interaction results (Scharpf, Reference Scharpf1997: 38–43). Thus, although institutional strength is not identical to impact or effectiveness, it is a good indicator and a necessary, if not a sufficient condition for it (Underdal, Reference Underdal2004: 29).
While ideally we would like to measure the effects of international institutions on how states actually use their monopoly of force, the literature on the effectiveness of international environmental regimes – arguably the most advanced discussion of the topic – highlights ‘unusually severe’ (Young, Reference Young2004: 11) methodological challenges. Analyses have to distinguish between specific and broader effects, provide a benchmark for assessing effects independent from the regime participants, and often have to use controversial counterfactual methodology or extensive process-tracing or case studies (Underdal, Reference Underdal2004). As a result, studies of effectiveness usually focus on single or few cases (e.g. Miles et al., Reference Miles, Underdal, Andresen, Wettestad and Skaerseth2002). In contrast, we want to depict broad patterns of institutional evolution in this field in order to cover as much as possible of the state monopoly of force.
The EU compliance literature also lends support to choosing strength as a proxy of effectiveness. One of its main empirical findings is that while there are substantive cases of non-compliance with EU law, all those cases were resolved over time (Tallberg, Reference Tallberg2002; Panke, Reference Panke2010). Thus, although frequently powerful states are against a specific rule, in the end they do what they are supposed to do, even if it takes a long time. Still, in addition to these conceptual arguments, we will provide empirical examples of the impact of those institutions, most notably on the European Arrest Warrant and the Anti-money Laundering (AML) Directive.
Research design and methods
Measuring the strength of international institutions
Since there is no generally accepted measure for the ‘strength’ of an international agreement or institution, we need to develop our own scales, drawing on attempts to code the properties of international environmental regimes (Breitmeier et al., Reference Breitmeier, Young and Zürn2006) and to measure the legalization of international politics (Abbott et al., Reference Abbott, Moravcsik, Slaughter and Snidal2000). In order to have a broad grasp, we look at the three classic functions of governance: legislation, execution, and adjudication (Zangl, Reference Zangl2006). All three functions are operationalized as ordinal scales each measuring the degrees of freedom a state has in the respective area: the higher the values, the stronger the international institution, the lower state autonomy and the higher the likelihood that individual state preferences are overridden. Table 1 provides an overview of our scales.
Measuring the strength of international institutions
The legislative function is best captured by looking at how agreements or other legal measures are adopted, that is, how decisions are made. Our scale of decision-making adapts Scharpf's discussion of forms of collective decision-making (Scharpf, Reference Scharpf1997). When states merely act strategically in reaction to the moves of other states, do not cooperate at all or react with unilateral ‘friendly behaviour’ (one state offers the other access to information but has not made a promise to keep doing so) there is no collective decision-making (value = 0). When states negotiate (value = 1), adopt measures in compulsory negotiations (value = 2), which might be followed by a vote with qualified majority (value = 3)Footnote 3 the likelihood of undesired outcomes for the state increases. Finally, hierarchy (value = 4) means that decisions are taken by somebody else who determines and controls a state's use of its monopoly of force.
For the executive function, we measure how much executive discretion states have by looking at the precision of agreements: the more precise an agreement, the more a state is bound to its rules. As this is a standard argument from the literature on the legalization of international politics, our measure is based on a scale originally developed in this debate (Abbott et al., Reference Abbott, Moravcsik, Slaughter and Snidal2000: 412–415). A value of 0 refers to very general statements that cover virtually any possible state behaviour, for example, a call for ‘appropriate measures’. Standards (value = 1) refer to wordings with general behavioural indications such as the requirement to transmit information in a ‘timely’ fashion. When more specific criteria are given to evaluate state behaviour (e.g. the specification of a time period of ‘at least 6 years’) this corresponds to ‘broad areas of discretion’ (value = 2).Footnote 4 When reference is made to precisely defined criteria, like specific lists, we use the category of ‘narrow areas of discretion’ (value = 3).
For the adjudicative function, we look at dispute settlement and measure the degree to which it is under the influence of states or independent from them. This scale is also built on a suggestion from the ‘legalization’ debate (Keohane et al., Reference Keohane, Moravcsik and Slaughter2000: 461; Zangl, Reference Zangl2006: 53–54). The higher its values, the lower the control of states over the outcome of dispute settlement. When states settle their disputes among themselves and the ‘accused’ state has a veto (value = 0), it does not have to fear much interference. When an ‘accused’ state does not have a veto, the outcome may be further away from its preference (value = 1). External adjudicators chosen by the disputants move the outcome yet further from their control (value = 2). An independent court is most remote from their control because it does not have to fear retaliation (value = 3).
