Mr David Thodey AO
Independent Review of the Australian Public Service
Department of the Prime Minister and Cabinet
PO Box 6500
Canberra ACT 2600
Dear Mr Thodey
The Centre for Independent Studies (CIS) welcomes the opportunity to make a contribution to this Review. While noting the broad scope of the review, in the limited time available we confine our submission to highlighting pre-existing work in two areas of particular importance.
A 2014 CIS report, Withholding Dividends: Better Ways to Make the Public Sector Efficient, noted excessive growth of the APS and made a number of recommendations to improve its efficiency and accountability. (A copy of the report is attached.)
Specifically, it stressed the need to:
increase competitive pressures in the provision of public services (for example: through enabling greater private sector contestability and the benchmarking of performance);
conduct regular independent reviews of agency functions and programs to assess whether they are meeting legitimate objectives and identify ways of improving their costeffectiveness.
While there appear to have been developments in both these areas, there is undoubtedly scope to
The CIS report also noted the need to be more discriminating in the application of the ‘efficiency dividend’. The blunt, across-the-board manner in which this budgetary measure continues to be applied, can produce excessive cuts in resourcing for some key areas (such as the Australian Bureau of Statistics), while perversely motivating new program proposals designed to offset the bottom-line impact of dividends.
The focus should instead be on ways of containing the growing functions and roles of the APS and making it more a ‘facilitator’ than ‘performer’ of public services, particularly where private or notfor-profit organisations provide a cost-effective alternative.
Another barrier to efficiency identified in the CIS report is the duplication of some functionsbetween the APS and state and territory bureaucracies, particularly in the health and education portfolios. For example, the 2014 National Commission of Audit noted the growing number of intergovernmental agreements, which create administrative functions for federal agencies to develop, review and report in areas that are the primary responsibility of the states and territories.
Enhancing policy capability and advice
While keeping costs down is important, so too is the quality of what the APS ‘produces’. Among the most important functions of the public service is providing ministers with advice that enables properly informed decisions about policies and programs; that is: advice that makes clear the source of a problem or need, the different options for addressing it (including the default option of not intervening), and their respective costs and benefits.
There are grounds for concluding the APS has not been discharging this core function adequately, and that this has contributed to the observed rise in policy failures over the past decade or so. In particular, reviews of the Home Insulation (pink batts) Program and NBN misadventures by
distinguished former senior public servants have identified deficiencies in technical skills, process and ‘craft’ capability within the APS. These reports and their recommendations, and the extent to which they have been acted on, should be addressed as a matter of priority by this Review.
The Independent Audit of the NBN Public Policy Processes, conducted by Mr Bill Scales AO, identified multiple failures in areas of core public service responsibility.1 It argued for “serious consideration” of whether the inability of the APS to have its views considered “was circumstantial or signals a more serious malaise that needs addressing.”
The Audit proposed, among other things, that:
departments have “knowledgeable subject matter experts” to assist in achieving strategic objectives; and
cost benefit studies be conducted for all large projects as a matter of course.
The 2015 report, Learning from Failure, by Professor Peter Shergold AC,2 was prompted by adverse judgments about APS performance and behaviour, identified by the Royal Commission into the
Home Insulation Program.
In his report, Professor Shergold observed: “it is apparent that public servants’ advice to ministers was, in many instances, poorly given, poorly received and poorly communicated.” He raised questions as to whether APS advice was sufficiently “frank and fearless”, including by “taking a
position on what is regarded as the best way forward”, based on robust analysis. And he found that“ accountability was blurred and risk poorly understood”.
His report to the government contains 28 recommendations, including:
Secretaries be held accountable for the quality of advice to ministers by their departments.
Significant advice to support policy deliberation be provided in writing [but not amended to suit the Office] and records kept – with the Freedom of Information Act amended to ensure the confidentiality needed for frankness and completeness.
A need to clarify who has “end-to-end” responsibility in program implementation, who has delegated authority and “where accountability resides.”
In his Garran Oration of 2013, Restoring Trust in Public Policy: What Role for the Public Service?, the then Dean of ANZSOG, Professor Gary Banks AO (now a Senior Fellow with the CIS), noted the challenges that recent developments in politics and executive government were posing for the
ability of public servants to discharge their duties to the public. The rising dominance of Ministerial staffers over departmental officers in policy formulation – with growth in adviser numbers coinciding with a decline in policy expertise – and “the subtle erosion of the capacity of our most senior public servants to speak truth to power” were seen as key issues to be addressed if real improvements in policy (and public trust) were to be achieved.
Elaborating in a 2017 CIS seminar, Professor Banks – who in a speech on leaving the Productivity Commission issued a ‘to do list’ of pro-productivity reforms proposed a new ‘to do list’ directed at
restoring essential policy capabilities in the public service. In summary, he saw a need for the public
service leadership to:
nurture a culture of ideas and respect for evidence;
build and maintain analytical critical mass;
not rely on consultants for ‘core business’;
be proactive in assessing issues and developing options;
make use of existing ‘due process’ provisions (regulatory impact analysis, cost benefit analysis);
establish ground rules for dealings with Ministerial offices;
uphold the need for written advice; and
hone the craft of ‘speaking truth to power’.
Transparency and consultation processes in the Review
The CIS notes and supports the recommendation by former APS Commissioner, Professor Andrew Podger AO, in his submission to the Review, that an issues/discussion paper be released that sets out
the Review’s early thinking about focus and direction. The CIS believes it is even more important that there be a draft report (or series of drafts) containing the Review’s preliminary findings and recommendations. This phase of the Review should make adequate provision for feedback,
discussion and subsequent deliberation prior to the finalisation of any recommendations to government. Experience suggests that without such ‘stress testing’ of proposals, the risks of impracticality and unintended consequences loom larger.
As noted at the outset, the scope of this Review is very wide. Such breadth has tended to militate against depth in past reviews, at the cost of their utility. (It needs to be recalled that a similar wideranging review of the APS was conducted as recently as 2010, with the ambitious title, Ahead of the Game: Blueprint for the Reform of Australian Government Administration.) There would accordingly be advantages in focusing on key issues where the payoff from reform appears greatest. The CIS
believes that the need to enhance the APS’s efficiency and policy capability clearly passes this test.
