A development towards a �European Administrative Space� (EAS), understood as convergence on a common European model, can be seen as a normative program, an accomplished fact, or a hypothesis. Here, focus is on the hypothesis and the search for analytical models with explanatory power. The approach raises several questions: What is �convergence� and what criteria can be used to decide whether an EAS exists? To what extent has there been convergence? Which aspects of administrative systems have converged and towards what? Through what processes have convergence taken place and how can it be explained?

�Administrative convergence� is a term without a clear and agreed-upon meaning (Pollitt 2002), but convergence on a common model implies a reduction of variance and disparities in administrative arrangements. Different administrations develop along the same path in a way that produces more homogeneity and coherence among formerly distinct administrations. A European Administrative Space denotes that public administration operates and is managed on the basis of common European principles, rules and regulations uniformly enforced in the relevant territory (Cardona 1999, OECD/Sigma 1999: 15). [1]  Common, or similar, arrangements drive out �local� principles, organizational forms, rules, standards and practices. A development towards an EAS stands in contrast to national administrative systems as �solid bedrock for nationalism�, that is, idiosyncratic arrangements where the structure of public administration reflects the identity, history and traditions of a specific state and society (Nizzo 2001: 2).

The aspiration of the paper is to suggest a way of thinking about administrative convergence, more than to give a firm answer about the extent of European convergence. Answers are of limited value if they are not based on theoretical ideas about how public administrations change over time, including ideas about under what conditions administrative convergence is likely and what factors influence change.

The hypothesis that a specific European Administrative Space is emerging will be held up against two competing, or supplementing, hypotheses: a global convergence hypothesis and an institutional robustness hypothesis. Recently, developments in public administration have been interpreted by means of two generic models: the �classical� or weberian public administration and the New Public Management (Verheijer and Coombes 1998, OECD/Sigma 1998a, Goetz 2001b). A favorite diagnosis has been a paradigmatic shift �from Old Public Administration to New Public Management� (Dunleavy and Hood 1994). The NPM, however, stands in contrast to the idea of a unique European convergence. It suggests that convergence is global, or at least common for OECD-countries (Pollitt 2002). NMP also assumes an �inevitable shift rather than a temporary fad� and that the change represents progress toward a more advanced administration (Osborne and Gaebler 1992: 328). Furthermore, administrative change is part of a development from �government� to �governance�, or governance without government, both within (Rhodes 1997) and among states (Rosenau and Czempiel 1992).

The vision of a European or global convergence can be supplemented by the institutional robustness hypothesis. [2] Here the basic assumption is that the two others overestimate the likelihood, extent and speed of convergence, and that Europe and the rest of the world are likely to continue for quite some time with a variety of administrative models. Furthermore, both the �Old Public Administration� and the �New Public Management� portray the administration as a tool for an external principal - a branch of government controlled by the legislative and judicial branches, or by shifting external circumstances. In contrast, the robustness hypothesis assumes that administrators are powerful actors in public policy making and administrative change. Likewise, public administration is a collection of partly autonomous institutions with identities, traditions and dynamics of their own.  

Under what conditions can we then expect administrative convergence? An old, and recurring, idea is that there is a single best way of organizing administration. Both scholars and practicians have looked for, and prescribed, context free, universal organizational forms that because of their comparative efficiency are expected to spread all over the world, independent of national, and other, differences. [3] There are, however, good reasons for being skeptical to universal prescriptions that are not time- and place-bound. Such principles tend to have a proverbial character (Simon 1946) and empirical research suggests that  �good administration� and �good governance� hinges on specific definitions of ends, purposes and values. If so, there can be no truly universal generalizations about public administration without a profound knowledge about the varying political and social characteristics that impinge on the administration (Dahl 1947, March 1997).

An alternative idea is that contexts matters and that one context is dominant in a specific time period. One can, for example, assume that growing international interdependencies and improved communication currently make diffusion of �best practice� increasingly efficient, globally or in Europe, reducing the significance of historical and cultural contexts as well as the distinction between public and private administration. A third possibility is to assume that environments are more or less homogeneous and imperative, administrative reformers are more or less powerful and institutions are more or less robust. Then we have to understand the conditions under which introducing, modifying or abolishing rules or material incentives will govern administrative conduct and change, the role of shifting contingencies, as well as the conditions under which existing institutional arrangements contribute to continuity or differentiation, rather than convergence. Likewise, we have to distinguish between attractiveness, where convergence emerge because one model is generally seen as superior, and imposition, where a model is preferred by a winning coalition and dictated to others.

Global convergence then can follow if administration is a context-free, technical activity with a single best solution, and if the global environment is currently dominant. European convergence can follow if context matters and the European context is dominant for administrations within, but not outside, the region. Robustness could be observed if context is not dominant, if administrations face different environments and established arrangements have some degree of autonomy.

