Accountability, Performance and Legitimacy in the Welfare State

If accountability is the answer, what was the question?

Authored by: Werner Jann

The Routledge Handbook to Accountability and Welfare State Reforms in Europe

Print publication date:  November  2016
Online publication date:  November  2016

Print ISBN: 9781472470591
eBook ISBN: 9781315612713
Adobe ISBN: 9781317044208




Accountability is one of the most widely discussed concepts of public administration research and teaching in the last decade. But why is this the case? Obviously, accountability is, like its counterpart transparency, a ‘magic concept’, and an indispensable part of the prominent and omnipresent discourse on ‘good governance’, as well as a significant element in debates about public sector reform (Pollitt and Hupe 2011). The same holds true for performance, which has been a magic and contested concept ever since New Public Management (NPM) entered the discourse about ‘modern’ processes and structures of the public sector. But the third term in the title of this paper, legitimacy, even though it is one of the basic concepts of political science and democracy and is at the heart of Max Weber’s theory of bureaucracy, has been surprisingly absent from current debates about the challenges of modern public administration, and for that sake also about the future of the welfare state. In this chapter, I will argue that different concepts of legitimacy lie at the heart of most debates about accountability and performance, and that a better understanding of the relationships between accountability, performance and legitimacy can clarify some of the puzzles of contemporary research.

 Add to shortlist  Cite

Accountability, Performance and Legitimacy in the Welfare State


Accountability is one of the most widely discussed concepts of public administration research and teaching in the last decade. But why is this the case? Obviously, accountability is, like its counterpart transparency, a ‘magic concept’, and an indispensable part of the prominent and omnipresent discourse on ‘good governance’, as well as a significant element in debates about public sector reform (Pollitt and Hupe 2011). The same holds true for performance, which has been a magic and contested concept ever since New Public Management (NPM) entered the discourse about ‘modern’ processes and structures of the public sector. But the third term in the title of this paper, legitimacy, even though it is one of the basic concepts of political science and democracy and is at the heart of Max Weber’s theory of bureaucracy, has been surprisingly absent from current debates about the challenges of modern public administration, and for that sake also about the future of the welfare state. In this chapter, I will argue that different concepts of legitimacy lie at the heart of most debates about accountability and performance, and that a better understanding of the relationships between accountability, performance and legitimacy can clarify some of the puzzles of contemporary research.

This chapter builds on recent research on how contemporary public sector reforms have been affecting accountability relationships in modern welfare states. A main focus in this research has been to describe, understand and explain how the balance between organizational autonomy and traditional forms of democratic accountability is changing, and how the trade-offs between political, administrative, managerial, legal, professional and social accountability are affected. What happens with political accountability as welfare states face reforms? Accountability in this kind of research is treated as a mechanism (and not as a virtue) (Bovens 2010), but still mainly as a ‘dependent variable’, because the aim is both to map and explain the changing nature and content of these mechanisms and their effects (Christensen and Lægreid 2001).

Empirical findings, among others, in labor market, welfare, health and migration policies have offered a wealth of information about changing mechanisms of accountability, the changing role of traditional political accountability, and about the impact of NPM and post-NPM reforms on accountability relationships (Jann and Lægreid 2015; see also contributions in this volume). But accountability, transparency and other magic concepts are not ends in themselves. The questions remain: What are they for, and why are they so popular?

My assertion is that the underlying rationality (i.e. the underlying aim or even ‘causal model’ of many management reforms), albeit mostly not very clearly expressed and perhaps not even well understood, was to enhance accountability and through that the performance and legitimacy of welfare arrangements. The term legitimacy was seldom used, but the background of many reforms was a growing concern that welfare states are seen as inefficient, do not achieve their objectives, are not sustainable in financial terms, and are therefore losing popular support. All kinds of New Public Management (NPM) inspired reforms were suggested and introduced, granting more autonomy to welfare agencies, transferring responsibilities to public managers, or even introducing some kind of market instruments and competition with private providers. All this so the underlying assumption would create more or ‘better’ accountability mechanisms, and thus better performance and in the end more support for welfare arrangements and the welfare state. This kind of reasoning was, for example, very prominent in Germany, where NPM reforms were and still are supported by many social democrats and were even proposed and supported by the trade union of the public sector (Jann 1997; Lægreid 2014).

The question is whether these assumptions hold true. In this chapter, I will use changing accountability relationships as the ‘independent variable’, trying to show whether and how they were supposed to and have influenced performance and legitimacy. In order to do this, I will look briefly and rather schematically into three different relationships (i.e. between accountability and performance, accountability and legitimacy, and performance and legitimacy). I am interested in legitimacy as the ultimate aim of managerial reform and accountability changes and I will try to clarify how these three ‘magic concepts’ interact, both normatively, theoretically and empirically.

Accountability and performance

The relationship between accountability and performance is supposed to be close. Sometimes performance, accountability and transparency are treated as nearly identical and simply axiomatic, ‘a good thing in itself’, a virtue (Talbot 2005; Bovens 2010), and as Barbara Romzek (2015: 28) states, ‘In its simplest sense accountability is answerability for performance, which, if it is working properly, should result in a reward or a sanction’ and ‘to do accountability you need performance information; accountability without performance information is a hollow concept’. But for some time both the empirical and theoretical relationships have been widely discussed in the literature and are contested. Even the direction of causality is unclear, some assume accountability enhances performance, while others claim it is the other way around (for summaries, see Pollitt 2011; Lægreid 2014; Christensen and Lægreid 2015; Jantz et al. 2015).

