From Public to Private Accountability in Norwegian Local Government

This article discusses the consequences of changing views on accountability in democratic decision-making. Trends in Norway indicate that Norwegians are evaluating local democracy increasingly in terms of service performance and output, rather than in terms of political input from citizens. While traditional process evaluation is associated with governmental hierarchies and how voters can make elected representatives accountable for their policies, performance evaluation has connections with the logic of the market. It represents a shift from collective political control to individual consumer satisfaction, and consequently, from public to private accountability. Some highly prized democratic values, such as vertical political accountability, informed discussions of the totality of interests, and traditional democratic will-formation, could be lost on the way.


Introduction
Traditional public administration has the unitary state as its core concern, with integrated policymaking and policy implementation constituting a closed system within government. Policies are decided at the top, by elected politicians, and implemented by public managers. Hierarchies are the key resource-allocation mechanism, and vertical-line management ensures accountability. Accountability is usually understood as the obligation owed by agents to their principals, the ultimate powerholders, roughly on the model of shareholders in a firm who can, at least in theory, exercise direct authority over officeholders (Uhr, 1993, p. 2). It is about how voters can make elected representatives answer for their policies, how legislators can revise the actions of public servants, and how the public can seek redress from government agencies' officials (Mulgan, 2000, p. 556). The concept of public accountability sees public officials as representatives of the people and accountable to the public for executing their designated tasks (Uhr, 1993). It rests on the premise that the account giving is open or at least accessible to citizens (Bovens, 2005). The idea that governments are the supreme actors in public policy is, however, being challenged, and so is the idea that the people can control policy by holding elected politicians accountable. These challenges rise from changes in relationships within government and changes between government and the private sector.
Although this has been a gradual development spanning many decades, the changes were most recently and most explicitly conceptualized in the ideas embedded in the New Public Management (NPM) and governance schools of thought (S. Osborne, 2010;Peters & Pierre, 1998;Pierre & Peters, 2000). NPM is essentially about boosting efficiency and saving money by applying the principles of the economic market and by applying business management techniques in the public sector. Governance, as situated as a paradigm of service delivery, finds its foundation in institutional and network theory. It focuses on interorganizational relationships, where multiple independent actors contribute to the delivery of public services and where multiple processes inform the policymaking system (Osborne, 2010, p. 9). 1 However, the importance of networks and the understanding of the governmental actor as one of several actors in policymaking is however not new, as regime and urban scholars have long argued that city governing involves exchanges with private actors (Kantor et al., 1997;Stoker, 1995;Stone, 1989Stone, , 1993. Today's urban governance takes one further step when analyzing the governance process, as it considers complexity and multiple intervening actors that operate on multiple scales without a structuring authority (e.g., Kaufmann, 2008).
Although governance and NPM vary in terms of focus, they share three basic features. The first is that accountability remains a tangled and unresolved issue; the second is that what is produced (output) is the main concern; the third is their downplaying of the role of elected officials. Moreover, these trends seem to share a de-emphasis of the traditional, lay democratic input side in politics, and a tendency to legitimize political decisions increasingly by the efficiency and effectiveness of the output side, or service delivery. Hence, the quality of democracy is increasingly measured by its output rather than by the opportunities it provides for public participation (Rothstein, 2011;Scharpf, 2003). It is accompanied by what we call a transition from public to private accountability, meaning that the specific modern form of accountability is for public agencies to comply with individual or group demands rather than with instructions and sanctions conveyed down the political chain of command. Although much of this development may be inevitable in complex, modern societies, it still seems to have some highly problematic implications for important democratic values. Of interest here, therefore, is to further understand the democratic consequences of the development from public to private accountability in democracies with long traditions for lay democratic input in politics: What democratic consequences might be involved in a shift from public to private accountability?
To discuss these questions, we choose to one of the Scandinavian countries. The Scandinavian countries, Sweden, Denmark and Norway, are small and unitary, known for strong welfare state institutions and an outreaching welfare policy. They are consensus-oriented democracies, have longlasting corporatist tradition for public and private collaboration, and there is a high degree of devolution of public tasks to the local level. Within these countries, we find high levels of transparency and trust in public institutions, and a large amount of social capital among the citizens (Pedersen & Kuhnle, 2017;Sørensen & Torfing, 2019). Norway is chosen as a critical case because we here find particularly high levels of trust in politicians (Stein et al., 2019), which indicate that people are satisfied with democracy. The high degree of devolution of public tasks to the local level should spur civic engagement and a favorable condition for citizen participation in democracy. This makes it particularly interesting to set focus on accountability changes at the local level and local democracy in Norway.
