Show Summary Details

Page of

date: 11 April 2021

The Extended Scope of Accountability in Public Administrationfree

  • Richard MulganRichard MulganDepartment of Public Policy, Australian National University


The accountability of governments to their citizens is usually framed within a relationship of principal and agent in which the government, as agent, is obliged to answer to the citizens as agents. It is also commonly located within a structure of representative democracy where political leaders are elected by, and answerable to, the voters. However, these two theoretical frames do not adequately capture the relations of government to their citizens or the parameters of government accountability.

Governments increasingly operate through non-hierarchical networks that are not subject to the vertical accountability assumed in principal-agent theory. Instead, networks offer alternative, informal accountability mechanisms based on horizontal relationships. These are evident, for example, in the responsiveness of professionals to their clients and the mutual accountability of network members to one another. These mechanisms have a sufficient share in the characteristics normally associated with accountability, including the obligations to inform, discuss, and accept consequences, for them to count as mechanisms of accountability in the usual sense. Redefinition of accountability, for instance to exclude the requirement of answering to another person or body, while understandable, is not essential.

Accountability mechanisms also function without the support of effective democratic elections. For instance, formal institutions of horizontal accountability, such as courts and anti-corruption agencies, can operate in non-democratic regimes and are better seen as conditions of representative democracy rather than as consequences. Partially democratic or authoritarian regimes also exhibit various forms of social accountability in which civil society organizations call governments directly to account without recourse to state-based agencies of accountability. Large authoritarian regimes can encourage limited accountability processes as a means of bringing public pressure to bear on recalcitrant cadres. To be effective, however, all such measures require at least some legally robust support from government institutions.

The Core Assumptions of Accountability

In its core meaning, accountability in public administration is usually understood as the obligation of government officials to answer for performance to some legitimate authority (e.g., Bovens, 2007; Dubnick, 1998; Gregory (2017); Mulgan, 2003). It thus involves a relationship between two parties, an accounter (the answering official) and an account-holder or forum (the authority), and typically involves three elements or stages: the giving of information (transparency); discussion or debate; and consequences, including rectification. Standard institutions of accountability include democratic elections, legislative scrutiny, judicial review, audit, ombudsmen, regulation, media inquiry, and so on, in which governments are held to account by, or on behalf, of the people.

Academically, accountability has been closely linked with the relationship of principal and agent, in which a principal sets tasks for an agent to perform on the principal’s behalf. The expectation that an agent will answer to the agent for performance accords closely with the core relationship of accountability—that of an accounter obliged to answer to an account holder. Principals and agents also feature in agency theory, a branch of rational actor theory, based on self-interested principals seeking to control equally self-interested agents. Not surprisingly, therefore, agency theory has yielded important academic analyses of public accountability (Schillemans & Busuioc, 2015).

For example, when agency theory is applied to representative democracy, voters and governments take on the roles of agent and principal, respectively, and elections become mechanisms for holding governments accountable (e.g., Przeworski, Stokes, & Manin, 1999). Similarly, within government, chains of delegation such as elected leaders to government agencies, and within agencies, bureaucratic heads to subordinates, can be analyzed as a cascading series of principal–agent relationships, with officials at each level being accountable to their immediate superiors (see, e.g., Bovens, Schillemans, & ’t Hart, 2008; Strøm, 2000). Although the systematic application of agency theory to public accountability remains a minority interest among public administration scholars, the basic concepts of principal and agent frequently recur in academic discussion of accountability because they neatly capture what has often been seen as the central features of the accountability relationship.

Within liberal democracies, public accountability relationships and institutions have also been framed within the overarching structure of the government as accounter/agent answerable to the people or citizen body as joint account holders/principals. Particularly in political science, which pays great attention to elections and the connection between voters and elected politicians, accountability is closely linked with the process of democratic elections. For example, the “democratic deficit” in international organizations, due to the lack of supra-national elections (e.g., Dahl, 1999), has also been characterized as an “accountability deficit” (e.g., Held & Koenig-Archibugi, 2004).

Public administration, with its focus on non-elected bureaucracies, has always looked beyond elections to other accountability mechanisms, including not only “vertical” accountability to bureaucratic superiors and elected leaders but also “horizontal” accountability to specialized accountability agencies such as legislative committees and auditors, and judicial accountability to constitutional courts court. Nonetheless, these other avenues of accountability have usually been understood, explicitly or implicitly, as nested within the general structure of the accountability of democratic government to the people who elected it. Vertical bureaucratic accountability is clearly anchored to the democratic authority of the elected leadership. Agencies of horizontal accountability, though lacking the power of enforcement themselves, often provoke the elected executive to take remedial action through the power of adverse publicity and potential reputational damage. In this sense, they operate “in the shadow of hierarchy” (Schillemans, 2008), relying on the background potential of democratically legitimated hierarchies to enforce their assessments.

In turn, judicial accountability, a form of accountability with both vertical and horizontal elements in which courts exercise the power of coercion but outside a structure of overall hierarchical control (dubbed “diagonal” accountability; Bovens, 2007), is also underpinned by principles of constitutional law designed to protect the democratic rights of citizens. In this way, all the different mechanisms of public accountability have been seen as subsidiary parts of a complex system of accountability, each contributing to the overall democratic purpose of keeping the people’s government accountable to its citizens.