Analysing changes in the state monopoly of force
An analysis of the monopoly of force requires particular attention to common conceptual pitfalls which would predetermine the empirical results. We need to avoid highly abstract concepts which leave a huge gap between theory and empirical research (Thomson, Reference Thomson1995). In addition, we should avoid using concepts in a dichotomous manner which excludes gradual variations and sets a very high threshold for change (Caporaso, Reference Caporaso2000: 3–4). Most importantly, the monopoly of force should not be reduced to the actual use of handcuffs or police truncheons. Weber already underlined the legitimacy aspect in addition to usage. For Michael Mann, more indirect ‘infrastructural power’ is at least as important as direct ‘despotic power’ (Mann, Reference Mann1987: 112–114). For these reasons, we distinguish between three dimensions of the monopoly of force (Friedrichs, Reference Friedrichs2007: 5–7):
• Legitimation concerns the definitions of the circumstances in which a state can legitimately use force. In our empirical study, this includes the definition of terrorist offenses as opposed to normal criminality and of illicit as opposed to legal use of drugs.
• Methods refer to the means by which legitimate goals in relation to the monopoly of force can be pursued. In our study, we look at the proscription or propagation of information sharing or specific investigation techniques.
• Authorization defines who has the right to allow the use of force. Although the authorities of a state usually authorize their own agents, this authorization may also come from another state or from an international institution. Empirically, we look at the European Arrest Warrant and ‘joint investigation teams’ (JITs).
All three dimensions are relevant for understanding changes in the monopoly of force because they cover different aspects of this complex phenomenon. Each can change without affecting the others. Looking at all of them prevents us from excluding potential change conceptually rather than empirically. We do not look, however, at the actual use of force beyond the state because there are no signs of a supranational police force. This does not imply that the state monopoly of force is intact and without serious challenge. When international institutions are strong in the legitimation, methods, and authorization dimensions they shape the way states can use their monopoly of force without replacing or supplementing it.
A similar phenomenon is known from EU politics in other areas. The EU has strong regulatory powers over economic, social, environmental, and other policies in the member states (Majone, Reference Majone1996). It does so despite the fact that it is fiscally weak and lacks the means to ‘enforce’ its rules against dissenting member states. Even in the field of taxation, another core state power, the EU regulates the member states’ use of their monopoly to tax without having itself the power to tax (Genschel and Jachtenfuchs, forthcoming). Disaggregating the monopoly of force into three dimensions allows for the possibility that international institutions regulate the state monopoly of force without exercising it themselves.
Given the centrality of the monopoly of force, change may occur very slowly. In order to be able to see such long-term changes, we will compare issues from the 1960s and 1970s with issues from the last decade. We focus on the fight against terrorism and drugs because both are of high salience to the state and have long been on the agenda. If we take these criteria together, we get a matrix of 12 fields. In each field, we analyse the most salient issue and the most relevant institutional context instead of privileging the EU from the outset (Table 1). We coded the strength of international institutions using the scales developed in section ‘Measuring the strength of international institutions’.
Taken together, our database is composed of about 450 documents of different types, such as treaty texts, legislation, speeches, internal notes, or minutes of meetings. Although our condensed end results (Table 3) consist of 36 data points (12 issues measured with 3 scales), these findings rely on an extensive coding procedure. To systematize the coding process, we used the MAXQDA software (http://www.maxqda.com).
Overview of issues
Note: The abbreviations in italics in each cell serve as unique identifiers for the issues throughout the paper.
The strength of international policing institutions
While combining qualitative with quantitative methods in one design, we encountered the well-known presentation and reliability problems of studies working with large data sets (Creswell et al., Reference Creswell, Plano Clark, Gutmann and Hanson2003). Unlike typical quantitative studies, we cannot simply present the results because the coding process cannot be taken for granted. However, our data set is too large to display the coding process in detail to the reader. While the methodological discussion on this point is not settled, we made a compromise between presentation of data and discussion of results (cf. Seale, Reference Seale1999: 154–157). In the following, we will therefore present the tip of the iceberg in a condensed, narrative fashion. A complete list of documents analysed as well as our MAXQDA file and a detailed discussion of the research process are available on the journal website.
Fighting terrorism
Legitimation
The main legitimation issue in the fight against terrorism has been to agree on an international convention defining terrorism (Friedrichs, Reference Friedrichs2006). Efforts in this direction started in the late 1960s at the level of the United Nations (UN) when terrorists sought to internationalize the Middle East conflict. However, due to fundamental disagreement mainly on the role of national liberation movements, no agreement was reached. Instead, the UN has adopted a series of issue-specific conventions.Footnote 5 Despite these efforts, none of them includes an explicit definition of terrorism, nor do they amount to one if taken together.
In 1999, the UN resumed its work on a comprehensive convention on terrorism.Footnote 6 As, however, the underlying conflicts have not changed even after 9/11, there is still no agreement. Faced with the stalemate in the UN on the one hand and in need of a definition for concrete measures like the European Arrest Warrant on the other hand, the EU adopted its own definition of terrorism in 2002 (2002/475/JHA). It lists a number of criminal offences that are considered terrorist acts, if committed with certain goals in mind (Article 1).