In the next two sections, we discuss private accountability on the basis of two models of democratic legitimation. Here we discuss the concept of accountability in its traditional meaning and how it is interpreted today. The following section looks at the new understanding of private-public relations, as conceptualized by NPM and governance schools of thought. We then sum up our theoretical discussions regarding the concept of private accountability, before turning our attention to the local level in Norway. After a brief look at Norwegian local government, we try to identify possible links between changes taking place at this governmental level and the changing concept of accountability. Finally, we discuss the democratic implications of this transition to new forms of accountability, before a short conclusion.

Two Models of Democratic Legitimation
Our understanding of how accountability can be achieved depends to some extent on what we see as the sources of political legitimacy. In this respect, Fritz Scharpf's (1999, p. 6) distinction between what he calls two dimensions of democratic self-determination becomes relevant. The first is input-oriented legitimization, which is expressed through the idea of "government by the people." It can be associated with Rousseau's theory of political decision-making, where the emphasis is on broad public participation in a process of rational collective will-formation, culminating in a consensus on which political steps to take. Since absolute consensus, in reality, is unattainable in most political matters, the input-oriented model is usually interpreted as legitimizing decision-making through the use of majority rule. But even this principle, according to Scharpf (1999), presupposes a "thick" collective identity, because (as he puts it citing Claus Offe) "my duty to accept the sacrifices imposed in the name of the collectivity rests on my trust in the benevolence of my fellow citizens" (p. 8). Thus the government by the people model has its limitations, especially in a loosely associated polity like the European Union (EU).
The second dimension, output-oriented legitimization, is explained as "government for the people" and is associated with the Federalist Papers and the way of thinking about democratic institutional design as expressed in the U.S. Constitution. We should not assume that citizens and officials make their political choices out of benevolence or solidarity or a common identity but, rather, out of interests that might unite larger or smaller factions of society. The aim of institutional design must be to block any faction intent on replacing the public interests with their private interests (i.e., misuse of political power). The output-oriented model derives legitimacy from its capacity to solve collective problems which cannot otherwise be solved, in other words, from its efficiency.
Political democracy is mainly built around our statuses as citizens. The status of political citizenship is primarily expressed through the perspective of input-oriented democratic legitimization that Scharpf talks about. The development of new aspects of public responsibility has, however, gradually encouraged the creation of new roles in the relationship between people and authorities. We are addressed sometimes as clients of welfare state agencies, other times as users of public services, and in some cases even as customers of commercial public service delivery (Eriksen & Weigård, 2000). These roles call for different ways to legitimize political decisions and equate more with Scharpf's concept of output-oriented legitimization. The concept of accountability becomes relevant for both perspectives on democratic legitimization and for all four types of relationships between citizens and the public sector. But as we will discover, as the character of this relationship has changed over time, so has our understanding of accountability.

Traditional and New Forms of Accountability
A core sense of the concept of accountability, which most would agree is the original sense, is that associated with the process of being called "to account" to some authority for one's actions (Jones, 1992, p. 73;Mulgan, 2000, p. 555). In the context of public politics, the traditional liberal and democratic interpretation envisages a chain of command with the electorate at the top, passing through the elected political bodies and various hierarchical levels of bureaucracy, down to the frontline civil servants at the bottom. At each level, actors can and will be held to account by superior levels in the system, ensuring that citizens-in their capacity as voters-retain ultimate control over political issues (Bovens et al., 2008, pp. 230-231). Thus, accountability is an important aspect of the concept of representation (Pitkin, 1967, pp. 55-59;Weale, 2007, p. 133). The idea is that the election is at the heart of the democratic process, an arena where representatives must answer for their actions in office. Here, their fate lies in the hands of the voters.
If this is the original meaning of accountability, its interpretation quickly spawned offshoots and spread in various directions (Uhr, 1993). Mulgan (2000, p. 556) identifies four main new ways of understanding the concept: accountability as professional or personal 'responsibility'; accountability as "control" (the checks and balances of a political system); accountability as "responsiveness" to citizens' or politicians' demands, and accountability as "dialogue" between citizens on political matters.
Building on Mulgan and others, Erkkilä (2007, p. 8) creates a typology of four "traditional" types of accountability and two "new," or "alternative," types. The traditional types are political, bureaucratic, personal, and professional accountability. The new types, on the other hand, are accountability through performance or through deliberation. Erkkilä links these new forms of accountability measures to the structural changes to which public administration in European countries has been subjected since the late 1980s. These changes have been characterized by the state losing power as a single center of authority while international organizations, local government, and/or private actors and nongovernmental organizations are seeing their power increase (Erkkilä, 2007, p. 9). Part of this development is the diffusion of, on the one hand, the doctrines popularly known as NPM, and, on the other, of governance networking. Erkkilä thinks attempts to hold public officials accountable by scrutinizing their performance or outcomes must be seen as a response to the changes in the way the public sector operates in the NPM era. The emphasis on governance and deliberative accountability can be seen as a counterweight to the NPM focus (Erkkilä, 2007, pp. 18-19).