It has become increasingly evident, however, that these two theoretical frames, the principal–agent relationship and representative democracy, do not adequately capture the relations of government to their citizens or the parameters of government accountability (Olsen, 2015). Decades of public sector reform have made the institutions serving public functions more fragmented and more diversified as well as less amenable to former mechanisms of accountability. Within the European Union, the emergence of powerful bureaucracies beyond the direct control of elected governments has directed attention to non-hierarchical processes of accountability. At the same time, international disillusionment with elections as a recipe for fair and effective administration is forcing a reconsideration of alternative mechanics of accountability and of the possibilities for accountability within authoritarian modes of government. This article explores how accountability studies have extended the focus of the subject to meet these challenges.


One important challenge to traditional understandings of accountability has come from the growing emphasis on non-hierarchical groupings responsible for performing public duties and providing government services. These groupings, referred to generally as “networks” and also including “partnerships” or “collaborations,” are found in a number of settings, most notably intergovernmental relations, where different government institutions with independent legal status voluntarily cooperate for joint purposes, and across the public–private divide, where public and private organizations and individuals share in producing public benefits.

Cooperative relationships between independent bodies have been a long-standing feature of all governments, particularly in federal systems that rely on joint action between legally distinct levels of government. Similarly, shared public–private activities are nothing new. However, since the late 20th century, a number of developments have brought these relationships into greater prominence.

Within federalism, the growing institutional role of the European Union as a form of multi-tiered government, though locally confined to Europe, has given renewed salience to issues of multilevel government (e.g., Bache & Flinders, 2004). More generally, the worldwide upsurge in government reliance on private-sector organizations to provide public services, through various forms of outsourcing, public–private partnerships, and community cooperation, has transformed and complicated the way in which governments interact with non-government sectors. At an academic level, these trends have prompted speculation about new models of government to replace apparently outdated models based on institutional hierarchy. In particular, new theories of governance emphasize networks and informal relationships across institutional boundaries as the drivers of government action (Pierre & Peters, 2000; Rhodes, 1997).

The Accountability of Networks

Non-hierarchical networks collectively providing public services pose well-known problems for accountability (Agranoff & McGuire, 2001; Mulgan, 2003; Popp, Milward, MacKean, Casebeer, & Lindstrom, 2014; Papadopoulos, 2007; Rhodes, 1997). The sharing of responsibility between members of a network implies that no one person or group takes overall responsibility for the collective action. In other words, there is no obvious government agent acting on behalf of the public as principals. This absence of a single point of authority in turn means that there is nobody to fulfill the various accountability functions for the network as a whole by, for instance, by informing the public, offering explanations and, where necessary, implementing remedies. The problem of “many hands” (Bovens, 1998; Thompson, 1980), which is endemic in attributing responsibility for any large-scale collective action, is particularly acute in networks because there is no official mechanism for seeking collective answers, and concerned members of the public have no obvious point of contact or redress. When the network fails to deliver, its individual members have an incentive to pass the buck to one another, frustrating public expectations that someone responsible will be held to account.

These problems of diffused responsibility occur classically within federal systems of government where constitutionally independent levels of government co-operate in the provision of certain public functions such as transport, health, and education, but neither level has final authority over the other. They have also appeared in the European Union, in which powerful intergovernmental bureaucratic institutions are not subject to the levels of scrutiny that normally occur in unitary constitutional systems (Harlow, 2002). When compared with unitary governments, multilevel governments such as federations are not exposed to the same range of horizontal accountability agencies to cover their joint activities, because such agencies, for instance, legislatures, auditors, and ombudsmen, are typically restricted to scrutinizing the actions of only one level of government.

Across the public–private divide, the problems are more varied and complex because of the range of possible relationships between public and private organizations (Alford & O’Flynn, 2012). The clearest case of a network is when responsibility is genuinely shared, for example, if governments enter into joint arrangements with charitable organizations to cooperate in meeting a social need such as community safety or aged care. In such instances, both parties contribute as genuinely co-equal partners, with neither party taking a lead, and neither can be held accountable for the joint outcome.

In sharp contrast, much private-sector provision of public services is organized in terms of contracts between governments as purchaser and private-sector organizations as providers, arrangements that are clear examples of principal–agent relationships. Even many so-called “public–private partnerships” (PPPs), for example, those commonly used for constructing public infrastructure, are not genuine partnerships between coequals, but rather open-ended forms of relational contract where many details are left unspecified, but governments retain overall responsibility for the outcome.

Contracting out (outsourcing) certainly leads to difficulties over holding contractors to account, such as how to monitor performance and how to manage the clash between public service values and private profit, problems that have received much attention from public administration scholars (e.g., Lane, 1999; Vincent-Jones, 2006). But these are typical examples of well-known difficulties facing principals trying to control opportunistic agents, in principle similar to those that occur within any complex organization, including within government bureaucracies. The government as purchaser/principal retains ultimate responsibility for the contract and can therefore be held accountable for any outcomes. Institutional separation of purchaser and provider may extend the span of control and give rise to more complex problems of securing compliance. But these problems of accountability are still to be understood and managed within a principal–agent relationship.

