74 Abstract The article pursues two related ambitions. Firstly, it seeks to give an overarching view of post-transition Hungary’s agency field by giving a qualitative description and analysis of the development of the legal-institutional framework, and by characterizing the proliferation of agencies with quantitative information. Secondly, and more specifically, the article explores certain aspects of agencies’ politicization; in particular, it seeks to identify the extent to which central governmental policies related to agencies serve the latent function of strengthening political control over agencies. The analysis corroborates the results of an earlier, more limited one: the hypothesis regarding the existence of such political control functions is supported by the empirical evidence. Keywords: Hungary, agencies, politicizatism, political control. AGENCIES AND THE POLITICS OF AGENTIFICATION IN HUNGARY HAJNAL György HAJNAL György Hungarian Academy of Science, Institute for Political Science, Corvinus University of Budapest, Budapest, Hungary Tel.: 0036-14-825.032 E-mail: gyorgy.hajnal@externet.hu Transylvanian Review of Administrative Sciences, Special Issue, pp. 74-92 75 1. Introduction The ambitions of this article are twofold: firstly, to give an overarching, synthesizing view of the development of the Hungarian agency landscape and secondly, to delve into the specific problem area of political control and politicization of agencies. The first ambition is justified by the more general ambition of the current special issue of the Transylvanian Review of Administrative Sciences; namely, to offer a comparative insight into the largely unexplored area of Central and Eastern European agencification. The latter ambition of this article is justified by the fact that the relationship between politics and agencies is one of the focal problem areas, on which the current special issue focuses. Moreover, this issue area seems to be particularly relevant in understanding the dynamics of Hungarian agencies. As to the empirical data utilized in the study, two sources deserve to be mentioned. The legal-institutional framework of agencification is analyzed mostly on the basis of the basic legal measures that have defined the field throughout the past 20 years. In addition, published analyses of different aspects and problems of the legal-institutional framework were relied upon (for an extended overview see Hajnal and Kádár, 2008; Hajnal, 2010a). In addition to that, systematized empirical data on Hungarian agencies were collected and analyzed. Information regarding organizational changes and functions was obtained from diverse sources of information, including government data files, legal data bases and other publicly available information (for technicalities see Hajnal and Kádár, 2008). This data-collection resulted in a database containing dynamic, longitudinal information (including data on ties between antecedent and successor organizations). This database – and thus the empirical basis of the study – currently extends to the 2002-2009 period. The significance of this data set is that there has been little systematic study of agencies in Hungary; almost all of the existing writings were confined to the analysis of specific legal problems or legal measures from a purely legal positivist (and practitioner oriented rather than academic) perspective. Empirical studies are rare, and with a few exceptions largely descriptive, single-case studies (Nyitrai, 1996; Budai 2008a-b; Papházi and Rozgonyi, 2008). Therefore the ability to characterize and analyze agency dynamics in broad, quantitative terms constitutes an important step towards a better empirical grasp of a number of key issues. The next section outlines the conceptual and theoretical framework of the study, and locates more precisely the above, broad research ambition in that context. The third section aims at giving an overarching overview of the dynamics of Hungarian agencies, while the last section presents the analyses related to the politicization of agencies. The article ends with a brief section on conclusions. 2. Agencification and political control of agencies 2.1. The quest for political control: previous findings on the driving motives of agency policy in Hungary A first question of agency research aims at explaining the phenomenon of politicians refraining from the exercise of some of their powers and delegating certain 76 competencies to agencies. In the North American and West European contexts a number of such explanations were proposed and, partly, supported by empirical evidence (for an overview see Christensen and Laegreid 2006; Pollitt et al., 2004, pp. 20-21). Empirical results are distinctly scarcer in the Hungarian context. Beside the generally modest volume of such research this scarcity has much to do with some deeper, more structural factors, which may be referred to as the lack of the agency concept in the Hungarian professional and academic discourse. Prior to 1989-1990 agency-like organizations operated in a context that was fundamentally different from the politico-administrative landscape of the 1990s and the 2000s (Beblavy, 2002). In a society where the vast majority of social and economic activities occurred within the realm of the state “whenever the political leadership wished to make the direction and control of a sphere more direct the affected tasks were declared state administrative tasks and these tasks were assigned to a newly created administrative organ” (Nyitrai, 1996 apud Papházi and Rozgonyi, 2008, p. 96). Therefore central government organizations in the Socialist era cannot be easily treated as the functional equivalents of today’s agencies and should therefore not be part of the “agencification story” – even though they meet the agency definition. The analysis of agencies and agencification in post-transition Hungary continues to pose some specific conceptual and methodological problems that are missing in other (Western) countries. In the 1990s and 2000s – although there was a broad public discourse about administrative policy and reform – central government administrative structures changed, with certain exceptions, according to short-term, and mostly party political or personal considerations. Terms such as “agency” or “agencification” – either in English or in Hungarian – did not appear at all in the public or professional discourse or only marginally. Neither did agencies in the strict sense – that is, characterized by some extent and form of contractual, performance-based relationship between the agency and its parent ministry – appear in the organizational landscape of Hungarian government. In sum, the ideal type “autonomous agency model” is practically non-existent in Hungary as one of its defining characteristics – output control and contractual arrangements, as opposed to hierarchy-based, input and process oriented command- and-control – was throughout the examined period absent. Moreover, agencies are not seen as a structural solution contributing to organizational or policy performance, the achievement of which would be based on the application of such