Normatively, this relationship is at the core of the NPM agenda. NPM assumes that new forms of performance management (i.e. output and outcome measurement, contract and result steering, incentives and sanctions) between and in public organizations will enhance the performance of these organizations. Thus, new forms of managerial accountability between principles (e.g. ministers and ministries) and agents (e.g. new forms of public agencies and new types of public managers; i.e. not traditional bureaucrats and bureaucracies) would make these agencies perform better by clarifying organizational goals, diminishing problems of information asymmetries, reducing drift, shirking or whatever undesirable behaviors of agencies or bureaucrats were suspected. This is also the core agenda of principal-agent theory.

The assumed relationship works in two steps. On the one hand, it is presumed that performance reporting and management will make public organizations more accountable and responsive, both to political principles but also ultimately to clients, citizens and the general public (performance enhances accountability). By providing information not only about inputs but also about intended and achieved results (i.e. outputs and outcomes), new forms of evidence and data are generated, which improve and alter the debate between administrative actors and different forums, which are at the core of each accountability relationship (Bovens 2007: 452). At the same time, by providing new and better information, ultimately also positive or negative incentives or sanctions can become better targeted and more successful. Combined with more managerial autonomy, this will give managers and specialized, single purpose agencies clearer goals, more flexibility, more incentives and better instruments to achieve desired results (accountability enhances performance).

Both assumptions are empirically contested and the relationship between accountability and performance is characterized by tensions, ambiguities and contradictions. There is no straightforward coupling between accountability and performance (Christensen and Lægreid 2015). Even friendly observers conclude that:

the positive effects of performance management on the steering and controlling of public sector organizations and on their performance have been overestimated . . . different variants of misuse and several dysfunctional effects of performance measurement exist, for example symbolic use of PIs, negligence of qualitative aspects, misuse of data for personal interests, and so on.

(van Helden and Reichard 2013: 13)

Most of the empirical evidence about the problematic and uncertain relationship between accountability and performance can be summarized under the broad and well-known concepts of bounded rationality, opportunistic behavior and unintended consequences.

First, performance measurement and management are confronted with problems of task complexity, contradictory goals and values, and ambiguous causality. It is very often hard to define and measure government performance, and it is even harder to establish clear causalities between management instruments, organizational behavior, and policy outputs and outcomes. At the same time, more managerial accountability may lead to accountability overload or even ‘multiple accountabilities disorder’ (Koppell 2005), and in the end to opportunistic behavior, to manipulating numbers (hitting the target and missing the goal), gaming, target ratcheting and other forms of opportunistic behavior (Jann and Jantz 2008). The more information is produced, the larger are the incentives to manipulate this kind of information. And finally, there is the unintended effect of what Pollitt (2011) has called the ‘tyranny of light’. The more information is produced and processed, the more distrust between actors develops (Flinders 2011).

If the relationship between (managerial) accountability and performance is complex and blurred at best, the question remains: Why are both concepts so popular, and why especially has accountability become such an important concept of public sector reform? Recently, the ‘why’ and ‘what for’ of accountability have been addressed from the perspective of ‘reputation management’ (Busuioc and Lodge 2015). The assertion is that accountability is not about reducing informational asymmetries, reducing drift, shirking or whatever, but it is about managing and cultivating one’s reputation vis-à-vis different audiences. It may be more ‘path-dependent’ and appropriate than consequential. It is about being seen as competent, reliable, fair, performance-oriented, etc. This way, it is possible to explain why ‘those supposedly holding to account are not particularly interested in this task, while those supposed to give account do so through distorted information, and/or with motivation-depleting results’ (Busuioc and Lodge 2015: 3). Accountability ‘serves as a way to justify one’s existence and can therefore become central to an organization’s . . . sense of identity’ (Busuioc and Lodge 2015: 4). In this understanding, reputation is a source of bureaucratic power, allowing organizations to ‘strengthen their autonomy, build alliances, enlist political support and ultimately, help ensure their survival’ (Busuioc and Lodge 2015: 5).

This definition of reputation comes very close to legitimacy (see also the distinction of technical, moral, procedural and performative competencies, which look very much like Suchman’s different forms of legitimacy) (Suchman 1995). The main difference seems to be that reputation can be understood as manipulative or manipulated, it is probably not deserved, it is only about appearing to be successful, reliable, fair and so on, like in the famous quip of Groucho Marx, ‘The secret of life is honesty and fair dealing. If you can fake that, you’ve got it made’.

But this points towards a rather narrow and manipulative concept of accountability and legitimacy. Reputation and legitimacy are seen as something that can be managed like anything else, so in the end this appears to be only a new twist of the managerial revolution, where you do not try to manage performance, but reputation. Obviously, the relationships between accountability and legitimacy are more complex, and also legitimacy is a broader concept than reputation, so in order to understand the problematic relationship between accountability and performance, we have to come to a better understanding of the concept of legitimacy.

Accountability and legitimacy

Opposite to the contested relationship between accountability and performance, the relationship between accountability and legitimacy has not been discussed in great detail (but see March and Olsen 1995: 143ff.; Moes 2009; Rothstein 2012; Olsen 2013; Moore 2014; Olsen 2015). The reason seems to be the very fundamental, complex and comprehensive nature of the legitimacy concept. There are, as mentioned, some ‘practitioners’ theories’ assuming that more accountability will not only lead to better performance, but will also enhance the legitimacy of public decisions, policies, organizations and, in the end, the legitimacy of the modern welfare state, but these are based more on ‘intuition’ than on systematic theoretical or empirical evidence. Considine and Afzal (2011: 375) suppose that ‘those elements of NPM that increase responsiveness, tailoring of services, and efficiency may help drive stronger attachment of citizens and élites to key state institutions’. But at the same time, they also speculate ‘those that push services out to contractors who then care little for creating public value and only seek to maximize profits will have the opposite impact on legitimacy’.