Many public–private arrangements fall between these two contrasting poles of co-equal cooperation and purchaser–provider subordination. For example, some shared operations between governments and charitable organizations involve transferring public funds to the community organization through a form of contractual arrangement even if the charity continues to make an independent contribution based on its own core values. The conflict between these two sets of imperatives, the duty to the government as paymasters and the duty to pursue the charity’s core mission, has led to frequent internal pressures within nonprofit contractors (Sullivan & Skelcher, 2002) and consequent clashes of accountability. With such hybrid arrangements, a key issue for accountability is whether—and, if so, to what extent—the private-sector partner retains a share of separate responsibility and can therefore be held to share accountability in any eventual joint outcome.

Faced with an undoubted accountability deficit because of the decentered nature of networks and the absence of a clear principal–agent relationship, analysts of networks have sought to focus on alternative aspects of accountability that do not depend so immediately on centralized enforcement. In particular, the first two stages of accountability (information and discussion) can certainly apply across jurisdictional boundaries (Papadopoulos, 2007). For example, the news media can investigate and report on the outcomes of joint programs where responsibility is shared between different regimes (Schillemans, 2011). So too can consumer groups and community organizations (Michels & Meijer, 2008). Individual citizens can also share their own experiences of government networks through social media platforms.

In terms of transparency and access to information, the comparative openness of networks due to their plural membership can be an advantage. This can be seen most obviously in federal systems where the different levels of government sharing responsibility for a particular function offer alternative channels of inquiry and communication for interested members of the affected public (Galligan, 1995). Although individual bureaucratic organizations can more successfully resist demands for public engagement—the obverse consequence of their centralized power structures—shared responsibility between more than one organization—provides opportunities for inquiring citizens to have another bite at the cherry. If rebuffed at one level of government, they may have recourse to another level, a tactic regularly employed by well-organized interest groups. Networks, that is, have the potential to be more transparent than monolithic hierarchies. Much depends, however, on the nature of the network, on the extent to which its members, collectively and individually, are open rather than closed and outward looking rather than inward looking (Mulgan, 2003).

One accountability strategy building on transparency and potentially compensating for the absence of a single identifiable agent has been the application of performance measurement. If members of a network can agree on their joint objectives and how to measure success in meeting these objectives, the network as a whole can be held publicly accountable for its performance. This approach is implicit, for example, in the United Kingdom’s use of targets to monitor the performance of networks tasked with providing health and social welfare (Skelcher, 2010). It has also been adopted in a number of federal systems, including the United States, Canada, and Australia, as a means of coordinating and monitoring outcomes in policy areas such as health and education where two levels of government share responsibility. However, in spite of optimistic rhetoric in terms of commitments to “partnerships” and “transparency,” experience has often fallen short. Performance targets are notoriously difficult to define and measure (e.g., Anderson & Findlay, 2010; Radin, 2009), while network members, such as state and provincial governments, face their own political and constitutional imperatives that militate against any de facto surrender of sovereignty beyond temporary arrangements of convenience.

Moreover, as with many so-called “partnerships” between public and private organizations, the relationships are often far from equal. The Anglo-sphere federations, for example, all exhibit significant “vertical fiscal imbalance,” in which the central government raises more revenue than it needs, allowing a surplus to be distributed to state and provincial governments for designated purposes. This casts the central governments more in the role of paymaster, purchasing services from the states and provincial governments, and turns the relationship more toward that of principal and agent. Indeed, the initiative for the complex regimes of performance measurement usually comes from central governments wanting to control how their funds are spent.

Nonetheless, performance measurement remains a potentially useful accountability device for genuine networks based on shared values and cooperation (Negoita, 2018). In such cases the joint objectives and required performance information develop incrementally from continuous negotiation and adjustment by the various network members rather than being the product of top-down fiat (Conteh, 2016). As with transparency generally, information sharing does not imply any direct mechanism for rectification. But, through the force of publicity and the fostering of inter-organizational trust, it may help to stimulate positive responses from those with the necessary powers of decision.

Network accountability can be particularly effective at the local interface between street-level bureaucrats and the public. Particularly in the area of social policy, “micro-networks” of committed public-service professionals from different organizations often share in providing complementary sets of services to individual members of the public and local communities (Hupe & Hill, 2007). Provided they are given sufficient discretion to make their own decisions, they can cooperate with one another in meeting their clients’ needs and can offer their clients more direct accountability for government services than is provided through standard hierarchical chains of command (Hendriks, 2009; Stoker, 2006). Because they are closely engaged with one another and with their communities, groups of street-level bureaucrats can respond quickly and effectively to individual grievances without having to negotiate detailed and often inappropriate instructions from their separate bureaucratic superiors. This grass-roots accountability sometimes occurs through local groups and associations where officials and their clients discuss issues of common interest. Such relationships exhibit a form of mutual “participatory accountability” (Hupe & Hill, 2007), which borders on co-production because clients are joining with officials to deliver services.

Although hierarchical direction is never totally absent from street-level bureaucracy, motivation for this bottom-up collective responsiveness to their clients depends largely on other factors, for instance, the officials’ commitment to values of professionalism and concern for their reputation. To these can also sometimes be added market-based incentives in the case of network members employed under commercial contracts (Thomann, Hupe, & Sager, 2018). Wishing to renew their existing contracts and to attract additional business in a competitive market, commercial contractors have a strong motivation to cooperate with other partners in collectively responding to the clients they are paid to serve.

