Abstract
The aim of this article is to explore the opportunities for and constraints of the establishment of a central coordination instrument in a highly fragmented central government. The article is based on a case study which examines the creation of a development system for Estonian senior civil servants. It describes and explains shifts in the coordination mechanisms within this initiative over the last decade. The study concludes that networks may offer an alternative to hierarchy for the creation of new coordination practices in fragmented administrative systems. However, networks may prove to be insufficient and require further institutionalization and formalization through the inclusion of hierarchical elements once the coordination instrument matures. The study also shows that the funding arrangements may trigger a shift in basic coordination mechanisms. It is therefore suggested that future research should systematically examine funding practices as an integral part of coordination arrangements.
Points for practitioners
This article shows how civil servants can launch a major coordination initiative without the input of the legislator, and build a governmental ‘island of excellence’ with the help of the EU’s structural funds. The establishment of a sound basis for the initiative in the form of a competency model and subsequently an assessment system was crucial in the evolvement of the senior civil service development system in Estonia. Constant learning and the gradual winning of the target group’s trust and support laid the groundwork for institutionalizing a coordination instrument initially designed to operate on a voluntary basis. The article also shows how addressing one coordination problem may raise new issues of accountability, sustainability and coordination.
Introduction
As a response to the increasing complexity of the public sector and its fragmentation, the need for coordination and integration has become evident all over the Western world. The emergence of such post-New Public Management (NPM) models as ‘whole-of-government’ and ‘joined-up-government’ demonstrates the importance of tackling the ‘wicked issues’, fighting ‘departmentalism’, ‘tunnel vision’ and governmental ‘silos’ (Christensen and Lægreid, 2007). Central coordination is particularly vital in fragmented and decentralized administrative systems in order to strengthen the collective ethos and culture that have been weakened by the recent NPM reforms which stressed the performance of individuals and individual organizations (OECD, 2003: 16).
The existing research shows that fragmentation springs from, and difficulties in cooperation lie within, not only specialization and decentralization but also in the individual agendas of ministers and senior civil servants (SCSs). Working together in a coherent manner requires the development and implementation of shared plans and common programmes, which in turn assumes that SCSs act on the basis of shared values (Bourgault, 2007: 258). Consequently, systematic development of SCSs with the aim of achieving common role perceptions as well as of paving the way for cooperation across organizational borders offers a solution for overcoming fragmentation, for the integration of public services and for addressing the ‘wicked issues’ (Bouckaert et al., 2010; Verhoest et al., 2007: 344). The leadership, professionalism and common values of the senior civil service (SCS) and the ability of SCSs to work together are key preconditions for successful coordination (Hansen et al., 2012). However, although the practices of formally recognizing SCSs as a separate group have been mostly studied in Anglo-American countries (Bhatta, 2001; Bourgault, 2007; Halligan, 2012; Hansen et al., 2012; Kim, 2007; OECD, 2008), the topic has been largely neglected in Western and in Central and Eastern Europe.
This article explores the evolution of a development system for Estonia’s SCSs. The focus is on the process of institutionalizing the relevant arrangements and on the causes and actors behind the emergence of this central coordination instrument. The timespan covered is from 2004 to 2013. In addition to casting light on the evolution of a new coordination practice, the study of the Estonian SCS development system is relevant for coordination studies because of the political and administrative context in which the Estonian SCS operates. Central government in Estonia is characterized by a high level of fragmentation and segmentation compounded by poor horizontal coordination practices. The problems of such an organizational setup and the related silo effects have become particularly visible in the last decade (OECD, 2011; Sarapuu, 2011; Tõnnisson and Randma-Liiv, 2008). Moreover, an entrenched segmentation creates considerable hurdles for the introduction of new central steering procedures. The case of the evolution of the Estonian SCS development system, and, in particular, the shift in this process from coordination by an ad hoc network to a formalized system offers an intriguing example. The authors aim to contribute to the academic literature on coordination by offering a systematic analysis of the explanatory factors behind the emergence of a new central coordination instrument, and of the changes in the mix of the basic coordination mechanisms that are at the core of the instrument (see Bouckaert et al., 2010). While recent evidence cited in the coordination and policy network literature points to the tendency of networks to supplant hierarchies (e.g. Bevir and Richards, 2009; Bouckaert et al., 2010; Klijn and Koppenjan, 2000), the case analysed in this article illustrates the opposite shift – from network towards hierarchy.
The article addresses the following research questions: (1) how did a new horizontal coordination instrument emerge in an environment characterized by strong segmentation and poor central steering? (2) how did the mix of the basic coordination mechanisms change in the course of the evolution of the instrument? (3) how can a shift from network to hierarchy be explained?