The studies referenced in this submission provide insights into how this might be achieved, and the CIS commends them to the Review.
Senior Policy Analyst
1Scales, B. 2014. Independent audit of the NBN public policy process. Available at:
2Shergold, P. 2015. Learning from Failure: why large government policy initiatives have gone so badly wrong in
the past and how the chances of success in the future can be improved. Available at:
3Hanger, I. 2014. Report of the Royal Commission into the Home Insulation Program. Available at:
4Banks, G. 2013. Restoring Trust in Public Policy: What Role for the Public Service? The Garran Oration 2013.
Institute of Public Administration Australia. Canberra. 21 November 2013. Available at:
5Banks, G. 2012. Productivity Policies: the 'to do' list. Address to Economic and Social Outlook Conference,
‘Securing the Future’. Melbourne. 1 November 2012. Available at: https://www.pc.gov.au/newsmedia/
Better Ways to Make the
Public Sector Efficient
National Library of Australia Cataloguing-in-Publication Data:
Author: Philipatos, Alexander, author.
Withholding dividends : better ways to make the public sector efficient / Alexander Philipatos.
Series: Target30 ; T30.10,
Target30 (Online); T30.10,
Subjects: Civil service--Australia--Salaries, etc.
Centre for Independent Studies (Australia), issuing body.
Dewey Number: 351.940247
Better Ways to Make the
Public Sector Efficient
CIS TARGET30 publications
T30.09 Jennifer Buckingham, School Funding on a Budget (2014)
T30.08 Simon Cowan (ed.), Submission to the National Commission of Audit (2014)
T30.07 Robert Carling, States of Debt (2014)
T30.06 Stephen Kirchner, Strengthening Australia’s Fiscal Institutions (2013)
T30.05 Simon Cowan (ed.), Emergency Budget Repair Kit (2013)
T30.04 Robert Carling, Shrink Taxation by Shrinking Government! (2013)
T30.03 Jeremy Sammut, Saving Medicare But NOT As We Know It (2013)
T30.02 Andrew Baker, Tax-Welfare Churn and the Australian Welfare State (2013)
T30.01 Simon Cowan (ed.), Towards Smaller Government and Future Prosperity (2013)
Executive Summary............................................................................................. 1
Efficiency in the public sector...........................................................................3
Exceptions to monopoly in the public sector.......................................................3
What is the efficiency dividend..........................................................................4
Scope and application......................................................................................4
Trends in federal expenses...............................................................................5
Has the efficiency dividend curbed the cost of the public service?..........................7
Problems with the efficiency dividend................................................................9
Measures for greater public sector efficiency.................................................... 11
Efficiencies through enterprise bargaining?....................................................... 11
Competition in public service delivery.............................................................. 11
Regular agency review.................................................................................. 12
During the 2000s, agency running costs in the federal Getting greater efficiency out of the public sector
government grew from $32 billion to $52 billion depends on getting better value for money
(23% in real terms), despite an efficiency dividend of out of essential services, but it also requires
at least 1% applying to departmental appropriations. decommissioning inefficient or ineffective
programs and agencies. The efficiency dividend
Going back further, the public service in Australia
does not address this significant problem. In
has grown top heavy since the early 1990s. The Senior
fact, it allows ministers to sidestep the important
Executive Service (SES) has grown by over 50% and
decisions about which programs and agencies
the number of Executive Level (EL) employees has
the government needs, and which should be cut.
more than doubled.1 In 1991, managerial employees
The decision as to where savings are made is
(EL and SES) constituted 15% of the public service;
instead left to the heads of departmental/agency
today they constitute 30%.2
There are far more managers at the top drawing
The government needs a more targeted approach
large salaries, and fewer at the bottom delivering
to driving efficiency and reducing costs in the public
services. Those of the public service who are
service. This report recommends two solutions:
delivering services are doing so at higher pay grades.
In addition, salaries have grown significantly, 1. T
he government should look to increase
particularly at the top end of the public service. Base competitive pressures in providing public
remuneration for SES level workers has grown between services. The type of competition will depend on
25% and 35% in real terms since 2004. the service:
The efficiency dividend, which is an across-the-board a. g
reater private sector involvement through
cut to the funding that agencies receive for running the use of vouchers for public services
costs, has fundamentally failed to drive efficiency in the
public sector. b. c
ompetitive tender, where private companies
compete for the right to provide public
It has also failed to stem rising costs in the public services for a designated period
sector, but the efficiency dividend has further problems:
reater contestability, where the
• he efficiency dividend is a blunt instrument for
T performance and efficiency of public sector
driving efficiency as it applies equally to efficient agencies are benchmarked against the
and inefficient agencies. private sector so that if agencies do not
deliver on outcomes, or are too inefficient,
• More than larger agencies, smaller agencies government can contract with the private
are put under greater pressure because sector.
they have difficulty achieving economies of
scale, and have fewer resources to apply for 2. C
onduct regular review of agency functions
additional funding (outside the reach of the and programs through an independent body
efficiency dividend). (such as the Productivity Commission’s Review
of Government Services) to determine which
• The efficiency dividend encourages gaming, agencies/programs are meeting objectives and
where instead of cutting back on running costs, how they can become more efficient. These
as is the intention of the dividend, agencies will reviews should also be used to cull programs
submit new policy proposals so that the funding and agencies that are continually failing their
granted for these proposals can be used to cover objectives, operating at too high a cost, or more
existing as well as new costs. This also fuels the appropriately provided for by the private market.
growth of unnecessary and ineffective programs. This process will require greater measurement
of outputs and outcomes, and so the build-up of
• New policies, programs and agencies are a performance indicators will aid in benchmarking
fundamental driver of the growth of government. across government services.
Whether in providing for defence, public infrastructure, An efficient public sector has always been a goal
health care or welfare, Australians rely on government of the Australian government, as was evident in
to provide necessary public services. But the manner the parliamentary debates on the legislation that
in which those services are provided, and the cost established the public service in 1901.6 As pressures on
those services impose on the community, needs the government’s budget become more acute, achieving
close attention. greater efficiency in the public sector will become a
common feature in public debate.
