The Limits to Public Value, or Rescuing Responsible Government from the Platonic Guardians

Abstract

In various guises, public value has become extraordinarily popular in recent years. We challenge the relevance and usefulness of the approach in Westminster systems with their dominant hierarchies of control, strong roles for ministers, and tight authorising regimes underpinned by disciplined two-party systems. We start by spelling out the core assumptions behind the public value approach. We identify two key confusions; about public value as theory, and in defining ‘public managers’. We identify five linked core assumptions in public value: the benign view of large-scale organisations; the primacy of management; the relevance of private sector experience; the downgrading of party politics; and public servants as Platonic guardians. We then focus on the last two assumptions because they are the least applicable in Westminster systems. We defend the ‘primacy of party politics’ and we criticise the notion that public managers should play the role of Platonic guardians deciding the public interest. The final section of the article presents a ‘ladder of public value’ by which to gauge the utility of the approach for public managers in Westminster systems.

In his seminal book Creating Public Value, Mark Moore writes that ‘public managers are seen as explorers who, with others, seek to discover, define and produce public value’ (Moore 1995:20–21). They should not focus on ‘simply devising the means for achieving mandated purposes’. Rather, they should ‘become important agents in helping to discover and define what would be valuable to do’. If traditional views of public administration ‘discouraged bureaucrats from exercising much imagination about the proper purposes of government and prevented them from taking any responsibility for defining them’, then we ought to construct ‘an alternative idea about how public managers should think and act’. What Moore attempts in his book is ‘to work out a conception of how public managers… could become more helpful to society in searching out and exploiting opportunities to create public value’. Elsewhere he has expressed his vision in more instrumental terms imploring public managers to ask ‘how best to develop a constituency that values what the manager wants or conceives’ (Moore 2006).

We find this approach to public management disturbing when applied to Westminster systems of government such as Australia, Canada and Britain. It misdiagnoses the function of management in the modern public sector and invents roles for public servants for which they are not appointed, are ill-suited, inadequately prepared and, more importantly, are not protected if things go wrong. It asks public managers to supplant politicians, to become directly engaged in the political process, and become the new Platonic guardians and arbiters of the public interest. In this article, we criticise the ‘public value’ approach, focusing on Moore's work as representative of the genre.

In various guises, public value has become extraordinarily popular in recent years. Many educational programs for public managers spread the word – from the Kennedy School of Government at Harvard University, to the Warwick Business School in the UK and the Melbourne Business School in Australia. It is also central to the educational program taught by the Australia and New Zealand School of Government to public managers across six jurisdictions. A previous issue of AJPA (Volume 63(4), 2004) was devoted to a symposium on public value in which several authors recommended adopting the public value approach in Australian government. Public value is up and running.

Moore's original message was an alternative normative vision for American public managers, operating in an environment relatively hostile to state intervention (see Moore 1994). We concede the public value approach might work in American government where much scope exists for policy entrepreneurs and policy brokerage, and policy actors have to build coalitions of support to make headway. However, it is far less relevant in parliamentary systems with dominant hierarchies of control, stronger roles for ministers, and tight authorising regimes underpinned by disciplined two-party systems. Surprisingly, Moore's message resonated far more forcefully in Westminster systems such as the UK, Australia and New Zealand than in the US. Why? The key is the shared tradition of constitutional bureaucracy. Public value sustains constitutional bureaucracy and helps it to rediscover itself after the predations of New Public Management (NPM) and the preference for the ‘contract state’ to deliver services (Davis and Rhodes 2000).

Scholars, public servants and policy advisers of many hues in Westminster systems have been quick to endorse its logic. For Stoker (2006:42, 46) public value is an ‘emerging management paradigm’ that is preferable because it goes beyond the market-oriented NPM and ‘offers a broad framework in which to comprehend the management challenge posed by networked governance’. ‘Public value management’, as he calls it, is suitable to modern post-bureaucratic polities in which the ‘emphasis is on challenge and change’, on democratic ‘openness’ and on ‘shared responsibility’ governance. Benington (2005:2) regards public value as offering ‘a clearer conceptual framework and overall strategic purpose for public service improvement and reform’ than earlier Keynesian or neo-liberal models. For Kelly, Mulgan and Muers (2002:7–8) public value is a broader ‘perspective’ that incorporates collective preferences and ‘shared aspirations’, while providing a more inclusive ‘yardstick for measuring performance’. Smith, Anderson and Teicher (2004:15) argue public value is ‘a more orderly framework for future debate about the public sector’. Smith (2004:79) regards Moore's message as universal that ‘could apply in Westminster as well as in Washington’.

For all theses authors, ‘public value’ encouraged a more positive view of both government and public service provision and innovation in countries dominated by neo-liberal ideological agendas and the rhetoric of small government. It also focused attention on issues such as performance, service quality, real outcomes, and trust. It stressed the importance of implementation in improving public policies. Public servants seized on the approach as a new belief system that provided a rationale for their existence. It was a justification for activism that helped them to rediscover their self-worth as public managers. Government was back in fashion; intervention and innovation were legitimate once again.

We have no qualms with these immediate benefits. Indeed, we welcome the renewed attention to government and public administration. But we believe the public value approach is flawed when applied to Westminster systems. So, we present a critique of the assumptions behind ‘public value’, not only to add to the scholarly debate, but also to provide public managers with a different set of spectacles through which they can view and judge the approach. We start by spelling out the assumptions and arguments behind the approach. We then focus on two assumptions. We argue against downgrading the ‘primacy of party politics’. We criticise the notion that public managers should play the role of Platonic guardians deciding the public interest. In the final section of the article, we present a ‘ladder of public value’ by which to gauge the usefulness of the approach for public managers in Westminster systems.

