Introduction
This article argues that due to the inbuilt institutional bias of the multilevel European Union (EU), the policy instruments applied are partially decoupled from underlying normative ideas. Supranational policy-making is strongly based on cognitive and in particular program ideas, whereas normative ideas, especially ideas of public sentiments, appear detached from policy-making. This entails two relevant consequences. First, the decoupling of abstract normative ideas from the pragmatic ideas promoted in day-to-day policy-making implies, in principle, problems for elites in terms of gaining public support and hence legitimacy. Second, legitimacy and public support only become a hindrance for policy-makers where they attempt to re-couple operational cognitive and abstract normative ideas. As long as these remain decoupled – that is, as long as public sentiments are not confronted with programmatic ideas and are not seen to undermine basic norms or values – policy-makers actually gain room for maneuver because they can advance new policy instruments in the shadow of old ideas. The main competing ideas of EU governance are linked to the principles of state sovereignty vs. increasingly centralized supranational administration. This general tension is applied to public administration. Whereas domestic civil services are mostly framed as autonomous, scrutinizing policy instrumentation illustrates the increasing shift to forms of integrated administration. The latter is based on pragmatic ideas that are not re-coupled to an explicit norm or value of a desirable, more state-like EU. Instead, it is bound to a pragmatic functional paradigm of inevitable requirements to keep multilevel governance working.
The analysis of policy instruments offers a fruitful empirical angle to tackle the question of ideological discrepancies and their implications in multilevel governance because analytically ‘public policy instrumentation is a major issue in public policy, as it reveals a (fairly explicit) theorization of the relationship between the governing and the governed: every instrument constitutes a condensed form of knowledge about social control and ways of exercising it’ (Lascoumes and Le Gales, Reference Lascoumes and Le Gales2007: 3). Accordingly, policy instruments reflect power relations and different instruments produce distinct kinds of relations between policy-makers and policy-takers or, in our case, supranational and national or subnational public administrators. The case of policy instrumentation in public administration illustrates how instrumentation performs the balancing act between competing pragmatic cognitive ideas and fundamental public sentiments. In the case in hand, these competing ideas are not reconciled. Re-coupling is explicitly evaded. Instead, the mix of instruments that satisfy both underlying ideas allows for the ‘peaceful co-existence’ and hence acceptance of pragmatic ideas despite inconsistencies with wider public sentiments.
The article is in three parts. The next section conceptualizes the link between policy ideas and policy instruments by drawing on Campbell's typology of the effect of ideas on public policy. To render these concepts viable for our analysis, the ideas that underpin the EU polity at large and the competing actor strategies of decision-makers within this context are presented. Against this background, the third part turns to the case of pubic administration in the EU's multilevel polity. Clarifying how the different types of ideas play out empirically highlights the systematic contradictions between cognitive and normative ideas. The dominant frames of an overly technocratic ‘Eurocracy’ and widely held public sentiments of autonomous national civil services are juxtaposed with policy paradigms that build on the notion of a genuine ‘European administrative space’ and the large toolbox of administrative policy instruments derived from the pragmatic program ideas that guide policy-making. It is therefore concluded that the EU actually has a substantive public administration policy that tangibly changes domestic civil services. Yet, due to successful decoupling, integrated bureaucratic practice has evolved largely unnoticed and is therefore hardly legitimized on the normative level.
Conceptualizing the role of ideas in empirical analysis
How are policy instruments linked to underlying policy ideas? Campbell (Reference Campbell1998, Reference Campbell2002) differentiates between types of ideas in public policy, which can be fruitfully applied to the EU. As will be shown, the multilevel structure of the EU enables policy-makers to refer to pragmatic ideas and frames when acting on the supranational level without necessarily promoting genuine values or attitudes in support of the EU. To start with, ideas and their role in social interaction need to be defined. Following in the footsteps of Max Weber, ‘[n]ot ideas, but material and ideal [ideological] interests, directly govern men's conduct. Yet very frequently the ‘world images’ that have been created by ‘ideas’ have, like switchmen, determined the tracks along which action has been pushed by the dynamic of interest’ (Weber, Reference Weber1948: 280). Assuming such an interplay of ideas and interests in which ideas are the matrix for interests, which then guide actors’ decisions, the analytical challenge is to identify the causal mechanisms ideas trigger in order to understand how they influence the making of public policies.