The case study is based on the in-depth analysis of the Estonian civil service training system which was commissioned by the Estonian Ministry of Finance (Sarapuu et al., 2011). It makes use of a range of sources and research methods: desk research on the relevant academic literature and government documents, followed by 31 in-depth interviews with Estonian civil servants and a focus group analysis. All co-authors have been involved in participant observation: one served as Secretary-General of a ministry during the initiation of the SCS development system, and two have participated in the design and implementation of several development activities as part of this system.
Analytical framework
Coordination mechanisms
Public sector coordination has been defined as a purposeful alignment of units or actors to achieve a defined goal (Verhoest and Bouckaert, 2005: 95). Coordination as a process is brought about with the help of specific activities or structures – coordination instruments (Bouckaert et al., 2010). Three basic types of coordination mechanism have been outlined in the academic literature that underlie different instruments: hierarchy, market and networks (Bouckaert et al., 2010). These mechanisms differ in their reliance on various types of incentives provided by means of voluntary agreements, common norms and culture, formal regulations or even coercion. In hierarchy-type coordination mechanisms, interaction is based on formal authority which derives from legislation, administrative orders, common standards, the rights of inspection and intervention. The purpose of using hierarchical power for coordination is to consciously design and control the goals of different actors as well as the procedures used for achieving these goals. Such an approach is characterized by considerable advantages which, among other things, are manifested in terms of firmness and speed, the potential for equity and objectivity, and clear responsibilities (Bouckaert et al., 2010: 269). The disadvantages include a lack of legitimacy and ownership, the potential mismatch between rigid organizational structures and complex and dynamic environments, administrative overload, and other bottlenecks afflicting formal bureaucracies (Bouckaert et al., 2010: 35).
The underlying logic of market-type coordination mechanisms stems from neo-institutional economics. Market-type coordination is based on exchange and competition. Guidance and control is provided by the ‘invisible hand’ through supply and demand, price mechanisms, and the self-interest to earn a profit and avoid losses. The main resources in such mechanisms relate to information, bargaining and power (Bouckaert et al., 2010: 35). The potential advantages of markets (either natural or artificially created) are incentives to performance, contractual clarity, and the balancing of supply and demand. The downside of market-type coordination includes excessive competition between actors, conflicting interests of organizations, instability of the system, asymmetry of information, and potential loss of economies of scale (Verhoest and Bouckaert, 2005: 35).
The theoretical basis for networks as coordination mechanisms comes from policy network theory (Bouckaert et al., 2010: 34) which assumes that policies are formed and implemented in complex interaction processes between a large number of interdependent actors (Klijn and Koppenjan, 2000: 139). The key characteristics of policy networks are the actors involved and their relationships (Klijn et al., 1995: 439). Interaction derives from resource dependence and assumes that actors are willing to exchange their resources in joint processes (Rhodes, 2007). This means that the attractiveness of the interaction process is of critical importance. Consequently, network-based coordination relies on cooperation and solidarity between actors whose relations are shaped and controlled by mutual interdependencies, trust, shared values, and reciprocity (Bouckaert et al., 2010: 35–36). The main advantages of networks are legitimacy and validity in decision-making, emphasis on shared norms and a high level of trust (Verhoest and Bouckaert, 2005). Networks possess good potential for the inclusion of diverse partners and for learning (Koppenjan and Klijn, 2004: 127–128), for coping with dynamic and unstable environments (Kickert et al., 1997: 54 ), and for inducing innovation (Goes and Park, 1997). However, networks also exhibit important disadvantages – slow and difficult decision-making processes, lack of clear responsibility and enforcement capacity (Verhoest and Bouckaert, 2005), limited scope of authority and the existence of power asymmetries (McGuire and Agranoff, 2011: 280).
In practice, different combinations of the basic coordination mechanisms are used, which represent options that are complementary rather than alternative (Christensen and Lægreid, 2008: 112). Governments are constantly searching for an optimal mix of coordination mechanisms, and empirical findings show significant fluctuation in the dominance of specific coordination mechanisms over recent decades (Bouckaert et al., 2010). While in the 1980s the trend was towards the dismantling of hierarchies and the strengthening of market-type coordination mechanisms, in the 1990s the focus shifted to the use of networks. However, since the beginning of the 2000s, a number of formal elements have been added to network-based mechanisms, such that one may now talk about the partial come-back of the hierarchies (Bouckaert et al., 2010: 262–263). The increasing use of such combinations reflects the need to counterbalance the shortcomings of one mechanism with the strengths of the other. By focusing on the question to what extent networks have replaced hierarchies as the new governing mode, policy network theory has significantly contributed to understanding relationships between hierarchy, markets and networks. On the one hand, representatives of the power-dependence approach argue that networks are an ‘alternative to, not a hybrid of, markets and hierarchies’ (Bevir and Richards, 2009: 5). On the other hand, Koffijberg and his co-authors (2012: 265–273) describe the dynamic combinations of hierarchy and networks as hybrids, i.e. structures in which hierarchy can intervene in a network by introducing different types of coordination strategies.