The Key Confusions and Core Assumptions

We identify two key confusions: about public value as theory; and in defining public managers. We identify five linked core assumptions in public value: the benign view of large-scale organisations; the primacy of management; the relevance of private sector experience; downgrading the primacy of party politics; and public servants as Platonic guardians.

Public Value as Theory

To evaluate any approach we need to know what the author claims for it. Is it a new empirical theory? Is a set of techniques for managers? Supporters of public value are inconsistent in their claims. For example, Moore is unclear whether he offers a new theoretical framework, a concept, a heuristic device, or an operational tool of management. He drifts between these meanings depending on the argument or case. He claims to present a normative theory of managerial behaviour that ‘lays out a structure of practical reasoning to guide managers of public enterprises’ with the aim of telling managers what they ‘should think and do’ (Moore 1995: 1–2). Yet, he slides in seamless progression from operational improvements at the workplace to higher levels where public value is an aspiration for the governmental system. He combines empirical evidence (usually derived from illustrative cases of bottom-up innovation) with normative notions for reforming public organisations that will produce better public policy and augment the public interest in unambiguous ways (see below).

Various proponents argue public value should be seen as a paradigm (Stoker 2006; Benington 2005); as a concept (Kelly, Mulgan and Muers 2002; Horner 2005); a model (O’Flynn 2005); a heuristic device or even a story (Smith 2004). The term has been contrasted with rational choice and neo-liberal theories and with economic and individualistic theories of consumption (Alford 2002; O’Flynn 2005; Benington 2005). Benington (2005) contrasts ‘exchange value’ (private choice derived from neo-classical economics) with ‘public value’ that includes economic as well as social and political value. It has also been linked to related concepts such as ‘consensual policy-making’ (Marton and Phillips 2005) and the ‘public interest’ (Bozeman 2002). Perhaps the ambiguous nature of public value and its various applications fuels its popularity – it is all things to all people. Obviously advocates of public value hold diverse positions. Most important, it is all too often unclear whether they are developing a normative argument or an empirical-operational one. It matters. The criteria for evaluating aspirations differ from those that seek to assess evidence.

Defining Public Managers

In Moore's account the definition of the ‘public manager’ is stretched way beyond its conventional meaning in the Westminster system. He includes elected, unelected and self-appointed actors including politicians, presidents, governors and mayors, political staff, commissioners and directors, senior civil servants, administrators, experts, supervising agents, judges, lobbyists and interest groups leaders, and even private managers if they produce mainly for the government (Moore 1995:2–3). The cast is inclusive and expansive and a source of potential confusion when applied to Westminster systems. In sharp contrast to the US, officials in Westminster systems are rarely if ever elected. Indeed, there is also no analogy to presidential appointees. The political appointment of public officials is the exception not the rule in Australasia, Canada and Britain.¹ So, the elected politicians have a legitimate role in determining the public interest under the conventional norms of representative democracy, but non-elected officials do not share that role. There is a widely shared understanding that the roles of elected politician and appointed official are sharply demarcated. Each has distinctive rights, responsibilities and bases of legitimacy. So, we need a theory that encompasses that separateness.² Yet, despite his all-encompassing definition, most of Moore's real-life examples talk to, and are centred on, the unelected individual public manager, not the elected and formally accountable politicians.

Benign Organisations

The growing literature on public value assumes a naive ‘common-weal’ view of politics and governance. Illustrative cases of public value are apparently laced only with good intentions (Horner 2005). Actors such as governments, public organisations and interest groups are assumed to want to do good things. They are motivated only to assist and please citizens, to create public value and improve the public good. There is, thus, a benign view of all large-scale organisations (both public and private). Actors in these contexts are considered to be interested in the well-being of others, not exploitative, cabalistic or predominantly self-interested. Yet, public organisations do not exist solely to perform pleasant functions. Many exist to enforce state rules, to push people into work, to achieve targets at the lowest cost, or to ration public goods and services.

Power imbalances both in the bureaucratic chain of command and between state officials and clients, citizens or interest groups are assumed away or played down. Power rarely figures in the explanations or reported cases (Kelly, Mulgan and Muers 2002). Often there is an assumption public value produces ‘win-win’ situations, because, if all the actors are indispensable to the achievement of a ‘successful’ outcome and all have their own power resources, then they all share the incentive to make trade-offs to achieve the collective desired intention. This stance ignores the asymmetrical power structures between state officials, groups and individual clients or citizens.

Also the emphasis falls on benign aspects of managerial action. We hear of displays of public-minded commitment or of successful innovation. We are told stories about managers developing new capabilities or initiating collaboration, and about individual enthusiasm and courage. Other managerial tasks such as control, direction, constraint management, rationing, sanctioning or enforcing penalties are conspicuous mainly for their absence. In short, the ‘dark side’ of the state and its administrative arms is largely ignored in the ‘public value’ vision.

