![]()
TOM CHRISTENSEN
Introduction
The central governmental apparatus in Norway is currently going through a period of extensive reform under a CentreāConservative cabinet. This reform wave is mainly inspired by New Pubic Management (NPM) and reflects those reforms which have been sweeping through many countries during the last two decades ā in particular the Anglo-Saxon countries (Pollitt and Bouckaert 2004). Whether these reforms will transform the central political-administrative apparatus in Norway ā a latecomer ā and its culture, remains to be seen. Reforms like NPM can quite easily become symbols more than practice, something that also makes it necessary to focus on continuous changes in government that are not tightly coupled to reforms (Christensen and LƦgreid 2003a).
The ideas behind the current reforms in Norway, and the more specific reform features, are historically atypical. When the modern Norwegian state was born in 1814 it was primarily a centralised state and has mostly been elaborated as one ever since, encompassing a strong hierarchical apparatus and much central power relative to subordinate levels and institutions (Christensen 2003). It was developed after some time into an ambitious large state, encompassing many organisations, functions and tasks. Historically, it was also widely accepted as appropriate by the citizens and cultural acceptance or trust in state institutions has been relatively high. Collective goals and values were very important. And when interest groups started to make an impact on governmental decision-making processes, mainly after World War II, they were also seen as participating in a collective state-led effort: to build the Norwegian welfare state.
The aim of this article is to analyse the development and eventual transformation of the central government in Norway, or more specifically the central administrative apparatus, and its relationship to the executive political leadership. The focus will be on structural reforms in the central state, but also on more general development features and partly on other reform elements. First, the main features of development and reforms in the period 1814 ā 1980 are analysed, based on a combination of three analytical models: āthe centralised state modelā, āthe institutional state modelā, and āthe corporatist state modelā (Olsen 1988). How and to what extent is the centralised state gradually strengthened and eventually weakened partly due to internal reasons and partly as a result of institutional and corporatist features? Second, and most important, reforms arising from NPM and āthe supermarket state modelā (Olsen 1988; Self 2000) at the central level during the last two decades are discussed and analysed. The reasons for how and why Norway has traditionally been regarded as a reluctant reformer, and currently a more eager reformer, are discussed and it is asked whether this process is eventually transforming the state (Tranvik and Selle, this issue). Third, the question of convergence and divergence is raised. Does Norway traditionally belong to a family of Scandinavian countries with certain common features? Have NPM-inspired reforms in Norway, furthered by global and European trends, made Norway more similar to other countries through a process of transformation, or are some of the traditional features still evident?
Theoretical Elaborations
Four analytical models ā state models ā are used to analyse development features in Norwegian central government in general, and reform processes more specifically. All models are characterised by some general ideas about democracy and government, some features about the relationship between executive political leaders and the central administrative apparatus, and some aspects of the internal organisation of the central administrative apparatus.
The basic model used in the analysis is the centralised state model or āthe sovereign, rationality-bounded stateā as Olsen (1988) labels this model of governance, characterising a centralised, hierarchical state with a large public sector, and emphasising standardisation and equality. The point of departure in this model is āthe parliamentary chainā (Olsen 1983), meaning that the sovereign people elect party representatives to political bodies and give them a popular mandate. Executive power is based on a majority in these bodies, and the executive has at its disposal a neutral civil service with a great deal of professional expertise ā a civil service that prepares and implements public policy, including reforms. In this model the public are first and foremost citizens and voters.
In the centralised state model politicians are held accountable at the next election for what they, and those responsible to them, have done during the previous period of office. The concept of accountability is primarily related to hierarchical accountability and ministerial responsibility, underlining the principle of upward accountability to the political sovereign (Peters and Wright 1996). In principle, ministers are accountable for everything that goes on under their guidance. The typical organisational form of the centralised state model, showing the attention to structural-instrumental factors, is the ministry with subordinate agencies that are fairly tightly controlled.