In the last half century, government (particularly,
the federal government) has grown significantly relative But the public sector operates quite differently to
to the size of the economy. This has happened despite the private sector, and these differences make it more
the microeconomic reforms during the 1980s and difficult to pursue improvements in efficiency.
1990s reducing the role of government in managing
the economy. Tariffs were reduced to open up product In most instances, the public sector operates as a
markets, financial deregulation opened up capital monopoly. The absence of output prices and a ‘market’
markets, and the labour market has been partially in the conventional sense means consumers do not
deregulated. In addition, the federal government sold get a valuation of services. Consumers (or taxpayers,
off state-owned enterprises such as Qantas and Telstra, rather) cannot exercise preference for one type of
and state governments sold off utilities. But at the service over another because there are no service
same time, government has embarked on a wider social providers competing with the government.7 Where
agenda focusing on health, welfare and education. government acts as a monopoly service provider, there
is no basis for comparison with industry best practice,
Australia’s ageing population will put continued and little means of distinguishing which methods are
stress on federal and state budgets. The 2010
most efficient.8 Inefficient practices can persist longer
Intergenerational Report projects that by 2049–50,
without the interruption of bankruptcy because the
total federal spending will increase by approximately
government can, and does, simply allocate more money
$60 billion in today’s terms.3 Two-thirds of the projected
to programs and agencies as they become more costly.
increase is expected to come from health costs.4
Spending on pensions and income support payments Measuring public sector performance is
is projected to rise from 6.5% of GDP to 6.9% on a fundamentally difficult since there are seldom output
‘no policy change’ basis, and economic growth is prices and often no competition. In the private sector,
forecast at an unimpressive 2.7% per year.5 At the the presence of prices for products and services means
same time, a lower proportion of the population will be statisticians can compare the cost of inputs (capital,
working-age contributors to the tax system. In short, wages, rent, etc.) to the price of outputs (final product/
government will be facing a new era of budget scarcity, service). The ratio of outputs to inputs—productivity—is
and the public service will be expected to deliver its readily identifiable in the private sector and is used to
services with fewer resources. compare the performance of firms within a given sector
or across sectors. That same comparison cannot be
Since the late 1980s, federal and state governments
made in the public sector for two reasons. First, the lack
have used the efficiency dividend as one of the chief
of output prices makes it impossible to get a measure
instruments to control costs in the public sector. This
of productivity. Second, no comparisons can be made
report analyses the effectiveness of the dividend as a
driver of efficiency in the public service and suggests where government is a monopoly service provider.
other means of pursuing efficiency gains in the public Having said that, there have been some improvements
sector. The report focuses on the operation of the to measuring the efficiency of specific services, such as
federal efficiency dividend as it has been applied to public health and education, which have more readily
federal budgets longest, and has been subject to several identifiable outputs.9
reviews. But the recommendations apply equally to
state and territory governments. Exceptions to monopoly in the
Efficiency in the public sector There are important exceptions to the aforementioned
In the private sector, competitive forces and the profit/ monopoly issues in providing public services.
loss motive create powerful incentives to reduce costs.
Public sector administrators do not face the same 1.
Where government provides a service
incentives and pressures. However, taxpayers rightly already provided for in the private market,
expect value for their tax dollars, and government and is competing with private enterprise.
departments and agencies cannot expect to draw upon Prime examples are the ABC and SBS, which
an ever-increasing pot of budget appropriations. These compete with private free-to-air channels,
pressures imply a need for some form of cost control. particularly in broadcasting news and current
In addition, as technological innovations lower costs in affairs. It is much easier for government to
the private sphere, it follows that if these innovations benchmark performance and cost efficiency of
are applied in the public sector similar efficiencies can the public broadcasters since there are obvious
be gained and costs lowered for government. comparable inputs and outputs to measure.
Where governments ‘marketise’ the Since 2008, the government has introduced two
provision of government services. Though additional one-year rates. The public service has
much of government operates as a monopoly, been generally operating on a base 1.25% efficiency
some functions can still be subjected to limited dividend since 2008. There was an additional impost of
market pressures. For example: 2% applied in the financial years 2008–09 and 2012–13.
tate governments have the sole
responsibility for prisons but contract Scope and application
out the operation of individual prisons to
As mentioned earlier, the efficiency dividend is an
across-the-board cut, in real terms, of the funding
tates also run hospitals but occasionally that agencies receive for their overall running costs.15
outsource patient care to private hospitals.10 Agencies may receive money either from appropriations
Public hospitals often have private wings, handed out in the budget, or from taxes and other
and medical practitioners working in these fees levied on industry. Some agencies do not receive
hospitals are often contractors.11 any of their funding from budget appropriations,
and are exempt from the dividend. In 2011, the
overnments also have franchising Commonwealth efficiency dividend applied to 66%
arrangements for public transport services of agencies.16 Roughly 30% of agencies were not
whereby private companies are contracted subject to the dividend because they did not receive
for maintenance and operations, such as budget appropriations.17
the ferries in Brisbane, or the trains and
buses in Melbourne.12 The agencies that receive appropriations from
the budget receive appropriations for departmental
expenses and administered expenses.
What is the efficiency dividend?
Departmental expenses refer to an agency’s running
The efficiency dividend is ‘the most readily acknowledged costs and include employee wages, supplier expenses,
across-the-board budget mechanism for promoting depreciation/amortisation, and other operational
improvements in agencies’ efficiency.’13 It is designed to expenses.18 Administered expenses relate to funding
create general pressure on the budgets of government within programs such as grants (e.g. for the arts);
agencies so that administrators continually look for subsidies (e.g. industry assistance); and benefit
cost savings and efficiencies, and redirect funds to payments (e.g. pensions).
higher-priority activities. The rationale is that:
The efficiency dividend is applied to departmental
expenses, not administered expenses. This is because
As the public service continually becomes
the dividend is not designed to cut into funding
more productive, there is room for cutting
designated for government programs but to affect the
public sector inputs by the rate of increase
funding that government agencies use to administer
in productivity (or something less) without those services. Agencies have little control over which
changing the level of output.14 programs and payments they administer on behalf of
government, but do have control over the resources
Essentially, government expects that over time,
they use to administer programs.