There is an obvious resemblance between Erkkilä's performance-type accountability and Mulgan's responsiveness-type. In both cases, the idea seems to be that people acting on behalf of government can and should be judged and sanctioned in relation to their contribution to satisfying actual wants and needs. The link is also clear between Erkkilä's deliberation form and Mulgan's dialogue form of accountability, as both emphasize the potential for political oversight and control embedded in citizens' communication on public matters. Of Erkkilä's four traditional types, it is likewise clear that both the political and the bureaucratic forms of accountability easily can be identified with Mulgan's description of the original meaning of the term.
That public officials should be responsive to the wishes of their political masters is nothing new, of course; it has long been at the center of the debate on how to achieve democratic accountability in modern administrative systems. But responsiveness can also mean that agencies, which deliver services to members of the public, should be responsive to the needs of their clients in a way analogous to private sector firms' sensitivity to consumer demands (Mulgan, 2000, pp. 566-567). This is a relatively new development, an offspring of the managerial reform movement-best known as New Public Management. It seeks to encourage a stronger client focus in public services: for officials to be less concerned with following fixed procedures and bureaucratic guidelines and more concerned with responding to the needs of their clients.
This has been described as a new type of accountability-an accountability "outwards" directly to the publicin contradistinction to the traditional type of bureaucratic accountability which ascended "upward," through the hierarchical chain of command. Still, it is also debatable-as illustrated by Mulgan (2000, p. 568)-how suitable it is to see it as an extension of the accountability concept. After all, it seems in part to express a general political imperative to provide better public services, but without the possibility of scrutiny and sanctions usually associated with accountability. There is a well-established (mainly American) tradition in which accountability is understood as a virtue of public agents rather than as an institutional mechanism whereby these agents can be held accountable (Bovens, 2010). But because we intend to use the concept in its latter interpretation-as a mechanism-it may be doubtful whether accountability outwards actually qualifies. On the other hand, these public-sector reforms also bring with them new institutions (or increased emphasis on older institutions) like complaint procedures, ombudsmen, and administrative tribunals, which are genuine accountability mechanisms. However, these institutions also illustrate the problematic analogy between private consumer markets and public service delivery. Private market relationships are not primarily regulated by accountability mechanisms but through (to use Hirschman's expression) exit opportunities: if customers are dissatisfied with the services of one provider, they simply move to another. Because it is difficult to graft the idea of supply diversity onto the public sector, however, compliance must be secured through accountability mechanisms, which are strategies of "voice" (Mulgan, 2000, pp. 568-569).
Just as Scharpf draws a line between input-and outputoriented ways of obtaining democratic legitimation, both Mulgan and Erkkilä distinguish between interpretations of accountability based on, respectively, input-and the outputoriented aspects of the governmental process. But although these phenomena may be related, securing legitimacy and securing accountability are still not the same thing. Political legitimacy is mainly a matter of winning support and acceptance for decisions, policies, and the government itself. In principle, at least, they can be achieved whether or not efficient channels of accountability are in place through which citizens can hold their officials to account. The purpose of reliable procedures of accountability is, no doubt, to secure legitimate government; but government can sometimes be perceived as legitimate even without those procedures in place.
What we can extrapolate from both Mulgan's and Erkkilä's discussions is that the meaning of "accountability" has changed over time and that this change is closely linked to the reforms in the way the public sector is organized and operates. If we observe this kind of empirical development, therefore, we know it will have implications for (as well as being the result of) how accountability is understood and pursued. And this, again, will influence the way democracy works.

Changes in Public-Private Relations
As we have seen, democratic legitimization and good government are increasingly associated with output results rather than with political input from citizens. Relations between government and the private sector are said to be changing, and what matters today is to get something done, either across or in defiance of blurred borderlines between hierarchies, markets, and networks. These ideas are captured in NPM and the governance schools of thought.