The Scandinavian countries have traditionally (although presently experiencing change) had less developed control and scrutiny systems, because the executive and the civil service can be trusted to work autonomously and use their own discretion (Nordby 2000). There are more āmeta-rulesā and shared norms and values that are important for control than there are detailed, rule-based systems (Christensen and Peters 1999).
In the institutional state model, the state is seen as a āmoral communityā built on historical-cultural traditions and āpath dependencyā and supported by norms, values and institutions that encompass rules of āappropriate behaviourā (Christensen and RĆøvik 1999; March 1994; March and Olsen 1989). The role of the state is to guarantee the moral and political order, and its citizens have defined rights and obligations that protect the individual and minorities against abuse from the more powerful groups in society (Olsen 1988). Such institutional standards, like constitutional constraints, are ideally superior to the values of majority rule and cannot legitimately, at least in the short run, be changed by a majority. The role of politicians is to protect Rechtsstaat values and rules, collective standards of appropriateness, and justice. In this model the concept of collective accountability or responsibility is central, meaning that every member of complex public organisations can be held personally accountable for the conduct of the organisation. There is a moral dimension to collective responsibility (March and Olsen 1995). A main channel for influence for the public as citizens is indirect through legal rules and procedures, and the values of a rule-bound civil service (Aberbach and Christensen 2003).
Under the institutional state model the civil service attends to general norms and values, but different parts of the civil service also have their own identity and mission. They are not neutral instruments for the political leadership. The civil service will be āhistorically ineffectiveā (Brunsson and Olsen 1993), i.e. will not change automatically as a result of political efforts at reform, but will adapt through a gradual historical process. The typical organisational structure in such a state would be the quasi-judicial independent body (like independent regulatory agencies with much legal competence), encapsulated in the political and moral order and insulated from day-to-day political control. In the Scandinavian countries, for example, there is a long tradition of having different independent Ombudsman bodies.
The corporatist state model is a modification of the vision of a centralised state and reflects a pluralist society and a heterogeneous political-administrative system. The state is characterised by several competing authority and power centres that engage in negotiations based on self-interest (Kvavik 1976; Olsen 1988). The main status of citizens in such a state is as members of interest groups. The typical organisational form of the corporatist state is a collegiate corporatist structure consisting of interest representation in government where public and private actors are accorded equal status.
Under this model the authority and power of political leaders is modified. The state itself is heterogeneous and a conglomerate of different interests, and characterised by an ongoing political tug-of-war between the government and interest groups. The corporate state and integrated organisational participation in government can potentially add a valuable element to democracy, but can also be seen as undemocratic, because it may undermine the electoral channel, but also because it may create bias among interest groups with different resources (Olsen 1983, 1988).
Olsen (1988: 241 ā 42) labels one alternative model of democracy and political-administrative control as āthe supermarket stateā. This model presumes that the government and the state have a service-provision role with an emphasis on efficiency and quality, and conceives of the people as consumers, users or clients (Hood 1998: 98). The supermarket state primarily attends to economic values and norms, meaning that other traditional legitimate considerations must be downgraded, making this model one-dimensional (Self 2000). According to this model administrative reforms are apolitical or even anti-political (Frederickson 1996: 268). Thus, political bodies and politicians tend to be seen as almost illegitimate actors who obstruct efficiency (Christensen and LƦgreid 2001a). The centralised state is seen as overloaded and inefficient (Boston et al. 1996: 8, 11). A perspective of the balance between citizenship roles and consumer roles is lacking (Aberbach and Rockman 2000). The supermarket model focuses on individual and institutional managerial accountability; public administrators are supposed to be autonomous and entrepreneurial and pay attention to signals coming to them from their clients or customers.
