Orphaned responsibility

Contracting out the duties of government

UTOPIAN IDEALS ARE ordinarily more ambitious, and romantic, than the desire to see a constitutional system functioning as it should. The latter sounds more like a defence of the status quo than a reimagining of political possibility.

Anyone who lived through 2020 in Australia, however, might beg to differ. The year during which the world seemed to fall off its axis saw numerous examples of apparent confusion on the part of Australian political actors about what our constitutional systems actually demand of them, and of head-scratching bewilderment among the rest of us about how this could be so.

A telling instance of this phenomenon was when the federal member for Wentworth, Dave Sharma, suggested on the ABC’s Q&A that some of the more scandalous recent decisions made by government agencies had no connection to ministers of government. He was immediately schooled by a visibly surprised fellow panellist, The Ethics Centre’s Simon Longstaff, on the fundamentals of the constitutional doctrine of ministerial responsibility. This in turn prompted former Commonwealth Treasury head, Ken Henry, to comment that in his observation the doctrine of ministerial responsibility had effectively disappeared by the late 1990s. The Shadow Minister for Education, Tanya Plibersek, took the discussion from there, offering various recent examples of government or ministerial failure that, accountability-wise, had met with no meaningful response.

There have been far more dramatic examples to suggest that political accountability in Australia might have entered a dark age than this three-minute discussion. But those three minutes resonated because, at around the same time, I had been closely following the ultimately futile quest of the Victorian COVID-19 Hotel Quarantine Inquiry to identify who was responsible for the decision to outsource frontline hotel quarantine enforcement to private security providers: an action that proved causally critical to Victoria’s devastating ‘second wave’ of COVID-19 infections, precipitated nearly four months’ lockdown and cost Victorian taxpayers tens of millions of dollars. 

This time last year I could not have imagined that I would end up spending so much of my intellectual energy trying to decipher what went wrong within, and what we need to learn from, something as esoteric as outsourced security services for hotel quarantine. I’d been thinking about what was amiss with contracted-out government service delivery for some time, but had not previously done the kind of thinking that this example prompted me to do. Like an endless onion-peeling operation, every time I’d said what I thought needed to be said about one or other feature of it, something else would reveal itself that also needed addressing. And then there was the inquiry itself, an often gripping drama that existed somewhere on a scale between Shakespeare and Beckett.

It cannot be that senior political and governmental officials these days are not as good as they used to be, or universally care less about what happens on their watch. There might be some truth to both propositions in some cases. But given that the problems we are witnessing are neither partisan nor confined to any one level of government, one can only conclude that something larger and more pervasive must be going on.

Every governmental epoch has particular philosophical and institutional characteristics. Ours is shaped by neoliberalism, which calls for institutions of government to be redesigned in accordance with market-based principles, and which is committed to the use of tools such as contracting out (often also described as outsourcing) as core techniques of government. Much has been written on the history and characteristics of neoliberalism as an ideology. But much less attention has been paid to the way that specific techniques of neoliberal government, like outsourcing, fit within and are answerable to the demands of Australia’s constitutional systems. This question of fit – or misfit – is larger than brief reflections about a specific example can possibly address. But if one has to start somewhere, what unfolded before the Victorian Hotel Quarantine Inquiry is as good a place as any.


FEW IN AUSTRALIA had heard of the idea of hotel quarantine until the weekend in late March 2020 when we learnt that this was how we were going to handle the threat of the novel coronavirus. Following a consensus decision by the newly formed National Cabinet, and using legal powers provided under relevant public health legislation, states and territories immediately got on with the job of putting hotel quarantine systems in place. The burden of this endeavour fell particularly heavily on the two states that carry the lion’s share of international passenger arrivals to Australia, New South Wales and Victoria.

Beyond general concern for the welfare of those detained within this new system, little attention was paid to how it actually worked until the news broke in June that something had gone very wrong in Victoria. There had been an outbreak of the virus – a significant one – and it appeared the key transmission point had been the contracted security guards installed within the hotels’ corridors to ensure that those detained did not leave their rooms.

Facing an anxious and angry public, on 2 July 2020 Victorian Premier Daniel Andrews announced an inquiry to get to the bottom of how the infection-control breaches had occurred. The Chair of that inquiry, retired judge Jennifer Coate, undoubtedly expected to uncover some governance problems in a high-stakes public-health operation that had been announced, designed and commenced over the course of a single frantic weekend. What she probably did not anticipate was that her investigation would expose a minefield of points of disconnect between fundamental constitutional demands on government action and the way the work of government was done in this instance.

