Stephen Bartos "The Senate and Public Sector Performance"

Papers on Parliament No. 62
October 2014

Stephen Bartos "The Senate and Public Sector Performance*"

Prev | Contents | Next

Gary Banks, Dean of the Australian and New Zealand School of Government, observed in his 2013 Garran Oration that Australians have lost trust in politicians and public servants. Trust was also the theme of Professor Andrew Markus’ last Senate Occasional Lecture, on politicians and the political system. My work in public sector governance has been concerned primarily with ways to improve public sector performance. This lecture therefore addresses the pivotal role the Senate can play in improving public services, by holding managers to account.

Performance and trust are inextricably related. One of the ways organisations build trust[1] and confidence is to deliver, reliably and consistently, what clients or customers want. In the case of the public service this means delivering good advice to ministers, quality services to the public and cost-effective regulation.

A 1992 evaluation of a decade of public service reform found that a random sample of the public expected that the private sector would do a better job than the public sector, but where the respondent had contact with the public service their perceptions were far more favourable.[2] It would be interesting to see whether a similar story applies today, in light of the media coverage of the Royal Commission into the home insulation program, Centrelink waiting times and the like. While it would be relatively easy for the public service to conduct a similar survey and compare it with the 1992 baseline I am not sure it would want to—but more on that later. The key point from that survey is that when public servants deliver to the public, it builds support and trust.

Let us take as a given that we all do want better public services. They are hugely important to Australia. As shown in the 2014–15 budget papers, Australian Government payments are 25.3 per cent of the nation’s gross domestic product—or a quarter of the total of all of the country’s goods and services produced. Despite everything you hear about budget cuts, the decline over the forward estimates period is small, to 24.8 per cent. For the foreseeable future, the federal level of government is about a quarter of all activity in the country. These payments go to social security, health, education, defence[3] and numerous other functions. It is in all of our interests to see that these funds are spent wisely.

We elect a government to do this. Politicians and political parties compete for our votes based on who we think will make the best choices.[4] The literature on governance tells us however that we cannot rely on managers alone to deliver good results. A system of accountability that holds them to account for performance is also vital.

There are numerous differences in systems of corporate governance worldwide, but a common characteristic of effective boards is that they exercise independent and objective oversight of the management of the company.[5] In Australia the ASX Corporate Governance Council[6] suggests that the board will usually be responsible for ‘overseeing management’s implementation of the entity’s strategic objectives and its performance’. Good boards[7] devote much of their time to quizzing the CEO and management on performance; and for their part good managers welcome this scrutiny. It helps to ensure that they deliver the best results possible. A tough board that asks difficult questions about proposed strategy and risks, and whether promised results have been achieved, helps the company succeed. That is an imperative in a competitive market: the alternative is bankruptcy.

The public sector does not face the same commercial pressures and incentives. Public sector bodies have a variety of different objectives, achievement of which can be difficult to measure. A self-interested minister or public servant who is allowed to choose between meaningless waffle or genuine performance information against which he or she can be held to account will choose the former.[8] These are well recognised problems. They are overcome through effective institutional arrangements for scrutiny and accountability. In Australia chief amongst these is the role of the Senate. The Senate can be a highly effective institution in ensuring the public service is held accountable for performance.[9] Its ability to do so is limited by the support and information it receives; while it has some of the enabling or supporting factors in place, there is room for further improvement.

I have characterised accountability not as a set of rules but as a relationship. There must be ‘a person or body who is held accountable, and a person or body to which they account’.[10] In the Australian public sector, accountability of public servants is to ministers, and the ministers’ accountability is to the parliament. Nevertheless there is also a duty on public servants to explain and justify their actions directly to the parliament. In this week in particular, as the Senate legislation committees conduct their estimates hearings, the accountability relationship comes to the fore. The task of ensuring accountability applies in practice, as well as on paper, rests with both parties to the relationship: the Senate and the public service.

The Senate’s capacity to exercise its role in the accountability relationship has had ups and downs over the past 115 years. It depends on the composition of the Senate, the strength of its committees, the structures and advice which support it, and the political environment. There are two factors in play at present which have the potential to weaken Senate scrutiny: blurring of the lines between public servants and ministers, and lowered standards of performance information.

Our system of government has traditionally recognised a difference between the roles of public servants and of ministers. Australia has a hybrid system of government with elements inherited from the colonial Westminster legacy, from United States constitutional thinking (which itself derived from earlier concepts from Montesquieu and other European political philosophers) and a commitment to federalism that cemented states’ rights in the Constitution. Although our Constitution specifies a separation of powers—that is, a strong distinction between the legislative, executive and judicial branches of government—Australia has maintained a commitment to an independent, apolitical public service. This is a contrast to the United States where the executive branch comprises the President, his or her staff, the Cabinet and the senior levels of the public service. Australia’s tradition was inherited from the United Kingdom, which had introduced an independent, merit-based public service in a series of reforms over the mid to late 1800s.[11],[12]

While Australians have never had a high regard for politicians, public servants were traditionally seen as different and attracted a higher level of respect. In recent years as the distinction between the public service and the government of the day has eroded, the public service is closing the ratings gap on trust that it previously had with politicians. The conflation in the public mind of government ministers and the public servants in their departments has been a source of problems for both, and has diminished the ability of the Senate to hold public servants to account.