Table 1 reproduces Campbell's typology that distinguishes the effect of different ideas on policy-making. According to the author, ideas work either in the background as underlying and at times taken-for-granted assumptions, or in the foreground as concepts or theories, which political elites articulate explicitly. The second dimension distinguishes between a cognitive level that rests on descriptions or theoretical analyses that specify cause/effect relationships, and a normative level embodied by values and attitudes (Campbell, Reference Campbell1998: 384). In order to analyze the link between policy ideas and policy instruments in a particular system such as the EU, we need to identify which program/pragmatic ideas, policy paradigms, symbolic frames and public sentiments shape the general polity context. Before turning to how different ideas operate with respect to public administration, I will therefore briefly depict the broader ideas the EU polity rests upon.
Table 1 Types of ideas and their effects on policy making
Source: Campbell (Reference Campbell1998: 385).
The multilevel context of decoupling in the EU
In the background of the policy-making process, the political system of the EU is founded on two competing ideas: a supranational state-like entity vs. intergovernmental cooperation between sovereign member states. In line with these underlying ideas, the first generation of integration research was divided between neo-functional/supranational and intergovernmental/state-centered ‘grand theories’ that argued for an emerging finalité of one kind or the other (Keeler, Reference Keeler2005).
This debate has largely been put to rest in favor of more mid-range theorization and the recognition that the EU is (and will remain for some time at least) a hybrid polity commonly described as a system of multilevel governance. Accordingly, the ‘bicephalous’ (Tömmel, Reference Tömmel2007: 265) nature is not a transitional state but a constituent feature of the EU, largely confirmed by the latest reforms contained in the Lisbon Treaty (2007, in force since 2009), which do not turn the EU into a ‘super-state’ but sustain the EU's hybridity. The notion of multilevel governance offers a heuristic to describe policy-making inside the non-state-like polity. The popularity of multilevel governance ‘among theorists can be attributable to its descriptive neutrality and, hence, its putative compatibility with virtually any of the institutionalist theories and even several of its extreme predecessors’ (Schmitter, Reference Schmitter2003: 49). The approach is accordingly silent when it comes to the role of ideas in policy-making, the main focus of our attention here. Filling the multilevel image with ideational meaning, the distinction between cognitive and normative ideas in the background indicates how to place the competing ideas into the policy-making context of the hybrid EU system.
The underlying competing ideas the EU rests on are best illustrated by the debate surrounding the last reform treaty, which culminated in an upfront confrontation. The reform treaty was first signed by the Union's heads of state and government as the Treaty establishing a Constitution for Europe in 2004. Two failed referenda on the ratification of the treaty in France and the Netherlands caused the ‘constitutional treaty’ to fall in 2005 and triggered some debate about the underlying core ideas of the EU polity. This debate was not resolved in favor of one or another idea that would, in consequence, have led to a more federal or intergovernmental institutional architecture for the EU. On the contrary, public debate was eventually circumvented by renaming essentially the same text as the Treaty of Lisbon, which, despite another failed referendum in Ireland in 2008, was finally ratified in 2009 after Irish public consent was reached in exchange for further guarantees safeguarding Irish sovereignty. In a nutshell: in the background of the debate on the most fundamental reform treaty since the foundation of the EU, normative ideas expressed in public sentiments were dominated by a tenor of national sovereignty, while on the cognitive level, political elites across the EU acted according to a supranational paradigm that argued the inescapable necessity of advancing the elite-driven process of integration. In the reflection period between the failure of the constitutional treaty and the mandate for the Lisbon Treaty, decision-makers’ interests incrementally converged with respect to a number of central functional issues. These included first and foremost the timeframe for finalizing the reform process and the reform measures that were to ensure the day-to-day functioning of the enlarged Union. Fundamental decisions on the polity development were instead excluded from the debate.Footnote 1
The example of the treaty reform process thus shows (a) strong public assumptions about desired national autonomy, and (b) how political leaders, at the domestic level first and foremost, nonetheless managed to regain the scope to push the reform process ahead by evading a confrontation with normative ideas and framing the Lisbon Treaty as a merely technical necessity. As long as EU policies were almost exclusively the preoccupation of political elites, the underlying ideational contradictions in the founding treaties were of no major relevance for legitimizing EU policies to citizens. The first decades of EU integration ‘were years of permissive consensus, of deals cut by insulated elites. The period since 1991 might be described, by contrast, as one of constraining dissensus. Elites, that is, party leaders in positions of authority, must look over their shoulders when negotiating European issues’ (Hooghe and Marks, Reference Hooghe and Marks2009: 5). In this changed context, according to the main argument, the decoupling of program ideas and public sentiments has become key to EU policy-making. The failed referenda and debate around the so-called Constitutional Treaty (in the Netherlands and France) and subsequently the Lisbon Treaty (in Ireland) in particular have revealed that public sentiment does not support the idea of a supranational polity as a legitimate solution per se. The main actors in this changed context remain the same elites, namely party leaders in positions of authority and most prominently the representatives of national governments. The agency provided by these actors is decisive for the de- and re-coupling strategies outlined in the next section. The actual policy instruments dealing with competing ideas that follow these strategies will be discussed subsequently when mapping actual administrative procedures.
