Go to top of page

Delivering performance and accountability

Please note - this is an archived publication.

Commissioner’s foreword

Accountability is fundamental to good government. It is one of the cornerstone values of an open democratic society. In April 2008, in an address to heads of agencies and the Senior Executive Service, the Prime Minister, Hon. Kevin Rudd MP, said that the Government’s agenda included ‘rebuilding a culture of accountability across all levels of government’.

A contemporary challenge for the APS is to improve accountability but in a manner that does not constrain our ability to provide advice and services in an innovative and agile way. The recent Cutler Review of the National Innovation System concluded that experimentation in innovative policy and administration should be a major theme of the current refashioning of federal relations, and that modern innovation requires the loosening of hierarchical routine.

If we lived in a certain world, we would always be able to work out in advance the best way of achieving our objectives and work methodically to do so.

Highly prescriptive approaches served us well in the past, and remain appropriate in some cases, but in a complex and uncertain world, full of shadowy doubts and possibilities, decision-makers can’t always know whether their strategies are appropriate or what the consequences of those strategies might be.

Governments around the world are recognising that networks and markets provide flexibilities and rapid learning that are not available in traditional rules-based organisations. These mechanisms allow for levels of responsiveness and citizen engagement that are not always possible in a rigid hierarchy.

We need to learn how to maximise the opportunities provided by these new mechanisms. In some cases that means removing prescriptive and rules-based approaches that are barriers to innovation, or finding new ways of measuring and evaluating performance that are open to uncertainty and experimentation.

But we need to do this in ways that preserve proper levels of accountability for the way that public funds are used.

The APS has twin challenges: removing unnecessary obstacles to innovation where this is required to improve the quality of outcomes achieved in complex and uncertain policy areas; and developing more variegated accountability and performance arrangements that are better suited to new modes of policy implementation and the needs of a diverse, informed and networked society.

I hope that this paper, which has benefited from the contribution of Mr Roger Fisher, will stimulate discussion and encourage public sector managers to reflect on these important challenges.

Lynelle Briggs
Australian Public Service Commissioner

Executive summary

Governments and the citizens they represent have long expected the public service to provide advice and deliver services in a transparent and accountable way.

Accountability is fundamental to good government. It is one of the cornerstone values of a modern, open society.

Over time, complex and sophisticated systems of accountability and performance management have developed to ensure that money is spent in accordance with the wishes of the Parliament and that public servants are properly accountable for the ways in which they perform and behave.

Australia’s accountability and performance management systems have been updated and extended over the years. They now incorporate components of horizontal accountability that were not present when the last major review of the public service was undertaken in the mid-1970s.

Though accountability arrangements have evolved over the years, they still reflect the techniques and values of the industrial era in which they were developed. They are based on hierarchical modes of decision-making and sequential approaches to problem solving, and they require single points of accountability.

These arrangements have served Australia well, and continue to be appropriate to many if not most of the functions of government.

But governments are facing new policy challenges, such as climate change, water scarcity, Indigenous welfare, and diseases linked strongly to lifestyle, problems which traditional techniques do not seem able to address effectively. These problems are difficult to identify and solve as they have multiple causes interacting in complex ways that are not well- understood. As the Prime Minister, Hon. Kevin Rudd MP, has stated, ‘… a business as usual approach … is not working. Most old approaches are not working. We need a new beginning.’1

In recent years, governments have attempted to deal with problems of this kind by employing a range of new public policy implementation models. Examples include whole of government arrangements, the use of networks of public and private sector providers working together to deliver services, and the application of market mechanisms to achieve public policy objectives. These models, which tend to work more through networks than linear approaches, often include social processes in distributed decision-making rather than hierarchical authority. Indeed, the current Council of Australian Governments (COAG) reform agenda is likely to encompass a variety of these new models for delivering services and programmes.

The new modes of policy implementation are collaborative and can seem unstructured or messy. They require levels of risk taking, experimentation and engagement with communities that do not fit comfortably within current accountability and performance management arrangements.

This paper examines how the current accountability and performance management arrangements are dealing with these new modes of policy implementation.

It points to accountability gaps that have emerged as the new modes of policy implementation have failed to achieve the standards of transparency and accountability that the public expects from governments.

The paper also contends that the current arrangements may constrain innovation. They are forcing public servants back into traditional problem solving techniques, even when they are being asked explicitly to be more adaptive and inclusive.

A one-size-fits-all accountability and performance management framework may be seen as anachronistic in light of the diversity of policy implementation modes and the accountability and performance gaps emerging in the current framework.

A flexible and adaptable public service needs to be able to devise new ways of tackling difficult policy problems and delivering services, and be prepared to experiment without always being sure of what their likely effects will be or how well they might work.

In addressing complex policy problems, decisions about accountability and performance management need to be a key part of policy design, because accountability and performance management arrangements play an important role in defining how problems are understood and addressed. Accountability systems that punish public servants for their mistakes will constrain policy innovation, and limit the public service’s capacity to deal flexibly with new and emerging problems.

The paper canvasses more extensive options for reforming accountability and performance management frameworks, but argues for a gradual, case-by-case approach in applying fit for purpose accountability and performance management arrangements to specific programmes and policy initiatives.

The solution to designing for innovation and flexibility does not lie in giving up on goals and objectives, but in formulating different ways of assessing performance and adjusting expectations in relation to how performance information is interpreted and applied.

The paper acknowledges the obstacles to be overcome in realising a more varied model of accountability and offers a framework to assist in designing fit for purpose accountability and performance management arrangements. It also makes some concrete suggestions for modifying the current framework which take into account the decision-making context and the selected mode of policy implementation.

The accountability framework

The concept of accountability

Accountability is fundamental to good government. It is one of the cornerstone values of a modern, open society.

The legitimacy of our institutions rests on confidence that public servants and elected representatives will provide a full and honest account of their actions when called to do so and be held accountable. Australians rightly see a high level of accountability of public officials as one of the essential guarantees and underpinnings of efficient, impartial and ethical public administration. Accountability is a central pillar of our political tradition.

At its simplest, accountability involves being called to account to some authority for one’s actions.2 In a democratic state, the key accountability relationships are between citizens and the holders of public office, and between elected politicians and bureaucrats.3

This model of accountability is captured in a Management Advisory Board/Management Improvement Advisory Committee (MAB/MIAC) report, which stated: ‘In the context of the relationship between public servants, secretaries of departments and ministers, and ministers and the Parliament, accountability is defined as existing where there is a direct authority relationship within which one party accounts to a person or body for the performance of tasks or functions conferred, or able to be conferred, by that person or body.’4

This definition incorporates a number of features: accountability is external, in that it is owed to someone or something other than the person who is being held accountable; it involves social interaction and exchange, though this may be somewhat one-way; and it implies rights of authority, in that the body to whom accountability is owed asserts authority over those who are accountable.5

However, the concept of accountability does not have a ‘fixed or precise meaning’.6 The relatively simple idea of being called to account for the stewardship of funds has been added to over time and accountability is now often associated with other ideas that relate to the performance of government. In many countries, accountability has become entwined with the concept of managerial performance.7 In addition to being accountable to an external authority for the public money they oversee, public officials are also held to account for how this money has been spent, and for the quality of the decision-making that underpins such spending.

Some commentators have suggested that performance management has emerged as a rival to traditional forms of democratic accountability.8 Accountability and performance can be seen to operate as rival concepts, in the sense that accountability arrangements impose costs that reduce flexibility and innovation. In Australia at least, accountability and performance management have not been presented in official documentation as rival standards. But the traditional idea of accountability has sometimes been subsumed within a performance management framework that emphasises results rather than process and is expressed in terms of managerial assessments of progress against outcomes, outputs and benchmarks.9

The concept of accountability has also come to include the responsibility that individuals might have to their professional ethics and/or values as well as their personal morality. This extends the concept of accountability beyond the traditional idea of external scrutiny; it refers to an inward rather than an outward obligation and, in the case of professional values and/or ethics, it can proceed without a visible process. The Public Service Act 1999 (the Act) seeks to fuse these very different notions of accountability by setting out the values which public servants need to uphold (APS Values (s. 10(1) of the Act)), and a code of conduct for public service employees (APS Code of Conduct (s. 13 of the Act)), as well as external processes through which public servants can be held to account.

Vertical and horizontal models of accountability

The traditional model of accountability describes a vertical chain that provides a continuum of accountability relationships between the electorate, the Parliament, the Government and the public service. Developments in administrative law over recent decades, however, have extended and strengthened the horizontal accountability of public servants and Ministers.

Freedom of Information (FOI) legislation, the Administrative Decisions ( Judicial Review) Act 1977 (the ADJR Act) and administrative appeals processes have created new channels for bringing information on the actions of public servants into the public domain. Public servants are now legally accountable for their actions to institutions other than the Parliament, and can be obliged to explain their actions and justify their decisions to the courts and quasi-judicial bodies. As a 1993 MAB/MIAC report indicated:

Traditionally, the primary accountability obligation of public servants is to the government of the day. Ministers are accountable to the Parliament for the exercise of ministerial authority, while public servants are accountable to ministers for the exercise of delegated authority. But these hierarchical relationships have now been complemented by public servants’ duty to explain or justify their actions to various review bodies and directly to Parliamentary committees where the minister neither knew nor should have been expected to know.10

The legislation establishing these arrangements confers an authority or measure of control over the actions of public servants that is legally enforceable. This has had implications for the way in which decisions are made, and for the cost and complexity of the systems and processes that are maintained to support decision-making.

Figure 1 sets out some of the key accountability relationships that bear on the work of Australian public servants.

Figure 1: Public service responsibility and accountability framework11

As the model shows, public servants are accountable mostly through a vertical and hierarchical chain, but it does include some horizontal accountability to external review bodies. This accountability map includes formal processes for reporting to the Parliament such as annual reports, portfolio statements, appearances before Senate committees and the system of performance reporting on programme outcomes. In practice, public servants are also subject to corporate and management cycles within their organisations that cascade down from high-level targets and feed performance information up into the accountability chain.

Other horizontal accountability arrangements are emerging. Some are less formal or confer less legal authority on the party to whom the accountability is owed, but they can nevertheless provide powerful checks on the activities of public servants.

A number of public service reforms over recent decades have sought to make public servants more responsive to their clients. Australian public servants are encouraged to be responsive to the needs of members of the public to whom they provide services. This has been supported through mechanisms such as Citizens’ Charters, charging users for some services, and performance benchmarking. As Richard Mulgan notes, public servants are now seen as accountable ‘outwards’ to the public as well as being accountable ‘upwards’ through the hierarchical chain of managerial command.12 These horizontal relationships are likely to be of increasing importance to the new and emerging modes of policy implementation being developed to deal with complex and intractable problems.

This idea of accountability is at present relatively weak and needs to be differentiated from the traditional Westminster concept. To the extent that any accountability relationship exists between a public servant and a Centrelink client, for example, it is one in which the client has no direct power to call the public servant to account for his or her actions. Aggrieved citizens may lodge complaints with the service provider, or complain to their local MP, but redress through these forms of complaint ultimately relies on the good faith of the service provider. The citizen has no power to sanction the public servant other than by refusing to accept the service on offer. The power of the citizen to call the service provider to account is even weaker in the case of government services delivered by an outsourced provider. This is a significant challenge for emerging modes of policy implementation based on distributed and devolved government.

Accountability trade-offs

Accountability is essential to the democratic system, but it is only one of a number of qualities necessary to an effective system of government.

Traditional systems of accountability can be expensive to maintain and are sometimes complicated to work within. Rules-based accountability systems and standard operating procedures evolved as a means of reducing the incidence of corruption, nepotism and waste.

They narrow the scope for personal discretion and allow Parliaments to locate the source of decisions and trace responsibility to the officeholders who need to be held to account.