Estimates hearings have always had elements of political theatre and parochial self-interest. It used to be the case that the Department of Finance provided advisers to each of what were then known as the estimates committees of the Senate, to assist them in understanding the budget estimates they were examining. I was appointed to head the Communications section in the Finance department in 1987 and prepared diligently for the first estimates hearing in which I was to exercise this advisory role. I knew expenditure details for the Communications portfolio back to front. I had a salutary lesson in the realities of political priorities when it turned out none of the senators on that committee needed any of the materials that I prepared. The main issue of substance I recall being discussed was the number and timing of ABC broadcasts of Tasmanian horse races. I believe the senator concerned was worried that horses in his state were not getting a fair go. Indeed, Tasmanian senators are particularly noted for pursuing their state’s interests. So although I am about to discuss current problems in our accountability systems, I do not want you to think that I am nostalgically harking back to an ideal time when every question went to the heart of a major national interest. The Senate, as it should, has always reflected a wide variety of concerns.

Some things have however changed. Up until the public sector reforms of the mid to late 1980s the public service asserted to itself a right to pursue its activities independently from the politicians of the day. Ironically, this gave public servants themselves greater political freedoms. The Department of Finance for a couple of years prided itself on employing the ACT president of both the Liberal and Labor party. It proved they were even-handed. I should here disclose that many years ago I was active in the latter; I allowed my membership to lapse after I was appointed to the APS senior executive service in 1989. I took the view that while in theory it is fine for public servants to participate in politics, in practice it becomes increasingly difficult at senior levels. Having worked at a senior level for the Hawke, Keating and Howard governments I saw successes and failings in all of them, and today am neither affiliated with or lean towards any of the major parties. One thing though worth emphasising is that despite what uninformed commentators may say, ministers from all sides (or at least the Finance ministers I worked with) were personally and sincerely committed to making well-informed decisions in the best interests of the country. I remain proud of having provided advice to John Fahey as Finance minister that helped him to the best ever track record[13] in managing government expenditure so as to achieve sustainable government finances.

Through the 1970s and 1980s it was still possible for public servants to express independent views at variance from the government. Public comment was not monitored or controlled. It was common to see uncensored journal articles from Treasury economists, health, social security and other experts contributing factual information to public policy.[14] That has been overtaken by a new approach which appears de facto to be that public servants should say little, and when they do say anything it is only to explain government policies.

In a similar vein, it was once expected that public service departments would provide their own independent evidence to Senate and other parliamentary inquiries into questions of public administration, including expressing divergent views on how best to regulate for performance reporting and accountability.[15] Today we see only one government view expressed, generally by a central agency. For example in the recent inquiry on the Public Governance, Performance and Accountability Act regulations the only dissenting views came from the Australian Public Service Commission and the Auditor-General: this is, from independent statutory bodies. This is despite the significant misgivings many departments and agencies have about progress with those regulations.

The diligence with which public servants remain close lipped and keep disagreements behind closed doors today means that the Senate, and the public, is often unaware of undercurrents of policy and public administration. A diversity of views might provide early warnings of potential problems—but today these rarely reach the public domain.

The drivers of this increased secrecy and closing of the ranks are not necessarily the public servants involved in the policies concerned. The public service has always had division between those who seek to avoid accountability and those who welcome it. Their influence ebbs and flows. Which tendency gains ascendance depends on the stance adopted by the government and senior public service leadership. It seemed for a while that the proponents of openness were gaining the upper hand. The public service had become increasingly open from the time of the Coombs Royal Commission into Australian government administration in 1996. Administrative law reforms such as the Ombudsman Act and freedom of information under the Fraser Government were followed by publication of estimates and performance information in the 1980s under the Hawke Government, and the introduction of program budgeting. These changes received strong support from Senate committees.

Recent reversals are driven not simply by personal preferences but by more fundamental forces: most importantly, the relationship between government and media. We have seen greater centralisation in the offices of successive prime ministers (from both sides) and a desire to control media messaging on a daily basis. The level of control requires every public statement to be ‘on message’. Given that, as noted earlier, public servants should be responsive to ministers it, has become increasingly difficult for them to participate in public debate.

It has not always been like this. The Senate has in the past been one of the nation’s strongest advocates of transparency in government and civil liberties. One wonders for example what Senator Alan Missen might have made of the secrecy surrounding so-called ‘on water’ matters or the reported ban by the Prime Minister’s Department on public servants commenting in social media even outside of their work. However we must recognise the environment has changed. My aim is not to lament the changes but to set out possible ways in which they might be addressed.

It is also important to note that one of the underlying reasons for greater identification of public servants with ministers is highly desirable—a culture in the public service of greater responsiveness. The days of public service mandarins who ruled vast bureaucratic empires and regarded ministers as a passing inconvenience are long gone. Few would welcome their return. One of the main objectives of public sector reform has been to improve the responsiveness of the public service to governments, of whatever political persuasion. It has been specified as one of the public service values in the Public Service Act 1999.[16]

It is surely better for government priorities to be determined by elected politicians than by unelected public servants. It is called democracy. The trade-off is that responsiveness inevitably means that public servants and their work are more closely identified with the priorities of the government of the day. They are no longer seen to be pursuing a separate agenda.