Several authors have realized that the capacity of a network to deliver results and to solve problems depends on the implementation process, i.e. how the interaction between network participants is managed (Cole and John, 1998; Kickert et al., 1997; Klijn et al., 1995; Rhodes, 2007). Previous studies reveal that some hierarchical elements in network administration are inevitable (Keast et al., 2006; Koffijberg et al., 2012). For example, when networks are managed centrally, certain horizontal relationships are transformed into vertical ones and the shadow of hierarchy rears its head. The centre of a network is often established which is responsible for, and retains control over, most important issues such as strategic decisions, the setting of key performance indicators or allocation of resources (Diefenbach and Sillince, 2011). The (financial) resources available for the central actor may trigger the introduction of certain hierarchical elements, for instance, by making a central network actor accountable for the use of such funds. Despite the widely used ‘family’ metaphor for networks, Diefenbach and Sillince (2011: 1527) conclude that just as in a family, some roles and responsibilities within networks are likely to get formally defined and hierarchically organized. This has been elaborated by the concept of ‘network centrality’ (Cole and John, 1998) which tries to capture the network actors in terms of their links with others, and analyses the crucial role of central actors in network decision-making and information processing.The use of hierarchical elements within networks can be empirically studied by looking at the dynamics of the actual functioning of networks over time. Since there are fewer formal rules, regulations, structures, procedures and policies in place in the early phase of network development, networks are shaped even more than more formal organizational types by the actual activities of their members (Diefenbach and Sillince, 2011). Any mix of coordination mechanisms underlying a specific instrument is seldom static – constant balancing between different actors and their interests may lead to dynamics and change. As the network evolves, its level of institutionalization is likely to increase. The level of institutionalization reflects to what extent the interaction between the actors has rooted, and whether certain kinds of rules of interaction and values have emerged (Dowding, 1995). There is a shortage of empirical studies outlining the actual functioning of such hybrids of coordination mechanisms and exploring and explaining their dynamics over time.
Furthermore, so far only a few studies have focused on the feasibility of different types of coordination mechanisms in particular contexts (Bouckaert et al., 2010). In order to be operational, coordination mechanisms have to reflect the characteristics of the environment, among them, the type of administrative structure and the division of decision-making power within that structure. The increase in fragmentation and complexity of administrative systems brought about by NPM reforms has also been reflected in the coordination instruments employed by those systems, for example, in the wider use of networks, partnerships, and collaborative arrangements (Bouckaert et al., 2010; Keast et al., 2006; Rhodes, 2007). Yet, there is still considerable room for analysis on the feasibility of specific coordination mechanisms in specific contexts. Consequently, by looking at the evolution of an Estonian central government coordination instrument, the article at hand aims to contribute to the better understanding of factors which shape the dynamics of different coordination mechanisms and of the relationship between those coordination mechanisms and their surrounding context.
Senior civil service as a coordination instrument
The development of a coherent group of senior civil servants can be seen as one instrument for facilitating horizontal coordination in government (Halligan, 2012; Verhoest et al., 2007: 344). The SCSs usually include (administrative) heads of ministries, departments, bureaus and agencies within the core civil service and other senior officials as designated within the central government of each country (Halligan, 2012: 116). The establishment of systems for developing common values and role perceptions among SCSs as well as stimulating the development of their identity as a coherent group has been seen as an important instrument for overcoming fragmentation and addressing the ‘wicked issues’ (Bouckaert et al., 2010; Verhoest et al., 2007: 344). With respect to the promotion of a shared culture in the civil service, SCSs are seen as crucial actors. Their personal leadership and example represent important enablers and preconditions for encouraging cooperation and coordination across the central government (Bourgault, 2007; Kim, 2007).
However, the ‘cement’ that SCSs are expected to bring to the civil service has been diluted by the strong individualization tendencies entailed in the managerial approach to public administration (Hansen et al., 2012). The focus on the performance of separate organizations may overlook broader goals and values of government (Bouckaert et al., 2010: 20). Consequently, along with the developments stressing coordination, a broader understanding of the roles and competencies of the SCS has evolved, which perceives the role of SCSs as extending far beyond their area-specific expertise (Kim, 2007: 137). Modern public managers are seen to serve not only the particular minister in charge of the area of government concerned but also the entire government as a ‘corporation’ (Bourgault, 2007: 258).
