AustLII Home | Databases | WorldLII | Search | Feedback

Melbourne University Law Review

Melbourne Law School
You are here:  AustLII >> Databases >> Melbourne University Law Review >> 2021 >> [2021] MelbULawRw 4

Database Search | Name Search | Recent Articles | Noteup | LawCite | Author Info | Download | Help

Ng, Yee-Fui --- "Towards Constitutional Consonance: Institutional Adaptation and the Administrative State" [2021] MelbULawRw 4; (2021) 44(3) Melbourne University Law Review 889


Yee-Fui Ng[1]*

The contemporary executive in Australia is a large, unwieldy and complex beast — very different from what it was at Federation. Over the years, the structure of the executive has changed dramatically through devolution from traditional governmental departments to agency and corporate structures, as well as privatisation and contracting out. This process was accompanied by a proliferation of Ombudsmen, auditors and tribunals tasked with monitoring the executive. This article analyses how Australian parliamentary and judicial institutions have adapted to significant changes to executive structure and operation by reference to three critical junctures at which the Australian administrative state was significantly reconfigured: (1) the introduction of the ‘new administrative law’ package, including statutory judicial review, tribunals and oversight bodies such as the Commonwealth Ombudsman; (2) the advent of the ‘new public management’ movement that led to large-scale corporatisation, privatisation and outsourcing in the public sector; and (3) the recent trend of automation of government decision-making, which may pose new challenges to the
reviewability of government decisions.
This article shows that at the first critical juncture, the Australian judiciary and legislature adapted the increasingly unwieldy executive to the constitutional system through judicial interpretation that strictly separated Commonwealth and state tribunals and the courts, co-option of oversight bodies by the legislature as supports for responsible government, and constitutionalisation of judicial review. However, the seismic changes to executive structure and operation in the form of privatisation and outsourcing have only been partially adapted to the constitutional scheme, with a drastic reduction of legal accountability via judicial review. This article suggests that the latest challenge of technocratisation of the bureaucracy needs to be carefully monitored to ensure that automated systems in the Australian public sector are compliant with responsible government and legal


The contemporary executive in Australia is a large, unwieldy and complex

beast — very different from what it was at Federation. The framers of the Constitution in the late 1800s could not have imagined the size, scale, complexity and power of the new regulatory state. Over the years, the structure of the Australian executive has changed dramatically through devolution from traditional governmental departments to agency and corporate structures, as well as privatisation and contracting out.[1] More recently, the operation of government has been fundamentally transformed by the technological revolution, which has increased the prevalence of automation of government decision-making.[2] The question is: how have Australian parliamentary and judicial institutions adapted to seismic changes in executive structure and operation over the last few decades to achieve consonance with constitutional principles?

To address these questions, this article will consider how the Australian legislature and judiciary have adapted to three critical junctures that have significantly reconfigured the structure and operation of the executive: (1) the introduction of the ‘new administrative law’ package, including statutory judicial review, tribunals and oversight bodies such as the Commonwealth Ombudsman; (2) the advent of the ‘new public management’ movement that led to large-scale corporatisation, privatisation and outsourcing in the public sector; and (3) the recent trend of automated government decision-making.

As we enter a new decade, it is timely to undertake a broad ranging retrospective stocktake of the evolution of the structures and operations of the modern administrative state, focusing on the adaptability of parliamentary and judicial institutions in building ‘constitutional thought, values, concerns, and law into highly varied day-to-day administrative practices’[3] across various critical junctures in the last four decades. Although there is a well-established literature on the Australian executive,[4] much of this is doctrinal, focusing on case law rather than the broader canvas of Australian public administration. There has not been a broad ranging analysis of how Parliament and the judiciary have adapted to evolving governmental structures and technological change. This article makes a novel contribution by shaping a reorientation of Australian public law scholarship away from a narrow doctrinal focus on the scope of judicial review and executive power, and towards larger questions of public administration and the administrative state within the context of the Constitution.

This article draws upon the theory of historical institutionalism to illuminate the ways in which the Australian legislature and judiciary have adapted to changes in the administrative state to achieve consonance with the constitutional system. Historical institutionalism is an intellectual tradition that considers how institutions shape long-term political development by examining how ‘temporal processes and events influence the origins and transformation of institutions that govern political and economic relations’.[5] In particular, the historical institutionalist theory of path dependence is instructive in explaining the causes and triggers of institutional change over the long run,[6] where contingent events may set in train institutional trajectories that have moderately deterministic properties.[7] A key concept is that of a ‘critical juncture’ — a compressed period of time, typically following an exogenous shock, during which a particular choice or event may lock in an institutional trajectory, making it difficult to change absent another exogenous shock to the system.[8] In providing a framework for analysing institutional adaptation over time, historical institutionalism may elucidate the trajectory taken by the Australian legislature and judiciary in adapting to fundamental changes in the administrative state. It should be noted that in undertaking this analysis of institutional adaptation, this article necessarily traverses an extensive history of Australian administrative law — a complex and contested terrain that may contain features that some commentators may disagree with. However, providing such a history is a necessary foundation for the analytical interventions that this article seeks to make.

Part II of this article will outline the minimalist constitutional framework of the executive, including the principle of responsible government. It will then analyse three critical junctures that have inflicted exogenous shocks to the composition, structure and operations of Australian institutions. The first period, outlined in Part III, was the introduction of the ‘new administrative law’ package in the 1970s that encouraged the proliferation of tribunals, auditors and Ombudsmen, and resulted in the constitutionalisation of Australian administrative law. These reforms were sparked by the recommendations in the reports of the Kerr and Bland Committees, which led to watershed reforms in Australian administrative law in the 1970s that increased the avenues of legal control of government action.[9] Part IV discusses the second critical juncture: the advent of the ‘new public management’ (‘NPM’) movement, which was part of a worldwide impetus for public sector reform that swept through democracies such as the United Kingdom (‘UK’), United States (‘US’), Australia and New Zealand in the 1980s, and sought to insert business-like principles into public administration.[10] The NPM movement significantly restructured the Australian administrative state, setting off a wave of privatisations, outsourcing, and contracting out of government services, which in turn led to a reduction of legal controls in government. In Part V, the article will examine the use of new technologies in government. Significant technological advances in artificial intelligence (‘AI’), machine learning and big data analytics over the last two decades have enabled the widespread automation of government decision-making, which arguably represents a new critical juncture. It is argued that, although the legislature and judiciary have successfully adapted oversight bodies, judicial review, and Commonwealth and state and territory tribunals into the constitutional fabric, the legislature’s and judiciary’s response to changes to the structure and organisation of the executive from the NPM movement has been somewhat less successful in preserving the principles of responsible government and legal accountability via judicial review. This suggests that the latest challenge of technocratisation of the bureaucracy, which represents a fundamental change in governmental operations, needs to be carefully monitored to ensure that automated systems in the Australian public sector are compliant with responsible government and enable individual legal redress.

II The Constitutional Framework of the Executive

When the framers gathered to draft the executive chapter of the Constitution, they faced a conundrum. They were intent on incorporating the principle of responsible government into the Constitution.[11] Yet this would involve enshrining, in fixed constitutional rules, practices relating to the selection, organisation and tenure of the executive that were, in the UK and colonial constitutions, left as unwritten ‘understandings’ — understandings that were themselves ‘not always understood’.[12] Rather than embarking on this highly fraught task, the framers drafted 10 sections that occupy barely one page of type, which were not intended to be taken literally, and did not reflect the practical reality of the

operation of the executive.[13]

The constitutional framework of the executive is hence characterised by large gaps and eloquent silences. The stipulations of the Constitution regarding the executive are found in ch II. According to s 61, the Commonwealth’s executive power is vested in the Queen and is exercisable by the Governor-General. However, apart from the reserve powers, the Governor-General exercises executive power on the advice of Ministers.[14] Thus, in practice, executive power is exercised by Ministers who administer government departments under s 64 of the Constitution.[15] Section 67 states that until Parliament otherwise provides, the Governor-General is able to appoint civil servants as ‘officers of the Executive Government of the Commonwealth’.[16] Public servants are distinguished from elected Members of Parliament by s 44, which prohibits a person holding any office of profit under the Crown (other than Ministers) from being Members of Parliament. Under s 64, the Governor-General is able to appoint Ministers to administer Commonwealth departments, and these Ministers must be elected to Parliament. These provisions are said to embody the concept of responsible government, which holds that the executive must be led by Ministers who command the support of a majority of the lower house of Parliament, and that those Ministers must answer to Parliament for all actions of the executive, both collectively, in relation to the actions of the government as a whole, and individually, for what is done by the departments they control.[17]

The only executive institutions contemplated by the Constitution are thus: the Governor-General, who exercises the executive power of the Commonwealth on behalf of the Queen (s 61); and Ministers of State, who are appointed by the Governor-General to administer departments of State (s 64) staffed by public servants (s 67). These provisions may look unproblematic, but if we attempt to read them literally and in light of the notion of responsible government, difficulties emerge. The Constitution does not contemplate any machinery other than ministerially administered departments through which the executive power of the Governor-General may be exercised. Since colonial times, however, the Australian Parliament has routinely passed legislation conferring functions that are neither legislative nor judicial in character — and thus are prima facie executive — not on the Governor-General, nor on Ministers appointed under s 64, nor on officials within departments, but on statutory oversight bodies with legal personality separate from that of the Commonwealth and not under direct ministerial control.[18] In addition, the complex administrative apparatus that later emerged in the 1970s did not fit within the neat schema envisaged by the Constitution. How can these practices be squared with ss 61 and 64 of the Constitution?[19]

The general answer is that responsible government was designed to be a malleable concept that could adapt to suit the political exigencies of the day, meaning that its content and characteristics could change over time depending on law, convention and political practice.[20] Using this rationale, Saunders has argued that ‘[w]hile the provisions of the Constitution fix some aspects of responsible government, they leave considerable freedom to the Commonwealth in structuring the executive branch’.[21] Likewise, a High Court majority has held that the executive should be given the flexibility to organise its own administrative arrangements in the interests of good government, including appointing multiple Ministers to administer a department, and that permitting such flexibility would involve no derogation from responsible government as the Ministers would remain answerable to Parliament.[22] This suggests that so long as responsible government is preserved, the executive is able to flexibly structure its administrative arrangements, including by introducing new political


The next questions are: how were these extra-constitutional institutions adapted to the existing constitutional structure? What was their position

in relation to a constitutional structure that envisaged a tidy tripartite separation of powers? How did various exogenous shocks affect Australian

institutions — and were the resultant changes consistent with the key principle of responsible government? The next Parts will examine the three critical junctures, at which major shocks occurred to the structure and operations of the executive, and analyse how the legislature and judiciary have adapted these new institutions and developments, with varying degrees of success, into the

constitutional structure.

III The 1970s as a Baseline for the Australian Administrative Apparatus

The 1970s was a watershed moment for Australian administrative law, with the introduction of the ‘new administrative law’ package incorporating the establishment of a generalist administrative tribunal, the office of the Commonwealth Ombudsman, statutory judicial review, and later, in the 1980s, freedom of information (‘FOI’) and privacy legislation.[23] These reforms were introduced following the recommendations of the Kerr and Bland Committee reports, with the aim of increasing the opportunities and avenues for citizens to challenge governmental action, and thereby enhance the accountability of the executive, and promote good decision-making within government.[24] Dennis Pearce hailed the new system as the ‘vision splendid of the means by which an affected citizen [would] be able to test Commonwealth government decisions’.[25]

The introduction of fledgling new institutions, such as tribunals and Ombudsmen, recalibrated the existing system and created a brand new trajectory for Australian institutions. In this vein, it is instructive to adopt the historical institutionalist theory of ‘path dependence’, which explains the causes and triggers of institutional change over the long run, and in which contingent events may set in train institutional trajectories that have moderately deterministic properties.[26] As foreshadowed in the introduction, a key concept in historical institutionalism is that of a ‘critical juncture’ — a compressed period of time, typically following an exogenous shock, during which a particular choice or event may lead to what is called ‘increasing returns’ that lock in an institutional trajectory, making it difficult to change absent another exogenous shock to the system.[27] One of the more important mechanisms of the increasing returns processes is said to be the legitimation of past contingent choices through their acceptance by significant social actors and the general public.[28]

In the language of historical institutionalism, the introduction of the ‘new administrative law’ package was a critical juncture, that is, a reasonably compressed period of time during which divergent institutional trajectories opened up, none of which were determined by the existing legal tradition. There was no indication in the Constitution that Ombudsmen or tribunals would exist, much less how they would function. There needed to be new rules for these institutions, and a reconfiguration of the existing system to accommodate them.