This creates a practical difficulty for the workload of Senate committees. Because public servants are now closely identified with ministers and government policy it becomes very difficult for a government senator to quiz a public servant on performance. If the scrutiny either deliberately or inadvertently reveals a performance failing then it reflects back adversely on the minister. This is less of a problem immediately following a change of government, but the longer a government remains in office the more likely it is that any problems raised by a senator will be sheeted home to a government decision. This means that only half a committee asks searching questions—diminishing its capacity to improve performance.

The other problem I want to cover is that of incomplete, and sometimes incomprehensible, performance information. Our accountability arrangements depend on the provision by the public service of clear and reliable performance information to the Senate. Ten years ago when teaching at the Australian National University to an executive cohort from the public service I distributed, for amusement, a fictional confidential briefing on how to escape accountability. It is at Appendix A. Among other things it suggested constantly changing objectives and performance information, together with reallocating organisational responsibility for programs, so nobody could ever be held to account. The course participants told me it was not fiction, it was pretty much a description of the way they operated. Little has changed since.

The literature on accountability reveals that from an academic perspective not much has changed either. Professor Richard Mulgan[17] has written extensively and perceptively on the topic. With a few changes of names and cases, his articles and books from 10 or more years ago could apply equally today. Resistance to stronger accountability through external oversight persists in our system of governance, even though the evidence gathered by Mulgan (and others in the field of public administration) demonstrates that it is necessary and desirable. Good performance information is an essential component.

The former Management Advisory Board (MAB), the key advisory body on Commonwealth public administration from the late 1980s until its replacement with a Management Advisory Committee following the passage of the Public Service Act 1999, devoted enormous effort to defining and obtaining service-wide agreement to concepts of accountability. It issued an exposure draft in June 1991 and then final report on accountability in June 1993. This remains a foundation document, referenced in current Australian Public Service Commission online guidance for public servants.

The MAB report noted that ‘the quid pro quo for the devolution of greater authority has been the expansion of accountability mechanisms’. Key components of the accountability system include ‘clear statements of the government’s objectives and the organisation’s role in achieving them’ and ‘management information systems ... to monitor and report on program performance’. Presciently it noted that ‘officials will increasingly be required to provide support to government in its parliamentary accountability activities’.[18]

Twenty years on, the public service still struggles to account for performance. Last year’s report by the Australian National Audit Office (ANAO), The Australian Government Performance Measurement and Reporting Framework—Pilot Project to Audit Key Performance Indicators, noted that ‘it is time for greater attention, investment and resourcing to be given to the quality and integrity of KPIs used by public sector entities to inform decisions about the performance of government programs’. Of 31 KPIs examined, five did not even meet the Finance department definition of ‘measuring the impacts ... on the target group’ and were descriptions of activity instead. Of the remainder, 22 met at least one of the criteria of being focused, understandable, measurable and free from bias, but there was only one that met them all. This is an important stream of work for the ANAO, which has been continued.[19] As noted in its report of February 2014, alas, the ‘continuation of the pilot project observed little change within the guidance promulgated by Finance for the 2013–14 financial year and observed that agencies’ implementation of performance measurement and reporting requires further development’.[20] Progress on improving performance information has been slow.

It is highly unlikely that this is because public sector performance is unmeasurable. Experience shows that where agencies put their minds to it they can develop clear and measurable objectives and very effective performance indicators. The more likely explanation for progress having taken one step forward, two steps back, is other changes in the political and economic climate.

The 1996 National Commission of Audit commented that ‘performance information is crucial in assessing whether policy goals have been achieved and how effectively the public sector has performed’. Among other things it recommended that government ‘adopt a strategic cross-program and organisational approach to setting key results areas for portfolios’ and that ‘Ministers should publish and report against annual plans that clearly explain strategies for restructuring and reducing costs’.[21] Very little of that audit became official policy. Its focus on reducing costs was seen as too extreme. However, the 1996 and 1997 Commonwealth budgets did put a heavy emphasis on the cost savings from privatisation and contracting out. A consequence was that performance-reporting mechanisms were seen in market terms and some of the previous government approaches such as mandatory evaluation were dropped.[22] The logic was internally consistent: if all activities could be put to the market, competitive processes would provide all the indicators of performance needed. The government over the course of its remaining terms in office decided it had no appetite for wholesale marketisation, but did not reinstate firmer performance management regimes.

Another factor was the 1999–2000 reforms to put the budget onto an outcomes, outputs and accruals basis. With some modifications over time it remains the Commonwealth budgeting system. The intention was to improve performance reporting; in practice results were mixed. Presenting the budget on an accruals basis brought it into line with annual reports, reported on that basis from the early 1990s. We saw for the first time an estimate of Commonwealth net worth, an important measure of the government’s stewardship of the economy and its own resources. Balance sheet information, better statement of risks, and more comprehensive financial accounts added to the capacity of senators and others to judge the budget. At the same time other information was lost. In some cases it was for practical rather than sinister reasons: for example graphs in the budget papers showing historical expenditure trends by function could not be produced due to a break in the series. The idea was that these could be reinstated once sufficient time series data had been collected, or if past years’ data could be converted.[23] That could be done now should government decide to do so.

