Content uploaded by Amanda-Jane George
Author content
All content in this area was uploaded by Amanda-Jane George on Jan 20, 2022
Content may be subject to copyright.
Forty Years of FOI: Accountability, Policy-making and The National Innovation and
Science Agenda
AJ George, Julie-Anne Tarr, Susan Bird*+
ABSTRACT:
Executive power in policy-making has been the subject of longstanding jurisprudential and political
debate. Innovation policies aimed at driving collaborative government and industry outcomes sit very
much at the intersection of this tension. In the 1990’s Christopher Arup highlighted legitimacy
concerns around ‘corporatist’ innovation policy involving greater government-corporate alliancing
and selective policy measures, nominating procedural reform and audits to check policy-making
power. However, the development of the National Innovation and Science Agenda shows these
mechanisms to be less than effective. With the passage of 40 years since the Freedom of Information
(‘FOI’) Bill was considered by a Senate Committee, it is timely to reconsider the role of public scrutiny
in policy-making. While increased scrutiny is at least part of the answer to better policy, the FOI
regime faces significant obstacles in achieving its objectives in the innovation policy space, if not at
broader levels within government policy development. Against the backdrop of recent calls for
greater confidentiality in the policy-making process, it is argued that increased secrecy is not the
answer.
1. INTRODUCTION
Executive power is an evolving topic of enduring interest. Former Chief Justice Robert French recently
noted ‘particular anxiety’ around the nature of executive power, especially non-statutory power.1 While
lawyers have long advocated for more effective fetters to avoid its ‘democratic deficit’,2 political scientists
believe it cannot be checked simply by regulation, given its deeply political origins.3 This article uses the
National Innovation and Science Agenda (NISA)4 as a case study to highlight issues around policy
development inherent in this larger theoretical debate.
*We gratefully acknowledge Mr Peter Timmins’ assistance around FOI processes and challenges in the
development of this article. We also thank our anonymous reviewers for their generous and insightful feedback in
developing the final version.
+Dr AJ George (LLM (London), PhD (UQ)), Senior Lecturer, CQUniversity; Prof Julie-Anne Tarr (PhD (UQ),
LLM (Monash), JD (Cornell)), QUT Business School; Dr Susan Bird (PhD (Victoria)), Lecturer, CQUniversity.
1 Robert French, ‘Executive Power In Australia – Nurtured And Bound In Anxiety’ (2018) 43(2) University of
Western Australia Law Review 16, 16; Cheryl Saunders, The Scope of Executive Power (Papers on Parliament No.
59, April 2013); Charles Lawson, Regulating Executive Power Under the Australian Commonwealth Framework
(Black Jettie, 2011). Cf. KM Hayne, ‘Executive Power’ (2017) 28 Public Law Review 236.
2 Margit Cohn, ‘Medieval Chains, Invisible Inks: On Non-Statutory Powers of the Executive’ (2005) 25(1) Oxford
Journal of Legal Studies 97, 104; see also Anne Cossins, ‘Revisiting Open Government: Recent Developments In
Shifting The Boundaries Of Government Secrecy Under Public Interest Immunity And Freedom Of Information
Law’ (1995) 23(2) Federal Law Review 226, 256-258.
3 Patrick Weller, ‘Parliamentary Accountability for Non-Statutory Executive Power: Impossible Dream or Realistic
Aspiration?’ (2005) 16 Public Law Review 314.
4 Commonwealth, National Innovation and Science Agenda Report: Welcome to the Ideas Boom, November 2015
(‘NISA’).
2
Policy-making is an integral part of the executive’s constitutional power to govern. Policies may be
proposed by a Minister or an agency, and many significant social policies have originated outside
government.5 However, it is the agencies that generally develop a policy, implement it, and evaluate its
performance. 6 Agency staff have the difficult mandate of serving the government of the day, yet remaining
apolitical.7 Policy development can thus be a challenging process that is impacted by relations between
Ministers, advisers, stakeholders, and agency staff, and the exigencies of party politics.
These dynamics were highlighted in the most recent review of the Australian Public Service (APS).
Consultation participants acknowledged the APS’ core business of policy-making, along with its heavy
responsibility in safeguarding the Australian national interest.8 However, agency staff keenly feel the
tension between this and serving the government interest.9 Ministerial respect for and adherence to due
Cabinet process when developing policy is critical as is, not uncommonly, fearless advice from
bureaucrats to Ministers throughout. Several key reports, including the recent APS review final report,
acknowledge this.10
Undoubtedly, success stories in Australian policy development are more the norm than the exception. As
Leutjens et al observe, ‘across many public policy domains, the bulk of public projects … perform not so
badly at all, and sometimes even highly successfully.’11 However, the last decade has been marked by a
number of high profile, large-scale public policy failures including the Home Insulation Program (HIP)
and National Broadband Network (NBN) policy.12 Calls to learn from these failures 13 have taken varied
forms within the media, academic research and political inquiries but with limited success if re-occurrence
of similar outcomes is the metric adopted.
Of relevance in the context of this paper are ineffective policy initiatives in the innovation space which
have arisen over the last 25 years, albeit on smaller scales than the HIP or NBN failures. Over this period,
5 Freedom of Information: Report by the Senate Standing Committee on Legal and Constitutional Affairs on the
Freedom of Information Bill 1978, and aspects of the Archives Bill 1978 (1979) (‘FOI Report’) 218.
6 Contracted service providers are often part of the implementation network, which raises its own issues: Myles
McGregor-Lowndes and Amanda McBratney, ‘Government Community Service Contracts: Restraining Abuse Of
Power’ (2011) 22 Public Law Review 279; Janet McLean, ‘The Crown in Contract and Administrative Law’
(2004) 24(1) Oxford Journal of Legal Studies 129.
7 APS Commissioner, Commissioner's Directions 2016 (2016) (‘Commissioner’s Directions’), Directions 13(j),
17(a).
8 Inside Policy, Report on the APS Independent Review Consultations (Stakeholder Consultations Report, 31
December 2018) 6, 9 (‘APS Consultations Report’) v, vi, 3, 6, 9, 23.
9 Ibid vi, 23.
10 See for example Peter Shergold, 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 (Independent
Review, 12 August 2015) v, 9, 25-30 (‘Shergold Report’); Department of the Prime Minister and Cabinet, Our
Public Service Our Future: Independent Review of the Australian Public Service (2019) 121 (‘Thodey Review’).
11 Joannah Luetjens, Michael Mintrom and Paul ‘t Hart (eds) Successful Public Policy: Lessons from Australia and
New Zealand (Australian National University Press, 2019) 3.
12 As discussed in Shergold (n 10).
13 Shergold (n 10); John Howard, ‘Public Policy Drift’ (Institute of Public Administration Australia Public Policy
Discussion Paper, 4 April 2012).
3
more than 150 reports14 and a plethora of policies have been introduced to improve outcomes in Australia’s
innovation performance, including the highly publicised NISA.
The NISA policy process was audited by the Australian National Audit Office (ANAO) in 2017.15
Findings highlighted difficult timeframes, and shortfalls in basic standards of evidence-based and
consultative policy development. Of concern to the auditors, the agencies conceded that little impact
would likely be registered from the $1.1 billion spent. The flawed policy-making process cost Australian
taxpayers dearly. Given the renewed focus on APS policy work, the difficulties in innovation policy-
making and the recent NISA experience are a worthwhile case study for examining the tensions around
checks on policy-making, and public scrutiny.
Christopher Arup’s seminal 1993 work on innovation law and policy16 identified various mechanisms to
curb innovation policy-making power, including audits and procedural reform. However, these
suggestions arguably belie a more fundamental issue: the dearth of effective public scrutiny at agency
level impedes sound, fearless advice, allowing ineffectively planned innovation policy to continue. The
Freedom of Information Act 1982 (Cth) (‘FOI Act’) – a key legislative response to this problem – is
therefore reviewed in this article through a socio-legal lens.17 It is contended, in this context, that despite
the passage of more than 40 years since a Senate Standing Committee (‘Committee’) first examined the
draft legislation18 it still faces significant obstacles in delivering on its intent – in the innovation policy
space, if not at broader levels within government policy development.
The paper commences in section 2 with an overview of the policy-making process which, in practice,
enjoys a low level of scrutiny from parliament and the courts. Section 3 reviews Arup’s early examination
of innovation policy-making,19 and the legitimacy problems that closer alliances between industry and
government may present. In this ‘corporatist’ realm, with little scrutiny, policy initiatives often fail to
achieve impact. Arup explores mechanisms to check this power in the wider administrative law
framework,20 identifying the possibility of audits and procedural reform.21
14 Howard Partners, Australia 2030 Prosperity Through Innovation: Report on the Analysis of Stakeholder
Consultations (2017) 3 (‘Howard Partners Report’), and the reports listed there.
15 Auditor-General of Australia, Design and Monitoring of the National Innovation and Science Agenda
(Australian National Audit Office (ANAO), Performance Report, Report No. 10 2017-2018, 27 September 2017)
(‘ANAO Report’).
16 Christopher Arup, Innovation, Policy and the Law (Cambridge University Press, 1993).
17 Consistently with Arup’s work, the socio-legal approach considers the practical operation and significance of the
law: Arup, ibid 5.
18 FOI Report (n 5).
19 Arup (n 16).
20 Lawson (n 1) 1, 27 denotes ‘core’ administrative law reforms as: Administrative Appeals Act 1975 (Cth),
Ombudsman Act 1976 (Cth), Administrative Decisions (Judicial Review) Act 1977 (Cth), Freedom of Information
Act 1982 (Cth), Privacy Act 1988 (Cth); here ‘administrative law framework’ includes the broader regulation of
agencies and the APS, including the Public Service Act 1999 (Cth) (‘PS Act’), the Public Governance,
Performance and Accountability Act 2013 (Cth), Auditor-General Act 1997 (Cth), Commissioner’s Directions,
agency directives and guidelines.
21 Arup (n 16) 58, 295, 305.
4
Section 4 discusses Arup’s mechanisms as exemplified by the NISA audit and its recommendation of
procedural reform. There are difficulties with these mechanisms. The prospect of audit scrutiny does not
appear to check innovation policy-making. Moreover, even where excellent policy-making procedures
exist they will be of little assistance if APS compliance, and its concomitant of fearless advice to Ministers,
cannot be secured. It is therefore posited that at least one response must be more effective public scrutiny,
to facilitate APS compliance with its administrative duties, including fearless advice.