The Primacy of Management

Public value approaches take for granted the primacy of management in government organisations. Management is seen through the eyes of the individual manager who has a set of managerial tools at his or her disposal. Agency and individual choices are stressed over structures and organisations. Individual managers, not organisations, devise and make strategies to ‘gain virtue’ (according to some measure of ‘improvement’ or satisfaction) even if they rely on other people to make them work. In Stoker's words (2006:55), this pushes managers into ‘doing politics’. Moore (1995:chapters 4, 5) repeatedly makes the point that managers must exercise political management within the constraints of their ‘authorising environment’ to ‘fashion legitimacy and support for themselves, their policies, or their organisational strategies’ (Moore 1995:113). His view of the manager presupposes an institutional environment in which managers exercise a degree of autonomy and entrepreneurialism that is not typical of public servants in Westminster systems or welcomed by their political masters.

As entrepreneurs, the onus is then placed on these managers to define or find the areas where public value can be created. They are responsible for identifying potential opportunities and initiating action. The role of the public manager is to act as a ‘public explorer’, a ‘discoverer’, an ‘entrepreneurial advocate’, a ‘creator’, and a ‘steerer’ rather than the ‘rower’ or functionary in government agencies. In doing so, they are shaping the operational definition of the public interest and largely determining whether the state ought, or is motivated, to invest in such value. They are playing small ‘p’ bureaucratic politics as they ‘become strategists’. In Moore's (1995:20) words: ‘they engage the politics surrounding their organisation to help define public value as well as engineer how their organisations operate’.

Although other actors may subsequently become involved (when managerial initiatives float up to the ‘authorising regime’), the initiation of attempts to create public value expressly falls to public officials. ‘Society needs value-seeking imaginations (and associated technical skills) from its public sector executives no less than from its private sector managers’ (Moore 1995:21). Public managers become internal lobbyists for particular ‘valuable’ strategies, and by providing ‘more formal channels through which managerial ideas about opportunities to create public value could be properly expressed’ they become less ‘vulnerable to self-interest’. Managers can be ‘taught’ about ‘how to search for and define public value more properly and effectively than they do now’, to ‘allow society to have the benefit of the experience and imagination of public sector managers without having to yield to their particular conceptions of the public interest’ (Moore 1995:21). Yet, these managers are still the principal formulators of notions of public value.

Moore is careful to avoid his managers confronting large ‘P’ politics, moral dilemmas, or major value conflicts in his illustrative cases. He chooses real-life examples that highlight the ordinary and mundane areas of public policy and policy implementation, such as public libraries, refuse collection, correctional services and environmental management. Who can object to latchkey kids being cared for by a sympathetic librarian or a sanitation commissioner devising new ways to help citizens keep cities clean? In such examples, the notion of public value is less controversial because we can identify some societal consensus exists about what constitutes the ‘social benefit’. Such agreement often does not exist.

Moore's choice of policies that create public value focuses on benign service provision where an agreed ‘good’ can be produced and the relevant stakeholders approve. In the American context, this emphasis on approval is all too easy to understand given the long-standing backdrop of hostility to government intervention. Citizens believe public officials are self-interested. So it is important to counter this view by identifying interventions that command agreement. We do not find adherents of the public value approach making cases for widening the net of tax collection, hospital managers selling human organs, better wheel clamping of parking offenders, quicker and more available abortions whenever the woman desires, incentive payments to police to catch more speeding motorists, or bounty-hunting for escaped criminals. The assumption is that a consensus can be identified or engineered by managers. It is not clear how entrepreneurial officials are to exercise their own moral compass in creating public value where there is conflict. It is not clear what role party politics will play in managing that conflict and identifying policies. It is not clear how public officials work with the political parties (if at all) or how they will contribute to their policy processes.

Learning from the Private Sector

Curiously, Moore's response to conflict, and we suspect it is a peculiarly American response, is not to turn to an examination of the political process but to private sector experience. He advises public officials seeking to create public value to copy the private sector and behave like corporate executives. He argues the public sector should adopt successful private sector approaches to corporate strategy to identify and create public value. His argument presupposes, on the one hand, that there are few significant differences between the public and private sectors, or that ‘constitutional’ differences can be overcome with appropriate leadership and corporate strategic planning. On the other hand, his argument presupposes the virtues and desirability of the rational actor model of decision-making and corporate leadership. So, he advances the argument that:

a high-quality decision is one that has a large measure of both process and substantive virtues. A decision gains process virtues by emerging from wide consultation in which all interested parties have a chance to be consulted, and all the formal legal rules governing the process of decision-making have been met. A decision acquires substantive legitimacy to the extent that the decision has been able to draw into the policy-making process as many relevant facts as are available, arrayed in an analytic framework that accurately represents the important values at stake in the decision, the principal alternatives open to decision-makers, and the likely consequences of the alternative choices (reckoned in dimensions of significance to the decision-makers) (Moore 1995: 163–164).

We are sceptical of these arguments. First, on the public-private distinction, Moore (1995:64) argues it is ‘easy to exaggerate the significance of these differences’ between the roles of private and public executives (Moore 1995:64). Curiously, he cites Allison (1983) on the distinctiveness of the public sector but ignores the significance of these differences for public enterprise managers (see Moore 1995: 343–344, note 25; 356, note 3). There is a substantial international literature that explores the substantive differences between public and private arenas (see Rainey and Chun 2005; Ranson and Stewart 1989). Moore downplays their significance.

Second, if private sector techniques offer such obvious and available ways to manage, then why is so little implemented across government? It is not because public managers are ill-trained, stupid or venal, but because private sector techniques do not fit the context, can be neutered by both bureaucratic and political games, and are not subjected to the same accountability as public management. Public sector officials also do not share the same risks and rewards.