In other cases the reduction in information was, unfortunately, occasioned by the reform. In theory, appropriations to outcomes, identifying the purposes to be achieved, was more in line with the Constitution and more informative than the previous system of appropriating the inputs to agency activity.[24] In practice many agencies took the opportunity of the introduction of a new system to reduce their reporting points, present vague objectives, and develop correspondingly by the performance measures. It reveals the limitations of devolution. In retrospect, it would have been better to have had greater central direction to preserve the integrity of performance reporting.[25]

Just as important was the experience of the mining boom of the 2000s, when government finances were under no pressure and the public service grew rapidly. The financial imperative to justify every dollar spent was no longer a driver of better performance reporting. In these sleepy years of Commonwealth public administration, performance reporting standards slipped.

The March 2010 report on public service reform, Ahead of the Game, does not specifically address the failings in performance information. It does have a chapter that asks ‘How is the APS performing?’ which looks at citizen engagement and calls for greater openness and transparency. It ducks the question of agency efficiency, saying there is no reliable data and more work is needed; although one of the positive outcomes arising out of this has been a series of capability reviews of major departments. While the largely diagnostic earlier chapter says the public service needs to ‘strengthen the accountability of APS leaders’, there are no recommendations on how to do this in the later sections of the report, except perhaps tangentially in the calls for more open government.[26] It is hard not to conclude from the report’s absence of commentary on accountability that the public service at the time saw little need to become more accountable to the parliament for performance. It is perhaps not surprising to see some degree of complacency and introspection given that the report was prepared by an advisory group dominated by public servants and with a secretariat drawn entirely from the public service.

Technological change has also played a part. At first sight it might appear that we have more information than ever, available through department and agency websites. In practice only a tiny proportion is actually useful for accountability: that is, providing evidence on how well agencies have delivered against the objectives and at what cost. Much of what is available online is in effect publicity material—useful for those wanting to find out what the agency does but not helpful for Senate scrutiny of performance. The sheer bulk of online material acts as a barrier to scrutiny because few stakeholders have the time or energy to trawl through it for useful information.

Some—of course by no means all—of the current problems in political trust stem from the decline of performance information. Without wishing to prejudge current Royal Commission processes, it seems likely that a better appreciation of performance indicators (or more simply, what works and what doesn’t) would have been helpful in designing financial stimulus programs. The previous government could have been saved a deal of the criticism it attracted.

Some—again far from all—of the government’s present troubles in explaining the May 2014 Budget arise from the lack of decent baseline performance information and lack of data on trends. If these were more widely available they would help illustrate the budget sustainability problem to a wider audience than the economists and other commentators who enjoy reading through tables and numbers in the budget.

What this illustrates is that holding public servants to account, while clearly a vital institutional function of the Senate, one which delivers benefits to the nation, is also a positive for government. Without it, government effectiveness declines.

So the question is, how can the Senate be enabled to ask the tough questions about performance?

Some answers lie within the government’s own control. The recent National Commission of Audit addressed this head on, and one of its recommendations (Recommendation 9 of its second report, Improving information on government programmes and public sector performance) was that:

Australians should have useful information about the objectives of government programmes, how much the government plans to spend, what it actually spends, and what it achieves. To improve information and drive better public sector performance, the Commission recommends that:

a. all information on programmes be provided in portfolio budget statements with appropriate scope and depth;

b. more meaningful key performance indicators be developed for each programme and be included in portfolio budget statements;

c. the Australian National Audit Office undertake regular audits of each department’s ‘programme performance information’ and its relevance, as contained in portfolio budget statements, including the efficacy of key performance indicators and the quality of the reporting against each indicator; and

d. the Department of Finance develop and maintain a central register of all programme expenditure on a programme-by-programme basis to better inform ministerial decision-making.[27]

It also recommended reinstating mandatory evaluation (recommendation 10, second report); the one weakness of the recommendation was in relation to reporting the results, suggesting ‘final evaluation reports being provided to the Department of Finance on completion’.[28] This approach means that if an evaluation finds problems it can be buried; which while it has obvious short-term political appeal does not lead to long-term performance improvement. A much better option would be for the evaluation reports to be conducted independently, and published on completion. This would allow not only the Senate but also the public to be better informed. It is after all public money used both for the program being evaluated and the evaluation itself.

The government is yet to respond to these recommendations, but there seems no earthly reason why it should respond anything but favourably.

There could also be better support to the parliament itself (particularly the Senate) in understanding the estimates that it examines after the budget and in additional estimates processes. In my report to the Business Council of Australia on budget integrity, published as an attachment to their 2011–12 budget submission[29], I noted that a parliamentary budget office could be tasked not only with costing proposals but also with fiscal sustainability reports, evaluation of major areas of spending that are difficult for government to address internally, and reviewing tax expenditures. I also suggested that it should play a role in explaining and commenting on budget and economic matters—which presently are far more of a mysterious black box than they should be. In the end, the Parliamentary Budget Office (PBO) has largely been confined to the costings role. It may do work behind closed doors on budget analysis, but there is no apparent education and explanation that is released publicly. It appears to play no role in assisting parliamentarians scrutinise performance. A PBO that does little outside costings is of some but only limited benefit. In many respects it takes workload off the public service, particularly Finance and Treasury, which aids the executive rather than legislative branch of government. The PBO is however a creation of the parliament: so it would be within the scope of the parliament to structure it to deliver more value through independent advice on performance measures and how well agencies are delivering against them.