agencies should deliver to the public the same quality
and quantity of service (or perhaps greater) with fewer Each year, the appropriations agencies receive for
resources. The efficiency dividend has been applied, departmental expenses are adjusted. This is based
at various rates, until today since it was introduced on the previous year’s appropriations, and is adjusted
by the Hawke government in the 1986–87 Budget. for changes to the agency’s functions (in terms of the
Table 1: Annual rates of Commonwealth efficiency dividend since 1987
Period Annual rate
2008–09 (base rate of 1.25% + additional single-year 2% impost) 3.25%
2012–13 (base rate of 1.25% + additional single-year 2% impost) 3.25%
Source: Nicholas Horne, The Commonwealth Efficiency Dividend: An Overview (Parliamentary Library, 2012), 7.
policies and programs they administer) and indexed Trends in federal expenses
for minimum wage growth. The dividend is applied to
The Commonwealth Review of the Measures of Agency
this portion of appropriations. It has been in place for
roughly 25 years at the federal level, but most state Efficiency in 2011 found that departmental expenses
and territory governments also apply, or have applied, (the part of agency expenses subject to the dividend)
efficiency dividends to their budget appropriations. grew from $32 billion to $52 billion (59%) between
2000–01 and 2009–10.22 Administered expenses grew
Apart from agencies that receive revenue from $147 billion to $292 billion (98%). Until 2007–08,
exclusively from taxes and other fees, some agencies, growth in departmental and administered expenses
despite receiving budget appropriations, have been
was roughly in line with growth in nominal GDP
exempted from the efficiency dividend. These include
(Figure 1). When the effects of changes in the terms
the ABC and SBS, due to electoral commitments to
of trade are accounted for, the trends are even closer.
‘maintain the real level of funding for each broadcaster,’
and Safe Work Australia, due to co-funding by federal However, things changed after 2007–08 when
and state/territory governments.19 Other agencies there was a large run-up in administered expenses.
with partial exemptions include:20 The review points to the government’s fiscal stimulus
response to the global financial crisis as the key driver
• Commonwealth Scientific and Industrial Research of the spike in administered expenses since 2007–08.
Organisation (CSIRO) The proliferation of spending initiatives, and the large
• Australian Institute of Marine Science (AIMS) amounts spent on those programs (cash grants to
families and individuals as well as other initiatives
• Australia Council for the Arts such as the Building the Education Revolution (BER)
• Australian Customs and Border Protection Service program) during the global financial crisis, accounts
for the growth in administered expenses.
• Australian Nuclear Science and Technology
Organisation (ANSTO) Once again, the statistics show a large departure
in administered expenses coinciding with the global
• Department of Defence (DoD)
financial crisis. More so, the growth in nominal
After accounting for exempted agencies and scope, administered and departmental expenses has grown
the efficiency dividend applied to approximately 6.5% substantially over the past decade. Despite continual
of total federal budget outlays in 2011.21 operation of the efficiency dividend, departmental
Figure 1: Cumulative growth in nominal expenses (departmental and administered)
compared to GDP (2000–01 to 2009–10)
Source: DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency
(Canberra: March 2011), 14.
expenses have still grown 23% in real terms over to households (in addition to other spending),
the decade, and administered expenses grew at administered expenses grew much faster than
roughly the same rate before spiking during the global departmental expenses. The cash payments required
financial crisis.23 minimal additional public service resources apart from
the actual grants being paid. Developing and
New programs and initiatives appear to be driving implementing an entirely new program would have
the need for more and more spending. Indeed, at meant a far greater impost on the public service.
the onset of the global financial crisis, when the Instead, the main cost was the actual grants to families,
Rudd government gave large one-off cash payments which is reflected in administered expenses.
Figure 2: Cumulative real growth in nominal expenses (departmental and administered) compared to
population growth (2000–01 to 2009–10)
Source: DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency
(Canberra: March 2011), 14.
Figure 3: Number of ongoing and non-ongoing APS employees (1990–2013)
Source: APS (Australian Public Service), Statistical Bulletins (Canberra: 1991 to 2013).
Has the efficiency dividend curbed the cost of the public service?
The federal public sector employs 248,000 people, 67% The composition of the ongoing (i.e. permanent)
(or 167,257) of whom belong to the Australian Public APS workforce has changed significantly since the early
Service (APS). 1990s (Figure 4). Entry-level positions such as trainees
and APS classifications 1 and 2 dropped precipitously
Since the mid-1990s, the size of the APS has changed
throughout the 1990s, and continued to decline until
considerably. Large cuts in the service enacted in the 2013. Since 1991, the number of APS1 employees has
1990s by the Keating and Howard governments were reduced from 30,437 to 776 (98%) and the number
completely reversed from 2000 to 2007. The ongoing of APS2 employees has shrunk from 18,983 to 3,282
workforce decreased by approximately 30% from 1992 (83%). Higher up the APS, the trend is reversed. The
to 1999. By 2007, employee numbers had risen back to number of APS6 employees grew from 18,172 to 32,837
1992 levels, but with one distinct difference. (81%) over the same period.
Figure 4: Growth of ongoing APS1 to APS6 employees (1991–2013)
Source: APS (Australian Public Service), Statistical Bulletins (Canberra: 1991 to 2013).
Figure 5: Growth of ongoing EL and SES employees (1991–2013)
Source: APS (Australian Public Service), Statistical Bulletins (Canberra: 1991 to 2013).
The growth is even more pronounced in the public employees compared to two decades ago, they
executive level (EL) and special executive service are paying far more for those public servants.
(SES) classifications. Over the same period, EL1
Lindsay Tanner, then shadow minister for finance,
classifications grew from 11,191 to 28,634 (156%) and
EL2 classifications from 7,262 to 13,087 (80%). SES highlighted the problem in 2007:
classifications grew from 1,788 to 2,736 (53%).