Accountability Within Networks

How networks are accountable is not just a question of their external accountability to the publics they serve. It also relates to their internal accountability: how the various members answer to one another within the network. Indeed, much of the literature on network accountability adopts a primarily internal focus (e.g., Agranoff & McGuire, 2001; Considine, 2002; Romzek, LeRoux, Johnston, Kempf, & Piatak, 2014), concentrating more on the problems of the relationships between members of the networks involved with service delivery, and how they are mutually accountable, rather than the larger issue of the relationship between networks and the public. Internal and external perspectives, however, regularly overlap when much of the subject matter of internal communication concerns dealings with external clients, as discussed in “The Accountability of Networks,” and when members of the network hold one another accountable for how their clients are treated. Moreover, the lines between internal provider and external client can disappear altogether in forms of co-production where members of a community cooperate with public officials as members of a network jointly producing policy outcomes, and thus perform a dual role of both holding to account and being accountable (Levasseur, 2018; Tuurnas, Stenvall, & Rannisto, 2016).

Internal network accountability is built on relationships of mutual cooperation between network members in which each accepts obligations to answer to the others on issues relating to their joint enterprise. The relationships are often informal in structure, built on a variety of connections, such as personal contacts, common membership of inter-organizational committees, memberships of professional associations, and so on (Romzek et al., 2014). Such informal networks occur at all levels of complex organizations, both within individual organizations and between members of different organizations. Within public administration, most attention has been directed toward networks of professionals providing services to the public and their mutual relations of professional accountability (Day & Klein, 1987; Romzek & Dubnick, 1987). Many social-work networks, for example, include formally trained professionals who share a common set of norms and attitudes that foster effective cooperation and informal monitoring. Though some professions, such as medicine, clinical psychology, and law, include their own formal accountability mechanisms to deal with professional malpractice, professional accountability extends much more widely than hierarchical disciplinary regulation and includes informal communication, which can have a similar effect of reinforcing professional norms and standards. It can also stimulate continuous learning (Conteh, 2016), which is recognized as one of the objectives of accountability (Bovens et al., 2008).

Because internal network accountability depends on personal familiarity and regular contact, it is most effective in small groups, such as between colleagues in street-level service centers. At the same time, it is also vulnerable to countervailing pressures on individual members that undermine interpersonal trust, for instance, frequent turnover of staff, the need to follow hierarchical directions from home organizations, and inter-organizational turf battles (Romzek et al., 2014). Reliance on network-based accountability can also encourage inward-looking cliquishness, in which the interests of professions are preferred to those of the publics they are supposed to serve. Again, a key factor is the extent to which individual network members are themselves externally accountable to other sources of authority.

Extending the Meaning of “Accountability”?

The types of accountability in networks discussed so far, by dispensing with hierarchical authority, lack a distinct relationship between principal and agent. But they show a clear family resemblance to the accepted understanding of accountability by requiring an obligation to answer to someone else for duties performed. The obligation may not be directly enforceable, but it is often powerfully encouraged by social, political, and personal pressures. On this view, accountability both by networks and within networks unequivocally counts as genuine forms of accountability (Popp et al., 2014). Difficulties still arise in realizing the practical benefits of accountability, as they do within hierarchical structures of accountability, but there should be little doubt about applying the general concept.

More contestable, however, is the extension of the term “accountability” to cover types of responsiveness that do not depend on an obligation to answer to others. The contrast can be seen within different understandings of professional accountability. In the sense referred to above, professional accountability, like other forms of network-based accountability, is social in operation, depending on interpersonal relationships between individuals? It obliges professionals to answer one another in formal and informal settings, discussing professional issues of mutual concern and learning from one another’s experiences.

This type of professional accountability can be distinguished from another view of professional accountability, closer to professional responsibility, which depends solely on personal values and choice. In this sense, the individual professional may be acting in the interests of his or her clients, but without actually answering to these clients or to professional colleagues. Any “answering” or “responding” is purely internal, for example, to the person’s professional conscience (Sinclair, 1995).

Strict defenders of the traditional understanding tend to exclude these cases as taking the concept beyond its basic sense of obligation to answer to an outside authority (e.g., Mulgan, 2003). On this view, social workers who choose to act in what they judge to be the interests of their clients may be acting responsibly and professionally toward their clients, but unless they are being actually held to account by these clients or by colleagues, they are not acting accountably.

Similarly, on the strict interpretation, market accountability refers only to those aspects of market transactions, such as sale and purchase, which give rise to contractual obligations where purchasers can hold vendors to account. When market players take their own initiatives to retain or attract customers they may be aligning their behavior with the interests of actual or potential customers or clients, but they are not acting accountably to these customers or clients. To equate such voluntary responsiveness with accountability is to downplay the importance of obligatory response to others and the threat of potential consequences that are at the heart of the accountability relationship.

On the other hand, such a conceptual limitation may be too restrictive in excluding actions and behavior that are often counted as genuine instances of accountability. Certainly, many public officials who conscientiously choose to act in the interests of the public regularly claim to be acting accountably even when they are under no external obligation to answer for their performance. If asked to whom or what they are accountable, they are likely to give some internal authority, such as their ethical standards or professional conscience. To deny their accountability would be equivalent to denying their integrity as professionals. In this sense, as Bovens has argued, accountability is understood less as an administrative mechanism and more as an organizational virtue, similar in meaning to “responsiveness” and “a sense of responsibility” (Bovens, 2010). In the interests of clarity, it may sometimes be useful to require that accountability stand for some form of external relationship and obligation, even one without a clear principal and agent. But the emotional tug of the concept is so strong that such limits will often be ignored and need not be insisted on.