In terms of decision-making, there was no cooperation in the 1970s,Footnote 7 but the recent EU definition constitutes a case of negotiation. As the EU Treaty does not require EU action on this issue, member states wishing to prevent any impact on their freedom of manoeuvre could simply veto the proposal.Footnote 8 In terms of precision, it leaves only narrow areas of discretion as it is based on a detailed list of offences and goals.Footnote 9 As there was no agreement in the 1970s, states were not restricted either in this regard or with regard to dispute settlement.Footnote 10 After the Lisbon Treaty entered into force, the European Court of Justice (ECJ) became responsible for dispute settlement.Footnote 11
Methods
Faced with the rise of domestic terrorism in the 1970s, European states made several attempts to create for the exchange of information because information was considered crucial in combating terrorism. The most important forum was TREVI, an informal arrangement of EU ministers of the interior founded in 1975. It created several administrative working groups dealing with the exchange of information and with the coordination of the fight against terrorism (Guyomarch, Reference Guyomarch1997: 130–131). All TREVI member states established liaison offices in 1977 in order to link their police forces and their secret services. On the level of civil servants, police officers of EU member states and some third countries met in the Police Working Group on Terrorism founded in 1979. It mainly facilitated the exchange of intelligence and promoted the secondment of officers to other countries. Even more secretive groups emerged over the years (Benyon et al., Reference Benyon, Turnbull, Willis, Woodward and Beck1993: 84, 89). Thus, states perceived a strong need for information from other states although they were reluctant to reveal sensitive information in turn.
With the Maastricht Treaty, TREVI and cooperation on Justice and Home Affairs, in general, became integrated into the EU as the so-called third pillar. While only slow progress was made in the beginning, numerous measures have been agreed in recent years. The Lisbon Treaty further strengthened this sector. Today, cooperation against terrorism is all about the gathering of intelligence, that is, information on terrorist suspects suitable for prosecution authorities. In this respect, the EU is now of central importance for its member states: It has created its own organization for the processing and analysis of crime-related data, Europol (Occhipinti, Reference Occhipinti2003), and the Schengen Information System as an additional institutionalized network for information exchange. Over the years, a number of concrete measures have been adopted. Examples include directive 2006/24/EC which foresees the retention of data on electronic communication such as phone calls, decision 2008/615/JHA which provides far-reaching access options to these data, and decision 2005/671/JHA which deals with information exchange and cooperation concerning terrorist offences (Geyer, Reference Geyer2008).
Despite the sensitivity of the information at stake, states did not act unilaterally, but created an informal international institution as early as the 1970s. In terms of decision-making, TREVI was clearly based on negotiations at the ministerial level which were prepared in working groups of civil servants (Lavenex and Wallace, Reference Lavenex and Wallace2005: 459).Footnote 12 The recent EU agreements are cases of negotiation as well.Footnote 13 At the time of decision-making, police, and judicial cooperation in criminal matters were regulated in Title VI TEU and remained subject to intergovernmental negotiation. The Europol convention was a partial exception as it was foreseen in the EU Treaties. Negotiation on it was thus compulsory.
With respect to precision, the cases are more difficult to code. As the texts of the TREVI meetings were kept secret, we had to rely on indirect sources. One police official lists categories of objects for which the exchange of object-related intelligence had been agreed (Wiesel, Reference Wiesel1985: 212). Another agreement stated that information on terrorist incidents had to be communicated to other member states within 24 hours (Anderson, Reference Anderson1989: 54). These provisions are quite precise. As, however, no common definition of terrorism existed to which they could have been applied, we classify them as standards.Footnote 14
As regards the recent EU measures, we can find paragraphs for each of the different levels of discretion. For instance, decision 2008/615/JHA defines the conditions under which information will be provided in such a long-winded way that it is impossible to determine whether conduct complies or not.Footnote 15 At the other extreme, Article 5 of the Data Retention Directive lists in intricate detail which information is to be retained, for example, ‘the date and time of the log-in and log-off of the Internet access service, based on a certain time zone, together with the IP address’ (2006/24/EC, para. 1(c), 2(i), leaving only narrow areas of discretion. Despite these different degrees of discretion, the central provisions leave only narrow areas of discretion.Footnote 16
No rules on dispute settlement seem to have existed in the 1970s.Footnote 17 Lately, however, the Lisbon treaty has granted jurisdiction to the ECJ. This also applies to Europol which is no longer governed by a convention but has been brought fully into the institutional architecture of the EU by Council Decision 2009/371/JHA.Footnote 18
Authorization
Our authorization cases deal with the extradition of terrorist suspects. Extradition treaties and treaties against terrorism traditionally contain political exemption clauses which make it difficult to seek extradition for terrorist offences (Bassiouni, Reference Bassiouni2003). The requested state could choose to refuse extradition by arguing that the offence was political. During the ultimately inconclusive UN debate on a comprehensive convention on international terrorism, the European Convention for the Suppression of Terrorism (ECST) was signed in 1977. Its purpose – to make terrorist offences extraditable by defining them as non-political offences – was compromised by far-reaching opt-out clauses (Article 13) and the possibility to denounce the convention with immediate effect (Article 14).