The governance perspective expresses a widespread belief that the state increasingly depends on other institutions to deliver its policies and secure its intentions (Bevir, 2008). It assumes that networks may be able to facilitate the capacity of markets and hierarchies with criteria that are not strictly economic and not strictly hierarchical (Pierre & Peters, 2000, p. 202). The perspective emphasizes that governments, with their single hierarchical chain of command, are not the only players in the game. Some theorists associate governance with developing social capital, strengthening of civil society, and achieving high levels of participation (Hirst, 2000;Kooiman, 1993;Sørensen & Torfing, 2008), others associate governance with empowered public entrepreneurs (D. Osborne & Gaebler, 1992) and market-based approaches to government (Donahue & Nye, 2002;Kettl, 1993). Some are concerned with multilevel governance and the negotiated, nonhierarchical exchanges between institutions at the transnational, national, regional, and local levels (Hix, 1998;Peters & Pierre, 2001;Smith, 1997). There are various theories of governance, but there is generally an understanding of the state as less capable of command and an understanding that there must be an exchange of resources if they are to achieve their goals (Bevir, 2008). Moreover, multiple actors can operate on multiple scales without a structuring authority to achieve their goals in multilevel governance (e.g., Hooghe & Marks, 2003;Kaufmann & Sidney, 2020). The multilevel governance concept is adapted to the urban perspective where two multilevel governance dimensions are proposed: A vertical dimension that analyzes the interaction between governments at multiple levels in policy making, and a local dimension that examines the involvement of nongovernmental actors in the policymaking process (Young, 2012). Consequently, multilevel governance is defined as "a mode of policy making that involves complex interactions among multiple levels of government and social forces" (Horak, 2012, p. 339).
This decision-making is based on deliberations and negotiations in more or less formal frameworks. At the local level, this can involve municipalities cooperating with each other, it can involve municipalities cooperating with actors in the civil sector, it can involve cooperation with other levels of government, and it can involve actors from both different government levels and sectors (Røiseland & Vabo, 2016). Governance theorists might argue that traditional channels of accountability can be replaced by processes like consumer choice, dialogue, and stakeholderism. Relevant stakeholders are organizations and groups directly affected by decisions made by governance networks, and these could promote interaction and exchange of knowledge in a way that enables the sanctioning of network policies and decisions (Aarsaether et al., 2009, p. 583).
NPM is a rather loose concept inspired by a combination of newer institutional economic theory and newer management theory (Boston et al., 1996). As a reform wave, it is meant to improve governments' economy and efficiency, to raise the quality of public services, and to improve the chances that policies implemented would be effective. It does so by grafting market principles and business management techniques onto the public sector. It offers to slim down municipal administrations and comes with a toolbox, the NPM "shopping basket" (Pollitt, 1995). Tools include precepts of professional management, competition, explicit standards, emphasis on output control, parsimony in resource use, and disaggregation of units in the public sector (Hood, 1991). Politicians are increasingly expected to be professional and to focus on overall principles and objectives (Pollitt & Bouckaert, 2004). The reform is also said to serve intermediate ends like strengthening political control of bureaucracy and freeing managers to manage. Its strong strategy of fragmentation is supposed to bolster innovation and efficiency and to enhance accountability, transparency, and legitimacy, while making roles less ambiguous (Christensen & Laegreid, 2001;Pollitt & Bouckaert, 2004, p. 6). Taking an agency out of a multipurpose body will allow a more professional management and higher-quality services (D. Osborne & Gaebler, 1992). The NPM school of thought finds that accountability is taken care of in their model as service delivery is directly linked to market demand instead of to political decisions, service providers will be immediately evaluated (Peters & Pierre, 1998, p. 67;Pierre & Peters, 2000, p. 228). They would also argue that the targets, benchmarks, and other indicators serve as a basis for evaluating performance.

The Transition to Private Accountability
We can now sum up what we mean by an alleged movement from public to private accountability in the municipal sector. Classical democracy builds on input-oriented democracy, as discussed in the second section, i.e., a strong feeling of membership, solidarity and duties to a group (thick collective identity) and has emphasize on democratic participation and public accountability. The neoliberal democracy on the other hand, builds on outputoriented accountability and private accountability, where common interests unite citizens and where efficiency matters. Citizens are often addressed to as clients, users, and customers of public services; they can cooperate but their primary driver for doing so is not based in solidarity and duties to a group. There is an alleged movement from classical democracy to neoliberal democracy, a movement from public to private accountability whereby the traditional role of the politically active citizen and the inputoriented form of legitimation are gradually being replaced by a situation wherein people are perceived and dealt with by their authorities as clients, users, or customers (i.e., stakeholders and a thin collective identity). The authorities want a more direct, output-oriented form of legitimation of their work. This desire has much to do with attempts to give the public sector a private-sector makeover with a view to improving the overall efficiency and quality of public services. Because traditional forms of public accountability are less and less relevant under these changed circumstances, the conception of accountability has had to be reinterpreted as well. These new types of accountability are private rather than public inasmuch as people interact with authorities as private individuals, taking care of their own particular interests rather than taking part in a collective, public process, forming society's political will. Table 1 summarizes the basic theoretical ideas and indicators linked to the concepts of public and private accountability.