Another step forward that may arise from the government’s consideration of the Commission of Audit and from the implementation of the Public Governance Performance and Accountability Act 2013 is more information in the budget papers themselves. In the 2014–15 Budget there were some positive developments. A table showing all agency outcomes was published in Budget Paper 4, Agency Resourcing. Tables showing staff by agency were moved from an appendix to Statement 6, Budget Paper 1, into Statement 4. Presenting the information in this way is a good move. Hopefully it will over time translate into better organised and more informative portfolio budget statements.

Although this week reminds us that the budget is an important occasion on which to examine estimates, it is not the most important time for addressing performance. This is because budget measures affect only a tiny proportion of total revenue and expenditure. Between the election and the Mid-year Economic and Fiscal Outlook, policy decisions on expenses had a $2 billion impact on the 2014–15 bottom-line, and May budget policy decisions a further $1.9 billion. Total Commonwealth expenses were $415 billion.[30] That is, the policy decisions that have generated so much heated debate this month amount to less than 1 per cent of total budget expenses. It is even less, a mere $673 million, measured in underlying cash terms.[31]

In any one year budget decisions are not even the tip of the budget iceberg but the seagull sitting on top of it. Admittedly the budget can often announce changes in direction that are small in the first year but end up shifting large amounts of spending over time. In the case of the most recent budget, for example, decisions on indexation have a compounding effect that means they will make a significant difference to savings in later years. Past government decisions on health and education have been identified in the most recent budget as having the reverse effect, loading large amounts of additional spending into later years. Even so, these amount to only some 10 per cent of the budget over the longer term. The vast bulk of Commonwealth spending churns on regardless of budget decisions.

Generally speaking this has been true of all budgets from all sides of politics (with some exceptions, including the introduction of the GST and major budget restructuring in the late 1980s).

The Senate does have an opportunity to examine the underside of the iceberg in its hearings on additional estimates, when it has the benefit of agency annual reports. It is perhaps a dubious benefit—annual reports vary considerably in the extent to which they reveal anything at all meaningful about performance. There are other ways to supplement this information. Evaluations, mentioned previously, will help. There is a further recommendation in the Commission of Audit second report that the Department of Finance conduct ‘rolling strategic reviews of major spending programmes’ (recommendation 11). As with the evaluation recommendation, this one has the defect of recommending a secret process inside government, with ‘results and any recommendations to be brought forward by the Minister for Finance as part of the annual Budget process’. That said, if the government agrees to the recommendation when it further considers the Commission report, there will always be the prospect of the Senate Finance and Public Administration Committee asking Finance to tell it what reports it has done, and then farming out the work of investigation of the reports to the legislation committees covering the identified programs.

That though brings us to the elephant in the Senate committee room—how can this be done in a way that genuinely addresses performance rather than it turning into a political circus?

There is a possible answer. Something the Commonwealth lacks, but is effective in other jurisdictions, is a clear statement from government as to the matters of public administration for which it is responsible, and a corresponding statement from the public service about the matters for which it takes full responsibility. Allusions to the distinction between policy and administration in the official witness guidelines are no substitute for a clear, officially endorsed delineation of roles and responsibilities.

We need look no further than New Zealand, which was a leader in codifying statements of expectations and intent and, while retaining the basics of that approach, has moved on to a set of key result areas identified by government with clearly allocated responsibilities and accountabilities from the public service for implementation. Prime Minister John Key has set out 10 priority results and targets to be achieved, and the public service reports on its progress collectively through the State Services Commission and through individual departmental reports. This approach appears from the outside to be working[32], and to have helped that country address some of its underlying budget problems.

Clarification of this nature will be a threat to those public servants who hide behind the blurred lines between them and government to avoid accountability. For public administration more broadly, it would be a plus.

There will be practical barriers. Some ministers prefer not to have clear statements of roles, because it gives them greater freedom to operate. In many activities, there is a genuine overlap where it can be difficult to determine with any clarity a dividing line between the role of ministers and the role of the public service: examples include the conduct of economic policy, the design and development of regulation and other areas of policy development. Fortunately these are only a tiny percentage of public service activity; most of the work of the public service is in program or regulation administration. It is an instance where attempting to achieve a complete coverage would doom the whole initiative to failure, whereas a rough and ready approach with some exceptions would provide greater certainty in almost all cases.

A possible model would be a cascade downwards from an overall statement of what items the public service could be expected to be accountable directly, perhaps issued by the Public Service Commissioner, supplemented by additional information from each agency.

Amongst the beneficiaries would be ministers themselves, already struggling to cope with workloads.[33] They will be reassured that they do not have to be on top of every administrative detail inside their portfolio agency.

The end result could be a committee hearing where a senator questions a public servant about matters that are clearly a public service, not minister’s, responsibility. If problems are identified, it might be quite feasible and proper to hear a ministerial response along the following lines:

Thank you senator for helping identify my department’s difficulties in administration of program X. That is great news: we can now deal with the problems. It will help improve public services for all Australians.

They could go on to say either:

I am shocked, heads will roll

Or preferably:

Heads will not roll. We need the knowledge the public servants involved now have inside those heads about how to avoid the same mistakes in future. They know they made mistakes, and learned valuable lessons.

Let us not get too carried away or optimistic. Senate hearings will always contain politics, because of course all senators are politicians. However, a clear and open statement of what public servants are really responsible for, together with performance information that can hold them to account, will at the very least be an improvement on what we have today.