Accountability Without Democracy Through Institutions of Horizontal Accountability

Analysis of network-based accountability has assumed a context of stable democratic states providing an overall framework of democratic accountability through regular free elections and a bureaucratic hierarchy accountable to elected leaders. This framework sets strong limits to the discretion of networks and their members while also guaranteeing reliable processes within which they can operate. However, most states do not meet these conditions. Apart from the thirty or so established democracies, a much larger number of regimes can be classified as partial democracies where political leaders have been successful in a nationwide election but where significant doubts remain about the integrity of the electoral process, including the opportunities offered to alternative candidates and parties, as well as the administration of the actual ballot. In such circumstances, elections fail to deliver their promise of holding political leaders properly to account and of making the bureaucracy indirectly accountable to the voters.

Much of the discussion of accountability in countries that lack effective electoral processes has been framed within theories of “developing” or “transitional” government in which regimes are viewed as progressing toward the status of fully-fledged representative democracies. Analysis of accountability therefore becomes a matter of how far non-democratic governments fall short of this model and what steps need to be taken to get closer to it. However, the unilinear assumptions behind this thinking are increasingly being called into question as influential authoritarian states, particularly China, assert their own independence and their right to be considered as legitimate forms of government rather than as failed imitations of Western models.

Regardless of what normative perspective is adopted, representative democracy is not the only constitutional structure within which accountability can occur. The absence of effective democratic accountability through the electoral chain of principals and agents still leaves room for alternative accountability arrangements that can help to make governments answerable to their citizens. In this way, accountability can be provided by institutions of horizontal accountability, such as courts and government auditors, which are independent of the executive arm of government and are capable of being developed separately from the electoral process. The significance of these institutions has been stressed within the subfield of development politics, which did much to popularize the concept of horizontal accountability within political science (e.g., O’Donnell, 1998, Schedler, Diamond, & Plattner, 1999). Although mainstream political science, with its focus on developed democracies, had concentrated on elections as the defining institutions of democratic accountability, scholars in development studies saw that institutions of horizontal accountability were worth nurturing on their own. Not only were they vital for creating effective representative democracies, but they also had their own independent value in offering protection for citizens from governments that abused their power.

The value of institutions of horizontal accountability is connected with liberal-constitutional theories of government, which are incorporated into modern representative democracy but are conceptually distinct from the democratic principle of popular sovereignty. Historically, for example, in Europe some key mechanisms of horizontal accountability, such as judicial review of executive action, emerged well before the development of voting rights for all adult citizens. Some present-day analyses of democracy place emphasis on the role of democracy in monitoring and constraining the power of government rather than its responsiveness to public opinion (e.g., Keane, 2009).

What institutions count as providing horizontal accountability has been a matter of disagreement. Analyses stressing legal and constitutional constraints on government tend to focus on constitutionally established institutions such as courts and tribunals as well as watchdog officials such as ombudsmen and government auditors (e.g., O’Donnell, 1998). Those looking more broadly for general scrutiny of government, on the other hand, tend to include institutions such as organizations of civil society and the media (e.g., Schillemans, 2011).

Anti-corruption agencies provide a good illustration of how legally horizontal accountability institutions can operate independently of democracy. Such agencies became common during the 1990s and early 2000s as responses to concerns from international agencies, such as the World Bank, about weakness of governance structures and problems of corruption in developing countries. Their popularity owed much to the success of two pioneering examples, Singapore’s Corruption Practices Investigation Bureau (established in 1952) and Hong Kong’s Independent Commission Against Corruption (established in 1974), both, significantly, products of non-democratic regimes.

Subsequent anti-corruption agencies have met with mixed success, often proving incapable of making significant inroads into widespread corruption because they have lacked sufficient resources and support from surrounding institutions, including the political leadership (de Sousa, 2010). By contrast, the political leaders in Singapore and Hong Kong were fully committed to their respective campaigns against corruption (Quah, 2010). Anti-corruption agencies can help reduce corruption, but they need the backing of politicians, who are more often a major part of the problem, especially in regimes with weak legal systems. They also require an effective judicial system to underpin their proceedings and help enforce sanctions. International experience with anti-corruption agencies illustrates the pitfalls of assuming that an accountability agency that has been effective in one constitutional and political context will be equally successful in another. Much depends on the local conditions, especially the relative robustness of administrative and legal structures.

Social Accountability

In response to the perceived weaknesses of state-based accountability institutions, both vertical and horizontal in structure, advocates of improved accountability in developing countries have sometimes looked beyond the state altogether, to the institutions of civil society. The World Bank, in an influential analysis of how to make service providers accountable to their clients (World Bank, 2004), distinguished between two routes to accountability. The “long route” required citizens to seek accountability through state actors, including politicians, regulators, and bureaucrats, and was often unreliable. The “short route,” by contrast, was more direct, allowing citizens to bypass the state procedures and demand accountability directly from providers. This latter type of direct accountability, often referred to as “social accountability,” an admittedly imprecise term (Odugbemi & Lee, 2011), relies on various forms of citizen engagement and empowerment and provides a central role for NGOs (non-government organizations). It requires citizens to shift from a passive stance as receivers of services to a more active role as informed participants, either as individuals or collectively in social groups. Though social accountability dispenses with the standard state mechanisms of accountability, including democratic elections, it is still typically constructed within a principal–agent framework, with citizens as principals seeking answers directly from frontline service providers as their agents.