More recently, traditional extradition has been replaced by the European Arrest Warrant (EAW; 2002/584/JHA). It is based on the principle of mutual recognition of judicial decisions between EU member states. In essence, arrest warrants for an offence listed in the EAW framework decision issued in one member state are executed in another one and the arrested suspect is then transferred to the issuing member state. Under the traditional extradition regime, such requests were not only checked by the courts but also by the executive which enjoyed large discretion regarding the final decision, thus clearly retaining sovereignty in the use of their monopoly of force. With the new procedure, surrender can no longer be blocked by the executive.
In terms of decision-making, the ECST was the result of negotiations.Footnote 19 The EAW is a case of compulsory negotiation, foreseen in Article 31 (b) TEU.Footnote 20 With respect to precision, the provisions of the ECST that terrorist offences are extraditable (Article 1) and that suspects who are not extradited due to an exception must be prosecuted nationally (Articles 6 and 7) leave little room for interpretation. However, there are several exit options. Some are clear-cut (e.g. Article 14 on denouncing the Convention), others serve their purpose by having been formulated at the standards level of precision only. Article 5 refers to the right to asylum and qualifies as a standard. Article 13 contains a provision that is supposed to make a refusal of an extradition request less likely, and which is also formulated as a standard. As these provisions weaken the more clear-cut provisions, we coded the ECST as a standard.Footnote 21
The EAW instead lists 32 offenses to which it applies, including terrorism. The only provisions that leave broader margins of discretion are those on multiple requests (which qualify as standards) and those on the determination of the competent national authorities, which leave this decision completely to the member states. In terms of precision, the EAW thus leaves only narrow margins for interpretation.Footnote 22
With respect to dispute settlement, the ECST stipulates that the European Committee on Crime Problems of the Council of Europe was to discuss problems so that they could be solved in a friendly manner (Article 9). If unsuccessful, the dispute could be referred to arbitration. The disputants chose the arbitrators who then nominated a referee.Footnote 23 As regards the EAW, the Lisbon Treaty grants the ECJ jurisdiction.Footnote 24
In effect, the EAW is a true revolution in European criminal justice cooperation. It replaces a sovereignty-preserving extradition regime full of exceptions and involving political decisions by governments. Instead, the EAW introduces a quasi-automatic judicial procedure with few exceptions that involves only prosecution authorities and local courts. Police officers in Europe now arrest suspects – even if they are their own nationals – based on foreign warrants and hand them over to the requesting member state. In addition, the EAW is not an isolated act but the blueprint for a series of measures built on the same principles.Footnote 25
Practitioners and courts are well aware of the significance of the changes (e.g. Keijzer and van Sliedregt, Reference Keijzer and van Sliedregt2009). Unsurprisingly, putting the EAW into practice took several years. The surrender of own nationals especially has sparked opposition. In 2005, the constitutional courts of Germany, Poland, and Cyprus ruled against the extradition of nationals. Meanwhile, most of the resistance has been overcome and the EAW directly affects an increasing number of people. In 2008, national authorities executed almost 700 EAWs against their own citizens.Footnote 26 From 2004 to 2008, the number of EAWs literally exploded from about 3700 to almost 14,700. Even for the ‘eurosceptic’ United Kingdom, this number rose from 96 in 2004 to 515 in 2008. Illustrating the relevance and speed of the surrender, an EAW was used by the United Kingdom to secure the return of Osman Hussain. The terrorist suspect had fled to Rome after the failed 21 July 2005 London bombings. He was arrested in Italy on July 29, returned to the United Kingdom on September 22, and convicted to 40 years of prison in 2007. The number of arrests at the request of another member state has grown at an even faster pace – from 10% of EAWs issued in 2004 to 20% in 2008. The refusal rate is constantly rather low at about 11%. By and large, courts accept arrest warrants by foreign judicial systems. Surrender procedures now take only several days to a few weeks. Overall, more than 4000 persons were surrendered to another member state in 2008. In the areas covered by the EAW, traditional extraditions with their political reservations have been replaced by a surrender procedure between judicial authorities.