A Administrative Tribunals

Although administrative tribunals have become integral to Australia’s system of administrative law, they are not explicitly mentioned in the Constitution, which only envisages ch III courts as part of the judiciary. This critical juncture opened up several divergent potential institutional trajectories — one that judicialised tribunals and treated them as courts, another that did not draw a strong distinction between tribunals and courts, and finally one that made such a distinction. The first option eventuated in the UK, where tribunals have gradually become judicialised through integration into the court system to form ‘a single twin-track system of adjudicatory institutions headed by the two highest central courts’.[29] Thus, tribunal members in the UK are seen to be part of the judiciary and can exercise judicial functions. The US adopted the second approach where tribunals are situated within the executive, but are seen to

exercise judicial functions.[30]

The Australian solution, which emerged when administrative tribunals were introduced and was solidified incrementally through case law, was the third option: drawing a strong line that explicitly distinguished tribunals from courts, both in function and powers. In 1971, Kerr’s report gave birth to a ‘truly novel’ body: the Administrative Appeals Tribunal (‘AAT’), which enabled review of government decisions over a diverse range of areas as a peak generalist tribunal.[31] The Kerr and Bland Committees also conceived of the new and unique concept of ‘merits review’, which was exercisable by tribunals as part of the executive,[32] as opposed to judicial review, which was exercisable by courts. Essentially this means that the tribunal would make the ‘correct or preferable’ decision de novo standing in the shoes of the decision-maker on the basis of up-to-date facts.[33] This led to Australia’s distinctive tribunal structure, within which Australian courts have held that tribunals are constitutionally different from courts and are thus unable to exercise federal judicial power.[34] Due to this differentiation in form and function, in Australia, tribunals are seen to be part of the executive branch of government.[35]

Thus, the Australian Parliament put into place a unique new generalist tribunal, which was touted as structurally and functionally different from courts. This in turn led to a path dependency that separated tribunals from courts. This institutional choice was then legitimised by its acceptance by significant social actors such as the courts and the general public. Following the introduction of the AAT, the Australian courts, with their strict doctrine of separation of federal judicial power, locked in a trajectory that inexorably led to tribunals being precluded from exercising federal judicial power. According to this orthodoxy, Commonwealth tribunals are not ch III courts under the Constitution, and are unable to exercise federal judicial power.[36] The position of state tribunals is more vexed, with an increasing preponderance of recent case law following a similar path to the federal sphere in deciding that state administrative tribunals (for example the Victorian Civil and Administrative Tribunal and New South Wales Civil and Administrative Tribunal) are not ‘courts of the state’, with the outlier being the Queensland Civil and Administrative Tribunal, whose constitutive statute designates it as a ‘court of record’.[37]

Australian tribunals are thus seen to exercise administrative functions, that is, merits review of administrative decisions. Nevertheless, tribunals have judicial characteristics and the ‘trappings of a court’,[38] as they conduct adjudicative proceedings utilising judicial processes, apply legal standards prescribed by relevant legislation to facts, and have the power to make orders that can be enforceable when registered in courts.[39] However, there is an inherent tension between tribunals being part of the executive branch of government on the one hand, and their aim to be fair and credible in their independent review of decisions on the other. There is a risk that the government may seek to control them, particularly in controversial areas such as the migration jurisdiction, where Ministers have the power to issue directions to tribunals.[40]

Therefore, the adaptation of tribunals within the constitutional system was done by structurally and functionally differentiating tribunals from courts within the constraints of the tripartite separation of powers, with tribunals situated in ch II and courts situated in ch III. The introduction of the AAT and the brand new concept of merits review set in train a strong trajectory that led to tribunals being categorised as part of the executive branch of government despite their quasi-judicial functions. This contingent institutional choice was reinforced and legitimised by subsequent court decisions establishing a strict demarcation of judicial power, leading to a solidification of the definitive distinction between tribunals and courts at the federal level, and an increasing recent trend of case law that disallows state tribunals from exercising federal

judicial power.

B Oversight Bodies as Delegates of Parliament

In contrast to the trajectory of administrative tribunals, the institution of the Commonwealth Ombudsman in the 1970s took a vastly different path. The original vision of the Commonwealth Ombudsman by the Bland Committee located the Ombudsman ‘outside the legal system of administrative review, oriented towards resolution of individual complaints and generally better at swatting flies than hunting lions’.[41] The Kerr Committee’s rejected ‘General Counsel of Grievances’ was likewise focused on individual complaint resolution.[42] Over the years, however, the Commonwealth Ombudsman has drifted from its original focus on granular individual complaints-handling towards a broader aim of investigation and audit of systemic issues within government administration;[43] in other words, successive Ombudsmen have moved closer to

‘hunting lions’.

Statutory oversight bodies have had a long history in Australia, existing since colonial times and in early Federation. The most significant of these is the Auditor-General, created in 1901,[44] whose audits of departmental and agency spending, encompassing both regularity and quality of performance, are an essential support for parliamentary supervision of public finance. The introduction of the Commonwealth Ombudsman joins this collective of oversight bodies set up to scrutinise governmental activity.[45] Following the introduction of the Ombudsman, there has been a proliferation of other oversight bodies and office-holders to monitor the executive. For example, Information Commissioners have been introduced more recently to police freedom of information regimes, while the states have introduced anti-corruption bodies tasked with investigating corrupt conduct by government officials.[46] Parliamentary budget officers play an important role in furnishing Parliament with independent analysis of the budget cycle, fiscal policy, and financial implications of policy proposals.[47] Within their legislative mandate, these watchdog bodies have a continuing function of review or enquiry in relation to particular aspects of executive activity. These arm’s length scrutineers are in effect given delegated authority by Parliament to oversee executive action and can be seen as alternate supports to responsible government. Oversight bodies undoubtedly contribute to political accountability of the executive, as an aspect of ‘monitory democracy’.[48] They have gained widespread public acceptance and their role is increasingly significant, particularly with the decline of Parliament as an effective locus of executive accountability: an effect of strong party discipline in Australia, which allows a majority government to easily pass legislation, notwithstanding even strenuous objections by the opposition and cross-benchers.[49]

As with tribunals, the existence of these oversight bodies was not envisaged in the Constitution, which has led to uncertainty about the appropriate operation of these bodies, and their position in the separation of powers. To enable them to perform their watchdog function effectively, oversight bodies require a degree of independence from the executive. However, oversight bodies in Australia have been attacked by some government administrations that have not been persuaded of their merits or enamoured of their activities through defunding, legislative amendment to reduce their powers, and even personal attacks, with a particularly ignominious example being the Australian Information Commissioner being defunded in 2014 to the extent that he had to shut down his office and work from home.[50] This antagonism has led to calls in Australia for the recognition of an ‘integrity branch’ of government.[51] One aim of such proposals is to underline the importance of the part played by bodies other than courts in the control of government.[52] Beyond that, such proposals might lead to the clarification and strengthening of the guarantees of their independence from government, if not to the creation or recognition of a distinct branch.

This article suggests that rather than the recognition of a new branch of government, in Australia we have seen the incremental co-option and supervision of oversight bodies by Parliament as an extension of their scrutiny functions, a phenomenon that is more prominent at the state level. Thus, the independence of oversight bodies from the executive has been enhanced through a closer relationship with Parliament in terms of appointments, legislative protections from removal, and parliamentary scrutiny. This is done through several mechanisms: the legislative designation of ‘officer of Parliament’,[53] the involvement by Parliament in the appointment processes of the integrity officers,[54] legislative protections from removal,[55] and supervision of their functions through parliamentary committees.[56] The term ‘officer of Parliament’, however, has not been defined in Australia and it is unclear what criterion is required to be an officer of Parliament and what additional status it confers beyond signifying a closer relationship with Parliament.[57] There is less attention given to the implications of such a title in Australia,[58] compared to the more systematic attention in Canada,[59] New Zealand,[60] and the UK.[61] In addition, because the Australian processes have been incremental, they have been implemented in a piecemeal

fashion inconsistently across oversight bodies and across the federation.[62]

Australian Parliaments have also undertaken direct supervision of the activities of oversight bodies. Particularly in the Australian states, the question of ‘second order accountability’ or of ‘who guards the guardians’ has attracted some attention with dedicated parliamentary committees, allied in some cases to inspectorates, set up in several states to scrutinise the work of integrity agencies.[63] These committees are generally tasked with reviewing the performance of the functions and exercise of powers of the integrity bodies, investigating complaints and conducting inquiries into the integrity bodies and referring any pertinent matters to Parliament. This may be coupled with periodic independent review of the operations of the integrity body by an independent auditor that reports to the committee. Yet Parliament is an imperfect source of accountability, as parliamentary committees may be focused on partisan issues or fiscal restraint, rather than accountability matters.[64] Oversight bodies, in other words, face two potential threats — from Parliament itself and from the

executive through Parliament.

Therefore, the trajectory of oversight bodies in Australia has diverged from that of tribunals. While the courts have been the initiators of the trajectory for tribunals that led to a strong line drawn to distinguish them from courts, Parliament has been the instigator of the piecemeal changes that have gradually co-opted oversight bodies to assist their scrutiny processes as part of the principle of responsible government. Parliament has also undertaken direct supervision of the activities of these oversight bodies. These changes have occurred incrementally and inconsistently across jurisdictions without an overarching framework. Nevertheless, we can see a gradual meandering in the direction of Parliament bringing these oversight bodies within their purview through direct supervision, designation as ‘officers of Parliament’, and legislative protection of their appointment and removal processes.

C Judicial Review

Another main element of the ‘new administrative law’ package was the introduction of statutory judicial review in the form of the Administrative Decisions (Judicial Review) Act 1977 (Cth) (‘ADJR Act’). The ADJR Act was intended to simplify the convoluted processes and remedies at common law by redefining the availability of judicial review to decisions of an administrative nature made under an enactment.[65] As Debbie Mortimer has explained, the intention of these reforms was to ‘de-constitutionalise’ administrative law away from the mechanism in s 75(v) of the Constitution through a more streamlined statutory process and enumerated grounds of review.[66] Initially the ‘de-constitutionalisation’ of administrative law was successful, with administrative law litigation proceeding almost exclusively under the ADJR Act in the first decade of its enactment, a body of jurisprudence developing based on each ground of review in s 5 of the ADJR Act and an increased volume of judicial review at the

federal level.[67]

However, in the 1990s, a divergent path developed, when governments hostile to asylum seekers enacted new draconian legislation that sought to override the jurisdiction of the courts through the mechanism of privative clauses, and excluded migration decisions from the ADJR Act. This in turn reinvigorated judicial review under s 75(v) of the Constitution due to the large volume of migration cases.[68] The courts have fought to protect their ability to grant legal redress through s 75(v).[69] The intensification of the battle between the courts and legislature over jurisdiction has led the courts to incrementally develop a new narrative for judicial review — one that inserted the Constitution into the centre of judicial review.[70] The re-constitutionalisation of administrative law in Australia culminated in the case of Plaintiff S157/2002 v Commonwealth, which affirmed the centrality and entrenchment of federal judicial review under s 75(v) of the Constitution based on the concept of the rule of law, and the rise of jurisdictional error as an ascendant concept.[71] The 2010 case of Kirk v Industrial Court of NSW held that judicial review for jurisdictional error is also entrenched at the state level;[72] bringing a symmetry to judicial review at the federal and state levels. The High Court’s reasoning for entrenching judicial review for jurisdictional error was the need to prevent the emergence of ‘islands of power immune from supervision and restraint’,[73] and ‘distorted positions’[74] isolated from the mainstream of Australian law. As Peter Cane observed, in a system of highly concentrated legislative and executive power such as Australia, ‘extensive immunity of a certain class of administrative decisions from judicial control would seriously disadvantage citizens against government’.[75]

The constitutionalisation of Australian administrative law is one of the most visible examples of adaptation of the administrative state to the constitutional scheme. The tussles between the judiciary and Parliament have led to the judiciary delineating the boundaries of judicial review jurisdiction which Parliament cannot trespass. Tied to this constitutionalisation is the expansion of the now constitutionally entrenched concept of jurisdictional error — with Australian courts finding new distinct grounds of review falling within this ambit, such as irrationality,[76] which is separate from the UK concept of Wednesbury unreasonableness,[77] meaning that irrational and unreasonable governmental decisions can be challenged by judicial review. This also indicates the Australian courts’ greater willingness to engage with the substantive reasoning of governmental decisions — skirting closer to the boundary between legality and merits.