Those who followed the inquiry will not soon forget the scale of the ‘whodunnit’ operation that saw a parade of senior political and governmental officials each strenuously deny any association with the decision to contract out the frontline task of hotel quarantine enforcement. If the ultimate failure of this exercise could be explained in terms of successive efforts to evade responsibility for the mess, we would be left with a sorry picture of the state of contemporary government – but at least it would be a simple one.

That is not what was going on. It appeared that this ‘decision’ had just happened, and no one seemed to have noticed anything strange about that.

Enter the orphan. 

When the inquiry’s final report was published in December, the line that attracted the most media attention was Coate’s description of the decision to engage private-security contractors to perform this frontline work as ‘an orphan’. As far as metaphors go, this ‘orphan’ hit the nail on the head. It summed up a scene in which a highly consequential decision had been made ‘without proper analysis or even a clear articulation that it was being made at all’, with no ‘contemporaneous rationale’ and ‘with no person or department claiming responsibility for having made it’. All of this, Coate was at pains to add, was ‘at odds with any normal application of the principles of the Westminster system of responsible government’.


THE ‘ORPHAN’ AT the centre of Victoria’s first hotel quarantine experiment offers a fitting frame for this essay because it represents the dystopia to which we must respond. In that dystopia, to borrow another metaphor, the cart seems too often to be put before the horse. The structural requirements and patterns of conduct associated with favoured governmental techniques such as contracting out seem to be given priority over the demands of our system of government. It’s as if the former should lead the latter, rather than the other way around.

So how are things meant to work?

Victoria’s first hotel quarantine program was instituted against the backdrop of the ground rules set by our Westminster system for all government action. Within that system, government actors are answerable to the parliament, and through the parliament to us, the electorate. The doctrine at the centre of this idea – ministerial responsibility – is key to the system’s operation. In the event of government failure in an area within their assigned portfolio, ministers of government are expected to demonstrate their responsibility through a range of mechanisms. These might range from pledging to undertake a positive program of reform through to resigning from office or from politics altogether.

The reason government action is answerable to the parliament in this way is simple. In our constitutional system the executive branch – the government – is drawn from the democratically elected parliament. A government holds office because parliament is prepared to continue to support it. The government’s engine room – the Cabinet – comprises ministers who have been assigned responsibility for particular portfolios of activity. Accountability mechanisms such as the doctrine of ministerial responsibility are therefore about closing the circle. They are meant to ensure that the authority entrusted to the government is used in the ways that parliament intended or can accept.

Government functions authorised by parliament obviously need to be performed by someone. That there is a domain of choice as to who performs these functions derives from the fact that this is generally a matter for the executive branch to decide. The accountability requirements, however, do not change. A government remains responsible for the performance of its functions irrespective of whether tasks associated with those functions are discharged by public officers appointed under legislation to perform certain roles, such as the police, or by personnel drawn from the private sector.


THE EFFORT OF the Victorian COVID-19 Hotel Quarantine Inquiry to find a parent for the decision to contract out frontline hotel quarantine enforcement to private security providers proved fraught because the dots upon which the application of these Westminster principles rely kept refusing to join up. This ultimately led to a historical ‘first’ for the state of Victoria: a sitting premier being called to give evidence at a judicial inquiry. Counsel assisting put it to Daniel Andrews that Victorians really should be able to know who made such a consequential decision – shouldn’t we? – and he agreed. But like all those who had given evidence ahead of him, the premier personally could not shed any light on the matter.

It’s one thing to grasp the circle of authority, responsibility and accountability that lies at the heart of Australia’s constitutional systems as a matter of theory. It’s quite another to understand how these imperatives might be shaped, in practice, by how a government function is actually performed. My argument is that the ‘orphan’ decision at the centre of Victoria’s hotel quarantine fiasco can only be explained by examining this interaction more closely, and that this examination makes it clear that the failures of the ‘normal application of the principles of the Westminster system of responsible government’ in this case were neither accidental nor exceptional. They were the product of certain features of the design and performance of contracting out that operate to undermine the very structure of our constitutional systems. Orphans are made, not born.