Prohibiting drugs
Legitimation
While international efforts to prohibit drugs can be traced back to a series of conferences from 1907 onwards, today's drug prohibition regime was formed when UN member states agreed on the Single Convention on Narcotic Drugs in 1961 (McAllister, Reference McAllister2000). Since then, the key issue in legitimating the fight against drugs has been to find agreement on a common definition legitimating global measures against both supply and demand. Within little more than 10 years, a universal definition of drugs as well as the respective measures were adopted. With respect to supply, states were obliged to strictly control the trade of licit drugs and, in particular their diversion to illicit channels. They also had to establish a national agency in charge of drug control and to adopt national drug legislation in line with the conventions. Concerning demand, measures pertain to education, after-care, rehabilitation, and the social reintegration of drug addicts. However, the demand side has never been at the heart of the regime, as the insight that the drug problem is about trade and addiction only evolved over the years. Thus, provisions on demand have been considerably less precise in all UN agreements. By 1972, two conventions and one protocol had been adopted. With the 1988 UN Convention Against Illicit Traffic in Narcotic Drugs and Psychotropic Substances, they constitute a comprehensive regime with near-universal validity (184 member states at present).Footnote 27
Although the UN regime has been the result of negotiation as decision-making mode,Footnote 28 we now witness a global drug prohibition regime within which the core issue in legitimating the fight against drugs – the definition of the substances counting as drugs – is controlled by international institutions. Thus, with respect to precision, states experience a highly constraining global regime: the 1961 Single Convention as well as the 1971 supplementary Convention of Psychotropic Substances established a number of schedules which lists more than 250 substances considered as drugs. These lists not only limit the room of manoeuvre for the signatory states with regard to illicit drugs but also concerning the licit use of drugs, for example, for medical purposes. The provisions are very precise and leave only narrow areas of discretion for the signatory states.Footnote 29
This conclusion has to be nuanced with regard to the demand side. During the 1980s, states – in particular European states – have been increasingly aware of the social problems of drug consumption and criticized the dominant US-inspired law enforcement approach focussing mostly on supply within the UN regime. They favoured stronger demand-side measures, most importantly harm reduction (Elvins, Reference Elvins2003: 31). However, as attempts to include such measures in the 1988 convention faced fierce US opposition, the agreement is very imprecise in this respect (see the ‘soft’ formulations of Article 4 (b–d)). Thus, in comparison to the supply side, UN provisions for demand are considerably more flexible.
This situation changed only slightly when the EU emerged as a player in international drug prohibition. After the Maastricht Treaty entered into force in 1993, the UN conventions were incorporated into the EU acquis communautaire, that is, into the undisputed set of rules all new member states must adopt.Footnote 30 While this inclusion has led to a strengthening of the supply-side of the UN regime, it has not led to common demand-side measures in the EU. Although the EU adopted its Drug Strategy 2005–2012 and two action plans (EU Council, 2004, 2005, 2008), matters of public health – to which demand-side measures predominantly belong – still primarily fall under the competence of the member states. In addition, as not all member states are willing to deviate from the UN drug regime, the EU strategy and plans use imprecise categories for substantive goals and leave much of the implementation to the member states. The result of these differing stances is an EU too divided on the question of appropriate demand-side measures to agree on a uniform approach. Thus, in comparison to the UN supply side, EU member states have regained broad areas of discretion with regard to demand, as measures are now again within the national competence of the member states (Boekhout van Solinge, Reference Boekhout van Solinge2004: 54–55). This explains why the issue is the only in our sample with decreasing strength.Footnote 31
Apart from this newly acquired leeway for member states, the increasing strength of international institutions in legitimizing drug prohibition is clearly visible in changes in the dispute settlement mechanisms. In 1961, Article 48 of the Single Convention provides that disputes should be settled among the states concerned. Only if all disputants agree, they can turn to the International Court of Justice (United Nations, 1973). Thus, states remain completely free on how to solve their disagreements.Footnote 32 In the EU, this freedom has been severely reduced: as the UN regime is part of the acquis communautaire, this entails dispute settlement by the ECJ.Footnote 33
Methods
In the 1960s and 1970s, US-inspired enforcement techniques dominated the methods dimension, while the fight against money laundering has been central since the 1990s. During the 1960s and 1970s, the United States took an active interest in promoting its enforcement techniques and exported them to Europe through bilateral channels (e.g. Cusack, Reference Cusack1974: 242–244). During the 1970s, the United States introduced various infiltration techniques to Europe. Infiltration meant the use of informants and undercover agents, the observation of suspects, and controlled deliveries of illicit drug consignments (Nadelmann, Reference Nadelmann1993: 207–246).
From the early 1980s onwards, controlling money laundering became the key method in the international drug prohibition regime. Formally established with the 1988 UN Convention Against the Illicit Traffic in Narcotic Drugs and Psychotropic Substances, the idea was to make drug trafficking more risky by getting hold of the profits derived from it (Mitsilegas, Reference Mitsilegas2003). The regime involved a growing number of international organizations besides the UN, for instance, the Financial Action Task Force, an independent international body working on the basis of a series of highly influential recommendations on how to counter money laundering. The regime developed further when the Council of Europe adopted the Convention on Laundering, Search, Seizure, and Confiscation of the Proceeds from Crime in 1990. Only 1 year later, the EU agreed on an AML Directive that was expanded and tightened several times (2005/60/EC).