Section 5 examines the current FOI regime, which faces significant challenges in facilitating public
scrutiny of policy-making. Section 6 discusses Chancellor Professor Peter Shergold’s Learning from
Failures report which suggests that the risk of policy-making failures may be lowered by amending the
FOI Act to strengthen secrecy for government decision-making. The report draws on notions of
Westminster government and fearless advice, reprising many arguments dismissed by the Committee in
1979. Section 7 discusses the various government reviews that have shaped progress in this area thus far.
The case study highlights the difficulties of the traditional Westminster response to policy-making;
confidentiality buttresses the sequestered space within which innovation policy-making failures may
continue. Generational change in the APS does not appear to have ameliorated the bureaucratic ‘culture
of secrecy’ 22 or ‘malaise’ regarding FOI matters.23 The ‘continuing resistance’ observed by the ALRC in
199524 is still evident today. It may be that strengthening the mandatory information publication scheme
provides a way forward.
2. EXECUTIVE POWER, POLICY-MAKING AND ACCOUNTABILITY
In Australia’s ‘evolving democratic experiment’25 the Constitution formally vests executive power in the
Queen, but it is exercised by the Governor-General with advice from the Federal Executive Council.26
However, the Council’s work is essentially to implement Cabinet and Ministerial recommendations, so in
reality, it is Cabinet (and the Prime Minister as Chair of Cabinet) that wields executive power.27 The
precise ambit of executive power, particularly non-statutory executive power, has long been debated and
22 Australian Law Reform Commission, Open Government: A Review Of The Federal Freedom Of Information Act
1982 (ALRC Report 77, December 1995) (‘Open Government Report’) [4.16].
23 Daniel Stewart, ‘Assessing Access to Information in Australia: The Impact Of Freedom Of Information Laws
On The Scrutiny And Operation Of The Commonwealth Government’ in John Wanna, Evert A Lindquist and
Penelope Marshall (eds), New Accountabilities, New Challenges (Australian National University Press, 2015) 140.
24 ALRC (n 22) 4.13.
25 Lawson (n 1) 3.
26 Commonwealth of Australia Constitution Act 1900 (Imp) 63 & 64 Vict, c 12 (‘Constitution’) s61-63, see also
Acts Interpretation Act 1901 (Cth) s 16A.
27 Lawson (n 1) 87; Catherine Althaus, Peter Bridgman, Glyn Davis, The Australian Policy Handbook (Allen &
Unwin, 6th ed, 2018) 175-176; Dan Meagher, Amelia Simpson, James Stephen Stellios, Fiona Wheeler, Peter
Hanks, Hanks Australian Constitutional Law: Materials and Commentary (Butterworths, 2016) (‘Hanks’) [7.4.9]-
[7.4.10C], citing FAI Insurances Ltd v Winneke (1982) 151 CLR 342, 382 (Aickin J).
5
continues to cause concern.28 For present purposes, it is sufficient to note that among these powers is the
government’s ability to engage in policy-making.
In our system of parliamentary supremacy, the legislature is traditionally viewed as a significant check on
policy-making power. The executive effectively controls the legislative program of parliament, but any
policy implemented by legislation must weather the scrutiny of both Houses of Parliament, particularly
the Senate, where a government may not hold a majority.29
Yet there are numerous ways to bypass the strictures of legislative scrutiny. For example, a government
may implement policy by simple financial appropriations, or cuts, which have ‘never been much of a
hurdle in the budget context’.30 A salient example is the Liberal government’s failed legislative attempt
to abolish the Office of the Australian Information Commissioner (‘OAIC’) in 2014, then the simple
removal of funding for it (later reinstated).31 Other avenues include policy implementation under ‘broad
permissive statutory frameworks’ and ‘informal administrative structures’.32 Political scientists are
generally sceptical of the ability to render the executive accountable to parliament.33
Policy-making, particularly when implemented by non-statutory means, also ‘minimises and partly
eradicates’ the possibility of judicial scrutiny.34 The Administrative Decisions (Judicial Review) Act 1977
(Cth) and Administrative Appeals Tribunal Act 1975 (Cth) apply to administrative decisions made
pursuant to statute rather than policy-making although, admittedly, there is no bright line between the
two.35 Other sources of jurisdiction in the Constitution and Judiciary Act 1903 (Cth) are confined to
‘matters’ where a court must establish an immediate right, duty or liability.36 If challengers to a policy do
not seek such a decision, the court’s jurisdiction is not enlivened. As Sir Gerard Brennan has noted, given
the concern for individual litigants’ interests, judges are ‘not very good at formulating or evaluating
policy’.37 Further, some decisions are simply not viewed as justiciable due to the constitutionally different
roles of the courts and executive. Where the exercise of power involves ‘complex policy questions’
28 See George Winterton, ‘The Limits and Use of Executive Power by Government’ (2003) 31(3) Federal Law
Review 421; French (n 1) 22. French notes some believe it now has little guidance as to boundaries: at 16, 32, 37;
see also Francis Gurry, ‘The Implementation of Policy Through Executive Action’ (1977) 11 Melbourne
University Law Review 189, 193. Cf. Hayne (n 1).
29 Althaus, Bridgman and Davis (n 27) 20; Hanks (n 27) [7.3.27C]; Cohn (n 2) 117.
30 Saunders (n 1); Arup (n 16) 226; Gurry (n 28) 199.
31 Maureen Henninger, ‘Reforms to Counter a Culture of Secrecy: Open Government in Australia’ (2018) 35
Government Information Quarterly 398, 402-403.
32 Arup (n 16) 226; see also Sue Taylor, Julie-Anne Tarr and Anthony Asher, ‘Australia’s Flawed Regulatory
Impact Statement (RIS) Process’ (2016) 44 Australian Business Law Review 361.
33 Weller (n 3) 324. See also Arup (n 16) 293; Hanks (n 27) [7.3.12].
34 Saunders (n 1).
35 Administrative Decisions (Judicial Review) Act 1977 (Cth), s3; Administrative Appeals Tribunal Act 1975 (Cth),
s 25; Peter Bayne, ‘The Court, the Parliament and the Government - Reflections on the Scope of Judicial Review’
[1991] Federal Law Review 1, 12.
36 In re Judiciary and Navigation Acts (1921) 29 CLR 257, 265; Judiciary Act 1903 (Cth) s 39B; Constitution ss
75(iii), 75(v); Chris Horan, ‘Judicial Review of Non-Statutory Executive Powers’ (2003) 31(3) Federal Law
Review 551.
37 As quoted in Bayne (n 35) 1.
6
Cabinet decisions have been held non-justiciable.38 So while the rule of law may be ‘strongly invoked’ in
other exercises of executive power, ‘the executive, as opposed to the judiciary … [possesses] a monopoly
of the policy-making function’.39
The next section discusses Arup’s concerns regarding the legitimacy of innovation policy-making given
the low levels of practical scrutiny. Arup presaged closer relations between state and industry, and the
ineffective nature of innovation policy measures, suggesting various checking mechanisms.
3. ARUP ON INNOVATION POLICY
Over 25 years ago, Arup observed the government’s adoption of innovation as a central policy concern in
Australia’s transformation from a resource-based to a ‘high-technology’ economy.40 Arup suggests that
the loosely regulated policy-making space allows government to pursue what he calls a ‘corporatist’41
approach; policy is freed from rule-based review through the use of statutory privileges, flexible
regulation, administrative action and collaborative decision-making.42 In this space, instead of the
traditional free market ‘hands off’ approach and primacy of the rule of law, government enlists the aid of
large companies, peak bodies, and trade associations to selectively ‘discipline industrial activity towards
modernising production and renewing growth’.43 Or, in the more exciting words of the NISA, to ‘drive
smart ideas that create business growth, local jobs and global success’. 44
Where corporatism is taken to extremes, Arup sees the law reduced to provision of a shell – property law,
contract law, law of associations – within which the government pulls its policy levers, ‘fashioning
working standards and individual decisions to further innovation objectives under fluctuating
conditions’.45 The line between public and private sectors is blurred. Power is shifted to the executive’s
discretionary programs over rule-based regimes.46 Apart from placing a large financial burden on the state,
legitimacy problems include allegations of favouritism and discrimination.47
Arup anticipates a constitutionalist call for innovation policy favours to be allocated legislatively, with
decisions to be measurable against legal criteria to open them up for review, to provide accountability to
parliament and the courts.48 However, in the end and despite some intellectual frustration, he concludes
that the liberal facilities of property and contract will co-exist in a complex interaction with the (then)
38 Minister for Arts Heritage and Environment v Peko-Wallsend (1987) 15 FCR 274, 278-9 (Bowen CJ). Sankey v
Whitlam (1978) 142 CLR 1 (‘Sankey’) made ‘inroads’ regarding common law judicial review, although policy is
still ‘at the heart’ of tensions between the judiciary and executive: Bayne (n 35) 6, 7.
39 McLean (n 6) 130.
40 Arup (n 16) 7.
41 Ibid 17, 31.
42 Ibid 50.
43 Ibid 25.
44 NISA (n 4) 1.
45 Arup (n 16) 50.
46 Ibid 56.
47 Ibid 29-30.
48 Ibid 31, 285, 293-294; see also Cohn (n 2).
7
newly emerging corporatism, its inducements and closer relations with interested parties.49 Perhaps
political scientists may conclude Arup simply acknowledged the growing drivers of regulatory evolution.
Yet Arup’s analysis of early innovation policy is a remarkably insightful depiction of innovation policy-
making as it has eventuated over the past quarter century. While the constitutionalist struggle against
corporatist policy-making failed to gain significant traction,50 Arup accurately notes the degree of
difficulty still encountered in the innovation sector: choosing the correct policy instruments often appears
to devolve to ‘picking winners’.51 Arup canvasses various positive adjustment measures (such as research
and development (R&D) grants and the R&D tax incentive), concluding that they generally seemed small-
scale and did not really influence change.52
With some ambivalence, Arup suggests a number of mechanisms to check executive policy-making
power, including the possibility of an Auditor-General’s performance audit,53 and procedural reforms to
inject ‘suitable standards of internal administrative rationality and participative democracy’.54
Interestingly, these mechanisms were exemplified in the 2017 Auditor-General’s performance audit of the
NISA. The next section discusses the NISA, its audit and outcomes.