Contracting out is a practice that uses a legal device – a contract – to separate the delivery of a government function ‘out’ from the public sector for performance by the private sector. As a technique of privatisation, outsourcing is based on the assumption that the functions of government can be performed by anyone with the training and skills necessary to perform them. First, however, the relevant government function must be redescribed as a ‘service’ that can be the subject of a contractual arrangement between the government party and a private-sector service provider. This translation of functions to ‘services’ might occur in whole or in part. In the Victorian hotel quarantine example, it occurred in part: private contractors were engaged to deliver ‘security services’ associated with the dual detention and infection control imperatives of quarantine.

While the contractual arrangement is in place, the government’s role is typically limited to contract management and oversight; direct control over performance of the activity is relinquished. As for the obligations of the private contractor, these are owed only to the government party and are limited to what is specified in the contract. There is no legal, or political, relationship between the contractor and the members of the public who stand at the receiving end of any relevant ‘service’.

The starting assumption of much contemporary governmental practice in Australia is not just that the work of government can be contracted out, but that it should be. This position is likely one part ideological commitment and one part pragmatic necessity driven by the radical reduction of the permanent public service that has travelled alongside the ever-expanding practice of outsourcing. Yet despite its ubiquity, how contracting out actually works often seems to be as opaque to insiders as it is to outsiders. ‘Commercial in confidence’ requirements undoubtedly play a role here. But this opacity is not helped by the tendency of commentators to accept the idea that any effort to analyse the practice in general terms is inherently perilous. The ‘apples and oranges’ argument is always close at hand, emphasising that every instance of outsourcing has specific features and arises in a specific context.

Giving specificity its due is undoubtedly important in academic analysis and the ‘real world’ alike. But it’s equally important to ensure that attention to those different contexts does not forestall efforts to identify and analyse patterns that might cut across them. Looking for those patterns is especially critical when trying to illuminate problems at the interface between contracting out and the accountability requirements at the heart of the design of our constitutional arrangements.


A NUMBER OF matters that arose before the Victorian Hotel Quarantine Inquiry serve to illustrate the need not just for reflection, but also reform, at precisely this interface.

The first relates to what we learnt about the internal bureaucratic practices that have developed around outsourcing. Once it was settled that the ‘security’ aspect of hotel quarantine would be performed by private contractors, the task of securing those contracts was assigned to an agency – the Department of Jobs, Precincts and Regions. This department was otherwise unconnected to the agency – the Department of Health and Human Services – that supported the statutory officers who held the legal powers of quarantine.

Separating the business of contract creation from the agency responsible for the discharge of the quarantine function ensured that the frontline of hotel quarantine enforcement was only ever tenuously tethered to the possibility of ministerial responsibility. Coate’s final report observes that those who negotiated and supervised the contracts ‘were doing so without any clear understanding of the role of security in the broader Hotel Quarantine Program’ and with ‘no experience in security issues or infection prevention and control’.

It’s worth reflecting here on what it means to be ‘responsible’ for government action. Where ‘accountability’ looks backwards, ‘responsibility’ speaks as much to the requirement that someone undertake a particular function as it does to their obligation to answer for how it was performed. In the hotel quarantine example, however, this foundational constitutional requirement was at large from the outset. The one politician who ultimately did take ‘responsibility’ was the then Minister for Health, Jenny Mikakos, who resigned from Cabinet and the Legislative Council in September 2020. The evidence to the inquiry did reveal multiple problems in the governance structure of her department that had contributed materially to the operational flaws of the quarantine system. But as a response to the central drama at the point when she stood down – the quest to identify who made the decision to outsource hotel quarantine enforcement – Mikakos’ gesture of political responsibility was abstract at best. She had nothing to do with that decision, was at none of the relevant meetings and had no say in how this aspect of the program was designed and performed.

While the remedy to this may be to insist that the business of contracting out take place within the same portfolio agency that is responsible for carrying out a particular government function – surely a good start – there are still obstacles to overcome. Rationalised by concerns about conflicts of interest, a departmental ‘wall’ goes up between the process of contract creation and the minister politically responsible for the functions associated with the contracted-out service. Describing this practice before the inquiry, Martin Pakula, the minister responsible for the Department of Jobs, Precincts and Regions, gave evidence that it was not typical for him to be aware of contracts made by his department. Indeed, it was rare. Pakula did not look especially uncomfortable conveying this point because this was – and is – simply how things work, even if the effect is to render the idea of ministerial responsibility theoretical at best.