Following our reasoning, we expect to find evidence of the classical form of democracy developing into a more neoliberal type in Norwegian local government. Simultaneously, this evolving landscape has enabled, and to some extent necessitated, a shift from a public to a private interpretation of accountability. More specifically, we will focus on • • Signs of weakening input-oriented channels of democratic legitimation, such as falling voter turnout and lower numbers of local council members and • • Signs of changes in forms of municipal organization consistent with a strengthening of output-oriented democratic legitimation, such as measurement of outputs, changes in organizational structures to improve efficiency, focus on consumer satisfaction rather than clear-cut arrangements that make elected politicians answerable for their actions and decisions.

Norwegian Local Democracy: Ideals and Realities
Local governments are meant to create the variety of services that reflect existing preferences within a territory. It thickens democracy insofar as citizens themselves are increasingly empowered to make their views known on public affairs, and it enhances efficiency because local knowledge is used in problem-solving and in responding rapidly to new circumstances (see, for instance, Clark, 1984;D. Osborne, 1988;Page & Goldsmith, 1987). Norwegian local governments have grown in complexity and the services they provide have expanded tremendously. All the over 400 municipalities in Norway (in 2019) are responsible for running primary and lower secondary schools, nurseries/kindergartens, care for the elderly and disabled, and social services. They are also in charge of local planning, environmental affairs, local roads, waste disposal, water supplies, and sewers.
While the local government is supposed to promote efficiency by offering varied responses and solutions to challenges, the state is pursuing equality and uniformityall grounded in legal rights and the rule of law, the need to even out differences, the regulation of public spending, and macroeconomic stability (Hansen, 1985;Musgrave, 1959). Not only the local governments are responsible for ensuring the welfare of citizens wherever they live in Norway, and for protecting the environment and future generations. The state is too. In other words, municipal tasks and responsibilities are intricately entangled with the tasks and responsibilities of the state (Kjellberg, 1991). The lines of responsibility, and thereby accountability, are blurred. Control and finger pointing at people to blame for poor schools and substandard elderly care become difficult given the entangled web of central and local government hierarchies. Local governments' accomplishments in this complex web can to, a certain degree, be measured by importing neoliberal and output accountability mechanisms.

Some Developmental Trends in Norwegian Local Government
In what follows, we briefly comment on indications of a decline in input-oriented legitimization before discussing some of the important manifestations of neoliberal ideas and governance structures and their conceptualization of accountability in local government.

Reduced Input-Oriented Legitimization
As discussed earlier, Scharpf (1999) talks about input-oriented and output-oriented democratic legitimization, while Mulgan (2000) and Erkkilä (2007) point to changes in the conceptualization of accountability that may very well be linked to a movement away from the more traditional, input-oriented form of accountability, and toward the more modern, output-oriented form of holding public agencies accountable. This evolution is reflected in the organization and workings of the public sector. There is evidence of this process at the local level too, at least in Norway. The issues discussed below, however, do mainly focus on formal politics. In urban/local politics, we acknowledge, participation often goes beyond formal politics and are present in such as infrastructure development in urban/local development and within the civil sector.
There is significant evidence of a rather depressing lack of interest among the public in participating in democratic processes; citizens do not seem to care much about communicating their ideas to decision-makers or joining in to set the agenda and to articulate values. They participate less in local elections. Voter turnout in municipal council elections was 72% in the early 1980s and was 65% in the latest election (2019). The abstainers are the largest group. Around the turn of the century, less than six in 10 cast their vote in local elections and people under 30 were most likely to show indifference to elections.
A second sign of indifference to local politics is the decline in political party membership. Willy Martinussen (2003) has been looking at different forms of political activity for a long time . He has studied party membership rates and participation at nomination meetings. The closer we come to the present, the sharper the fall in the membership, especially around the turn of the century. Membership numbers of and activity in political parties are waning, and as a percentage of the population, membership is down by about half since the beginning of the 1990s. It is currently around 8% (Barstad & Hellevik, 2004;SSB, 2017).
A third sign of the declining importance of the input side is the falling number of councilors or elected representatives on local councils over the past few decades. The total number of local council representatives at the end of the 1980s was more than 13,500; at the 2011 election it was 10,800; the number has decreased by more than 10%. Since then, Norway has had the highly emphasized Municipal Reform 2015-2016 that targeted Local government amalgamation. The reform was nationwide (it is still ongoing, but is less emphasized) and primarily relied on voluntary decisions in each local government. It was however strongly promoted by the central government and involved financial incentives. The number of municipalities are, by the already confirmed decisions, to be reduced from 428 to 356 by 2020 (i.e., by 17%) (Norwegian-Government, 2019). This reduction is clearly going to affect the total number of local councilors in Norway and will reduce input-oriented legitimization.