4. DEVELOPMENT OF THE NATIONAL INNOVATION AND SCIENCE AGENDA
4.1 Policy proliferation, the rapid rise and fall of NISA
Innovation is still seen as a crucial driver of economic prosperity,55 and innovation policy-making has
proliferated in the last 25 years. However, as Arup predicted, it has largely been ineffective. Despite many
selective policy measures and hundreds of government-funded reports56 Australia has not achieved a more
competitive position on innovation indices.57 Indeed, there is little consensus even on indices, and a review
panel is examining just how Australia should assess its innovation performance.58 There is a vast business
social science literature on innovation policy, the disadvantages of traditional regulatory approaches and,
increasingly, the need to view and regulate innovation as a complex system.59 It poses a challenging and
49 Arup (n 16) 59.
50 Cf. Cohn (n 2).
51 Arup (n 16) 28, 233, 282. See also John Howard and Roy Green, Challenges for Australian Research &
Innovation (Discussion Paper, 10 April 2019) 12, 35.
52 Ibid 278; nb. Bill Ferris, Alan Finkel, John Fraser, Review of the R&D Tax Incentive (Review Report, 4 April
2016).
53 Arup (n 16) 295-296.
54 Ibid 59, 305.
55 NISA (n 4) 1; Innovation and Science Australia (ISA), Australia 2030: Prosperity Through Innovation (Policy
Framework, November 2017) (‘2030 Report’) 9.
56 Howard Partners Report (n 14) 3.
57 NISA (n 4) 1, 4, 10.
58 Department of Innovation, Industry and Science, Improving Innovation Indicators (Consultation Paper, March
2019). Final report yet to issue.
59 For a seminal discussion see Mark Dodgson, Alan Hughes, John Foster and Stan Metcalfe, ‘Systems Thinking,
Market Failure, and the Development of Innovation Policy: The Case Of Australia’ (2011) 40 Research Policy
1145; Howard (n 13). More recently, see 2030 Report (n 55) 21-22; ISA, Performance Review of the Australian
Innovation, Science and Research System (Report, December 2016) (‘ISA Performance Review’) 11.
8
entrenched policy problem; the level of complexity means it is ‘not feasible to map the … system in detail
and quantify all of the variables, interactions and feedback loops.’60
When former Prime Minister Malcolm Turnbull was appointed in September 2015, he immediately placed
innovation squarely on the political agenda.61 As for policy-making, he noted –
We need to restore traditional cabinet government. There must be an end to policy on the run and
captain’s calls. We need to be truly consultative with … the wider public. We need an open
government …62
Surprisingly, then, the NISA package was developed and launched by December 2015. One year on, the
government announced Australia had ‘come a long way’ and that the NISA was having ‘a significant
impact’.63 Another year, and work was underway on a different innovation policy framework (that largely
failed to gain traction),64 and Australian officials signalled to the International Monetary Fund (IMF) that
the NISA would be wound down.65 The IMF’s assessment, echoing Arup, was that the NISA was small-
scale and did not reflect a longer-term vision.66 Literature on the NISA was not favourable.67 The
government then advised that, after reassessment, it would pursue a ‘new policy focus’ on the space sector
and advanced manufacturing.68 The ‘violent swings’69 around corporatist innovation policy-making thus
seem largely unchecked.
4.2 What did the NISA promise?
The NISA promised an ideas boom, injecting $1.1 billion between 2015-2019. It comprised 24 measures
administered by nine different portfolios. Expenditure was divided between four ‘pillars’ as follows:
60 ISA Performance Review (n 59) 11.
61 Malcolm Turnbull (Prime Minister), ‘Changes to the Ministry’ (Media Transcript, 20 September 2015)
<https://www.malcolmturnbull.com.au/media/Ministry>.
62 Malcolm Turnbull, ‘Malcolm Turnbull’s Speech In Full: “We Need a New Style of Leadership”’, The Guardian
(Online News, 14 September 2015) emphasis added <https://www.theguardian.com/australia-
news/2015/sep/14/malcolm-turnbulls-speech-in-full-we-need-a-new-style-of-leadership>.
63 Greg Hunt and Malcolm Turnbull, ‘National Innovation and Science Agenda is Having a Significant Impact One
Year On’ (Media Release, 7 December 2016) <https://www.minister.industry.gov.au/ministers/hunt/media-
releases/national-innovation-and-science-agenda-having-significant-impact-one>.
64 The government accepted only 17 of 30 recommendations in the new proposed policy framework (2030 Report
(n 55)): Commonwealth, Australian Government Response to Innovation and Science Australia’s ‘Australia 2030:
Prosperity Through Innovation’ Report (May 2018).
65 International Monetary Fund, Australia: 2017 Article IV Consultation (IMF Country Report No. 18/44, February
2018) (‘IMF Report’) 22. Some NISA initiatives continue.
66 Ibid 20.
67 Sinclair Davidson and Jason Potts, ‘A New Institutional Approach to Innovation Policy’ (2016) 49(2) The
Australian Economic Review 200, 205; Yuliani Suseno and Craig Standing, ‘The Systems Perspective of National
Innovation Ecosystems’ (2018) 35 Systems Research and Behavioral Science 282; Harry Bloch and Mita
Bhattacharya, ‘Promotion of Innovation and Job Growth in Small- and Medium-Sized Enterprises in Australia:
Evidence and Policy Issues’ (2016) 49(2) The Australian Economic Review 192.
68 James Riley, ‘Labor “hypocrisy” on R&D: Andrews’, InnovationAus (Online Innovation News Forum, 10 May
2019) <https://www.innovationaus.com/2019/05/Labor-hypocrisy-on-RD-Andrews>.
69 FOI Report (n 5) 54.
9
Source: ANAO Adaptation of information from the National Innovation and Science Agenda70
Despite the large outlay, it should be emphasised the NISA package was significantly less than the 2015-
2016 government spend of around $5.751 billion on the R&D Tax Incentive Scheme and higher education
sector grant funding.71 So, as noted by the IMF, the NISA was small-scale and short-term, but in terms of
expenditure of public funds, it was considerable.
4.3 The Auditor-General’s Report
The Auditor-General subsequently conducted a performance audit of the Department of Prime Minister
and Cabinet (PM&C), the Department of Industry, Innovation and Science (DIIS) and Innovation and
Science Australia (ISA), to report on the policy design and monitoring process for the NISA.72 The report
is careful to point out it does not purport to examine the merits of policy positions, nor government
decisions. Nevertheless, the media was quick to discern its less than favourable assessment of agency
policy-making.73
Auditors assessed policy briefing documents and submissions against agency guidance material, the APS
Values,74 APS Commissioner’s Directions,75 policy procedure manuals and the like.76 In light of scale of
expenditure and expected impact, auditors found the material did not provide effective benchmarks for
the standard of analysis and evidence.77
70 Reproduced under Creative Commons Licence, Attribution-NonCommercial-NoDerivs 3.0 Australia (CC BY-
NC-ND 3.0 AU)) <https://creativecommons.org/licenses/by-nc-nd/3.0/au/>.
71 In 2015-2016 the R&D Tax Incentive scheme cost $2.951 billion in tax foregone: The Centre for International
Economics, R&D Tax Incentive Review (Final Report, 29 March 2016) 112. The government spent $2.8 billion in
research sector grant schemes: NISA (n 4) 3.
72 ANAO Report (n 15).
73 See for example Primrose Riordan, ‘PM’s “Agile” Agenda Fails the Auditor Test’, The Australian (Online
News, 28 September 2017) <https://www.theaustralian.com.au/nation/politics/prime-ministers-agile-innovation-
agenda-fails-the-auditor-test/news-story/9cfd806d3a08b22da624091be30adac0>.
74 Section 10 of the PS Act (n 20).
75 Commissioner’s Directions (n 7).
76 ANAO Report (n 15) 24.
77 Ibid 24-25.
10
The speed with which the NISA was developed also raised concern. PM&C could not justify 1) the short
timeframe, and 2) limited or entirely absent advice around implementation risks, governance, or
evaluation. Indeed,
… ‘[i]t was unclear how the timeframe … took account of the Department’s own better practice
guidance on the successful implementation of policy initiatives, which emphasises that
implementation should be considered at every stage of policy development.’78
This over-responsiveness to the Prime Ministerial agenda contributed to subsequent problems. Auditors
found a lack of evidence both for the four barriers/Pillars of NISA79 and the measures to address these
barriers. Much of the policy advice failed to answer even the most basic questions in policy development
templates; it was:
… general in nature and did not present quantitative or in-depth analysis of problems … A number
of proposals that involved significant expenditure aimed at transforming parts of the innovation
system relied on assertions rather than evidence.80
Evaluation problems then emerged. The NISA Taskforce did not include ‘any modelling, forecasting or
other analysis’ of expected impacts in the design phase.81 Moreover, PM&C conceded it would be hard to
find any authoritative way of modelling given the disparate nature of the NISA measures.82
Ultimately, the DIIS produced a complex map of ‘outcome chains’ to measure impact, but acknowledged
that many of the measures and effects were ‘relatively small’ and ‘unlikely to be captured at a whole-of-
economy (macro) level’. 83 PM&C also noted the NISA package would be ‘too small, even in aggregate,
for an economy-wide modelling or analytical framework to register an economic impact beyond the
materiality thresholds used by the Treasury’.84 These comments reflect the IMF’s later assessment.85
Opacity was another significant issue. Consultation occurred for some NISA measures,86 but in at least
one instance time pressure impacted on consultation levels. The ‘targeted consultation’ was adequate only
in light of the short timeframe and because some measures were canvassed in prior reviews and
consultations. However, while drawing on prior consultations to reduce stakeholder burden may be
appropriate, this presupposes a transparent link between stakeholders’ views and the proposed measures.