The shift of focus to the performance of the government as a whole has also been reflected in the training and development of SCSs (Bhatta, 2001; Bourgault, 2007; Kim, 2007). In general, it has been recognized that issues related to common values, ethics, and the joint management of processes can be handled more effectively in a centralized way (Christensen and Lægreid, 2007: 162). Factors which facilitate the development of SCSs include a shared vision supported by appropriate HRM systems (such as reliance on a uniform competency framework and appraisal system), a common administrative culture which acts as the cement that creates coherence across separate organizations, and central leadership from the Cabinet Secretary or the like (Bhatta, 2001; Bourgault, 2007: 272). Scholars have also underlined trust and the feeling of ownership on the part of the target group as a precondition for success (Bhatta, 2001; Christensen and Lægreid, 2008: 114).
Although several authors recognize the network of the SCSs in facilitating horizontal coordination in government, the application of specific coordination mechanisms in the setup and functioning of the SCS is still insufficiently studied. A coherent SCS is likely to be most beneficial in fragmented and/or decentralized administrative contexts by providing the necessary ‘glue’ that would tie different parts of the system together. However, the existing literature is silent about how such a centralized SCS system can be established in a fragmented administrative context where actors with their individualized goals are likely to be resistant to centralizing initiatives. At the same time, there are limited incentives for individual organizations and senior civil servants for voluntary collaboration. The case study below looks specifically at challenges related to the creation and evolution of an SCS in such a fragmented administrative system.
Case study: the Estonian senior civil service development system
The Estonian central government
Since regaining independence in 1991, Estonia has experienced major political, economic and administrative reforms. As a result of the disappearance of the Communist Party as the central coordinating power and of the need to build up a functional political and administrative system, all governments of the past two decades have faced the challenge of reforming the administrative structure and the coordination mechanisms within it (see a thorough overview of the post-communist development of the Estonian administrative structure in Sarapuu, 2011).
The Estonian central government operates a highly segmented administrative system in which responsibility for public policies and programmes lies with individual ministries. Although the functionally autonomous ministries are relatively small, they operate as strong administrative actors who have considerable leverage in issues falling in their area of government. The highest coordinating power is vested in the political executive – the Cabinet of Ministers. The Cabinet and the Prime Minister are supported in their work by the Government Office which is presided over by the apolitical Secretary of State. The Government Office, whose role has largely been that of a technical support unit of the Cabinet, has recently strengthened its coordinating functions. It includes units which deal with the coordination of EU matters, and with strategic planning and the development of senior civil servants. In addition, the Ministry of Finance, which is responsible for the budgetary process, also has at its disposal strong coordination instruments. However, the central coordinating units in the system only enjoy limited coordinating powers and are often constrained by limited financial and human resources as shown below in the case of the civil service (see also Järvalt and Randma-Liiv, 2010; Sarapuu, 2011).
This fragmented system has been reluctant to vest coordination powers in any central unit. There have been several initiatives for strengthening central coordination of which all but a few have been characterized by painstakingly slow progress and eventually proved futile. The existing horizontal coordination instruments (e.g. procedures for consultations on draft legislation and for elaboration of government policy in EU affairs) are mostly based on network-type cooperation and in that way reinforce the central role of the ministries as decision-makers.
Problems related to the fragmentation of the administrative system have become more and more evident during recent years (Sarapuu, 2011). As pointed out by the OECD, the Estonian administrative system’s ability to work in a ‘joined-up fashion’ has manifested itself in times of crisis or when an immediate policy response has been needed, rather than in everyday work (OECD, 2011). Under pressure, coordination has relied heavily on personal contacts and informal networks. In times of plain sailing, however, there is an increasing need for better horizontal integration of policy sectors and for the whole-of-government approach.
Evolution of the senior civil service development system
The Estonian public service is composed of the civil service (central government) and local government service. This study focuses solely on the civil service. Only the core of the public administration (ministries, government organizations and county governments) is covered by the open, position-based civil service system established in 1995. The general fragmentation of the Estonian central government also applies to civil service policy which is characterized by ‘a case of no central human resource strategy’ (Järvalt and Randma-Liiv, 2010). By 2013, there was neither a comprehensive civil service strategy nor any systematic steering effort with respect to the development of the civil service. No single institution at the level of central government has the power and necessary resources to develop and implement government-wide civil service policy. Each ministry and executive agency is responsible for the recruitment, training, performance appraisal and pay of its officials. As there is no central training institution within the civil service, each organization depends substantially on the private sector training market.