APPENDIX A: A confidential guide to senior managers on how to keep performance free from scrutiny and accountability

All senior managers have at one time or another faced the annoying problem of parliament or the public seeking to know if they are doing a good job. How can you avoid these embarrassing moments?

1. Make sure that your performance reporting structure changes each year. This makes it almost impossible for you to be held to account. If possible, change your outcomes and outputs descriptions so that nobody is clear about what it is you are meant to be achieving. Frequently it does not matter if you make any real changes to your activities as long as the definitions are changed. If you can change the actual objective without anyone noticing, make sure it is changed to one that is a lot easier to achieve. Keep it as vague as possible.

2. If you can’t change the outputs and outcomes structure, at least use a different set of performance criteria each year—that way nobody will ever find out if you are improving or getting worse over time.

3. If you have been forced by difficult senators into providing performance measures that are consistent year on year, ensure that they are impossible to measure in any one year. Goals over the course of the next decade are ideal—by the time the decade is over nobody will remember who was involved.

4. It is important never to be fully responsible for something. It is easier to ‘assist’ or ‘contribute to’ an objective. Even if your only contribution is correcting the spelling in another department’s policy statement you can include it in your performance information as ‘contribution to’—that way you get to share the credit if the policy is a success. Conversely, if the policy is an abject failure you can always say ‘all we did was correct the spelling, we did not write the policy itself, that was entirely up to department x’.

5. Make sure that you reallocate responsibility for delivering each outcome and output to a different organisational unit each year. The best way to do this is to send only parts of each output to another area, leaving a small part remaining in the original area responsible—that way with any luck neither of them will ever have to answer any questions about performance.

6. If all else fails, change the managers. The new management will have two years in which they can blame any failings in performance on the previous regime while taking any credit for improvements themselves. You know there will be an election called sometime inside three years, a time when nobody pays any attention to performance management anyway.

Question — Both sides of politics commit to running a budget surplus over the economic cycle. I was wondering if you could explain to me what that actually means and in doing so could you refer to structural budget figures because they were in the budget papers for a short period of time. They seem to have disappeared for the last few years and I was wondering if the Commission of Audit looked at structural budget figures at all in their recommendations?

Stephen Bartos — That is a number of different questions but let’s get first to that issue of running a surplus over the course of the economic cycle. Traditional notions of the economy are that we go through periods of growth and then decline. Interestingly in Australia we have gone through a period now of 23 years of nothing but growth but even so in some periods the growth has been stronger than others. The idea of maintaining the budget in balance over the course of the cycle is that when times are good you save some money and when times are bad you can then have the ability to spend some more money. So maintaining the budget either in balance, that is on an even keel, or in a slight surplus over the course of the cycle, enables you to deal with any economic downturns that might come much more effectively.

Now there are real difficulties in measuring that. So economists who look at what is the structural state of the budget have to take in to account what might have happened without government spending, what is happening in the rest of the world and working that out is in fact an incredibly complex thing to do. In Australia it has been made a little more difficult by the fact that with that really long period of economic growth what really is the structural state of the budget? To illustrate that difficulty, Treasury, at the time of the global financial crisis, argued we needed a lot of government spending in order for us not to dip into a recession. In effect what they did was use government spending to build a little bridge over the chasm. That was seen as a desirable thing in the Treasury documents of the time. It has also since been criticised by what was then the opposition, now the government, for having done too much of it. So working it out exactly is incredibly difficult.

I suspect that is why we do not have as much information about the structural state of the budget in the budget papers themselves but I would encourage, if anyone here is from Treasury, some of that material that Treasury works on about trying to calculate that to be released into the public domain perhaps through the Treasury website. The Parliamentary Budget Office (PBO) has done a paper on this as well and that is publicly available.

The other thing to mention is that the budget at the moment is not in a state of real difficulty short term. The thing that is worrying policy makers is the longer term effects of trends particularly in health spending and in social welfare expenditure as the population ages.

Question — My question relates to the role of the Senate in relation to accountability for delivery of services. Government is responsible for the delivery of a wide range of services. Many services that used to be delivered directly by the public sector are now contracted out to the private sector and of course a wide range of issues arise in relation to the way those services are delivered by the private sector. What should be the role of the Senate in pursuing accountability for delivery of services where they are delivered by the private sector? Should it be confined to questioning officials on their management of the contracts? Should the Senate go further and seek to call private sector contractors who are delivering services under contract?

Stephen Bartos — That is a very tricky question because in terms of an in-principle answer you would think that they ought to be able to call the private sector contractors but in practice I do not think that they would be able to go there. It would just be practically too difficult and it is not something institutionally that is built into the thinking of the providers. It might be extremely difficult to get people willing to provide those services if they were subject to that kind of process, but even more importantly, I do not think it would be feasible to retrospectively apply that. So, in other words, if people who have currently got a government contract for delivery of service were suddenly told you didn’t know at the time you bid for it but we are now going to make you appear before the Senate committees. I think this wouldn’t be something you could do.

You could introduce it prospectively for the future but one of the ways around it, and I think this is something that can be done, is to actually hold the public servants to account for the performance of those contractors. All of those contracts do have performance information imbedded in them that is reported to the public servants. To ask the public servants how are your contractors performing against that performance information and getting them to talk that through I think is your other option.