Actor strategies and institutional structures: decoupling and re-coupling
It is not unusual for normative and cognitive ideas to be in discord. Part of the art of policy-making is to bridge or cover up the distance between program ideas and symbolic framing. Two opposed strategies are available in cases where cognitive and normative ideas clash. Policy-makers either take part in an exercise of ‘communicative discourse’ (Schmidt, Reference Schmidt2008, Reference Schmidt2010) to reconcile the cognitive and normative levels by changing ideas, or they opt for the opposite strategy, namely to decouple pragmatic actions from legitimizing talk (Meyer and Rowan, Reference Meyer and Rowan1977; Brunsson, Reference Brunsson1989; Lipson, Reference Lipson2007). Pointedly worded, the latter strategy inverts the logic of advocacy coalitions in the policy process. Unlike the usually assumed prerequisite of creating consistency between core beliefs and actions (Sabatier and Jenkins-Smith, Reference Sabatier and Jenkins-Smith1993, Reference Sabatier and Jenkins-Smith1999), decoupling suggests the contrary: no ideational change but inconsistent behavior.
The above illustration of the ideas that underpin the bicephalous institutional architecture of the EU creates a structural bias among European political elites for relying on strategies of decoupling program ideas and public sentiments. The reason for this lies in the trajectory of integration dynamics, intentionally based on cognitive ideas – that is mainly arguing for functional necessities – in order to circumvent deadlock and conflict over normative ideas. Moreover, the very institutional architecture of the EU polity attributes different cognitive and normative roles to different institutional actors. While the European Commission is endowed with the pursuit of the shared common good of the Community, member state representatives are responsible for protecting specific national interests within the Councils (Majone, Reference Majone2006). Both roles are well established, leading not least to the notion of the Commission being an overly functional technocracy, mainly concerned with widening Community competences, whereas the national governments try to counteract these tendencies in the Council.
The third main institutional player, the European Parliament (EP), has gained considerably more powers as the direct representative body of citizens over the past decades. In parallel, voting behavior inside the EP has increasingly grown to reflect traditional left–right patterns (Hix and Noury, Reference Hix and Noury2009). Yet each increase of parliamentary power has been accompanied by a decrease in electoral turnout for EU elections, and the persistent second-order electoral status demonstrates that the apparent politicization inside the EP has not rendered the Parliament a noteworthy actor able to steer an EU-wide communicative discourse. To the extent that the formal institutional lack of democratic participation has been tackled, the actual ‘lack of democratic politics’ has obviously increased (Hix, Reference Hix2008: 595). There is hence a mismatch between formal powers and the way actors, be they electorates or decision-makers, make use of them.
This lack of politics owes to the institutional architecture and actor strategies dominant political elites rely on; both are deeply intertwined. The lack of an EU government proper and the failure to establish the EP as a political forum despite a real increase in powers implies that ‘we cannot organize opposition in the EU – we cannot appeal for votes against a government in elections or in parliament – because the EU itself has been depoliticized. Opposition demands political debate, and, as Vivien Schmidt puts it, the EU is a polity without politics’ (Mair, Reference Mair2007: 7–8). The competing background ideas thus show in the institutional architecture as ‘depoliticized politics’. The organizational context implies that ‘a “political deficit” currently exists in the European Union, as actors such as political parties, trade unions, or even the media that traditionally dominate the political scene at national level, and therefore provide useful markers to voters, are generally weaker at European level’ (Dehousse, Reference Dehousse1998: 115). In the conceptual framework referred to: policy-making is dominated by pragmatic instrumentation but lacks tools to effectively couple these instruments with normative ideas.