In order to clarify this entangled relationship, it is helpful to systematize what is meant by legitimacy, and how different meanings possibly can be operationalized. As with most fundamental concepts in the social sciences, such as democracy, authority or autonomy, legitimacy is a contested concept. But legitimacy is also a central concept in neo-institutional theory and in the most comprehensive survey of the large and diverse literature on organizational legitimacy, the author concludes that ‘many researchers employ the term legitimacy, but few define it’ (Suchman 1995: 572). In order to overcome this unsatisfactory state of affairs, it makes sense to start from Suchman’s broad summary definition of legitimacy as a ‘generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions’ and further distinguish legitimacy as a relationship between an actor or organization and an audience (or forum) based on self-interest (pragmatic), on normative approval (moral), or on taken for grantedness (cognitive) (Suchman 1995: 574; see also Scott 2013).

In addition, it is helpful to introduce a distinction, originally suggested by Fritz Scharpf (1970, 1999) between input and output legitimacy, and its recent extension by Vivien Schmidt towards throughput legitimacy. In this approach, the general definition and distinction of legitimacy builds on Weber (and later Easton), and relates to the extent to which (input) politics, (throughput) processes and (output) policies are acceptable to and accepted by the citizenry, such that citizens voluntarily comply with government acts and decisions, even when these go against their own interests and desires (Scharpf 1999; Schmidt 2013). Finally, it is important to remember that, again following Weber, ‘legitimacy is a belief, an empirical phenomenon to be established as other socially relevant subjective phenomena, and not something to be inferred from compliant behavior nor deduced from the presence or absence of its presumable determinants’ (Mayntz 2011: 138).

Unfortunately, this broad concept of legitimacy is very hard to operationalize. Concomitant concepts that are often used as proxies are ‘trust’, in institutions, organizations and persons, or even generalized social trust, ‘confidence’ in or ‘satisfaction’ with institutions and actors, or, as mentioned, their ‘reputation’ (Newton 2001; Christensen and Lægreid 2005; Van de Walle 2013; Greiling 2014). Suchman concludes his survey of the literature with the not very optimistic observation, that researchers currently possess little systematic knowledge and evidence comparing the effects of organizational activities on multiple types of legitimacy, or comparing the effects of multiple types of legitimacy on organizational outcomes. Legitimacy as a concept is not homogeneous, and the different facets of legitimacy are not always compatible (Suchman 1995: 602). This paper cannot solve the complex puzzle of empirical ‘measures’ for legitimacy, but in order to better understand the composite nature of legitimacy and to point towards possible future research, it makes sense to more clearly distinguish its main characteristics.

Following Scharpf (1999: 7ff.; see also Peters 2013), input legitimacy thus refers to the participatory quality of the democratic process leading to government policies, laws and rules as legitimized by the traditional ‘majoritarian’ institutions of electoral representation. It encompasses the democratic principle of ‘rule by the people’, meaning that political decisions are derived from the preferences of the population in a chain of accountability, linking those governing to those governed. Democratic legitimacy thus requires mechanisms or procedures to link political decisions and government policies with citizens’ preferences. In modern democracies, these mechanisms are mostly reflected in representative institutions in which political decision-makers can be held accountable by the means of elections. The democratic doctrine of the German constitutional court, for example, frames this as the ‘uninterrupted chain of legitimacy’, from voters via elections, parliaments, governments, ministers, ministries and agencies, to policies and governmental acts. Input legitimacy, as Schmidt stresses, depends on citizens expressing demands institutionally and deliberatively through representative politics while providing support via their sense of identity and community.

But this is not the only form of legitimacy, even in modern democracies. It is accompanied by output legitimacy (i.e. the capacity of a government or institution to solve collective problems and to meet the expectations of the governed citizens). ‘Output legitimacy requires policies to work effectively while resonating with citizens’ interests, values and identity’ (Schmidt 2013: 7), or, as Scharpf (1970) argues, democracy would be an ‘empty ritual’ if the democratic procedure was not able to produce effective outcomes (i.e. ‘achieving the goals that citizens collectively care about’). Citizens expect governments to solve collective problems, or at least alleviate them. A system that is exemplarily democratic, meeting the highest participatory standards, but which cannot guarantee acceptable policies, for example levels of safety or social standards, will not be legitimate and will not be able to survive. On the other hand, a system that ‘provides the goods’ but does not meet at least minimum democratic standards is also not seen as sustainable. This idea of the ‘good dictator’ is rejected on normative-theoretical grounds, since citizen participation in public affairs is seen as a value in itself, as an indispensable part of human emancipation. It is also rejected on analytical-theoretical grounds, since only open discussions, negotiations and communicative rationality will generate mutual understanding, adaptation and improvement through exchange of reasons and arguments and, in the end, experience-based learning. A core assumption of modern democratic theory is that democracies have a unique ability to learn from experience (Olsen 2013, 2015). This may be contested by the output achievements of countries that hardly meet democratic criteria such as Singapore or China, but also there citizen reactions to policy failures appear to have some influence on governments, but without threatening the domination of the ruling party (Peters 2013: 366).

Vivien Schmidt adds to this the concept of throughput legitimacy. Different from the performance-oriented legitimacy of output and the participation-oriented legitimacy of input, throughput legitimacy is process-oriented, and based on the interactions – institutional and constructive – of policymaking actors:

Throughput, in short, encompasses the myriad ways in which the policy-making processes work both institutionally and constructively to ensure the . . . accountability of those engaged in making the decisions, the transparency of the information and the inclusiveness and openness to ‘civil society’. As such, it constitutes a third and distinct criterion in the normative theoretical analysis of democratic legitimacy, alongside output and input.

(Schmidt 2013: 7)

Throughput legitimacy, in other words, demands governance processes that are characterized among others by transparency, inclusiveness, openness and fairness. More generally, one could argue that a decision, a policy or an organization possesses legitimacy if it is arrived at or if it is decided in a formal and procedurally correct and accepted fashion (see Weber’s legal-rational legitimacy and Luhmann 1983).