In short, the positioning of new institutions into the constitutional structure has been a product of power struggles between the executive, Parliament and the judiciary over the last few decades. The upshot of the adaptation of the ‘new administrative law’ to the constitutional system is that the judiciary has slotted tribunals into the executive branch to protect federal judicial power and Parliament has initiated a gradual co-option of oversight bodies to assist them in undertaking their scrutiny functions, and has undertaken supervision of these bodies. Meanwhile, the courts have firmly wrested judicial review for

jurisdictional error into a protected constitutional position.

IV Privatisation and Corporatisation as Disruptive Forces

The next critical juncture in Australia was the seismic disruption wrought by the NPM movement in the 1980s, which fundamentally transformed the structure and operations of the executive by dispersing the distribution of public power to various entities within and outside of the executive.[78] NPM is premised on a reconceptualisation of the way governments should be managed, with an increased focus on reducing costs, managing programmes more efficiently, and making public service managers more accountable for results.[79] Although statutory authorities and government corporations have long existed in Australia,[80] the extensive restructuring of government due to the NPM movement and the shift of governmental policy towards corporatisation created more arms-length bodies, such as statutory authorities and government corporations, alongside a wave of privatisations in the 1990s.[81]

Following a period of fragmentation, the new millennium witnessed attempts to reverse the devolutionary tendencies of earlier reforms by focusing on ‘integrated governance’ and ‘whole-of-government’ initiatives,[82] and the rationalisation of the number of arms-length bodies from its zenith in the 1990s.[83] Despite this, there remain a large number of statutory authorities and

government corporations in Australia today.[84]

The traditional method of organising government is through departmental structures with Ministers, who are responsible to Parliament, at the apex. The funnelling of power and responsibility hierarchically through a single person, the Minister, was seen to be the best method of ensuring that administration was conducted according to the wishes of Parliament, as representatives of the electorate.[85] Thus, the Constitution only envisages a bilateral relationship

between Ministers and the departments they administer.

However, major changes to governmental structures in the last few decades have undermined the neat linear relationship by which Ministers are responsible to Parliament for a public service department. The restructuring of government through the NPM movement has resulted in public services being delivered through a bewildering array of arms-length statutory authorities, government companies, contractors and private bodies.[86]

The differentiation of governmental structures into various new legal forms (statutory authorities, government companies, contractors) raises questions as to where the boundaries of the executive lie and how responsible government is to be maintained over core governmental functions. The notion of responsible government only clearly applies where the executive has some level of ownership or control over the entity, as it is difficult for the executive to be responsible for an entity it cannot control. Privatisation cleanly slices off all government ownership and control over an entity, thus excluding the operation of responsible government completely. The government has no ownership over their contractors, and control has to be exerted through contractual mechanisms, leading to regulation primarily by contract and thus a more limited application of responsible government.[87] The government does retain some level of ownership and control over statutory authorities and government companies; thus, these entities can be brought within the schema of responsible government. In short, an emphasis on responsible government leads to a focus on ownership and control by the executive. This leads to the exclusion of privatised entities from its ambit, as well as the regulation of contractors by contractual

mechanisms rather than public law mechanisms.

One question is how the post-constitutional ‘agentification’ and corporatisation of government in the form of statutory authorities and government companies has been adapted to the constitutional structure, as the executive does retain a level of ownership and control over these bodies. This article argues that the Australian legislature has enfolded these arms-length bodies within the financial accountability framework through the Public Governance, Performance and Accountability Act 2013 (Cth) (‘PGPA Act’), a product of the evolution of financial management legislation over 30 years, which sets up a comprehensive framework for corporate governance and financial accountability for public sector entities in Australia. The PGPA Act distinguishes between Commonwealth companies subject to the disciplines of the Corporations

Act 2001 (Cth) (‘Corporations Act’) and all other Commonwealth entities, which are classed either as non-corporate Commonwealth entities — such as departments and non-statutory agencies that are legally part of the Commonwealth — and statutory corporations that, by reason of their legislative foundation, enjoy separate legal existence.[88]

The PGPA Act provides a comprehensive regime for the management of public funds, requiring departments, statutory authorities and government companies to be financially accountable to Parliament.[89] It imposes both ex ante and ex post accountability requirements on government entities that differ based on the legal and ownership status of the corporation, with the most onerous requirements applying to Commonwealth entities (including departments, statutory authorities and statutory corporations),[90] and descending tiers of requirements that apply to wholly-owned government companies and government-controlled companies.[91] For instance, Commonwealth entities must provide an annual performance statement to be tabled in Parliament.[92] Commonwealth entities must also prepare an annual report following each reporting period and provide it to their portfolio Minister, who must then table the report in Parliament.[93] Government owned and controlled companies must furnish their portfolio Minister with the company’s financial report, directors’ report, and auditor’s report as required under the Corporations Act, which the Minister must also table in Parliament.[94] The PGPA Act therefore gives flesh to the principle of responsible government in its prescriptive requirements for the executive’s financial accountability to Parliament.

However, this is not to be sanguine about the effect of the devolution of the executive (from traditional departmental structures to statutory authorities and government companies) on parliamentary accountability. According to the traditional concept of responsible government, there is a clear vertical line of authority by which the Minister is responsible to Parliament for the actions of their department. This tidy arrangement has been disrupted by the plethora of statutory authorities, government companies and contractors, thus dispersing responsibility among Ministers, heads of statutory bodies, boards of directors and contractors. This means that parliamentary accountability and thus responsible government can be subverted by the obfuscation of Ministers’ responsibilities due to the fragmentation of responsibility between Ministers and arms-length bodies. For instance, Graham Taylor remarked that the message that came through from ministerial replies in New Zealand in connection with the activities of arms-length bodies was that: ‘If it’s bad news it’s commercially sensitive and I’ll refuse to answer; if it’s good news or trivial it’s not sensitive and I’ll take the credit; if I’m not sure which it is I’ll refer it to the corporation’.[95]

Besides parliamentary accountability, the privatisation and corporatisation of government also created a new critical juncture for Australian courts, which have had to determine whether privatised and corporatised bodies would be subject to judicial review. There were two main options: (1) to look at the nature of the power (ie determine whether the function exercised was public in nature); or (2) to determine whether the source of the power was public, that is, derives from statute or the prerogative.[96] The UK adopted the ‘public function’ test, as exemplified by the case of R v Panel on Take-Overs and Mergers; Ex parte Datafin Plc (‘Datafin’), where judicial review hinged on the nature of the power being exercised (ie the public function of the body).[97] This meant that, in the UK, decisions of private bodies (including completely private companies without any link to government) could be subject to judicial review if they undertook a public function, such as an exercise of regulatory or statutory power.[98]

By contrast, the Australian High Court’s approach to judicial review has focused on the legal form of the body, rather than the function that the body undertakes (eg whether it is a ‘public function’). This is despite the Datafin precedent in the UK, and some level of flirtation with the Datafin test by Australian intermediate courts at the state level.[99] In part, this can be attributed to the fact that Australian courts were more constrained than the courts in the UK as to the path they could have taken, as the wording of the ADJR Act and s 75(v) of the Constitution dictated their approach to judicial review.

The ADJR Act only applies to decisions of an administrative character made under an enactment.[100] Due to the narrow reading by the High Court of the phrase ‘under an enactment’ in Griffith University v Tang, requiring that the decision be expressly or impliedly required or authorised by statute, and that it must itself confer, alter or otherwise affect legal rights and obligations,[101] many decisions of statutory authorities and government companies (and other public entities) will be excluded. This includes decisions made under contract (or even decisions made under a statutory power to contract).[102] The formalism of the High Court’s approach is demonstrated by its majority decision in NEAT Domestic Trading Pty Ltd v AWB Ltd, holding that a decision made by a private company, enjoying de facto monopoly power, that operated as a precondition of a legislatively-authorised administrative consent, was not made ‘under an enactment’ because the company’s power to make decisions derived not from the statute but from its incorporation and the applicable companies

legislation.[103] Judicial review under the ADJR Act was therefore unavailable.[104]

So far as judicial review under s 75(v) of the Constitution is concerned, corporate entities, including statutory corporations and government companies, are not susceptible to such review, because it is limited to acts of ‘officers of the Commonwealth’. This has been interpreted by the courts as requiring a formal appointment of a natural person, and excluding artificial persons.[105] A statutory authority, however, might be subject to judicial review under s 75(iii) of the Constitution where it is considered to be part of ‘the Commonwealth’,[106] with a remote possibility of extending this to wholly-owned government

companies — although the court has not made such an extension to date.[107] Statutory authorities will be considered to be part of the Commonwealth and thus susceptible to judicial review where the statute evinces an intention that the Commonwealth operates in a particular field through a body created for that purpose, which depends on the degree of control that the Commonwealth has over the body and the purpose of the body.[108]

Australian courts have therefore adopted a formalist approach that precludes both common law and statutory judicial review of decisions of corporatised entities such as government companies. This approach limits judicial review over statutory authorities, leading to great consternation among academic commentators due to the significant reduction of legal accountability.[109] The historical institutionalist approach may explain why Australian courts have taken this unfortunate direction. The High Court has long maintained a narrow interpretation of ‘officer of the Commonwealth’ under s 75(v) of the Constitution as excluding artificial and corporate entities,[110] which predated the advent of the privatisation and corporatisation of government. The precedential die was thus cast long before this critical juncture ensued. Likewise the wording of the ADJR Act enacted in the 1970s places an emphasis on the source of the power as its basis for jurisdiction for undertaking judicial review, that is, decisions have to be ‘under an enactment’. This led to a judicial trajectory that placed primacy on the legal source of power (ie that there needed to be statutory authority for the power). Section 75(iii) remains a terrain whose contours are yet to be clearly mapped; nevertheless, the concept of governmental control is the broadest test that allows certain statutory authorities to come within its ambit, though it is still narrower than the UK’s public function test.

Thus, the advent of the privatisation and corporatisation of government has had deep ramifications in terms of the legal control of government action. Australian courts had previously set in motion precedential legal reasoning that was difficult to diverge from, even following the seismic shifts in executive structure and operation due to widespread privatisation and corporatisation, which consequently entrenched the inability of individuals to access judicial review. This may indicate that the legislature should seek to more fully adapt the administrative scheme to the constitutional landscape to ensure that individual legal redress is maintained in the face of privatisation and corporatisation. Although decisions made by statutory authorities and government companies that are expressly authorised by statute may be subject to ADJR Act review,[111] only a few specific statutory schemes explicitly enable judicial redress[112] or tribunal review[113] against specific government companies, indicating that the legislature has only minimally facilitated the availability of legal review in limited and particular instances. A more complete solution would be to amend the ADJR Act such that the availability of review no longer depends solely on the decision being made under an enactment, and potentially incorporating other contextual factors such as the exercise of public power (such as regulatory power) or the government’s level of control over the body. The constitutional formulation in s 75(v) is difficult to amend, but the judiciary has the latitude to more clearly develop the jurisprudence under s 75(iii) to more broadly enable review over corporatised bodies controlled by government.

Although the prospects for judicial review in Australia are dismal due to privatisation and corporatisation, the legislature has adapted other aspects of the administrative law system to enable accountability for government contractors. For instance, the Australian Parliament has enacted legislation enabling the Commonwealth Ombudsman to investigate private entities that provide goods and services to the public under a contract with a government department or agency,[114] and the Auditor-General has, since 2011, been empowered to conduct performance audits of government contractors, with broad powers to compel the production of documents from Commonwealth contractors and to enter the premises of these contractors.[115] This enhanced external oversight has been accompanied by increased disclosure requirements. Ministers are required to disclose details of government contracts under a Senate Order,[116] while the Auditor-General is required to audit and report to Parliament annually any inappropriate use of confidentiality clauses in a sample of government contracts.[117] There is a public right to access information held by government contractors relating to government services through FOI legislation.[118] Despite this, FOI requests can be circumvented if a government overzealously utilises the ‘commercial in confidence’ exemption.[119]

In sum, the advent of the privatisation and corporatisation of government has posed major challenges to the effectiveness of mechanisms of legal and political accountability. Privatisation completely severs the link to government, leading to a separation from the tenets of responsible government. The fragmentation of the structure of the executive has also dispersed responsibility amongst statutory authorities, boards and government companies, thus providing opportunities for Ministers to evade parliamentary accountability by blaming others. Nevertheless, the legislature has partially integrated government contractors into the constitutional system by enabling oversight bodies to investigate government contractors, and subjecting government contractors to FOI requirements. The legislature has also integrated statutory authorities and government corporations into the system of responsible government by incorporating prescriptive mechanisms of financial accountability to Parliament for statutory authorities and government companies via the PGPA Act. However, the legal control of government action through judicial review has been adversely impacted due to judicial interpretation that excludes corporatised entities from the ambit of both statutory judicial review and s 75(v) of the Constitution, and limits availability of judicial review through s 75(iii), resulting in the broad-ranging circumvention of legal accountability.