In the decade or more since the theory of social accountability was first popularized, international donors have allocated extensive funds toward projects that can be loosely classified under that heading. Overall, however, experience has generally failed to match the optimistic rhetoric (Brinkerhoff & Wetterberg, 2015). There have been some success stories, for example, in the Philippines and Indonesia, where schemes to devolve responsibility to community groups have been able to tap into well-established traditions of grass-roots activism. In many countries, however, civil society is too weak to offer meaningful engagement with government agencies. Concerned individuals face familiar collective action problems of converting their isolated personal interests into effective joint associations, particularly in the face of apathetic or hostile officialdom.

Moreover, instances of successful citizen interaction of citizens with government generally depend on some level of positive support and encouragement from the government side. For example, civil society groups can certainly make good use of government performance information to pinpoint areas of weakness, but only if the government collects accurate information and is prepared to make it publicly available. More broadly, any active involvement by civil society depends on government willingness to safeguard a space for public discussion, through at least minimal commitment to freedom of association and communication.

For these reasons, social accountability does not offer a clear alternative to state-based accountability mechanisms. Civil society can certainly assist in making governments more publicly accountable in a great variety of ways, through public criticism and direct engagement with government, including arrangements of partnership and co-production. Like all watchdog institutions, associations of civil society with a focus on government can keep an eye out for inefficient or corrupt behavior by government officials. But their capacity to make a significant difference is conditional on the functioning of certain elements of government. These elements vary according to the national context but include aspects of law enforcement, government transparency, consultation procedures, and so on. Civil society associations cannot provide an effective, direct link between citizen and government quite separate from the state’s own institutions of accountability, which remain essential to any effective system of public accountability.

Government-Sponsored Accountability in Authoritarian States

As the world’s leading authoritarian state, China provides evidence of how accountability mechanisms can operate within a government structure where a single political party retains central control over both executive government and the judicial system. Though the central authority of the party is unquestioned and beyond overt challenge, members of the ruling elite are not unaware of the benefits of subjecting government officials to open scrutiny. They are therefore prepared to tolerate, and indeed sometimes encourage, processes of accountability that make officials accountable to the public without directly challenging the dominant role of the party. In this way, accountability can be seen as a means of co-opting public opinion to exert control over wayward bureaucracies, thus cementing the regime’s legitimacy.

For example, as part of opening up its program of economic liberation, China adopted a freedom-of-information mechanism (Regulations on Open Government Information) that requires all levels of government to establish open government systems for the agencies in their jurisdiction, allowing access to certain categories of administrative and economic information (but not information relating to party matters or the courts) (Piotrowski, Zhang, Lin, & Yu, 2009). The main rationale has been to improve the administrative and economic efficiency of government agencies, including state-owned enterprises, though experience has been patchy because of the devolved responsibility for implementation and the traditional secrecy of professional bureaucrats.

A similar but more comprehensive scheme covers environmental issues, requiring the publication of information on environmental protections, public complaints, and violations of regulations. The intention is to reduce the number of public protests against pollution and other types of environmental damage by harnessing civil society groups as watchdogs. Although civil society is often weakly resourced, international NGOs have sometimes lent their assistance by pressuring multinational companies operating in China to insist on improved environmental standards from the more local levels of government, especially in townships and villages, where central party control is both more attenuated and less at risk (Tan, 2014).

Other accountability initiatives have also occurred at the local level. For example, the country-level city of Wenling in Zhejiang Province has established consultation assemblies in each of its constituent townships and village (Hsu, 2009). These assemblies discuss a number of local governance issues, including the annual budget and official salaries as well matters of community importance. Although their decision-making power varies and is potentially subject to overrule from above, the assemblies provide forums at which local cadres must explain their actions to citizens. Failure to make a good impression at the meetings is likely to tarnish the reputation of individual cadres and thus affect their future careers. Another type of grass-roots accountability occurs in rural China, where small groups based on shared values and interests exhibit “accountability without democracy” (Tsai, 2012) by embedding local officials within them as members. The social solidarity binds the officials to the rest of their groups and ensures that they are responsive to their local communities.

These initiatives, though far from universal and often local in focus, provide a model of how certain types of accountability are possible within an authoritarian regime. Much depends on the strength and independence of civil society organizations and the media, including social media, which can foster forms of social accountability (Ma, 2012). This in turn depends on the willingness of the state to tolerate public scrutiny of its activities. Enlightened self-interest may encourage governments to accept some level of public criticism not only as a means toward more effective administration by their officials but also, more broadly, as a device to enhance the regime’s legitimacy with its citizens. Accountability, as a legitimating mechanism, is more usually associated with liberal-democratic governments grounded in the individual rights of citizens, whereas non-democratic regimes typically depend on other strategies for affirmation, such as mass mobilization and propaganda (Beetham, 2013). However, stable authoritarian states may see some advantage in making certain accountability rights available to their citizens. At the same time, there will always be limits to such openness set by the regime’s over-riding imperative for survival and by the personal ambitions of individual cadres unaccustomed to answering directly to the public.