78 Ibid 21-22.
79 Ibid 26-27.
80 Ibid 28.
81 Ibid 27.
82 Ibid.
83 Ibid 41.
84 Ibid 27.
85 IMF Report (n 65) 20.
86 ANAO Report (n 15), 23, 30-31, 37.
11
Several NISA measures were not those recommended in earlier reviews,87 and the need for further
consultation was flagged in relation to others.88
As the final bottom line, auditors made only two recommendations: that DIIS make proper evaluation and
monitoring arrangements for NISA measures, and that both PM&C and DIIS revise their policy
development material.89 Recommendation 1(c) suggested that this improved policy material include
mechanisms designed to ensure its consistent application moving forward.90
Arup’s mechanisms thus appear to be less than effective. As to audits, as Arup (and the auditors) conceded,
auditors must appraise agency performance and financial efficiency, not the merits of government
policy.91 In any event, the prospect of an audit does not appear to check peremptory policy-making
conducted largely in secrecy. While the ANAO’s audit program is an active one, it devotes only 2% of
audits to policy development.92
As to procedural reform, the weakness in this mechanism is encapsulated in ANAO recommendation 1(c):
even best practice policy development guidelines will be ineffective absent APS compliance, and
provision of Ministerial advice that is quality, evidence-based, and fearless. Agency staff, under the
administrative framework, are of course already required to do this;93 but questionable selective policy
continues in practice.
John Howard, long-time consultant to government on innovation policy matters, also advocates procedural
reform via a 10-step ‘business case’ policy framework, on grounds that ‘changing guidelines, processes
and practices is usually easier than changing human nature’.94 However, as ‘[t]he cookbook does not create
the meal’,95 Howard advocates using the framework alongside measures such as APS capacity-building,
a focus on systems and design thinking, inclusiveness, openness, and engagement with stakeholders and
the wider community. Notably, if not with a level of irony, Howard’s brief analysis of a predecessor policy
to NISA is based only on a ‘desktop review of known circumstances’,96 because it was ‘unlikely [the
agency] would disclose their internal documents and deliberations’. The downside of greater stakeholder
collaboration with government, Howard cautions, is that it runs the risk of yielding evidence that is ‘open
87 Ibid 31.
88 Ibid 30.
89 Ibid 10, 29, 51.
90 Ibid 29.
91 Arup (n 16) 295-296.
92 ANAO, Annual Audit Work Program 2019-2020, (Web page, 1 July 2019) <https://www.anao.gov.au/work-
program/overview>.
93 The PS Act sets out the APS Values (s 10) and Code of Conduct (s 13), including the requirement to comply
with applicable Acts, instruments, and directions by authorities (ss 13(4), 13(5)). Commissioner’s Directions (n 7)
require compliance with laws, standards and the Code (14(d)) and evidence-based, frank advice (17) (in 2013,
Directions 1.3, 1.6; in 2004 2.5, 2.7).
94 Howard (n 13), 25.
95 Ibid, 4.
96 Ibid, 17.
12
to question’,97 susceptible to capture, and system gaming.98 In other words, the precise problems identified
by Arup in the corporatist approach to innovation policy-making.99
The complex, dynamic, challenging nature of the innovation policy ‘dance’100 – like policy-making
generally – necessitates a multi-faceted approach to reform.101 Calls for procedural reforms, APS capacity-
building, wider engagement, and (most recently) a growing focus on localism in policy solutions, are not
only welcome but coincide with both design thinking and a systems approach to policy-making.102
Nevertheless, it is posited that Howard’s ‘human nature’ conundrum demands these initiatives be
accompanied at all stages of policy-making by greater levels of scrutiny,103 with audit being only one
cross-check and point of accountability.
This approach can be seen in the 2019 call by Terry Moran (Chair of the Centre for Policy Development
and former secretary, PM&C) to expand our notion of democratic institutions beyond public institutions
to include ‘private and community institutions’ and also, importantly, to focus on the ‘accountability of
public administration and the institutions of which it is comprised.’104 Driving heightened levels of
scrutiny in this way delivers double value. At the policy front end, scrutiny incentivises APS capacity-
building and compliance with procedural reforms, as well as enhancing citizens’ trust in a more open
policy process. At the implementation/delivery end, it promotes stakeholder (and public) efficacy by
enhancing access to the information needed to assess performance and ask questions.
This approach also coalesces with a more corporatist policy-making framework around difficult areas such
as innovation. If both participation and accountability are extended to early-stage agency policy work as
well as later development of initiatives, selective measures may achieve greater legitimacy and
citizen/stakeholder buy-in. Examination of the current operation of the FOI regime is thus required. As
discussed subsequently, it faces significant practical challenges in delivering the requisite, intended levels
of public scrutiny.
5. PUBLIC SCRUTINY OF POLICY-MAKING POWER
There is a ‘complex tapestry’ of statutory and non-statutory regimes regulating access to government
documents,105 including the Cabinet convention of secrecy, although the latter has been increasingly
97 Ibid, 31.
98 Ibid, 32.
99 Ibid, 16.
100 Ibid, 5.
101 Shergold (n 10) (discussed below); Howard (n 13); Commonwealth, Independent Review of the Public Service
(Online Review Portal) <https://www.apsreview.gov.au/> (‘APS Review Portal’).
102 Terry Moran, ‘The Next Long Wave Of Reform – Where Will The Ideas Come From?’, The Mandarin (Online
News, 25 March 2019); Howard (n 13); Dodgson et al (n 59).
103 Howard (n 13), 28 suggests all phases of policy-making ‘be open, transparent, consultative and accountable’,
although these themes are not the focus of the paper; they are discussed in Attachment 1, 29-32.
104 Moran (n 102).
105 Moira Paterson, Freedom of Information and Privacy in Australia: Government and Information Access in the
Modern State (LexisNexis, 2005) [1.2]; Australian Law Reform Commission, Secrecy Laws and Open
Government in Australia (Report No. 112, December 2009).
13
questioned before the courts.106 This section focuses on the FOI regime, which since 2010107 has the object
of providing Australians with a ‘right of access to documents’ to increase ‘public participation in
Government processes’ and ‘scrutiny … of the Government’s activities’.108 This right is qualified by a
number of exemptions to allow Ministers and agencies to maintain (what has long been viewed as) a
necessary level of confidentiality in Westminster-styled Cabinet government.
Under s 34 of the FOI Act, Cabinet documents enjoy a ‘class’ exemption from access if they have the
dominant purpose of submission to Cabinet and were, or were proposed to be, so submitted.109 This first
exemption is formidable and reflects the long tradition of Cabinet confidentiality: it is not subject to a
public interest test because the test is said to be ‘implicit in the purpose of the exemption itself’. 110
Policy documents are likely protected under s 34 given that all major policy initiatives are referred to
Cabinet. Policy officers are therefore counselled by the PM&C to document the possibility of Cabinet
consideration, to allow FOI decision-makers to assess the applicability of the exemption.111 Purely factual
material is not exempt, but if reports are ‘shot through’ with policy considerations the exemption may
stand.112
The other exemption of interest is for ‘deliberative matter’. Sections 11A(5) and 47C of the FOI Act
comprise a ‘content’ exemption for documents disclosing matter involved in Ministerial or agency
deliberations, including policy-making and implementation.113 The 1979 Committee noted submissions
indicating this exemption could harbour ‘a vast potential for frustration’ of the legislation.114 It only
‘reluctantly’ concluded it be left in the Bill unchanged, placing great weight on proper application of its
public interest test.115 Accordingly, the Information Commissioner’s Freedom of Information Guidelines
(‘FOI Guidelines’) view it as confined to, for example, reflections on the ‘wisdom and expediency of a
proposal’.116 Access must be provided unless contrary to the public interest.117
106 Commonwealth v Northern Land Council (1993) 176 CLR 604, 615-616; Mark Rodrigues, Cabinet
Confidentiality (Parliamentary Background Notes, 28 May 2010).
107 Freedom of Information Amendment (Reform) Act 2010 (Cth), Schedule 1.
108 FOI Act, s 3.
109 FOI Act, s 34(1)-(3).
110 Commonwealth, FOI Guidelines: Guidelines Issued by the Australian Information Commissioner Under s 93A
of the Freedom of Information Act 1982 (January 2019) (‘FOI Guidelines’) [5.55]. However, the objects clause that
stated exemptions were necessary to protect ‘essential public interests’ was repealed in 2010. New Zealand’s
publication of Cabinet documents suggests otherwise: Rodrigues (n 106) 8-12; Robert Hazell and Ben Worthy,
‘Assessing the Freedom of Information’ (2010) 27 Government Information Quarterly 352, 358.
111 Department of the Prime Minister and Cabinet, FOI Guidance Notes (July 2011) 6-7.
112 Section 34(6), see Kim Rubenstein, ‘The Extended Reach of Cabinet Documents: Lessons from Victoria and
Queensland’ (1996) 3 Australian Journal of Administrative Law 134, 136.
113 FOI Guidelines (n 110) [6.60].
114 FOI Report (n 5) 213, 214.
115 Ibid 218.
116 FOI Guidelines (n 110) [6.57]-[6.58].
117 Section 11A(5); FOI Guidelines (n 110) [6.4]-[6.27]; see Simon Murray, ‘Freedom of Information Reform:
Does the New Public Interest Test for Conditionally Exempt Documents Signal the Death of ‘the Howard
14
Unlike other conditional exemptions no harm is specified in s 47C, but if no harm would eventuate,
disclosure is unlikely to offend the public interest.118 Also, the public interest test in s 11B does not specify
factors weighing against disclosure. This concerns some,119 but in practice the decision-maker must be
guided by the FOI Commissioner’s comprehensive list.120 Factors favouring access include that it would
promote the objects of the Act, inform debate on a matter of public importance, or promote oversight of
expenditure. Irrelevant factors include embarrassment to government, risk of misinterpretation, and
causing confusion or unnecessary debate.
These ‘irrelevant’ factors abrogate some of the public interest factors weighing against disclosure in Re
Howard and the Treasurer.121 However, two Howard factors – development of policy, and inhibition of
frankness and candour – remain.122 Discussion of these controversial factors is often intertwined, but
candour is unlikely to be upheld as the sole basis for non-disclosure.123 The FOI Guidelines state –
Agencies should start with the assumption that public servants are obliged by their position to
provide robust and frank advice at all times and that obligation will not be diminished by
transparency of government activities. … In particular, the FOI Act recognises that Australia’s
democracy is strengthened when the public is empowered to participate in Government processes
and scrutinise Government activities (s 3(2)). In this setting, transparency of the work of public
servants should be the accepted operating environment and fears about a lessening of frank and
candid advice correspondingly diminished.124
The 1979 Committee voiced similar sentiments, adding that increased scrutiny would lead to greater
recognition of the effectiveness of the APS, reduce distrust, and possibly lead to greater appreciation of
the role of public servants.125 It would also go some way towards countering Arup’s legitimacy questions
Factors’?’ (2012) 31(1) The University of Tasmania Law Review 58, 77; J R Henman, ‘The Urgent Need for
Reform of Freedom of Information in Australia’ (Conference Paper, Public Right to Know Conference, University
of Technology Sydney, 21 August 2004).