V Automation as a New Grand Challenge

In the last couple of decades, a new grand challenge has arisen, which is to maintain accountability in the face of the government’s use of automated technologies in decision-making. As early as 2003, the Administrative Review Council listed a large range of major federal government agencies, including Comcare, the Department of Defence, the Department of Veterans’ Affairs, and the Australian Taxation Office, that made use of automated systems in governmental decision-making.[120] Since then, there have been major advances in

technology, including big data analytics, AI and machine learning, which provide new opportunities for government authorities to develop automated decision-making tools. The Australian public sector is already using technology-assisted decision-making in a wide range of contexts, with Centrelink’s

automated debt raising and recovery system[121] and the Australian Border Force’s SmartGate identity-checking at Australian airports[122] being some of the best-known examples.

The use of automation in the administrative processes can improve efficiency, certainty, predictability and consistency.[123] Automated systems have the capacity ‘to process large amounts of data more quickly, more reliably and less expensively than their human counterparts’ and have a useful role to play ‘when high frequency decisions need to be made by government’.[124] New technologies are also being assessed for their potential to improve access to justice and

provide support for marginalised groups.[125]

However, the use of emerging technologies is rarely unproblematic. The utilisation of automated decision-making raises a range of rule of law issues, in particular regarding procedural fairness,[126] the transparency of decision-making,[127] the protection of personal privacy,[128] and the right to equality.[129] The disastrous rollout of Centrelink’s online compliance initiative (dubbed ‘Robodebt’), in which errors of methodology of government decision-making resulted in incorrect or inflated debt calculations for over 200,000 individuals, is illustrative.[130] This involved an online machine learning method for raising and recovering social security overpayment debts, extrapolating from the Australian Taxation Office’s data matching information about the total amount and period over which employment income was earned and applying that average to every separate fortnightly rate calculation period for working age


Controversially, from July 2016, the online compliance scheme automatically targeted and raised debts for every case where the person could not disprove the possible overpayment, effectively shifting the onus of proof from the department to the individual.[132] These large-scale incorrect calculations have reduced public trust in computer-supported government decision-making and led to grave repercussions for vulnerable low-socioeconomic debtors, including individuals experiencing severe mental health issues, with reports of suicide in the affected population.[133] The Australian judiciary is thus faced with a new grand challenge in adapting technological decision-making to the administrative state. The courts have several options of determining legal reviewability: (1) examining the process of decision-making, eg requiring mental processes of deliberation; (2) considering the identity of the decision-maker; or (3) ascertaining the nature of the power, ie focusing on whether the function being performed is public in nature. As the identity of the decision-maker and process of decision-making is different for humans compared to AI, the third option is the most promising in allowing automated decision-making to fall within its ambit. For instance, as discussed above, the UK has adopted the ‘public function’ approach based on the Datafin test,[134] meaning that the key issue is whether a body is undertaking a public function, which would make it seem immaterial whether the decision is undertaken by a human or AI.

Despite the obvious attractions of the public function approach for enabling redress for technological decision-making, there are early indications that the Australian judiciary is adopting the first option of interpreting the ADJR Act by reference to the process of decision-making. The majority decision of the Full Federal Court in Pintarich v Deputy Commissioner of Taxation throws doubt on whether automated decisions are reviewable under the ADJR Act, as Moshinsky and Derrington JJ held that a ‘decision’ made under the ADJR Act has to involve a mental process of deliberation.[135] It is notable that Kerr J provided a significant dissent that recognised the difficulties in imposing a requirement that human mental processes need to be engaged for an act to be a ‘decision’ under the ADJR Act, particularly in the context of automated decision-making systems:

The hitherto expectation that a ‘decision’ will usually involve human mental processes of reaching a conclusion prior to an outcome being expressed by an overt act is being challenged by automated ‘intelligent’ decision-making systems that rely on algorithms to process applications and make decisions.
What was once inconceivable, that a complex decision might be made without any requirement of human mental processes is, for better or worse, rapidly becoming unexceptional. Automated systems are already routinely relied upon by a number of Australian government departments for bulk decision-making. Only on administrative (internal or external) and judicial review are humans

Justice Kerr objected that the legal conception of what constitutes a decision ‘cannot be static; it must comprehend that technology has altered how decisions are in fact made and that aspects of, or the entirety of, decision-making, can occur independently of human mental input’.[137] While this view is preferable, the High Court has refused special leave to appeal this case,[138] meaning that the majority’s decision remains the final word on the issue.

This may mean that if a human makes a decision guided or assisted by automated systems, this would still be a decision under the ADJR Act under the majority’s interpretation, as it would still involve a mental process of deliberation and cogitation by a human decision-maker. Where the decision is actually made by an automated machine without any human involvement, this may not constitute a decision under the ADJR Act, as the majority’s test presumes that a human brain is involved in a mental process. This may lead to a perverse incentive for departments and agencies to automate in order to avoid judicial review. Similar problems may also arise for tribunal review at the AAT if the same requirement of deliberation over a decision applies, where the enabling legislation provides for such review referring to the term ‘decision’,[139] although this has not been challenged to date.

Another potential avenue of challenge is via s 75(v) of the Constitution, which allows for judicial review in all matters where constitutional writs are sought against an officer of the Commonwealth. There is no issue if a computer is merely assisting a human, and the human, who is a Commonwealth officer such as a public servant, made the actual decision. However, if the computer made the actual decision, it is more difficult to argue that automated decisions fall within the scope of s 75(v), as courts have read in a requirement of a formal appointment of a natural person, and a prohibition against artificial persons.[140]

On the other hand, it may be argued that it is still possible that s 75(v) review would be available, but not for the decision itself. As observed by Yee-Fui Ng and Maria O’Sullivan:

[t]he focus of s 75(v) is on the decision-maker, rather than the method of decision-making (via a decision). It may thus be argued that the review exists for any actions that the Commonwealth officer tries to take acting in reliance on that decision (eg deducting payments from a pension following a computer ‘decision’ that a debt was owed).[141]

This suggests that as long as a human remains in the decision-making loop, challenges of automated decisions via s 75(v) may remain a viable option.[142] In short, it is still uncertain whether judicial review under s 75(v) of the Constitution will be available for automated decisions made by computers, and this issue remains to be clarified by case law.

Presuming there is jurisdiction for the court to adjudicate on an automated decision, the next question is which grounds of review might be utilised to challenge automated decisions. There are several possible grounds relating to the design of an automated decision-making process. For example, the automated system would need to be designed carefully so that the discretion of the decision-maker (if any) remains unfettered in exercising their power under the relevant legislation, policy or procedure.[143] Where the automated system was designed in a manner that included irrelevant considerations, those could be a ground for challenge as well.[144]

In addition, where interim steps in a decision-making process are automated, there may be potential for a decision to be affected by jurisdictional error where there is an error in the automation process.[145] For example, the Federal Court consent orders in the Robodebt litigation, Amato v Commonwealth, declared that an automated decision made on the basis of Robodebt methodology was irrational and thus unlawful.[146] There is a kernel of potential in the Court’s use of irrationality as a basis for their invalidation of the Robodebt decision; an irrational decision constitutes a jurisdictional error which, as discussed in Part III(C), is a constitutionally entrenched aspect of judicial review. This holds promise, suggesting future irrational automated decisions will be struck down by Australian courts. However, one should avoid undue

optimism — the consent order on the decision’s unlawfulness proceeded on the narrow basis of irrationality of the methodology of the decision-making, which means that future automated decisions made on the basis of different, more reliable data points may be held to be valid. The litigation did not raise the bar with regard to any procedural safeguards, such as by requiring more transparency or an explanation of the decision-making, which will affect the

opportunities of future debtors to seek redress.

Thus, where the court’s jurisdiction for judicial review is enlivened, there are a range of grounds that can be utilised to successfully challenge the decision. However, the prospects for passing the jurisdictional hurdle are not promising for the ADJR Act and are yet to be ascertained for s 75(v) of the Constitution.

Therefore, it can be seen that the automation of decisions is likely to preclude judicial review under the ADJR Act, and may possibly exclude judicial review under s 75(v) of the Constitution or tribunal review, leaving individuals unable to challenge automated decisions by government. This is because Australian administrative law has historically focused upon human decision-makers, as reflected in the framework for judicial review. In addition, as discussed above in Part IV, Australian courts have adopted a formalist rather than a purposive approach to interpreting the ADJR Act and s 75(v) of the Constitution. The courts’ requirement of a process of deliberation for a ‘decision’ to be made under the ADJR Act has left a vacuum where decisions are automated, while the narrow reading of s 75(v) of the Constitution to only include natural persons has excluded the ability of individuals to challenge the decisions of corporate and potentially technological entities.[147] By contrast, the UK courts’ focus on public function, rather than legal form and the requirements of a decision,[148] is more amenable to reviewability within a changing technological environment, as it focuses on the function being performed, rather than the legal form of the body or what constitutes a ‘decision’.

In contrast to the formalist approach of the courts, the Australian legislature has shown a proactive inclination to integrate automated decisions into the administrative system through the introduction of deeming provisions across a broad range of statutes, including migration, business registration and social security legislation.[149] As these provisions deem any decisions made by AI to be that of the legal decision-maker, it may be argued that Parliament intended to preserve review rights and to enable enforcement action for such automated decisions. For example, these provisions may potentially enable review under

s 75(v) by deeming that a decision effected by the use of AI is attributed to an individual such as the Secretary.

In addition, the Australian legislature has been active in instigating parliamentary committees to investigate automated government decision-making. Parliamentary committees have been vigilant in investigating the government’s use of automated systems in the areas of migration,[150] veterans’ affairs,[151] and social security.[152] For instance, the Senate Community Affairs References Committee closely examined the use of a targeted compliance framework involving automated sanctions that applied to more than 75,000 welfare recipients of parenting payments (‘ParentsNext’). These automated sanctions include payment suspensions that are automatically triggered by a person receiving demerit points, leaving some families without money for daily essentials.[153] Approximately 1 in 5 participants in the programme were subject to an automated sanction.[154] An intensive version of the ParentsNext programme specifically targets regions with high concentrations of Aboriginal and Torres Strait Islander populations.[155] The committee found that the programme was causing ‘anxiety,

distress and harm’ for many parents, including women escaping violence.[156]

Further, in 2017, the same Senate Committee raised early concerns about the rollout of Centrelink’s automated data matching system (ie Robodebt) regarding procedural fairness, the reversal of onus of proof and the design of the system of data matching that resulted in a discrepancy or error rate of 20%. In addition, the Senate Committee brought to light the difficulty experienced by recipients contacting the Department of Human Services to discuss their debt matter, with their calls going unanswered, and the emotional trauma, stress and humiliation faced by recipients.[157] The Committee recommended that the programme be put on hold until the procedural fairness flaws were addressed, and that the department resume full responsibility for calculating verifiable debts (including manual checking) relating to income support overpayments, which should be based on actual fortnightly earnings and not an assumed average.[158]

Other mechanisms of administrative law, such as oversight bodies, have also been vigilant in investigating the use of automated systems. The Commonwealth Ombudsman played an instrumental role in investigating the government’s deficiencies in administering the Robodebt system, with reports in 2017[159] and 2019[160] on this issue. The Ombudsman’s 2017 report identified issues with the ‘fairness, transparency and usability of the online system’, and found that ‘many of these issues could have been avoided by better project management, design, user testing and support for users of the online system’.[161] The Ombudsman’s follow-up investigation in 2019 found that the Departments of Social Services and Human Services had made significant progress in implementing their recommendations.[162] These illustrate the Ombudsman’s strong monitoring role in investigating the Robodebt issue and scrutinising the departments’ implementation of its recommendations, which has resulted in changes in departmental practices.