De- and re-coupling strategies are determined by this institutional context, one that leaves main supranational actors without the capacity to convincingly re-couple policy ideas – that is, to create coherence between beliefs in the background and political action in the foreground. The Commission, on the one hand, is defined as a functional, technical organization whose legitimacy is born precisely from its apolitical ethics. Any substantive politicization of the Commission without rendering it a government proper would thus deprive it of its very raison d’être; however, without politicization, it cannot credibly represent competing political ideas. The EP, on the other hand, is not perceived as a relevant actor even though it may be said to have become one. A substantive shift of political attention from national parliaments to the European would stir up major political stakes among the third set of political actors that likewise hold a monopoly over the political discourse on the EU: national governments. Given that vital policy-making powers have been conferred to the EU and are shaped outside the strict zone of direct domestic influence, national cabinets and parliamentarians cannot have a real interest in systematic re-coupling. Re-coupling would essentially mean referring electorates to supranational decision-making; most consequentially, to publicly promote participation in EP elections and to call on the Commission to effectively take over governmental responsibilities. Systematic re-coupling entailing the redirection of political attention from national to supranational public actors is hence threatening these actors’ own political survival. Systematic decoupling, in contrast, allows them to stay in the driving seat of the political process by keeping control over the political world, even if actual control over policies and action has moved elsewhere.
Recalling the main arguments developed by organizational theorists, Tilcsik summarizes how the creation of legitimacy can be systematically decoupled from pragmatic actions to serve the interests of actors who dominate the political discourse:
According to Meyer and Rowan (Reference Meyer and Rowan1977), decoupling enables organizations to gain external legitimacy while also maintaining the internal flexibility with which it can address practical considerations. […] Building on this framework, Elsbach and Sutton (1992) added that decoupling may help organizational spokespersons provide plausible excuses and justifications in case the violation of an institutional mandate is revealed. More recently, Westphal and Zajac (2001) suggested that decoupling occurs not because it is functional for an organization but because it serves the interests of organizational leaders (Tilcsik, Reference Tilcsik2010: 1474, emphasis added).
In sum, the institutional structure of the EU that balances the accommodation of both competing background ideas creates a systematic bias for decoupling – that is, for saying one thing in the national arena and doing another at the supranational level. All the features stressed in the above citation play a role in this. First, the external legitimacy of the EU is bought by stressing its non-state-like nature, while internal flexibility is sustained as findings on ‘subterfuge’ and ‘informal institutionalization’ demonstrate (Héritier, Reference Héritier1999; Farrell and Héritier, Reference Farrell and Héritier2007). Second, the multilevel system lends itself to excuse and justify failure by shifting blame onto supranational actors. As the empirical analysis below will elaborate, the Commission in particular is systematically framed as the scapegoat, even in the case of corruption, where defection is at least equally problematic within the member states where most money is actually lost. Third, the interests of national political leaders – the dominant ‘organizational leaders’ in EU decision-making – matter rather than functional considerations. This is the very core of the decoupling strategy, namely to safeguard political stakes while at the same time trying to achieve supranational coordination. Keeping control over the legitimizing talk grants national politicians their continued political role, independently of pragmatic actions. Decoupling in the multilevel setting furthermore offers opportunities for national politicians to realize policies otherwise impossible to effect in the domestic context due to political opposition. In order to illustrate how this basic pattern plays out in a central field of public authority, I will turn to bureaucratic politics. This is suited to the empirical examination of how the decoupling of ideas plays out because it directly captures the discrepancy between the strong regulatory integration of specific policy issues in the EU and the lack of institutional design that one would expect to follow hand in glove. In other words, if policy-makers decide on pragmatic policy solutions, decisions should be executed via an accompanying administrative apparatus to guarantee the intended outcomes. However, and indicative of the EU's institutional architecture as just described, public administrations remain formally under the control of the member states. In addition, national bureaucracies are deeply entrenched in their national traditions that vary between Anglo-American, Napoleonic, Germanic and Scandinavian across the EU (Painter and Peters, Reference Painter and Peters2010). Therefore, investigating the instruments, that is the legal steering tools, applied in the interplay between supranational decision-making and national administration offers decisive insights into the decoupling of competing ideas in multilevel systems.