This case study looks into one particular aspect of civil service functioning: the establishment and institutionalization of the SCS development system. In Estonia, there are approximately 90 positions which fall into the category of ‘senior civil servant’. These positions included in 2011 the Secretary of State (1), secretaries-general of the ministries (11), deputy secretaries-general of the ministries (50), and directors-general of the executive agencies (various boards and inspectorates exercising executive functions) (28). These 90 senior civil servants make up 0.4 percent of the total of 22,286 (as of 2011) civil servants in Estonia. This percentage falls into the internationally recognized proportion of SCSs which is expressed as the range between 0.13 to 2.1 percent of the civil service, and which in many cases is under 1 percent of the civil service (Halligan, 2012: 116).
The first Public Service Act (in force from 1996 to 2013) did not make provision for a central development system of SCSs. The only central tool at that time was the Selection and Evaluation Committee of Senior Civil Servants at the Government Office. The committee had an advisory role in screening and short-listing candidates for certain SCS positions (such as those of deputy secretaries-general of ministries and directors-general of executive agencies). However, final decision-making power over the selection of the corresponding SCSs was left to the relevant minister. Training and development of SCSs was handled in a highly decentralized way by each public sector organization until 2004. The adoption of a new Public Service Act in 2012 turned a new page in SCS-related matters by introducing novel formal elements of central administration in the development of SCSs.
The establishment of an SCS development system has been a dynamic process covering the period from 2004 to 2013. Below, the evolution of this coordination instrument is described with reference to its three principal stages. Each stage is characterized by addressing its main developments, actors, resources, and coordination mechanisms.
In May 2004, the Cabinet adopted the Public Service Development Concept as the main background document for the preparation of a new Public Service Act. Although adoption of the envisaged new Public Service Act eventually failed due to political disagreements (which did not concern the SCS), some ideas outlined in the concept were developed further. Among those, the concept emphasized the need to develop SCSs who were perceived to lack the management skills that their positions required. The concept proposed launching a programme for the development and training of SCSs on the basis of a uniform competency model and central assessment system under the responsibility of the Government Office.
The Government Office convened a special task force which prepared a competency model for the SCS and its implementation plan by 2005. The aim of the competency model was to serve as support for the self-development of individual SCSs, and for the selection of new SCSs (Järvalt and Veisson, 2005: 5). The role of the Government Office was envisaged as that of a strategic partner for SCSs, to coordinate the development system and to offer support and advice, whereas responsibility for achieving development goals remained with individual SCSs and their immediate supervisors. Based on the competency model, a variety of one-off development activities was proposed to the target group. These included specially designed training and development programmes, individual coaching and mentoring, and future leaders’ development courses. A special electronic environment which was called the e-Competence Centre and which used the 270-degree appraisal method was created a few years later to facilitate the assessment of the competencies as well as the keeping of the records of planned and completed development activities. The Government Office offered support to individual SCSs in discussing assessment results and planning development activities for the following period. In addition, the first steps were taken towards adopting the model as reference material in the selection of new SCSs.
The funds available during the formation stage were limited, although some ‘seed money’ was allocated to finance the work of the project team from the EU’s pre-accession instruments and from the budget of the Government Office. In 2005, a special budget item was dedicated to SCSs’ development under the general programme Central Training 2005–2008. It was largely funded from the EU’s structural funds (85 percent), which guaranteed the funding required for the core team at the Government Office. However, these funds were insufficient for initiating extensive development activities.
This stage focused on the individual development of SCSs and did not seek to develop SCSs collectively as a coherent group. The main actors behind this initiative were non-political officials of the Government Office, led by the Secretary of State. Initially, an important role was also played by the task force which was presided over by an external consultant and included 15 representatives of the target group and a few experts from academia. The task force relied strongly on shared values and consensus-building. The important role played by the task force was crucial to instilling a sense of ownership and acceptance in SCSs. Consequently, network-type coordination practices were predominant in the setting up of the central SCS development system. The task force together with key officials from the Government Office formed the initial network, which also included the Government Office as an entity that served as the network’s enabler without having considerable resources, specific expertise or decision-making power over the network. Since the initiative as such did not have any basis in law, inclusion of the target group depended on the personal approach and on making the development activities attractive to SCSs.
The government that took office in 2007 re-launched the preparation of civil service reform. Document analysis indicates that the new draft of the envisaged Public Service Act included more detailed rules on the SCS but, once again, the corresponding bill did not pass in the Parliament due to political disagreements. However, the need to develop SCSs was clearly recognized by that time, and the corresponding part of the new bill was not questioned in the Parliament.
In this period, the competency model was revised in order to address competency needs in a more detailed way. Even more importantly, significant changes which took place during that period started to institutionalize the SCS development system and strengthened network centrality (Sarapuu et al., 2011).