The widespread use of automation in governmental welfare programmes points to a broader issue in terms of the ‘digital welfare state’.[163] In a targeted welfare system, a heavy-handed approach that adopts automated systems, rather than personalised processes with appropriate oversight, adversely affects vulnerable people the most, as they are most likely to be in the cohort of social security recipients. The United Nations Special Rapporteur on extreme poverty and human rights, Professor Philip Alston, warned about the risks of the digital welfare state, where ‘digital data and technologies ... are used to automate,

predict, identify, surveil, detect, target and punish’.[164]

This trend towards the use of automated technologies to target vulnerable populations points towards the need for both individual redress and independent scrutiny over punitive governmental programmes. The challenge of adapting automated decisions into the administrative state is thus a paramount issue for our time. Technological decision-making is an important issue that will continue to gain increasing prominence in the future as government authorities are increasing their use of automation. The Commonwealth government has adopted a digital transformation strategy that aims to use automated systems, where possible, to eliminate manual processing and case management, reducing the need for bespoke systems.[165] The New South Wales government has said it will start to ‘[t]est AI/cognitive/machine learning for service improvement’ and implement ‘[f]ull automation where appropriate’.[166]

Although it is still too early to tell how the Australian legislature and judiciary will fully respond to the increased use of automation in government decision-making, the Robodebt controversy and budding case law provide a few early indications. It appears that the Australian courts will continue on their previous formalist trajectory in interpreting the ADJR Act and s 75(v) of the Constitution, which may well pose issues for judicial review of automated government decision-making, and may also impact negatively upon merits review. The historical institutionalist approach indicates that the path dependency of the Australian courts that was exhibited during the critical juncture of privatisation and outsourcing is likely to continue in the new era of automation, resulting in a lack of judicial redress for automated decisions and thus a further diminution of legal control of government. If the jurisdictional hurdles are surmounted, in terms of grounds of judicial review, the potential use of jurisdictional error to invalidate automated decisions opens up the possibility for individual legal redress, given the constitutionalisation of judicial review for jurisdictional error discussed above in Part III(C). By contrast to the courts, the Australian legislature has been proactive in integrating automated decision-making into the administrative state through the introduction of legislative deeming provisions and the active scrutiny of automated government decision-making through parliamentary committees. In addition, oversight bodies such as the Commonwealth Ombudsman have played an active role in scrutinising automated government decisions. In a context where the government is intensifying its use of automation to achieve efficiency gains, it is necessary to vigilantly monitor these technological developments to ensure that they are consistent with the principles of responsible government and legal accountability.

VI Conclusion

The spartan and sparse nature of the executive chapter in the Constitution belies a complexity in the structure, organisation and control of the modern executive. The modern regulatory state has increased in complexity over the last few decades, with a range of new bodies including tribunals, Ombudsmen, commissioners and auditors, and the significant restructuring of the executive from traditional departmental structures to a range of statutory authorities and government companies, alongside a raft of privatisations and outsourcing of

governmental functions. Although the contours of the judiciary and legislature are well-defined, the executive has become increasingly plural, fragmented

and dispersed.

According to the traditional tripartite separation of powers doctrine, every governmental function that is not assigned to the legislature or judiciary is to be discharged by the executive,[167] and that executive consists of all those parts of the governmental apparatus that are not the legislature or the judiciary. However, some external oversight mechanisms do not fit neatly into the traditional conception of the separation of powers. This conundrum is posed by the parliamentary creation of statutory authorities that are subject to no or limited ministerial control, such as independent regulators and statutory officers. Administrative tribunals that provide for appeal on the merits from executive decisions (including decisions of independent regulators) also sit uneasily within this structure as they have judicial characteristics but are part of the executive. While their approach to decision-making may be highly judicialised, they are seen to exercise administrative functions.

This article has shown that the Australian legislature and courts have had varying degrees of success in building ‘constitutional thought, values, concerns, and law into highly varied day-to-day administrative practices’[168] across multiple critical junctures from the 1970s to the present. By and large, despite a Constitution that does not envisage this great plurality of bodies within the executive, the legislature and the courts have adopted incremental adjustments to bring public administration into greater consonance with Australian constitutional democracy. The institutional developments witnessed in Australia in the long run have served to reinforce the traditional tripartite separation of powers, rather than to create a new fourth branch to accommodate the new arms-length scrutineers that emerged in the last few decades. The historical institutionalist theory of path dependency suggests an explanation for this: Australia started its legal tradition with an institutional trajectory shaped by its Constitution that neatly delineated between the legislature in ch I, the executive in ch II and the judiciary in ch III, and this institutional foundation led to a self-reinforcing trajectory of institutional development where the legislature and courts either pushed new post-constitutional bodies into the executive (for tribunals) or

co-opted them to assist the legislature (for oversight bodies).

However, recent developments have challenged the effectiveness of legal and political mechanisms of control in Australia. The ‘new public management’ era of business-like government, which has led to a series of privatisations and outsourcing of governmental functions, has only been partially adapted to the constitutional system of responsible government. The legislature has enhanced scrutiny over statutory authorities and government companies via the

PGPA Act, and increased accountability over government contractors through legislative amendments providing oversight bodies with jurisdiction to investigate government contractors and extending FOI requirements to these contractors. Nevertheless, the judiciary’s emphasis on legal form over function has led to the exclusion of corporatised entities from the ambit of judicial review, which has fundamentally undermined legal redress for governmental action.

The newest critical juncture for the Australian administrative state is the increased trend towards automation of government decision-making. Although it is too early to tell how automated decisions will be adapted into the constitutional scheme, early indications are that legal redress for individuals may again be negatively impacted by the automated decision-making due to the possible exclusion of judicial review, but that parliamentary committees and oversight bodies will be vigilant in investigating automated decision-making.

Another observation that can be made is that the trend over the last 30 years has shown a gradual reduction of legal control over the exercise of public power, as changes in the distribution of public power have resulted in the decline of judicial review. The high point of legal control over governmental action was exhibited in the ‘new administrative law’ phase in the 1970s, where the structures of governmental power were left unchanged, while legal controls on government power were significantly increased in terms of judicial review, oversight and disclosure obligations. In the second phase of privatisation and outsourcing, there were significant changes to the distribution of public power, which was dispersed across a range of institutions. This has led to the decline of legal controls on government power in the form of judicial review. Finally, the technological revolution has changed the way in which public power is exercised, which may further reduce the applicability of legal controls in terms of judicial review. The general trend towards the reduction in legal controls over the last few decades means that other forms of executive oversight have become paramount, such as Ombudsmen, auditors-general, and parliamentary


In conclusion, to achieve economic and cultural growth, as well as attain societal order and wellbeing, a nation’s institutions of governance must adapt to an ever-changing world. Edmund Burke encapsulates this challenge for all governments in his aphorism: ‘A state without the means of some change is without the means of its conservation.’[169] The Westminster constitutional context of the executive is steeped in fluid convention, which allows for the gradual evolution of the practices, organisation and control of the executive. Australian reformers have shown remarkable ingenuity in creating new institutions that comprise the administrative state, which the legislature and judiciary have, over the years, adapted to the constitutional scheme. An examination of various exogenous shocks to the constitutional system over the long term has shown that the Australian legislature has sought to maintain its system of responsible government by enacting legislation that subjects new administrative institutions to parliamentary scrutiny or investigation by oversight bodies, which Parliament has co-opted to assist with their scrutiny functions. On the other hand, the judiciary’s primary preoccupation has been the preservation of judicial independence and jurisdiction, with less attention paid to the evolving legal forms of the executive. As a result, precedential decisions have created a path dependency that has precluded judicial review in the era of privatisation and corporatisation, with a potential negative impact on automated government decisions as well. There is thus a distinct contrast between the ability of parliamentary institutions to accept, and to some extent to assimilate (to their profit) new ideas about administrative organisation and accountability — so as to preserve the essence of constitutional expectations regarding accountability — and the very uneven record of the judiciary in relation to the analogous preservation of such expectations regarding legality.

To sum up, this article highlights the dynamic, evolving nature of Australian institutions over the last four decades, and the roles of the legislature and judiciary in adapting these institutions to the existing constitutional structure. This process of adaptation is not always complete nor fully satisfactory, as shown by the reduction in legal controls over the years, but Australia’s system of checks and balances in terms of parliamentary and oversight mechanisms does provide a degree of counterweight to executive power. Ultimately, this article shows that Australian judicial and parliamentary institutions are able to, sometimes imperfectly, adapt to administrative change.

* Senior Lecturer, Faculty of Law, and Deputy Director, Australian Centre for Justice Innovation, Monash University. The author would like to thank Peter Cane, Terence Daintith, Stephen Gray and Ben Saunders for their helpful comments on a draft of this article.

[1] See Terence Daintith and Yee-Fui Ng, ‘Executives’ in Cheryl Saunders and Adrienne Stone (eds), The Oxford Handbook of the Australian Constitution (Oxford University Press, 2018) 587; Yee-Fui Ng, ‘In the Moonlight?: The Control and Accountability of Government Corporations in Australia’ [2019] MelbULawRw 16; (2019) 43(1) Melbourne University Law Review 303 (‘In the Moonlight?’).

[2] See generally Monika Zalnieriute, Lyria Bennett Moses and George Williams, ‘The Rule of Law and Automation of Government Decision-Making’ (2019) 82(3) Modern Law Review 425; Yee-Fui Ng et al, ‘Revitalising Public Law in a Technological Era: Rights, Transparency and Administrative Justice’ [2020] UNSWLawJl 37; (2020) 43(3) University of New South Wales Law Journal 1041; Carol Harlow and Richard Rawlings, ‘Proceduralism and Automation: Challenges to the Values of Administrative Law’ in Elizabeth Fisher, Jeff King and Alison L Young (eds), The Foundations and

Future of Public Law (Oxford University Press, 2019) 275.

[3] David H Rosenbloom, Rosemary O’Leary and Joshua Chanin, Public Administration and Law (Routledge, 3rd ed, 2010) 1, 44.

[4] See, eg, Janina Boughey and Lisa Burton Crawford (eds), Interpreting Executive Power (Federation Press, 2020); George Winterton, Parliament, the Executive and the Governor-General: A Constitutional Analysis (Melbourne University Press, 1983); Anne Twomey, ‘Pushing the Boundaries of Executive Power: Pape, The Prerogative and Nationhood Powers’ [2010] MelbULawRw 9; (2010) 34(1) Melbourne University Law Review 313; Peter Gerangelos, ‘The Executive Power of the Commonwealth of Australia: Section 61 of the Commonwealth Constitution, “Nationhood” and the Future of the Prerogative’ (2012) 12(1) Oxford University Commonwealth Law Journal 97.

[5] Orfeo Fioretos, Tulia G Falleti and Adam Sheingate, ‘Historical Institutionalism in Political Science’ in Orfeo Fioretos, Tulia G Falleti, and Adam Sheingate (eds), The Oxford Handbook of Historical Institutionalism (Oxford University Press, 2016) 3, 3.

[6] See Elizabeth Sanders, ‘Historical Institutionalism’ in Sarah A Binder, RAW Rhodes and Bert A Rockman (eds), The Oxford Handbook of Political Institutions (Oxford University Press, 2009) 39. For an application of historical institutionalism to Australian constitutional interpretation, see Theunis Roux, ‘Reinterpreting “the Mason Court Revolution”: An Historical Institutionalist Account of Judge-Driven Constitutional Transformation in Australia’ (2015) 43(1) Federal Law Review 1.

[7] Roux (n 6) 11.

[8] Ibid 11–12. See also Giovanni Capoccia, ‘Critical Junctures’ in Orfeo Fioretos, Tulia G Falleti, and Adam Sheingate (eds), The Oxford Handbook of Historical Institutionalism (Oxford University Press, 2016) 89.

[9] See generally Commonwealth Administrative Review Committee, Parliament of Australia, Report (Parliamentary Paper No 144, August 1971) (‘Kerr Committee Report’); Committee on Administrative Discretions, Parliament of Australia, Final Report (Parliamentary Paper No 316, October 1973) (‘Bland Committee Report’).

[10] Peter Cane, Controlling Administrative Power: An Historical Comparison (Cambridge University Press, 2016) 438–58 (‘Controlling Administrative Power’).

[11] Michael Crommelin, ‘The Executive’ in Gregory Craven (ed), The Convention Debates

1891–1898: Commentaries, Indices and Guide (Legal Books, 1986) 127, 136–45.

[12] Sidney Low, The Governance of England (T Fisher Unwin, 1906) 12.