Historical Development of the Central Norwegian Governmental Apparatus
The Norwegian Constitution of 1814, coinciding with the transition from 400 years of Danish rule to nearly 100 years of Swedish rule, was built on a Montesquieu-like concept of public governance (Arter 1999; Elder et al. 1988). The Swedish king and his advisers formed the executive. The parliament ā the Storting ā was relatively weak, being elected by the few and dominated by a class of civil servants. From 1814 to 1884 the Storting was not based on political parties; i.e. Norway was a pre-party state. From the beginning, judicial power was relatively weak and has continued to be of low political significance. In 1814, the central administrative apparatus changed from a collegial to a hierarchical structure, dominated by six or seven ministries and by judicial expertise (Christensen 2003).
Two major changes occurred in the central political-administrative system during the nineteenth century. From the 1820s professional groups demanded independent administrative bodies outside the ministries. The breakthrough for such bodies came in the 1850s when there was a rapid increase in their numbers, and a second wave came in the 1870s. These directorates or central agencies grew faster than the ministries and imitated the Swedish model with its relatively autonomous bodies (Christensen 2003). Second, there was organised opposition to the king in the Storting by a coalition of urban radicals and nationalist-minded farmers. This opposition established the parliamentary principle in 1884 and coincided with the establishment of the first two parties, the Conservatives and the Liberals (Christensen 2003).
The centralised state model is built on a structural view of governance, emphasising the central control of political and administrative leaders and the conscious design and redesign of public structures to further collective goals (Egeberg 2003; Gulick 1937; Simon 1957; March and Olsen 1983; Olsen 1992). The period 1814 ā 84 is characterised both by the establishment of main features of this model through the hierarchical central civil service controlled by the king and administrative leaders, and also by challenges to this conscious structural design and control, predominantly by professional groups demanding independent agencies and the opposition groups in the Storting, a challenge that led to a more elaborate central structure.
The year 1884 marked the breakdown of the personal power of the king, potentially creating an alliance between the Storting and the cabinet. This increasingly close connection between the executive and the legislature consolidated and increased the strength of the central state, and delivered a blow to the partly independent power base of the civil service. When the union between Norway and Sweden was dissolved in 1905, Norway had a well-established centralised state tradition. Structurally, the period between 1884 and 1940 was accordingly characterised by more reluctance to accept independent agencies, because that could undermine central control.
After World War II the Labour Party stood at the forefront of a strong build-up of the centralised state, retaining a parliamentary majority in parliament between 1946 and 1961. The party had close connections to the new group of key civil servants ā the national economists ā and shared a belief in technical rationality, strong central capacity and central planning as measures to solve societal problems (Slagstad 1998). One problem with such a development was that it overloaded the central political leadership, creating governance problems (Christensen 2003). This problem led to an elaboration and growth in the central state through such changes as more frame laws, delegation of authority from the legislature to the executive, delegation from the political leadership in the executive to the administrative leaders and to different ranks in the civil service. It also led to increased specialisation and complexity inside the central administrative apparatus (Olsen 1983). The central capacity problems, either related to problems of attention or fragmentation, resulted in a number of reorganisations of the civil service, combining structural differentiation and better coordination to cope with problems of specialisation (Christensen 2003). There was a rapid growth in central agencies from the mid-1950s until the early 1970s, stemming from the new administrative doctrine that technically-oriented tasks and institutions should be hived off, and become independent administrative agencies, to increase the capacity of the political executives.
The institutional state model tells a historical-cultural story of the state as furthering integration by emphasising collectivity, a common heritage, and a shared future (Christensen and Peters 1999; March and Olsen 1989, 1995; Peters 1999; Selznick 1957; Thelen and Steinmo 1992). Politicians, civil servants and citizens are aware of their obligations and rights and they learn what is āappropriateā behaviour (Christensen and RĆøvik 1999; March 1994). Features of the institutional state model can be seen quite early in the development of the centralised Norwegian state. In his analysis of the national strategists, Slagstad (1998) describes not only their programmes and support for the central state but also the parallel development of the institutional state. Thus, already in the 1830s and 1840s, the liberal values of the Rechtsstaat were emphasised. The Rechtsstaa...