The programme document Development of the Senior Civil Service 2008–2009 was officially approved at the beginning of this period, and stated as the general goal of the exercise ‘the development of competent SCSs who advance the achievement of strategic goals of the state’. This signalled a shift in civil service development objectives from the enhancement of individual managerial skills to the achievement of the government’s strategic goals. With this programme document, the development of SCSs became formally separated from the central civil service training programme. The document also became a semi-formal basis for continuing the work on the SCS development system. Moreover, the adoption of the programme document secured generous funding for the development activities offered to SCSs with 85 percent of the money coming from the EU’s structural funds. This funding scheme has continued to date. Comparing the resources of the Government Office with the resources at the disposal of the Ministry of Finance for funding centrally planned training activities for the rest of the public service, the top executives were clearly a priority. For instance, in 2010–11, 1.07 million euros were allocated to the development of 90 senior civil servants whereas only 0.65 million was spent on organizing central development activities for the rest of the public service (around 28,000 people). Thus, the development of SCSs became a separate island of excellence within the public service, offering SCSs glamorous development activities such as, for example, joint study trips for secretaries-general to foreign countries.
The most important structural change during this stage was the establishment at the Government Office in 2010 of a special unit for the development of SCSs under the direct supervision of the Secretary of State – the Centre of Excellence for Top Civil Servants (CETCS). The CETCS became an independent unit at the centre of government, unaffiliated to any particular ministry. The CETCS was allocated the following tasks: providing support for the selection of SCSs, administering their development, and cooperating with the relevant institutions and networks locally as well as internationally. The CETCS was formed on the basis of the core group in the Government Office which had been leading the development of SCSs since 2004. At various points, the group had three to five members. The Secretary of State remained one of the key actors through his personal commitment and efforts to advance SCS development. The establishment of the CETCS was part of a larger structural change whose aim was to shift the main responsibility for shaping public service policy to the Ministry of Finance. As a result, a dual system was set up in the central administration of public service – the general steering of the entire public service is in the hands of the Ministry of Finance, while the development of SCSs is administered by the CETCS at the Government Office.
In this phase, the prevalence of network-type coordination continued, but institutionalization of the system with the establishment of a special unit and a special budget line as well as more elaborate application of the competency model in the recruitment and selection of SCSs suggested increasing network centrality and the emergence of a hierarchy. This was in sharp contrast with the loose network with very limited central steering characteristic of the previous period. The combination of the independent institutional position of the CETCS and of substantial funding from the EU’s structural funds equipped the CETCS with significant power, despite the fact that this power did not derive from formal authority. In addition, the CETCS gathered valuable information on the target group over the years and was able to accumulate specific in-house know-how which contributed to its increasing power.
The new Public Service Act was adopted in June 2012 and came into force in April 2013. The Act distinguished SCSs from the rest of the civil service as far as their recruitment, selection, assessment and development were concerned. The regulation on the recruitment, selection, assessment and development of SCSs entered into force in July 2013, making the participation in the SCS development programme compulsory and thus ending the era of voluntarism. The main mandatory elements included the obligation to submit to assessment at least once every two years, and mandatory participation in development activities.
The preparation of the regulation was led by the Secretary of State and involved consultations and negotiations between CETCS, the ministries and members of the target group. During the previous years, SCSs themselves had become more active in the design and evaluation of development activities, and were also involved in discussing the draft regulation. Voluntary cooperation between CETCS and SCSs, and among SCSs themselves, had established mutual interdependencies, collective ownership and shared vision about the future of the SCS development programme. Moreover, those SCSs who actively participated in the development activities had gradually become informal opinion leaders among SCSs and managed to consolidate support to further formalization of the SCS.
Despite such a participatory process, the preparation of the SCS regulation was perceived as a top-down initiative orchestrated by the Government Office (but still approved by the majority of SCSs). Although the political level had previously taken precious little interest in the SCS, during the official Cabinet meeting the ministers raised questions with respect to the educational requirements proposed for candidates to SCS positions, and the requirements were eventually removed from the regulation.
With the adoption of the new Public Service Act and the corresponding SCS regulation, the hitherto dominant coordination mechanism – networking – was complemented and partly replaced by a hierarchical arrangement which gave the Government Office a certain authority over the target group. The power and formal authority of the CETCS increased considerably. The newly adopted SCS regulation made participation in the development and assessment activities compulsory for all SCSs, which itself demonstrates a clear shift towards hierarchy.
Discussion
The establishment and gradual institutionalization of the SCS development system has led to the recognition of the SCS as a special group within the Estonian public service. The empirical study illustrates how a network-type coordination practice gradually changed into a coordination instrument with clear hierarchical elements. The fact that the Estonian SCSs development system was initiated and elaborated without a solid legal basis placed considerable constraints on the tools available to the Government Office. In 2004, it was impossible to develop relationships with SCSs on the basis of a formal hierarchy, considering the general reluctance towards centralization initiatives in the fragmented administrative culture (Sarapuu et al., 2011). A hierarchically structured development system of SCSs could only have been imposed in the presence of a strong political will which was not the case. Under such circumstances, a network approach was the only realistic option as it involved voluntary non-binding participation: an opportunity but not an obligation for SCSs. One can argue that the political and administrative context largely determined the use of a network as the initial coordination mechanism. Thus, the case study suggests that in fragmented administrative contexts where the use of a hierarchy-type approach may be complicated due to political and/or administrative resistance, networks may offer a more feasible alternative for the creation of a new central coordination instrument.
Voluntary participation was crucial for the emergence of a supportive atmosphere among the SCSs. It established a foundation of informal relations between the CETCS and the SCSs, and helped to build shared values among network actors. In this sense, the delay in the civil service reform was also useful for the development of mutual collaboration and trust among the SCSs, and between the Government Office and SCSs. However, voluntary participation also had its drawbacks: for instance, the rate of participation in different development activities never rose above 65 percent of the target group (Sarapuu et al., 2011: 62). Consequently, the fact that participation was voluntary hindered uniform achievement of results due to the incomplete coverage of the target group, and indicated that not all SCSs were necessarily interested in the assessment and development of their knowledge and skills.
At the same time, voluntary collaboration proved that the majority of SCSs were able to detect common interests and were willing to develop joint activities and goals despite the usual scepticism towards such joint views and action in highly fragmented systems. Due to the voluntary participation, the performance of the CETCS depended on the other actors – it was forced to treat SCSs with great respect, professionalism and delicacy in order to make development activities attractive to the target group. Nevertheless, the more ambitious the goals and activities of the SCS development programme became, the more power the central actor gained. Yet, high network centrality with its centralized communication channels and resource allocation was not perceived negatively by the SCSs. The participatory process in the design, development and evaluation of the SCS development system by the Government Office and individual SCSs built a collective ownership and trust, and ultimately led to the acceptance on the part of SCSs to give the Government Office even more authority in the future. Consequently, voluntary participation resulted in a critical mass of committed SCSs which, in turn, paved the way for a gradual institutionalization of the coordination instrument, with the majority of the SCSs supporting the increase in network centrality and the ultimate ‘hierarchization’ of the administration of SCS development. The commitment of the participating SCSs and the presence of a working institutional framework made it possible to force the remaining SCSs to join their peers through what Koffijberg et al. (2012) call ‘negotiated hierarchical intervention’.
This study shows that any combination of coordination mechanisms underlying a specific coordination instrument is hardly static. Constant balancing between different actors, their interests and interaction – especially characteristic of networks – requires a long-term approach to studying the dynamics of coordination mechanisms and their combinations. This case represents an incremental process in which the new coordination instrument did not come about as the result of a wave of the political wand, but was implemented on the basis of the initiative and continuous leadership exercised by the Government Office, and personally by the Secretary of State. Although earlier findings show that political push is often needed to produce effective coordination (Bouckaert et al., 2010), the case study described above shows that the role played by the administration can be crucial. The personal commitment of the leaders of the Government Office made it possible to construct a new coordination instrument despite the absence of a formal basis in the law. However, the Government Office had a vested interest in the further institutionalization and formalization of the SCS development system as a step that recognized and confirmed its strategic coordination role within the executive. While network-based instruments can be regarded as a temporary measure, formalization is likely to ensure the (financial) sustainability of the initiative, which was clearly in the interests of the Government Office.
The backing of the SCSs for the centralization and formalization of the instrument could be explained by their interest in the continuation of the attractive development programme. In addition, SCSs themselves may have realized that the enhancement and institutionalization of merit practices based on more objective evaluation of their competencies could protect them against potential political intervention. Consequently, the combined interests of the Government Office and SCSs strongly facilitated the shift towards hierarchy. Altogether, effective networking and largely shared views on the administrative level enhanced legitimacy of the SCS and also made it easier for politicians to pass the new legislation through which the hierarchical approach became cemented.
Interestingly enough, previous studies have not looked sufficiently at the role of financial resources in the shaping of coordination mechanisms. For example, when describing the characteristics of hierarchies, markets and networks, one of the contemporary classics on coordination (Bouckaert et al., 2010) neglects financial resources. The general feature of ‘resources’ is considered from the perspective of the actors’ inherent powers (authority, information, trust), but is not seen to include funding. However, when analysing the shift from network to hierarchy in the Estonian SCS development system, it is hard to avoid the conclusion that the corresponding initiative was largely based on the new and generous funding scheme available from the EU’s structural funds. Funding new coordination instruments through structural funds is particularly relevant for the new EU member states in which their respective allocations make up a substantial proportion of the national budget (one-fifth in the Estonian case) and where this money is often used to launch new initiatives.