Increased Output-Oriented Legitimization
Since the 1980s, many Organisation for Economic Co-Operation and Development (OECD) countries have reformed their public services in line with ideas from the private sector and neoliberal economic theory. The ideas of NPM fell on especially fertile ground in the Anglo-Saxon countries, where they gained a rapid foothold. There were strong economic and institutional pressures, few constitutional obstacles, and many parliamentary conditions that favored such strategies (Pollitt & Bouckaert, 2004). The Scandinavian countries, however, were reluctant implementers of these reform strategies, given their stronger egalitarian values, greater constitutional obstacles, and often minority coalition governments. These obstacles made it difficult to roll out a radical reform strategy (Christensen, 2003). Still, some of these neoliberal ideas seem to have found sponsors in local governments in Norway, and the purpose has always been to improve efficiency in municipal service delivery and in responding to citizen demands.
New ideas in municipal management. First of all, we find changes in the organizational and management sides in the municipalities. The Municipal Act, adopted in 1993, allowed local authorities to organize themselves much as they wished, producing in its wake a wide variety of organizational forms (Bukve & Offerdal, 2002). Organizational freedom also spurred organizational disaggregation. Local government is increasingly divided into more or less independent units; municipal leaders purchase services from these units and bring in more or less private firms to manage their affairs. There are a number of areas of decisionmaking where democracies have depoliticized decisions and handed them over to autonomous bodies. In 2008, for instance, about 43% of the municipalities had two managerial levels, with the lower level acting as a relatively autonomous municipal "single purpose" service unit (Hovik & Stigen, 2008). Municipal activities used to be divided among departments and agencies (etater) under the jurisdiction of the chief executive officer. Nowadays, municipal companies run by a manager and governed by a board answering to the municipal council are increasingly the preferred way of doing council business (Torsteinsen & Bjørnå, 2012). In just over 10 years, between 1997 and 2008, the number of registered municipal companies in Norway tripled, from 850 to 2,552. In 2015, we find 2,418 municipal companies registered owned by local and regional governments in 2015 (Bjørnsen et al., 2015). They are vast in number compared with municipal companies in Sweden and Denmark (Aars & Ringkjøb, 2011). Performance in these units is evaluated by managers, owners, and users and through market competition; the general public and the collective body of elected representatives in the individual local government have a lesser say. Such units focus on aims, tasks, and budgets. They do not have to pay attention to collective responsibilities, for instance in urban planning or welfare development, and are not transparent and accountable to the public. Output is what matters. Publicly owned agencies that work fine and plow money back into the community make people pleased. The agency model is based on minimal contact between leaders at different levels. Nor are senior officials or senior politicians supposed to intervene in the everyday business of these units. The most vital elements of traditional democracy-participation in decision-making by involved parties and the search for common ground-have a different form in these autonomous structures where user evaluations and user boards are the predominant evaluation standards (Bogason, 1996). User evaluations, not participation in decision-making, provide the most significant input to these self-contained units.
Competition and performance measurements. Second, "competition" is a keyword in the neoliberal trend; it is supposed to boost efficiency. For Norwegian municipalities, competitive procurement can take many forms, including some kind of quasi-competition between public companies, benchmarking, competitive tendering between public and private companies, and so on (Johnsen et al., 2004). For the municipalities, the performance measurement database on local and central government activity is of the utmost importance. 2 Local authorities have had to report key indicators on the economy, education, health, culture, environment, social and technical services, transport, and so on to a national electronic register. These indicators are available to the public and used by the media. This reporting requirement in itself promotes efficiency because the municipalities easily can compare each other's performance (Johnsen & Larsen, 2015;Johnsen, 2007). Indeed, the register implies a sort of quasicompetition and is one of the most important measures in the central government's oversight and control of municipalities. This is however a system that only reports and compares output. It says little about the quality of what is produced and whether this quality is in accordance with what local people want. Standards and indicators serve as a basis for monitoring and auditing municipalities' performances.