However, the accountability systems that prevent fraud, waste and abuse also reduce the scope for independent action. Rules and procedures can slow decision-making and constrain the choices available to decision-makers when they are confronted by new or unusual circumstances. Transparency increases the probability that poor decisions will be exposed. This can help to prevent foreseeable and preventable errors, but it can also encourage risk avoidance and conservative decision-making.13

The performance costs of control systems are well-documented. In 1991, James Fesler and Donald F. Kettl pointed out that:

Excessive controls can disrupt consistent administration and produce inequities. Excessive controls multiply requirements for review of proposed decisions, increase red tape, and delay action. So much energy can be spent attempting to control administrative activities, in fact, that little time or money is left to do the job at hand. Excessive controls, therefore, may dull administration’s responsiveness to its public.14

Agencies with high levels of accountability often display low levels of innovation and flexibility.15 Innovation usually involves risk. But in the public sector, where there is a high level of scrutiny of how things are done, the disadvantages of taking a risk greatly outweigh the possible benefits of success. Departures from standard procedures are discouraged because mistakes are highly visible and success is difficult to measure.

The day-to-day incentives created by traditional accountability arrangements favour adherence to process over innovation, and routine over experiment. These arrangements encourage officials to become highly adept at producing uniform products and operating within ‘routinised’ processes. That isn’t always a bad thing; as Geoff Mulgan points out, society doesn’t want too much experimentation with benefit payments, the operation of traffic lights, school curricula or court procedures.16

It would be unfair to say that accountability and performance management arrangements are the sole determinant of the level of innovation within an agency, or that it is not possible to innovate in an agency that is strictly performance managed. Much depends on the clarity of expectations, where the points of control are applied, how performance is measured and the underlying culture and values. The culture of an organisation is shaped by its formal accountability structures, but it is also influenced by the quality of relationships and leadership. Courageous leaders and talented staff can innovate even within a rigid accountability framework, but an overly rigid and prescriptive framework can make this more difficult. In social policy areas, including employment and the Indigenous sphere, various trials and pilots have been conducted to determine the way forward; however, in view of current demands for greater innovation and even more new ideas, it is important to ensure that there is a supportive environment for innovation.

Citizens in market democracies demand more from their governments than standard, one-size-fits-all solutions. Successful governments also need to create customised services that respond to the wide variety of circumstances and the highly differentiated needs of citizens living in complex, diversified urban societies.17

In meeting these needs, a flexible and adaptable public service needs to be free to innovate, and to sometimes try things without always being sure of their likely effects, or how well they will work. On occasion, the best policy option will involve uncertainty or even some deliberate trial and error. Accountability systems that punish public servants for unforeseen or unpreventable errors will constrain policy innovation and organisational learning, and limit the public service’s capacity to deal with new and emerging problems.18

Case study

How accountability arrangements influence behaviour and shape the way problems are addressed

Natural resource management and land management problems defy traditional bureaucratic and analytic approaches to problem solving.

Adaptive management is one of the few management approaches with the potential to work successfully with the factors of complexity and uncertainty. However, attempts to apply adaptive management to complex problems in the natural resources area have been frustrated by the demands of traditional accountability and reporting structures, and the behaviours that they reinforce.

The Heartlands project was funded from 2001 to 2004 to address the management of dry land salinity. It was explicitly badged as an attempt at adaptive management.

Heartlands planners aimed to design and implement landscape scale land use changes in order to learn from those changes. The project was funded by the Australian Government and State Governments and managed by an alliance of scientists, agency employees and farmers.

Despite the explicit objective of applying adaptive management in this project, reporting guidelines required information on progress against targets, and the continuation of annual funding was known to be contingent on acceptable performance in the previous year. A participant observer over the life of one of the major Heartlands sites reported a persistent anxiety amongst the programme managers about uncertainty, and a feeling that complexity added to uncertainty. The programme managers’ anxiety around complexity and uncertainty influenced their perception of the risk attached to the programme and its capacity to continue to attract funding. This was managed in a number of ways that limited the effectiveness of the ‘learning by experience’ that was fundamental to the adaptive approach.

The team sought to reduce ambiguity by commissioning narrow scientific investigations and restricting the flow of information. The project managers sifted information and requests for collaboration actively before passing them on to the wider community, in order to protect the communities from too much and too varied information, and to avoid losing credibility by being seen to be associated with a project that may fail. Management and decision-making were restricted to a core group within the management team, and reports to the funding bodies emphasised progress towards targets rather than lessons learned from surprises or mistakes.

One agency member of the project team noted at a planning meeting: ‘“… Government wants to keep an eye on what’s being spent. The pressure is on you guys.’” The observer noted that this articulated the subtle but constant pressure exerted by levels of government and the community to reduce the risk and treat the Heartlands project as a conventional management exercise rather than an exercise in adaptive management.19

Accountability and performance management frameworks shape organisations and influence the behaviour of people who work within them. They have an effect on the appetite for risk and the willingness of employees to adapt and innovate. Some public services need to be provided in an equitable and consistent way, but in delivering others, governments desire experiment and innovation over uniformity. An effective accountability and performance management framework needs to take into account the different trade-offs required to deliver the public services that society requires.

The accountability challenge for new modes of policy implementation

New modes of policy implementation

Many contemporary public policy problems seem remarkably immune to the conventional problem solving techniques that have evolved within government. The Prime Minister has made it clear, for example, that a ‘business as usual approach towards Indigenous Australians is not working. Most old approaches are not working. We need a new beginning.’20

Governments in most advanced economies are responding to this challenge by experimenting with new modes of public policy implementation. Elaine Kamarck identifies three emerging forms of policy implementation: reinvented government built around performance managed bureaucracies; government by network; and government by market.21 It is likely that all three forms of policy implementation will be used to support the agenda announced for COAG in late 2007. Kamarck includes cross-agency cooperation (referred to in Australia and the United Kingdom as whole of government and joined up government respectively) as an example of reinvented government.22

In Kamarck’s model, performance managed bureaucracies are public sector organisations in which central control mechanisms such as budget rules, personnel rules and procurement rules are traded off for enhanced flexibilities and replaced or supplemented by outcome or output performance measures. Performance measures act as market proxies and customer service is used to model organisational behaviour.

In government by network, the public service works with a wide variety of institutions. The Government funds other organisations to do the work it wants done. Government by network can encompass churches, universities, research laboratories and cooperative research centres, non-profit and for-profit organisations. Networks can also include state and local government agencies. In the Australian context, an example of government by network is the provision of health care and programmes to tackle homelessness.

For a number of reasons, governments choose to create networks to implement policy. Sometimes, the organisations in the network bring capabilities and a style of operation that does not always exist within the public sector. Jurisdictional or constitutional boundaries can also prevent a government body from providing services directly. In some instances, other levels of government or non-government bodies are closer to the community and are therefore better able to tailor their services to local needs. Governments also occasionally use networks because they have ideological or other preferences for locating work outside the government sector.

Government by market involves few, if any, public employees and little or no public money. The Government uses state power to create or support a market that fulfils a public purpose, where that kind of market would not exist in the private sector. Examples of government by market include emissions trading schemes, the use of tradeable licences to manage fisheries, and the transition in water policy from a system of extraction-based rights attached to land titles and permit licences to a market-based system of water trading.

New modes of policy implementation have sometimes been successful in dealing with complex problems because they have not been subject to the usual accountability constraints on innovation and adaptation. The OECD has observed that distributed governance has inherent risks for democratic control and accountability.23 Though it is not generally acknowledged, freedom from the usual rules on accountability is part of the attraction of these modes of policy implementation for decision-makers.24

This presents a challenge for the democratic state: how does it create the accountability and control mechanisms necessary for the successful operation of new and emerging modes of policy implementation?

Performance managed bureaucracies

Elaine Kamarck describes performance managed bureaucracies as government shorn of its public sector trappings, especially, rigid rules regarding budgets, personnel and procurement.25 She points out that many of the strategies of reinvented government are borrowed from the private sector.26 Kamarck’s idea of a reinvented government seems very similar to the new public management, with its ‘general cuts to public sector spending’ and ‘emphasis on private sector styles of management’.27 Kamarck refers to the ‘grand bargain’ between bureaucrats seeking to remake their organisations, and the elected representatives who hold them accountable within a new context in which strict, sometimes stultifying rules are replaced by performance measures.28 But this can be achieved only if the goals of public sector organisations are clearly articulated and measured.29

Many of the problems that governments deal with do have clear goals and a linear relationship between means and ends. When this is the case, problems can be managed effectively within traditional accountability and performance management arrangements without over-simplification or distortion. Governments also generally have good reasons for wanting issues to be dealt with in a way that provides a high level of security or confidentiality. Kamarck argues that where Governments prefer a high degree of control over a process, and where outputs and outcomes can be measured, public policy objectives are best met by reinvented or reformed public sector organisations operating within government. Examples include issuing passports and drivers’ licences, and determining eligibility for government benefits.

The existing accountability and funding arrangements were designed primarily for departments working independently.30 When performance managed bureaucracies are operating alone, and when they are dealing with relatively straightforward problems, the traditional accountability and performance frameworks face few challenges. Accountability problems arise when performance managed bureaucracies are asked to work across organisational or jurisdictional boundaries on joint problems that are complex in the sense of being decentralised or ambiguous.

Whole of government approaches

Whole of government approaches fall mainly within Kamarck’s description of performance managed bureaucracy,31 but whole of government approaches can also involve elements of government by network when non-government entities are closely involved in the policy design or delivery of services. The Northern Territory Emergency Response (NTER) is one example.

In Australia, a considerable investment has been made in developing a capacity to deliver policy in a more connected, or whole of government, way. The Management Advisory Committee (MAC) has argued that there are many imperatives which make whole of government work increasingly important. These include pressures on the public service to offer advice which comprehends a broad range of stakeholders’ views, and to respond innovatively to complex and interrelated policy challenges such as environmental or rural issues. There are also pressures for a more joined up delivery of programmes across areas as diverse as security threats, drug dependence and obesity, as well as rising community expectations for easier access to government that could be provided through better integration of service delivery.32

The MAC report, Connecting Government, has defined whole of government as public service agencies working across portfolio boundaries to achieve a shared goal and an integrated government response to particular issues. The report acknowledges that this could describe formal or less formal arrangements and that it encompasses policy development, programme management and service delivery. Whole of government can also include very different management styles and decision-making arrangements, from highly-centralised arrangements with tight coordination by a lead agency through to collaborative decision- making between partners where power is shared equally.

Whole of government approaches have worked well in emergency situations. Examples include the whole of government response to the 2004 Boxing Day Asian Tsunami, the rapid evacuation of Australian citizens from Lebanon in 2006 and the response to Cyclone Larry in Queensland the following year.

In each of these examples the critical nature of the problem has been acknowledged and the policy objective defined clearly. This has allowed leaders to draw separate agencies in to a centralised network in which they have been able to demand a high level of coordination and common effort, even from marginally affiliated players or competitors for resources and policy attention.

Whole of government arrangements have also been used to implement NTER, with the Taskforce leading the response comprising senior government officials as well as externally appointed experts.

The Australian accountability and performance management frameworks have flexibilities that can be used to support whole of government approaches. Ministers may bring joint submissions to Cabinet, and the Expenditure Review Committee may consider related proposals from separate portfolios in a single session and on the basis of a common briefing. The financial management arrangements also provide some flexibility. The current Department of Finance and Deregulation (Finance) guidance on outcomes and outputs sets out that where agency outcomes are similar, it may be appropriate for those agencies—even if they are in different portfolios—to agree to a single outcome statement within which they can establish their own outputs.

Agencies can also have funds appropriated to a central account, to be drawn on by staff in different agencies. A Northern Territory Flexible Funding Pool Special Account, for example, was established to support NTER. Departments also used many of these flexibilities in managing National Heritage Trust and National Action Plan for Salinity and Water Quality initiatives. Despite these flexibilities, the common perception amongst public servants is that appropriation and accountability arrangements do not fully support joined up approaches, and that implementing supportive budgeting and reporting approaches can be problematic in practice.

The Australian Public Service Commission has reported that only 27% of Senior Executive Service (SES) and Executive Level (EL) employees, who had dealt directly with people from other public service agencies during the last 12 months, for example, agreed that current financial and accountability arrangements usually facilitated whole of government work. Twenty per cent believed that they only rarely facilitated whole of government work.33 Public servants involved in whole of government activities seem to believe that rigid funding arrangements and programme guidelines are barriers to better working arrangements with other departments.

These perceptions are supported by the work of the Australian National Audit Office (ANAO) and a number of commentators. Though agencies are able to have money appropriated to a common outcome, a recent survey of 44 agencies subject to the Financial Management and Accountability Act 1997 found that none of the surveyed agencies had cross-agency outcomes. The ANAO audit of whole of government Indigenous service delivery arrangements revealed that despite the establishment of flexible funding pools, there was no public reporting of the outcomes of the Shared Responsibility Agreements (SRAs) or the Community Engagement Implementation Assistance Program funded from the common pool. ANAO concluded that insufficient attention had been paid at the policy implementation stage to the original intention that funds would be used flexibly through joint funding arrangements.34

Case study

Indigenous trials—Jointly funded services

ANAO observed that a precursor to effective and efficient whole of government work in delivering services to Indigenous Australians is seamless joint funding of services and initiatives.

In its audit report on the Indigenous trials, ANAO concluded that at present there is no effective way of managing joint funding arrangements.

ANAO also found that, even where joint decision-making structures have been developed, staff drift away from solution brokering towards programme administration. In addition, they see themselves as being accountable to their home departments, because this is where their legal accountabilities rest.

Individual departmental accountability requirements detract from the whole of government focus. ANAO found that line departments have participated willingly in joint assessment processes on programme applications and then continued to make decisions in isolation as to whether or not to approve applications based on their individual assessment.

In some cases, where several government departments were signatories to a contract, each agency developed its own reporting, monitoring and acquittal requirements that were set out in separate schedules to funding agreements. This can impose significant costs on communities, which might be involved in developing separate financial and performance reports for the one project. It would also do little to build trust or cooperation between communities and government service providers.

In order to circumvent these arrangements, many groups developed smaller and simpler agreements that were narrower in scope and much easier to manage. However, these provided little funding flexibility and reduced the whole of government benefits.

Many of these problems relate to the training and capacity of staff. But it would be unfair to leave it at this. The staff involved in the trials were often volunteers and almost always highly motivated. Ultimately, the behaviours of the staff and the departments they represented reflected their understanding of the accountability arrangements, and the underlying incentives of the processes, systems and structures in which they worked.

ANAO believes that the existing outcomes and outputs framework has the flexibility to provide appropriate accountability for whole of government projects. It has suggested, however, that a lead agency should guide any large and complex project and that funding options could be more flexible and better tailored to meet the needs of Indigenous communities.35

These problems can be reduced by sensible reforms within the current accountability framework. ANAO has recommended that the concept of cross-agency outcomes be reviewed in light of its lack of application by agencies to date. ANAO has also suggested that clear guidance be developed for departments regarding how best to engage in joint funding of Indigenous services. An assessment of whether this approach was used to underpin NTER would be useful.

Agencies need greater guidance on the sorts of budgetary and accountability arrangements that are likely to suit different types of whole of government exercises. Departments should also be encouraged to take the initiative in this area more than they have done in the past; they certainly appear to have sufficient authority to align their internal financial management and accountability arrangements where there are good whole of government reasons for doing so. The current accountability or performance management framework does not prevent departments from facilitating whole of government outcomes by exchanging financial information and preparing joint reports against common whole of government objectives. To date, however, they have been reluctant to do so.

Departments have not, as yet, had sufficient motivation to move away from the traditional, vertically differentiated approach to working. Some agency managers appear to have internalised the ‘silo’ approach to their work that is embedded within the existing accountability and performance management frameworks. Accountability and performance systems reward individual and agency performance rather than collective effort; therefore, it is not surprising that managers should behave in a way that reflects the incentives of the system within which they work. Overall, there seems insufficient incentive to overcome organisational inertia and entrenched practice.

The accountability framework is probably not the only cause of this problem. The framework and the incentives for staff themselves to operate within silos reflect deeply-rooted assumptions about the way work should be organised. Whole of government activities require institutional capabilities that are not required for less complex tasks.36 Another significant issue is whether public servants trained to operate within vertical silos always have the skills and attitudes necessary to work effectively across boundaries and with communities that might not share their underlying values and beliefs.37

Efforts to encourage public servants to work differently and to train them in techniques for collaborative working may help, but they will not resolve the fundamental tension between the horizontal responsibilities in whole of government approaches and the vertical accountabilities embedded in the Westminster system of Cabinet Government, in which the existence of separate portfolio agencies reflects an underlying accountability of individual

Ministers to Parliament. Departments’ interests should reflect the concerns and approaches of their Ministers. In the Westminster tradition, these differences are resolved in the Cabinet or by the Prime Minister, but in whole of government approaches, the expectation is that they will be resolved within and between agencies.

It is not always clear, for example, who in a whole of government exercise should take responsibility if something doesn’t work, or if a Minister is not fully informed on matters within his or her portfolio. The attenuation of individual accountability that was evident in the failure of individual officials to correct misapprehensions about the behaviour of refugees in the October 2001 Children Overboard affair illustrates the kinds of accountability issues that can emerge with whole of government approaches.

Various countries have developed their own responses to these challenges and there remains something makeshift and provisional about whole of government arrangements in most countries. Nobody has yet found a means, or perhaps sufficient reason, to move away from a fundamentally vertical, departmental system of government. For most government activities, the concentration of functional expertise within separate agencies or portfolios offers many advantages that outweigh the occasional problems of coordination.

In 2003, the Canadian Public Service Modernization Act (PSMA) established a committee of Deputy Ministers—the Australian equivalent of Portfolio Secretaries—to advise the Secretary of the Treasury Board on whole of government issues.

In Australia, governments have experimented with various structures for managing whole of government initiatives, including task forces managed out of the Department of the Prime Minister and Cabinet, the nomination of lead agencies, the use of boards of management or Ministerial Councils to oversee the work of agencies, and the use of a service delivery agency to manage a loose confederation of service providers. In the United Kingdom, numerous strategies have been adopted to strengthen the centre and the role of the Prime Minister in leading joined up government initiatives. In most cases, the problem of joint accountability has been avoided by retaining a clear authority close to the centre of government.

Ultimately, whole of government approaches are principally about improved coordination. In hindsight, there is perhaps nothing very remarkable in the realisation that a class of problems has emerged whose causes are so complex and the solutions to which are so ‘multi-factorial’ that, as Tom Ling argues, they require a ‘multi-agency’ response.38 The challenge lies in bringing disparate government agencies together when necessary without them losing their capacity to see issues from different standpoints.

In its findings on the Indigenous trials, ANAO suggested that a lead agency should guide any large or complex whole of government exercise. This makes sense in terms of ensuring that someone has ownership of the issue and an incentive to drive it along. The problem with this approach, however, is that it anchors whole of government around the perspective of a single agency. Although a single lead agency can help improve coordination and consistency of process, this approach comes at the cost of fewer opportunities for applying multiple perspectives to a common problem, and fewer incentives for individuals to challenge, dispute and explore.

This doesn’t much matter when a national emergency has been declared and there is broad community support for a concerted effort to solve a problem that has become critical. It is an important limitation when the Government is seeking to deal with a problem that isn’t yet fully understood or well-defined.

More work needs to be done on the accountability and performance management frameworks to ensure that whole of government approaches contribute effectively as an alternative form of policy implementation. At this stage, it seems that they may be best suited to problems that cross jurisdictional boundaries but which are relatively well-defined and well-understood. Yet, they may have a part to play in conjunction with other techniques.

Government by network

Government by network involves collaboration with third parties outside government.

Government by network offers a means of dealing with problems that need customised solutions,or which are highly decentralised. There is no ideal way of solving some community health problems,or of finding jobs for the long-term unemployed.The health and attitudes of the individuals, the quality of their family or social support and the availability of local opportunities all have a bearing on the best approach in the circumstances.

Government by network allows a range of options to be brought to bear on a problem by engaging providers who are familiar with local conditions and have the opportunity to take different and innovative approaches to the issue. Government involvement in research and development is undertaken typically through government by network because it is not bureaucratic and offers opportunities for flexibility, experimentation and innovation. Government by network seems best suited to problems where a government values innovation so much that it is prepared to give up some level of control over the way things are done.39

Where government by network is undertaken with non-government bodies, it is usually achieved through a contract between the government purchaser and the provider. Control is managed in terms of a formal agreement between parties that are otherwise separate and independent. Accountability is determined by the nature and terms of the contract, with each party being accountable to the other for performance as specified.40

This should serve to enhance accountability. Service standards can be specified and performance can be monitored against explicit criteria. The separation of responsibility for supply from the assessment of performance is clearer and more transparent than for services provided in-house and should avoid the potential for conflicts of interest.

However, many contracts for government by network are more open-ended than this ideal model might imply because they deal with more complicated situations. In many contracts the actual requirements are deliberately focused on achieving broad, high-level outcomes, because there are too many potential aspects to be detailed in a legal document, the policy problem cannot be defined precisely, or solutions require a high level of flexibility and adaptation by the service provider. Open-ended contracts can allow for flexibility in the way services are provided, but they also reduce the amount of information available to third parties, including the Parliament, on how services are provided.

Accountability in networks can sometimes be problematic because of the difficulty of locating authority in a networked arrangement. In theory, government by network may alter the way things are done, but it should not change who is ultimately accountable for what is provided. Responsibility may be devolved, but accountability remains with the Minister and the official responsible for the expenditure of public monies. In practice, accountability can become blurred.

Case study

Locating accountability in a networked construction project

Robert Bevan has written recently about the impact of networked governance arrangements such as public-private partnerships on the quality of the architectural outcomes for major public works.

He suggests that, as a result of experience with cost over-runs and delays on large building projects such as the Sydney Opera House, Australia’s governments led the way in developing new ways of financing and managing the construction of public buildings. The previous hierarchical model by which an architect was employed to design and supervise the construction of a major building from within a government agency has been replaced by a networked arrangement in which project managers, fixed price tenders, design and construct contracts, public-private partnerships and financial institutions mediate between the architect and what is actually built.

Bevan argues that these arrangements have been successful in reducing costs and delays, but at the price of banal architecture, often poor value for money over time and buildings that often fail to meet the needs of the people who use them.

He cites the example of the Bathurst Base Hospital. Bevan asserts that the design and construction was so poorly executed that when it opened in 2008, the emergency paging system didn’t work, ambulances couldn’t fit in the car park, operating theatres proved to be completely inadequate, sewage leaked onto beds, and the mental health unit remained empty because of dangerous design features.

The outline design for the building had been developed by the NSW Government Architect’s Office. The design was then put to tender. The successful construction company worked up a detailed design and oversaw the construction of the facility.

The construction company now refuses to say whether it used its own designers or subcontracted the detailed architecture to a private practice. The NSW Government Architect’s Office also refuses to say how much of its design was used.

In this case, existing accountability arrangements were (and remain) unable to ensure that citizens can find out who was responsible for the appearance and design faults of a public building constructed with public money, let alone call those responsible for the debacle to account.

Bevan concludes that, because it is unclear who ‘owns’ the process or the outcome under these kinds of contracts, accountability is difficult to determine.41

Networks are more open and porous than traditional hierarchies. They also have more complicated information flows, multiple points of contact and interlocking responsibilities that lend themselves to buck-passing when things go wrong.42 In many contractual arrangements, there is scope for either the government entity or the contractor to do things that can affect outcomes.43 This sometimes renders it very difficult for users to determine who is responsible for the services provided or the overall outcomes achieved.

Third party providers in government by network face a far less extensive array of accountability mechanisms than do traditional bureaucratic agencies. Parliament and the Auditor-General do not have direct access to the third party providers in government by network, and FOI laws are not applicable to non-government providers. However, the Ombudsman is empowered to investigate the delivery of goods or services to members of the public by a contractor on behalf of the Government. Some commercial-in-confidence information is typically not provided to the public at large, though Parliament has in recent years reduced the scope for withholding contract information from Parliament itself. There are other exclusions from the normal accountability arrangements that apply to bureaucratic organisations operating within government.

Nevertheless, third party providers in government networks often complain that the reporting and acquittal requirements accompanying government contracts are onerous and burdensome. From the point of view of non-government organisations (NGOs) such as not-for-profit bodies, and even some small commercial firms, contracting with government can involve very high costs, and significantly higher than usual levels of accountability, both to the Government and the wider public. Participation in the network may require them to supply more information on their financial transactions and internal processes than they would normally have to generate. This can impose significant costs.

The dilemma is that, while increased accountability requirements may represent a significant operational cost, and may even be felt to compromise the organisation’s independence and core values, such requirements are likely to fall well short of those imposed on public service organisations.

Case study

Compliance costs—The problem of red tape

The leaders of many organisations involved in government by network believe that the accountability requirements attached to government funding are excessive and onerous.

Some of these requirements appear to disadvantage small and not-for-profit organisations. In some cases, they impose on NGOs the structures and processes of large bureaucratic organisations. A recent Australian Government tender, for example, required contractors to set out details of their recruitment policies, including position specifications and job descriptions, and to provide a copy of their organisation’s risk management plan.

An Australian NGO in the field of women’s health is obliged to report against 170 separate line items. It estimates that the cost of complying with external accountability requirements represents around 30% of its budget.

Research into the impact of new funding arrangements, such as contracting, on NGOs was commissioned in 2002 by the then Department of Family and Community Services (FaCS). A report on that research, which was completed in January 2008, documents the experiences of over 600 organisations contracted to provide services on behalf of the Department.

Most of these organisations were relatively small—the median number of full-time staff for all of the organisations was four. Seventy-five per cent of the organisations had an income of less than $1 million per year, and 60% had an annual income of less than $250,000.

The study explored the impact of contractual reporting and accountability requirements from a range of perspectives: type of reporting; frequency; staff hours allocated; and similarities between requirements across programmes.

Just over 50% of the organisations were required to provide financial data. About the same percentage were required to provide client data reports. Surprisingly, perhaps, only one-third of the organisations were required to provide performance reports, mostly on a half-yearly or yearly basis.

The collection and reporting of client data requires the greatest use of staff resources. Over one-quarter of respondents allocated in excess of eight days per month to the collection and reporting of client data. For organisations with a median full-time staffing level of just four, this requirement alone would consume 10% of the organisation’s full-time staffing effort.

Most respondents believed that additional client information was of little relevance to the management of services. The reporting was generally viewed as an obligation that primarily serviced the needs of the Department, rather than as a tool with which to engage in quality improvement.

Respondents indicated that reporting requirements in relation to public accountability improved the internal functioning of their organisation, but expressed more general concerns about the impact of the contracting arrangements on performance in delivering outcomes across the sector.

Non-government providers are critical of what they see as government by network’s focus on means rather than ends, and its neglect of long-term considerations of organisational diversity, sustainability, and the distinctive community foundations that give NGOs their unique social capacities.44

Contracting is sometimes criticised because it narrows the scope of what providers can be held accountable for.

The Job Network, for example, has delivered job placement services more efficiently and effectively than the former Commonwealth Employment Service—the budget cost of providing labour market services has been substantially reduced and exit rates from unemployment have improved slightly, compared to previous arrangements.45 The employment services arrangements allow the Government to subject private employment agencies to considerable scrutiny and to reward good performers and penalise bad ones. But the detailed results of this scrutiny and consequent actions remain commercially confidential and are not available to the Parliament or the public.46

By changing the type of information available, and to whom that information is provided, government by contract has enhanced the traditional vertical chain of accountability supported by a range of external review bodies such as the Ombudsman and the Administrative Appeals Tribunal (AAT), and produced a stronger system of managerial accountability between the contracted parties. Arguably, this leads to more innovative and efficient solutions for complex and distributed problems, but it leaves an accountability deficit relative to internal service providers.

The most significant accountability challenge for government by network lies in accountability for the performance of the network as a whole. The accountability focus in most government by network arrangements is understandably on the performance of individual providers.

However, public concern about the treatment of detainees in immigration detention centres and nursing home residents suggests that the community has expectations about the way in which publicly funded services that go beyond the strict requirements of contracts are provided. The traditional accountability framework for government by network is effective in auditing contracts to ensure that money is being spent appropriately and that services are being provided as specified in the contract, but it does not seem to extend very easily to evaluation of what the network as a whole is achieving, or what impact the performance of individual providers might be having on the rest of the network.

In 2006, the then Shadow Minister for Employment and Workforce Participation, Senator Penny Wong, criticised the Job Network because of what she perceived as the insensitivity of the contracting arrangements to wider social impacts. She argued that Job Network providers were focused on matching job seekers with available jobs, but did not have strong incentives to raise participation rates or to address the social causes of long-term disadvantage.47 On 13 May 2008, the Rudd Government announced changes to the employment services model aimed at addressing some of these concerns by providing more support for disadvantaged job seekers and increasing the focus of providers on addressing skills shortages.48

There is a similar challenge in systematising learning from the wide variety of organisations and innovations operating in the network.49 Bilateral contracts provide powerful performance incentives for individual providers and, over time, competitive contracting rewards organisations that adopt the best practices. Where contracts operate for several years before they are retendered, however, this can be a slow and uncertain way of disseminating lessons across the system. Even where services have been devolved successfully to third party providers, governments have a continuing interest in the operation of the network as a whole.

Although government by network provides an effective way of devising innovative solutions to complex and highly-distributed problems, it is vulnerable to some serious accountability deficits under the current accountability and performance management arrangements.

Participative governance—Partnerships with communities

Government by network can involve collaboration with communities to achieve common goals.

Meredith Edwards has argued that progress on complex issues, ‘and particularly sensitive ones such as indigenous policy and delivery issues, will require moving beyond consultation to a more active engagement with those likely to be affected by decisions’.50 She identifies several necessary ingredients for the successful creation of participatory policy mechanisms: strong leadership; trusting relationships; and shared decision-making power.

Case study

New ways of engaging citizens in policy making

The question of who participates poses a challenge for networked decision-making.

The Internet has opened up opportunities for citizens to contribute to policy in their own right, and not only through organised representation.

The NSW Government convened an alcohol summit in 2005 for the purpose of devising strategies to reduce the harm caused by the misuse of alcohol. Part of the summit process involved direct contributions by citizens through an online channel.

Several local councils in Victoria are also exploring Internet-based channels to engage with communities on issues such as development controls and urban planning, waste management and other environmental matters.

In the United Kingdom, domestic violence legislation was influenced materially by feedback from women who contributed to the process only because the online channel provided the anonymity that allowed them to tell their stories in safety.

Participants in these processes do not carry any formal accountability for the quality of their input or the decisions taken. Under current accountability arrangements, governments generally retain formal and legal accountability for their actions, even where they succeed in using participative processes to build support for their initiatives. This can limit the extent to which governments are prepared to devolve policy design to the participants in collaborative processes.51

The practice of participatory governance is fraught with difficulties. Participatory processes are expensive and time-consuming, and require sophisticated institutional support. They make heavy demands on the skills and training of the policy makers who are required to design processes, select public representatives, set procedures and prepare reports while maintaining the integrity, openness and credibility of the processes.

Participative decision-making requires trusting relationships. Without trust, genuine collaboration is impossible. However, the concept of trust and how to design policy processes in order to maximise trust is still a developing area in public policy. Our current accountability arrangements do not rely on trust, but on the demonstration and acquittal of responsibilities. It is difficult to persuade a stakeholder that they are fully trusted when they are denied the opportunity to decide how money is spent and obliged to account for monies they receive.

Meredith Edwards makes the point that it will be difficult to build and maintain trust with communities that have suffered long-term disadvantage and discrimination.52 In reviewing the experience of participative governance in Indigenous policy, Edwards notes that there is a deep divide between the types of structures and processes governments use to ensure accountability, and traditional Indigenous governance arrangements. She also observes that there have been many examples where there was good intent and a genuine engagement at the initial stages of a project; however, this has reverted to old patterns and relationships when government resumed a more dominant role or when officials felt constrained by their accountability obligations.

Edwards raises a series of questions about whether the principles of individual and collective responsibility and accountability to the taxpayer through Parliament can be achieved in the terms in which they are currently expressed when the boundaries between what governments and communities can do are blurred. Tom Fitzpatrick argues in a Canada School of Public Service paper that distributed governance through partnership arrangements involves complex accountability relationships that increasingly require multiple accountabilities and frameworks.53 However, this would seem to imply a significant shift in the way that power relationships and accountability between governments and community partners are defined.

There are concerns that the institutional adjustments needed to make participative governance work would blur traditional lines of ministerial accountability and increase the risk of misappropriation of funds. Some analysts even fear that too much collaboration could undermine government accountability altogether.54

A form of joint accountability is theoretically possible within the current Australian legislative framework in regard to the arrangements that exist for third party drawing rights. Yet, this particular flexibility is heavily conditional and, under the present framework, ultimate authority and accountability for the way in which funds are used rests with the public servant, not with the third party beneficiary. Overseas experiments with collaborative partnerships have often involved joint decision-making and forms of horizontal management, but they have generally stopped short of any formal system of joint accountability.

Case studies at Murdi Paaki in Western NSW have shown that there is scope for forms of collaborative government or participatory governance within the existing accountability framework.55 However, there are limits to how far this can progress in the absence of changes in the underlying model of accountability. Community ownership of solutions is likely to be conditional if ultimate authority to make decisions and to be held accountable for them continues to reside entirely with officials and if departments continue to make decisions in isolation from their consultative processes. ANAO has acknowledged the need to revisit the accountability framework in order to accommodate horizontal ways of working. It is not yet clear how far this would go.

Reports on experience in the Australian states and overseas have contended that for participatory governance to be effective, accountability and performance management arrangements will need to be adjusted so as to accommodate joint accountability and a clarification of horizontal accountabilities. Some reports have suggested that an effective legislative framework acknowledging joint accountability is needed.56

Government through markets

Of all the new modes of policy implementation, government by market presents the most profound accountability challenges. In government by market state power is used to create or support a market that fulfils a public purpose, where that market would not exist in the private sector. Government by market provides an alternative means of implementing public policy, but it does not absolve a government from criticism for the consequences of poor market design or remove its accountability for the outcomes that are achieved.

Elaine Kamarck suggests that government by market is the best, and sometimes the only, realistic implementation option when a policy consensus is reached that requires many hundreds of businesses or many thousands of people to change their behaviours.57 Government by market is attractive when citizens place a high value on personal choice, when there is no clear consensus about the causes of the problem, or when scientific or technical knowledge of the problem changes so quickly that the Government’s programmes or regulations cannot keep up.

Government by market is a very powerful alternative to traditional modes of policy implementation because it provides rapid adjustment to price signals and allows an unlimited number of individual adaptations towards the achievement of an overall public benefit. High levels of innovation are possible because there are few procedures to follow once the broad market design has been put in place. However, in government by market, transparency in relation to individual transactions is extremely low. It is impossible to monitor what millions of individuals are doing in response to the market, except in the aggregate. Over time, the stocks in a fishery, the amount of water in a catchment, or the level of greenhouse gases in the atmosphere, can be measured. But it is impossible to hold individuals to account for every transaction in the market, and accountability can only be applied to the market as whole, not to its parts.58

The accountability deficit for government by market is similar to the gap in government by network, but more pronounced. While the public expects that services funded from the public purse should meet certain standards of probity, fairness and transparency, privately owned firms operating in a competitive market may apply different standards and values to their operations. They may also be less concerned about overall social impacts than a democratically elected government would be.

Commentators have pointed to the loss of public accountability for the way in which water is allocated in the transition to a national market for water trading in Australia. Water is seen by many Australians as part of the commons that should be available to all, with access managed by the Government to ensure fair access and sustainable use. One rural landholder has argued:

‘This process should be controlled by, and answerable to, the democratic process.’59 The introduction of a national water market may help in moving water to higher value uses, and ensure more efficient use of a limited resource, but this may be at the cost of a loss of accountability for the way in which water is used in particular markets and river systems.

In government by market, it is impossible for the Government to prescribe behaviours separately, or to monitor the performance of every transaction within a market. The market will adjust quickly to take advantage of any opportunities and reward high performers, even if this involves windfall gains for individual firms arising from imperfections in the design of the market. Concerns about water harvesting practices, or reactions to the windfall gains by some large European firms in the first round of the European Union Emissions Trading Scheme (EU ETS), however, suggest that public support for government by market can be eroded by poorly-designed markets that allow or encourage behaviour which falls short of public expectations about what is fair or appropriate.

Case study

Government through markets—Emissions trading in Europe

Emissions trading schemes are an example of government by market.

Theoretically, governments could seek to reduce greenhouse emissions in a number of ways—through public awareness and publicity campaigns, investment in more efficient urban design or transport, programmes and tax incentives to convert to greener technologies, a carbon tax, the prohibition of certain kinds of economic activity, or via market mechanisms. The Kyoto Protocol included commitments to emissions trading.

In cap and trade schemes governments establish a legal framework within which businesses can buy and sell emissions credits. This creates a market with incentives for businesses to find the cheapest and most efficient ways of reducing emissions.

The European Union Emissions Trading Scheme (EU ETS) was introduced in 2005. Experience with the EU ETS shows that poor market design or unexpected gaming by market participants can undermine the original policy intentions and even make things worse.

The original decision by EU governments to give the EU ETS allowances away rather than auction them made the scheme easier for the power generators and other polluters to accept. But it also handed them lots of cash: they simply passed the extra costs on to consumers and pocketed the money. According to a report by the British firm, IPA Energy Consulting, Britain’s power-generation sector alone made a profit of around £800 million ($AUD1.5 billion) from the scheme in its first year.

Critics have argued that the scheme has shifted the production of damaging gases to the developing world. HFC-23 is one of the most potent greenhouse gases. Its global warming effect is 11,700 times greater than that of carbon dioxide, but it is cheap to capture—burning it off costs less than €1 for the equivalent of one tonne of carbon dioxide.

The very cheapness of cutting emissions of HFC-23 makes the trade controversial.

Credits costing less than €1 to produce have been sold on the market for up to €11.The returns on emission trades create incentives for Chinese factories to continue to produce HFC-23 and mean that European companies find it more profitable to buy credits from the destruction of HFC-23 rather than reduce their own coal sourced emissions.

The Chinese Government now applies a 65% tax on emissions reduction credits. As a consequence, European companies are paying many times the actual cost of reducing emissions. The price they pay is passed on to European consumers.

There are perverse incentives in the scheme for companies in developing countries to understate the greenhouse reductions that can be achieved for a dollar of investment and for European companies to overstate the greenhouse savings from the credits that they purchase. Individual governments also seek to manipulate the system for their own advantage. This means that the scheme needs to be underpinned by expensive monitoring arrangements that generate considerable red tape and new areas of government activity.

Despite the existence of the EU ETS, European emissions overall do not seem to be falling, which suggests that there may not be as much switching out of coal, or as much technological innovation, as had been hoped. It is possible that Chinese credits are too cheap and the carbon price too low and too volatile to create the right incentives to reduce European emissions.

Government by market does not mean that governments are no longer accountable for results, or that they are removed from the system. European Governments continue to refine the EU ETS and have reduced planned allocations to industry for the period 2008–2012.60

Public servants can be held accountable for the quality of the market design, but the results of system-wide evaluations are usually only available some time after the event, often after all the officials responsible for the market design have left their positions, and in highly- complex systems the evaluation may be ambiguous or uncertain because of unintended consequences or secondary effects in the system.

While government by market is a very effective technique for addressing some kinds of complex problems, there is potential for significant gaps to emerge between community expectations about the underlying ethos or values that will be applied by participants in the market and the commercial pressures and history of those organisations.

Delivering accountability and performance in contemporary public policy

Fit for purpose

There is no single solution or methodology for dealing with the complexity of the new and emerging issues confronting governments; similarly, no one accountability or performance management approach could accommodate all new modes of public policy implementation.

The OECD has suggested that, in view of the different functions performed by government and the diversity of government programmes, consideration should be given to adopting an approach to performance management that is flexible enough to allow for a range of methods in ensuring accountability and evaluating performance.61

The US Government Accounting Office made a similar point in seeking to caution Congress about the consequences of imposing a uniform regime of performance reporting on federal grants programmes that had been designed to incorporate varying levels of flexibility and accountability for performance.62

The differences between programmes can be significant, even within a single mode of policy implementation. Research and development programmes often involve multi-year grants or financial contributions, support for untested procedures, and the use of peer review processes to determine which applications will be funded. Though research programmes are funded from the same budgetary source as programmes for the direct supply of a monopoly service such as passport administration, the programmes incorporate different levels of uncertainty, and it makes sense to apply different kinds of performance management.

The US Government eventually modified the Government Performance and Results Act of 1993 in order to differentiate between different kinds of programme.

The Australian accountability and performance management frameworks are flexible enough to accommodate similar fit for purpose arrangements.

Case study

Adaptive harvest management of waterfowl

In natural resource management, the traditional, sequential approach to decision- making has been beset with difficulties arising from disagreement about appropriate management objectives and uncertainty surrounding the effect of management actions.

Structured decision-making processes and adaptive management are useful approaches to dealing with these difficulties.

In 1995, the US Fish and Wildlife Service first implemented an adaptive approach to the management of the mallard duck population. This approach was developed in collaboration with stakeholders, including scientists, environmentalists, hunters and policy makers, and has been followed successfully for over a decade.

The elements of this process include an agreement of objectives for the forthcoming year, a set of available management actions, models of system response to management actions, and a monitoring programme to estimate system state and other relevant variables.

In late summer of each year, a decision is made about the hunting regulations to be implemented in the upcoming hunting season. In the following spring, mallard population size and wetland numbers are again estimated, and the population estimate used to estimate new model weights.

Stakeholders may wish periodically to modify or develop new programme objectives, as perspectives and values evolve over the course of time. Scientists and managers may also conclude that existing models should be modified or new models developed as learning progresses.

The adoption of a formal adaptive management approach to dealing with the harvest of mid-continent mallards has been successful enough for the arrangements to have survived for well beyond a decade. Mallard populations have fluctuated between about six and 12 million birds over the decade. Management decisions have been transparent and disagreements have shifted from topics that conflated ecological uncertainty, partial controllability, and objectives, to a clear focus on objectives and management alternatives.

A number of important lessons have been learned from this experience. One lesson is that where uncertainty exists, eliciting and balancing stakeholder values and opinions is an important element of the management process. While many management strategies may have achieved biologically sustainable results, an optimal strategy required a social process reflecting stakeholder input and social values. Experience with this programme also shows that periodic renewal of strategies outside the normal process of technical adaptation is necessary in order to maintain stakeholder and political support.

One surprising lesson was that the management uncertainty was so great that efforts to fine tune strategies for the purpose of exploiting additional small sources of variation in harvest potential were not especially effective. One implication of this is that changes to objectives, models and management alternatives should be considered over long rather than short time periods.

Adaptive management was developed in this case as a way of dealing with policy differences over objectives and the scientific uncertainty associated with any natural system. The approach has been successful and appears preferable in conditions of uncertainty to the traditional hierarchical and sequential approach to policy implementation.63

Outcomes-based performance management as it has been applied seems to work only fora limited sub-set of the modes of policy implementation. For government by network and government by market it may be more useful to look at performance information as contour lines on a map, giving a guide to where a traveller stands and possible future paths, rather than as milestones and targets on the way to a fixed destination.

In some cases, comparative information on performance, choice of suppliers, and flexibility as to how things are done may provide more relevant forms of accountability to citizens and stronger incentives for managers to find better ways of delivering services than detailed reporting on outputs and outcomes.

Leadership and choices

The ways in which leaders influence communities in defining problems are central to how accountability and performance management are applied to public policy problems.

Leaders have opportunities to define the nature of problems and help shape the form of policy response that follows.64 Public sector leaders may need to recognise that there are no ready answers to a problem and be prepared to open processes up to collaborative decision-making and adaptive learning.

This could include, for example, exempting adaptive projects from some of the usual performance assessment rules, instituting reporting arrangements that are better suited to conditions of uncertainty and complexity, such as longer reporting cycles, and reporting on lessons learnt, rather than on results achieved.

However, in a democratic and transparent political system subject to intense external scrutiny, there are great pressures on decision-makers to avoid uncertainties and act as though problems can be solved by applying known and proven techniques.65

New modes of policy implementation offer greater flexibility and scope for innovation, and for this reason are attractive in some cases. However, governments, Parliament (including and perhaps especially through the Senate Estimates process), the media and the public retain high expectations of public servants’ accountability for what is achieved and how it is achieved.

A more variegated model of accountability requires decision-makers to acknowledge uncertainty where they see it and to move into a more uncertain decision-making mode. That is often difficult enough. But leaders also need to manage the expectations of others, and persuade them that they are following the right course. Keith Grint argues that the critical test of leadership is not the ability of decision-makers to understand the situation confronting them, but their success in applying persuasive mechanisms to render situations more tractable and compliant to the way they believe problems ought to be addressed.66

In practice, this often means becoming an active participant in the way that problems are perceived and defined, and helping to shape expectations about acceptable tolerances for inconsistencies, mistakes and failures in how complicated problems are tackled.

Risk management is one tool that might help decision-makers as they move into more complex forms of policy implementation. The new modes of policy implementation are riskier than traditional bureaucratic approaches because they involve experimentation and greater collaboration with parties that have incentives to challenge standard procedures. But while it may not be possible (or desirable) in the new modes of implementation to control the way things are done, it is possible to manage and mitigate the risks.

Some agencies are developing more systematic approaches to risk management to help them deal with increasing complexity in the delivery of programmes. One Portfolio Secretary has emphasised the importance of having a rigorous project management framework in place under which people are held accountable and management kept informed. He has said that by this he does not mean one-size-fits-all approaches—‘… for complex projects we need complex risk management and project design. [For] others less complex.’67 In the right management environment, and with the right kind of leadership, some of the risks surrounding the complexity and uncertainty of new modes of implementation can be mitigated.

It takes courage, perseverance and skills, but good leaders can make choices that support innovation around new modes of policy implementation.

New models of accountability

Accountability and performance management arrangements need to be taken into account in designing policy, because accountability and performance management arrangements influence how problems are understood and addressed.

Elaine Kamarck suggests that the different modes of policy implementation, or the way that governments choose to do things, and the accountability and performance regimes that accompany them, should be determined by the nature of the policy goals that the Government wants to pursue. She says: ‘It is well established that certain kinds of policy goals require certain kinds of implementation approaches.’68 If governments want consistency and security, they will choose a department or agency and apply a traditional accountability framework. If they want innovation and fit for purpose solutions to a known problem they will choose a network, and so on.

The following table attempts to relate policy goals to the new modes of policy implementation and the accountability challenges that they present. The table sets out policy objectives in terms of levels of innovation and consistency of product and then provides comment on the design of accountability and performance mechanisms for each of the modes of policy implementation as identified by Kamarck. Whole of government falls mainly within performance managed bureaucracy, but it can also include elements of government by network, depending on the extent to which NGOs are involved in designing policy or delivering services.

There may well be issues that require responses which will utilise all three of Kamarck’s models for intervention at the same time. Government responses to climate change, for example, include elements of performance managed bureaucracy, government by network and government by market.

Table 1: A More Diversified Approach to Accountability and Performance Management69
Performance managed bureaucracy Government by network Government by market
Innovation Low Medium to High High
Consistency High Medium to Low (but can be influenced by the form of contracting) Low (but is a function of market design and regulatory arrangements)
Accountability mechanisms

Performance measures and objectives replacing some rules and regulations

Compliance with process

Overall goals for the network

Individual performance management of third party providers

Shared performance for outcomes



Performance measured in the aggregate and only after some time has passed

Individual performance is not relevant


Management goals

Trade-offs between compliance and performance

Managing inputs and outputs in a way that ensures efficiency and effectiveness

Understanding causality in the network and managing the parts around a goal for the entire network

Defining successful network performance

Getting the right incentives

Preventing cheating and gaming

Setting process correctly and adjusting as necessary

Ensuring information symmetry

Preferred mode for accountability Performance management and reporting

System evaluation

Periodic reviews of overall outcomes

Performance management of third party providers

Infrequent systemic review and evaluation
Major challenges

Improved specification of outcomes

Clearer and simpler documentation

Recognition that different programmes may benefit from different performance and reporting arrangements

Providing for citizen engagement

Building capability for stakeholders and public servants managing network relationships

Building bottom-up accountability

Sharing lessons

Addressing accountability deficit in relation to how things are done

Market design

Addressing accountability deficits around process and fairness

Accountability to the public for process and system outcomes

Actions required Executive action by central and line agencies

Executive action to allow adaptive approaches to goal setting and resource allocation

May require legislative change to accommodate dual accountabilities

Governments need to be willing to evaluate outcomes and redefine property rights or re-enter the market to secure public policy objectives

This model does not require a wholesale replacement of the existing accountability and performance management frameworks, but a modification of the frameworks to take account of the decision-making context and the preferred mode of policy implementation. Most of the proposed changes could be made by executive action. However, in some cases it would be necessary to support the changes with specific legislation, for example, when clarifying the nature of horizontal accountabilities or exempting officials from the application of existing accountability requirements.

The new modes of policy implementation make significant demands on the skills and capacity of public servants. Collaborative governance is difficult for a highly-trained specialist with technical knowledge not available to the people he or she may be working with. Public servants trained to operate within a hierarchical system of accountability and a simplified model of problem solving may need assistance in developing the skills required to manage group decision-making in a highly uncertain environment.

A continuing investment needs to be made in providing practical support for public servants dealing with intractable problems. New modes of public policy implementation require capabilities in problem framing and boundary setting, the ability to generate fresh thinking on intractable problems, methods for working across organisational and disciplinary boundaries, and techniques for effective decision-making in situations characterised by high levels of uncertainty. Public servants will need to tolerate rapid change in the way problems are defined and to engage stakeholders as joint decision-makers rather than as providers or recipients of services.

Not all public servants will need to work in these ways all the time, and some staff may not be affected at all, but many public servants will be confronted by ambiguous and complex problems at some point in their career, and others will be required to support or assist agencies struggling to find a way through new problems and develop innovative ways of working. It is important that the senior levels of the public service are able to exercise the kinds of leadership that these problems require.

Opportunities for action

This paper argues for a gradual, case-by-case approach in applying fit for purpose accountability and performance management arrangements to specific programmes and policy initiatives.

The solution does not lie in no longer requiring officials to be accountable for their actions, even for the most complex and ambiguous problems, but in looking for different, less traditional, ways of assessing performance and adjusting expectations of the way in which performance information should be interpreted and applied.

A number of agencies are already experimenting with more flexible approaches to accountability and performance management while operating within the existing legal and administrative frameworks.

Case study

Use of qualitative information on performance

The Commonwealth Ombudsman’s Office is seeking to supplement traditional accountability and performance information in its Annual Report by also including case studies.

The case studies provide a short narrative that sets out the circumstances of particular cases, how the Ombudsman’s Office dealt with the matter and the eventual outcome.

The case studies complement traditional forms of activity and outcome reporting.

They provide additional information that allows the Parliament and members of the public to understand how the Ombudsman’s Office operates and the ways in which complaints can be resolved.

Traditional quantitative reporting of complaint handling can be ambiguous because a reduction in complaints could indicate either a level of dissatisfaction with how complaints are handled, or an improvement in the underlying processes as a result of successful interventions in the past.

Stories provide a powerful means of communicating complex ideas because they connect with the way that human beings process information. Stories are easier to remember than raw numbers and allow a more nuanced message to be communicated.

The Ombudsman’s report does not attempt to generalise from the case studies or make wider claims about their relevance.

The case studies are not intended to provide a complete or comprehensive account of the work of the Ombudsman. Like all information, they give a partial account of a complex reality. However, they do bring out points of interest and provide a starting point for further questions about the work of the Office that complements existing quantitative performance information.70

The Rudd Government has expressed an interest in finding new ways of delivering on its social reforms. It has implied that accountability and performance management arrangements need to be looked at as part of the development of new approaches. Addressing the question of Indigenous disadvantage, the Prime Minister has stated that the time has come for new approaches to enduring problems:

A new beginning which contains real measures of policy success or policy failure. A new beginning, a new partnership, on closing the gap with sufficient flexibility not to insist on a one-size-fits-all approach for each of the hundreds of remote and regional Indigenous communities across the country but instead allows flexible, tailored, local approaches to achieve commonly-agreed national objectives that lie at the core of our proposed new partnership.71

The approach outlined in the Prime Minister’s statement retains a whole of government flavour but also includes elements of collaborative government and traditional top-down performance management. There is a very heavy emphasis on problem definition and targeting; Mr Rudd stresses the need to work on ‘clear goals, clear thinking’ as well as ‘defined projects’ and refers to ‘real measures of policy success or policy failure’. The Prime Minister has promised to report to Parliament on the first sitting day of every year on the Government’s efforts to improve the living conditions, health and education of Aboriginal people.

The combination of clear leadership, tightly-managed performance and whole of government seems to be in line with international thinking on collaborative government. The approach mirrors those of the United Kingdom and other OECD countries that have adopted broad ‘societal indicators’ as a basis for aligning government effort with larger societal trends.

A Treasury Board of Canada Secretariat paper on collaborative government argues that horizontal commitments to common societal outcomes and performance measures improve the alignment of effort across agencies and avoid many of the costs usually associated with attempts to coordinate activities across multiple jurisdictions.72 The challenge in participative governance is to find ways of ensuring that the parties agree on the broad measures to be adopted and how they are to be measured, even if they remain separately accountable through their own institutions for the way in which they have delivered on those commitments.

Case study

Collaborative partnerships

Collaborative partnerships are an example of government by network.

Experiments with collaborative partnering in the United Kingdom and Canada suggest that a strong accountability framework is a critical requirement for effective partnerships across different types of collaborative arrangement. Collective accountability to the partnership governance structure needs to be addressed as well as individual accountability of each participant to its own governing body. Elements often considered to constitute effective accountability include agreed performance expectations, clear roles and responsibilities, balanced expectations and capacities, credible reporting and reasonable review, evaluation and audit.

There are many forms of collaborative partnership, but they typically involve agreement about the results that are to be achieved. Canada’s Centre for Collaborative Government says that in collaborative partnerships, the Government and the partner(s) negotiate a framework of outcomes, principles, objectives and indicators.

The partner(s) then commit to having their performance evaluated against the agreed indicators. At the same time, they acquire some flexibility regarding the design and delivery of services, and possibly some lower-level policy making capacity.

The flexibility and elements of joint responsibility in the collaborative partnership often involve the delegation of some decision-making authority to non-government actors. How much depends on the agreement and the depth of the collaborative partnership. In practice, a clear distinction between consultation and collaborative partnerships or citizen engagement may be hard to draw.

The OECD has endorsed some basic principles for active citizen engagement set out by Canada’s Institute on Governance in 1998. These include: shared agenda-setting; a relaxed time frame for deliberation; an emphasis on value sharing; and consultative practices based on inclusiveness, courtesy and respect.73

The existing accountability arrangements have many advantages. They remain appropriate where governments value consistency or want to avoid risk. Even where it may not be possible to set precise outcomes in advance for a specific policy intervention, programme objectives have value in providing information on the Government’s purpose, and a general sense of the direction in which the Government is heading.

The Australian accountability framework has flexibilities that are not fully exploited. But there is a range of evidence, including from ANAO, researchers outside government, and surveys of public servants conducted by the Australian Public Service Commission, that the current accountability and performance management arrangements aren’t always supporting new ways of working.

ANAO recommended in its review of the whole of government Indigenous service delivery trials that the concept of cross-agency outcomes be reviewed in light of its lack of application by agencies to date and that clear guidance be developed for departments as to how best to engage in joint funding of Indigenous services.74

In responding to this report, Finance and ANAO have undertaken separately to prepare guidance material and revise aspects of existing arrangements.

Finance has also sought to clarify arrangements in two pieces of recent legislation. In the Explanatory Memorandum to the Financial Framework Legislation Amendment Bill 2008, Finance said that:

52. The reference in the note to a Chief Executive being able to enter contracts ‘in relation to the affairs of the Agency’ is to be read in broad terms. In particular, the Government will generally expect agencies to work cooperatively in a range of areas, including the implementation of whole-of-Government policies.75

The Explanatory Memorandum sets out a number of examples where the current arrangements provide flexibilities for agencies to cooperate on whole of government exercises. These include the opportunity for one agency to enter into a contract on behalf of the Commonwealth, where the services can be accessed by other agencies. Finance also suggests that agencies may decide cooperatively to share arrangements, such as through a request for tender that allows for the inclusion of other agencies.

These comments address some of the issues raised in the ANAO report on the whole of government Indigenous trials and should go some way towards reassuring officials that under certain circumstances the accountability framework allows them to cover costs that might fall within other areas of government, or to devise common approaches to procurement and tendering when implementing whole of government policies.

It would be helpful if these comments could be given wider coverage and further practical explanation. It would also be useful to know, for example, whether the statement that departments involved in these transactions might later be reimbursed for their services is a suggestion or a mandatory requirement, and whether the failure to do so would attract adverse comments by ANAO in subsequent audits or performance reviews. More advice on the question of reporting against shared outcomes would be useful as well.

ANAO’s response to its own recommendations for revising the existing accountability framework in order to facilitate whole of government arrangements is not yet clear.

The Government has established Operation Sunlight to improve the transparency and accountability of Commonwealth public funding and expenditure.76 The Minister for Finance and Deregulation, Hon. Lindsay Tanner MP, has stated: ‘Reporting on outcomes is inadequate and some outcomes are so broad and general as to be virtually meaningless for budget accounting purposes.’77

Changes were made to the presentation of the 2008–2009 Budget Papers, including enhanced programme (or sub-functional) information in Statement 6 and a more strategic focus in the Portfolio Budget Statements (PBS). The PBS now includes a Strategic Direction statement, which provides the reader with a brief overview of the agency and its strategic intent, the strategies to be implemented, and the goals and results to be pursued during the Budget cycle.

The Minister for Finance and Deregulation has also announced a review by his department of all outcomes, which is due to be completed and ready for consideration in the 2009–2010 Budget. This review will seek to improve the specificity of outcomes and their consistency across government.78

It is important to ensure that Budget decisions and agency funding are reported in a transparent way. This review could provide an opportunity to take up the Prime Minister’s challenge to reflect on whether a one-size-fits-all performance management framework continues to produce the right balance between accountability and innovation for all programmes, and whether there might be scope for a slightly more differentiated approach in relation to some activities.

The idea of earned autonomy has some currency in the United Kingdom.79 This involves the relaxation of vertical accountabilities as the practitioners of new modes of policy implementation demonstrate enhanced capacity—in terms of improved performance—in tackling problems and ensuring greater transparency in decision-making processes.

The authors of a paper for the UK Cabinet Office argued against accountability arrangements that are overly reliant on process driven inspection and audit regimes. They advocated a more rounded accountability, which faces outwards towards users and citizens, as much as upwards.80 An accountability framework that includes reporting on the openness of processes, the levels of citizen engagement, the quality of participation, and levels of satisfaction, might supplement and add meaning to reporting on the outputs produced or the outcomes that the Government and its advisers believe may have been achieved.

The Minister for Education, Hon. Julia Gillard MP, has championed the use of benchmarking for school performance as a way of providing more useful information to parents and communities.81 Calling on the states and territories to agree to greater transparency of school results and features, Ms Gillard has been reported as saying:

Giving full information to people would mean that they can actually know what’s going on and, rather than judging individual schools or school systems on the basis of myths, rumour, prejudice or perception, people would have the facts.82

Benchmarking and evaluation are useful because they can provide information on performance in forms that are relevant to the citizens who experience the services provided by government.

Benchmarking provides snapshots of outcomes at points in time only. It is not able to explain why outcomes are achieved, but it can help in identifying issues for further consideration. Evaluations are a more powerful tool for understanding the effects of government programmes, though they often involve methodological and conceptual complexities that make them expensive to undertake and difficult to apply to practical decision-making.83

Case study

Evaluation of International Development Assistance

This year, the budget of the Australian Agency for International Development (AusAID) will rise from $3.2 billion to $3.7 billion. In a May 2008 article in the Australian Financial Review, Andrew Leigh asked the question ‘Does aid work?’

Leigh noted that at an aggregate level it is extremely difficult to identify the impact of foreign aid on a country’s level of development.

During the 1990s, a series of studies reached ambiguous conclusions. In a well-known paper published in 2000, Craig Burnside and David Dollar found that aid enhanced growth, as long as the recipient country had sound fiscal, monetary and trade policies.

Three years later, William Easterly, Ross Levine and David Roodman questioned these results, and argued that they did not hold up when more years of data were added to the analysis. Subsequent research by Paul Collier indicates that even in poorly-governed parts of Africa, poverty rates would today be much higher had the continent received no aid.

Most recently, Jeffrey Sachs has asserted that, rather than discriminating against countries with badly-performing institutions, richer nations should ensure that they don’t provide the wrong kinds of aid.

AusAID is taking the question of aid effectiveness seriously. In its first annual report, the Office of Development Effectiveness (ODE), located within AusAID, carried out what it called a ‘health check of the entire Australian aid program’, in which it critiqued aspects of the programme in surprisingly robust terms.84 ODE pointed out that Australia spends twice as much as other rich donors on technical assistance, and warned that some of our technical assistance programmes may even be undermining the capacity of poor countries to govern themselves.

The ODE report emphasised the importance of carrying out more impact evaluations—asking not only whether the programme was properly administered, but also whether it had helped to improve the lives of the poor.

In his article, Andrew Leigh argued that this approach is particularly valuable—given that overseas aid is an area where too many advocates have focused on inputs (aid as a percentage of national income) rather than outputs (poverty reduction).85

These forms of accountability, such as benchmarking and evaluation, are better suited to services that are delivered through networks and markets compared to traditional outcomes/ outputs reporting. This is because they are open to the rapid adaptation and adjustment that characterise new modes of policy implementation. They are also useful in circumstances where there are high levels of complexity or uncertainty about the impacts of policy interventions. While the outcomes/outputs framework requires a clear causal link between the intervention and what is achieved, benchmarking and evaluation can be applied to outcomes or outputs, without necessarily assuming anything about causality.

The greatest challenge and perhaps most radical way of reforming accountability arrangements to support new ways of working would be the transition to an accountability framework that acknowledges the pressures for shared decision-making power and funding authority through the development of collective accountability for joint governance.

Catherine Allan, Meredith Edwards and ANAO have each observed in the case studies provided with this paper that attempts to deliver policies in collaborative ways can default to traditional ways of working when public servants involved in these arrangements are confronted by the demands of the current accountability and performance management arrangements.

To create the environment in which officials are able to perform adequately in the joint delivery of services and also to operate safely within the law, it may be necessary to develop an effective legislative framework acknowledging joint accountability. The simplest and safest way to do this would be to devise specific legislation acknowledging joint accountability for some programmes.

Governments and Parliaments understandably have some reservations about any move which would mean that the normal lines of accountability could not be traced to a single source. But joint accountability doesn’t necessarily mean that there is no ultimate point of accountability. As explained earlier in this paper, a theoretical form of joint accountability already exists in the current financial framework, in the form of third party drawing rights. There is scope for more guidance material from central agencies to help public servants understand how this could work.

However, other forms of joint accountability are currently possible within the existing accountability framework.

One common form of joint accountability involves the creation of an intermediate jointly-owned entity to receive Australian Government funds, and funds from other sources, with funding to the entity and disbursement from the entity subject to different legal and community accountabilities.

Case study

Canada’s Urban Aboriginal Strategy

Canada’s Urban Aboriginal Strategy (UAS) represented a response to frustration with fragmented service delivery through multiple agencies and programmes. The UAS involved community groups as well as provincial, municipal and federal governments in integrating the narrower objectives from a range of separate programmes within a single, overarching policy goal or ‘mission’, which was to ‘close the gap in life chances of urban Aboriginal people and the mainstream population’.

This overarching objective allowed stakeholders to view previously unconnected programme objectives, such as those aimed at homelessness, illiteracy or substance abuse, as converging on a common goal.

In an effort to break through the policy silos, the UAS agencies then used the overarching goal as a basis for developing a set of ‘horizontal’ terms and conditions to replace the existing terms and conditions of the separate programmes. The UAS treats these experiments as pilot programmes and establishes them through a Memorandum of Understanding (MOU) with each participating agency or jurisdiction.

In essence, the MOU defines a collaborative arrangement in which departments, provinces, municipalities, and community organisations agree to work together. It outlines their various responsibilities and sets out a funding framework for the pilot.

As a rule, each department contributes to a fund by pooling a portion of its urban Aboriginal programme budget. The MOU also establishes a role for community-level representatives in developing the pilot.

The MOU provides a form of shared decision-making while addressing concerns over lost or blurred accountability. It does this through three steps:

  1. the partnership council as a whole is assessed for the contribution it makes to the overall objective of ‘closing the gap in life chances’
  2. the partners must set specific programme-like objectives and performance measures for the pilot so that the partnership council can be assessed for the effective management of its resources
  3. each Minister whose department is involved is held accountable for the quality of the collaboration agreement, including its suitability to achieve the agreed objectives.

Under this model, a Minister whose department is participating in a pilot would still be able to account for his or her programmes in much the same terms as before.

Managers, for example, are still required to report on specific programme objectives and targets have to be met. The difference is that the specific programme objectives and targets are arrived at through negotiation between the partners, who must then agree to be responsible for them.

Though they provide a form of joint accountability, the new arrangements do not appear to diminish traditional lines of accountability. Advocates of this approach argue that the model strengthens accountability by virtue of the additional requirement that each pilot must report on its progress towards the overarching goal.86

The Prime Minister has stated: ‘Policy innovation … is at the heart of being a reformist government.’87 This paper argues for reform and adaptation of the existing accountability and performance management frameworks in order to produce more flexible and contemporary arrangements that provide better accountability to the public and more support for the public servants who are seeking to take up the Prime Minister’s call for policy innovation.

In the meantime, there are some potential paths to flexibility within the current accountability and performance management frameworks that can improve the public service’s capacity to deliver better services to Australians.


The current accountability and performance management arrangements were designed in a different era to support forms of policy implementation that addressed the needs of the time.

They are still suitable for many of the issues dealt with by governments, especially those problems where it is clear what action should be taken, what the effects of that action will be, or where the Government prefers consistency and uniformity over innovation and creativity.

The arrangements as they stand, however, are putting too many obstacles in the way of innovation. They have resulted, for example, in accountability gaps that are reducing public support for new ways of working. This is forcing public servants back into traditional problem solving techniques, when more adaptive and inclusive approaches are required. In addition, such accountability deficits are undermining support for new modes of policy implementation.

The solution does not lie in giving up on accountability and performance management, even for the most complex and ambiguous problems, but in looking for different ways of assessing performance and adjusting expectations about how performance information should be interpreted and applied.

This could be a starting point for sensible refinement of our accountability and performance management arrangements.

1 Hon. Kevin Rudd MP, ‘Apology to Australia’s Indigenous Peoples’, House of Representatives, Parliament House, Canberra, 13 February 2008, p. 3, <http://www.pm.gov.au>

2 G. W. Jones, ‘The Search for Local Accountability’ in S. Leach (ed), Strengthening Local Government in the 1990s, Longman, Harlow, Essex, 1992, p. 73.

3 R. Mulgan, ‘“Accountability”: An Ever-Expanding Concept?’, Public Administration, Vol. 78, No. 3, 2000, p. 556.

4 Management Advisory Board/Management Improvement Advisory Committee 1993, Accountability in the Commonwealth Public Sector, Australian Government Publishing Service, Canberra, p. 1.

5 R. Mulgan, ‘“Accountability”: An Ever-Expanding Concept?’, Public Administration, Vol. 78, No. 3, 2000, p. 555.

6 D. Chalmers et al 2006, European Union Law: Text and Materials, Cambridge University Press, Cambridge, p. 312.

7 A. Schick, ‘The Performing State: Reflection On An Idea Whose Time Has Come But Whose Implementation Has Not’ (Paper prepared for the 33rd Session of the Public Governance Committee, Public Governance and Territorial Development Directorate, OECD, Paris, 6–7 April 2006), p. 12, <http://www.oecd.org>

8 W.T. Gormley, Jr. & S. J. Balla 2004, Bureaucracy and Democracy: Accountability and Performance, Washington, D.C., CQ Press, pp. 14–15.

9 Department of Finance and Administration 2000, The Outcomes and Outputs Framework Guidance Document, p. 5, <http://www.finance.gov.au>

10 Management Advisory Board/Management Improvement Advisory Committee 1993, Accountability in the Commonwealth Public Sector, Australian Government Publishing Service, Canberra, p. 4.

11 Management Advisory Board/Management Improvement Advisory Committee 1993, Accountability in the Commonwealth Public Sector, Australian Government Publishing Service, Canberra, p. 25.

12 R. Mulgan, ‘“Accountability”: An Ever-Expanding Concept?’, Public Administration, Vol. 78, No. 3, 2000, p. 568.

13 For a more extensive treatment of this idea, see F. Anechiarico and J. Jacobs 1996, The Pursuit of Absolute Integrity: How Corruption Control Makes Government Ineffective, University of Chicago Press, Chicago, especially, pp. 174–176 and R. D. Behn 2001, Rethinking Democratic Accountability, Brookings Institution Press, Washington, D.C.

14 J. W. Fesler & D. F. Kettl 1991, The Politics of the Administrative Process, Chatham House Publishers, Chatham, NJ, p. 321.

15 B. A. Radin 2006, Challenging the Performance Movement: Accountability, Complexity and Democratic Values, Georgetown University Press, Washington, D.C., pp. 44–45.

16 G. Mulgan, ‘360 Degree Improvement and the Imperative of Social Innovation’ (Address to the UK Public Services Reform Conference, 6 June 2006, National School of Government), p. 5.

17 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, pp. 66–67.

18 P. G. Thomas 2006, Performance Measurement, Reporting, Obstacles and Accountability: Recent Trends and Future Directions, ANU E Press, Canberra, p. 61.

19 C. Allan, ‘Can Adaptive Management Help Us Embrace the Murray-Darling Basin’s Wicked Problems?’ in C. Pahl-Wostl, P. Kabat & J. Moltgen (eds), Adaptive and Integrated Water Management: Coping with Complexity and Uncertainty, Springer, Berlin, 2008, pp. 61–75.

20 Hon. Kevin Rudd MP, ‘Apology to Australia’s Indigenous Peoples’, House of Representatives, Parliament House, Canberra, 13 February 2008, p. 3, <http://www.pm.gov.au>

21 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, pp. 16–18.

22 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, p. 72.

23 OECD 2005, Modernising Government: The Way Forward, OECD, Paris, p. 118.

24 R. Mulgan, ‘Accountability Issues in the New Model of Governance’ (Paper for Political Science Program Seminar, Graduate Program in Public Policy, ANU, 3 April 2002), p. 21.

25 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, p. 28.

26 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, p. 64.

27 E. Anderson, G. Griffin & J. Teicher, New Public Management and Industrial Relations in the Australian Public Service: An Overview, Department of Management, Faculty of Business and Economics, Monash University, Working Paper No. 32/03, May 2003, p. 2.

28 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder CO, pp. 28–29, 68.

29 ANAO/CPA Australia, Monitoring and Reporting Financial and Non-Financial Performance of Australian Government Organisations, Issues Paper Series No. 5, February 2008, especially, pp. 20–21. A number of writers have questioned the extent to which the performance of government can be measured effectively, even in relation to comparatively simple functions.

  • See, for example, H. Ergas, ‘Reforms Could Hit Target, Miss Point’, The Australian, 3 September 2008, p. 12; B. A. Radin 2006, Challenging the Performance Movement: Accountability, Complexity and Democratic Values, Georgetown University Press, Washington, D. C. Other writers have pointed to the unintended consequences of a strict adherence to performance regimes.
  • See, for example, L. Kerr, E. Carson & J. Goddard, ‘Contractualism, Employment Services and Mature-Age Job Seekers: The Tyranny of Tangible Outcomes’, The Drawing Board: An Australian Review of Public Affairs, Vol. 3, No. 2, November 2002, pp. 89–91 and S. Jenkins, ‘Brokenshire’s Bathboard: Review of D. Campbell-Smith, Follow the Money: The Audit Commission, Public Money and the Management of Public Services, 1983–2008 (2008)’, Times Literary Supplement, 4 April 2008, p. 7.

30 ANAO 2007, Whole of Government Indigenous Service Delivery Arrangements, Performance Audit Report No. 10, 2007–08, Commonwealth of Australia, Canberra, pp. 61, 107, <http://www.anao.gov.au/>

31 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, p. 64.

32 Management Advisory Committee 2004, Connecting Government: Whole of Government Responses to Australia’s Priority Challenges, Commonwealth of Australia, Canberra, pp. 1–5

33 Australian Public Service Commission 2007, State of the Service Report 2006–07, Commonwealth of Australia, Canberra, p. 232

34 ANAO 2007, Whole of Government Indigenous Service Delivery Arrangements, Performance Audit Report No. 10, 2007–08, Commonwealth of Australia, Canberra, p. 105, <http://www.anao.gov.au>

35 ANAO 2007, Whole of Government Indigenous Service Delivery Arrangements, Performance Audit Report No. 10, 2007–08, Commonwealth of Australia, Canberra, pp. 11–38, 105–108, <http:/www.anao.gov.au>

36 Management Advisory Committee 2004, Connecting Government: Whole of Government Responses to Australia’s Priority Challenges, Commonwealth of Australia, Canberra, p. 53. The report also identifies the importance to whole of government work of leadership, organisational agility, resolving cultural differences, supportive structures, appropriate capabilities, and workplace support (See pp. 43–57).

37 The need to develop capacity in these areas was identified in the Australian Public Service Commissioner’s State of the Service Report 2005–06 (p. 210) and the Management Advisory Committee’s report, Connecting Government (2004) (pp. 14–15).

38 T. Ling, ‘Delivering Joined Up Government in the UK: Dimensions, Issues and Problems’, Public Administration, Vol. 80, No. 4, 2002, p. 622.

39 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, p. 103.

40 R. Mulgan, ‘Accountability Issues in the New Model of Governance’ (Paper for Political Science Program Seminar, Graduate Program in Public Policy, ANU, 3 April 2002), pp. 12–13.

41 R. Bevan, ‘Cliques and Mortar’, Australian Financial Review, 13–15 June 2008, pp. L19–L21.

42 R. Mulgan, ‘Accountability Issues in the New Model of Governance’ (Paper for Political Science Program Seminar, Graduate Program in Public Policy, ANU, 3 April 2002), p. 18.

43 OECD 2005, Modernising Government: The Way Forward, OECD, Paris, p. 136.

44 M. Rawsthorne, ‘Contracting—The Impact on Non-Government Organisations’, Address to the Social Policy Research Centre, University of New South Wales, 2003, <http://www.sprc.unsw.edu.au>

45 P. Oslington, ‘Contracting-Out of Assistance to the Unemployed: Implications of the Australian Experiment’, Economic Record, Vol. 81, No. 252, March 2005, pp. 30–37.

46 R. Mulgan, ‘Accountability Issues in the New Model of Governance’ (Paper for Political Science Program Seminar, Graduate Program in Public Policy, ANU, 3 April 2002), p. 14.

47 Senator Penny Wong, quoted by P. Karvelas in ‘Change Tack on Jobs: ALP’, The Australian, 23 October 2006, p. 2. See also ALP, ‘Reward for Effort: Meeting the Participation Challenge’ (A Discussion Paper on Australia’s Workforce Participation Issues), released by Senator Penny Wong, November 2006.

48 Hon. Brendan O’Connor MP and Hon. Julia Gillard MP, ‘More Flexible, Better Targeted Employment Services’, Joint Media Release, 13 May 2008.

49 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, p. 37.

50 M. Edwards, ‘Participatory Governance and the Indigenous Agenda: From Rhetoric to Reality’, Centre for Policy Development, 11 March 2008, p. 1, <http://www.cpd.org.au>

51 M. Stewart-Weeks, ‘From Control to Networks’ in H. K. Colebatch (ed.), Beyond the Policy Cycle: The Policy Process in Australia, Allen & Unwin, Crows Nest, NSW, 2006, pp. 184–203.

52 M. Edwards, ‘Participatory Governance and the Indigenous Agenda: From Rhetoric to Reality’, Centre for Policy Development, 11 March 2008, p. 6, <http://www.cpd.org.au>

53 T. Fitzpatrick 2000, Horizontal Management: Trends in Governance and Accountability, Canada School of Public Service, Ottawa, <http://www.csps-efpc.gc.ca>

54 D. G. Lenihan et al, ‘What is Shared Accountability?’, Policy, Politics and Governance, Vol. 5, November 2003, <http://www.ppforum.ca>

55 ANAO 2007, Whole of Government Indigenous Service Delivery Arrangements, Performance Audit Report No. 10, 2007–08, Commonwealth of Australia, Canberra, p. 104, <http://www.anao.gov.au>

56 M. & P. Henderson & Associates 2004, Partnerships for Service Delivery: Review of the Research and Practice Literature (Report to the Department of Communities (Queensland)), p. 7, <http://www.getinvolved.qld.gov.au>

57 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, p. 20.

58 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, pp. 39–40.

59 B. Haigh, ‘Water Trading Plan “Feudal”’, Canberra Times, 28 April 2008, p. 9.

60 The Economist, 7 September 2006, p. 14, 31 May 2007, p. 37, 17 May 2008, pp. 73–74; M. Lockwood, ‘Carbon Trading: A Good Idea But Not Easy to Implement’, Australian Financial Review, 9 February 2007, pp. 6–7; K. Orchison, ‘EU Model Struggling to Shake Off “Failure”Tag’, The Weekend Australian, 11–12 October 2008 (Clean Energy Supplement), p. 1.

61 OECD 2005, Modernising Government: The Way Forward, OECD, Paris, p. 75.

62 B. A. Radin 2006, Challenging the Performance Movement: Accountability, Complexity and Democratic Values, Georgetown University Press, Washington, D.C., pp. 44–45.

63 J. D. Nichols et al, ‘Adaptive Harvest Management of North American Waterfowl Populations: A Brief History and Future Prospects’, Journal of Ornithology, Vol. 148, No. 2, Supplement 2, December 2007, pp. 343–349. See also J. D. Nichols, F. A. Johnson & B. K. Williams, ‘Managing North American Waterfowl in the Face of Uncertainty’, Annual Review of Ecology and Systematics, Vol. 26, November 1995, pp. 177–199.

64 C. Allan, ‘Can Adaptive Management Help Us Embrace the Murray-Darling Basin’s Wicked Problems ?’ in C. Pahl-Wostl, P. Kabat & J. Moltgen (eds), Adaptive and Integrated Water Management: Coping with Complexity and Uncertainty, Springer, Berlin, 2008, p. 70. Allan includes a most useful reference to K. Grint, ‘Problems, Problems, Problems: The Social Construction of “Leadership”’, Human Relations, Vol. 58, No. 11, 2005, pp. 1467–1494. See also S. Bouwhuis, ‘Leadership Harvard Style’, Australian Journal of Public Administration, Vol. 66, No. 4, December 2007, pp. 507–512.

65 K. Grint, ‘Problems, Problems, Problems: The Social Construction of “Leadership”’, Human Relations, Vol. 58, No. 11, 2005, p. 1475.

66 K. Grint, ‘Problems, Problems, Problems: The Social Construction of “Leadership”’, Human Relations, Vol. 58, No. 11, 2005, p. 1492.

67 D. Borthwick, ‘Responding to a Changing Policy Agenda: Achieving Outcomes and Managing Risk’, Keynote Address, Comcover Senior Executive Forum, 12 June 2008, p. [6].

68 E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, p. 15.

69 Adapted from E. C. Kamarck 2007, The End of Government … As We Know It: Making Public Policy Work, Lynne Rienner, Boulder, CO, Chapter 3, ‘Democratic Accountability’, pp. 25–45.

70 Commonwealth Ombudsman 2007, Annual Report 2006–07, Commonwealth of Australia, Canberra, Chapter 5, ‘Challenges in Complaint Handling’ (pp. 38–49) and Chapter 7, ‘Looking at the Agencies’ (pp. 59–113).

71 Hon. Kevin Rudd MP, ‘Apology to Australia’s Indigenous Peoples’, House of Representatives, Parliament House, Canberra, 13 February 2008, p. 3, <http://www.pm.gov.au>

72 L. Bellefeuille-Pregent & T. Wilson 2004, Societal Indicators and Government-wide Reporting in the Government of Canada, Treasury Board of Canada, Ottawa, <http://www.tbs-sct.gc.ca>

73 Australian Public Service Commission 2007, Tackling Wicked Problems: A Public Policy Perspective, Commonwealth of Australia, Canberra; M. & P. Henderson & Associates 2004, Partnerships for Service Delivery: Review of the Research and Practice Literature (Report to the Department of Communities (Queensland)), <http://www.getinvolved.qld.gov.au>; Institute on Governance, ‘A Voice for All: Engaging Canadians for Change’ (Conference on Citizen Engagement, Ottawa, 27–28 October 1998), <http://www.iog.ca>

74 ANAO 2007, Whole of Government Indigenous Service Delivery Arrangements, Performance Audit Report No. 10, 2007–08, Commonwealth of Australia, Canberra, pp. 105–108, <http://www.anao.gov.au>

75 Department of Finance and Deregulation 2008, Explanatory Memorandum to the Financial Framework Legislation Amendment Bill 2008, <http://www.comlaw.gov.au>

76 Hon. Lindsay Tanner MP, Media Release, 13 May 2008, <http://www.financeminister.gov.au>

77 Hon. Lindsay Tanner MP, ‘Address to the CEDA 2008 State of the Nation Conference’, Canberra, 5 June 2008, p. 3, <http://www.financeminister.gov.au>

78 Hon. Lindsay Tanner MP, ‘Address to the CEDA 2008 State of the Nation Conference’, Canberra, 5 June 2008, p. 3, <http://www.financeminister.gov.au>

79 See, for example, S. Lee & R. Woodward, ‘Implementing the Third Way: The Delivery of Public Services under the Blair Government’, Public Money and Management, Vol. 22, No. 4, 2002, pp. 49–56.

80 G. Kelly, S. Muers & G. Mulgan 2002, Creating Public Value: An Analytical Framework for Public Service Reform, UK Cabinet Office, London, <http://www.cabinetoffice.gov.uk>

81 A. Patty, ‘Tell-all Report Cards to Compare Schools’, Sydney Morning Herald, 12 August 2008, pp. 1–2; Hon. Julia Gillard MP, Media Release, ‘Address to the Western Chance Breakfast’, Melbourne, 21 August 2008.

82 J. Ferrari & D. Nason, ‘Gillard Admits School Choice is “Guesswork”’, The Weekend Australian, 23–24 August 2008, pp. 1, 22.

83 C. O’Faircheallaigh, J. Wanna & P. M. Weller 1999, Public Sector Management in Australia: New Challenges, New Directions, 2nd edn, Macmillan Education Australia, South Yarra, Vic., pp. 206–207.

84 Office of Development Effectiveness 2007, Annual Review of Development Effectiveness 2007, Commonwealth of Australia, Canberra, p. 1,

85 A. Leigh, ‘Sorting Good Aid from Bad’, Australian Financial Review,20 May 2008,p.70.The other authors referred to in this Case Study were quoted by Andrew Leigh in his AFR 20/5/08 article (i.e. C. Burnside, P. Collier, D. Dollar, W. Easterly, R. Levine, D. Roodman and J. Sachs).

86 D. G. Lenihan et al, ‘What is Shared Accountability?’, Policy, Politics and Governance, Vol. 5, November 2003, <http://www.ppforum.ca>

87 Hon. Kevin Rudd MP, ‘Address to Heads of Agencies and Members of the Senior Executive Service’, Great Hall, Parliament House, Canberra, 30 April 2008, p. 6, <http://www.pm.gov.au>

Last reviewed: 
12 June 2018