Third, the rational comprehensive model of policy-making is an idealised notion largely removed from the reality of public policy-making. Politics, value clashes, interests, cultures, symbolic imperatives, processes and accountability requirements all make the rational actor model untenable in public policy decision-making. These limits have been spelt out so often, they need no repetition here (see, for example, Lindblom 1990; Wildavsky 1979).

Downgrading Party Politics

Politics is portrayed as a ‘problem’ in public value accounts, almost as an illegitimate interference standing in the way of good management. It prevents managers from ‘exercising much imagination about the proper purposes of government’ or developing interests in new ideas or improving the delivery chain (Moore 1995:21). Tales abound about the practical difficulties of doing anything because ‘politics’ gets in the way or exercises a veto, or because it is hard to attract attention higher up, or because it is difficult to get politicians to focus on the micro politics of production. The solution is to transfer real political responsibility to front-line officials, and have them negotiate approval for their proposed actions.

For anyone in Australasia, Canada or Britain reading Moore on politics, it is obvious he is talking about the American system of government and its pluralist politics. There is nary a mention of either party politics or the constitutional conventions of ministerial and collective responsibility. There is no account of constitutional bureaucracy (see Parris 1969; Rhodes, Wanna and Weller, forthcoming; Sossin 2006). There may be no agreed definition of Westminster or even parliamentary government, but most versions embrace the ideas of centralisation and hierarchy. For example Strøm and his colleagues (2003:chapters 3, 23) conceive of parliamentary democracy as a chain of delegation from principals to agents; from voters to their elected representatives, from legislators to the chief executive, from the chief executive to ministerial heads of departments, and from ministers to civil servants. Gamble's (1990: 407) definition encompasses a unitary state; parliamentary sovereignty; strong cabinet government; and majority party control of the executive (that is, prime minister, cabinet and the civil service). Moore and his several followers provide no assessment of the impact of a clear political and administrative hierarchy for the practices of creating public value. To be fair to Moore, he is often the victim of his followers’ enthusiasm. We can find only one point where he discusses the applicability of his ideas in other governmental systems (Moore 1995:4; 313, note 16). His fault is to believe his supporters. That conceded, it is clear that ideas such as hierarchy, strong cabinet government; and majority party control of the executive, ministerial control of officials and merit appointment of neutral officials fit ill with the public value approach. Whatever public value may be about it is never about majority party centralisation and control.

Public Officials as Platonic Guardians

The inherent danger with ‘public value management’ is that public managers are asked to serve as the Platonic guardians and arbiters of the public interest.³ They are charged with imagining value and defending their notions of the ‘public good’ against other conceptions. So, although public value writers often stress the importance of collective preferences and participatory involvement of the community, public value approaches are premised on a fundamentally non-democratic notion. Bureaucracy and bureaucratic interests predominate and displace democratic preferences.

In public value management there is no difference between public value and the public interest. They are interchangeable. If public managers are benign, as Moore believes, then they can interpolate public preferences and enhance the public interest by initiating improvements in delivery to mutual advantage. Value adding is a one-way process – organisations can keep on doing what they do or add value through innovation. But who says it is of ‘value’, and what if innovations lead to deteriorations in delivery? Moore insists the ‘public’ or community are somehow the arbiters of ‘value’. But because he distrusts the formal electoral process (Moore 1995:52–56), managers have to turn to other surrogate measures of endorsement – such as client satisfaction or feedback, willingness to co-produce, increases in trust of the public enterprise or its officials (Moore 1995:186–187).⁴

Worryingly, this model is one-sided. There is no real conception of ‘negative public value’, where even well-intentioned innovations can lead to deteriorations in delivery chains, or create increasing expectations that agencies cannot meet. There is no conception of the ‘mobilisation of bias’ (Bachrach and Baratz 1970) and the ways in which minorities can be organised out of the political arena. The simple point is that Moore envisages a one-way benevolent conception of public value. But the public interest is endlessly contested. Value conflict lies at its heart. Consensus is elusive. To remove it from the formal political process is to attenuate accountability and to place public servants in a position of power without responsibility.

Throughout our discussion of the confusions and assumptions in the public value approach we have stressed we are talking about Westminster systems. Our last two points about the primacy of party politics and the public interest lie at the heart of our argument. We believe the inbuilt hierarchy of Westminster systems coupled with the disciplined party politics imposes severe limits on both the role of public servants and the utility of the public value approach. So we now return to these two arguments in more detail.

The Primacy of Party Politics

To draw an elementary distinction, Moore takes large ‘P’ politics or party politics and consigns it to the authorising environment. He then focuses on the small ‘p’ politics of the bureaucracy and the way in which public managers manage their environment. But in Westminster systems with disciplined party systems, and in sharp contrast to the coalitional parliamentary systems of Western Europe (Rhodes 2006) and loose pluralist party system of the US, large ‘P’ politics sets the decision structures for small ‘p’ politics.

In effect, for Moore, politics operates in its own insular realm removed from the operational efforts of managers. Politics is distanced from the operational leadership role of the manager. This view downgrades politics. It allows politics to enter his examples only as retrospective agreement or as a belated form of authorisation (tied to after-the-fact legitimacy). Although he talks often of politics and stresses the need for managers to recognise (or read) the political landscape, he is uncomfortable with the politics of liberal democracy. He is critical of the conventional theories of democratic administration on many grounds at many points in the book (Moore 1995:16–17; 31–36; 48–49). Formal politics works to its own expedient logics and does not fit easily with the managerial endeavour. It does not ‘deliver’ for the public good but ‘seems focused on winning votes by satisfying special interests rather than on finding and producing something publicly valuable’ (Moore 1995:19). Formal big ‘P’ politics is risk averse and does not encourage the quest for public value.