In each of our three categories, the strength of international institutions has strongly increased over the years. In terms of decision-making, ministers and even heads of state met in the 1960s and 1970s to negotiate possible forms of cooperation (e.g. Ministère de l'Intérieur, 1969). The first major agreement was the Franco-American protocol of 1971 by which France and the United States set up joint teams, sent French and US police officers to the partner country for workshops and regulated the modalities of information exchange on methods and training (Ministère de l'Intérieur, 1971). Similar bilateral agreements between the United States and other states followed. They emerged in negotiations.Footnote 34 This has fundamentally changed in the EU: money laundering became part of the first pillar of the EU, and hence subject to negotiation with a subsequent vote.Footnote 35
Regarding precision, we also witness an increase. The 1960s and 1970s agreements exhibit differences in the degree of discretion: on the one hand, the Franco-American protocol makes detailed provisions about the number and location of US drug enforcement liaison officers in France (Article 7). However, the substance of cooperation is circumscribed only in relatively general terms (e.g. Article 3 on information exchange). In a similar agreement between Germany and the United States (Rebscher, Reference Rebscher1981: 168–169), German police officers aimed to model their organizational structures according to the US drug enforcement agencies (Busch, Reference Busch1999: 23–24). On the whole, we have a patchwork of mostly bilateral agreements that set standards that make sense only in a specific situation.Footnote 36 With the EU entering the scene, precision drastically increases. The AML directive contains very precise provisions on sensitive topics such as the lifting of the professional secrecy of lawyers (Articles 2 and 28) or the shifting of the burden of proof to the suspect (Article 9).Footnote 37
Things are similar with regard to the question of dispute settlement. While in the 1970s protocols remained mute on dispute settlement,Footnote 38 the inclusion of money laundering in the first pillar of the EU automatically implies that dispute settlement is performed by the ECJ.Footnote 39
In terms of impact, the AML directive is as important as the EAW although it shows a different pattern. While in the EAW, states use each other's means of force in order to achieve their goals, they use a broad range of private actors as ‘deputy sheriffs’ (Levi and Reuter, Reference Levi and Reuter2009: 371) in the AML case. Banks and other financial institutions had to introduce a systematic scanning of financial transactions and reporting of potentially suspicious cases to public authorities. Together with the obligation of customer identification, this has entailed a partial lifting of ‘banking secrecy’. Lawyers also have to report suspicious deposits or transactions by their clients. Together with the banks, they have unsuccessfully protested against being incorporated into intelligence activities (Montebourg, Reference Montebourg2002: 625–635 and 647–658). Similar requirements apply to auditors, external accountants, tax advisors, real estate agents, or casinos.
As the requirements are backed up by sanctions, they entail a significant impact on the procedures, systems and controls of banks, and a variety of other actors in the member states (Katz, Reference Katz2007: 210). Even more significantly, the fight against money laundering, a major criminal justice issue, is now a firmly established EU competence. A directive that was tightened and expanded several times, very precise provisions, decision-making by majority voting, and adjudication by the ECJ oblige the member states to use their monopolies of force to implement a common European policy against major domestic economic interests as well as against protests by civil rights organizations, and make it difficult for single member states to resist any further tightening of the respective provisions.
Authorization
The cases in this dimension deal with the management of trans-border investigations. The 1960s and 1970s were characterized by mostly bilateral cooperation concerning the exchange of personnel for joint investigations and joint raids against drug producers and traders, often in border areas (Ministère de l'Intérieur, 1970; Bundeskriminalamt, 1974). These instances of informal cooperation were later formalized, for instance in the Franco-American and German-American protocols mentioned above. Here, German and French police officers were seconded to both producing and transit countries as well as to the United States, and US officers came to the continent (Nadelmann, Reference Nadelmann1993: 189–249).
Over the years JITs turned into standard instruments and were considered to be effective. Consequently, the EU formalized their use in a framework decision on JITs (2002/465/JHA). This decision allowed member states to set up teams of law enforcement agents permitted to operate on territories other than their own (though not disposing of executive powers). However, in the first years of its existence, the instrument did not live up to the high expectations of member states (Rijken, Reference Rijken2006). The main reasons for this reluctance can be found in sovereignty concerns with respect to foreign officials on one's territory, a lack of awareness of JITs as an investigative option, and, eventually a lack of funding. In recent years, these problems have been slowly tackled and the number of JITs is constantly growing. By 2008, around 40 teams had been formed (cf. Riegel, Reference Riegel2008: 83). This is a highly interesting development from our perspective: despite JITs’ significant interference with national sovereignty concerns (as not only foreign officers but also Europol or Eurojust officials can be part of the team), states increasingly make use of the new instrument. In February 2010, the Council adopted a resolution on a model agreement for setting up JITs in the attempt to take advantage of the roughly 8 years of experience with them (2010/C70/01).