Instrumentation in EU multilevel administrationFootnote 2
In the case of public administration, the tension between the underlying competing ‘grand ideas’ is markedly explicit. On the normative level, ideas about civil services continue to be strongly based on the formal legal viewpoint that since national public administrations ‘remained strictly an area of national sovereignty, there cannot be any European policy since there is no community competence in this area’ (Mangenot, Reference Mangenot2005: 4, italics in original). Nonetheless, and due not least to the small size and limited resources of the European Commission (that counts no more than some 33,000 staff),Footnote 3 national civil services are key actors in formulating and implementing supranational policies. The theory developed above suggests that faced with competing ideas, decoupling the cognitive and normative levels offers a viable political strategy. The guiding hypothesis is that short of formal EU competences in administrative policy, far-reaching bureaucratic policies have been established on the basis of pragmatic program ideas. This set of policy instruments exists in the shadow of decoupled symbolic frames and public sentiments. To illustrate how this systematic decoupling of cognitive and normative policy ideas plays out, I will discuss the different dominant policy ideas on EU bureaucracy illustrating both the empirical expressions in terms of public discourses and hands-on policy instruments. Moreover, the most relevant actors promoting the various types of ideas are identified (see Table 2). Methodologically, the focus is on the different legal provisions that delineate the policy instruments applied to grasp, as pointed out above, the ways that social control is actually exercised.
Table 2 Policy ideas and instruments in EU policy-making
Starting with the most general notion, public sentiments about EU public administration are based on the formal independence of national administrations as enshrined in the founding treaties. As is customary for governance by international organizations, the EU system also initially rested on the principle of indirect administration: supranational rules spell out common goals and states are obliged to comply with these but the choice of administrative means remains a domestic prerogative. The above quote on the absence of formal competences echoes this sentiment. The most explicit expression of this notion is the European Union Public Administration Network (EUPAN), an informal network of the Directors General responsible for public administration in the 27 member states and the Commission. The EUPAN promotes voluntary cooperation between EU member states while explicitly circumventing unintended pull-effects towards formalized harmonization. Notably, the member states keep emphasizing that all actors involved ‘acknowledge the informal nature of the European Public Administration Network’ (European Council, 2008) although in practice there is great overlap with the institutionalized bodies of the EU, especially the Council presidency. Although the EUPAN has existed as an informal network between ministers responsible for administrative affairs since the 1970s, questions about real output in core areas of activity (Human Resource Management, Innovative Public Services, e-Government) are generally raised, not least by the network itself (EUPAN, 2008: 4). The EUPAN, which has attracted considerable attention and expanded its scope of activities to all relevant areas of administrative coordination in recent years (Demmke and Bossaert, Reference Demmke and Bossaert2006), thus reflects most clearly how strong public sentiments of truly informal cooperation imply both a great lack of de facto policy outputs and an almost absent awareness of any kind of common policy, given that in its almost 40 years of existence the EUPAN has remained virtually unnoticed.
More important for national bureaucratic independence is the original logic of indirect administration, initially complemented by only a few delegated rights to direct administration by the Commission. Indirect policy execution is based on two principles: the autonomy of national administrations (or structural subsidiarity), which leaves day-to-day management to national authorities, and effectiveness (efficient, effective and uniform enforcement) according to which national ‘administrations, by virtue of a duty of loyal co-operation under the Treaty (Article 10), are required to take all necessary steps to ensure full implementation of these provisions’ (Nizzo, Reference Nizzo2001: 5–6). In a formal legal sense, indirect administration remains the dominant mode of policy execution. Accordingly, in the academic literature ‘in the first thirty years after the Treaty of Rome, the common wisdom was that European administrative law did not exist’ (Bignami, Reference Bignami2004: 3–4; see also Chiti, Reference Chiti2004).