[13] See Daintith and Ng, ‘Executives’ (n 1) 587–8.

[14] Sir Samuel Walker Griffith, ‘Notes on Australian Federation: Its Nature and Probable Effects’ (Paper, 1896) 17–18; New South Wales v Commonwealth [1975] HCA 58; (1975) 135 CLR 337, 364–5

(Barwick CJ); Hayden v Ditfort (1988) 93 FLR 131, 136 (Kirby P, Hope and Clark JJA). See generally R v Kirby; Ex parte Boilermakers’ Society of Australia [1956] HCA 10; (1956) 94 CLR 254, 275

(Dixon CJ, McTiernan, Fullagar and Kitto JJ); Australian Communist Party v Commonwealth [1951] HCA 5; (1951) 83 CLR 1, 221–2 (Williams J); Amalgamated Society of Engineers v Adelaide Steamship Co Ltd [1920] HCA 54; (1920) 28 CLR 129, 146–7 (Knox CJ, Isaacs, Rich and Starke JJ).

[15] Ryder v Foley [1906] HCA 61; (1906) 4 CLR 422, 432–3 (Griffith CJ). See PH Lane, Lane’s Commentary on the Australian Constitution (Lawbook, 2nd ed, 1997) 433–4.

[16] The Public Service Act 1922 (Cth) was held to be Parliament’s provision for the appointment of civil servants under s 67 of the Constitution: Barratt v Howard [1999] FCA 1132; (1999) 165 ALR 605, 609

[9]–[10] (Hely J).

[17] See generally RS Parker, ‘The Meaning of Responsible Government’ (1976) 11(2) Politics: Australian Journal of Political Science 178; Winterton (n 4) 71–85. For examples of the many judicial references to these ideas, see FAI Insurances v Winneke [1982] HCA 26; (1982) 151 CLR 342, 349

(Gibbs CJ), 364–5 (Mason J); Egan v Willis [1998] HCA 71; (1998) 195 CLR 424 (‘Egan’).

[18] Statutory oversight bodies include the Auditor-General, Ombudsman and Information Commissioners: see below Part III(B).

[19] See generally Senate Standing Committee on Finance and Government Operations, Parliament of Australia, Statutory Authorities of the Commonwealth (Report No 5, September 1982)


[20] Re Patterson; Ex parte Taylor (2001) 207 CLR 391, 403 [17] (Gleeson CJ) (‘Re Patterson’);

Williams v Commonwealth [2012] HCA 23; (2012) 248 CLR 156, 351–2 [516]–[517] (Crennan J); Egan (n 17) 451 [41] (Gaudron, Gummow and Hayne JJ).

[21] Benjamin B Saunders, ‘Responsible Government, Statutory Authorities and the Australian Constitution(2020) 48(1) Federal Law Review 4, 27.

[22] In Re Patterson (n 20), the High Court confirmed that it was constitutionally permissible for multiple Ministers to be appointed to administer a department, including Assistant Ministers or Parliamentary Secretaries: at 403 [17] (Gleeson CJ), 415 [66], 457–60 [205]–[211]

(Gummow and Hayne JJ), 498 [320] (Kirby J), 519 [380] (Callinan J).

[23] Administrative Appeals Tribunal Act 1975 (Cth); Administrative Decisions (Judicial Review) Act 1977 (Cth) (‘ADJR Act’); Ombudsman Act 1976 (Cth); Freedom of Information Act 1982 (Cth); Privacy Act 1988 (Cth).

[24] Dennis Pearce, ‘The Fading of the Vision Splendid?: Administrative Law’ (1989) 58 (April) Canberra Bulletin of Public Administration 15, 15. See also Peter Cane, Administrative Tribunals and Adjudication (Hart Publishing, 2009) 182–3, 257. See generally Kerr Committee Report

(n 9); Bland Committee Report (n 9).

[25] Pearce (n 24) 15.

[26] Roux (n 6) 11. See also Sanders (n 6) 39.

[27] Roux (n 6) 11–12.

[28] James Mahoney, ‘Path Dependence in Historical Sociology’ (2000) 29(4) Theory and Society 507, 508, 513, 523–4. See generally Paul Pierson, ‘Increasing Returns, Path Dependence, and the Study of Politics’ (2000) 94(2) American Political Science Review 251.

[29] Cane, Administrative Tribunals and Adjudication (n 24) 71–2.

[30] Ibid 72–82.

[31] Peter Cane, ‘Merits Review and Judicial Review: The AAT as Trojan Horse’ [2000] FedLawRw 11; (2000) 28(2) Federal Law Review 213, 214.

[32] Kerr Committee Report (n 9) 68–75 [228]–[249], 115 [16]; Bland Committee Report (n 9) 30–1 [171], 35–6 [180], 47 [xxxiii].

[33] Drake v Minister for Immigration and Ethnic Affairs (1979) 2 ALD 60, 68 (Bowen CJ and

Deane J).

[34] See Brandy v Human Rights and Equal Opportunity Commission [1995] HCA 10; (1995) 183 CLR 245, 259–60 (Mason CJ, Brennan and Toohey JJ) (‘Brandy’).

[35] Kerr Committee Report (n 9) 68–75 [228]–[249], 115 [16]; Bland Committee Report (n 9) 30–1 [171], 35–6 [180], 47 [xxxiii].

[36] Brandy (n 34) 260.

[37] Meringnage v Interstate Enterprises Pty Ltd [2020] VSCA 30; (2020) 60 VR 361, 387–8 [80], 408–9 [148] (Tate, Niall and Emerton JJA), holding that the Victorian Civil and Administrative Tribunal was not a ‘court of a state’ within ch III of the Constitution and was unable to exercise state judicial power against the Commonwealth. In Burns v Corbett (2018) 265 CLR 304, 334 [39], 337 [46], 339 [50] (Kiefel CJ, Bell and Keane JJ), the High Court held that the New South Wales Civil and Administrative Tribunal was not ‘a court of a State’ and was prevented from adjudicating disputes involving any of the matters set out in ss 75 and 76 of the Constitution, even when the dispute involves the application of state legislation. However, in Owen v Menzies [2012] QCA 170; [2013] 2 Qd R 327, 334 [10], 338 [20] (de Jersey CJ, Muir JA agreeing at 357 [103]), 344–5 [48], 348 [61] (McMurdo P), the Queensland Civil and Administrative Tribunal was held to be a ‘court of a State’; a strong factor was the designation in its constitutive statute as a ‘court of record’. For a detailed discussion of the case law regarding state tribunals, see Rebecca Ananian-Welsh, ‘CATs, Courts and the Constitution: The Place of Super-Tribunals in the National Judicial System’ [2020] MelbULawRw 11; (2020) 43(3) Melbourne University Law Review 852.

[38] Shell Co of Australia Ltd v Federal Commissioner of Taxation (1930) 44 CLR 530, 543 (Viscount Dunedin LC).

[39] Yee-Fui Ng, ‘Tribunal Independence in an Age of Migration Control’ (2012) 19(4) Australian Journal of Administrative Law 203, 206.

[40] Ibid 205–7, 211.

[41] Senate Standing Committee on Finance and Public Administration, Parliament of Australia, Review of the Office of the Commonwealth Ombudsman (Report, December 1991) 14 [2.43].

[42] Kerr Committee Report (n 9) 93–4 [313].

[43] Anita Stuhmcke, ‘“Each for Themselves” or “One for All”?: The Changing Emphasis of the Commonwealth Ombudsman’ [2010] FedLawRw 5; (2010) 38(1) Federal Law Review 143, 144. See also Anita Stuhmcke, ‘Australian Ombudsmen: Drafting a Blueprint for Reform’ (2017) 24(1) Australian Journal of Administrative Law 43 (‘Australian Ombudsmen’).

[44] Statutory bodies established in early Federation include the Auditor-General (Audit Act 1901 (Cth) ss 3–15), Public Service Commissioner (Commonwealth Public Service Act 1902 (Cth)

ss 5–11), and Commissioner of Land Tax (Land Tax Assessment Act 1910 (Cth) ss 4–9). See also RL Wettenhall, ‘Commonwealth Statutory Authorities: Patterns of Growth’ (1977) 36(4) Australian Journal of Public Administration 351, 361; RL Wettenhall, ‘Commonwealth Statutory Authorities: Exploring the Field’ (1979) 38(2) Australian Journal of Public Administration 176. For statutory authorities in colonial times, see Paul Finn, Law and Government in Colonial Australia (Oxford University Press, 1987) 58–61, 95–102, 128–32.

[45] See AJ Brown, Chaos or Coherence?: Strengths, Opportunities and Challenges for Australia’s Integrity Systems (National Integrity Systems Assessment Final Report, December 2005) 31–2.

[46] Zoe Mitchell, Shannon Merrington and Peter Bell, ‘A Comparative Analysis of the OECD Anti-Corruption Models (Asia & Europe) and Australia’s Existing Anti-Corruption Platform’ (2014) 4(3) International Journal of Business and Commerce 1, 5–7.

[47] ‘About the PBO’, Parliament of Australia (Web Page) <>, archived at <>.

[48] See generally John Keane, The Life and Death of Democracy (Simon & Schuster, 2009).

[49] See John M Carey, ‘Competing Principals, Political Institutions, and Party Unity in Legislative Voting’ (2007) 51(1) American Journal of Political Science 92, 99–100.

[50] Gabrielle Appleby, ‘Horizontal Accountability: The Rights-Protective Promise and Fragility of Executive Integrity Institutions’ (2017) 23(2) Australian Journal of Human Rights 168, 174; Scott Prasser, ‘Australian Integrity Agencies in Critical Perspective’ (2012) 33(1) Policy Studies 21, 29.

[51] JJ Spigelman, ‘The Integrity Branch of Government’ (2004) 78(11) Australian Law Journal 724. See also Bruce Ackerman, ‘The New Separation of Powers’ (2000) 113(3) Harvard Law Review 633, 693–6.

[52] See John McMillan, ‘Re-Thinking the Separation of Powers’ [2010] FedLawRw 18; (2010) 38(3) Federal Law Review 423, 433.

[53] See, eg, Auditor‑General Act 1997 (Cth) s 8(1), designating the Commonwealth Auditor-General as an ‘officer of Parliament’; Ombudsman Act 2001 (Qld) s 11(2), designating the Queensland Ombudsman as an ‘officer of Parliament’; Auditor General Act 2006 (WA) s 7(1), designating the Western Australian Auditor-General as an ‘officer of Parliament’. See also Constitution Act 1975 (Vic) ss 94B, 94E, 94F, which constitutionally entrenches the Victorian Ombudsman, Auditor-General and Electoral Commissioner as independent officers of Parliament. However, the legal effect of this entrenchment is open to question: see Gerard Carney, ‘State Constitutions’ in Cheryl Saunders and Adrienne Stone (eds), The Oxford Handbook of the Australian Constitution (Oxford University Press, 2018) 277, 301–3. The New South Wales (‘NSW’) Ombudsman was described as an ‘independent officer of the New South Wales Parliament’ in K v NSW Ombudsman [2000] NSWSC 771, [25] (Whealy J).

[54] See, eg, Ombudsman Act 1974 (NSW) ss 6, 6A, 31A31BA, providing that the NSW Ombudsman is appointed by the Governor, on the recommendation of the Minister, with the Committee on the Ombudsman, the Law Enforcement Conduct Commission and the Crime Commission able to veto the appointment; Ombudsman Act 1972 (SA) s 6(1), providing that the South Australian Ombudsman is appointed by the Governor ‘on a recommendation made by resolution of both Houses of Parliament’.

[55] All oversight bodies have legislative protections from removal. For example, the Commonwealth Ombudsman can only be removed on address from both Houses of Parliament on specified grounds, such as misbehaviour or physical or mental incapacity: Ombudsman Act 1976 (Cth) s 28(1).

[56] For example, the Joint Committee of Public Accounts and Audit supervises the activities of the Commonwealth Auditor-General: Public Accounts and Audit Committee Act 1951 (Cth) s 5; NSW has the Parliamentary Joint Committee on the Office of the Ombudsman, the Law Enforcement Conduct Commission and the Crime Commission: Ombudsman Act 1974 (NSW) s 31B; the Victorian Inspectorate oversees the Independent Broad-Based Anti-Corruption Commission and staff, Victorian Ombudsman and staff, Victorian Auditor-General and staff, and Victorian Information Commissioner and staff, in addition to the specialised Independent Broad-Based Anti-Corruption Commission Committee: Victorian Inspectorate Act 2011 (Vic) s 11.