Though government accountability to the public is often understood and analyzed within a framework of principal–agent obligation in representative democracy, it is increasingly seen to be operating outside this framework. For example, cross-organizational networks without any clear agent, and often also lacking a clear principal, are still expected to be accountable to members of the public they serve. In addition, internally within particular networks, individual network members are accountable to one another without any formal subordination. To be complete, accountability requires not only transparency and discussion but also, where necessary, the imposition of sanctions and other consequences, which are most readily guaranteed within a structure of hierarchical authority. In the absence of such authority, other motivations for compliance are called into play, including concern for reputation, professionalism, and economic self-interest. In practice, most network members also belong to their own to bureaucratic hierarchies, which can offer supplementary reinforcement. But, even considered on their own, systems of network-based accountability are important mechanisms in the general arsenal of accountability weaponry.

Accountability can also be found in regimes that lack effective elections and a chain of bureaucratic accountability from rank-and-file officials upward to elected leaders. For instance, constitutional institutions of horizontal accountability can provide citizens with the means of holding governments to account through independent legal processes that do not depend on democratic legitimation. Civil society is also a potent source of accountability pressures, particularly on frontline bureaucrats. However, to be effective, mechanisms of accountability usually require the support of robust legal systems and the willingness of the executive government to submit to legal constraints, requirements associated with the rule of law. Indeed, if there is any constitutional doctrine that is essential for accountability, it is the rule of law rather than electoral democracy.

These wider applications of accountability confirm that the core thrust of public accountability, the demand that governments answer to those affected by their decisions, has arisen in many more types of institutional setting than is sometimes acknowledged. Although analysis of particular accountability situations will need to be flexible and context dependent, there is no need to posit new meanings for the wide-ranging concept of accountability itself.