118 FOI Guidelines (n 110) [6.55]-[6.56].
119 Shergold Report (n 10) 22.
120 FOI Guidelines (n 110) [6.20]-[6.22]. Under s 11B(5) of the FOI Act agencies/Ministers ‘must have’ regard to
the FOI Guidelines.
121 [1985] AATA 100.
122 FOI Guidelines (n 110) [6.78]; Stewart (n 23) 114.
123 Candour must be related ‘to some particular practice, process, policy or program’: FOI Guidelines (n 110)
[6.81] (and authorities cited there), but for s 47C candour may ‘possibly’ be the sole factor if the public interest is
‘clearly, heavily weighted’ against disclosure or it impedes ‘efficient functioning of government’: at [6.82] (no
authorities cited). Candour must be approached ‘cautiously’ in accordance with ss 3 and 11B: at [6.85]. The ‘sole
factor’ concession was added in 2016. The 2011 Guidelines at [6.77] stated that both the policy and candour
grounds were ‘not, in those terms, consistent with the new objects clause … and the list of public interest factors
favouring access’. The 1979 Committee’s view was that, after Sankey (n 38), candour must be ‘seriously
question[ed]’ as a public interest consideration: FOI Report (n 5) 63, 221; Justice Mason in Sankey said the
possibility that disclosure would adversely impact candour was ‘so slight it may be ignored’ (at 40).
124 FOI Guidelines (n 110) [6.83]-[6.84], emphasis added. They clarify (to some extent) Bannister’s concerns:
Judith Bannister, Accountability or Participation? Disentangling the Rationales for FOI Access to Deliberative
Material (Adelaide Law School Research Paper No. 2016-12).
125 FOI Report (n 5) 26.
15
around corporatist innovation policy-making. Unfortunately, a narrow view of disclosure in the ‘public
interest’ persists in the senior APS, drawing on Westminster notions of responsible government, collective
responsibility, and anonymity.126 Chancellor Professor Shergold is a strong advocate of confidentiality in
this regard. 127
If these FOI exemptions do not apply, there are other avenues to refuse access to policy documents.
Agencies may rely on the ‘practical refusal’ ground in s 24(1) of the FOI Act, if the disclosure exercise
would ‘substantially and unreasonably divert the resources of the agency’.128 Controversially, this ground
was recently used to excuse agency non-compliance with its own mandatory disclosure log
requirements. 129 Otherwise, fees for access may be a sufficient deterrent for many seeking access to policy
documents.130 Henman discusses one notable example of a quoted charge of $605,284.72 for a single
application which, after negotiation, was reduced to $284.131 There is a reported increase in the number
of applications withdrawn for agencies with high quoted fees, with the suggestion that high quotes are
used as a deterrent.132 In 2017-2018, the government notified FOI fees of $383,531, but collected less than
one third ($115,863).133 Stewart notes that agencies prefer to rely on high fees rather than ‘practical
refusal’, but Moon notes these provisions may be used consecutively, causing significant delays.134
Timeliness itself is another problem, given the media (with its time-critical deadlines) is instrumental in
raising awareness of policy issues. The FOI Act requires a decision on access or review of refusal to be
made within 30 days,135 but this does not apply to reviews by the Information Commissioner. In 2018-
2019, the Commissioner received 928 applications for review (up 82% since 2015-2016);136 of the
completed reviews a significant 61.7% of agency decisions were set aside137 (up from 37% 2017-2018;138
22% in 2016-2017139). Average time to complete was 7.8 months140 (up from 6.7 months 2017-18; 6.2
126 Stewart (n 23) 114-115.
127 Shergold (n 10) 15-24, discussed below.
128 FOI Act, s 24AA(1)(a)(i).
129 Stephen Easton, ‘The Department of Energy Has Been Breaching the FOI Act for 10 Months Now’, The
Mandarin (Online News, 16 October 2019).
130 See ss 8D(4)-(5) 29, s 94; Freedom of Information (Charges) Regulations 2019 (Cth).
131 Henman (n 117) 63; Easton (n ) discusses a quote from the Department of Defence for $2,515 to disclose costs
of ministerial travel; searching for 45 hours, deliberating for 97 hours. Cost was ultimately reduced to zero.
132 Henman, ibid; Peter Timmins, ‘Submission to Allan Hawke’ (Submission to the Review of the Freedom of
Information Act 1982 and Australian Information Commissioner Act 2010 (‘Hawke Review’), 1 July 2013) 19
<https://www.ag.gov.au/Consultations/Documents/ReviewofFOIlaws/Hawke%20review.pdf>.
133 Commonwealth, Freedom of Information Statistics: FOI Requests, Costs and Charges 1982-2018
<https://data.gov.au/dataset/ds-dga-b0771c28-09cc-4c4e-9e61-9a96f6e3d040/details>.
134 Stewart (n 23) 146; Danielle Moon, ‘Freedom of Information: User Pays (and Still Faces Delays)’ (2018) 43(3)
Alternative Law Journal 192, 193-195.
135 FOI Act, ss 15(5)(b), 15(6) and (8), 15AA, 15AB and Part VII.
136 OAIC, Annual Report 2018-2019 (12 September 2019) (‘OAIC 2019 Report’) 77.
137 Ibid 78.
138 OAIC, Annual Report 2017-2018 (17 September 2018) (‘OAIC 2018 Report’) 77.
139 OAIC, Annual Report 2016-2017 (14 September 2017) 86.
140 OAIC 2019 report (n 136) 14.
16
months 2016-2017141). In 2017-2018, complaints rose 72%.142 They remain at this level, but are taking
longer to resolve (average 7.2 months, up from 5.8 months); 18% take more than 12 months.143 Complaints
focus on charging, practical refusal, timeliness, and poor service – most commonly a failure to reply to
correspondence.144 Pro-access commentators agree that the timeliness problem strikes at a key tenet of the
regime to facilitate informed representative democracy.145
Early ‘bureaucratic prophesies of doom’ regarding the FOI Act146 have not crystallised. Despite significant
revisions in 2010,147 the Act offers ample scope to maintain confidentiality of non-personal information
such as policy documents. In 1995, the ALRC noted FOI requests for policy information are the ‘real test’
of whether the regime increased government accountability and citizen participation – but its costs,
inconvenience and frustration resulted in few applications.148 In 2018-2019, there were 4,585 requests for
non-personal information (15.2% of all requests). Of these, 40% were refused, 22.8% were granted in full,
and 37.2% granted in part.149 This represents an improvement on the 54.1% refusal rate in 2017-2018,150
but little has changed over a five-year horizon. Despite the small proportion of non-personal requests, the
‘deliberative matter’ exemption was fifth most frequently claimed.151 So, while the FOI Act may ‘work
well’ to facilitate access to personal information, this is not so for policy-related information.152
141 OAIC 2018 Report (n 138) 14.
142 Ibid 83.
143 OAIC 2019 Report (n 136) 15.
144 Ibid 84.
145 Murray (n 117) 58, 77; Henman (n 117) 62, 63; Stephen Easton, ‘FOI Laws: Fixing the Chilling Effect on
Frank Advice’, The Mandarin (Online News, 18 June 2015), citing Peter Timmins’ view that, in view of timeliness
and other FOI issues at federal level ‘it’s pretty dark days … about FOI generally, and we’re yet to see a penny
drop with the government that this is a cause of real concern’ <https://www.themandarin.com.au/40043-abbott-
takes-secrecy-new-heights-public-servants-care/>.
146 Ernst Willheim, ‘Recollections of An Attorney-General’s Department Lawyer’ (2001) 8 Australian Journal of
Administrative Law 152, 157; see also Stewart (n 23) 97-98.
147 The 2010 legislation (n 107) revisions included a new objects clause, amended Cabinet documents exemption,
recasting of other exemptions; the Freedom of Information (Removal of Conclusive Certificates and Other
Measures) Act 2009 (Cth) removed conclusive certificates.
148 Open Government Report (n 22) [2.11].
149 OAIC 2019 Report (n 136) 173.
150 OAIC 2018 Report (n 138) 160.
151 OAIC 2019 Report (n 136) 176-178.
152 Stewart (n 23) 149.
17
One further difficulty is that the FOI regime is generally reactive. Critical policy decisions will usually
predate FOI requests. Made against existing non-disclosure provisions, formulating successful FOI
requests may become a de facto ‘needle in a haystack’ exercise. As an important counterpoint, the
information publication scheme was broadened in 2010 to move the FOI regime towards proactive ‘agency
driven publication’.153 Publication is mandatory for a range of material, to facilitate access to information
and potentially reduce the number of FOI requests.154 Yet, counterintuitively perhaps, the scheme has
narrowed requirements for publishing policy-related information: agencies now need only publish
sufficient details to allow public comment on ‘specific policy proposals’155 rather than, as previously,
details of ‘any arrangements that exist’ for public participation by way of consultations, representations
‘or otherwise’ in an agency’s policy work.156 This appears out of step with the general move to encourage
early stage public engagement in the policy process.
On a practical level, then, we have a cloistered policy-making environment. To the political scientist and
many Senior APS Executives this is entirely appropriate, given the turbulent context of policy
development. As Althaus, Bridgman and Davis note, ‘[t]he policy dance is sometimes seemingly random
… The interplay of politics, policy and administration is a hurly-burly, pulling sometimes this way,
sometimes that’.157 Policy success ‘is often built on paradox and good fortune’158 – conventions and
secrecy simply allow the executive the necessary flexibility to deal with the exigencies of ‘political
patterns’. 159
For the public lawyer and the electorate witnessing the growing number of large-scale policy failures,
however, the position may be more concerning. The next section examines a major and well received160
153 Explanatory Memorandum, Freedom of Information Amendment (Reform) Bill 2010 (Cth) 6.
154 Hawke Review (n 132) 98. Section 8(2) lists mandatory material; s 11C complements the IPS, mandating
publication of ‘disclosure logs’ of material accessed under the FOI regime (cf. Easton’s concerns (n 129)).