Don’t get me started on how much of that might be classified as ‘on water’ or commercial-in-confidence or some other way of avoiding scrutiny. That is an increasing problem as well and is an unresolved question where I, to be honest, think that the Senate does have a role. The commercial-in-confidence excuse for not talking about the performance of your contractors tends to fall apart in reality because most of the contractors have agreed that that can be made available. It is not really commercial-in-confidence; that is just a smokescreen. So I think that there is a proper and important role for the Senate in quizzing about that. You have been observing these things for a long time, yourself, so what is your view?

Questioner — There is a gap in the public accountability in the whole administrative law area and in this area and my own view is that contracts for performance of public services should include provisions to ensure proper accountability.

Rosemary Laing — May I add something? On a technical basis, the estimates process itself is limited under the standing orders to senators asking questions of ministers and officers about the items of expenditure. But that is only the estimates process. In every other mode of operation Senate committees have the power and the ability to call basically anyone before them. So if there was an inquiry into the delivery of a particular service there is no inhibition on the power of the Senate and its committees to call private contractors as well as government officers.

Now the reality, as we have mentioned, is that the commercial-in-confidence blanket that people hide under can be a deterrent to the provision of information to the Senate for our benefit. But there are some rules in the Senate about a requirement that if you are going to make a commercial-in-confidence claim you need to provide an explanation of the actual harm to the public interest, and in that case, what commercial harm will result from the disclosure of the information. In fact the Auditor-General has been working for many years on government contracts, contracts for provision of services, to ensure at the behest of the Senate that they do not contain unnecessary confidentiality provisions that would prevent the provision of information to parliament. So I think that there are some solutions there.

Stephen Bartos — You are quite right. Because it has been the news of the week, we are looking a lot at the estimates hearings but the legislation committees have inquiries that are wide-ranging on all sorts of other things and that is a very important mechanism of the Senate.

Question — I just have a question in light of Senate estimates. As you would be aware a very common response that public servants do give is ‘that is a matter for the government’ in order to avoid providing further information on that topic. Do you have any comments about whether in your experience you have seen that increasing in recent years and also any suggestions you might have for, when that sort of response is given, what the role of the Senate should be in prosecuting that matter further?

Stephen Bartos — Partly that was why I was advocating something that gave more clarity around that. I did mention the guidelines to official witnesses and they do provide that public servants do not comment on matters of policy, they comment on matters of administration. There is quite extensive guidance in those guidelines for witnesses and they change from time to time but that particular set of provisions has been around for as long as I can remember.

Rosemary Laing — Since 1989.

Stephen Bartos — So that is an area where it is a pretty well established part of the way things work. Does it happen more often now? Look, I think what I am seeing is a greater reluctance on the part of public servants to provide any information. Not just because it might be policy-related but for the other reason we were just talking about of commercial-in-confidence or for various other reasons. That is not because they have become inherently more secretive but because a lot of the incentives on them, as I was trying to explain in the lecture today, are to actually try and protect ministers. So I think the incentives have changed a bit so we probably are seeing more of that behaviour of trying to avoid questions.

Rosemary Laing — Might I add something on that as well? Any of these claims by ministers or witnesses or public officials that they can’t answer that question because it is a matter for the minister, that is not a conclusive answer for a Senate committee. A Senate committee may, if it chooses, press the witness and insist on an answer. It can also put the alternative view: ‘Listen, it’s not for you to tell us what information we are allowed to have—the elected representatives of the people here—it is for us to determine whether we accept the ground you are putting forward for not answering that question’. That is certainly becoming more accepted as a method of going about these things.

Stephen Bartos — It is certainly the case that the Senate has always had amongst the ranks of senators people with a really strong commitment to openness, accountability and transparency and in the formal lecture I mentioned people like Senator Andrew Murray or Senator Alan Missen. I think Alan Missen would be horrified at some of the secrecy and also some of the assaults on personal freedoms at the moment.

Question — We have focused very much in your talk and in the discussion about accountability around public administration. That still leaves us a long way from accountability for real outcomes. In the news this week we see that Australia is now almost equivalent to the US in terms of obesity rates at the same time the government is abolishing the preventive health agency and making it harder to go to the GP, these sorts of things. I am interested in whether you see any future where the Senate might take more of a role in thinking about and taking some role in accountability around outcomes and whether that would be a good thing, whether it is ever likely to happen and what some mechanisms for that might be?

Stephen Bartos — Absolutely one hundred per cent the Senate should be concentrating on whether outcomes are being achieved. That was really one of the goals of a lot of the public sector reform of the late 1980s, even through to the mid-1990s, to try and get a greater focus on outcomes, better performance reporting against outcomes rather than just processes or activities that were being conducted. It is really important for the focus to be on ‘well, yes, fine, but does it make a difference to the lives of Australians?’ That is vital.

Now, as I indicated in the lecture, I do think much of that focus did fall away in much of the 2000s. It really is quite important that the work of the Australian National Audit Office on trying to get better accountability against outcomes, and the work of Senate committees does get supported because really the purpose of the scrutineers is to ask ‘have you achieved any results?’ That is the most important thing to actually ask questions about.

* This paper was presented as a lecture in the Senate Occasional Lecture Series at Parliament House, Canberra, on 30 May 2014.