The government have expanded the Public
So while the lower levels of the APS have been
Service back out to where it was but with
hollowed out, outgoing employees have been replaced
one important caveat, and that is that there
by higher-grade workers earning larger salaries. The
public service has become top heavy. There are many are fewer workers, fewer people actually
more managers, more highly trained executives, and delivering services on the ground, and a
fewer entry-level employees. This has big implications lot more chiefs, a lot more fat cats, a lot
for the public services’ wage bill. While Australian more people at the top end earning very
taxpayers are not necessarily paying for many more high salaries.24
Figure 6: Cumulative real growth in median base salary (2002–12)
Sources: APS (Australian Public Service), Remuneration Report 2012; ABS, ‘Consumer Price Index,’
Cat. No. 6401 (March 2012).
Figure 7: Cumulative real growth in median base salary (2002–12)
Source: APS (Australian Public Service), Remuneration Report 2012; ABS (Australian Bureau of Statistics),
Consumer Price Index (various years).
The changes in the composition of the APS go some associated with an ambitious Council of
way to explaining why, for example, between 2007–08 Australian Governments reform agenda,
and 2012–13, the total number of employees in the and an increasingly fraught and complex
federal public sector increased by 4.8% while employee international environment.31
wages and salaries rose by 28.7%.25
So while more is expected of the public service, there
But there has also been growth in employee are also unnecessary programs and regulations (and
remuneration. Complete data on remuneration does not unnecessary staff) that continue to consume valuable
trace as far back as employee head counts gathered in resources. Bartos also points to easy money and a
the APS Statistical Bulletins. However, data show that lack of ministerial oversight as important contributors.32
in the 10 years from 2002, employee base salaries for
graduates and APS classifications grew between 15% One possible explanation for the growth in
and 20% in real terms (adjusted for inflation).26 The APS 5 and APS 6 (and the decline in APS 1 and APS 2)
only exception was APS 1 classifications, whose base classifications is that employees at lower classifications
salaries grew by 8%. are being promoted as a means of giving them a pay
rise over and above wage caps. Governments impose a
Again, the most significant story is at the higher restriction on wage increases for the public sector—no
classification level. While base salaries have grown at a wage increases over and above inflation (CPI or another
healthy pace for EL employees (each around 15% in real wage index) without productivity improvements. An
terms), SES salaries have grown much faster (Figure 7). employee must apply for a job at a higher classification
Base salaries for SES employees have grown between to get a substantial wage increase. If this is happening
25% and 30% in real terms over the past decade. Not on a larger scale, it means the wage cap is being
only has the number of high-grade employees grown at gamed, and that tasks that used to be performed by
an alarming rate, but the growth of their salaries is also workers at APS 1 and APS 2 classifications are now being
a worrying trend. The combination of these two trends performed by workers on higher pay classifications.
has significant implications for the budget.
At first glance, there is no obvious reason why the Problems with the efficiency dividend
APS requires so many more high-level employees. The The efficiency dividend is a relatively simple instrument
2010–11 APS State of the Service report notes that governments can apply to appropriations. This is
from 1984 to 2002, the size of the SES was relatively perhaps its chief advantage. It is predictable and the
stable with minor variations reflecting budgetary savings made are obvious insofar as they affect the
changes or organisational change.27 Since 2002, the budget bottom line. The dividend applies pressure
SES has risen by 53%. The public service played an indirectly, rather than directly, and provides incentives
important role in policy formation and giving advice to to managers to find savings in areas of their choosing.33
the government during the microeconomic reform era
of the 1980s. But as Professor Stephen Bartos, former Despite these advantages, the efficiency dividend
deputy secretary of the finance department, notes:28 has several flaws.
Scope and size: The efficiency dividend is applied
The numbers of senior executives we needed
to appropriations for an agency’s departmental
during 18 of the most turbulent years of
expenses (see p. 4). It is also not applied to all
APS history (microeconomic reform, the
agencies. In 2011, it applied to roughly 66% of
recession we had to have, opening up the
federal agencies, which equated to just 6.6%
economy to international competition, sales
of the federal budget.34
of large government assets, APS reform,
devolution, market-testing) remained more
Indiscriminate: Perhaps the chief criticism
or less the same … But, in the comparatively
of the efficiency dividend is that it does not
easier eight years since 2003, the number
specifically target wasteful, unnecessary or
of senior executives has blown out by
inefficient agencies. It is applied to most
50 per cent.
agencies without taking into account differences
A review of the SES in 2011 found strong links in funding arrangements.35 Instead of cutting
between the growth in the SES and the growth in back known inefficiencies or irrelevant agencies,
the number of programs administered by the APS.29 the efficiency dividend applies to efficient and
The review also pointed to other factors such as inefficient agencies alike. It has the perverse
the heightened national security environment, the effect of punishing agencies that have already
complexities of climate related issues, the global achieved efficiencies, and rewarding those that
financial crisis, and immigration.30 can hide additional savings to use at a later
time, such as during 2008–09 or 2012–13 when
However, the report also found: an additional single-year impost was applied
to the baseline dividend percentage.36 This fact
Growing regulation, increasing scope and is simply part of the dividend’s design and the
use of judicial review, increasing frequency difficulties that exist in measuring productivity
of cabinet meetings outside Canberra, and efficiency in the public sector. If it were
increasing number of high-level negotiations possible to measure productivity in the public
sector the way it is measured in the private Agency size: An agency’s size can have
sector, it would be possible to apply different considerable effect on the agency’s ability
efficiency dividends for different agencies, or to to find efficiencies and deal with a cutback
different departments. It would also be much in budget appropriations. Smaller agencies
easier to compare productivity levels with the have two distinct disadvantages compared to
private sector. As it stands currently, there is no larger agencies:
clear connection between the dividend and any
a. Poorer economies of scale
actual efficiency measures.37
ower ability to obtain funding for new policy
ne of the efficiency dividend’s positives is that
it forces some accountability on agency managers
while still allowing them the flexibility to prioritise In a submission to an inquiry into the effects of
functions and expenses. However, this also allows the efficiency dividend on smaller agencies, the
the responsible minister to avoid making tough Australian Electoral Commission (AEC) noted
decisions about which programs are working that for some smaller agencies, ‘Overheads
more efficiently than others, which programs are such as IT, property and security comprise a
not meeting objectives, and which programs disproportionate share of their budget.’39 As
are not worth the resources they are consuming.38 a result, a large proportion of expenses are
The dividend allows the minister to simply fixed. They then have a very small amount of
outsource these decisions to agency heads and their budget from which they must try to create
let them make the cuts. That may be successful savings to cope with the efficiency dividend.