155 Section 8(2)(f).
156 Section 8(1)(a)(ii) of the 1982 legislation.
157 Althaus, Bridgman and Davis (n 27) 52; Luetjens, Mintrom and ‘t Hart (n 11) 10.
158 Althaus, Bridgman and Davis (n 27) 248.
159 Ibid 21.
160 Luetjens, Mintrom and ‘t Hart (n 11) 19.
0
2000
4000
6000
2014-2015 2015-2016 2016-2017 2017-2018 2018-2019
FOI non-personal request
outcomes 2014-2019
Total requests Refused Granted Granted in part
18
2015 report on large-scale policy failures, and the traditional Westminster-styled response, that policy
deliberations ‘need to be kept in confidence’.161
6. THE SHERGOLD REPORT
The Shergold Report, Learning from Failure, was commissioned to consolidate the lessons from the HIP
and NBN failures.162 Regarding the HIP ‘disaster’,163 Shergold notes ‘mistakes were manifold: … flawed
program design, rushed implementation and inadequate monitoring. … [T]he advice provided by public
servants to ministers was, in many instances, poorly given, poorly received and poorly communicated.
Consultation … was all but absent.’164 The Cabinet process was subverted, and the rush led to ‘crucial and
material compromises’ in design and implementation.165 The NBN failures also involved ‘closed-door’
policy design, no consultation with industry, no cost-benefit analysis or business case and poor
understanding of risk.166
Shergold makes 28 recommendations,167 many of which provide a keen vision for a more open,
collaborative, innovative APS. Of most interest for present purposes is Shergold’s discussion of the APS’
failure to provide robust policy advice. On this issue, Shergold highlights the need for ‘analytically
rigorous’ evidence-based policy-making, which should be informed by multiple perspectives, and that
implementation issues must not treated ‘as an afterthought’. He suggests consultation is often conceived
‘far too narrowly’, and encourages policy ‘co-design’.168 However, Shergold’s discussion of frank and
fearless advice, and the need for confidentiality, raises concern.
As a longstanding proponent of confidentiality in aid of frank advice, Shergold believes sufficient
accountability is achieved via Parliamentary committees, audits, the wider administrative law framework,
and the FOI Act.169 Controversially, his use of the old conclusive certificate process to prevent access to
deliberative matter went before the High Court.170 In Shergold v Tanner171 the High Court considered
whether the Federal Court had jurisdiction to review Shergold’s decision to sign certificates after a date
161 Shergold Report (n 10) 3.
162 Ibid.
163 Ibid 11.
164 Ibid 8.
165 Ibid 9.
166 Ibid 12.
167 Ibid iii-xi.
168 Shergold Report (n 10) 16-17; compare the Commissioner’s Directions (n 7) on APS Values.
169 Stewart (n 23) 129-130.
170 Under old s 36(3) (and s 33A(2)) of the FOI Act, abolished in 2009 (n 147).
171 Shergold v Tanner [2002] HCA 19.
19
for review by the AAT of the decision to deny access had been fixed.172 It was determined that it did,
although it was a pyrrhic victory.173 Tanner later withdrew from the dispute due to costs and delays. 174
In setting the context for confidentiality in the report, Shergold details the difficulties for public servants
in providing frank and fearless advice. Expediency under ministerial pressure may be appealing, but the
expectation that the APS serve the public interest requires ‘steely resolve’ – although, he adds,
obstructionism is not an option, either.175 He notes that Ministers should create the environment conducive
to encouraging and receiving the best possible advice,176 which should be written,177 and Secretaries
should be responsible for policy quality.178 And then – in a seemingly ambitious attempt to wind back
advances in the notion of open government – Shergold suggests the APS should be the beneficiary of
amendments to the FOI Act to increase the level of secrecy for agency advice.179 To this end, Shergold
makes a number of arguments around Westminster-style government that require closer analysis.
First, Shergold discusses the fragile trust between public servants and ministers, and the need to distance
the theoretically apolitical public servant from political debate (neutrality argument). Frankness in
sensitive policy advice, Shergold contends, is made much harder if public accessibility is part of the
equation (candour argument).180 If disclosure occurs, advice becomes a political issue, trust is eroded, and
public servants are dragged into a public political debate (need for anonymity argument). Public servants
will thus temper their advice and/or provide less considered oral versions away from public scrutiny.181
Clearly these are valid points of concern but they were also ones that were considered and directly rejected
for articulated reasons by the Senate Committee in 1979.182 There appears to be little reason to now doubt
the validity of either this body’s reasons or conclusion, even with the passage of time.
On neutrality, the 1979 Committee noted that no submission or witness had suggested the FOI legislation
would have any significant impact, and it shared this view.183 Subsequently, neither the 1995 ALRC
Report184 nor the 2013 Hawke Review185 suggested the FOI legislation impacted adversely on neutrality.
To the extent comparisons may be made with the United Kingdom, Hazell and Glover’s 2011 study of the
172 Ibid [8]. Earlier, the ALRC advised this practice was ‘an abuse of the certificate provisions’: Open Government
Report (n 22) [8.20].
173 Shergold v Tanner [2002] HCA 19, [40]: ‘the content of a requirement to provide natural justice to the person
aggrieved by the decision may be very limited’; see also Judith Bannister, ‘Case Notes: McKinnon v Secretary,
Department of Treasury’ (2006) 30(3) Melbourne University Law Review 961, 970.
174 Bannister (n 173) 970.
175 Shergold Report (n 10) 18.
176 Ibid 19.
177 Ibid 18, 19.
178 Ibid 19.
179 Ibid 20-22.
180 Ibid 20.
181 Ibid. Although this would contravene APS requirements in the broader administrative law framework.
182 FOI Report (n 5) 44-55, 214-218.
183 Ibid 46.
184 Open Government Report (n 22).
185 Hawke Review (n 132).
20
impact of FOI legislation on Whitehall confirms that civil service impartiality ‘has remained largely
intact.’186
As to anonymity, Shergold’s arguments were similarly made by witnesses before the 1979 Committee,
who posited a strained relationship between ministers and officials187 and tempered or reduced quality
advice.188 The Committee nevertheless resolved that ‘the political reality has far outdistanced the pristine
theory [of Westminster system anonymity]’.189 The names and views of many senior officials were aired
before parliamentary committees, and appeared in the media daily.190 Contemporary media practices and
senior bureaucrats’ increasing willingness to speak out on topical issues make this even more true today.191
The 1979 Committee also discounted candour as a convincing argument for non-disclosure under what is
now the deliberative matter FOI exemption.192 While disclosure may entail a more tempered style of
advice, and it may be less agile, the ‘specious or expedient advice … may well vanish’,193 thereby
improving quality.194
Indeed, the candour/confidentiality argument loses some of its force if the need is for quality advice rather
than blunt or ‘frank’ advice. Directions 13-17 of the APS Commissioner’s Directions now oblige the APS
to provide quality advice. The header reference in Direction 17 to ‘frank’ advice is not repeated in the
substance of the clearly drafted provision, which requires that advice be, among other things: non-partisan,
objective, evidence-based, relevant, comprehensive, unaffected by ‘fear of consequences’, and which does
not withhold ‘important facts or bad news’.195
In the same vein, the 2013 Hawke Review noted that while candour/confidentiality was not ruled out as a
consideration in the s 11B(4) public interest test for disclosure of s 47C deliberative matter, ‘officials
should be happy to publicly defend any advice given to a minister’ – if not, they should rethink it.196 One
caveat in this respect was extended to exemptions for incoming government briefs.197
186 Robert Hazell and Mark Glover, ‘The Impact of Freedom of Information on Whitehall’ (2011) 89(4) Public
Administration 1664, 1672-1673.
187 Ibid 50.
188 Ibid 51.
189 Ibid 49.
190 Ibid 49-50. See also Hazell and Glover (n 186) 1672.
191 Cf. OAIC, Disclosure of Public Servants’ Names and Contact Details (Discussion Paper, July 2019); Kieran
Pender, ‘“Silent Members of Society”? Public Servants and the Freedom of Political Communication in Australia’
(2018) 29 Public Law Review 327.
192 Ibid 216-218.
193 Ibid 53, 216.
194 Ibid 216.
195 Commissioner’s Direction 17(e) (n 7).
196 Hawke Review (n 132), 48.
197 An exemption was recommended for incoming government and minister briefs, question time and estimates
hearings briefings: Hawke Review (n 132), 49. In Crowe and Department of the Treasury [2013] AICmr 69 a
candour submission for an incoming government brief, for a party unable to form government, was upheld.
Influential factors included the ‘unique’ context of needing to establish immediate agency/ministerial rapport, the
immediate responsibility of an incoming minister for their portfolio, and the ‘special feature’ of a document written
21
Distilling the candour/confidentiality argument to its core contention better balances this debate. The
bottom line question is not whether, under the guise of Westminster tradition, the comfort of increased
secrecy will help public servants better negotiate political pressure to deliver better policy outcomes:
history seems to deny this given the present high level of confidentiality in policy-making, and continued
policy failures. Rather, the question is whether an increased prospect of disclosure might incentivise public
servants to comply with their administrative and procedural obligations, emboldening them to refuse
expediency in the face of ministerial pressure.198 Better policy outcomes arguably may flow as a
consequence along with, as noted, potentially greater trust and appreciation by the public.199
Shergold’s second argument revolves around inconvenience. He says openness and transparency must be
subject to the government’s requirement for confidentiality in policy-making because: ‘[n]ot to do so
burdens ministers and their advisers in a way that other decision-makers are not.’200 Shergold draws
support from the fact that CEOs are not required to disclose board deliberations, nor are courts expected
to circulate draft opinions or discussions prior to final judgment.
