A Working Paper

Andrew Paul Klein

Abstract

This paper examines the structural relationship between two apparently distinct developments in Australian governance: the emergence of “whole of government” as an organising principle of public administration from the late 1980s and its formal codification in 2004, and the parallel embrace of neoliberal free-market theory by both major political parties from the mid-1980s. It argues that the two were not merely contemporaneous but structurally complementary. The New Public Management reforms of the 1980s and 1990s fragmented the public service into silos. Whole-of-government then supplied the coordinating architecture required to manage the marketised service delivery that the silos produced. The result was a system in which market-based governance operated at scale, while the silos and the coordination framework diffused responsibility for the outcomes. The paper traces the timeline of convergence, identifies the mechanisms of mutual reinforcement, examines the language of cohorts and categories that accompanied the process, and applies Bandura’s mechanisms of moral disengagement and the Confucian principle of zhengming to assess what was produced. It concludes that the perception of credibility was manufactured, and that the lived experience of the Australian public was something else.

1. Introduction: The Question of the Dovetail

Two things happened to Australian governance in the space of about twenty years. The public service was fragmented — split into agencies, business units, and contracted-out functions under the banner of New Public Management. Then it was reintegrated — brought together across portfolio boundaries under the banner of whole of government.

At the same time, both major parties converged on a set of policies known as neoliberalism: deregulation, privatisation, marketisation, competition, outsourcing. The state withdrew from direct provision and reconstituted itself as a market designer and steward.

The question this paper asks is whether these two developments were related. Not in the sense of a conspiracy, but in the sense of a structural dovetail. Did the fragmentation of the public service, and its subsequent coordination under whole-of-government language, create the architectural conditions required for neoliberal governance to operate at scale? And did neoliberalism, in turn, provide the ideological justification that made the coordination appear technical rather than political?

The paper argues that the answer is yes to both.

2. The Parallel Timelines

2.1 The Fragmentation: New Public Management (1983–1996)

The structural transformation of the Australian Public Service began well before “whole of government” was formalised.

The Hawke government’s 1987 machinery-of-government changes were the pivot. In July 1987, the number of Commonwealth departments was reduced dramatically — from 28 to 16 — creating “mega-departments” and significantly strengthening the Department of the Prime Minister and Cabinet. The reform was presented as a rationalisation. In practice, it centralised political control while leaving the operational structure of agencies intact.

This was followed by the New Public Management (NPM) reforms of the late 1980s and early 1990s. NPM introduced:

· Competitive tendering and contracting out.

· Purchaser-provider splits (the state buys services rather than delivering them).

· Output-based budgeting and performance measurement.

· Commercialisation and corporatisation of government functions.

· Devolution of operational responsibility to agencies while centralising strategic control.

The 1996 election of the Howard government accelerated the process. The creation of Centrelink commercialised welfare delivery. The Job Network introduced competitive contracting for employment services. The National Competition Policy payments tied state compliance to market reform. The Sibley Report (1997) and the Uhrig Report (2003) reviewed and refined the architecture.

The effect of NPM was to fragment the public service. Agencies became semi-autonomous units with their own performance measures, budgets, and accountability frameworks. The vertical program structures — welfare, employment, aged care, disability — became separate silos, each with its own logic, its own data, and its own contractor relationships.

Status: Established.

2.2 The Reintegration: Whole of Government (2002–2004)

The COAG trials from 2002 and the Management Advisory Committee’s Connecting Government report of April 2004 introduced the coordinating framework.

The report defined whole of government as “public service agencies working across portfolio boundaries to achieve a shared goal and an integrated government response to particular issues”. The stated aim was to address “wicked issues” — complex social problems that cut across vertical program structures.

But the report also contained a warning, quoted in full in the accompanying paper:

“There is also the danger of excessive political control constraining advice that does not reflect prevailing political orthodoxy: ‘whole-of-government’ can become a euphemism for everyone to be ‘on message'”.

The coordinating mechanisms that emerged — Indigenous Coordination Centres, whole-of-government taskforces, the Secretaries Board, the central agencies — did not rebuild the vertical structures that NPM had fragmented. They added coordination layers on top of the silos.