Thus, with the EU getting involved in the establishment of JITs, the room for manoeuvre for member states has significantly decreased in each of our categories. In terms of decision-making, the 1970s protocols were the outcome of negotiations,Footnote 40 but the EU framework decision on JITs constitutes a case of compulsory negotiation since Article 30, para 2 (a) TEU explicitly calls for them.Footnote 41 While the early protocols did not contain any provisions on dispute settlement,Footnote 42 the ECJ enjoys full jurisdiction after the Lisbon Treaty.Footnote 43 In terms of precision, the 1970s agreements proscribe standards,Footnote 44 such as sending additional officers in ‘special situations’ (Ministère de l'Intérieur, 1971: Article 8). Within the EU, however, the teams have encountered difficulties in their concrete operation, most notably because the framework decision neither specifies the areas in which they are allowed to act nor their time frame of operation (Friedrichs, Reference Friedrichs2007: 172–173). They may be created ‘for a specific purpose and a limited period, which may be extended by mutual consent’ (Article 1). Finally, the framework decision allows for the participation of a large variety of actors, such as Europol agents, Commission officials, as well as law enforcement agents of third states with explicit mentioning of the United States (Preamble, item 9). Overall, the framework decision leaves states broad areas of discretion in terms of precision.Footnote 45
Conclusion: European constraints of the state monopoly of force
Table 3 summarizes our findings. In 15 out of 18 issues, the strength of international institutions in the field of policing increased over time. This included all issues of dispute settlement. It remained constant in two issues of decision-making and decreased only in one issue of precision. This increase is statistically significant, with Cramér's V > 0.5 in decision-making and >0.8 in precision and dispute settlement.
Changes are smallest with respect to decision-making. Most decisions are still taken in negotiations with veto power or an exit option for each state. But meanwhile, this has radically changed with the EU's Lisbon Treaty (Ladenburger, Reference Ladenburger2008). Most future decisions will be taken by the ‘ordinary legislative procedure’, which amounts to voting (3) on our decision-making scale. Only policies touching upon operational cooperation remain subject to unanimous decision, that is, to negotiation (1) on our scale (Article 87, para. 3 TFEU).
Precision has increased strongly over time. In the 1960s, cooperation in the methods and authorization dimension left ample discretion to states. Currently, they have much less discretion. In the legitimation dimension in the fight against terrorism, EU states have moved from failed cooperation to very precise provisions. In the fight against drugs, the precision of legitimating rules has slightly decreased. This is the only decrease in our sample. It reflects the successful attempt of a significant number of EU member states to obtain more domestic policy autonomy on the demand side within the rigid UN drug control regime.
In dispute settlement, increase has been uniform and strong. Whereas hardly any dispute settlement mechanisms were in place in the early period, the ECJ now serves as an independent and compulsory dispute settlement body across issues.
Our data also reveal a specific relationship between decision-making on the one hand and precision and dispute settlement on the other. There are no issues with high scores for decision-making and low scores for precision or dispute settlement. States are reluctant to adopt more constraining modes of decision-making. However, they do not attempt to offset the constraining effect of higher modes of decision-making by imprecise agreements or by weak dispute settlement: Once they move beyond negotiation, they also conclude more precise agreements with independent dispute settlement. Even if states stick to negotiations in decision-making, they adopt more precise agreements and more independent dispute settlement. States do not shift the burden of interpreting imprecise agreements to dispute settlement bodies. Only if agreements are already precise, are disagreements solved by strong dispute settlement provisions.
Interestingly, states do not adopt less constraining provisions when the issue in question comes closer to the use of force either. In both time periods analysed, the values do not decrease from legitimation to authorization as one might expect. Instead, they are almost uniformly low in all three dimensions in the 1960s–1970s and almost uniformly high in all three dimensions in the 1990s–2000s.
These findings clearly contradict the first hypothesis that states would create only weak institutions in the field of the monopoly of force because they were driven by sovereignty concerns. Instead, they support the second hypothesis that states are mainly driven by efficiency concerns and create strong institutions when they perceive the need to solve collective action problems or to increase their problem-solving capacity. The institutions created by EU member states in the field of the monopoly of force are not weak or informal any more. Instead, they contain provisions that are very precise and leave much less room for interpretation compared to earlier ones, and they provide for compulsory and independent dispute settlement.