The second matter concerns the translational move from functions to ‘services’ that makes contracting out possible at all. That much more is at stake in this move than has likely been grasped was again well demonstrated in evidence at this inquiry.

Quarantine is a statutory function that combines detention under a legal order with the public-health imperative of infection prevention and control. In this instance performing that function was understood as requiring ‘security services’. The next consideration then appeared to be whether these ‘security services’ should be delivered by public officers (relevantly, Victoria Police) or private security contractors. The choice to be made between the two, to the extent that it was ever entertained, apparently came down to a judgement about the most efficient use of police resources.

This is all it took for the burden of a high-stakes and novel government operation to be shifted to people with no specialised skills in either detention or infection control. Indeed, the acquisition of those specialised skills was to be governed by contract or was simply assumed. Responsibility for training in infection prevention and control was assigned to the contractors, who would in turn provide training comprising a short online module. Some security guards did not even complete this before taking up their post in the hotels. As for the verbal de-escalation techniques to be used in the event a detainee became resistant, individual guards were assumed to have acquired these when qualifying for a security licence. This, too, proved incorrect in some cases.

The translation of government functions into ‘services’ also carries risks associated with the prevailing business model in the contracting industry. The key issue here is whether that business model leans heavily on subcontracting, or indeed sub-subcontracting. In this hotel quarantine example both were rife. This had a significant impact on the integrity of the quarantine frontline in terms of levels of training as well as the reliability of channels of communications about the job to be done. Coate made clear in her final report that there was no evidence of bad faith on the part of the contractors. But that was and is not the point. The point was that private security guards, drawn from a predominantly casualised workforce, ‘were not the appropriate cohort to provide security services in the Hotel Quarantine Program without close monitoring and extensive and continued training’.

The third matter concerns a question that floated above the drama of the inquiry rather than being the direct subject of its attention. What exactly did anyone expect to learn from finding out who made the decision to outsource the frontline of hotel quarantine?

The level of public interest in this pursuit was enormous. But even if this interest was born of nothing more than wanting to find someone to blame, to blame for what, exactly? For the fact that a major COVID-19 outbreak occurred on a government’s watch? Or for the fact that this same government didn’t give adequate thought to who should be installed at the frontline of such a critical governmental operation?

The fact that these two questions were closely related doesn’t change the fact that they speak to two different concerns. Nor does it change the fact that the second question will rarely be answered simply by finding out who made a particular decision. Identifying who made the decision to outsource hotel quarantine enforcement was never necessarily going to reveal why that decision was wrongheaded. So much was clear from evidence given by a range of senior government officials and politicians, each of whom indicated that they had no reason to think that private security was inappropriate for what one witness described as ‘this type of role’. It was only the Chief Health Officer who stated explicitly that ‘with the benefit of hindsight’ it was apparent that contracted security guards were the ‘wrong cohort’ for such a novel and high-stakes operation.

It is therefore necessary to untangle the difference between what we might have expected to learn from an answer about who made this decision – and what we need to learn from it.

To my mind, that lesson is simple: no one thought to question the decision because few ever question the suitability of contracting out, full stop. Coate’s observation that ‘[t]he people should be able to be satisfied that the action to proceed in this way was a considered one that addressed the benefits, risks and options available in arriving at such a decision’ speaks to the reflective void that was revealed before her.

It’s important to pause before assuming that what came to light during Coate’s inquiry provoked a collective epiphany about the need for a more critical approach to outsourcing. We would do better to ask whether those charged with responsibility for making such decisions ever think seriously about whether contracting out is, in a particular instance, appropriate. Public reactions to the hotel quarantine fiasco would suggest they should.

The kinds of questions that might produce a more principled evaluation of the appropriateness of outsourcing in a particular case are neither complex nor abstract. What is the government function to be performed? Is it novel? What kinds of public interest considerations might it engage? To what extent can it be broken down into specified tasks to make it amenable to being contracted out? What gaps or contingencies might be encountered if this is done? What is the character of the workforce that should be assigned? What lines of vision will be required? What is the rationale for engaging private contractors rather than public officers?

What matters, above all, is that these questions are asked at the outset rather than in hindsight. If answers suggest that outsourcing would be inappropriate in the particular instance, tinkering with the practice itself (by attempting a better ‘service specification’, for example, or more effective ‘monitoring’) is hardly going to do the trick.