The distinction to input and output legitimacy may not always be quite obvious, but in throughput legitimacy it is not (only) the democratic legitimacy of organizations and decisions that counts, but whether rules and expectations of legality, impartiality, financial and legal accuracy, expertise and due process have been followed. This is again closely related to the ‘quality of governance’ agenda, where bad throughput – consisting of oppressive, incompetent, corrupt or biased governance practices – is seen as the main reason of legitimacy deficits. Bo Rothstein has argued for some time that the ability to control corruption and respect for rule-of-law principles and impartiality are more important in explaining the perception of government’s legitimacy than are democratic rights and welfare gains, and that meritocratic bureaucracy and rule of law thus have made a strong comeback (Rothstein and Teorell 2008; Rothstein 2012; Pierre and Rothstein 2013). He even argues that social trust may run from trust in the quality of government institutions responsible for the implementation of public policies to trust in ‘most people’. Good public administration is thus a causal factor for high levels of public and generalized trust (Scandinavia as an example).

Procedural rationality, rule of law, fairness, impartiality, reliability and expertise are, as Johan Olsen has argued, the core ingredients of the classical concept of a Weberian bureaucracy (Olsen 2006, 2008). Bureaucrats are servants and guardians of legal and professional rules and a constitutional order, balancing three possibly competing principles, i.e. hierarchy (democracy and politicians as gatekeepers), rules (law and courts as gatekeepers) and expertise (knowledge and professions as gatekeepers). Rationality, and thus (legal-rational) legitimacy, is in this understanding an attribute of organizational structures and the procedures to reach decisions and outcomes, and not of the outcome itself. Schmidt’s observation of the rising importance of ‘throughput legitimacy’ echoes Olsen’s reflections on the ‘ups and downs’, and particularly the strong comeback and ‘rediscovery’ of bureaucracy. Bureaucracies are, as Olsen stresses, shaper and amplifier of cultural values, identities and standards of behavior, and it is exactly these values and standards that are at the core of throughput legitimacy.

Vivien Schmidt concedes that the different forms of legitimacy are sometimes difficult to disentangle, throughput has sometimes been discussed in output terms, where particular institutional or discursive processes (that is legality or fairness) are seen as preconditions for (or elements of) better output performance, and occasionally in input terms, where (again these) certain institutional processes or deliberative interactions are preconditions for (or again elements of) better input participation (Schmidt 2013: 14), but they are still helpful to understand why there are different forms of accountability, how they relate to different forms of legitimacy and how they interact.

Following from these three forms of legitimacy, we can distinguish three forms of accountability: traditional political and democratic accountability, which mainly follows chains of principal-agent relationship between citizens, politicians and bureaucrats and is concerned with the democratic chain of command; outcome- and output-oriented accountability, which is mainly interested in performance, effective policies and ‘value for money’ by managers and for citizens, clients and ‘customers’; and throughput-oriented accountability, where formal procedures, due process, fairness, expertise and similar values are at the center of attention again.

Exactly these distinctions have been at the core of recent debates about changing accountability regimes. Empirical studies indicate that traditional democratic input legitimacy is losing ground, while non-majoritarian organizations gain legitimacy. An indicator is the rising trust in central banks, courts, ombudsmen and independent regulatory agencies (the ‘age of the unelected’) (Olsen 2015), while traditional democratic actors, such as parliaments, governments and especially political parties and politicians, are losing acceptance. In nearly all Western democracies, we can observe declining numbers of voters and fewer party members and declining levels of trust towards government. So the overall hypothesis is that we experience declining levels of legitimacy through traditional representative democracy, hierarchy and principal-agent accountability (Peters 2013).

The rising importance of managerial forms of accountability through NPM-oriented reforms towards performance measurement and management are part of this development. They are again closely linked to a rising belief in output legitimacy. The emphasis on the central role of managers, on performance management, and on providing services to ‘customers’ all tend to depoliticize the processes of governing and to emphasize the production of services rather than the democratic mechanisms for selecting policies (Peters 2013: 368). We have thus a rising suspicion towards traditional forms of democracy and, supported by the diminishing support of old-fashioned social democracy, some evidence that the traditional welfare state is losing support. NPM reforms strengthen and reinforce this belief in output legitimacy and accountability.

But the most striking development in recent years has been the proliferation of other and new forms of accountability, sometimes labeled horizontal or diagonal accountability through audit institutions, administrative and specialized courts, but also professional associations and finally citizen and client organizations, user groups and the general public (Schillemans 2008). Using the concepts of accountability theory, the overall development can be characterized as declining relevance of traditional political and administrative fora, while in most countries legal, financial, professional and social fora become more prominent. At the same time, one can argue that legality, due process, fairness and expertise gain in importance as elements of accountability (i.e. the well-known characteristics of a Weberian bureaucracy). These values never disappeared, but were traditionally supervised and enforced through the processes of political accountability. Recently, all kinds of ‘diagonal’ and ‘horizontal’ forms of accountability have been added. All in all, there can be no doubt that it is throughput legitimacy and accountability that have been gaining ground.

It is important to realize that the nature of the internal or external forum, be it political, administrative, legal, professional or social, should not be confused with the nature of the accountability criteria (i.e. politicians, bureaucrats, judges and professionals are interested in input, output and throughput criteria) (Jann and Lægreid 2015). Our empirical studies show that elements of output accountability (concerning performance and results) have become more important in deliberations, but also elements of input accountability (traditional, hierarchical and democratic chains of legitimacy) are still going strong. They have certainly not disappeared. But in all areas, there is also clear evidence that elements of throughput accountability have become evermore imperative.

Accountability does not only mean that actors are judged on their responsiveness to participatory input demands, or that they can be held responsible for the output and outcome of their decisions. Modern accountability is strongly concerned with the quality of these decisions in terms of procedure and due process, and it is in this last dimension where most of the accountability revolution of the last years can be observed, at least in the rising importance of new forums, actors and concerns. It is obvious that the performance-based legitimacy of the ‘output’ variety is insufficient for legitimization of political systems, that outputs and outcomes also require ‘input legitimacy’ through democratic participation and control, but both also require procedural throughput legitimacy (i.e. substantive values and principles guiding performance and process that make the performance valued and trusted).

Whether and how accountability does have a strong positive impact on legitimacy is hard to establish. Vivien Schmidt argues that, unlike input politics and output policies, where more of either is at least likely to increase the public’s sense of democratic legitimacy, with throughput processes, more of it may have little effect on public perceptions of legitimacy, while less of it in the wake of corruption, incompetence and exclusion may bring down the whole house of cards. Violating throughput legitimacy through insufficient accountability, transparency, inclusiveness or openness can have a major negative impact on public perceptions of legitimacy.

Performance and legitimacy

But what else influences legitimacy? As we have seen, the concept of performance can be slippery and ambiguous. In its recent NPM understanding, it is only concerned with the problem-solving capacity of policies and decisions, with effectiveness, efficiency, citizen satisfaction and value for money (i.e. output and outcome accountability). But the performance of political systems and institutions can also be linked to the democratic quality of decision-making and to elements of legality, fairness, professionalism and expertise. It is thus necessary to try to better understand and conceptualize the relationship between all these forms of performance on legitimacy. In order to do this, it is again helpful to return to institutional theory.

Intuition and reasonable expectations suggest the better the performance of the welfare state and its organizations, the higher its legitimacy. But neo-institutional theory at least suggests other possibilities, especially using the concept of decoupling (Meyer and Rowan 1991). Maybe the relationship is not as close as suspected, and maybe even actors are actively engaged in loosening or obfuscating this relationship. And finally, the relationship could also work the other way: the more legitimate political-administrative systems are, the better their performance.

Suchman suggests a helpful distinction between pragmatic, normative and cognitive legitimacy that resonates with Scott’s three pillars of institutional theory (Scott 2013). Pragmatic legitimacy is closely related to the logic of consequentiality (i.e. organizations are judged as legitimate if they perform as promised). Neo-institutional theory suggests that the stronger the technical environment, the greater the need for pragmatic legitimacy (Scott and Meyer 1991). Whenever outputs are somewhat easy to measure or at least to observe (Wilson 1989), organizations’ legitimacy will depend on these outputs. But Suchman also suggests that:

in a world of ambiguous causality, the surest indicator of ongoing commitment to constituent well-being is the organization’s willingness to relinquish some measure of authority to the affected audience (to be co-opted, so to speak). Displaying such responsiveness is often more important (and easier) than producing immediate results.

(Suchman 1995, with reference to Selznick 1949; Meyer and Rowan 1991)

This is exactly what we observe concerning the growth of different forms of ‘social accountability’, engaging clients, customers, user groups and the environment in accountability arrangements at the expense of traditional political accountability, and also as a substitute for difficult or contested managerial performance measurement and management.

Neo-institutional theory is even more helpful in arrangements where the institutional environment is stronger than technical considerations, and here the assumption is the stronger the institutional environment, the greater the need for normative and cognitive legitimacy. This again means that managers are more interested in manipulating symbols and rituals, because that is much easier than to influence performance or ‘tangible real outcomes’. And again, this is what we observe in different forms of opportunistic management behavior (gaming and so on), and which is the basis of the new interest in ‘reputation management’ (Carpenter and Krause 2012).

At the same time, measures of performance may become morally proscribed (consequential legitimacy as rational myths). Public organizations create legitimacy by embracing socially accepted techniques and procedures, most significantly in the absence of clear output and outcome measures and causalities. This is again what we observe in the ‘audit revolution’ (Powell 1999), and which may be behind the growth of procedural or throughput legitimacy. It is very hard to measure performance and to establish clear causalities, but at least we are following all the procedural rules of performance management. Another similar way of conceptualizing these developments is ‘mimetic isomorphism’ (DiMaggio and Powell 1983).

Here, not only the procedures, but the structure, of public organizations become important again, most significantly the structure and characteristics of classical bureaucracies.


Institutionally prescribed structures convey the message that an organization is acting on ‘collectively valued purposes in a proper and adequate manner’ . . . structures, like procedures, serve as easily monitored proxies for less visible targets of evaluation, such as strategies, goals, and outcomes.

(Suchman 1995: 581; see also Meyer and Rowan 1991: 50)

Throughput legitimacy is thus not only procedural, but also structural, and may therefore be a good substitute for simple input and output legitimacy.

Finally, cognitive legitimacy can also play an important role. Even though ‘for things to be otherwise being literally unthinkable’ is the most slippery of all concepts of legitimacy, taken-for-grantedness represents both the most subtle and the most powerful source of legitimacy. If alternatives become unthinkable, challenges become impossible, and the legitimated entity becomes unassailable by construction. So Suchman advises that at the cognitive level, ‘accounts should be simple or even banal, not only explaining organizational behavior, but also making it seem natural and inevitable’ (Suchman 1995: 596).

This brief inspection of neo-institutional theory’s contributions towards the understanding of the sources of legitimacy can obviously only be very cursory. But it suggests that the importance of output and outcome performance for the institutional legitimacy of public sector organizations, and the public sector in general is probably more limited than usually supposed. Input and especially throughput performance may be more important, or at least easier to modify, influence and manipulate by managers and politicians, and also easier to monitor and evaluate by outside observers or ‘fora’.

The relationship between legitimacy and performance may even be reversed; the more legitimate institutions are, the better they can perform their tasks. Bo Rothstein argues in this direction, but also here the evidence is not easily summarized. The obvious observation that in nearly all rankings of ‘good governance’ those countries that are occupying the top ranks are also scoring highest on measures of democracy and welfare may hint into this direction (Pierre and Rothstein 2013; see also Holmberg and Rothstein 2015 for a discussion of a large number of indicators and their blurred internal relationships).

In the same direction point some results from the ‘sustainable governance indicators’ (SGIs) of the Bertelsmann Foundation. The SGI tries to measure, with the help of both OECD data and expert assessments, the quality of the policy performance, democracy and governance capacities in all OECD and EU countries ( Not very surprisingly, in all three categories, the Scandinavian countries, as well as Switzerland, New Zealand and Germany, reach top ranks. When we had a closer look at the relationship between the different clusters of indicators (Jann and Seyfried 2009), we found that policy performance showed the strongest relationship not with ‘executive capacity’ (measured with indicators such as strategic capacity, inter-ministerial coordination, evidence-based instruments and so on), but with ‘executive accountability’ (measured with data about citizens’ participatory competence, the media, and the professional and advisory capacities of intermediary organizations). Also, the status of democracy showed a strong positive relationship with policy performance, but when looking at the indicators in more detail we found that it was particularly the ‘rule of law’ indicator that explained most of the correlation. All these data are somewhat questionable, as are the significance of some of the correlations, which again are obviously open to different interpretations concerning the directions of causality, but all in all these data hint clearly towards the interpretation that policy performance is strongly dependent on the legitimacy of institutions, and here again mostly on what we have called input and throughput legitimacy, and not the other way round. Strong policy performance is probably much more the consequence of, rather than the reason for, strong legitimacy.

Discussion and conclusion

This chapter started from the reflection that in the widespread discourse about accountability, the aspect of ‘why accountability, and for what?’ has been somewhat neglected. We know a lot about different fora, forms and foci of changing accountability relationships (Jann and Lægreid 2015), but it is somewhat unclear why the concept has reached this kind of popularity. The answer, this chapter suggests, is the ‘missing link’ of legitimacy. Or, to rephrase the question in the subtitle, if accountability is seen as the answer to many problems of the contemporary state, the underlying question is about legitimacy. The rising concern for accountability is both a reaction to direct challenges of traditional input and output legitimacy, and to the underlying growth of throughput legitimacy.

Performance may affect output legitimacy through more and better government performance, but this link is very ambiguous, as is the link between accountability and performance. But, on the other hand, more traditional performance criteria, such as due process, fairness, impartiality and expertise, may strongly influence throughput legitimacy. But causality in these relationships is not easily established; a stronger demand for throughput legitimacy may enhance new forms of accountability, and vice versa. So at least for the time being, we are more observing mutually reinforcing processes and coevolution rather than clear causal processes.

In the literature, changing accountability mechanisms and their consequences have usually been related to the advent of NPM-style public sector reforms and their fundamental concern with performance measurement and management, as well as with the obvious difficulties of these reforms and the ongoing move into post-NPM kinds of reforms (Christensen and Lægreid 2013). This chapter argues that the underlying assumptions of principal-agent relationships (and their expectations of opportunistic behavior, economic incentives, close monitoring and so on) play an important role in the popularity of the concept of accountability, but that they are not sufficient to understand the rise and the changing empirical characteristics of the concept. While normative and theoretical arguments, based mainly on principal-agent assumptions, point towards a close relationship or even nearly Siamese-twin concepts (i.e. the more accountability, the better performance), empirical studies observe tensions, ambiguities, contradictions and unintended consequences. There is no straightforward coupling between accountability and performance. Instead, trust and collaboration may be weakened or destroyed through intensive monitoring, and instead of cooperation, compromise and integration, suspicion, confrontation and disintegration of public organization may develop (Olsen 2014: 112). So the appearance of new forms, fora and foci of accountability cannot be explained by their close relationship to performance.

The association between accountability and legitimacy is much less well researched. This has to do with the broad and encompassing concept of legitimacy, which does not lend itself to plain forms of operationalization and management. When distinguishing three main forms of legitimacy, input, output and throughput, it becomes obvious that accountability is concerned with all three of them, though with the help of different processes, organizations and actors. Input legitimacy is at the core of traditional political and hierarchical accountability. The main concern is that democratically elected actors representing the interests of citizens and voters have a direct influence on the actions of public organizations and public managers, and through that on the output and outcome of administrative activities and policies. Output legitimacy, on the other hand, is mainly concerned with the effectiveness, efficiency and ‘value for money’ of government policies and activities. It is not important whether they are democratically controlled, but whether they solve collective problems and meet the expectations of the governed. This is also the main concern of modern forms of output accountability, through advanced instruments of performance measurement and management.

But the most interesting development is the growing importance of throughput legitimacy, which is mainly concerned with the legitimacy of procedures and structures of public organizations, and the values they confer. It is not mainly the participative quality of public organizations and policies nor their performance that are at the center of attention, but the quality of how these policies have been formulated and especially implemented (i.e. values of legality, due process, fairness, expertise, professionalism and so on). This again relates to the growing prominence of what has been called horizontal or diagonal accountability (i.e. accountability fora and concerns that are not part of traditional hierarchical, democratic accountability, but instead rely on courts, citizen participation, watchdog agencies, independent audit institutions, professions and so on). Also, the recent ‘return’ of bureaucracy and the values that bureaucratic institutions strengthen and confer belong to these developments.

New forms of accountability, both output- and throughput-oriented, do not replace or supersede established ones, such as democratic accountability, but all of them coexist and sometimes reinforce each other. We observe processes of layering but not of replacement. How strong the impact of new accountability regimes on different forms of legitimacy is, is for the time being hard to establish. We lack sufficient empirical and theoretical studies. But it is highly plausible that the rise of new forms of accountability is strongly influenced by new or renewed legitimacy demands.

Finally, the relationship between performance and legitimacy has to be better understood. Here, we observe that forms of pragmatic legitimacy, stressing direct utility for participants (i.e. outputs and outcomes), are important and are becoming more so, especially in highly technical environments. But at the same time, we know from studies based on institutional theory that processes of decoupling may occur, and normative or cognitive forms of legitimacy may be much more important than pragmatic, consequential ones, especially in highly institutionalized settings. Legitimacy is conferred by appropriate, proper and desirable procedures and structures, or is even ‘taken for granted’. Citizens or clients are involved in accountability relationships, are even co-opted, conferring normative legitimacy, and proper procedures and structures are stressed. Again, this is what we have observed in changing accountability arrangements. In the end, the relationship between legitimacy and performance may even be reversed. Strong output and policy performance may be much more the consequence of than the reason for strong legitimacy.

All these observations should help us to understand the recent developments and puzzles of accountability arrangements, and should point to further, better-informed research. But we do not only need more empirical research; we also need more informed theorizing linking different forms, fora and processes of accountability with different forms of legitimacy and performance. It is rather obvious that new forms of governance (i.e. what has been labeled New Public Governance), stressing more and new kinds of actors, a more fragmented public sector, more public-private partnerships, more networks and negotiations, need new forms of legitimacy. This is what we have been observing in the last few years, with the growth first of output, and more recently with the renaissance of throughput legitimacy and the corresponding forms of accountability.

At the same time, we can also observe a comeback of bureaucracy and bureaucratic values of throughput legitimacy, accountability and ‘good governance’. But this will certainly not be enough to secure the overall legitimacy of modern political orders, which are ‘multi-levelled, multi-centered, hybrid, networked and fluid’ (Olsen 2015). So the relationship between changing public organizations and processes, public policies, reforms and their legitimacy, the old concerns of Max Weber, should be more on the forefront of our research and theorizing than they have been in recent years. The old, fundamental questions of our discipline are never solved, but have to be tackled again and again, as the famous Gruk tells us, ‘problems worthy of attack, show their worth by hitting back’.


Bovens, M. (2007) Analyzing 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, 23(5): 946–967.
Busuioc, M. and Lodge, M. (2015) The reputational basis of public accountability. In CARR Discussion Paper 78 (May 2015). London: CARR, pp. 1–23.
Carpenter, D. P. and Krause, G. A. (2012) Reputation and public administration. Public Administration Review, 72(1): 26–32.
Christensen, T. and Lægreid, P. (2001) New Public Management – undermining political control? In T. Christensen and P. Lægreid (eds), New Public Management: The Transformation of Ideas and Practice. Aldershot: Ashgate, pp. 93–119.
Christensen, T. and Lægreid, P. (2005) Trust in government: the relative importance of service satisfaction, political factors and demography. Public Performance and Management Review, 28(4): 487–511.
Christensen, T. and Lægreid, P. (2013) Beyond NPM? Some developmental features. In T. Christensen and P. Lægreid (eds), The Ashgate Research Companion to New Public Management. Farnham: Ashgate, pp. 391–404.
Christensen, T. and Lægreid, P. (2015) Performance and accountability: a theoretical discussion and an empirical assessment. Public Organization Review, 15(2): 207–325.
Considine, M. and Afzal, K. A. (2011) Legitimacy. In M. Bevier (ed.), The Sage Handbook of Governance. Los Angeles, CA: Sage, pp. 369–385.
DiMaggio, P. J. and Powell, W. W. (1983) The iron cage revisited: institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2): 147–260.
Flinders, M. (2011) Daring to be a Daniel: the pathology of politicized accountability in monitory democracy. Administration and Society, 43(5): 595–619.
Greiling, D. (2014) Accountability and trust. In M. Bovens , R. E. Goodin and T. Schillemans (eds), The Oxford Handbook of Public Accountability. Oxford: Oxford University Press, pp. 617–631.
Holmberg, S. and Rothstein, B. (2015) Good societies need good leaders on a leash. In C. Dahlström and L. Wängnerud (eds), Elites, Institutions and the Quality of Government. Basingstoke: Palgrave Macmillan, pp. 13–32.
Jann, W. (1997) Public management reform in Germany: a revolution without a theory? In W. Kickert (ed.), Public Management and Administrative Reform in Western Europe. London: Edward Elgar, pp. 441–560.
Jann, W. and Jantz, B. (2008) A better performance of performance management? In KPMG International (ed.), Holy Grail or Achievable Quest? International Perspectives on Public Sector Management. KPMG International, pp. 11–25.
Jann, W. and Seyfried, M. (2009) Does executive governance matter? Executives and policy performance. In Bertelsmann Foundation (ed.), Sustainable Governance Indicators 2009: Policy Performance and Executive Capacity in the OECD. Gütersloh: Verlag Bertelsmann Stiftung, pp. 145–186.
Jann, W. and Lægreid, P. (2015) The reform of welfare states: management, accountability, and per formance. International Journal of Public Administration, 38(13–14): 941–946.
Jantz, B. , Christensen, T. and Lægreid, P. (2015) Performance management and accountability: the welfare administration reform in Norway and Germany. International Journal of Public Administration, 38(13–14): 947–959.
Koppell, J. G. S. (2005) Pathologies of accountability: ICANN and the challenge of ‘multiple accountability disorder’. Public Administration Review, 65(1): 94–108.
Luhmann, N. (1983) Legitimation durch Verfahren. Frankfurt am Main: Suhrkamp.
Lægreid, P. (2014) Accountability and public management. In M. Bovens , R. E. Goodin and T. Schillemans (eds), The Oxford Handbook of Public Accountability. Oxford: Oxford University Press, pp. 324–338.
March, J. G. , and Olsen, J. P. (1995) Democratic Governance. New York: Free Press.
Mayntz, R. (2011) Legality, legitimacy and compliance. In N. C. Bandelow and S. Hegelich (eds), Pluralismus-Strategien-Entscheidungen. Wiesbaden: VS Verlag für Sozialwissenschaften, pp. 138–149.
Meyer, J. W. and Rowan, B. (1991) Institutionalized organizations: formal structure as myth and ceremony. In W. W. Powell and P. J. DiMaggio (eds), The New Institutionalism in Organizational Analysis. Chicago, IL: University of Chicago Press, pp. 41–62.
Moes, J. (2009) Accountability – eine Antwort auf Legitimationsprobleme in neuen Governancestrukturen? In S. Botzem et al. (eds), Governance als Prozess. Baden-Baden: Nomos, pp. 29–54.
Moore, M. (2014) Accountability, legitimacy, and the court of public opinion. In M. Bovens , R. E. Goodin and T. Schillemans (eds), The Oxford Handbook of Public Accountability. Oxford: Oxford University Press, pp. 632–646.
Newton, K. (2001) Trust, social capital, civil society, and democracy. International Political Science Review, 22(2): 201–214.
Olsen, J. P. (2006) Maybe it’s time to rediscover bureaucracy? Journal of Public Administration Research and Theory, 16(1): 1–24.
Olsen, J. P. (2008) The ups and downs of bureaucratic organizations. Annual Review of Political Science, 11 (June): 13–37.
Olsen, J. P. (2013) The institutional basis for democratic accountability. West European Politics, 36(3): 447–573.
Olsen, J. P. (2014) Accountability and ambiguity. In M. Bovens , R. E. Goodin and T. Schillemans (eds), The Oxford Handbook of Public Accountability. Oxford: Oxford University Press, pp. 106–123.
Olsen, J. P. (2015) Democratic order, autonomy, and accountability. Governance: An International Journal of Policy, Administration, and Institutions, 28(4): 425–540.
Peters, B. G. (2013) Responses to NPM: from input democracy to output democracy. In T. Christensen and P. Lægreid (eds), The Ashgate Research Companion to New Public Management. Farnham: Ashgate, pp. 361–374.
Pierre, J. and Rothstein, B. (2013) Reinventing Weber: the role of institutions in creating social trust. In T. Christensen and P. Lægreid (eds), The Ashgate Research Companion to New Public Management. Farnham: Ashgate, pp. 405–416.
Pollitt, C. (2011) Performance blight and the tyranny of light? Accountability in advanced performance measurement regimes. In M. J. Dubnick and H. G. Frederickson (eds), Accountable Governance: Problems and Promises. New York: M. E. Sharpe, pp. 81–97.
Pollitt, C. and Hupe, P. (2011) Talking about government: the role of magic concepts. Public Management Review, 13(5): 641–658.
Powell, M. (1999) The Audit Society: Rituals of Verification. Oxford: Oxford University Press.
Romzek, B. S. (2015) Living accountability: hot rhetoric, cool theory, and uneven practice. PS: Political Science & Politics, 48(1): 27–34.
Rothstein, B. (2012) Political legitimacy for public administration. In B. G. Peters and J. Pierre (eds), The Sage Handbook of Public Administration. Los Angeles, CA: Sage, pp. 407–419.
Rothstein, B. and Teorell, J. (2008) What is quality of government? A theory of impartial political institutions. Governance, 21(2): 165–290.
Scharpf, F. W. (1970) Demokratietheorie zwischen Utopie und Anpassung. Konstanz: Universitätsverlag.
Scharpf, F. W. (1999) Governing in Europe: Effective and Democratic? Oxford: Oxford University Press.
Schillemans, T. (2008) Accountability in the shadow of hierarchy: the horizontal accountability of agencies. Public Organization Review, 8(2): 179–295.
Schmidt, V. A. (2013) Democracy and legitimacy in the European Union revisited: input, output and ‘throughput’. Political Studies, 61(1): 2–22.
Scott, W. R. (2013) Institutions and Organizations: Ideas, Interests, and Identities. Thousand Oakes, CA: Sage.
Scott, W. R. and Meyer, J. W. (1991) The organization of societal sectors: propositions and early evidence. In W. W. Powell and P. J. DiMaggio (eds), The New Institutionalism in Organizational Analysis. Chicago, IL: University of Chicago Press, pp. 108–140.
Selznick, P. (1949) TVA and the Grass Roots. Berkeley, CA: University of California Press.
Suchman, M. C. (1995) Managing legitimacy: strategic and institutional approaches. The Academy of Management Review, 20(3): 571–610.
Talbot, C. (2005) Performance management. In C. Pollitt , E. Ferlie and L. E. Lynn (eds), The Oxford Handbook of Public Management. Oxford: Oxford University Press, pp. 491–517.
Van de Walle, S. (2013) NPM: restoring the public trust through creating distrust? In T. Christensen and P. Lægreid (eds), The Ashgate Research Companion to New Public Management. Farnham: Ashgate, pp. 309–420.
van Helden, G. J. and Reichard, C. (2013) A meta-review of public sector performance management research. Tékhne, 11(1): 10–20.
Wilson, J. Q. (1989) Bureaucracy: What Government Agencies Do and Why They Do It. New York: Basic Books.
Search for more...
Back to top

Use of cookies on this website

We are using cookies to provide statistics that help us give you the best experience of our site. You can find out more in our Privacy Policy. By continuing to use the site you are agreeing to our use of cookies.