The second hypothesis distinguishes between pooling (i.e. majority voting among states) and delegation to independent institutions as two distinct forms of institutional choice. In our sample, we find reluctance towards pooling of decision-making (only one value of 3 in Table 3). When adopting a specific measure which relates to the monopoly of force, states retain unanimity, and hence a veto on outcomes. Once the measure is adopted, however, dispute settlement is delegated to the ECJ in all cases in the 1990s–2000s. This delegation is of a particular kind. Unlike in other principal-agent relationships, states have to assume that a revocation is close to impossible because it would violate the rule-of-law principle that is fundamental both to the EU and to the member states. In this trusteeship or fiduciary type of delegation member states presuppose that the ECJ will use discretion. They delegate powers to an independent court in order to increase long-term policy credibility and to avoid short-term defection (Grant and Keohane, Reference Grant and Keohane2005: 32, 36; Majone, Reference Majone2005: 64–67).
The role of the ECJ is even further increased because it is not limited to interstate dispute resolution. The ‘preliminary ruling procedure’ (Article 267 TFEU), which after the Lisbon Treaty also applies to policing, opens a new opportunity structure for private litigants and domestic courts. In the past, this procedure has been one of the main causes for the increasing hierarchy of EU law and the enlargement of EU powers. Furthermore, the Commission can now sue member states for non-compliance with EU law (Article 259 TFEU). Although the EU cannot use force to ensure compliance, research shows that member states comply eventually (Panke, Reference Panke2010).
The low values in decision-making, that is, the prevalence of the unanimity rule and the reluctance towards pooling, do not offset this delegation, quite the contrary. Although unanimity prevents the adoption of rules that even a single member state finds too constraining, it also makes it more difficult to adopt new rules in order to counteract judicial activism for which the ECJ is famous. Unanimous decision-making thus does not protect member state autonomy but increases the discretion of the ECJ (Tsebelis and Garrett, Reference Tsebelis and Garrett2001: 369–370; Alter, Reference Alter2006). The introduction of the ordinary legislative procedure by the Lisbon Treaty in the areas analysed here increases the collective autonomy of the member states in this respect but also increases the likelihood for individual member states to be outvoted. It also makes the European Parliament a co-legislator together with the member states and thus strengthens another supranational actor with policy preferences independent from the member states. As a result, states will not be able to fully control policy development in important areas relating to the monopoly of force.
On the surface, the state monopoly of the legitimate use of physical force is still fully in the hands of EU member states. Police officers of one state or EU agents still cannot arrest citizens in another state. However, the main trend is not the emergence of a new layer of governance which could legitimately use physical force, comparable for instance to US federalism. Instead, it is the emergence of ever more constraining European and international rules on how member states can use their monopoly of force. These rules have become much more precise, and disputes over their application are settled by an independent court. They also have concrete effects. JITs are becoming more frequent, allow police agents of one state to operate on the territory of another and Europol officers to participate in these teams. With the AML Directive, member states use private corporations for policing purposes and interfere deeply with privacy rules and practices. With the European Arrest Warrant, the prosecution authorities of one EU member state can request the arrest and even the surrender of a citizen of another member state by the authorities of that very state in a quasi-automatic procedure, which bears little resemblance to traditional sovereignty-preserving extradition practices. With almost 4900 arrests made in 2008,Footnote 46 this has become a daily routine in Europe. All these measures are based on EU rules.
Our findings thus fall between the extreme positions of a decline or a resilience of the state. Instead, we see a middle way between both positions: The state monopoly of the legitimate use of physical force is still intact. However, the reasons for legitimating its use and the criteria for when, how, and under which conditions it can be used are increasingly shaped by the EU. Supranational European institutions do not replace the state monopoly of force or compete with it but increasingly embed it in a larger institutional framework and constrain and regulate its usage beyond the individual or collective control of the member states.
Although the institutional setting of the EU and the readiness of its member states to accept supranational institutions are unique, some careful generalizations are possible. First, we see strong international institutions touching upon the monopoly of force but without central enforcement powers. Second, institutional strength comes less from supranational decision-making but from highly precise agreements and independent dispute settlement. These conditions are not specific to the EU. Strong institutions in the field of the monopoly of force thus seem possible even beyond the EU when states unanimously adopt rules which are not only precise but also contain provisions on independent dispute settlement. In more homogenous contexts such as the OECD or in regional agreements, this combination offers the possibility to reconcile the concern for strong institutions in the fight against transnational crime with a concern for sovereignty.
Acknowledgements
This paper originated in a research project on the internationalization of the state monopoly of force within the Bremen TranState research centre (http://www.state.uni-bremen.de) which included the authors as well as Jörg Friedrichs and Holger Stritzel. We gratefully acknowledge their empirical and conceptual input to the project. The Deutsche Forschungsgemeinschaft provided generous financial support. We would like to thank Andrea Derichs, Axel Domeyer, Oliver Hübel, Thomas Müller, Mariya Shisheva, and Dana Trif for research assistance. We are also grateful to the anonymous reviewers, James Davis, Berthold Rittberger, Tanja Börzel, and most notably Bernhard Zangl for constructive comments on earlier versions of this paper.