 

The rest of the paper is divided into four parts. First, the theoretical ideas are developed a little further. Second, some observations of administrative convergence in the European Union are presented. Third, focus is directed towards enlargement and convergence among candidate countries, and fourth, some conclusions and puzzles are presented.

 

Administrative dynamics

Three hypotheses. The idea of an emerging European Administrative Space has roots in a European continental public law tradition and the �Old Public Administration�. The idea of a global convergence has affinities with the NPM, primarily an Anglo-American school of thought. The �Old Public Administration� and the �New Public Management� are different in many ways. The OPA assumes that the core of political life is law making, interpretation, implementation and enforcement. The public administration is a legally constituted instrument governed by law and hierarchy. Administrators are rule-driven bureaucrats executing and maintaining legal norms with integrity on the basis of Rechtsstaat principles, that is, in a neutral way and with the common good in mind. This perspective emphasizes reliability, consistency, predictability and accountability. Administrative dynamics is governed by legislators or announced and enforced by the courts  (Weber 1978: Ch. XI).  

The New Public Management portrays a centrally organized and rule-bound public administration as outdated. The public sector is not distinctive from the private sector and the slogan: �let the managers manage� implies fewer uniform, system-wide rules and procedures and more flexibility and decentralization. The public administration is primarily an instrument for efficient service production, governed by a performance-oriented culture with a focus on results, entrepreneurship and efficiency. An ideal model of private corporations operating in competitive markets inspires the NPM. Like everyone else, administrators are assumed to be self-interested actors governed by shifting material incentives and opportunity structures. In order to prevent opportunism and rent seeking, incentives have to be designed carefully. Important tools are market competition, auction systems, price systems, vouchers, contracts, performance appraisals, scorecards and benchmarking. Change follows from shifting circumstances, customers� demands and political objectives. A basic aim is to reform the public sector so that it will be �serving the economy better� and improve international economic competitiveness. [4]

The tension between a legal-bureaucratic and a managerial approach to public administration and the idea that private business administration provides an exemplary model for the public sector, are hardly new (Waldo 1956). Nevertheless, a focus on the differences between the two may disguise that they also share important assumptions. Both prescribe a single set of administrative principles for organizing the public sector. They assume that more efficient forms over time will replace less efficient ones, resulting in administrative convergence on an organizational model with superior technical efficiency, respectively a weberian bureaucracy and the private firm in competitive markets. Both are also based upon an instrumental or �machine� view of administration, shared by the management literature, public law and democratic theory. The administration is based on delegated competences and act as agents in conformity with rules, objectives and incentives originating from external principals. The principals� manipulation of such factors is supposed to generate intended and desired structural  and behavioral, as well as substantive results. 

In contemporary democracies administrative environments are not, however, so simple, coherent and imperative. They seldom provide public administration with clear competences, rules, objectives and incentives. On the contrary, the administration operates in a complex ecology of institutions, actors, goals, rules, interests, powers, principles, values, beliefs and cleavages. Politicians, judges, experts, organized groups, mass media and individual citizens are likely to hold different and changing � not coherent and stable - concepts of �good administration� and �good governance�. [5] They are likely to want the administration to serve a variety of changing and not necessarily consistent principles, goals and interests.  For example, administrators are expected to: 

·       ensure administrative loyalty to the incumbent government, and yet be politically neutral and loyal to all legitimately elected governments and the principle of representative government;

·       <![endif]> be effective policy entrepreneurs taking initiative and getting problems solved but also be accountable to elected leaders and the people;

·       <![endif]> be rule-driven and the guardians of the constitution, the rule of law principle and an impersonal legal arrangement, without ending up in excessive formalism;

·       <![endif]> act as experts and the users of the best scientific knowledge available - an administration based on merit, professional competence, loyal to the principle of enlightened government - but avoid technocracy;

·       <![endif]> be cost-conscious managers governed by the principle of economic efficiency, adapting to changing circumstances and consumer demands, using markets and price mechanisms, without producing social inequality, exclusion and protest;

·       <![endif]> protect specific constituencies and the principle that affected parties should be heard, without giving privileges to strongly organized interests.

Each concern is a possible source of legitimacy as well as criticism and tensions between them are a source of administrative dynamics. Throughout history conflicting role expectations have been reflected in struggles over the size, organization, procedures, competences, resources, recruitment and outcomes of public administration, as well as struggles over how the relations to other institutions, organized interests and individuals should be organized (Kaufman 1956, Jacobsen 1960,1964). The political-administrative order involves imbalances and collisions between principles and institutions (Orren and Skowronek 1994, March and Olsen 1989, Olsen 2001) and due to sequential attention and local rationality, where different institutions defend different principles and interests (Cyert and March 1963), different concerns are activated over time and across institutions. As the mix of concerns changes, so do conceptions of good administration and good administrators.

During periods of transition, in particular, conceptions of the exemplary administration are challenged and can be dramatically redefined. Conventional wisdom becomes heresy. Administrative virtues are turned into vices. Old expertise is scrapped and new types of knowledge, skills and training are demanded. Trust in institutions disappear or emerge. Organizational structures, roles and cultures are branded illegitimate and new ones are legitimized (Kaufman 1956, Jacobsen 1960,1964, March and Olsen 1989). Because tensions are enduring rather than temporary, any prescription based on hegemonic aspirations and the universalization of a concern is likely to foster criticism, countervailing forces and search for a new balance between institutions. Theorizing administrative dynamics requires understanding how balances are struck and administrations find their place in a political order.   

Such institutional balancing acts are usually constitutional and political and without a correct technical answer. Organizing public administration involves a power aspect. Long (1949: 257), for example, argued that �the lifeblood of administration is power� and Weber  (1978) observed that the political masters could easily become dilettantes facing a professional administration. An implication is that we need to avoid definitions of public administration that à priori assume specific authority and power-relations. [6] Historical processes have produced administrative systems and institutions with different identities, power balances and autonomies. Changing the public administration in order to create convergence implies intervention in large-scale, complex and dynamic institutions with pre-existing identities, structures and resources.

The institutional robustness hypothesis reflects such ideas and assumes that established administrative institutions are a key factor in understanding convergence. Administrative history is not �efficient� in terms of adapting institutions rapidly to environmental change or reform efforts. The most standard institutional response to novelty is to find a routine in the existing repertoire of routines that can be used (March and Olsen 1989: 34). External change and reform initiatives are interpreted and responded to through existing institutional frameworks. Convergence consistent with the identities and dynamics of institutions will be fairly smooth. Convergence in conflict with such factors will create conflict. [7] Sudden and radical change is most likely in situations with obvious performance crises. Change in reform rhetoric, formal rules and organizational charts, behavioral practices and substantive outcomes can be loosely coupled.

Convergence: attractiveness and imposition. Administrative convergence can take place due to attractiveness or imposition. Attractiveness signifies learning and voluntary imitation of a superior model. The receivers copy an organizational form because of its perceived functionality, utility or legitimacy. Likewise, a common model can emerge through joint deliberation, or each country facing the same challenges can independently develop similar solutions. Convergence as attractiveness is likely if a single administrative prescription is generally viewed as superior to other ways of organizing the public administration, globally or in the European context.

The perceived superiority can be based on technical-functional performance and comparative efficiency. For example, functional interdependence, as a result of economic or political integration, can lead to more interaction, exposure to others� solutions and agreement on a �best practice�. In a more moderate version variance is reduced as some organizational forms disappear because they do not perform in an acceptable way according to minimal standards. Existing institutions, budgets or staffs are not adequate to handle important tasks. There is erosion of problem solving capacity, performance is declining and crises have already occurred or they are expected. 

Superiority can also refer to normative attractiveness. Organizational templates then have a value of their own and a specific way of organizing public administration can be seen as most democratic, rational or modern even without evidence of its functional superiority. [8]   Then, convergence around appropriate organizational forms, processes and practices is likely if a single administrative form achieves a normative hegemony. An implication is that de-legitimization of forms and reduced pride in existing arrangements is likely to reduce resistance to change. Convergence might take place as external high status forms are copied without much analysis of their performance in the new setting.

Imposition signifies convergence based on the use of authority or power. A single model penetrates the territory and weakens or eliminates established institutions. When no single way of organizing public administration is seen as functionally or normatively superior, convergence by imposition is likely if there is a concentration of authority or power and if administrative issues are given attention and priority among the powerful. Authority signifies that some external source is accepted as legitimate when it comes to create, reform or dissolve administrative arrangements. Conditionality implies acceptance of organizational forms not for their inherent characteristics, but as a condition for membership, legitimacy or side-payments. Convergence based on coercive power signifies the ability to dictate solutions against the will of administrations. 

Understanding convergence through imposition requires a focus on the distribution of relevant resources and processes of mobilizing support. For example, if it is accepted that the main challenge is not to rewrite legal texts but to change and improve the actual working of public administrations (Ziller 1998: 138), we need to observe that there may be important differences in the resources needed to win battles over texts and formal arrangements and winning battles over behavioral practice, results and administrative identities and cultures. The latter is likely to have larger and more complex constellations of institutions, actors, cleavages and power relations, making convergence less likely or slower to achieve.

 

Conditions for an EAS. The European context has several characteristics that could be expected to promote administrative convergence and a European Administrative Space, but also a number of properties that could counteract this trend. One such factor is that the European order is characterized by long, strong and varied institutional histories, with different trajectories of state- and nation-building (Rokkan 1999). As a result, public administration structures and regulations vary among the current member states (OECD/Sigma 1998: 13) - an institutional pattern that is likely to create variations in both the motivation and capability for institutional convergence (Olsen and Peters 1996, Pollitt and Bouckaert 2000).

Traditionally the EU has also given modest attention and priority to administrative issues. There has been more focus on policy making and substantive results than on administrative arrangements. [9] The costs of changing the organization of domestic administrations have been seen as high (Hanf and Soetendorp 1998). The Commission crisis, reaching a climax when the Santer Commission resigned in the spring 1999 after accusations of mismanagement, fraud and nepotism (Committee of Independent Experts 1999a,b) moved administrative reform to the top of the Commissions agenda. Still, the Union has not developed an explicit and coherent administrative policy (Sverdrup 2002a).

The Treaties do not prescribe a specific model of administration and there is no acquis communautaire for public administration. On the contrary, the EU has throughout its history assumed that a variety of national administrative systems are legitimate and compatible with membership, and that different arrangements can do equally well in implementing EU legislation. [10] Furthermore, the Union has limited powers in administrative matters. Debates over the distribution of administrative competence have been linked to struggles over the general distribution and separation of powers in the Union. Member states have guarded their autonomy and have been reluctant to grant general organizational, supervisory and enforcement competence to European institutions. Most of them have been satisfied with laws that give the Commission direct administrative responsibilities only in specified domains and otherwise assume that the EU will not interfere with the internal administrative organization of the member states �as long as Union obligations are followed and common rules are implemented. Administrative instruments that leave discretion to member states have been more popular than those imposing specific administrative solutions (Schwarze 1996, Cardona 1999, Kadelbach 2002).

In addition to limited legitimacy the Commissions administrative capacities in terms of staffs and budgets are modest compared to its tasks. The Union�s administration is a �political administration�, in the meaning of a staff with significant discretion (Joerges and Dehousse 2002). Administrators are, however, part of complex networks across sectors and levels of governance involved in a joint exercise of coordination and authority. [11] Usually implementation takes place through national administration or by shared management. Sometimes it is organized through outside contractors such as firms and organizations. The growth in specialized autonomous European agencies has also been combined with a reluctance to give them discretion and national representatives have been placed on their governing boards (Chiti 2002, Dehousse 2002). As a result, it has been suggested that the idea of an administration �serving some common higher �authority� cannot be upheld in a non-unitary, non-state pluralist polity� such as the EU (Joerges 2002: 21). The conditions make the idea of a single �Principal� controlling administrative convergence less plausible.

There are also actors and forces in favor of convergence. Although no grand implementation crisis has emerged in the Union, the issue of an implementation deficit has been raised (Sverdrup 2002b). Member states� preference for administrative autonomy has to be balanced against the Union�s need for effective and uniform implementation. For the Court and the Commission the role as guardians of the Treaties provides a stronger power-base than their limited formal administrative policy competence (Kadelbach 2002). A high level of interaction and exposure to each others administrative thinking and solutions could also be expected to contribute to an EAS. European administrations have for centuries studied and copied each other in administrative matters and they share important characteristics (Barker 1944, Markussen and Ronit 2002). Nevertheless, increased economic and political integration, interdependency and interaction have created a qualitative new situation. Because institution building and policymaking in the Union have been unevenly developed across institutional spheres and policy areas, the adaptive pressure on national administrations is, however, likely to vary (Knill 2001).

Finally, European convergence may be more likely because the international normative reform climate has changed. The institutionalized pressure for global administrative convergence has weakened and the support for NPM is less enthusiastic than a decade ago. For example, the World Bank now expresses less firm beliefs than they did in the early 1990s. Then they �knew� that a paradigmatic shift was needed and the prescribed medicine included a shift from a centralized, hierarchical, rule-driven administration to one characterized by management and market orientation (World Bank 1991: 38). Now, the need for in-depth understanding of the specific situation in individual countries is emphasized. There are few answers that are right under all circumstances. No single recipe or imported one-size-fits-all notion of �best practice� will do. A good public sector is no longer a minimalist one. Rather, it has to match its capabilities (World Bank 1997:  iii, 11, 2000: xv).

In sum, the European context suggests that administrative convergence is more likely to follow from attractiveness than from imposition. Convergence is also more likely to be an artifact of substantive policies than the result of a coherent European administrative policy. If so, convergence could be expected to be incremental and putting constraints on national administrative solutions, rather than generating a full-scale exemplary model.

Observed administrative convergence. There are a limited number of systematic studies of European administrative convergence, a fact that invites tentative rather than firm conclusions. There are also few references to an EAS in Union documents and in speeches by European political leaders. The concept is mostly found in studies of public law and reports from OECD/Sigma in conjunction with the enlargement of the Union. Furthermore, there is no consensus in the literature on the extent of convergence on a common European model. On average, studies in administrative law report more convergence than studies by empirical social scientists.

A report from the Thematic Network in Public Administration argues that European public administration has a specific nature and that a gradual convergence of public administration systems in a �European Administrative Space� is under way. [12] Nizzo also concludes that �a �European administrative space�, in which the areas of convergence exceed the divergences, does now exist� (Nizzo 2001: 7). Likewise, a report from SIGMA observes that while an EAS is not an established fact, the driving forces of convergence are gaining speed and power. It is asked whether the process has reached a point where diversity is replaced by European administrative standards (OECD/Sigma 1999: 15).

A key argument is that while there is no formal acquis communautaire regulating European public administration, �a non-formalized� acquis has emerged. Member states with different legal traditions and different systems of governance have developed a common body of doctrine and shared understandings of principles of administrative law, standards of good practice and the requirements for unitary and efficient implementation of EU laws. The implementation of specific policies, and the need to achieve specified objectives and results, has required the member states to change national administrative law systems and modify administrative principles, structures and procedures (Kadelbach 2002). New administrative bodies and positions have been instituted and the concept of public administration, the public sector�s scope and extent, as well as its role in economy and society has converged. Calendars, deadlines, statistics, frames of reference etc. have been modified (OECD/Sigma 1999: 3, Sverdrup 2000, Nizzo 2001: 5). Such developments are illustrated by a study of the process through which the scope of exemption of �employment in the public service� was defined, that is, the areas where member states can impose restrictions on the free movement of workers and discriminate on grounds of nationality. Here a functional (type of work), not an institutional, definition was developed, national conceptions of �public service� were redefined, and during the process member states� preferences changed towards more willingness to recruit foreigners (Dimitrakopoulos 2001).

There are several nuances in the conclusions. For example, Schwarze (1996: 800) argues that the administrative structure appears to be destined to be particularly resistant to European influence. His study of 12 EU member states documents that there are substantive and structural differences between administrative law systems and he predicts that national administrative orders will not lose their particularized and dominant role in the near future. Nevertheless, national administrative law has undergone �remarkable change� under European influence, creating considerable coherence. The organizational structure of the administration has been more influenced in some member states than in others and absence of a national tradition has facilitated European influence.

Ziller observes that member states look to each other for inspiration. Becoming a EU member also means accepting some common administrative standards. Yet, Ziller makes a distinction between approximation and harmonization and argues that there is neither a project nor a need for having identical administrative systems and institutions throughout the Union (Ziller 1998a: 137).

Based on a study of the public service of the 15 member states, Bossaert and colleagues are even more reluctant to accept an EAS as an established fact. European law making and case laws in substantive fields have produced administrative change �more or less through the back door� (Bossaert et al 2001: 3, Demmke 2002: 11). However, diversity and specific national features still characterize public administration, as the part of the political-administrative system most strongly influenced by specific national traditions. There have been change but not harmonization and convergence. �We cannot yet speak of a European Administrative Space, although trends can be seen towards a Europeanization of the public services� (Bossaert et al. 2001: 248, 251-2). 

The approximation and harmonization of administrative law systems in the Union has primarily been the work of judges and lawyers (Shapiro 2001). It has taken place via the case law of the European Court of Justice, legislative acts and voluntary adaptation. An important challenge has been to develop a balance, which preserves the autonomy of national administration and at the same time, ensures that Community rules are administered properly (Cardona 1999, Nizzo 2001, Kadelbach 2002). Convergence on a European model is seen as driven by the need for national administrations to obey the same laws, and a widespread interaction amongst civil servants and politicians; processes that are assumed to inevitably change the ways in which governments operate (OECD/Sigma 1999: 16, Nizzo 2001).

Conclusions about convergence depend on whether they are based on studies of legal texts or actual practices. For example, the European Ombudsman complains that major achievements, such as new rules on access to documents and good administration as a fundamental right of citizenship, are implemented in a less than perfect way (European Ombudsman 2002). More generally, Dehousse argues that there is a mismatch between law and practice in the Union and that traditional legal categories have �ceased to reflect adequately the actual operation of the EU machinery� (Dehousse 2002: 208). Dehousse and several of his fellow students of law conclude that an improved understanding of convergence requires detailed information about how the EU political-administrative system works in practice. [13]

Empirical studies by social scientists by and large report little structural convergence. They observe a significant and persistent shift of domestic attention, patterns of interaction and allocation of administrative resources to European-level institutions. They also find convergence in policies. Yet, the main finding is (although again with many nuances) that no new harmonized and unified European administrative system has emerged; no significant domestic convergence towards a common model has occurred; and no single prescription, in functional or normative terms, has achieved hegemony.

Governments and administrative systems have differentially adapted to European pressures on their own terms. That is, adaptation has reflected resources and traditions, the pre-existing balance of domestic institutional structures, and �the broader matrices of values which define the nature of appropriate political forms in the case of each national polity� (Harmsen 1999: 81). Likewise, a study of ten smaller West European states - both member and non-member states - concludes that adaptations to the EU has been influenced by existing institutional arrangements and traditions (Hanf and Soetendorp 1998). Institutional learning across national borders has been limited (Kassim 2000: 242, Maurer, Wessels and Mittag 2000). Competitive selection on the basis of comparative efficiency has been a significant process in some sectors, like telecommunication (Schneider 2001: 78). Yet, selection has not in general secured convergence towards a �best practice� across Europe (Harmsen 1999: 84). 

While the New Public Management has had a significant impact in several member states, its role in the Union has been modest. The �open method of co-ordination� developed around the Lisbon process and the Lisbon Council in March 2000 has in particular been portrayed as an example of a new mode of network governance. [14] Héritier, however, questions the importance and newness of the open method. She observes that it is used in a limited number of policy-areas and primarily in areas where the Union does not have a clear mandate and where its competence is contested. Usually there are elements of bargaining over targets and indicators - bargaining taking place in the shadow of government, hierarchy and legislation.  The Commission has also used the method earlier when moving into new policy areas where its competence was problematic, often as a first step towards legislation (Héritier 2001).  

In sum, European-level developments have not dictated convergence on a single form of administration. Goetz (2001a: 220,227) even concludes that the literature �casts some doubt over the explanatory power of �European integration� as a major force driving domestic executive change�. In the Union there has been no support for imposing a unitary solution. Sector specific policies and court decisions have generated change. They have constrained the structural options available, yet they have created limited convergence. Considerable discretion has been left to domestic actors and structural domestic diversity has persisted in spite of intensive interaction among administrators and competition between national models. European signals have been interpreted and modified through domestic institutions in ways that have limited the degree of convergence and homogenization. EU arrangements have been compatible with the maintenance of different national institutional structures and practices. Established national patterns have been resistant but also flexible enough to cope with changes at the European level. [15]

 

Conditions for an EAS. Many of the applicant countries were in a different situation than were member states. With respect to the Central and Eastern European countries (the CEECs), several factors could be assumed to make convergence likely, through attractiveness as well as imposition. Established arrangements were de-legitimized and the CEECs faced a bargaining situation characterized by asymmetry and conditionality. If performance crisis, power asymmetry and high priority of an issue by the powerful part are important causes of change, the applicant countries could be expected to change and possibly converge on a single model imported from outside. There were, however, also factors working against convergence.

The European Union did no assessment of the existing administrative systems during the preparation for accession in 1973, 1980, 1986 and 1995 (Ziller 1998: 138). In contrast, the EU has since 1997 placed administrative issues high on the enlargement agenda. The Union has coordinated administrative policy and accession policy and the administrative capacity to take on obligations of membership has become an important criterion for membership. Candidate countries have been put under pressure to modernize their administrations, that is, to develop a professional civil service and build institutional capacity to implement and enforce legal norms. [16]  

While some actors are proud of their historic achievements and do their best to protect them, others are eager to get beyond �the burdens of the past�. The domestic administrative arrangements of the CEECs were characterized by low performance and legitimacy and were in need of radical reform. The EU was also in a stronger position vis-à-vis the applicant countries than in relation to members. The Union controlled entrance and could provide economic help. It could therefore formulate conditions for membership, including administrative conditions. Countries that wanted membership and at the same time were less attractive members in terms of economic and administrative performance or democratic legitimacy were in particular in a weak bargaining position. They could be expected to be candidates for administrative convergence.

Probably, the most significant factor impeding convergence on a model transferred from the EU has been the absence of an agreed-upon exemplary organizational model and a coherent reform policy in the Union. The EU has wanted reforms and improved capabilities, but exactly what kind of administration and capabilities are needed has not been clear. The Union�s conditionality principle, defining the conditions that will make the administrations of applicant countries compatible with EU standards and able to implement the acquis, gives the most detailed guide to EU preferences. Specifications of the structures required were presented in an informal working document produced by the Commission in September 2000 and Grabbe argues that the document�s informal status, explicitly not committing the Commission, is indicative of EU�s uncertainty about how far to go in providing institutional criteria and preferences (Grabbe 2001: 1023). While the convergence criteria have become more precise over time, the administrative obligations of EU membership are still not well specified. Inconsistent signals are also sent to applicant countries from the EU and from different member states (Grabbe 2001, Dimitrova 2002). Furthermore, due to change over time the CEECs have had to adapt to a moving target of performance standards (Fournier 1998: 135).

Part of the explanation of the EU reluctance is also that the dilemmas facing reformers are real. There are no obvious and simple solutions. For example, ensuring predictability, stability and continuity has high priority, but reforms must not hamper later revisions and modifications, as external conditions continue to change. Therefore, stability has to be combined with flexibility, dependability with openness, and continuity with adaptability (Hesse 1998: 170). After some enthusiasm with New Public Management principles, going �back to basics� and weberian bureaucracies have been assessed as more attractive. Several reports by OECD/Sigma have advised the CEECs not to copy business methods and NPM reforms in Western Europe. The CEECs have been recommended to �recognize their own potential� rather than borrowing from the West. Each country has to find its own way forward (Fournier 1998: 135) and reforms have to be adapted to tight budgetary constraints and a short time frame. It is �impossible to simply adopt Anglo-Saxon administrative cultures� and such prescriptions are likely to have �detrimental� and �disastrous� consequences for the CEECs  (Fournier 1998: 129, Hesse 1998: 176, Metcalfe 1998: 61).

Another factor is the expressed norm that the EU �is not entitled to set conditions for future members to which current members are not subject and which they would find unacceptable� (OECD/Sigma 1998a: 15). Imposition would also reduce the trust in, and legitimacy of, administrative arrangements in the CEECs (Dimitrova 2002). Still, the Union is seen to have a right to demand administrative solutions compatible with membership. It is legitimate to suggest �alternative models to serve as possible reform objectives� based on the institutions and practices of member states and other modern democracies. While the EU constitutional and administrative systems differ, it is argued that there are a limited number of models and they share important features (OECD/Sigma 1998a: 15).

The general tendency then is to portray administrative change as a question of attractiveness and introducing common standards of good practice. The role of the Union is to provide inspiration, give guidance and work as �catalysts� or �accelerators� of change. Convergence is not portrayed as a question of power and imposing a particular European model (Fournier 1998: 126, OECD/Sigma 1999, Nunberg 2000: 2).

Observed administrative convergence. The administrative development in the CEECs is an unfolding process, and Goetz and Wollman (2001:882) observe that �the present configuration of executive arrangements in the CEECs is a still fluid amalgam of inherited, imported and domestically developed institutional arrangements�. Administrative capacities have been built and a variety of principles, rules, standards and regulations shared by European democracies have been introduced, constraining the set of administrative options available to the CEECs.

Nevertheless, so far convergence has materialized to a limited degree. It is also argued that West-East knowledge transfer has had a marginal impact. The efforts of Western �management gurus� and �academic tourists� have been of little use and dissociation of law making from implementation has created a mismatch of intentions and real changes. New legislation has been adopted even when it was recognized that the administrative preconditions for giving effect to laws were lacking and could not be created in the near future (Hesse 1998: 175-6). The administrations of the ancient regime have proved surprisingly resistant to radical transformation. Practices from the communist era have survived �remarkably undisturbed, yet in de-ideologized form and the �new civil servant� is a rehabilitated version of the �old bureaucrat� (Nunberg 1999). Older national traditions have sometimes been mobilized to replace more recent and de-legitimized arrangements.

Lippert, Umbach and Wessels (2001) and Goetz (2001b) suggest that structural continuity in the applicant countries has primarily been a result of the weakness of EU influence. Several factors other than EU policy have influenced administrative developments in the CEECs, but the existing institutional differentiation and robustness of domestic institutions and traditions are not seen as decisive. Limited resources have, however, constrained reform effort.

In sum, convergence has been modest in spite of the presence of several factors that could be assumed to work in favor of such change. De-legitimized institutions and power asymmetry have been counteracted by the lack of an exemplary model agreed upon by EU-members, the absence of a coherent administrative policy in the Union and the EU norm that domestic organization should be controlled by member states. Of course, the time frame used may also be too short to conclude about convergence.

To what extent is there then support for the European convergence hypothesis and the vision of an emerging European Administrative Space? Variance and disparities have been reduced and convergence has occurred. Common institutions have been established. Domestic ones have been adapted to European standards and administrative law has been harmonized. Nevertheless, there is also continuity and differentiation and convergence has often involved discovering and codifying what member states� administrations already had in common. Attractiveness has been more important than imposition. Incremental constraints on the set of acceptable administrative solutions have reduced variance and may in the long run accumulate into an EAS. Yet, no unitary model has been imposed or has emerged.

Several factors make it difficult to give clear answers. In the literature, there are no shared concept and indicators of �administrative convergence�. There are few comparative empirical studies and when different mixes of convergence and continued differentiation are observed, no shared metric exists to decide to what degree convergence has taken place. Conclusions vary according to the countries studied, and because different parts of public administration have different dynamics, conclusions also depend on which aspects are studied. Examples are whether focus is on administrative law and formal organizational structures, objectives and standards, administrative doctrines and codes of conduct, roles and identities, behavioral practices and standard operating procedures, budgets, the size of the work force, recruitment criteria, careers, training, remuneration, labor relations or outcomes.

An institutional focus, therefore, must not be allowed to overshadow that administrative convergence depends on changes in administrators as well as in structures, rules and resources. The NPM assumes self-interested, utility maximizing administrators. The OPA assumes administrators socialized into an ethos of rule following and public service. Understanding administrative convergence, however, requires longitudinal studies of the processes through which a sense of administrative identity and role is developed, lost and redefined. Such understanding also requires insight into the conditions under which the administration is likely to develop European and domestic identities and internalize democratic, constitutional and professional norms. An example would be whether legal rules work as external incentives affecting behavior through calculation of the expected utility of available alternatives, or whether rules have normative power. In the latter case administrators are governed by internalized codes of right and wrong, true and false, legal and illegal and they act on basis of a logic of appropriateness. For democratic governments it is easier and more legitimate to change formal administrative, legal and political structures than to change moral structures and mentalities that influence whether people voluntary comply with rules of conduct. Yet, ignoring the role of change in moral and causal beliefs is likely to create surprises and disappointments in the CEECs (Cepl 2000) as well as in the West.

A firmer conclusion can be drawn when it comes to the global convergence hypothesis. The hypothesis gets little support. The NPM is part of the administrative rhetoric and practice in several member states. Still, it has played a modest role in European convergence and no general trend from �Old Public Administration� to New Public Management can be observed in the Union. The different perspectives of the OECD/SIGMA and OECD/PUMA reform programs, promoting respectively the �classical� model and NPM, illustrates the need to contextualize theories and principles, make their basic assumptions explicit and study their scope conditions. These differences also suggest a need to take into account sequential and cyclical attention to legitimate concerns, such as stability and flexibility, innovation and accountability, capacity building and slimming down, substantive results and constitutive normative principles, due process and rights. Sigma has advised the CEE countries to give priority to the rule of law, accountability, reliability and predictability and trust building over efficiency.  PUMA has in recommendations to western countries given priority to economy and efficiency. It would be no surprise if priorities change in both regions over the next few years, among advisers as well as governments.

For students of administration European developments revive old issues, such as the need to take the historical and cultural context into account and develop a specific public sector organization theory, as a counterweight to an a-historical and a-cultural, private sector management inspired NPM. Public administration is in the midst of a balancing act where its institutional role is redefined and reorganized. It is part of a long-term process of reorganizing inter-institutional relations and re-defining democratic and constitutional ideals in a multi-level and multi-centered Europe. The transformations involve a variety of change processes.  Even if some actors are more influential than others, no single �principal� is likely to control administrative convergence by redesigning laws or incentive structures.

The institutional robustness hypothesis differentiates between change that is consistent, or not, with an institution�s identity and between situations characterized by a performance crisis, or not. Verification/falsification is difficult because these distinctions do not predict precisely when convergence is expected to take place. The hypothesis is supported by the fact that many elements of the public administrations have survived both in member states and applicant countries. A shared European context, common policy making and extensive interaction have produced modest convergence. The often-alleged inflexibility of public administration has, however, not been observed. Administrations have shown willingness and ability to change, rather than rigidity, insulation and insensitivity to environments.

One reason may be that European involvements have provided interesting tasks and made administrators more central participants in policy making � changes that are unlikely to collide with institutional identities and impede change. Administrations have succeeded in coping with changing environments, but they have done so in ways influenced by existing administrative arrangements. In a world of weberian bureaucracy it is no surprise that formal structures are maintained, since it is irrelevant whether rules come from Brussels or national capitals. Yet, the institutional robustness hypothesis is supported by the observation that member countries have been able to cope with common challenges, both in Brussels and at home, through their different ways of organizing public administration. 

Observations of European developments create two puzzles for students of formally organized institutions. A foundational belief among students of formal organizations, as well as practicians, is that �organization matter�. The assumption is that purposes and substantive rationality are embedded in formalization. Well-organized institutions, working properly, govern behavior and achieve in practice the purposes they were designed to serve. Designing and redesigning institutions change behavior and improve substantive results (Goodin 1996, Offe 2001, Stinchcombe 2001). In this perspective, students of administrative convergence face two puzzles: first, loose coupling between formal organization and behavior and, second, loose coupling between formal organization and substantive results.

The first puzzle is that stability in organizational structures and behavioral flexibility have come together. Intense interaction among European administrators, changed patterns of attention and resource allocation have taken place without radical structural change and convergence. The observation of loose coupling is not new in the study of formal organizations. [17] It is well known that the degree of coupling, and the degree to which organizational structure govern behavior and shape results, vary.  On the face of it, the loose coupling of organizational structure and behavior also seems to support an interpretation of formal structures as �facades� and �myths� (Meyer and Rowan 1977, Brunsson 1989). A more careful consideration, however, shows that the observation in the European context is the opposite of the expectations of this literature. Here the prediction is that organizations facing changing environments will change their formal �facades� and protect their behavioral cores. In the European context formal organizational structures have been stable while practices have changed.

The second puzzle is that the European Union has been successful in terms of common policy-making and economic and political integration with rather modest administrative convergence. While it is uncertain exactly how unified domestic implementation of EU policies is, no implementation crisis has emerged. Differently organized administrations have been able to cope with new functional tasks and changing normative environments. They have succeeded in producing acceptable, and sometimes outstanding, substantive results under new conditions, basically within established administrative structures. 

Possibly, the two puzzles are false and due to the fact that students of European convergence lack adequate, fine-grained categories for analysing organizational structures, behavioral patterns, substantive results and their interrelations. It is, however, also possible that the puzzles are real. If so, they challenge students of public administration to rethink the varying relations between organization, behavior and outcomes and �how organization matters�.

Literature