He is also critical of the behaviour and motivations of politicians, and assumes society distrusts both politics and politicians (see Moore 1995:31–33, 54, 179, 330, 338, 349). Politics is venal and expedient (shades of Tammany Hall), and politicians are seen as villains who wield political power temporarily but do not defend the public sector. They may be able to authorise, fund and monitor, but often they cannot prioritize or give government agencies clear mandates or directions. So, managers need to fill this political void at their operational level. Politics also tries to impose a dysfunctional hierarchic mind-set on managers ‘downward toward the reliable control of organisational operations’ (Moore 1995:17). It acts to sap bureaucracy of imagination, making it reactive and small ‘c’ conservative.⁵ Politics impedes innovation in the public sector, imposes constraints on managers – they ‘have to be reined in more tightly than private sector managers’ (Moore 1995:19) – and it limits the scope for outward entrepreneurship to achieve ‘valuable results’. In one sense, if public managers adopt a public value approach they are being asked to rebel against standard politics and usurp the democratic will of governments. Moore comes closest to this when he advises managers to orient themselves upwards to renegotiate policy mandates (that is, overturn the political wishes of elected officials, Moore 1995:17). This point was recognised by Stoker (2006:55–56) as a ‘key dilemma’ when ‘management is king, and politics is sidelined’.

Moore also asserts that markets are preferable to government interference, because individual preferences can be satisfied, and that individual liberty is preferred to collective goals. In the US with its preference for smaller, cost-effective government, he suggests officials have to make a convincing case for intervention because any resources required ‘are only grudgingly surrendered’ (Moore 1995:29). In turn, this then means that:

it is not enough to say that public managers create results that are valued; they must be able to show that the results obtained are worth the cost of private consumption and unrestrained liberty foregone in producing the desirable results (Moore 1995:29).

Managers can dream but they must turn their dreams into reality, so:

if managers have substantially valuable ideas but are unable to attract political support or administer them feasibly, then those ideas must fail as strategic conceptions. Such ideas are ‘academic’ in the worst sense of the word (Moore 1995:71–72).

Moreover, in terms of political management:

building support and legitimacy for a policy, or of enhancing the effective claim that an official may make on the society at large, is what political management is all about… managers must work to fashion legitimacy and support for themselves, their policies, or their organisational strategies (Moore 1995:112–3).

Crucially, managers require sanction from authorising actors; they

need these ‘external’[authorising] actors for one (or both) of the following reasons: they need their permission to use public resources in pursuit of a given enterprise; or they need their operational assistance to help produce the results for which they are responsible (Moore 1995:113).

The eventual authorisers are external to both policy formulation and management.

Moore does not pretend we live in an apolitical world. He recognises other powerful stakeholders and the wider political context. But his main point here is that such ‘external’ politics should be ‘managed’ by the public managers. So, he treats politics by compartmentalising it into two discrete junctures with which managers need to engage. First, he recognises that public managers cannot be unilateral actors exercising unrestrained bureaucratic leadership, so some form of oversight and public legitimising is necessary. He writes in frustration that managers ‘know full well that there is no escape from the powerful mechanisms of political oversight’ (Moore 1994:296). So, he constructs a stage which he calls ‘the authorising environment’ and confines democratic (and authoritative) politics to this narrower realm. He defines the authorising environment as comprised of ‘external actors beyond the scope of their [the managers] authority’ (Moore 1995:113–114). Entrepreneurial managers then need to gain assurances they have permission (authority and money) and operational assistance. Politics now is not all pervasive but has become encapsulated as ‘approval’ and ‘continuing consent’.

Second, he talks about small ‘p’ politics taking place in the operational networks in which managers find themselves. Managers have to persuade other significant actors in the field of the merits of their case and the value of the ends they propose. They have to negotiate the micro-politics of place and opportunity, seeking agreement, aligning interests, and building coalitions. Politics is dialogue, but initiated by the line manager. It is about line managers promoting end strategies and crafting agreements. But if Lindblom (1990) is to be believed, agreement is the key criterion of success, and agreement will be about means not ends.⁶

This characterisation of politics may fit the US but we suggest that in Westminster systems politics are inherent in all the situations and choices managers make, and in the choices others make that impact on the manager. We should be clear, and managers should recognise for their own good, that they are being asked to play a high-risk political game. It can get nasty, bite them back, leave them exposed, and perhaps destroy their careers (see, for example, the cases in Rhodes and Wanna, forthcoming). They are not protected if they assume such political roles. Public value initiatives can just as easily serve powerful bureaucratic interests as they can the citizens’ interests; can just as easily impose costs as benefits. Community acceptance of the value-added is not the only test of value, it may be temporal, contextual, and it may be fickle (this point is recognised by Moore 1994:297; 1995:72).

Moreover, by distancing the authorising environment from the policy process, Moore's description resembles not the cut and thrust of political life, but an aloof judicial bench arbitrating on ideas or granting legitimacy to managerial proposals. In Westminster systems, the roles of members of the authorising environment are not confined to ‘authorising’ the actions of public bodies and non-elected officials. It is as likely they will argue, disagree, compete and behave inconsistently. The manager may need their cooperation, not just their ‘authorisation’. Yet the concept of the ‘authorising environment’ is used to encompass not only building constituencies and inter-agency collaboration, but also all the stages of formal accountability buried in the simple notion of ‘needing permission’.