However, as much as the image of indirect administration prevails, in real practice this assumption cannot be upheld any longer. Empirically we see, on the one hand, that actual administrative practice blurs the distinction between direct and indirect administration. Public sentiments clash with pragmatic policy-making based on cognitive program ideas. On the other hand, a number of policy paradigms have surfaced in the policy background that challenge the principle of administrative autonomy. The dominant actor in terms of defining policy paradigms has been the European Court of Justice (ECJ). In order to guarantee the consistent application of EU law, ECJ case law has come to practically subordinate the principle of autonomy to the principle of uniformity (Galetta, Reference Galetta2010: 59–68). This means that, should national procedures conflict with EU law, the uniform application of common rules takes precedence over the administrative autonomy of a single state. Over the years, legal practice established basic administrative law based on procedural rather than constitutional principles. The fact that these principles evolved from ECJ case law and statutes entailed a technocratic and depoliticized bias (Shapiro, Reference Shapiro2001). In short, administrative practice has led courts to define basic principles and hence an abstract body of uniform procedural law that structures administrative action on all levels of the EU system.
The second significant cognitive idea in the background is the notion of a ‘European administrative space’, which was explicitly invented as wildcard to display an essentially absent bureaucratic model of the EU. In the run-up to eastern enlargement (2004/2007), the candidate states’ ability to implement the large body of EU law was increasingly questioned, leading to the introduction of ‘administrative capacities’ as an accession criterion (European Council, 1995; see also: Heidbreder, Reference Heidbreder2011a: 67–73). However, in order to formulate concrete measures to monitor the extent to which the acceding states met this criterion, some template was necessary. References to an alleged common ‘European administrative space’, mainly by legal scholars, was rather ‘metaphorical’ (Meyer-Sahling, Reference Meyer-Sahling2009: 10) than based on substantive EU competences. This notwithstanding, the Commission used this paradigm to develop concrete policy claims and coercive leeway to impose administrative standards on the incumbent member states (Verheijen, Reference Verheijen2000, Reference Verheijen2007; Dimitrova, Reference Dimitrova2002, Reference Dimitrova2005). Notably, the paradigm of an administrative space was necessary on a cognitive level, not least to practically implement the administrative capacity criterion. It was not extended to public sentiments, but endured as a contested term in the academic debate (Siedentopf and Speer, Reference Siedentopf and Speer2003). The fact that administrative reforms introduced in the candidate states were barely sustained after accession is a clear indication of this (Meyer-Sahling, Reference Meyer-Sahling2009). The notion of a unified administrative space has not become a dominant discourse to label the administrative network that underpins EU policy-making. Rather than the illusive administrative space image, pragmatic program ideas underpin the multiple policy instruments that shape daily administrative practice.
Concrete policy instruments are based on program ideas. ‘In contrast to cognitive paradigms, which provide an overarching understanding of how the world works and, in turn, how political institutions and policy instruments ought to be organized in order to achieve broad policy goals, programmatic ideas are more precise guidelines about how already-existing institutions and instruments should be used in specific situations according to the principles of well-established paradigms’ (Campbell, Reference Campbell1998; Campbell, Reference Campbell2002: 28). Accordingly, based on the general paradigm of mutual interdependence within the EU and on the basis of general administrative law principles, a pragmatic preoccupation with domestic administrative systems can be observed. This is best reflected in the introduction of a general ‘right to good administration’ (Charter of Fundamental Rights and Freedoms, Art. 14) that gained legal force with the ratification of the Lisbon Treaty. Pragmatic concerns with the performance of national public administrations regard more than vertical control to ensure the rightful execution of common policies by national administrations. It is also a matter between member states that show increasing concern about horizontal guarantees of minimum administrative standards to ensure the equal treatment of citizens in other member states’ administrative systems in the deeply integrated Union.
It was already clear in the 1960s that member states were actually dependent on technical administrative support from the Commission. This marked the inception of the rapidly expanding comitology system that comprises some 200–300 expert committees of EU and national public officials to date (Vos, Reference Vos2009). While comitology has served as one of the prime examples of the ‘fusion’ of different levels of EU governance (Wessels, Reference Wessels1998), more recent analyses go much further when classifying administrative procedures. Over more than five decades a toolbox of instruments has developed that renders the distinction between direct and indirect administration unsuitable for the EU context (Hofmann, Reference Hofmann2009). The assumption of indirect administration is formally sustained, although various other forms of direct, shared or common administration have proliferated (Ziller, Reference Ziller2006) and in real bureaucratic practice the notion ‘is a simplified model that no longer can be maintained. Instead, the different policy fields have developed along a spectrum of joint administration in which pure ‘direct’ administration by EU institutions on the one hand and pure ‘indirect’ administration by MS [member state, the author] authorities on the other hand represent the two extreme ends. In reality there are hardly any examples for either direct or indirect administration without any forms of cooperation between the national and the EU levels’ (Hofmann and Türk, Reference Hofmann and Türk2006: 90). In short, public administration builds on a mix of non-hierarchical and hierarchical relationships that are both explicitly and implicitly concerned with the organization and administrative practice of domestic civil services.