Just don’t contract out at all.


IT’S WORTH OBSERVING, as a final point, that questions about the appropriateness of outsourcing have rarely, if ever, been the subject of significant parliamentary debate in any Australian jurisdiction. The advent and steady expansion of contracting out in this country has always been a project of the executive branch of government, acting of its own volition and operating on the evident assumption that the ways governments perform the functions we have entrusted to them is a neutral matter of technique with no particular political consequence.

Everything about this needs to be challenged. Outsourcing is clearly not neutral in its effects: how could it be when the point is to make government look and operate more like the private sector? As for its political consequences, Australian governments of all stripes have embraced the practice when in power. That, however, is not the measure of political consequence with which I am concerned. The point I want to make concerns the implications of contracting out for the political relationship between a government and the people it governs.

This political relationship is felt, in a concrete way, every time we interact with government – federal, state or local, and irrespective of the party in charge. It is found in the many small spaces in which we experience government power in action. And it is this political relationship that is most at risk of being rendered unintelligible when outsourcing intervenes. We remain subject to government power, but we are not actually interacting with government. If the scale of public interest in the ‘whodunnit’ question at the heart of Victoria’s Hotel Quarantine Inquiry is any indication, it would appear that this relationship could do with some urgent attention.

Because perhaps another lesson from the orphan is that people might care much more about how the work of government is done than contemporary governments wish to acknowledge. The way to find out is to ask. And the way we do that, in our constitutional system, is by putting the question to parliamentary debate.


IN THE END, there was no answer to the question of who had made the decision to engage private security providers for the frontline of Victorian hotel quarantine enforcement. As it turned out, the question itself was at odds with the starting assumptions and preferred practices of contemporary government. In that universe political actors seem increasingly to know little about how the work of government is done or of their ultimate responsibility for those choices. Much stands in the way of that understanding, by design, and few so far seem to have been troubled by the fact.

If this was the real problem behind the scenes that played out before the inquiry – and I think it was – it’s time to think seriously about what political accountability needs to mean in a governmental universe that operates this way. The spectacle of politicians resigning from their portfolios or from politics altogether undoubtedly has its place in the current era as much as any other. The problem is that the practices that trigger such accountability mechanisms tend to escape unreformed – and this continues to receive too little attention.

Some lessons clearly were learnt from the downfall of Victoria’s first hotel quarantine experiment. Closely following the recommendations of Coate’s interim report, Victoria’s program was completely ‘reset’ to ensure that it is overseen by a single agency that lies within the portfolio of a single responsible minister, with no central role given to private contractors. One might hope that lessons will equally be learnt from the implementation of Coate’s recommendation seventy-six that the Victorian Public Sector Commissioner examine the lines of accountability and responsibility between departmental heads and ministers and give guidance across the public service as to the obligations, both in law and practice, of all relevant public-office holders.

Yet there was something troubling about comments made by the Victorian premier on the day that the inquiry’s final report was published. Daniel Andrews suggested that the key lesson of the episode was not about who performs the work of government, but about the need for better ‘oversight’. If by oversight the premier had in mind the restoration of clear lines of ministerial responsibility for the performance of government functions, few could possibly disagree. If he had in mind better oversight of the work of contractors specifically, this too could hardly be argued with. But what of the starting point of the premier’s response: that lessons did not need to be learnt about who performs the work of government or how they perform that work?

This position cannot stand – at least not as a general proposition. We are no longer new to the game of contracted-out government service delivery. Its ubiquity is a sign of its vintage as much as its appeal to those who continue, if not to actively promote it, then at least to not question its use. It’s far from clear what is still to be learnt about the potential threat that such outsourcing can pose to the effective operation of our constitutional systems.

Surely, at this point, we would do better to simply accept that we have a problem, and that we will not see a ‘normal application of the principles of the Westminster system of responsible government’ until we begin to undo the patterns of thought and practice that have all but assigned those principles to the status of utopian ideals.

The principles of the Westminster system of responsible government are all we’ve got. We either protect them or we get used to a governmental universe in which the dots don’t join up, no one can explain, no one is responsible and no one who should be concerned seems to be especially worried.

Get the latest essay, memoir, reportage, fiction, poetry and more.

Subscribe to Griffith Review or purchase single editions here.

Griffith Review