Although superficially compelling, such comparisons are fundamentally distinguishable on accountability
grounds from the position of the APS. Shareholders have a statutory remedy for inappropriate Board
conduct via ss 232 and 233 of the Corporations Act 2001 (Cth). Disappointed litigants have access to
appellate courts. Ministers are (at least theoretically) accountable to Parliament. Agency policy
development, in contrast, has little in the way of practical accountability.201 Although Shergold’s answer
is to consider the quality of departmental advice in a Secretary’s performance review,202 time limitations
cast doubt on the depth of examination possible.203
Shergold’s stance on the FOI ‘burden’ resonates with many senior bureaucrats.204 Following the Report,
one agency head publicly noted they would be in a better position to ‘manage’ the FOI process if
exemptions were widened, particularly to ward off media requests: ‘most of them are from journalists
looking for a story, which I think is not misusing the FOI Act, but is that its original intent?’205
for a one-person audience that would be less useful if released simultaneously to the public at large (at [85]). See
also Cornerstore Legal [2013] AICmr 71, upholding a candour argument for external administrators’ incident
reports. In both cases, it was held candour arguments will not be accepted simpliciter; ‘class’ claims must be
contextualised by the particular documents at issue.
198 FOI Report (n 5) 26, 53.
199 Ibid 26, 218.
200 Shergold Report (n 10) 20, emphasis added.
201 Particularly given the demise of individual ministerial responsibility for agencies: FOI Report (n 5) 25-26, 39.
202 Plus strengthened ministerial responsibility, and adherence to Cabinet process: Shergold Report (n 10) 19.
203 Stephen Bartos, ‘The Shergold Report: Freedom from Political Mischief Can Trump Freedom of Information’,
The Sydney Morning Herald (Online News, 27 February 2016) <https://www.smh.com.au/public-service/the-
shergold-report-freedom-from-political-mischief-can-trump-freedom-of-information-20160227-gn5e6q.html>.
204 Stephen Easton, ‘Lonely Voice Challenges Top Mandarins Over ‘Open By Default’’, The Mandarin (Online
Public Sector Forum, 13 April 2016) <https://www.themandarin.com.au/63147-lonely-voice-challenges-top-
mandarins-open-by-default/>.
205 Ibid, emphasis added.
22
Media use of the FOI system as part of publication chains is well known, as is its anathema status for the
more risk averse ranks of senior APS, who may be tempted to ‘wrongly’ label the system ‘a diversion’.206
The media, however, is often the principal awareness-raising mechanism207 for a public that typically has
less time, money and motivation to spend negotiating a system that is often fraught with delay and
obfuscation. Indeed, the Australian media has initiated a Right to Know campaign calling for FOI reform,
which is ‘unusual’ in both its scale and broad-based support.208 Any further restriction of access to non-
personal information by amending an FOI system already facing significant challenges in achieving its
intent is, at least for the public lawyer, problematic.
The bureaucratic commentary above also raises further considerations. Firstly, the fact that senior
bureaucrats feel free to publicly articulate this tends to conflict with Shergold’s argument on the need for
historical Westminster system anonymity and shielding from public debate. Second, it defeats Shergold’s
point that a government policy decision once made devolves to the public servant’s role of implementing,
not questioning: ‘even if the Secretary believes the government is acting unwisely, the answer is
necessarily, ‘Yes, Minister’’.209 As Timmins notes, FOI ‘is still the law and public servants should respect
it as such’.210
On the issue of inconvenience, it is sufficient to note the 1979 Committee’s compelling conclusion:
The political system, whatever its form or nature, should exist to one end only: not the convenience
of the government, but the service of the people. To this end, no views about the supposed nature
of the Westminster system should prevent the strengthening of the accountability of all parts of
the government to the people from being achieved …211
Nevertheless, Shergold concludes the post-2010 FOI Act is a ‘significant barrier to frank written advice’212
– even though both HIP and NBN occurred prior to the 2010 revisions.213 He suggests a statutory
‘rebalancing’ exercise, 214 including within the Act variables such as a list of factors weighing against
disclosure (including candour), and/or an explicit candour exemption, and/or clarification of the harm that
the s 47C deliberative matter exemption is intended to avoid.215 Shergold draws some support from the
206 Hawke Review (n 132) 85.
207 Ibid 105.
208 David Crowe and Jenny Noyes, ‘Campaign for the Right to Know Fights the Darkness’ The Sydney Morning
Herald (Online News, 20 October 2019) <https://www.smh.com.au/politics/federal/campaign-for-the-right-to-
know-fights-the-darkness-20191020-p532gq.html>.
209 Shergold Report (n 10) 18.
210 Peter Timmins’ views, reported in Easton (n 204).
211 FOI Report (n 5) 55, emphasis added.
212 Shergold Report (n 10) 21.
213 Ibid.
214 Ibid 22. On the dangers of the ‘balancing’ concept in the FOI context, see McKinnon v Secretary, Department
of Treasury [2006] HCA 45, [19] (Gleeson CJ, Kirby J).
215 Ibid.
23
Hawke Review on the latter point although, as noted above, any wider support from that Review is
doubtful;216 his position also conflicts with the conclusions of the 1979 Committee.
Shergold concedes that ‘placing restrictions on freedom of information is extraordinarily sensitive’, but
his solution is revealing: that a ‘bipartisan group of former ministers, together with former Secretaries’ be
appointed to investigate and report to government on options and approaches.217 Such proposals lend
weight to Hazell and Glover’s concerns, that:
FOI can be seen as a ‘general interest’ reform … Its benefits occur to a diffuse group (the public)
and the costs affect a concentrated group (civil servants). Following this logic, it is ‘remarkable’
that FOI was enacted, but likely that it will not be ‘durable’, as the concentrated interest will
mobilize more effectively than the diffuse interest. 218
The Shergold Report, and its positive reception among senior public servants, suggests that the prevailing
view of bureaucrats on confidentiality around policy-making has not changed much since the Committee’s
1979 report. One can sympathise to some extent with the public servant’s plight in the face of difficult
Ministerial pressure and the exigencies of party politics. Divided loyalties to government and the national
interest have long been cited as the breeding ground for ambivalence towards the FOI regime.219
However, given the significant level of practical confidentiality already available, the risk of large-scale
public policy failure like HIP and NBN – or on a lesser scale in the innovation sector, the NISA – will not
be diminished by increasing statutory options for agency secrecy. New mechanisms to maintain the status
quo will not produce change. As Grube and Howard observe:
In a sense, Westminster has always acted as something of a façade – a veil behind which to hide
all the complexity, duplicity, and political difficulties of a parliamentary government.220
Perhaps polemically to Westminster traditionalists it is posited that the 1979 Committee’s analysis still
holds true. Increased public scrutiny and a greater commitment to open government must form part of the
strategy for better innovation policy-making. It is not contended here that FOI reform is a panacea for all
policy-making ills. The complex and challenging nature of policy-making – or at least innovation policy-
making – demands a multi-faceted approach. The many reviews and reports to date in this area
demonstrate the magnitude of the challenge.
216 Hawke Review (n 132) 48.
217 Shergold Report (n 10) 23, emphasis added.
218 Hazell and Glover (n 186) 1666.
219 Open Government Report (n 22) [4.13].
220 Dennis C Grube and Cosmo Howard, ‘Is the Westminster System Broken Beyond Repair?’ (2016) 29
Governance 467, 478.
24
7. GOVERNMENT, APS REVIEWS
The Shergold Report is just one in a long history of calls for APS reform. The ALRC’s 1995 Open
Government Report advocated, among other things, more action to ‘dismantle the culture of secrecy’ in
the APS.221 Successive reviews have envisioned a revitalised APS, allowing citizens to become active
participants rather than passive recipients,222 but as the Government 2.0 Report noted, this requires
‘[l]eadership and policy and governance changes … to shift public sector culture and practice to make
government information more accessible and usable, make government more consultative, participatory
and transparent’.223
In 2010 the Gillard Government responded to the Government 2.0 Report with its Declaration of Open
Government, promising strengthened rights of access to information, a pro-disclosure culture, and a
collaborative government.224 FOI reforms followed that year, but as noted the regime still faces significant
challenges, particularly regarding a pro-disclosure culture.225 While the Open Government Partnership
recently reported on the need for greater public participation in government deliberations,226 and a
proposed Engagement Hub is to ‘inform’ the government’s commitments in its Second Open Government
National Action Plan 2018-2020, there will be no central repository of government-commissioned or
conducted research, due to ‘varying policy and program priorities’.227
Similarly, the collaborative, evidence-based and accountable policy development culture may be
embedded in the administrative law framework, but policy development efforts such as the NISA indicate
it has not yet translated broadly into practice. As the Government 2.0 Report noted, there is now no
shortage of ‘invitations’ for citizen involvement; the problem is the lack of agency response in ways that
actually demonstrate appreciation of public involvement.228 As to procedural reform, both the Scales
Review and NISA audit noted the mere existence of guidance documents ‘does not provide an assurance
221 Open Government Report (n 22) 7.
222 Advisory Group on Reform of Australian Government Administration, Ahead of the Game: Blueprint for the
Reform of Australian Government Administration (March 2010) (‘Moran Review’) 38; Government 2.0 Taskforce,
Engage: Getting on with Government 2.0 (22 December 2009) (‘Government 2.0 Report’) 2. At least 18 reports
have issued 2003-2018, which informed the 2019 APS Review: Department of the Prime Minister and Cabinet,
APS Review: Priorities for Change (19 March 2019) (‘Priorities for Change Paper’) 54-55.
223 Government 2.0 Report, ibid.
224 Lindsay Tanner, Declaration of Open Government (media release, 16 July 2010)
<https://www.finance.gov.au/blog/2010/07/16/declaration-open-government/>.
225 Stewart (n 23) 151.
226 Open Government Partnership Practice Group on Dialogue and Deliberation, ‘Deliberation: Getting Policy-
Making Out From Behind Closed Doors’ (The Deliberation Series Volume I, May 2019) (‘OGP Deliberation
Series’) <https://www.opengovpartnership.org/wp-content/uploads/2019/06/Deliberation_Getting-Policy-Making-
Out_20190517.pdf>.
227 Commonwealth, Australian Government Feedback on Commitment Ideas for Australia’s Second National
Action Plan 2018-20 (Government Response, September 2018) 5; PM&C, ‘Enhance Public Engagement Skills in
the Public Service’, Second Open Government National Action Plan 2018-2020 (Government Commitment
Dashboard) <https://ogpau.pmc.gov.au/commitment/enhance-public-engagement-skills-public-service>.