Of course in Westminster systems significant actors exist who are charged to serve in these specific roles – ministers and prime ministers. Yet Moore puts these actors to one side or consigns them simply to oversight functions. Contrary to Moore's depiction, ministers and politicians are not merely end-users of public value that has been imagined by line managers, but are themselves significant initiators of policy choices. Leadership in disciplined party systems is well placed to impose its choices. Ministers are expected and expect to play the political game and shoulder the accountability. Ministers and their senior advisers would be perplexed (not to mention offended) to read they were mere authorisers of decisions formulated and framed elsewhere. Westminster systems are hierarchical, not pluralist, ministers believe in hierarchy, and the party system allows them to impose their choices. From their perspective, they are involved in agenda-setting, initiating policy proposals, formulating and lobbying for proposals, brokering deals, fighting for funds, and authorising and announcing. They are assisted by officials who owe their commitment and loyalty to the minister. Moore's account does not describe their world-view, let alone credit them with any of these functions.⁷ Nor does his view pay enough attention to the limits that loyalty to the minister places on the scope for initiative by public managers. Loyalty is a pre-eminent virtue. Disloyalty is simply not tolerated. One person's initiative is another's disloyal act. Bureaucratic entrepreneur beware (again, for examples see Rhodes and Wanna, forthcoming).

The Public Interest

For Moore the preferences of public managers about the public interest will normally prevail. It is their responsibility to initiate and manage the dialogue with stakeholders. Thus:

Society needs leadership from these managers to help it learn what is both desirable and possible to do in public domains for which these managers are temporarily responsible… in most cases there is more discretion than most public managers (and their overseers) acknowledge. Nearly always, the politics surrounding a public enterprise are sufficiently contentious to suggest several different plausible and sustainable conceptions of public value (Moore 1995:63).

But who gave these Platonic guardians the right to choose between these conceptions of the public good? How are they held to account when it goes wrong? This conundrum is the second dilemma at the heart of the public value approach.

The seeming popularity of Moore's approach may excite renewed debate over the notion of the ‘public interest’, which has long been out of academic favour. We should remind ourselves that Schubert (1960:199–206) identified three conceptions of the public interest – rationalist, idealist and realist. Rationalists postulate a commonweal expressed in the popular will, articulated through for example a two-party system, and faithfully executed by public officials. Idealists support the ‘true’ interests of the public that reside in a higher, natural law interpreted by public officials. Realists see the public interest as the result of pluralist bargaining. Moore's view is principally idealist with an unwelcome dab of realism.

From a behavioural or modernist-empiricist position, Schubert (1960:220) is critical of all three conceptions because they are not testable. To Schubert (1960:223), public interest theories do not describe how people behave and, therefore, ‘there is no public-interest theory worthy of the name’. But, we prefer the notion that the public interest has different meanings in different narratives (see Bevir and Rhodes 2003; 2006).

The concept of the ‘public interest’ can be treated as a verbal symbol (Goodsell 1990:102–107). When used by public officials, it evokes certain values and establishes a particular normative frame of reference. Six values can be identified: legality-morality; political responsiveness; political consensus; concern for logic; concern for effects; and agenda awareness. So, if the public interest has no specific content but is just a verbal symbol, public officials must follow the law and behave honestly and with integrity. They must defer to the majority wishes of citizens. Rival groups in the community must make a case broader than naked self-interest. Those claiming to act in the public interest should be able to show that their social goals are valid, the policy will advance the goals, and the policy is reasonable and coherent. The future effects of the policy must be beneficial, a claim which implies attention to both long-term considerations and all affected persons. And, finally, the public interest discourse stresses concern with the unarticulated needs in society. So, the notion of the public interest can have impact as a verbal symbol in political discourse. It is constructed within political traditions that offer distinct and distinctive narratives.

In the market tradition, as Peters (1996:43) argues, the public interest requires governments be judged according to how cheaply public services are delivered and how they respond to market signals. Citizens are seen ‘as consumers as well as taxpayers’, even if accountability is often difficult to identify. Hence, ‘the public interest can be served by allowing citizens to exercise freer choice in a market for public services’ (Peters 1996:44), meaning the public interest is ‘low cost’. According to Peters (1996:87) such flexible government has ‘the least clearly articulated concept of the public interest’.

In bureaucratic traditions, the public interest approximates Schubert's rationalist notion. The minister articulates the common will as a member of the elected majority party, and the role of the public servant is to put the minister's wishes into practice. This approach is a constitutional version of the public interest: ‘the duty of the individual civil servant is first and foremost to the Minister of the Crown who is in charge of the department in which he or she is serving’ (Armstrong 1985). Public duty takes precedence over private interests. The individual civil servant should contribute to the greater good and be anonymous and unselfish. Chapman (1988:38) comments that civil servants had ‘a disposition to find public affairs of interest; no desire or intention to take part in political life; and a readiness to work as a member of a team, rather than seek personal glory’. In Britain, for instance, the civil service culture is considered to be a blend of values including honesty, loyalty, impartiality, propriety and a respect for intelligence with conservatism, scepticism, elitism and arrogance (Butler 1992:8; Plowden 1994:21–23, 74).

A third approach to analysing the public interest stresses dialogue among actors and groups. For example, Wamsley et al. (1990) define the public interest as both an ideal and a process, meaning it is the product of a continuing dialogue between all affected individuals and groups that addresses long-range views and competing demands. Officials are trustees of the quest for the public good. It is their responsibility to look beyond the short-term, to stimulate reasoned debate, to involve citizens in that debate, and to expand the opportunities for involvement. They are the repository of ‘specialised knowledge, historical experience, time-tested wisdom and … some degree of consensus of the public interest’ (Goodsell 1985:155). Officials act as a counterweight to short-term political expediency and opportunism. They stand for integrity and probity against the possibility of partisan interest and corruption. However, although they are guardians of the dialogue and the process, this approach could be open to the same criticism levelled against Moore, that public officials are positioned as Platonic guardians acting in the broader public interest. As Stivers (1990:265) comments:

not being able to quantify the public interest or to reach full agreement about the correct translation of values into action are not the same thing as having nothing basic to which administrative discretion must hold itself accountable. One may never have full confidence that the choice itself is right, but one may approximate such confidence in the knowledge that one's claims have been checked against the interpretation offered by citizens.

So, dialogue and joint action with citizens play a key role in constructing a shared interpretation of the public interest values.

Just as the public interest is variously constructed, so is the idea of public value. Public value is not given nor is it merely about doing good things today. We suggest that notions of the public interest and public value need to take the following into account.

First, if public value has meaning, it is a shared meaning that operates within a narrative and its associated tradition. The shared meaning develops iteratively and collaboratively; it emerges through dialogue and reconfirmation in society.

Second, private definitions of public value (or assertions from single actors that this or that will add to public value) are flawed. No one actor can impose a definition of public value, and it is impossible to define a priori the substantive content of public value.

Third, government actors (politicians and officials) cannot unilaterally command public value, but must work through others, using the skills of diplomacy, negotiation, and mediation. There is an element of this last dimension in Moore's managerial model which introduces the ideas of ‘deliberation’ and ‘learning’. But little is made of it in his text and, again, he does not address the central issue of why public managers are the chosen ones constructing ‘solutions’ (Moore 1995:179–185).

Accordingly, instead of attempting to determine (and usurp) the public interest, officials may play a more valuable role in monitoring and overseeing the process of public dialogue. Although they may have their own preferences (and can engage in the shared narrative contributing to the meaning of the public interest), public official can become the guardians of process, not content, working with shared meanings of the public interest and shared understandings about their roles. They may be bearers of legality, disinterested in outcomes, and motivated by honesty and integrity. They can act as counterweights to partisan interests. They remain a repository of institutional scepticism. Officials can become trustees of knowledge, experience and expertise. They can serve as protectors of the longer view. They may also have a role in addressing procedural issues of equity involving the under-represented or the ignored (see Goodsell 1990; O’Toole 1997).

The News Isn't All Bad: The Ladder of Public Value

Our criticisms of public value management focus on its underlying assumptions, especially on its treatment of party politics, the requirement for ‘managers to do politics’, and the contested nature of the public interest. We accept public officials should seek to innovate and improve services. We also accept it is important for managers to increase the value of their services to the public. But as important, we need to know the limits to any approach. The key question is which bits if any of Moore's public value approach will work in Westminster systems?

So, we suggest looking at public value approaches along two dimensions. We distinguish between operational behaviour and normative vision; and between high and low levels of risk (see Table 1).

Table 1.  Forms of Public Value

	Operational	Normative
High risk	Transformative leadership	Adventurism
Low Risk	Incrementalism	Emergent strategies

Using this matrix, we construct a ‘ladder of public value’ identifying the most useful and applicable features of the approach while highlighting the most arbitrary, voluntaristic and risky features for public officials.⁸ If public value is mutual incremental adjustments to improve performance, or administrative technologies designed to create better relations (or production) between policy implementers and citizens, then it will have some validity. In short, lower levels of the ladder can work. So, for example, better information on the needs of clients is useful intelligence; collaborative forms of co-production can assist delivery mechanisms for all stakeholders (see Alford and Speed 2006); and the better alignment of incentive structures with program goals can all produce better outcomes. These innovations are relatively low-level and initiated by officials who are best placed to see the possibilities. But the higher one moves up the ladder, the more the role of the public servant approximates to that of Platonic guardian, the lower the probability there will be any effective reform (see Table 2).

Table 2.  The Public Value Ladder

Levels of Risk	Examples of Public Value	Key Questions
High risk	Ideas-driven management; searching for high degree of stakeholder consensus and engagement; coalition-building by public officials; finding constituencies that value what the manager wants.	Why do we want or need Platonic guardians? Should managers be ‘doing politics’? What are the risks of managers becoming brokers and ‘explorers’ in search for value? Who authorises them to redefine political priorities? Is there a threat to access and equity?
Driven by normative vision; highest political risk
Medium to high risk	Substantial engagement and empowerment of managers and clients in policy formulation and implementation; devolving decision-making capacities to clients; more complex innovations in delivery processes.	Does governments risk ceding control of public policy to the community rather than keeping it in the hands of accountable elected officials? Will there be limited oversight from ‘authorising actors’? Is there a risk of capture and clientelism?
Strong normative orientation; high-level of political risk
Medium risk	Ambitious forms of co-production; technical improvements in delivery chains; assistance to comply with obligations, self-administered modes; adoption of new technologies.	How do managers identify improvements in the delivery chain? Can they enhance operations without changing policy settings? Are the risks largely technical? Did it work? Did it produce the intended benefits to production?
Principally operational but with increasing elements of political risk
Medium-low risk	Direct consultation with clients over delivery and compliance systems; systematic use of evaluation data; public reporting on targets informed by client preferences.	Will operational initiatives threaten administrative coherence and consistency? Will active managers become disillusioned if prevented from innovating? How do we manage hidden costs and unintended outcomes?
Principally operational but more adventurous reconfigurations – some political risk
Low risk	Market survey information of clients; client feedback information, satisfaction ratings; better information on client needs-expectations.	Will politicians and clients see and appreciate the benefits of minor operational improvements? Will innovations lead to administrative savings or improved performance?
Marginal operational adjustments, low-level political risk

For two decades politicians in Westminster system have railed against ‘statesmen in disguise’ and demanded responsive public servants (see for example Rhodes and Weller 2000). They are not about to do an about-face and hand the initiative back to their officials. Any brave soul who exercises the entrepreneurial option will discover all too rapidly the dangers to individual officials.

Conclusions

Arguably, public value approaches have provided a much-needed boost to theories of public administration. Such a philosophical challenge has not been seen since neo-liberal and public choice theories revolutionised public management in the 1980s. Indeed, public value management, with its emphasis on public innovation and the worth of public managers, poses a fresh ideological challenge to the neo-liberal mantra governments have been peddling for over two decades. This esteem for the public sector partly explains why public value approaches have been so readily embraced by middle to senior public servants in constitutional bureaucracies who want to restore some pride in their profession, to be acknowledged for doing good work, and to regain some self-worth and intrinsic reward for their public service. We do not find governments and politicians lauding the praises of public value approaches. Rather, we find the discourse driven by academics genuinely committed to the ethos of public service, by think-tanks sympathetic to government provision and social justice, and by operational managers who see it as offering a new set of navigational reference points with which to navigate the ups and downs of public policy.

There is no inherent problem with highlighting the need for innovation and for political management as essential skills in modern public administration. They are necessary whether public managers address the issue of public value or not. There is value in low-risk operational innovations. So, we clear up Moore's two confusions by stipulating that ‘public value’ is best regarded as a tool used by public servants to identify and implement operational improvements at the workplace But, in addition, we must be clear in defining the expected role of the ‘public manager – a label that, in Westminster systems, refers only to non-elected neutral officials.

If we see public value as a theory that anoints not only elected but also unelected public managers as the Platonic guardians of the public interest, and normatively endorses this position, then the approach encourages managers to usurp the democratic will. It is a repackaging of bureaucratic self-interest. In public value approaches public officials are encouraged to bypass the conventions of democratic politics and seek to interpolate the public interest as a means of enhancing their significance and self-worth.

Politics is ever present. It cannot be contained or sidelined merely to an add-on ‘authorising environment’. Politics is not just something to ‘be managed’ by public managers. In reality, politics and management co-exist in given contexts and with specific actors. Politicians can exercise different dimensions of power, will have different motivations and pressures on them, and can use different tactics at different times. They are far from powerless arbiters of managerial entrepreneurialism and fundamentally remain accountable for public policy decisions. We agree with Aaron Wildavsky writing back in 1968 about the then fashionable management reforms of Planning, Program Budgeting Systems (PPBS) when he vigorously defended the primacy of politics:

political rationality is the fundamental kind of reason because it deals with the preservation and improvement of decision structures, and decision structures are the source of all decisions. … There can be no conflict between political rationality and … technical, legal, social, or economic rationality, because the solution of political problems makes possible an attack on any other problems, while a serious political deficiency can prevent or undo all other problem solving (Wildavsky 1968:393 citing Diesing 1962:198, 203–204).

Public value pretends public managers can supplant politics and play at being the new Platonic guardians of society. In Westminster systems the idea is fraught with enormous political risk. Ultimately the politicians remain responsible and accountable for whatever outcomes are attempted – even if it only becomes known some time later. Our message is that we cannot take the (party) politics out of public management in Westminster systems – and we should not try.

Acknowledgements

We would like to thank John Butcher, Mark Moore, and Gerry Stoker for their help and advice. The remaining defects are our responsibility.

Endnotes

• 1

  See for example the discussion in Plowden 1994; Podger 2007; Rhodes and Weller 2001; and RIPA 1987.

• 2

  We do not know whether Mark Moore would agree with the argument that there should be separate theories but he suggested the idea to us. Private correspondence April 2007.

• 3

  The term Platonic Guardians is derived from the Greek philosopher Plato's notion that it was the duty of virtuous officials to guard the societal good. They were arbiters of the public good and justice and presided over political processes.

• 4

  By contrast, some British writers who discuss public value examples talk more about value being created in a process of ‘iteration between politicians, public managers and the public’ where ‘value is ultimately defined by the public themselves… through a variety of means and refracted through the decisions of elected politicians’. See Horner 2005:36; Kelly, Mulgan and Muers 2002:4.

• 5

  It is not central to the argument of this article, so we observe only in passing that we do not accept Moore's account of rule-bound bureaucracies. Bureaucracy has many virtues and it is not inevitably or invariably reactive and conservative. See, for example, Goodsell 1985 and Olsen 2005.

• 6

  Although Moore cites Lindblom's work (1990), he pays little or no attention to the implications for his own analysis (see Moore 1995:369, note 84).

• 7

  On the roles of ministers see Marsh, Richards and Smith 2001; and Rhodes 2007.

• 8

  We borrow the idea of a ladder from Arnstein's (1969) ladder of participation.