Status: Established.

2.3 The Ideological Parallel: Neoliberalism (1983–Present)

The same period saw both major parties converge on neoliberal policy.

The Hawke-Keating governments floated the dollar (1983), deregulated the financial system, dismantled tariffs, and began privatisation. The Howard government deepened the process. The Rudd-Gillard-Rudd governments continued it. The Abbott-Turnbull-Morrison governments extended it. The Albanese government has not reversed it.

The Productivity Commission became the institutional architect. As documented in The Neoliberal Turn, the Commission has “led analysis on why, when and how market-based competition can improve human services, playing a central role in the design of key market schemes, including the NDIS and Child Care Subsidy“. It recommends that government should “recognise and actively engage in its role as a market designer and market steward”.

Status: Established.

3. The Dovetail: How the Two Reinforced Each Other

The timelines are not merely parallel. They are structurally interdependent. The following mechanisms of mutual reinforcement can be identified.

3.1 Fragmentation Required Coordination

The NPM reforms fragmented the public service into silos. But silos cannot deliver marketised services without coordination. A market-based NDIS requires a market steward. A market-based aged care system requires an assessment framework. A market-based employment services system requires contract management.

Whole-of-government supplied the coordinating architecture. It provided the language — “shared goals,” “integrated responses,” “wicked issues” — and the institutional mechanisms — central agencies, taskforces, coordination units — that allowed the market-based system to be managed.

Without whole-of-government, the silos would have produced chaos. With it, the silos could be coordinated in the service of market design.

Status: Inference.

3.2 Coordination Centralised Political Control

Whole-of-government did not distribute power across the silos. It concentrated it in the central agencies — the Department of the Prime Minister and Cabinet, the Department of Finance, the Treasury.

The Secretaries Board, established under the Public Service Act 1999 and strengthened after 2020, became a mechanism through which the Secretary of PM&C coordinates the Secretaries of other departments. The central agencies set the policy framework; the line agencies deliver within it.

This centralisation served the neoliberal project in a specific way. Market design is a central function. The Productivity Commission, the Treasury, and the Department of Finance are the institutions that design markets. The line agencies are the institutions that operate them. Whole-of-government gave the central agencies the authority to override line agency concerns.

Status: Inference.

3.3 Coordination Diffused Responsibility

This is the crucial structural point. The combination of silos and coordination produced a system in which no single entity is accountable for the outcomes.

· The silo is accountable for its own performance. It delivers its program. It reports its outputs.

· The coordinating body is accountable for coordination. It ensures the silos talk to each other.

· The central agency is accountable for policy design. It sets the framework.

When a program fails, responsibility is diffused. The silo says: “We delivered what was specified.” The coordinating body says: “We coordinated the coordination.” The central agency says: “The silos failed to deliver.”

This is the architecture documented in the Robodebt Royal Commission. The scheme was not built by any single agency. It was built through the interaction of the Department of Human Services, the Department of Social Services, the ATO, and the political level. Each bore some responsibility. None bore enough.

Status: Established.

3.4 The Market Required a “Non-Political” Justification

Whole-of-government supplied that justification. Coordination sounds technical. Integration sounds administrative. Shared goals sound pragmatic. The language is the language of public administration, not political ideology.

This allowed the market-based restructuring of services to proceed without appearing to be an ideological project. When government “reforms” aged care by introducing a market, it does so under the banner of “whole of government,” “client choice,” and “consumer-directed care.” The market is presented as the mechanism for delivering coordination. The politics is buried in the language of administration.

Status: Inference.

3.5 The Silos and Coordination Together Produced the Consulting Layer

The fragmentation of the public service — the loss of in-house capability, the outsourcing of policy and program design — created a demand for external expertise. The consulting firms supplied it.

The consulting firms also supplied the coordination narrative. They produced the reports, designed the frameworks, and conducted the reviews that justified the market-based restructuring. KPMG, PwC, Accenture, Nous Group, and the rest operate across the silos, providing the connective tissue that whole-of-government nominally provides. In practice, they are the connective tissue.

This is documented in the report The Architecture of AI Governance: “Management advisory services” account for 53 per cent of high-value consultancy contracts, “work that often overlaps with core functions traditionally undertaken by the Australian Public Service”.

Status: Established.

4. The Language of Cohorts and Categories

The combination of fragmentation and coordination produced a specific language form. The individual citizen is not addressed as a person. They are addressed as a member of a cohort.

· Welfare recipients become “jobseekers“.

· Disabled people become “NDIS participants“.

· Older people become “aged care consumers“.

· Veterans become “claimants”.

· Homeless people become “people experiencing homelessness”.

· Patients become “health care consumers“.

The function of this language is not merely administrative. It is categorical. It sorts individuals into groups that are managed by different silos, funded by different programs, and measured by different indicators.

The individual who does not fit the category — the person with fluctuating disability, the casual worker whose income varies week to week, the veteran with a complex claim — is not served by the system. They fall into the gap between the categories. The whole-of-government framework is supposed to catch them. In practice, the framework has never been designed to see them.

This is the architecture of the Robodebt scheme. The recipient is a discrepancy in the data. The individual is the gap.

Status: Inference.

5. The Analytical Frameworks

5.1 Bandura’s Mechanisms in Operation

Four of Bandura’s mechanisms of moral disengagement are structurally embedded in the architecture.

Euphemistic labelling. “Whole of government,” “client choice,” “consumer-directed care,” “market stewardship,” “reform,” “efficiency,” “sustainability.” Each term sanitises a change in the distribution of power and resources.

Diffusion of responsibility. As documented above. The silos and the coordination layers ensure no single entity is accountable for the outcome.

Displacement of responsibility. The market is presented as the source of the outcome. “The market will decide.” “The market will punish.” “The market will sort the efficient from the inefficient.” The political decision is displaced onto an entity that cannot be questioned.

Dehumanisation. The individual becomes a category. The category is processed. The processing produces an outcome. The person who is processed is not in the decision.

Status: Inference.

5.2 Zhengming: The Rectification of Names

The Confucian test is straightforward. Does the name correspond to the reality?

The name is “whole of government“. The reality, as documented in the accompanying paper, is fragmentation, turf wars, silos, and failed experiments.

The name is “market stewardship“. The reality, as documented in the six portfolios examined, is market design in the service of corporate interests: the NDIS being cut, aged care being marketised, the tax system operating on two tiers, data centres approved before regulation, veterans’ reforms implemented slowly.

The name is “shared goals“. The reality is that the goals of the central agencies are not the goals of the line agencies, which are not the goals of the contractors, which are not the goals of the citizens.

The rectification of names would require an honest account of what the language has produced. That account is documented in the record. What has not yet occurred is the honesty.

Status: Inference.

6. The Perception of Credibility

The dovetail produced something more than an architecture. It produced a perception.

The perception was that the Australian government was:

· Coordinated — the language of whole of government.

· Efficient — the language of NPM.

· Reformed — the language of continuous improvement.

· Modern — the language of market design.

· Accountable — the language of performance frameworks, KPIs, and reviews.

The perception was manufactured. It was produced by the language, by the consulting firms, by the think tanks, and by the mainstream media that reported the language without checking the outcomes.

The lived experience of the Australian public was different. The NDIS participant who cannot access support. The aged care resident who dies on a waiting list. The welfare recipient who receives an unlawful debt. The veteran whose claim is rejected. The homeless family on a waiting list. The small business owner processed by the ATO’s automated recovery system. The taxpayer whose data flows into Palantir’s analytics platform without consent.

The gap between the perception and the lived experience is not a failure of communication. It is a structural feature of the architecture. The architecture was designed to produce coordinated market governance. The perception of credibility is the byproduct. The lived experience is the cost.

Status: Inference.

7. Conclusion: The Dovetail and the Citizen

The question was whether “whole of government” and neoliberalism were related. The evidence supports the finding that they were dovetailed.

NPM fragmented the public service into silos. Whole-of-government supplied the coordination architecture required to manage marketised service delivery at scale. The coordination concentrated political control in central agencies, diffused accountability across the silos, and produced the language that made the market-based restructuring appear technical rather than ideological.

The result is a system in which:

· The market is sovereign.

· The state is the market designer.

· The citizen is the cohort.

· The outcome is the metric.

· The metric is the perception.

· The perception is the credibility.

The lived experience of the Australian public — the deaths on aged care waitlists, the unlawful debts, the veterans abandoned by the claims system, the homeless on the waiting list — is not in the metric. It is in the gap between the metric and the person.

The record is kept. The dovetail is named. And the perception of credibility stands in for the reality.

Claim- Status – Summary

# – Claim- Status

1 – Hawke’s 1987 machinery changes reduced departments from 28 to 16- Established

2 – NPM fragmented the public service into silos – Established

3 – Whole-of-government was codified in 2004 – Established

4 The 2004 report warned it could become a euphemism Established

5 – Both major parties converged on neoliberal policy from 1983 – Established

6 – The Productivity Commission is the market design architect – Established

7 – NPM fragmentation required coordination architecture – Inference

8 – Whole-of-government centralised political control- Inference

9 – Silos plus coordination diffuses responsibility – Established (Robodebt)

10 – Whole-of-government supplied non-political justification – Inference

11- The consulting layer supplies the connective tissue – Established

12 – The language of cohorts categorises the individual- Inference

13 – The dovetail produced a manufactured perception of credibility- Inference

14 – The lived experience of the public is the cost of the architecture – Inference

Method notes – Claims are classified throughout as Established, Inference, or Speculation. The paper examines whether the emergence of “whole of government” language and practice in Australian governance was structurally correlated with — and enabled — the application of neoliberal free-market theory. It does not argue that “whole of government” was designed as a neoliberal instrument. It argues that the two dovetailed, and that the architecture they jointly produced created the conditions in which market-based governance could be applied at scale, while the adverse outcomes of that application were obscured by the language and structure of the coordination framework.

References

1. Management Advisory Committee. (2004). Connecting Government: Whole of Government Responses to Australia’s Priority Challenges. Australian Public Service Commission.

2. Shergold, P. (2004, April 20). Launch of Connecting Government. Speech.

3. Productivity Commission. (2017). Data Availability and Use: Inquiry Report No. 82.

4. Productivity Commission. (2025). Competitive care: Why, when and how competition can improve human services. Conference paper.

5. Royal Commission into the Robodebt Scheme. (2023). Report. Commonwealth of Australia.

6. Australian Public Service Commission. (2024). Statement on the Robodebt Centralised Code of Conduct Inquiry.

7. Centre for Public Integrity. (2026, August 6). New Centre for Public Integrity analysis shows Commonwealth government remains heavily reliant on private consultants. Media release.

8. Bandura, A. (1999). Moral Disengagement in the Perpetration of Inhumanities. Personality and Social Psychology Review, 3(3), 193–209.

9. Confucius. The Analects (13:3).

10. Marks, G. (2008). Coercive Governance and Remote Indigenous Communities. Australian Indigenous Law Review, 12(1).

11. Sullivan, P. (2007). Strange Bedfellows: Whole-of-Government Policy and Shared Responsibility Agreements. Ngaiya: Talk the Law, 5.

12. Podger, A. (2022). Allan Barton Memorial Research Lecture. Royal Society of Australia.

13. Scamps, S. (2025, November 4). Ministerial ‘turf wars’ are stalling critical policy reforms. The Point.

14. The Neoliberal Turn: How Market Ideology Entered Human Services and Why It Required Measurement. (2026). The Patrician’s Watch.

15. The Architecture of AI Governance in Australia: Contracts, Infrastructure, and the Political Economy of Scrutiny. (2026). The Patrician’s Watch.

16. Australian Government Priorities: Corporate and Extractive Interests Before the Welfare of Citizens. (2026). The Patrician’s Watch.

17. The Rectification of Names: “Whole of Government” as Euphemism, Political Control, and the Failure of Coordination. (2026). The Patrician’s Watch.

Prepared for the record. The dovetail is named. The perception is documented. And the citizen is in the gap.