[1]     Not empirically proven. As Kieron O’Hara notes (Trust: From Socrates to Spin, Icon books, Cambridge, 2004) ‘we do not understand it [trust] ... sociologists, economists and philosophers have studied it, and agree on little, except that it is a mystery’.

[2]     Taskforce on Management Improvement, Commonwealth of Australia, The Australian Public Service Reformed: An Evaluation of a Decade of Management Reform, AGPS, Canberra, 1992, pp. 403–4.

[3]     Listed in order of size. Together these four functions account for some two thirds of budget spending (64 per cent) in 2014–15. See Table 3, Estimates of Expenses by Function, Statement 6, Budget Paper 1.

[4]     Not just on spending but also taxes and regulation.

[5]     Organisation for Economic Co-operation and Development, OECD Principles of Corporate Governance, OECD, Paris, 2004.

[6]     Corporate Governance Principles and Recommendations, 3rd edn, ASX Corporate Governance Council, Sydney, 2014, recommendation 1.1.

[7]     Richard LeBlanc and James Gillies, Inside the Boardroom: How Boards Really Work and the Coming Revolution in Corporate Governance, John Wiley and Sons, Ontario, 2005.

[8]     Not all public servants are self-interested; there are instances where public servants have provided clear and meaningful performance indicators for programs they run without any external prompting.

[9]     This is not to suggest it is equivalent to a company board in scrutinising performance (differences are too numerous to cover here) but it is just as important.

[10]    Stephen Bartos, Public Sector Governance Australia, CCH Australia, North Ryde, NSW, 2004.

[11]    Public Service and Merit Protection Commission, Commonwealth of Australia, Serving the Nation: 100 Years of Public Service, Public Service and Merit Protection Commission, Canberra, 2001. The 1854 Northcote Trevelyan report, although credited with establishing the modern British civil service, actually took more than 40 years to be implemented.

[12]    Although some characterise ours as a hybrid ‘Washminster’ system (a term coined in Elaine Thompson, ‘The “Washminster” mutation’, Australian Journal of Political Science, vol. 15, issue 2, 1980) it would perhaps be better characterised as a uniquely Australian system of government that draws elements from multiple traditions.

[13]    Measured in terms of the budget figures and final outcomes achieved; noting that in politics different players have very different yardsticks against which to measure success.

[14]    Then, as now, partisan political comment was not condoned. However presentation of factual material was acceptable even in cases where the facts in question did not support a particular government line.

[15]    Including proposing alternatives to ideas put forward by the Department of Finance—for examples, see numerous Joint Committee of Public Accounts and Audit and Senate Finance and Public Administration inquiries in the 1990s.

[16]    Section 10(1)(f) ‘the APS is responsive to the Government in providing frank, honest, comprehensive, accurate and timely advice and in implementing the Government’s policies and programs’.

[17]   ‘ “Accountability”: an ever-expanding concept?’, Public Administration, vol. 78, issue 3, 2000,
pp. 555–73.

[18]    Management Advisory Board, Accountability in the Commonwealth Public Sector, AGPS, Canberra, 1993, pp. 7, 15.

[19]    Australian National Audit Office, The Australian Government Performance Measurement and Reporting Framework: Pilot Project to Audit Key Performance Indicators, Australian National Audit Office, Canberra, 2013, pp. 21, 70.

[20]    Australian National Audit Office, Pilot Project to Audit Key Performance Indicators, ANAO, Barton, ACT, 2014, p. 21.

[21]    National Commission of Audit, Commonwealth of Australia, Report to the Commonwealth Government, June 1996, pp. 116, 120.

[22]    Although it should also be noted that the evaluation policies by this time had become stale and in many cases evaluations were conducted purely for compliance rather than to find out information on performance.

[23]    Retrospective conversion of past years’ numbers proved too complex and difficult a task.

[24]    There had previously been reporting known as program ‘budgeting’ which was nothing of the sort: it was reporting against artificially constructed collections of activities that rarely aligned with appropriations. Its value was however that it set out the objectives and performance measures for each program.

[25]    There is an alternative account that suggests there was a deliberate effort by central agencies at the time to encourage departments to avoid accountability.

[26]    Terry Moran, Ahead of the Game: Blueprint for the Reform of Australian Government Administration, Department of the Prime Minister and Cabinet, Canberra, 2010.

[27]    National Commission of Audit, Towards Responsible Government: Phase Two, National Commission of Audit, Canberra, 2014, p. xxiii.

[28]    ibid., p. xxiv.

[29]    Stephen Bartos, Enhancing Budget Integrity in Australia: An Options Paper for the Business Council of Australia, Sapere Research Group, January 2011.

[30]    Figures from Budget Paper No. 1, Statement 6, Table 2, Reconciliation of expense estimates,
p. 6–6.

[31]    There are significant differences between fiscal balance and underlying cash estimates—see
tables 5 and 6 in budget statement 3.

[32]    Noting that most of the reporting on it comes from those within the system, so a degree of caution about the reports is warranted. See State Services Commission, ‘Better public services: results for New Zealanders’,

[33]    Not a new phenomenon—see Patrick Weller and Michelle Grattan, Can Ministers Cope?: Australian Federal Ministers at Work, Hutchinson, Melbourne, 1981.

Prev | Contents | Next