The dominant actors in this case are the actual administrators, namely officials in the Commission and national or subnational bureaucracies, and a growing number of delegated agencies (Wonka and Rittberger, Reference Wonka and Rittberger2010; Egeberg and Trondal, Reference Egeberg and Trondal2011) who interact continuously in EU policy-making. Which specific relationship comes to bear varies substantially between different policy fields. In policies that are harmonized under EU law, the Commission has some far-reaching powers to inspect, supervise and monitor compliance with EU legislation (Cassese, Reference Cassese2004). To provide but one example, although agricultural payment agencies are national bodies, even the way they administer funds is narrowly prescribed by EU rules. Policies that are not harmonized but in which mutual recognition applies imply a different form of administrative cooperation, because domestic civil servants need to incorporate the rules of other member states. Here, the most far-reaching implications were introduced with the Services Directive, which introduced an obligation on administrative cooperation for the first time (Schmidt, Reference Schmidt2009). To illustrate, in order to enable the free movement of service providers, national bureaucracies need to recognize qualification certificates issued in other member states, which is one of the issues for which shared administrative tools have been set up to horizontally link the member states’ bureaucracies (e.g. by opening national registers to partner administrations). Community agencies, finally, have gained significant relevance in recent decades. Delegated executive and coordinating tasks, they form a third group of actors beside public bureaucracies. These examples provide only a very limited insight into the multiple common procedures. For the purpose of this article, it suffices merely to hint at the various facets of interlinked network administration. In brief, in order to practically implement EU policies, administrative networks have grown substantially beyond indirect administration and de facto a wide range of mixed administrative procedures dominates EU policy execution.
This far-reaching pragmatic preoccupation with public administrations on the cognitive level is not matched by symbolic framing. Especially where pragmatic ideas clash with well-established normative frameworks, framing is vital for re- or decoupling strategies. Illustrating the former in their study on the development of the Single Market Program of the European Community, Fligstein and Mara-Drita demonstrate how Commission President Delors succeeded in reviving the stalled integration process by framing the measures intended to better facilitate the exercise of the four basic economic freedoms already applicable among member states. They show how the Commission argued ‘within the constraints of existing institutional arrangements, provided a “cultural frame” and helped create an elite social movement’ (Fligstein and Mara-Drita, Reference Fligstein and Mara-Drita1996: 1).
In contrast to this feedback of symbolic frames into normative ideas by national policy-makers in the area of market integration, symbolic frames on European civil service have not led to normative moves among elites to legitimize an integrated administration to their domestic constituencies. What we observe is rather the incremental development of program ideas that are not matched by normative shifts but rather inverse framing that reinforces existing public sentiments. Concretely, civil service in the context of the EU is predominantly framed as a matter of the ‘Brussels bureaucrats’ (Stevens and Stevens, Reference Stevens and Stevens2001), represented by the Commission which is ‘widely perceived as a technocracy’ (Moravcsik, Reference Moravcsik2002: 604–605) – and with the term ‘Eurocrat’, referred to even by convicted proponents of the federal European idea (Spinelli, Reference Spinelli1966; see also: Pinder, Reference Pinder2007) and which has found its way into major dictionaries, that define a ‘Eurocrat – a staff member of the administrative commission of the European Union’ (Merriam-Webster). More so, some 44 percent of Austrians, Finns, Danes, Swedes, and Germans consider the bureaucracy as the symbol of the Union par excellence (Commission européénne, 2010: 139). There is a considerable mismatch between the actual strong role and considerable leeway domestic public administrators have in implementing EU policies, and the diffuse perceptions of the bureaucratic structures of the EU, framed almost exclusively as an isolated Brussels malaise, which leads Shapiro to conclude that ‘[i]ndeed, anti-E.U. sentiment is far more often expressed as anger with the Eurocrats than as distaste for its explicitly political organs’ (Reference Shapiro2005: 347).