Governance networks blurring political responsibilities. Third, there are increasing indications of governance processes in public policymaking that is, overlapping relationships and networks involving new actors, blurring the accountability line. Governance focuses on the interorganizational relationships, where multiple independent actors contribute to the delivery of public services and where multiple processes inform the policymaking system. Governance may on one hand be achieved by building strategic networks with actors external to local government. The prevalence of so-called development networks in the various municipalities has been the subject of many recent case studies (see, for instance, Bjørnå & Aarsaether, 2010;Fimreite & Aars, 2005;Vabo et al., 2004;Vabo & Røiseland, 2012). Such networks can be very efficient because they can avoid bureaucratic red tape and can set an objective that is much in line with what the population wants, but councilors are often sidelined in matters where they used to have a say. Fourth, a governance reform trend in the Norwegian social and health sector relies on cooperation between the municipalities and the national government (i.e., governance as involvement of different actors in the public sector). The two major health reforms in the new millennium have overarching aims that are to be achieved by increasing administrative capacity by cutting across administrative levels and policy fields. The welfare administrative reform (NAV reform) of 2005 aimed to get unemployed people back to work and to make the administration more user friendly, effective, and holistic (NOU-2004:13). The tools were a merged government agency, cooperation between central and local authorities, and a frontline office that served like a user-friendly one-stop-shop system for clients. The Coordination Reform of 2012 (Samhandlingsreformen) aimed to get more collaboration between first-tier health care in the municipalities and second-tier services, the specialist hospitals owned by the government. It is a reform prompted by what was considered to be too little emphasis on preventing diseases, a fragmented health service, an aging population with a growing need for health services, and a need to stimulate economic efficiency. The local sector-the municipalities-took over more of the responsibility, but not all of it. Economic incentives were provided and contracts were made between the municipalities and hospitals/secondtier services (St.meld.47, 2008(St.meld.47, -2009. Both these reforms were attuned to making the administration more user friendly, holistic, and effective and relied on partnership arrangements and complex interactions among multiple levels of government (multilevel governance). Services are to be better adapted to the user's needs. These reforms are illustrative of a view that collaboration and bridging are better than silo mentalities and are less focused on expertise and bureaucratic hierarchies that aim to clearly define the individual roles within the decision-making process. This interlevel organizational cooperation does not provide clear-cut arrangements such that particular officials and politicians are answerable to citizens for decisions and actions; the lines of accountability are not the traditional vertical lines to elected representatives of local or central government. Accountability lines are blurred.

Is Private Accountability a Welcome Substitute?
What all these empirical trends indicate is a decreasing focus on the local government as an arena of traditional democratic will-formation and consensus-building through the ordinary political channels.
In the reforms of the past few decades, we have seen the emergence of a new role for elected representatives. Councils have shrunk, and the current ideal is clearly for this body to decide the general direction and leave the details to the administrative levels. We have seen that municipal company organization, professional management of municipal core tasks, and intricate performance measurement systems are in vogue. We have indicated that public-private network strategies are broadly used by local governments and have argued that the main health and social reforms of today are based on collaborations that challenge vertical political accountability; user service satisfaction is what counts. We are witnessing, in other words, a change from input-oriented to output-oriented legitimization, as Scharpf has described. What we are getting is more government for the people than government by the people.
What are the consequences for real democratic accountability of these changed circumstances? As Mulgan (2000) and Erkkilä (2007) suggest, given the altered focus of the public sector, we will probably have to alter the way we understand accountability if it is to continue as a relevant concept. The traditional holding to account through the hierarchical chain of command, at the top of which the collective citizenry sits, seems less and less important. Instead, the crucial relationship today is that between the service-producing units at the lower levels of this chain of command and the groups of service consumers which they serve. These public agencies-according to this new, neoliberal interpretationare accountable insofar as they deliver the services people want and demand. In order not portray the situation it too stark terms, however, we should remember that democratic input and accountability can come in several shapes and forms. In the context of Norwegian local government, elected councilors typically have two roles. According to the traditional model, the municipal council has many members, each fulfilling a type of ombudsman role, with hands-on responsibilities for many detailed issues on behalf of their constituencies. The executive board (formannskapet) is proportionally elected from the council. Ideally, these executive boards are supposed to work out consensual solutions to the various issues. In this sense, the model seems to be rooted in a preparty political era, when cross-sectional agreements were more commonly obtained than today. When it comes to accountability, the Norwegian model is not easy on voters because majorities often shift from issue to issue and knowing who to hold to account on general performance and the overall policy can be very difficult.
However, the overall picture is that the old input-oriented form of accountability seems to be weakening in the face of the new output-oriented form. And there is a lot to criticize about the traditional way of securing accountability, as well as a lot to praise the reinterpreted concept for. Most importantly, the old chain of command is often very cumbersome in practice and many would probably say that it is an illusion to believe that it gives voters any significant degree of control over policy outcomes. Having to answer to senior levels in a hierarchical chain of command is normally associated with a system of process evaluation, but what really matters is the outcome of the process-whether people get what they want from their local government-which can be controlled by a system of performance evaluation. And performance evaluation is precisely what is offered by the new forms of accountability. Besides, the idea of "thick democracy," able to define a specific will of the people at elections, may seem more obsolete today than ever before, given the plurality of both interests and values even in most local communities.