228 Government 2.0 Report (n 222), 2.
25
of their use’ (especially in the absence of fearlessness bolstered by scrutiny). Again, cultural change is
required.229
The most recent review of the APS,230 chaired by David Thodey, was launched in 2018. The Thodey
Review’s Priorities for Change report acknowledged the APS is still perceived as a risk-averse closed
book.231 Consultation participants ‘overwhelmingly’ indicated that the overarching purpose of the APS is
serving both the public interest and the government of the day, developing and implementing evidence-
informed policy, and that APS interactions with the public should be ‘transparent, ethical and
accountable’. 232 The Review explored ways to empower senior officers to ‘lead by example in setting an
“openness by default” culture’,233 but as one stakeholder commented, “‘[o]penness by default’ is
completely antithetical to the mindset of senior officers’.234
Early concerns were raised regarding the independence of the Review’s panel; it was housed within
PM&C and its head was a Deputy Secretary appointed by the PM&C Secretary.235 The panel was to be
‘shadowed by an advisory group of current ministers’, and it was to report not directly to the Prime
Minister but to him through PM&C.236 Gourley put it in ‘starker terms’: it was like the Banking Royal
Commissioner Kenneth Hayne being assisted by senior officers from the banks under investigation.237
FOI concerns were then raised when the Thodey panel stated it would explore ‘the extent to which the
FOI regime is helping the APS balance openness with the importance of providing frank and fearless
advice to government’.238 This of course implies that fearless advice requires secrecy, in the time-
honoured Westminster tradition. The Shergold Report informed the review process.239
In response, a submission by two respected former secretaries suggested that claims about the negative
impact of FOI ‘seem to be exaggerated’, and that the Review should carefully examine whether the risk
averse APS attitude was due more to pressures from Ministers and advisers than to concerns about the
229 Bill Scales, Independent Audit: NBN Public Policy Processes (Audit Report, 25 July 2014) 68.
230 Thodey Review (n 10).
231 Priorities for Change Paper (n 222) 46.
232 Inside Policy, An Independent Review of the Australian Public Service: A Detailed Consultation Report
(Stakeholder Consultations Report, 3 December 2018) 3, 6.
233 Ibid.
234 Anonymous (Stakeholder comment, 21 March 2019) <https://contribute.apsreview.gov.au/transparency-and-
accountability/list>.
235 Paddy Gourley, ‘The “Independent” Review So Botched it Will Need to be Reviewed’, Sydney Morning Herald
(online news website, 4 June 2018) <https://www.smh.com.au/politics/federal/the-independent-review-so-botched-
it-will-need-to-be-reviewed-20180531-p4zio7.html>. See the leadership and governance arrangements outlined in
the Thodey Review (n 10) 15.
236 Gourley, ibid.
237 Ibid.
238 Priorities for Change paper (n 222) 46.
239 Ibid 54.
26
FOI Act.240 Terry Moran (former secretary, PM&C) went further, suggesting the FOI regime be reviewed
to reduce exemptions, and mandate publication of policy ‘business cases’.241
Moran’s suggestions are consistent with Howard’s earlier proposals and merit further consideration. In
particular, the existing mechanism of the ‘proactive’ information publication scheme represents a potential
way forward that may mitigate the seemingly intractable cultural change issue within the APS, decrease
the need for ‘reactive’ policy-related FOI requests, and clear the way for more open, inclusive, and
engaged policy-making.
However, such suggestions were not discussed in the Thodey Review’s final report, which recommended:
1) establishing a Charter of Partnerships, to set expectations around public engagement and ‘promote an
open APS’;242 and 2) a review of FOI, privacy and record-keeping arrangements.243
The first recommendation suffered from a number of problems. While talk of government partnerships is
‘powerful and evocative’,244 the ‘language of partnership has multiple applications in policy rhetoric’.245
It can describe command and control arrangements, or informal networks and collaborations.246
Unfortunately, such arrangements generally do not operate as partnerships in practice.247
Another difficulty was that, while the Review acknowledged work on a different APS-wide engagement
framework by the DIIS (not PM&C), and launched a month earlier,248 the Charter of Partnerships would
have duplicated or replaced that project. Accordingly, the government did not accept the Thodey
recommendation.249 The APS will apply the DIIS framework, which was built on solid research
undertaken to fulfil a government commitment in the First Open Government National Action Plan 2016-
240 Andrew Podger and Helen Williams, ‘Response to APS Review “Priorities for Change”’ (Submission to APS
Review, 26 April 2019) <https://contribute.apsreview.gov.au/submissions/view/sbm093823c039df683ea4965>.
241 Terry Moran (n 102). Moran also advocates for a new integrity commission to examine maladministration and
deficient policy advice.
242 Thodey Review (n 10) 122.
243 Ibid.
244 Myles McGregor-Lowndes, ‘Is There Something Better Than Partnership?’ in Jo Barraket (ed), Strategic Issues
for the Not-For-Profit Sector (New South Wales Press, 2008), 68.
245 Jo Barraket, ‘Introduction’ in Jo Barraket, ibid 8, citing Stephen H Lindner, ‘Coming to Terms with the Public-
Private Partnership: a Grammar of Multiple Meanings’ (1999) 43(1) American Behavioral Scientist 35. See also
Geert R Teisman and Erik-Hans Klijn, ‘Partnership Arrangements: Government Rhetoric or Governance Scheme?’
(2002) 62(2) Public Administration Review 197.
246 Barraket, ibid.
247 See generally McGregor-Lowndes (n 244), Teisman and Klijn (n 245).
248 Thodey Review (n 10) 118, Priorities for Change Paper (n 222) 46; Department of Industry, Innovation and
Science, APS Framework for Engagement and Participation (2019) (‘APS Engagement Framework’)
<https://www.industry.gov.au/data-and-publications/aps-framework-for-engagement-and-participation>. See also
submission by Damian Carmichael, ‘Submission’ (Submission to Thodey Review)
<https://www.apsreview.gov.au/your-ideas/submissions/damian-carmichael>.
249 Commonwealth, Delivering for Australians, A World-Class Australian Public Service: The Government’s APS
Reform Agenda (2019), 17.
27
2018.250 Time will tell whether, and to what extent, the DIIS framework can drive better APS engagement
practices.
As to the second recommendation, the Thodey Review simply accepted the Shergold contention that ‘the
Commonwealth FOI laws now present a significant barrier to frank written advice’,251 citing anecdotal
evidence by ‘former ministers and senior public servants’.252 The Review concluded it was ‘critical’ to
make APS advice supporting the government’s deliberative processes confidential, and to provide an
exemption from release under FOI legislation.253 The recommendation of an FOI review naturally
followed. Commentators have noted the lack of any detailed examination of evidence by the Thodey
Review;254 the same sentiments were aired following release of the Shergold Report.255
The government has not enlivened the calls for an FOI review, responding that its ‘principal focus’ was
to ensure that agencies effectively implement current requirements and address practical problems.
However, it did not close the door on FOI amendment in future, noting that ‘[a]ny further reform … would
be considered separately.’256
Thus it seems as though the 40-year battle for acceptance of the FOI regime and its place in facilitating
public scrutiny of, and engagement in, Australian policy-making is not over yet. Against the Thodey
Review’s most recent call for further confidentiality, the media’s Right to Know campaign and Moran’s
suggestions provide salient counterpoints. Ultimately, if the NISA in the innovation sector is any
indication (along with the HIP, NBN), increased secrecy for the APS is not the answer to better policy
outcomes.
8. CONCLUSION
Experience over the last decades has evidenced that Australian executive policy-making power operates
in a cloistered environment. From a political science perspective this arguably aligns with the British
Westminster style of Cabinet government. Given the growing list of large-scale public policy failures,
however, public lawyers and the electorate may be concerned that the checking mechanisms in this area
may benefit from reconsideration. As has been argued, the FOI Act and other elements of the
administrative law framework are operating neither optimally nor in the manner intended, undermining
250 Commonwealth, Australia’s First Open Government National Action Plan 2016-18 (2016) 10, 59-61. A review
of the new APS Engagement Framework is beyond the scope of this paper. See also Carmichael (n 248); OGP
Deliberation Series (n 226).
251 Thodey Review (n 10) 121.
252 Ibid.
253 Ibid.
254 Peter Timmins, ‘Government Quick to Hose Down Thodey Call for FOI Reform’, Open and Shut (blog, 13
December 2019) <http://foi-privacy.blogspot.com/2019/12/government-quick-to-hose-down-
thodey.html#.XlOo9c4zY2y>. See also the submissions by Podger and Williams (n 240).
255 Paddy Gourley, ‘The Shergold Report: No Hollywood Adaptation for this B-grade Script’, Sydney Morning
Herald (online news website, 27 February 2016) <https://www.smh.com.au/public-service/the-shergold-report-no-
hollywood-adaptation-for-this-bgrade-script-20160226-gn4kt1.html>.
256 Commonwealth (n 249) 17.
28
the levels of public scrutiny that may have benefitted policy-making functions around initiatives such as
the HIP, NBN and NISA programs.
The innovation sector provides a useful case study to highlight this issue. In 2015, Prime Ministerial
speeches emphasised innovation as a central economic policy focus, and also drew a clear connection
between poor economic policy design, captain’s calls, and a closed government. However, innovation
policy-making continued its corporatist trajectory observed more than 25 years ago by Arup. The NISA
continued the small-scale, selective intervention approach, and the agencies conceded little benefit would
be registered from its disparate measures. The NISA audit revealed the now-familiar policy-making
deficiencies elucidated by the HIP and NBN program failures. The case study, set against the context of
the current operational problems in the FOI regime, and the Shergold and Thodey Reports calling for
increased secrecy, is concerning. Secrecy in innovation policy-making is an enduring policy problem, not
a cure.
In 1979, the Committee concluded that the Public Service Board’s approach to the ministerial-public
servant relationship in the Westminster system meant the FOI Act would:
… only be effective if the public service as a whole, and especially at senior level, is prepared to
adopt attitudes which are more conducive to the free exchange of information than has been the
case in the past. 257
Nevertheless, it also observed that the Westminster system ‘is neither so rigid nor so weak that it has failed
to accommodate change … we have seen the system changing in helpful ways.’258 It is hoped that the
DIIS’ new engagement framework will spur positive steps forward in this evolution. In any future review
of the FOI system, Moran’s reduction of exemptions deserves serious consideration, along with
amendment of the information publication scheme to require release of an agency’s policy ‘business
cases’, as a further means to forge a way forward.
257 FOI Report (n 5) 46.
258 Ibid 26.