Summarizing the empirical picture, the observable expressions of the underlying ideas highlight a stark discrepancy between the normative and cognitive levels. The dominant public sentiment of national bureaucratic autonomy in the background is reflected in the symbolic framing of the EU administration being limited to ‘Brussels’. This image stands in sharp contrast to the highly interconnected administrative networks that execute EU policy in point of fact. It appears foremost in the great number of policy instruments applied to realize vertical and horizontal coordination between mutually dependent administrators. In the background, these pragmatic solutions based on concrete program ideas link to paradigms, that is, elite assumptions that circumscribe useful solutions available to policy-makers. Within the multilevel administrative setup these paradigms have incrementally evolved in the form of basic legal principles and the notion of an integrated administrative space. The analysis clearly shows that although paradigms and programs correspond on the cognitive level, they remain decoupled from frames and public sentiments on the normative level. This decoupling creates leeway to develop even far-reaching coercive EU policies, such as the demands for national administrative reform the Commission made of candidate states in the enlargement context.
Conclusions
The analysis of the linkages between policy ideas and policy instrumentation within the European administrative space shows a stable structure for policy-making in a context of contradictory normative and cognitive ideas. The policy paradigms that motivate policy-makers to further supranational cooperation compete directly with the public sentiments of administrative sovereignty of domestic civil services. The competing ideas of a supranational entity and national sovereignty are a constituent feature of the EU, one deeply embedded in the Union's institutional architecture, and not a transitional condition, which will in the short- or mid-term be decided in favor of one or another idea. Therefore, policy-making has to adapt to the context of contradictory cognitive and normative ideas. In the case of administrative policy, this balancing act is achieved by decoupling. In the background, cognitive policy paradigms are decoupled from public sentiments, and symbolic frames are decoupled from actual policy instruments. In the foreground, actual policy programs of increasing shared and integrated administration are detached from a symbolic framing that presents Brussels bureaucrats as completely detached from national civil services. Accordingly, public sentiments of domestic administrative sovereignty are responded to by framing the supranational public administration as an isolated Brussels phenomenon, while at the same time pragmatically strengthening a common administrative framework. The empirical analysis of the policy instruments based on the four types of policy ideas confirms a dominant pattern of normative and cognitive decoupling. This arrangement moreover proves stable over time, as the most recent reforms introduced by the Treaty of Lisbon confirm.
More generally, when confronted with competing normative and cognitive ideas, decoupling offers policy-makers an alternative to discursive strategies that aim to reconcile public sentiments and policy programs. Obviously, decoupling has its limits. Politicization that focuses the public debate on the question of consistency between policy programs and shared values and attitudes will inevitably raise elementary questions of legitimacy.Footnote 4 Furthermore, decoupling programs and frames implies problems of control and accountability because public awareness is directed away from the actual implications of policy programs. At the same time, blame for unpopular decisions can be shifted, which ‘is not just a problem for the EU in gaining acceptance among European citizens, and it does not affect the legitimacy of the European Union alone. It also raises questions about sufficient domestic accountability. If governments can contrive to blame the EU for what are, in fact, the deficiencies of their own domestic policies, then the lack of transparency at the European level has the effect of undermining the adequate democratic control of domestic institutions’ (Kumm, Reference Kumm2008: 116–117). Policy instruments often blur legal categories and the clear attribution of political responsibility. Multilevel governance is particularly prone to a decoupling of cognitive and normative ideas through policy instrumentation, because legitimacy is produced nationally while action is exercised on the supranational level. However, the use of policy instrumentation as a strategy of decoupling competing ideas should apply to policy-makers in any democratic polity. Returning to Weber's imagery: in the EU, governments’ conduct is guided by ideas and material interests but the two lack ‘world images’ that, like switchmen, determine a single track of action. Instead normative and pragmatic ideas are decoupled to run on different tracks.
Acknowledgements
The author would like to thank in particular Giliberto Capano, Michael Howlett, and Jeremy Rayner; Claudio Radaelli and the other participants of the ECPR workshop on policy ideas and instruments for their valuable inputs during the joint sessions in Münster in 2010. In addition, his special thanks for the thoughtful comments to the three anonymous reviewers.