On the other hand, there are also many problematic aspects to this new way of understanding and practicing accountability. First of all, it is questionable whether public officials' responsiveness to demands rightfully can be labeled "accountability." As pointed out by Mulgan (2000, p. 570), there is neither obvious superiority and nor obvious subordination in the relationship between officials and the public, and therefore no explicit possibility for sanctions. Dialogue on equal terms is fine, but what happens to personnel or units that still fail to fulfill consumer demands? What chances do citizens have to hold them to account if not through the democratic chain of command? Unlike options in a market economy, "exit," or transferring one's custom to a different provider, is rarely feasible when it comes to public services, unless one is prepared to vote with one's feet and move to a new local government territory. A similar criticism can be raised against governance networking. It is all very well that affected parties talk together to find smart solutions and get everyone on board, but who is ultimately responsible for what is done or not done? The answer to this basic accountability question is that responsibility is often fragmented among the networking partners.
Second, dialogue (Mulgan)-or deliberation (Erkkilä)-is suggested as one of the new accountability mechanisms in its own right. But as we just indicated, it is doubtful whether open communication between affected parties can qualify as a form of accountability as long as there exists no way of imposing sanctions. While dialogue in the old model takes place among political representatives or between representatives and their constituents, the new model prefer direct communication between service-providing officials and interested members of the public. In any case, deliberation will always be an essential part of a truly democratic form of political will-formation, especially one that can be associated with a thick democracy (Habermas, 1996).
Third, even if the accountability mechanisms were somehow to work well in this provider-consumer relationship, there is still a democratic problem in conceptualizing public policy formulation as a matter of supply and demand. If one thinks exclusively in utilitarian terms and see it as the goal of public policy to provide the greatest happiness to the greatest number, then a perfectly functioning responsiveness principle would perhaps be the adequate means to realize that goal. But with limited resources, it will never be possible to meet every demand and need; prioritizing will always be an essential part of politics. Utilitarianism tells us to prioritize policies that will benefit the greatest number, and a responsiveness principle might do just that. But the largest groups, with the power to make the strongest demands, do not necessarily have the most legitimate demands. There are rightful claims to be made on behalf of powerless groups, and it will always be the job of a truly democratic political system to give these claims the priority they deserve.
Undoubtedly, the output of public agencies which people find satisfying will produce its own legitimacy, and people are, as Scharpf (1999, pp. 26-27) says, often satisfied with policy outcomes. This is probably an even more important question at the municipal level, because many perceive municipalities as service-providing institutions. A local authority intent on good performance and responsive to citizens' demands, with affected parties consulted directly, will be seen by most as an accountable form of government. But in the long run, accountable government also depends on someone seeing the totality of interests and opinions and giving priority to some of them. In many instances, seeing the totality of interests depend on a strong input-oriented democracy, built on a thick collective identity. The fundamental aspect of democratic legitimization, therefore, is based on a well-functioning, input-oriented democracy. According to Adrienne Heretier (1999), outputoriented legitimacy in different specialized agencies is likely to be what she calls a welcome "substitute legitimation" (p. 271). For these reasons, it is unlikely that the modern performance type of accountability can ever fully replace the traditional processual type of accountability in systems of local democratic government. It will never be more than a limited supplement.

Conclusion
This article has discussed democratic consequences of a shift from public to private accountability, with trends in Norway as a case. Trends in Norway indicate that citizens are evaluating local democracy increasingly in terms of service performance and output, rather than in terms of political input from citizens. The traditional process of evaluation, the one where voters can make elected representatives accountable for their politics, has lost grounds to performance evaluations and fragmented responsibilities among network partners. The trends described from Norway represent a shift from collective political control to individual consumer satisfaction, and consequently, a shift from public to private accountability. We have argued that these trends affect some highly prized democratic values negatively. The ability citizens actually have to hold someone to account for what is done and not done is weakened, so is the conditions for traditional democratic will-formation. Moreover, the conditions for considerations of the totality of citizen interests are weakened and so are conditions for considerations of the legitimacy of citizen demands. Democracy has often been described-in Lincoln's famous words-as "government of the people, by the people, for the people." But as Scharpf highlighted, there might be a big difference between a type of government that is mainly by the people and one that is mainly for the people. And while we can imagine nondemocratic governments that are for the people, democratic governments alone are-in some way or another-by the people.
And if we look at accountability, the concept essentially lies in the (hypothetical or real) contractual relationship that exists between principal and agent, with the principal wielding ultimate power. Democratic accountability, therefore, must always involve the dimension of citizens being able to exercise political control over public services through their elected representatives. No privatized form of accountability can fully replace this understanding of the concept.