Generous funding from the EU Social Fund accelerated the institutionalization of the Estonian SCS development system. On the one hand, the allocation of financial resources proved to be of key importance not only in securing funding but also in providing a semi-formal basis for the SCS development programme. One can even argue that the presence of funding (and the need to administer the funds allocated) was one of the reasons behind the establishment of the CETCS. This shows the importance of a resource-dependency in networks (Rhodes, 2007), as in the absence of formal authority the power of the CETCS resulted to a large extent from its control of a specific budget item. On the other hand, the temporary character of the funding turned the development of SCSs into a project-based undertaking which raises questions about financial sustainability. A potential conflict was created by the new Public Service Act which, by and large, ensured the permanence of the arrangements instituted for the development of SCSs. The funding scheme described suited the network approach since it allowed the new coordination instrument to be gradually built up and improved. However, the recent shift towards hierarchy calls into question the suitability of funding the instrument on a project basis. The question that must be answered now is how to secure stable funding in order to guarantee the implementation of the SCS regulation.
This case demonstrates that financial resources could be more systematically linked to different coordination mechanisms in future research. While hierarchy-based instruments by their nature seem to assume the presence of stable core funding, networks tolerate more flexible financing schemes including project-based funding. One could also argue that this link is reciprocal and that project-based funding is suitable for the establishment of networks rather than hierarchies.
All in all, the establishment and institutionalization of the Estonian SCS development system may appear to be a success story when focusing only on the target group. However, when considering the wider civil service policy, some deficiencies become evident. For one, the SCS development system is not part of general civil service policy because the roles and responsibilities of the two central coordinators (Government Office and the Ministry of Finance) are weakly linked as they only occasionally share information, plans and activities for the development of SCSs (responsibility of the Government Office) and the rest of the civil service (where the Ministry of Finance is in charge). Consequently, the SCS development is not sufficiently tied to HRM processes on the organizational level. Thus, the emergence of a new coordination instrument has created new coordination problems. By integrating SCSs among themselves, the coordination instrument has unintentionally segregated them from the rest of the civil service. The shift from network to hierarchy has further institutionalized such segregation.
Conclusion
Hitherto, the academic literature on coordination has shown that hybrids exist and that different coordination mechanisms are increasingly combined, but has not yet provided a thorough analysis of the dynamics of such mixes of coordination mechanisms. The first finding of this study is related to the creation of a new coordination instrument in a fragmented administrative context. Due to the lack of a proper legal basis, the use of hierarchical authority was constrained at the outset, and the Estonian SCS development system was established on the basis of a network. Such a network-type non-binding and voluntary approach suited the Estonian administrative system because it duly reflected the general reluctance towards centralization initiatives. Therefore, it is argued that networks may offer a realistic alternative to hierarchy in creating new coordination practices in fragmented administrative systems in which centralization attempts may prove either politically or administratively difficult. At the same time, a network may outlive its usefulness once the coordination practice matures and requires further institutionalization and formalization through the inclusion of hierarchical elements.
Second, the Estonian case study illustrates how certain characteristics of the network pave the way for the introduction of hierarchical intervention. Voluntary involvement and participatory decision-making enable the construction of collective ownership and mutual trust among network actors which are not only necessary for the effective functioning of the network itself but may also become essential in securing support for the introduction of hierarchical intervention.
Third, the current study shows that by attempting to solve one coordination problem, others may be created. By establishing an institutionally separate development unit and integrating the SCSs among themselves, the coordination instrument has unintentionally segregated them from the rest of the civil service. The emergence of a new coordination instrument has thus created new coordination problems between the two central coordinating institutions, and between SCSs and the rest of the civil service.
Fourth, the case study draws attention to the importance of the funding arrangements which underlie the development of coordination instruments. In the Estonian case, the funding scheme appears to have been an important factor which triggered a shift in the coordination mechanisms involved. It is therefore suggested that financial resources are an essential part of the characteristics of any coordination mechanism and should be systematically considered when studying such mechanisms. While hierarchy-based coordination instruments are likely to assume stable core funding, networks allow for more flexible financing schemes including project-based funding. This is one of the research areas on which future investigations should focus. A question for further research is also how to operationalize change in the basic coordination mechanisms. This is not only important for understanding the link between networks and hierarchies, but also for comprehending the essence of the basic coordination mechanisms and the relationships between them.
Footnotes
Aknowledgment
The research leading to these results has received funding from the European Union’s Seventh Framework Programme under grant agreement 266887 (Project COCOPS), the Estonian Research Council's institutional grant IUT19-13, and from the Estonian Science Foundation grant 9435.