if there are obvious cuts to be made, but not so
easy if entire programs or even entire agencies So while a sizeable cut in appropriations may
ought to be cut. In addition, the priorities of be easily managed by larger agencies with
the minister may not always align with the larger economies of scale and more manageable
priorities of the agency manager. The manager overheads, smaller agencies are under greater
of an agency that has outlived its usefulness will pressure. In circumstances where agencies
not easily suggest abolishing the agency when have been unable to find additional efficiencies,
it means eliminating their own job, and those some have responded by reducing services and
of their colleagues. Nor will they easily abolish laying off workers.40 Reducing services via staff
programs they favour. reductions is not the intention of the efficiency
dividend. It may be the case that some agencies
Revenue base: An agency’s revenue comprises
- are overstaffed, but if that were the case a
budget appropriations and proceeds from other reduction in staffing should not compromise
taxes and fees. Since the efficiency dividend service obligations.
is applied to budget appropriations and not to
other forms of an agency’s revenue, there can Gaming: The efficiency dividend is prone to
be quite large differences between the proportion gaming by agencies unable or unwilling to drive
of an agency’s revenue that is subject to efficiencies. Agencies can obtain additional
the dividend. funding for new work/programs the government
seeks to implement by submitting new policy
An agency will feel more pressure from the proposals. The opportunity to advance new
efficiency dividend if a large part of its revenue policy proposals is available to all ministers and
comes from budget appropriations. Agencies the agencies in their portfolio.41 If approved,
who receive most of their revenue from taxes, these new proposals will be granted with
levies and other fees will find themselves to a additional funding. The agency can then use
greater extent immune from the cutbacks made at least part of these funds to help alleviate
via the efficiency dividend. This means through pressure from the efficiency dividend. This has
no fault of their own, and from factors entirely two negative effects: negating the intention
out of the agency’s control, certain agencies of the efficiency dividend, and leading to a
will find themselves under much greater fiscal raft of policies and programs that are not
pressure than others. genuinely pursued.
Measures for greater public sector efficiency
Though the efficiency dividend has advantages in its incentives, and unsurprisingly, will not easily offer up
simplicity and pressure on management, it is unlikely solutions that will lead to the termination of their own
to be the most appropriate means of driving efficiency (or their colleagues’) jobs. In addition, the inability to
in the public sector. accurately measure productivity in the public sector
creates additional problems for administrators looking
The growth in the cost of the public service has
to enforce agreed productivity improvements. This does
several drivers, and government will need to tackle
not mean any attempt to improve productivity/efficiency
these drivers if agencies are to continue to find
through enterprise bargaining negotiations should be
efficiencies. It is clear that one of the major drivers of
abandoned. Rather, it is not a suitable substitute for
the cost of the public service is the growth in new
the efficiency dividend.
programs and policies. While the public service has
a pivotal role in the formulation of new policies and
services, not all new programs will be successful, Competition in public service delivery
and not all will be worth the resources required to fund
One of the ways government can drive greater efficiency
them. Some programs should not have government
in the public sector is to introduce greater competition.
involvement at all. If the growth of new policies
Even where government has the sole responsibility for
and programs is also driving up the demand for
delivering public services to its citizens, it can inject
high-skilled (and high-salaried) employees, then these
competitive pressures into the delivery of these services
new proposals require initial and ongoing scrutiny.
via different means. Gary Sturgess points to three
Efficiencies through enterprise
Choice based markets: Customers (or
taxpayers) use funds provided by government
As early as 1992, some of the shortcomings of the to select from a number of approved suppliers.
efficiency dividend had been highlighted. The report These schemes operate essentially as vouchers,
The Australian Public Service Reformed: An Evaluation such as the Medicare card used locally to purchase
of a Decade of Management Reform noted that the health care services, or for the proposed National
dividend failed to consider the difference between Disability Insurance Scheme.
agencies.42 It suggested that the future of the efficiency
dividend lay in resource agreements and enterprise Commissioned markets: Private companies
bargaining, rather than arbitrary annual reductions bid for the right to provide services to taxpayers
in appropriations for running costs.43 This proposal through a tender. The winning company secures a
was echoed in the 1995 report Keeping the Customer contract and operates as a monopoly for a defined
Satisfied: Inquiry into the Devolution of Running Costs period of time, after which the government can
Government policy in several states stipulates that Contestability: The performance of service
the government will not award wage increases over and providers is benchmarked and they face the
above a specific percentage for inflation indexing unless threat that the government will source
offset by efficiency or productivity improvements. Those alternatives if the providers fail to deliver results.
productivity improvements would be compensated
Not all of the above options will be applicable to
by additional wage increase above indexation. The
all public service agencies. The appropriate option will
approach of attaining productivity/efficiency gains
depend on the service in question. For example, health
through enterprise bargaining negotiations comes from
care or education options can be served by vouchers,
attaining savings by removing restrictive work practices,
since consumers can choose from a range of private
or adopting new technologies/work practices.
sector options using public funds. This option is not
The government’s primary objective here may be to possible for the provision of national defence equipment
keep costs under control, but motivating the union and (fighter jets, armaments, submarines, etc.) since
the workforce to embrace a productivity agenda would government must be the sole purchaser and provider
allow departments to change workplace practices with of national defence. A commissioned market for defence
less resistance. equipment may be more appropriate.
The problem with this strategy is enforcing agreed Increasing the role of the private sector in
productivity improvements. Enterprise bargaining as providing public services has been a growing trend in
means for driving efficiency has been used in several recent decades, both nationally and internationally.
states, and in practice agreed efficiency improvements But it should be noted that there are already various
do not eventuate. Part of the problem here may be areas of our economy with significant private sector
that those involved in the bargaining process are involvement. At home, around 17% of Australian
far removed from the day-to-day workings of lower prisoners are held in privately managed correctional
level public servants, and thus, do not know where institutions.46 In Victoria, that proportion is 34%. In
savings can be made. Public servants have their own health, there are 768 public and 556 private hospitals,
with 40% of inpatients treated in private hospitals.47 outcomes become the indicators the government looks
There are often privately financed hospitals where to when deciding who or which agency to contract –
state governments have contracted a private company private, public or a combination of the two.
to design and construct a new hospital, or to manage
Second, involving the private sector in public
the operations of a public (or private) hospital.48
service provision will stimulate greater innovation. As
Roughly 35% of Australian children attend a private
companies compete for the rights to provide public
school, and 39% of secondary school students go to
services, or as several contracted companies compete
to outperform each other, they will experiment both
In sectors where public and private providers deliver with new approaches to delivering services or novel
similar services, it is possible to make meaningful ideas for reducing costs. There are incentives both to
comparisons regarding performance and cost between increase quality and to reduce costs. Though innovation
private and public provision. is possible in the public sector, it is less likely than in
the private sphere, just out of the sheer size and number
• In the realm of public transport, a study
of competing firms in the private sector.
comparing like-for-like rail services between
international operators and Sydney’s CityRail
(NSW government owned and operated) Regular agency review
indicated operating costs were 23% lower for
The government should also look to conduct
comprehensive yearly audits of agencies/portfolios for
• The Productivity Commission conducted a efficiency by an external and independent body, such
comprehensive study comparing 368 public as the Productivity Commission’s Review of Government
acute hospitals in Australia and 122 private Services. Regular review was proposed in the
hospitals (for-profit and not-for-profit).51 It government’s review of efficiency measures to develop
found that ‘In 2007–08, the general hospital alternative ways of promoting efficiency.53
cost per “casemix-adjusted separation” was Having regular review agency efficiency helps
about 30% higher in public hospitals compared collect performance literature and provide an evidence
to private hospitals.’ base that ministers and policymakers require to make
an informed decision. By giving greater attention to
• In 2005, the NSW Parliament’s Public Accounts
measuring results and outputs (rather than inputs),
Committee released a report titled Value for
the government can build a narrative around both the
Money from NSW Correctional Centres. This
efficiency and effectiveness of programs and policies.
report found that health services in the privately
It would also act as an opportunity to reassess the
operated prison in Junee were less than half the
cost of those in publicly managed prisons.52 It merits of spending initiatives, and decommission
also noted that the average daily expenditure on programs that are not meeting objectives or are
an inmate was half the cost in Junee than in the consuming too many resources.
public prisons. Though Junee had cost advantages For agencies that perform similar functions, such as
over the public prisons, the cost differential was those producing research or regulatory oversight, the
still quite stark. gradual build-up of performance metrics could allow
Government should extend competitive pressures agencies to benchmark performance against each other,
further into the public service than traditional and for best practice to be replicated. The benchmarking
areas such as health and education. This does not process could also assist the government in making
necessarily mean contracting with the private sector decisions about possible private sector involvement.
for providing government services. Governments Regular review would help overcome some of
should explore benchmarking the performance of public the major criticisms of the efficiency dividend. First,
sector agencies to compare them with private sector it would eliminate the indiscriminate nature of the
equivalents. Sometimes this step is enough to kick-start dividend by giving a performance appraisal specific
better performance since agencies know that if their to the agency. This would mean that smaller agencies
services are consistently more expensive than private with fewer options to increase efficiency would not
equivalents, or if they consistently fail to deliver be put under the same funding pressure as larger
adequate results, the government will go elsewhere.
agencies that can use economies of scale. Second,
There are two main benefits to introducing greater because the review would assess the effectiveness
competition and choice into the public service, and not of new policies, it would also reduce the ability of the
all relate to cost. agency to ‘game’ the new policy proposals system.
First, it motivates measurement. If the government The audits would report to the relevant minister,
is to consider contracting services to a private company, and the agency head tasked with providing information
or at least to benchmark public agencies against for the audit process would also be responsible for
private sector counterparts, it must focus attention implementing the review’s recommendations. Some
on outputs and outcomes rather than inputs. The additional resources would be required to increase the
government is forced to come up with a specific definition capacity of the Review of Government Services such
of the services it wants if it is to outline its expectations that it is able to conduct the sort of comprehensive
to a prospective private contractor. These outputs and review that is needed.
Pursuing efficiency in the public sector is difficult. Government should where possible increase
In the private sector, competition creates ongoing competition in providing public services by involving
pressure on businesses to reduce costs and pass on the private sector. This may take the form of publicly
those savings to consumers. The public sector lacks funded vouchers for taxpayers to buy from the private
these important incentives. market, involve contracting out some government
services to the private sector, or simply mean a
Australians expect government to deliver high-
contestable market where public agencies continue
quality public services; however, they are also acutely
to provide services, but where their performance
aware of the share of their income they sacrifice for
and efficiency are benchmarked against private
sector equivalents. This additional measurement and
The efficiency dividend has been the central competitive threat will help motivate better use of
instrument in government’s attempts to drive efficiency public resources.
and deliver value for money in the public sector. But it
Comprehensive yearly reviews of government
has fundamentally failed to contain costs in the public
agencies/portfolios are an essential part of ensuring
service, particularly in relation to the workforce itself.
value for money for taxpayers. The government needs
Since the early 1990s, the public service has become quality, independent advice to make informed decisions
top heavy. After the public sector cuts of the mid- to about which agencies and programs are worthwhile,
late-1990s, the number of EL and SES employees has which can become more efficient, and which should
ballooned. There has also been significant growth in be scrapped.
salaries, particularly at the top end. There are now many
The public sector is fundamentally different from
more high-skilled, high-salaried executives in the public
the private sector. Private businesses will grow and
service and fewer on-the-ground workers. This has
decline, and this ongoing process of renewal brings
serious implications for the cost of the public service.
regular change. The public sector on the other hand
Several reviews of the efficiency dividend have requires constant reform to keep it effective and
come back with the same conclusion—the government efficient. It is incumbent upon government to continually
needs more targeted and confronting measures to review its functions to ensure taxpayers are getting
drive efficiencies in the public service. their money’s worth.
1 APS (Australian Public Service), Statistical Bulletins (Canberra: 1991 to 2013).
2 APS (Australian Public Service), Statistical Bulletins (Canberra: 1991 to 2013).
3 Treasury, Intergenerational Report 2010 (Canberra: Government of Australia, 2010), 45.
4 As above.
5 As above, 58.
6 APSC (Australian Public Service Commission), Serving the Nation: 100 Years of Public Service
(Canberra: APSC, 2001), 8–9.
7 Francisco Pedraja-Chaparro, Javier Salinas-Jiménez, and Peter C. Smith, Assessing Public Sector Efficiency:
Issues and Methodologies (31 March 2005), 345.
8 As above.
9 DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency
as above, 9.
10 Productivity Commission, Public and Private Hospitals (Canberra: 2009), 46.
11 As above.
12 Alexander Philipatos, Free Trade Ferries: A Case for Competition, Issue Analysis 127 (Sydney: The Centre
for Independent Studies, 2011).
13 As above.
14 As above.
15 DoFD (Department of Finance and Deregulation), Finance Submission to the Joint Committee of Public
Accounts and Audit Inquiry into the Effects of the Ongoing Efficiency Dividend on Smaller Public Sector
Agencies (2008), 1.
16 Nicholas Horne, The Commonwealth Efficiency Dividend: An Overview (Parliamentary Library, 2012), 3.
17 DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency,
as above, 66.
18 Budget Measures: Budget Paper No. 4: 2012–13 (Canberra: Government of Australia, 2012), 8.
19 As above, 66–68.
20 Nicholas Horne, The Commonwealth Efficiency Dividend, as above, 9.
21 DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency,
as above, 20.
22 As above, 13.
23 As above, 14.
24 Lindsay Tanner, Shadow Minister for Finance, Second Reading, Hansard (Parliament of Australia, 12 February
25 ABS (Australian Bureau of Statistics), Employment and Earnings, Public Sector, Australia, 2012–13, Cat. No.
6248.0.55.002 (Canberra: ABS, 2013).
26 APS (Australian Public Service), Remuneration Report 2012 (Canberra: APS, 2012); ABS (Australian Bureau
of Statistics), Consumer Price Index (Canberra: ABS, 2002–12).
27 APS (Australian Public Service), State of the Service 2010–11 (Canberra: APS, 2011), 2.
28 Stephen Bartos, ‘PS cuts: The coward’s guide,’ The Canberra Times (5 December 2011).
29 APS (Australian Public Service), State of the Service 2010–11, as above, 3.
30 As above.
31 As above, 3.
32 Stephen Bartos, ‘PS cuts: The coward’s guide,’ as above.
33 Nicholas Horne, The Commonwealth Efficiency Dividend, as above, 11.
34 DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency,
as above, 20.
35 DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency,
as above, 11.
36 Stephen Bartos, ‘PS cuts: The coward’s guide,’ as above.
37 Nicholas Horne, The Commonwealth Efficiency Dividend, as above, 12.
38 DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency,
as above, 27.
39 AEC (Australian Electoral Commission), Submission to the Joint Committee of Public Accounts and Audit Inquiry
into the Effects of the Ongoing Efficiency Dividend on Smaller Public Sector Agencies (Canberra: AEC, 2008), 3.
40 DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency,
as above, 70.
41 ANAO (Australian National Audit Office), Submission to the Joint Committee of Public Accounts and Audit
Inquiry into the Effects of the Ongoing Efficiency Dividend on Smaller Public Sector Agencies (July 2008), 3.
42 Task Force on Management Improvement, The Australian Public Service Reformed: An Evaluation of a Decade
of Management Reform (Canberra: AGPS, 1992).
43 ANAO (Australian National Audit Office), Submission to the Joint Committee of Public Accounts and Audit
Inquiry into the Effects of the Ongoing Efficiency Dividend on Smaller Public Sector Agencies, as above, 63–64.
44 David Simmons, ‘Keeping the customer satisfied: Report on the inquiry into the devolution of running costs
flexibilities,’ House of Representatives Standing Committee on Banking, Finance and Public Administration
(Canberra: Australian Government Publishing Service (AGPS) 1995).
45 Gary Sturgess, ‘Diversity and Contestability in the Public Service Economy’ (Sydney: NSW Business
46 General Purpose Standing Committee No. 3, ‘Inquiry into the Privatisation of Prisons and Prison-Related
Services,’ NSW Legislative Council (3 June 2009), 7.
47 Gary Sturgess, ‘Diversity and Contestability in the Public Service Economy,’ as above, 10.
48 Peter Phelan and Jeremy Sammut, Overcoming Governance and Cost Challenges for Australian Public Hospitals:
The Foundation Trust Alternative, Policy Monograph 137 (Sydney: The Centre for Independent Studies), 12.
49 Gary Sturgess, ‘Diversity and Contestability in the Public Service Economy,’ as above, 13.
50 L.E.K. Consulting, ‘Cost Review of CityRail’s Regular Passenger Services’ (Sydney: Independent Pricing and
Regulatory Tribunal (IPART), June 2008), 3.
51 Gary Sturgess, ‘Diversity and Contestability in the Public Service Economy,’ as above, 88.
52 NSW Parliament Public Accounts Committee, ‘Value for Money from NSW Correctional Centres,’ Report No.
13/53 (Sydney: 2005), vii.
53 DoFD (Department of Finance and Deregulation), Report of the Review of the Measures of Agency Efficiency,
as above, 43.
Alexander Philipatos is a Policy Analyst in the Economics Program at The Centre for Independent Studies.
His research has included analysis of public transport infrastructure, but focuses predominantly on industrial
relations and labour market policy.
Alexander has a Bachelor of Economics (Hons) from La Trobe University; before joining the CIS in 2011,
he was a tutor at La Trobe teaching Introduction to Quantitative Analysis.
CIS TARGET30 • T30.10 • ISSN Print: 2201-7615 • ISSN Online: 2202-3240 • ISBN: 978 1 922184 28 3
Published May 2014 by The Centre for Independent Studies Limited. Views expressed are those of
the authors and do not necessarily reflect the views of the Centre’s staff, advisors, directors or officers.
© The Centre for Independent Studies, 2014
This publication is available from the Centre for Independent Studies.
PO Box 92, St Leonards, NSW 1590 Australia • p: +61 2 9438 4377 f: +61 2 9439 7310 e: firstname.lastname@example.org