[57] In fact, the Auditor‑General Act 1997 (Cth) s 8(2) disclaims the effect of being an ‘officer of Parliament’, stating: ‘There are no implied functions, powers, rights, immunities or obligations arising from the Auditor-General being an independent officer of the Parliament’.

[58] An exception is a Victorian inquiry into such implications: Public Accounts and Estimates Committee, Parliament of Victoria, A Legislative Framework for Independent Officers of Parliament (Report No 67, February 2006).

[59] See, eg, Jeffrey Graham Bell, ‘Agents of Parliament: A New Branch of Government?’ (2006) 29(1) Canadian Parliamentary Review 13.

[60] See, eg, Finance and Expenditure Committee, Parliament of New Zealand, Inquiry into Officers of Parliament (Report, 1989).

[61] See, eg, Oonagh Gay, ‘Officers of Parliament: A Comparative Perspective’ (Research Paper No 03/77, Parliament and Constitution Centre, House of Commons Library, 20 October 2003).

[62] For example, the Commonwealth Ombudsman is still seen to be part of the executive, while the Commonwealth Auditor-General is an ‘officer of Parliament’: Auditor‑General Act 1997 (Cth) s 8(1); Stuhmcke, ‘Australian Ombudsmen’ (n 43) 49.

[63] See, eg, Ombudsman Act 1974 (NSW), which establishes a joint committee of both houses of the legislature, called the Committee on the Office of the Ombudsman, the Law Enforcement Conduct Commission and the Crime Commission. The functions of the Committee include the exercise of a power of veto over the proposed appointment of a person as Ombudsman, the examination of the reports of the Ombudsman, as well as monitoring and keeping the legislature informed on the operations of the Ombudsman: at ss 31A31BA. A specific parliamentary committee of this nature does not exist at the federal level.

[64] For example, the NSW Ombudsman faced two parliamentary inquiries into the handling of a large investigation, ‘Operation Prospect’, which Stuhmcke contended were set up for partisan reasons: Anita Stuhmcke, ‘Australian Ombudsmen: A Call to Take Care’ (2016) 44(3) Federal Law Review 531, 547–50. Subsequently, the NSW Ombudsman chose not to seek reappointment due to the attacks on his office, stating that ‘I made the decision given the nature of the appalling comments that have been made about me, about my office and about the length of my term that it was in the best interests of my office for me not to seek a further term’: at 551, quoting Evidence to General Purpose Standing Committee No 4, Parliament of New South Wales, Sydney, 19 June 2015, 29 (Bruce Barbour).

[65] Sir Anthony Mason, ‘Administrative Review: The Experience of the First Twelve Years’ [1989] FedLawRw 3; (1989) 18(3) Federal Law Review 122, 123; Will Bateman and Leighton McDonald, ‘The Normative Structure of Australian Administrative Law’ (2017) 45(2) Federal Law Review 153, 161.

[66] Debbie Mortimer, ‘The Constitutionalization of Administrative Law’ in Cheryl Saunders and Adrienne Stone (eds), The Oxford Handbook of the Australian Constitution (Oxford University Press, 2018) 696, 702–3. See also Stephen Gageler, ‘Impact of Migration Law on the Development of Australian Administrative Law’ (2010) 17(2) Australian Journal of Administrative Law 92, 95.

[67] Mortimer (n 66) 702; Gageler (n 66) 95.

[68] ADJR Act (n 23) sch 1 para (da). See Gageler (n 66) 92; See Justice Ronald Sackville, ‘Judicial Review of Migration Decisions: An Institution in Peril?’ [2000] UNSWLawJl 59; (2000) 23(3) University of New South Wales Law Journal 190, 191–4.

[69] See, eg, the Migration Reform Act 1992 (Cth) ss 13, 33, which introduced mandatory detention for non-citizens and curtailed Federal Court judicial review; Migration Litigation Reform Act 2005 (Cth) s 15, which introduced privative clauses into s 474 of the Migration Act 1958 (Cth) seeking to oust the jurisdiction of the courts. See generally Gageler (n 66).

[70] Mortimer (n 66) 696; Daniel Reynolds, ‘The Constitutionalisation of Administrative Law: Navigating the Cul-de-Sac’ (2013) 74 (October) Australian Institute of Administrative Law Forum 73.

[71] [2003] HCA 2; (2003) 211 CLR 476. The High Court held that there was an ‘entrenched minimum provision’ of judicial review at the federal level, which was review for jurisdictional error: at 513 [103] (Gaudron, McHugh, Gummow, Kirby and Hayne JJ). See, eg, Leighton McDonald, ‘The Entrenched Minimum Provision of Judicial Review and the Rule of Law’ (2010) 21(1) Public Law Review 14; Will Bateman, ‘The Constitution and the Substantive Grounds of Judicial Review: The Full Scope of the Entrenched Minimum Provision of Judicial Review’ [2011] FedLawRw 17; (2011) 39(3) Federal Law Review 463. A further development was the High Court utilising the term ‘constitutional writs’ to refer to the remedies under s 75(v) of the Constitution, rather than ‘prerogative writs’: Re Refugee Tribunal; Ex parte Aala [2000] HCA 57; (2000) 204 CLR 82, 92–3 [19]–[21] (Gaudron and

Gummow JJ).

[72] (2010) 239 CLR 531, 581 [99]–[100] (French CJ, Gummow, Hayne, Crennan, Kiefel and

Bell JJ).

[73] Ibid 581 [99].

[74] Ibid, quoting Louis L Jaffe, ‘Judicial Review: Constitutional and Jurisdictional Fact’ (1957) 70(6) Harvard Law Review 953, 963.

[75] Cane, Controlling Administrative Power (n 10) 495–6.

[76] Minister for Immigration and Citizenship v SZMDS (2010) 240 CLR 611, 625 [40]

(Gummow ACJ and Kiefel J), 645–8 [124]–[132] (Crennan and Bell JJ); Minister for Immigration and Citizenship v Li [2012] HCA 61; (2013) 249 CLR 332, 366 [72] (Hayne, Kiefel and Bell JJ).

[77] Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1947] EWCA Civ 1; [1948] 1 KB 223.

[78] Cane, Controlling Administrative Power (n 10) 455–7.

[79] Christopher Hood, ‘A Public Management for All Seasons?’ (1991) 69(1) Public Administration 3, 4–5.

[80] See Finn (n 44) 58–61, 95–102, 128–32.

[81] Ng, ‘In the Moonlight?’ (n 1) 315–16.

[82] Cane, Controlling Administrative Power (n 10) 458.

[83] Ng, ‘In the Moonlight?’ (n 1) 317.

[84] As of January 2021, there are 99 Commonwealth agencies and statutory authorities, 71 statutory corporations and 18 Commonwealth companies: Australian Government Department of Finance, ‘Flipchart of PGPA Act Commonwealth Entities and Companies’ (Flipchart, 4

January 2021).

[85] Terence Daintith and Alan Page, The Executive in the Constitution: Structure, Autonomy, and Internal Control (Oxford University Press, 1999) 29.

[86] Cane, Controlling Administrative Power (n 10) 456–7.

[87] See Terence Daintith, ‘Regulation by Contract: The New Prerogative’ (1979) 32(1) Current

Legal Problems 41, 55.

[88] Public Governance, Performance and Accountability Act 2013 (Cth) ch 3 (‘PGPA Act’).

[89] Ng, ‘In the Moonlight?’ (n 1) 320–2.

[90] See PGPA Act (n 88) s 10.

[91] Ibid ss 89(1), 90. Ex ante accountability ‘is required before significant policy or resource decisions are made’, while ex post accountability operates after the event: Stephen Bottomley, ‘Government Business Enterprises and Public Accountability through Parliament’ (Research Paper No 18, Department of the Parliamentary Library, April 2000). See also Ian Thynne and John Goldring, Accountability and Control: Government Officials and the Exercise of Power (Law Book, 1987) 226.

[92] See PGPA Act (n 88) ss 38–9; Public Governance, Performance and Accountability Rule 2014 (Cth) s 16F.

[93] PGPA Act (n 88) s 46.

[94] Ibid s 97.

[95] Graham Taylor, ‘State-Owned Enterprises’ Accountability to Parliament’ (1987) 10(16) The Capital Letter 1, 1.

[96] See Terence Daintith and Yee-Fui Ng, ‘Legal Form and Function in the Public Sector: The Government-Owned Company in the United Kingdom and Australia’ (2020) 136 (April) Law Quarterly Review 292, 307 (‘Legal Form and Function’).

[97] [1986] EWCA Civ 8; [1987] 1 QB 815, 847 (Lloyd LJ) (‘Datafin’).

[98] Daintith and Ng, ‘Legal Form and Function’ (n 96) 307.

[99] See Justice Emilios Kyrou, ‘Judicial Review of Decisions of Non-Governmental Bodies Exercising Governmental Powers: Is Datafin Part of Australian Law?’ (2012) 86(1) Australian Law Journal 20, 22–5, 26–9. Cf Masu Financial Management Pty Ltd v Financial Industry Complaints Service Ltd [No 2] (2004) 50 ASCR 554, 559–60 [3]–[5] (Shaw J); Norths

Ltd v McCaughan Dyson Capel Cure Ltd (1988) 12 ACLR 739, 745–6 (Young J); Typing Centre of New South Wales v Toose (Supreme Court of New South Wales, Matthews J, 12 December 1988) 19–20, discussed in McClelland v Palms Surf Life Saving Club [2002] NSWSC 470; (2002) 191 ALR 759,

790–1 [115]–[117] (Campbell J); Chase Oyster Bar Pty Ltd v Hamo Industries Pty Ltd [2010] NSWCA 190; (2010) 78 NSWLR 393, 413 [81] (Basten JA); Khuu & Lee Pty Ltd v Corporation of the City of Adelaide (2011) 110 SASR 235, 241–3 [25]–[31] (Vanstone J); Mickovski v Financial Ombudsman Service Ltd [2012] VSCA 185; (2012) 36 VR 456, 466 [30]–[32] (Buchanan and Nettle JJA and Beach AJA).

[100] ADJR Act (n 23) s 3.

[101] [2005] HCA 7; (2005) 221 CLR 99, 130–1 [89] (Gummow, Callinan and Heydon JJ).

[102] General Newspapers Pty Ltd v Telstra Corporation [1993] FCA 473; (1993) 45 FCR 164, 173 (Davies and

Einfeld JJ); Australian National University v Burns [1982] FCA 191; (1982) 64 FLR 166, 174 (Bowen CJ and Lockhart J), 180–1 (Sheppard J); CEA Technologies Pty Ltd v Civil Aviation Authority [1994] FCA 1180; (1994) 51 FCR 329, 337 (Neaves J).

[103] [2003] HCA 35; (2003) 216 CLR 277, 289 [23] (Gleeson CJ), 296–8 [47]–[56] (McHugh, Hayne and

Callinan JJ).

[104] Ibid 300 [64] (McHugh, Hayne and Callinan JJ).

[105] Bodies held not to be ‘officers of the Commonwealth’ include the National Companies and Securities Commission, Australian Telecommunications Commission, Australian Postal Commission, Migration Agents Registration Authority, and Australian Medical Council: see, eg,

R v Murray; Ex parte Commonwealth [1916] HCA 58; (1916) 22 CLR 437, 452 (Isaacs J) (‘Murray’); Broken Hill Pty Co Ltd v National Companies and Securities Commission (1986) 61 ALJR 124, 127

(Dawson J) (High Court) (‘Broken Hill’); Businessworld Computers Pty Ltd v Australian Telecommunications Commission [1988] FCA 206; (1988) 82 ALR 499, 500 (Gummow J) (Federal Court) (‘Businessworld’); Post Office Agents Association Ltd v Australian Postal Commission (1988) 84 ALR 563, 575 (Davies J) (‘Post Office Agents’); McGowan v Migration Agents Registration Authority [2003] FCA 482; (2003) 129 FCR 118, 126 [26] (Branson J) (‘McGowan’); Australasian College of Cosmetic Surgery

Ltd v Australian Medical Council Ltd [2015] FCA 468; (2015) 232 FCR 225, 233–4 [40]–[43] (Katzmann J) (‘Australasian College’). See also Mark Aronson, Matthew Groves and Greg Weeks, Judicial Review of Administrative Action and Government Liability (Lawbook, 6th ed, 2017) 49–51 [2.160]. The High Court has not yet resolved whether government contractors are ‘officers of the Commonwealth’: Plaintiff M61 v Commonwealth [2010] HCA 41; (2010) 243 CLR 319, 345 [51] (French CJ, Gummow, Hayne, Heydon, Crennan, Kiefel and Bell JJ). See also Janina Boughey and Greg Weeks, ‘“Officers of the Commonwealth” in the Private Sector: Can the High Court Review Outsourced Exercises of Power?’ [2013] UNSWLawJl 14; (2013) 36(1) University of New South Wales Law Journal 316.