  • Agranoff, R., & McGuire, M. (2001). Big questions in public network management research. Journal of Public Administration Research and Theory, 11(3), 295–326.
  • Alford, J., & O’Flynn, J. (2012). Rethinking public service delivery. Basingstoke, U.K.: Palgrave Macmillan.
  • Anderson, L., & Findlay, T. (2010). Does public reporting measure up? Federalism, accountability and childcare policy in Canada. Canadian Public Administration, 53(3), 417–438.
  • Bache, I., & Flinders, M. (Eds.). (2004). Multi-level governance. Oxford, U.K.: Oxford University Press.
  • Beetham, D. (2013). The Legitimation of Power. 2nd ed. Basingstoke, U.K.: Palgrave Macmillan.
  • Bovens, M. (1998). The quest for responsibility. Cambridge, U.K.: Cambridge University Press.
  • Bovens, M. (2007). Analysing and assessing accountability: A conceptual framework. European Law Journal, 13(4), 447–468.
  • Bovens, M. (2010). Two concepts of accountability: Accountability as a virtue and as a mechanism. West European Politics, 33(5), 946–967.
  • Bovens, M., Schillemans, T., & ’t Hart, P. (2008). Does public accountability work? An assessment tool. Public Administration, 86(1), 225–242.
  • Brinkerhoff, D. W., & Wetterberg, A. (2015). Gauging the effects of social accountability on services, governance, and citizen empowerment. Public Administration Review, 76(2), 274–286.
  • Considine, M. (2002). The end of the line? Accountable governance in the age of networks, partnerships, and joined-up services. Governance, 15(1), 21–40.
  • Conteh, C. (2016). Rethinking accountability in complex and horizontal network delivery systems. Canadian Public Administration, 59(2), 224–244.
  • Dahl, R. (1999). Can international organizations be democratic? A skeptic’s view. In I. Shapiro & C. Hacker-Cordon (Eds.), Democracy’s edges (pp. 19–36). Cambridge, U.K.: Cambridge University Press.
  • Day, P., & Klein, R. (1987). Accountabilities: Five public services. London, U.K.: Tavistock.
  • Dubnick, M. (1998). Clarifying accountability: An ethical theory framework. In C. Sampford, N. Preston, & C.‑A. Bois (Eds.), Public sector ethics: Finding and implementing values (pp. 68–81). London, U.K.: Routledge.
  • Galligan, B. (1995). A federal republic. Cambridge, U.K.: Cambridge University Press.
  • Gregory, R. (2017). Accountability and responsibility. In Oxford Research Encyclopedia on Politics.
  • Harlow, C. (2002). Accountability in the European Union. Oxford, U.K.: Oxford University Press.
  • Held, D., & Koenig-Archibugi, M. (2004). Introduction. Government and Opposition, 39(2), 125–131.
  • Hendriks, C. M. (2009). The democratic soup: Mixed meanings of political representation in governance networks. Governance, 22(4), 689–715.
  • Hsu, S. P. (2009). In search of public accountability: The “Wenling model” in China. Australian Journal of Public Administration, 68(s1), S40–50.
  • Hupe, P., & Hill, M. (2007). Street-level bureaucracy and public accountability. Public Administration, 85(2), 279–299.
  • Keane, J. (2009). The life and death of democracy. New York, NY: Simon and Schuster.
  • Lane, J. E. (1999). Contractualism in the public sector. Public Management, 1(2), 179–194.
  • Levasseur, K. (2018). Co-producing accountability? Drawing conclusions from non-profit child care services in Manitoba. Canadian Public Administration, 61(1), 26–44.
  • Ma, J. (2012). The rise of social accountability in China. Australian Journal of Public Administration, 71(2), 111–121.
  • Michels, A., & Meijer, M. (2008). Safeguarding public accountability in horizontal government. Public Management Review, 10(2), 165–173.
  • Mulgan, R. (2003). Holding power to account: Accountability in modern democracies. Basingstoke, U.K.: Palgrave Macmillan.
  • Negoita, M. (2018). Beyond performance management: A networked production model of public service delivery. Public Performance & Management Review, 41(2), 253–276.
  • O’Donnell, G. (1998). Horizontal accountability in new democracies. Journal of Democracy, 9(3), 112–126.
  • Odugbemi, S., & Lee, T. (Eds.). (2011). Accountability through public opinion: From inertia to public action. Washington, DC: World Bank.
  • Olsen, J. P. (2015). Democratic order, autonomy, and accountability. Governance, 28(4), 425–440.
  • Papadopoulos, Y. (2007). Problems of democratic accountability in network and multilevel governance. European Law Journal, 13(4), 469–486.
  • Pierre, J., & Peters, B. G. (2000). Governance, politics and the state. Basingstoke, U.K.: Macmillan.
  • Piotrowski, S. J., Zhang, Y., Lin, W., & Yu, W. (2009). Key issues for implementation of Chinese open government regulations. Public Administration Review, 69(s1): S129–S135.
  • Popp, J. K., Milward, H. B., MacKean, G., Casebeer, A., & Lindstrom, R. (2014). Inter-organizational networks: A review of the literature to inform practice. Washington, DC: IBM Center for the Business of Government.
  • Przeworski, A., Stokes, S. C., & Manin, B. (Eds.). (1999). Democracy, accountability, and representation. Cambridge, U.K.: Cambridge University Press.
  • Quah, J. S. T. (2010). Defying institutional failure: Learning from the experiences of anti-corruption in four Asian countries. Crime, Law and Social Change, 53, 23–54.
  • Radin, B. A. (2009). Challenging the performance movement. Washington, DC: Georgetown University Press.
  • Rhodes, R. A. W. (1997). Understanding governance. Public networks, governance, reflexivity and accountability. Buckingham, U.K.: Open University Press.
  • Romzek, B. S., & Dubnick, M. J. (1987). Accountability in the public sector: Lessons from the Challenger tragedy. Public Administration Review, 47(3), 227–238.
  • Romzek, B., LeRoux, K., Johnston, J., Kempf, R. J., & Piatak, J. P. (2014). Informal accountability in multisector service delivery collaborations. Journal of Public Administration Research and Theory, 24(4), 813–842.
  • Schedler, A., Diamond, L., & Plattner, M. (1999). The self-restraining state: Power and accountability in new democracies. Boulder, CO: Lynne Rienner.
  • Schillemans, T. (2008). Accountability in the shadow of hierarchy: The horizontal accountability of agencies. Public Organization Review, 8(2), 175–194.
  • Schillemans, T. (2011). Does horizontal accountability work? Evaluating potential remedies for the accountability deficit of agencies. Administration and Society, 43(4), 387–416.
  • Schillemans, T., & Busuioc, M. (2015). Predicting public sector accountability: From agency drift to forum drift. Journal of Public Administration Research and Theory, 25(1), 191–215.
  • Sinclair, A. (1995). The chameleon of accountability. Accounting Organizations & Society, 20(2–3), 219–237.
  • Skelcher, C. (2010). Fishing in muddy waters: Principals, agents, and democratic government in Europe. Journal of Public Administration Research and Theory, 20(2–3), i161–i175.
  • de Sousa, L. (2010). Anti-corruption agencies: Between empowerment and irrelevance. Crime, Law and Social Change, 53(1), 5–22.
  • Stoker, G. (2006). Public value management. A new narrative for networked governance? American Review of Public Administration, 36(1), 41–57.
  • Strøm, K. (2000). Delegation and accountability in European democracies. European Journal of Political Research, 37(3), 261–289.
  • Sullivan, H., & Skelcher, C. (2002). Working across boundaries: Collaboration in public services. Basingstoke, U.K.: Palgrave Macmillan.
  • Tan, Y. (2014). Transparency without democracy: The effects of China’s environmental disclosure policy. Governance, 27(1), 37–62.
  • Thomann, E., Hupe, P., & Sager, F. (2018). Serving many masters: Public accountability in private policy implementation. Governance, 31(2), 299–319.
  • Thompson, D. (1980). Moral responsibility of public officials: the problem of many hands. American Political Science Review, 74(4), 905–916.
  • Tsai, L. L. (2012). Accountability without democracy. Cambridge, U.K.: Cambridge University Press.
  • Tuurnas, S., Stenvall, J., & Rannisto, P.‑H. (2016). The impact of co-production on frontline accountability: The case of the conciliation service. International Review of Administrative Sciences, 82(1), 131–149.
  • Vincent-Jones, P. (2006). The new public contracting: Regulation, responsiveness, relationality. Oxford, U.K.: Oxford University Press.
  • World Bank. (2004). Making services work for poor people: World development report. Washington, DC: World Bank.