means as extended autonomy and improved (ex-post) control mechanisms. Rather, agencies in a sense “happen to be there” as a result of some other driving factors not exposed or discussed either in the narrower professional or in the broader public discourse. Consequently, there is an idiosyncratic lack of correspondence between the conceptual framework (or cognitive map) related to agencies in Hungary on the one hand, and the “international mainstream” conceptualization, on the other. This lack of correspondence is probably characteristic for many other post-communist countries as well (see Moynihan (2006) in relation to Slovakia). It complicates a 77 discussion of (effects of) agencification, because the various effects attributed to the autonomous agency model are simply not expected by Hungarian policy makers and the broader public as they lie beyond the scope policy outcomes deemed relevant. Consequently the expectation that agency autonomy would lead to results like a substantial (positive) change in performance, transparency, coordination, credibility, accountability, legitimacy and so on is not discussed/analyzed in any significant form and extent in relevant government documents or academic research. Beside the above mentioned absence of the agency concept from the Hungarian discourse the reconstruction of the normative argument(s) underlying the creation and/or preservation of agencies is hindered by two additional problems. Firstly, as mentioned earlier there is little empirical evidence on patterns of agencification prior to 2002. And secondly, analyses, policy papers or broad policy statements with an even minimum extent of argumentative ambition and logic are extremely rare (a notable exception is Rozgonyi, 2008). However, despite the above bottlenecks four broad – partly hypothesized – motives of agency formation can be deduced from previous research done in this field (Hajnal, 2010a). Agencification may have been used as an anti-corruption and rule-of-law measure. Agencification usually affected functions related to the execution/enforcement of law. Moving such functions out of the ministry in charge meant that the agency became able to act as the administrative authority of first instance. In this new administrative setup the ministry could oversee administrative decisions of first instance as the administrative authority of second instance. This two-tier arrangement was expected to limit the possibility of “politicized” and possibly corrupt administrative practices. Moreover, it was seen as being more in line with liberal democratic, rule-of-law requirements vis-à-vis administrative decision-making. Agencification was in some cases imposed by external actors or donors. Moving (mostly implementation, information and support) functions out of ministries appeared as a frequent element of international technical assistance programs since for much of the period under study this was regarded as an integral component of the “international best practice”. Moreover, the EU accession process induced the creation of a limited number of agencies, as well (such as the ones in charge of distributing EU funds or of anti-discrimination measures). Agencification could be used to create more flexibility in HRM. There is a growing tradition in Hungary that – like in a number of other countries – government organizations are, from time to time, exposed to unexpected and harsh downsizing campaigns initiated from the top political leadership. These campaigns usually occur as responses to – or at least are usually justified by – acute budgetary crises. They involve most of all, an across-the-table cutback of personnel. As a counter-measure, ministries (and, to a lesser extent, other governmental organizations) have for quite some time been striving to create buffer organizations – most of all, agency type ones – where “downsized” personnel and management could be temporarily transferred to and retained until the downsizing campaign was over. 78 Finally, in contrast to the above motives having a chiefly “administrative” character, “agency policy” seemed to have served different, rather subtle, political ends as well. For example, the almost constant (re)creation and reorganization of agencies may serve as an instrument of politically controlling “broken-free” administrative apparatuses and as a means of building political and organizational Hintergrund. Moreover, agencies could function as a barter good in political transactions (for example, paying a political ally or neutralizing an opponent with a new, prestigious agency). Most of these motives – as well as numerous other ones – were, to a smaller or greater extent, exposed and examined in the international literature (for an overview see Pollitt et al., 2004, p. 20). In the following, I will focus on the latter, fourth set of triggering factors: the “political use” of agencies, or at least certain aspects thereof. 2.2. The “political use” of agencies As argued before, the “autonomous agency model” as it is understood in the international academic and practitioner discourse is, by and large, unknown in Hungary. Nevertheless – as the analyses presented above show – it is still possible and meaningful to examine “what has been going on” in relation to agencies, particularly for the past ten years or so. Most administrative reforms were not specifically related to agencies. But, throughout the past years there were some “grand themes” appearing – sometimes reappearing – in relation to ministries, agencies, local governments and other public organizations alike. Probably one – possibly the most frequent – of these grand themes was the politicization and the centralization of the central state administration. Increasing politicization was undergoing at least since 1998 (the starting year of the first Center- Right Orbán Cabinet). This development gained substantial impetus with the reforms initiated by the second (Center-Left) Gyurcsány Cabinet in 2006 extending, besides agencies, much or all of the central government machinery. These measures included, among others, the abolition of the permanent state secretary position in ministries; moreover, numerous other changes weakening the depoliticized, career type nature of the civil service system were implemented. The further centralization and strengthening of political control were to be achieved by, among others, transferring “sensitive” tasks to higher levels of the politico-administrative hierarchy. For example, public procurement, HRM and logistic functions – which have a key role in nurturing ministries and ministers’ informal, organizational power base (Papházi and Rozgonyi, 2008, p. 94) – were transferred to the Prime Minister’s Office. This pattern appeared on the level of agencies as well; many important tasks previously serviced by agencies were hived back into the parent ministries. This political and discursive context signifies the central role of the different, mostly informal and subtle, channels and fields of politicization in understanding the Hungarian agency field. In the following subsection, therefore, I will present a theoretical framework of politicization, which will serve as a basis for formulating the research question more precisely. 79 2.3. Politicization and political control: a conceptual framework Politicization of government apparatuses is often conceptualized simply as the politicization of the civil service; that is, as “the substitution of political criteria for merit-based criteria in the selection, retention, promotion, rewards, and disciplining of members of the public service” (Peters and Pierre, 2004, p. 2; Flinders, 2008). There are numerous attempts to broaden this conceptualization so as to embrace the many other ways, in which government apparatuses and thus the implementation of government policies can be influenced by informal societal, political or organizational values, interests and aspirations. For example, Eichbaum and Shaw (2008) differentiate between the extent to which (senior) appointees are or can be assumed to be partisan on the one hand, and the extent to which policy support and advice given to ministers by the (senior) civil servants are actually politically distorted. However, the focus of the current study is on organizations, rather than individuals, as primary units of analysis. Moving on to this higher level of analysis Peters (2001) differentiates between two kinds of politicization. Firstly, the interaction between „informal” political actors – pressure groups, societal interest groups and so on – on the one hand, and administrative apparatuses, on the other. Secondly, when administrative apparatuses interact with other, formally defined entities/actors of the state then what we have at hand is bureaucratic politics. It is this latter aspect of the politics-administration interaction, in which the focal issue of this study – political control – may be located. The formal political institutions of liberal democracies are organized around the concept of the chain of accountability – that is, elected politicians being accountable, through democratic elections, to the citizenry while administrative apparatuses being accountable to, and being controlled by, their political masters. In his research on (regulatory) agencies in four of the largest European countries Thatcher (2002) concludes that there are important channels of exerting direct, institutionalized, formal political control over agencies at politicians’ disposal. These include the nomination and early dismissal of senior staff, setting the resources – especially budgets – of the subordinate organizations, and the power to overturn (important) decisions of these organizations. Interestingly however his empirical analysis concludes that politicians – with some minor exceptions – don’t use their formal powers of controlling regulatory agencies. According to Thatcher this finding raises the question whether elected officials control agencies through “means such as creating resource dependencies and/or informal relationships” (Thatcher, 2002, p. 962). What are these informal channels of influence? Peters (2001, p. 244) goes into substantial detail in enumerating and delineating politicians’ – formal as well as informal – tools or, as he puts it, “ploys” of politically controlling bureaucratic apparatuses. These tools are as follows. – Special budgetary institutions (such as the Office of Management and Budget) controlling budgets and spending; – Organizational differentiation enabling executive politicians and legislature to follow the specialization and differentiation happening in bureaucracies; 80 – “Counterstaffs” providing independent sources of information (especially in political systems characterized by deep and lasting social, religious, ethnic or political divisions); – Managerialism – understood as the redefinition of the role of (senior) civil servants in a way that distances them from policy advice and especially policy formulation and, instead, emphasizes their role in the implementation of policies; – “Customer-driven government” understood as the redefinition of citizens’ role and entrusting them with some extent of control powers; – Control of staff, which involves what was earlier referred to as the politicization of civil service, that is, the totality of institutions, practices and techniques used by/enabling politicians to employ (senior) administrative personnel in line with their political/ideological dispositions, and to induce those in office to align with these values; – The existence of a strong ideological party, which especially – but not only – in single-party political systems may ensure compliance and/or reduce bureaucratic initiative; and – Finally, by threatening with the employment of physical coercion the military – constituting a special case of party governments – may increase, in certain political systems, the ideological and political compliance of the bureaucracy. Some of the above-mentioned “ploys”, such as the creation and use of special budgetary institutions, the politicization of civil service and – more lately, to some extent – customer orientation seem to be present in the toolbox used by politicians in Hungary. Other means, such as managerialism, “the Party” or the use of military threat are, by and large, absent from the Hungarian practice. On the basis of the sporadic qualitative evidence and earlier analyses performed on a limited data set (Hajnal, 2010a) it seems, however, that the above typology of the means and “ploys” employed by politicians does not include an important one: namely, the use of restructuring as a tool of political control. The ambition of the current paper is, therefore, to examine the Hungarian agency field in the view of the above analytical framework, in particular, to test the hypothesis that – in addition to those listed, with reference to the “Western” practice, by Peters (2001) – politicians in Hungary use another important tool of politically controlling agencies: the frequent reorganization of agencies. 3. Agencification in Hungary: the development of the legal-institutional framework and the proliferation of agencies The development of the legal-institutional framework of agencies in post-transition Hungary can be divided into two phases: the first one starts in 1990 and ends with the adoption of a new overarching legal framework in 2006; the second phase ends in the last year covered by the current study, 2010. These historical epochs are briefly overviewed in the below subsections. The third subsection characterizes the proliferation of agencies in quantitative terms, focusing on the 2002-2009 period covered by the data set underlying the current research. 81 3.1. Agencies between 1990 and 2006 The Hungarian central government subsystem is divided into ministries, the number of which ranged between 12 and 18 in the period under study. At the core of the central government structure lies the Prime Minister’s Office. Ministries are chiefly responsible for policy-making while most of the implementation tasks – especially those with a territorial dimension – are carried out by agencies (albeit that there are important exceptions to the above role, such as centralized public procurement, which is located in the Prime Minister’s Office). The local governmental system is a two-tier one involving, at the first tier, nineteen counties and the capital city Budapest, and, at the lowest, municipal tier almost 3200 local governments governed by elected councils. Local governments are responsible for another broad set of public service provision tasks, including child care, education, health, and local physical infrastructure services. In the Hungarian government apparatus there are various sorts of organizations possibly regarded as (and/or satisfying the definitional criteria of) agencies1. Because of the practical limitations of the research reported here – most of all, the rather dynamic nature of the institutional landscape and limited existence of, and access to, data – throughout the study the term “agency” is used in a somewhat restricted manner. By “agency” it is referred to public administration organizations (i) directly subordinated to the Government (i.e., either the Cabinet or a ministry) and (ii) falling into the organizational scope of Law on Civil Service. This latter restriction means that some clusters of agency type organizations that should “normally” be included in the analysis are omitted. These omitted clusters include: – the armed forces (military, police, customs guard, border guard and secret service organizations); and – certain public budget organizations staffed by public servants (as opposed to civil servants) exerting research, policy support, (higher) education, cultural, health, or social service etc. functions. Before 2006, there was no general legal regulation on the structural features of organizations in the executive branch. Competences and structures of individual ministries were regulated by various, and frequently changing, government decrees. The government had extensive freedom to establish the government structure within the broad limits specified by law and Constitution. Both academics and practitioners were dissatisfied with this unregulated nature of the central government organizational universe – and specifically that of agencies – and persistently argued in favor of common legal standards as a token of effective central government structure. It 1 The term “agency” is understood in a relatively narrow sense, as the sum of “Category 1” and “Category 2” agencies (van Thiel, this issue), whereby agencies are conceptualized as a public organization “(i) structurally disaggregated from the government and (ii) operat[ing] under more business-like conditions than the government bureaucracy” (Talbot, 2004, p. 5 apud van Thiel et al., 2009, p. 2). 82 is illustrative of the gross legal-institutional ambiguity that even ministry leaders are uncertain and legal documents are erroneous regarding the legal status of the organizations subordinated to their own ministry (Papházi and Rozgonyi, 2008, p. 86). Between 1990 and 2006, there has been only one significant event regarding the legal framework of agencies. The 2040/1992 Government Resolution (1992) created a three-category typology for agency-type organizations. However, the resolution did not classify the existing agencies accordingly. Moreover, the resolution was not to be applied in a consistent and compulsory way for subsequent organizational decisions. This measure could therefore not settle the problem of unregulated proliferation of agencies. In the following fourteen years a series of government resolutions repeatedly tried to eliminate opaque and overlapping agency structures and to develop and clarify the legal-structural principles underlying them (Hajnal and Kádár, 2008, p. 5) – with relatively little success. The typology outlined in the above-mentioned government resolution were, at least until 2006, also applied in the academic and practitioner literature dealing with this subject (Balázs, 2004; Sárközy, 2006; Vadál, 2006). It defines three agency types2: (a) the so-called OHSZ’s (országos hatáskörű szerv); (b) central bureaus (központi hivatal); and (c) ministry bureaus (minisztériumi hivatal). The list reflects a clear and institutionalized organizational hierarchy, going from the most to the least autonomous type of organization in terms of formal, legal-institutional autonomy, OHSZ’s and central bureaus being more structural autonomous (“Category 2” of the typology developed by van Thiel et al., 2009, p. 22), while ministry bureaus being less structural autonomous (being mostly “Category 1” agencies, ibid.). However, all three types of organization are separate legal entities with nationwide competence that perform some specialized public administration task(s). All organizations operate under the full authority of the appointed agency director. Some additional features of the three agency types are outlined in the table below. As the table shows increased structural autonomy – that is, the extent, to which an organization is able to entrench itself from its politico-administrative context – emanates from a variety of sources. Firstly, autonomy may depend on the “rank” of the politico-administrative entity deciding upon the agency at hand. For example, OHSZ’s are founded by the Parliament by a law; hence their statute (and any amendments thereof) involves a complex and highly visible legislative process involving numerous players and veto points. In contrast, central bureaus may be founded/modified/ended by Cabinet decisions while the fate of ministry bureaus can be simply decided by individual, across-the-table ministerial decisions. 2 Note that the category of ministry bureaus is, both conceptually and operationally, quite vague. Not only is this legal category not defined in the confidential and unpublished 1992 Government Resolution but it is also used in a somewhat vague and contradicting manner in the academic course. 83 Table 1: Typology of agencies based on their legal-structural features Structural features OHSZ(Type A) Central bureau (Type B) Ministry bureau (Type C) Founder/form of founding document (statute) Parliament / Law Government /Government Decree Minister / Ministerial Decree Superior organ Cabinet Ministry Ministry Appointment/dismissal of agency head - By the Cabinet / Prime Minister - Appointed for a term of 4-6 years - By the Minister - Appointed for an indefi nite period - By the Minister - Appointed for an indefi nite period Remuneration of staff (according to Law on Civil Service) Same as for ministry staff Less than for ministry staff Less than for ministry staff Participation in the governmental decision-making bodies - May participate in meetings of Permanent State Secretaries - May participate in Cabinet meetings* - May participate in meetings of Permanent State Secretaries - May not participate in governmental decision-making bodies Budgetary status (position in the Law on Budget) Separate section in the Law on Budget* Subsection within the Ministry’s section in the Law on Budget Not included explicitly in the Budget * This applies to a smaller set of OHSZs only. Secondly, the issue whether it is the Cabinet or the minister who is charged with the supervision of the agency has far-reaching significance. Naturally, the former has no appropriate apparatus for directing an organization effectively, thus by all means this kind of supervision is less operative and powerful than that of the ministry. From a functional perspective this is related to the fact that an OHSZ’s duties do not fall into the portfolio of any ministry (as they were established for providing functions besides ministries), therefore it is independent from the ministerial structure, as being typically subordinated to the Cabinet. This implies the following with regards to its autonomy: – There is no ministry which could absorb its duties. – The designated minister bears only minor supervisory functions over the OHSZ, but (s)he cannot issue instructions to carry out specific tasks or to rectify the ones already taken. – Its head is appointed and dismissed by the Prime Minister or the Cabinet, hence the Prime Minister (and not a minister) is responsible for the body’s activity before the Parliament. – The OHSZ can appeal directly to the Cabinet, and their leaders are invited to the Cabinet meetings, unlike other types of agencies represented only through their parent ministries. Unlike OHSZ’s, central bureaus’ duties fall into the portfolio of the parent ministry. Central bureaus have no direct relation with the Cabinet. Ministry bureaus have no separate scope of duties from that of the ministry’s. What differentiates them from the central bureaus is that they have no legally defined and addressed competence; they only perform the devolved/assigned competence and duties of the ministry. This 84 practically means that, in contrast to the case of central bureaus, the minister may freely reshuffle or even cease the ministry bureau on his/her own decision. With regards to the shielding and legal entrenchment of managerial positions, in the case of OHSZ’s, the leader is appointed for a term of four or six years, and can only be dismissed in extraordinary cases. Consequently, (s)he enjoys a significantly higher extent of employment security than the leaders of most other agencies. Some of the most autonomous OHSZ’s gained further privileges in the field of shaping government policy, for instance by signing international treaties or directly taking part in preparing draft bills. With regards to budget procedures, a set of the most autonomous OHSZ’s are entitled to take a direct role in the budgetary bargaining process (i.e., they directly negotiate with the finance ministry). In contrast, central bureaus are represented in this process only via their minister, while ministry bureaus are – as a general rule – simply not a party in the budget bargain as they simply receive what the minister decides to allocate to the given agency. The extent of autonomy differs also with regards to agencies’ ability to modify approved budgets. Many features of the aspects discussed above continue to characterize agencies’ legal-institutional setup even after the 2006 reforms marking the end of the first, long period of agency development. Despite this significant extent of continuity there are, however, a number of important differences between the pre- and post-2006 legal framework, too. A brief overview of these follows in the next subsection. 3.2. The 2006 reform to central government apparatus Given the scope and limits of this paper it is possible to enumerate only some of the most important elements of the policy related to the period 2006-2010 and the political developments characterizing this extremely hectic period. By 2006, the overly ambitious expenditure programs and welfare policies launched in 2001 and strengthened further in the subsequent years have bumped into harsh and clear fiscal limitations. Although the “statistical make-up” performed by the government for the spring elections in 2006 could postpone the open facing with these realities the new cabinet had to take immediate and severe restrictive steps. These included a series of steps downsizing public sector personnel, and ceasing or merging numerous public organizations. These government policies faced an extraordinarily effective and coordinated strategy of resistance pursued by a broad host of diverse, otherwise often strongly contradictory, societal and political actors. This resistance included, among others, elements ranging from unprecedentedly violent street protests to popular referenda, and to the opposition’s fierce and persistent parliamentary and communication campaign. The economic and fiscal crisis starting autumn 2008 and the resulting further series of harsh restrictions required by the IMF came on top of the aforementioned crises (and counter-crisis measures). The last observation included in the database referring to 1 January 2009 can be regarded as reflecting the end stage of the Gyurcsány era: after the crisis and the dissolution of the (Center-Left) Socialist-Free Democrat coalition 85 the government resigned and a new minority Socialist cabinet was inaugurated in April 2009, which already lies outside the temporal scope of the present study. As part of its attempts to (re)gain control over government apparatuses in 2006 the incoming, second Gyurcsány cabinet initiated a law regulating the basic structural features of agencies, in an attempt to create some order after more than a decade of mushrooming central government organizations and organizational units. The key element of the new law created an overarching, uniform regulation of central government organizations, including agencies3. This regulation included a clearly defined, closed agency typology, which all existing and new agencies had to comply with. The new typology in part followed the pre-existing, albeit non-compulsory, typology reflected in the above table. From our current perspective, however, two differences deserve mentioning. Firstly, the law abolished the category of ministry bureaus, all of which were (meant to be) converted into some other type (usually central bureau). However, this measure was not implemented in practice to a full extent as a small number of agencies continued to exist – de jure and/or de facto – in the form of ministry bureau. Secondly, the terminology has slightly changed as one of the remaining two categories – “OHSZ” – was renamed “Government bureau”. Besides their fundamental impact on the agency field the 2006-2007 administrative reforms deserve attention also because of their unusual thrust and doctrinal clarity. Market type mechanisms, (multinational) corporate management practices, performance orientation, monetary incentive systems, competition, and contracting out are all core elements of the NPM toolbox and rhetoric but had never appeared so strongly and consistently on the Hungarian reform agenda before then. From the point of view of agencies this implied, interestingly, not agencification in the usual sense of the word but in some respects the opposite thereof. A core element of the reforms was downsizing both in terms of organizations and staff (the original plans foresaw a staff decrease of about 10% for central administrative organs). The integration of administrative organizations affected the agency population particularly strongly, because sometimes as many as six organizations were merged into one. Interestingly however, the stated objective of amalgamations was different from the one(s) usually appearing in other countries’ post-NPM reforms taking place in the name of “joined up government”. Rather, the objective here is to achieve economies of scale, thereby realizing cost savings. However, whether or not this hypothesis holds true was very rarely argued for (or even questioned), let alone empirically tested. 3.3. The proliferation of agencies Agencies operate in quite diverse functional/policy fields. However, much of them are concentrated in a few areas. In 2006, for example, almost half of them were involved in sectoral (34%) or general (12%) economic affairs, while one fourth of them provided general public services. 3 Law on Central State Administrative Organizations and on the Legal Status of Cabinet Members and State Secretaries (Law LVII/2006) 86 As regards agencies’ proliferation quantitative data are not available for the period 1990-2001. On the basis of available, sporadic evidence it seems however that there has been an upward trend in agencies’ number and importance throughout much of this period. This is corroborated by an unpublished government report cited by Papházi and Rozgonyi (2008, p. 96) which states that between 1995 and 2004 a 23% increase occurred in the number of “central state administrative organizations” (an analytical category not clarified in the report). The figures below show that this trend continued until 2006, the first year of the second Gyurcsány cabinet. Table 2: Proliferation of agencies between 2002 and 2009 (broken down by legal status) Count Legal status 2 Total OHSZ/Gov. bureau Central bureau Ministry bureau YEAR 2002 2003 2004 2005 2006 2008 2009 Total 14 17 20 24 19 7 7 108 27 30 29 26 25 31 30 198 11 14 15 15 17 4 3 79 52 61 64 65 61 42 40 385 Between 2002 and 2006, the total number of agencies – as operationalized in this study – increased from 52 to 65. This upward trend was present in all three subsets of agencies. However, as the table shows in the course of 2006-2007 a sharp U-turn occurred in the patterns of agency proliferation as the number of agencies decreased dramatically. This spectacular trend change was composed of such elements such as hiving tasks back into the parent ministries, a very strong strife to merge existing agencies into larger units, and granting them less and less formal autonomy (Hajnal, 2010). Interestingly, these elements are just the opposite of many of the mainstream NPM agencification principles. As already noted above in relation to the changes in the legal-institutional framework, these patterns in fact have more in common with “whole of government” attempts to counterbalance the perceived limitations and dysfunctions of agencification reforms in certain other countries (Pollitt, 2003). It is plausible to presume that these changes are, at least in part, attributable to similar factors as in other countries. Especially the (perceived) weakening of the political center and the dilution of its control capabilities seems to be a major underlying factor. At the same time, some other factors seem to play an important role in the shifting patterns of Hungarian administrative policy. Firstly, the alarming fiscal pressures – threatening, at some points, with the collapse of government finances – was and continues to be the rhetorical cornerstone of most of the structural changes. According to this argument the amalgamation of different agencies is expected to create serious economies of scale by allowing for the elimination 87 of overlapping functional (IT, HR etc.) units. Secondly, tendencies to diminish the scope and the role of civil service appeared especially for the first two years of the second Gyurcsány cabinet (in office between 2006 and 2009). Condemnations of the “rigid, inefficient, lazy bureaucracy” (thereby referring to the professional civil service) and praising of corporate style governance were frequent elements of the government rhetoric of the time. The opposite trend of extending the scope of civil service regulations to new organizational areas can be assumed to be motivated by such factors as more advantageous employment conditions (salary, work intensity, employment security) and the increased organizational prestige associated with civil service. The below table illustrates the organizational weight of different administrative organizations within the broader public administration structure, using the most recent data available. Table 3: Number of executive branch organizations and civil servants falling within the scope of the Law on Civil Service4 Organization type Organizations Employees4 B1 – Ministries 14 6,675 B2 – Autonomous state admin. organizations 4 806 B3 – Government bureaus (previously: OHSZ’s) 6 2,303 B4 – Central bureaus 33 12,270 B5 – National HQ’s of armed bodies 4 304 BT – Territorial state admin. organizations 179 40,708 BH – Local state admin. organizations 196 4,862 C – Local government admin. organizations 1,922 43,366 Total 2,358 111,294 Source: Compiled from KIM (2010) * Staff fi gures include full time and part time employees as well As of January 2010 approximately 111,000 civil servants were employed (a marginal number of additional civil servants were employed by organizations outside the executive branch). Approximately 60% of civil service personnel worked in organizations hierarchically subordinated to central government while the remaining 40% was employed in local government offices. Agencies in the broader sense involve organization types B3 to B5. In terms of staff these organizations (B3 to B5) employ about 22% within the central state apparatus. 4. Restructuring as political control 4.1. Operationalizing the research question As discussed earlier at the heart of the agencification lies the phenomenon of executive politicians granting subordinate agencies some extent of autonomy. Autonomy itself, however, is a multifaceted concept involving such diverse sub- 4 Staff figures refer to staff employed in civil servant status only. Total staff may include employees employed by regular labor contract framework and, in the case of armed bodies, a large number of military/police personnel. 88 concepts as formal, legal, structural, policy and managerial autonomy, to mention but a few (Verhoest et al., 2004). The following investigation focuses on the relationship between formal autonomy and de facto autonomy. Formal autonomy is understood in the sense of what Thatcher calls formal institutional framework of controlling agencies (Thatcher, 2002, p. 958). Referring to yet another conceptual framework, it is the formal autonomy (Yesilkagit and van Thiel, 2008), which legally codifies channels and measures through which executive politicians (most of the time the minister in charge) are able to control the behavior of the agency on the one hand, and the barriers and constraints of this controlling capacity, on the other. This concept is, in the context of Hungarian agencies, well operationalized by the legal status of agencies (see above section on the legal framework of agencies). Reflecting the fundamentally formal, institutional focus of the current study (and the underlying data) the conceptualization of de facto autonomy is confined to a specific and narrow aspect sense: the extent, to which agencies are able to retain their basic legal-structural features in the face of political turbulence and pressures of the day. Political rhetoric and the broader public discourse in Hungary and abroad, and a significant proportion of academically oriented models as well as explanations based on “less noble motives” (Pollitt et al., 2004, pp. 19-20) involve the precept that politicians grant different levels of formal autonomy to agencies (thus different legal statuses) in order to ensure their different positions on the “full political control” versus “full entrenchment from political control” imaginary axis. Indeed, it is not easy to explain the existence of agencies with different degrees of formal autonomy (i.e., having different legal statuses) otherwise as on the basis of politicians’ intention to create – for one reason or another – agencies with different extent of autonomy. In practice, however, the relationship between formal and de facto autonomy is not that straightforward. For example, Yesilkagit and van Thiel (2008) found that while there are significant differences of de facto autonomy between (groups of) agencies characterized by different formal levels of formal autonomy, the relationship is oftentimes not linear: formally more autonomous agencies may have – in some respects and in certain cases – less de facto autonomy, and vice versa. Previous research on Hungarian agencies (Hajnal, 2010a) found an even more counter-intuitive – one may say perverse – relationship between agencies’ formal and de facto autonomy. Namely, instead of the rather haphazard relationship found by Yesilkagit and van Thiel there seemed to be a clear and linear – however clearly negative – relationship between the two. That is, more formal autonomy seemed to imply less de facto autonomy. This finding – reached in an earlier phase of the research underlying the present article – was based on a more limited data set of agencies ranging from 2002 to 2006 only. On the basis of this limited data set two hypothetical explanations of this strange finding were formulated (Hajnal, 2010a, pp. 60-65). The first explanation is related to the, so to say, “political value” of the organization. This value – not to be mixed up with policy saliency – is determined by such factors as the capacity they offer to their political masters (i) to place senior party cadres or 89 clientele into prestigious and well-paid positions, and (ii) to exert influence on larger and more important apparatuses, budgets and policy fields. Moreover, (iii) these capacities extend to the future possibility of a political shift in the governing coalition or party, thereby creating an organizational Hintergrund for party cadres and clientele. Agencies having higher structural status (and thus political value) are likely to be used more often as a kind of “exchange good” in political transactions – i.e., to “pay the price” of political deals between allies and adversaries alike – more frequently than structurally less autonomous ones. However, as a result of its structural entrenchment such “payments” usually require some sort of seemingly elaborate and well-founded structural re-arrangement; simply substituting the agency head with one belonging to another party or power grouping would be overly obvious and difficult to explain. The second explanation – not necessarily excluding but possibly supplementing the previous one – is based on the idea of using restructuring decisions as a means of political control and accountability. It seems that in a number of cases the extent of political control over agencies and/or charismatic agency heads tends be perceived insufficient by the agency’s political masters. Because of the limited effectiveness of institutionalized, formal control arrangements an important way of regaining control over such organizations is to “punish” and remove agency management by means of reorganization. It is this latter explanation, which is put to further scrutiny. However, in comparison to the previous analyses (e.g. those presented in Hajnal, 2010a) the analysis that follows is extended in two respects. Firstly, the empirical basis is substantially broadened as the time period covered by the data set is extended from 2006 to 2009 (the starting date continues to be 2002). This temporal extension had the consequence of increasing the number of observations by about 50%, to n = 152 organizations. Secondly, more elaborate analytical tools are utilized: survival times of agencies – that is, the length of time elapsed between two organizational changes affecting the agency’s basic structural features – are calculated and used for the analysis5. 4.2. Empirical findings The results of an extended analysis of the above problem are presented below. The research question may be re-stated as follows: to what extent does agencies’ formal autonomy go hand-in-hand with their de facto autonomy? Formal autonomy continues to be operationalized as legal status of the organizations. De facto autonomy is operationalized, in line with the principally formal-structural focus of this study, as the extent to which a given agency’s formal structural features remain unaffected by political decisions. 5 Put simply, the survival time of an agency is the length of time elapsed between the change event (i) starting, in a given form and legal status, the organization at hand, and (ii) the event ending it. Thus the survival index can be interpreted as a measure of the extent, to which an organization is/was able to retain its key structural features in a highly dynamic and hectic environment. 90 The below figure visualizes the relationship between formal autonomy and agencies’ average survival times in the three groups defined by legal status. Legal status Ministry bureauCentral bureauOHSZ S ur vi va l ( ye ar s) 3,4 3,2 3,0 2,8 2,6 2,4 2,2 2,0 Figure 1: Average survival times of agencies with different legal status (2002-2009, n = 152) As the figure shows the formally most autonomous organizations (OHSZ’s) “survive” on average 2.2 years without being restructured, whereas the formally less autonomous central bureaus’ life span is 2.5, and that of the (formally least autonomous) ministry bureaus is as much as 3.3 years. In order to test the statistical significance of the differences ANOVA test was performed. The difference between the three groups proved to be statistically significant at the p = 0.48 level. These results corroborate the conclusions based on earlier analyses performed on a significantly more limited data set (Hajnal, 2010a). That is, contrary to both: – the intuitive, normative assumption of a positive relationship between formal versus de facto autonomy, and – the empirical finding of a haphazard relationship between these two variables established by Yesilkagit and van Thiel (2008). Hungarian agencies exhibit a clear negative relationship between their formal and factual structural autonomy. This finding supports the present study’s hypothesis about the existence of – supposedly relatively new – means of agencies’ political control not identified/described by pre-existing scholarship. As a note of caution it has to be noted nevertheless that – as it is usually the case – the pattern identified in the empirical data may be consistent with a number of other explanations, most notably the “political value”-hypothesis elaborated elsewhere (Hajnal, 2010a) and briefly discussed earlier. Still it seems somewhat difficult to develop any alternative interpretation of the empirical finding presented above not involving the quest for subtle and informal means of political control and/or realizing political benefits on the part of executive politicians. 91 5. Conclusions “Agencies” and “agencification” – fervent topics in the international public admi- nistration discourse – figured only marginally, or even less, in the Hungarian government practice and scholarship throughout the past decades. Despite this conceptual – and even terminological – gap it is both theoretically relevant and empirically possible to analyze the proliferation and dynamics of agency-like organizations. One of the key conclusions reached thereby is that the importance and proliferation of agency-like organizations grew throughout much of Hungary’s post-transition administrative history. Moreover, the development pattern found in Hungary, including the reversal of the agencification trend in the second half of the 2000’s, is in several respects noticeably similar to the path followed by other post-transition countries of the region (Randma-Liiv, Nakrosis and Hajnal, this issue). Another important finding relates to the special, to some extent possibly even idiosyncratic, ways, in which agencies seem to be utilized by executive politicians. The finding that agencies’ formal and de facto autonomy (or at least some important aspects thereof) are in a clear and almost perfect linear, but nevertheless negative, relationship highlights some of the possible “misuses” (Hajnal, 2010a) of agencies. These and similar kinds of other misuses – such as those reported by Moynihan (2006) in relation to Slovakia – are likely to play quite important roles in the functioning of governments and should therefore be a key topic on the research agendas of the future. References: 1. Balázs, I., ‘A központi közigazgatás különös hatáskörű szerveinek szabályozási kon- cepciója’, 2004, Magyar Közigazgatás, vol. 54, no. 9, pp. 513-528. 2. Beblavy, M., ‘Understanding The Waves of Agencification and The Governance Pro- blems They Have Raised in Central and Eastern European Countries’, 2002, OECD Journal on Budgeting, vol. 2, no. 1, pp. 121-139. 3. Budai, G., ‘A második Gyurcsány-kormány által meghirdetett államreform a sajtó tükrében’, in Rozgonyi, T. (ed.), Államreform, közigazgatás, háttérintézmények, Bu- dapest: Gondolat, 2008a, pp. 50-81. 4. Budai, G., ‘Változás és alkalmazkodóképesség’, in Rozgonyi, T. (ed.), Államreform, közigazgatás, háttérintézmények, Budapest: Gondolat, 2008b, pp. 19-49. 5. Christensen, T. and Laegreid, P., ‘Agencification and Regulatory Reforms’, in Chris- tensen, T. and Laegreid, P. (eds.), Autonomy and Regulation: Coping with Agencies in the Modern State, Cheltenham: Edward Elgar, 2006, pp. 8-52. 6. Eichbaum, C. and Shaw., R., ‘Revisiting Politicization: Political Advisers and Public Servants in Westminster Systems’, 2008, Governance: An International Journal of Policy, Administration, and Institutions, vol. 21, no. 3, pp. 337-363. 7. Flinders, M., ‘The Politics of Patronage in the United Kingdom: Shrinking Reach and Diluted Permeation’, 2008, Governance: An International Journal of Policy, Adminis- tration, and Institutions, vol. 22, no. 4, pp. 547-570. 8. Hajnal, G. and Kádár, K., The Agency Landscape in Hungary: An Empirical Survey of Non-Departmental Public Bodies 2002-2006, ECPR Standing Group on Regulatory Governance, Annual Conference, Utrecht, The Netherlands, June 5-7, 2008. 92 9. Hajnal, G., ‘Failing Policies or Failing Politicia ns? Policy Failures in Hungary’, 2010b, World Political Science Review, vol. 6, no. 1, Article 13. 10. Hajnal, G., ‘The Agency Landscape in Hungary 2002-2006’, in Laegreid P. and Ver- hoest, K. (eds.), Governance of Public Sector Organizations. Autonomy, Control and Performance, Houndmills/New York: Palgrave Macmillan, 2010a, pp. 44-65. 11. Közigazgatási és Igazságügyi Minisztérium, Adatok a magyar közigazgatás személyi állományáról a 2010. Január 1-jei állapot szerint, Budapest: Ministry of Public Ad- ministration and Justice, Unpublished government report, 2010. 12. Moynihan, D.P., ‘Ambiguity in Policy Lessons: The Agencification Experience’, 2006, Public Administration, vol. 84, no. 4, pp. 1029-1050. 13. Nyitrai, P., ‘Adalékok az országos hatáskörű szervek rendszerében 1990-94 között bekövetkezett változások elemzéséhez’, in Fogarasi, J. (ed.), A magyar közigazgatás korszerűsítésének elvi és gyakorlati kérdései: Oktatási anyag, Budapest: Unió, 1996, pp. 146-153. 14. Papházi, T. and Rozgonyi, T., ‘Minisztériumi háttérintézmények a teljes átalakulás szélén’, in Rozgonyi, T. (ed.), Államreform, közigazgatás, háttérintézmények, Budapest: Gondolat, 2008, pp. 419-465. 15. Peters, B.G. and Pierre, J., Politicization of the Civil Service in Comparative Perspective: The Quest for Control, London: Routledge, 2004. 16. Peters, B.G., The Politics of Bureaucracy, London: Routledge, 2001. 17. Pollitt, C., Talbot, C., Caulfield, A. and Smullen, A., Agencies. How Governments Do Things through Semi-Autonomous Organizations, New York, NY: Palgrave Macmillan, 2004. 18. Pollitt, C., ‘Joined-Up Government: A Survey’, 2003, Political Studies Review, vol. 1, no. 1, pp. 34-49. 19. Randma-Liiv, T., Nakrošis, V. and Hajnal, G., ‘Public Sector Organization in Central and Eastern Europe: from Agencification to De-agencification’, 2011, Transylvanian Review of Administrative Sciences, Special Issue, pp. 160-175. 20. Sárközy, T., Államszervezetünk potenciazavarai, Budapest: HVG-Orac, 2006. 21. Talbot, C., ‘The Agency Idea: Sometimes Old, Sometimes New, Sometimes Borrowed, Sometimes Untrue’, in Pollitt, C. and Talbot, C. (eds.), Unbundled Government: A Criti- cal Analysis of the Global Trend to Agencies, Quangos and Contractualisation, London/ New York: Routledge, 2004, pp. 3-21. 22. Thatcher, M., ‘Regulation after Delegation: Independent Regulatory Agencies in Eu- rope’, 2002, Journal of European Public Policy, vol. 9, no. 6, pp. 954-972. 23. Vadál, I., A közigazgatási jog kodifikációja – Stabil kormányzás, változó közigazgatás, Budapest-Pécs: Dialóg Campus, 2006. 24. van Thiel, S. and CRIPO Team, The Rise of Executive Agencies: Comparing the Agen- cification of 25 Tasks in 21 Countries, EGPA, Annual Conference, Malta, September 2-5, 2009. 25. Verhoest, K., Peters, G.P., Bouckaert, B. and Verschuere, B., ‘The Study of Organisa- tional Autonomy: A Conceptual Review’, 2004, Public Administration and Development, vol. 24, no. 2, pp. 101-118. 26. Yesilkagit, K. and van Thiel, S., ‘Political Influence and Bureaucratic Autonomy’, 2008, Public Organization Review, vol. 8, no. 2, pp. 137-153.