[106] Section 75(iii) of the Constitution confers original jurisdiction on the High Court for actions in which the Commonwealth is a party, including some statutory authorities: see

Maguire v Simpson [1977] HCA 63; (1977) 139 CLR 362, 389 (Gibbs J), 391 (Stephen J), 398–9 (Mason J), 407 (Murphy J).

[107] The possibility of s 75(iii) applying to government companies was left open in SGH Ltd v Federal Commissioner of Taxation [2002] HCA 18; (2002) 210 CLR 51, although in that case the High Court held that SGH Ltd was not the State of Queensland, because the rights of control which the State had over SGH Ltd, by reason of its majority shareholding in SGH Ltd, were restricted by the existence of members of the public who were minority shareholders and whose rights were safeguarded by the general law: at 71 [26] (Gleeson CJ, Gaudron, McHugh and Hayne JJ). But see Commonwealth v Bogle [1953] HCA 10; (1953) 89 CLR 229, 256 (Webb J), 267 (Fullagar J, Dixon CJ agreeing at 249, Kitto J agreeing at 274), in which a majority of the Court held that the Commonwealth Hostels Ltd was not to be treated as the Crown for the purposes of immunity from state legislation although it was a wholly-owned government company. Darrell Barnett has argued that wholly-owned government companies should come within s 75(iii): Darrell Barnett,

‘The Commonwealth: A Multitude of Manifestations’ (2007) 81(3) Australian Law Journal 195, 210–11.

[108] Bodies held to be within the jurisdiction of s 75(iii) include the Commercial Trading Bank of Australia, Commonwealth Bank of Australia, Repatriation Commission, and Australian Atomic Energy Commission: see, eg, Inglis v Commonwealth Trading Bank of Australia [1969] HCA 44; (1969) 119 CLR 334, 341–2 (Kitto J); Bank of New South Wales v Commonwealth [1948] HCA 7; (1948) 76 CLR 1, 274–6 (Rich and Williams JJ), 322 (Starke J), 367 (Dixon J); Repatriation Commission v Kirkland [1923] HCA 18; (1923) 32 CLR 1, 20 (Rich J); Kathleen Investments (Australia) Ltd v Australian Atomic Energy Commission [1977] HCA 55; (1977) 139 CLR 117, 133 (Barwick CJ). See also Barnett (n 107) 212–14.

[109] See, eg, Mark Aronson, ‘A Public Lawyer’s Responses to Privatisation and Outsourcing’ in Michael Taggart (ed), The Province of Administrative Law (Hart Publishing, 1997) 40; Margaret Allars, ‘Private Law but Public Power: Removing Administrative Law Review from Government Business Enterprises’ (1995) 6(1) Public Law Review 44; Nick Seddon, ‘The Interaction of Contract and Executive Power’ [2003] FedLawRw 21; (2003) 31(3) Federal Law Review 541.

[110] See Murray (n 105) 452 (Isaacs J). See above n 105 and accompanying text.

[111] See, eg, Telstra Corporation Ltd v Kendall [1995] FCA 1481; (1995) 55 FCR 221, 231 (Black CJ, Ryan and Hill JJ).

[112] See, eg, Aboriginal and Torres Strait Islander Act 2005 (Cth) s 200(6), enabling review under the ADJR Act (n 23) of decisions made by Aboriginal Hostels Ltd; Competition and Consumer Act 2010 (Cth) s 152CJE, enabling judicial enforcement by affected individuals of an NBN corporation’s obligations to supply National Broadband Network services; National Electricity (South Australia) Act 1996 (SA) s 70, subjecting the Australian Energy Market Operator to judicial review.

[113] See Corporations Act 2001 (Cth) s 921V(7), enabling AAT review of decisions by the Financial Adviser Standards and Ethics Authority to refuse to approve a foreign qualification.

[114] Ombudsman Act 1976 (Cth) ss 3(4B), 3BA, 5(1).

[115] Auditor-General Act 1997 (Cth) ss 18B, 323.

[116] Senate, Parliament of Australia, Procedural Orders and Resolutions of the Senate of Continuing Effect (at January 2020) o 13.

[117] See, eg, Auditor-General, Senate Order for Departmental and Entity Contracts (Financial Year 2016–17 Compliance) (ANAO Report No 49, 25 June 2018).

[118] Freedom of Information Act 1982 (Cth) s 6C.

[119] Ibid ss 47, 47G.

[120] Administrative Review Council, Automated Assistance in Administrative Decision Making

(Issues Paper, 2003) 11–15.

[121] Commonwealth Ombudsman, Centrelink’s Automated Debt Raising and Recovery System: A Report about the Department of Human Services’ Online Compliance Intervention System for Debt Raising and Recovery (Report No 2/2017, April 2017) 5 [1.1] (‘OCI Report’).

[122] ‘SmartGates’, Australian Border Force (Web Page) <>.

[123] Zalnieriute, Moses and Williams (n 2) 425.

[124] Justice Melissa Perry, ‘iDecide: Administrative Decision-Making in the Digital World’ (2017) 91(1) Australian Law Journal 29, 30.

[125] See, eg, Paul Gowder, ‘Transformative Legal Technology and the Rule of Law’ (2018) 68(1) University of Toronto Law Journal 82.

[126] See generally Harlow and Rawlings (n 2) 275.

[127] See, eg, Jenna Burrell, ‘How the Machine “Thinks”: Understanding Opacity in Machine Learning Algorithms’ (2016) 3(1) Big Data & Society 1.

[128] Robert van den Hoven van Genderen, ‘Privacy and Data Protection in the Age of Pervasive Technologies in AI and Robotics’ (2017) 3(3) European Data Protection Law 3; Algorithms and Human Rights: Study on the Human Rights Dimensions of Automated Data Processing Techniques and Possible Regulatory Implications (Council of Europe Study DGI(2017)12, March 2018) 12–16.

[129] See generally Cathy O’Neil, Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy (Crown, 2016).

[130] See Senate Community Affairs References Committee, Parliament of Australia, Design, Scope, Cost-Benefit Analysis, Contracts Awarded and Implementation Associated with the Better Management of the Social Welfare System Initiative (Report, June 2017) 1 [1.2] (‘Social Welfare System Initiative Report’).

[131] See Terry Carney, ‘Vulnerability: False Hope for Vulnerable Social Security Clients?’ [2018] UNSWLawJl 27; (2018) 41(3) University of New South Wales Law Journal 783, 810.

[132] OCI Report (n 121) 5 [1.1], 6 [2.5].

[133] Social Welfare System Initiative Report (n 130) 5–6 [1.23]–[1.25].

[134] See Datafin (n 97) 847 (Lloyd LJ).

[135] [2018] FCAFC 79; (2018) 262 FCR 41, 67 [140] (‘Pintarich’). See also Yee-Fui Ng and Maria O’Sullivan, ‘Deliberation and Automation: When Is a Decision a “Decision”?’ (2019) 26(1) Australian Journal of Administrative Law 21.

[136] Pintarich (n 135) 48–9 [46]–[47].

[137] Ibid 49 [49].

[138] Pintarich v Deputy Commissioner of Taxation [2018] HCASL 322.

[139] See, eg, Hazardous Waste (Regulation of Exports and Imports) Act 1989 (Cth) s 57, which provides that ‘[a]pplications may be made to the Administrative Appeals Tribunal for the review of decisions of the following kinds ...’ (emphasis added).

[140] According to Murray (n 105) 452 (Isaacs J), an ‘officer of the Commonwealth’ has to have an office of some conceivable tenure, be directly appointed by the Commonwealth, accept office and salary from the Commonwealth, and be removable by the Commonwealth. See also Broken Hill (n 105) 127 (Dawson J); Businessworld (n 105) 500 (Gummow J); Post Office Agents

(n 105) 575 (Davies J); McGowan (n 105) 126 [26] (Branson J); Australasian College (n 105) 233–4 [40]–[43] (Katzmann J); Aronson, Groves and Weeks (n 105) 49–51.

[141] Ng and O’Sullivan (n 135) 32.

[142] As Perry has pointed out: ‘If automated systems were used in cases [requiring discretion or evaluative judgment], not only may there be a constructive failure to exercise the discretion; by their nature they apply predetermined outcomes which may be characterised as pre-judgment or bias’: Perry (n 124) 33.

[143] Australian Government Information Management Office, Department of Finance and Administration (Cth), Automated Assistance in Administrative Decision-Making: Better Practice Guide (Report, February 2007) 14.

[144] See Craig v South Australia [1995] HCA 58; (1995) 184 CLR 163, 179–80 (Brennan, Deane, Toohey, Gaudron and McHugh JJ); Minister for Immigration and Multicultural Affairs v Yusuf (2001) 206 CLR 323, 351 [82] (McHugh, Gummow, and Hayne JJ). See also Dominique Hogan-Doran, ‘Computer Says “No”: Automation, Algorithms and Artificial Intelligence in Government Decision-Making’ (2017) 13(3) Judicial Review 345, 355.

[145] Hogan-Doran (n 144) 355.

[146] Order of Davies J in Amato v Commonwealth (Federal Court of Australia, VID611/2019, 27 November 2019) 6 [9].

[147] See Ng, ‘In the Moonlight?’ (n 1) 327.

[148] See Daintith and Ng, ‘Legal Form and Function’ (n 96) 307.

[149] See, eg, Migration Act 1958 (Cth) s 495A; Business Names Registration Act 2011 (Cth) s 66; Social Security (Administration) Act 1999 (Cth) s 6A.

[150] See, eg, Parliamentary Joint Committee on Human Rights, Parliament of Australia, Human Rights Scrutiny (Report No 11, 16 October 2018) 78–81.

[151] Senate Foreign Affairs, Defence and Trade Legislation Committee, Parliament of Australia, Veterans’ Affairs Legislation Amendment (Digital Readiness and Other Measures) Bill 2016 [Provisions] (Report, February 2017) 5–11 [2.2]–[2.32].

[152] Social Welfare System Initiative Report (n 130).

[153] Senate Community Affairs References Committee, Parliament of Australia, ParentsNext, including Its Trial and Subsequent Broader Rollout (Report, March 2019) 55 [3.19]–[3.22]

(‘ParentsNext Report’).

[154] Ibid 13 [1.52].

[155] Ibid 29 [2.54]. See also ‘Morrison Government Must Scrap ParentsNext and Stop Making Women’s Lives Harder’, Human Rights Law Centre, (Web Page, 29 March 2019) <>, archived at <>.

[156] ParentsNext Report (n 153) 71 [4.1], 74 [4.25].

[157] Social Welfare System Initiative Report (n 130) 33–4 [2.89]–[2.93], 36–8 [2.104]–[2.114], 79–84 [4.46]–[4.77], 107 [6.2]–[6.3].

[158] Ibid 108–9 [6.9]–[6.11].

[159] OCI Report (n 121).

[160] Commonwealth Ombudsman, Centrelink’s Automated Debt Raising and Recovery System: Implementation Report (Report No 1/2019, April 2019).

[161] Ibid 1.

[162] Ibid 2, 27 [4.1]–[4.5].

[163] See Philip Alston, Report of the Special Rapporteur on Extreme Poverty and Human Rights, UN Doc A/74/493 (11 October 2019).

[164] Ibid [3].

[165] Digital Transformation Agency, Vision 2025 (Strategy Paper, 2018).

[166] New South Wales Government, Digital NSW: Designing Our Digital Future (Strategy Paper, 2019) 3.

[167] As the High Court acknowledged in Williams v Commonwealth [No 2] [2014] HCA 23; (2014) 252 CLR 416, 467–8 [78] (French CJ, Hayne, Kiefel, Bell and Keane JJ), partially accepting a proposition put in argument by the Commonwealth, ‘executive power is all that power of a polity that is not legislative or judicial power’.

[168] Rosenbloom, O’Leary and Chanin (n 3) 44.

[169] Edmund Burke, Reflections on the French Revolution, ed Charles W Eliot (PF Collier & Son, 1909) 161.

AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback