What It Costs

Tax, Debt, and a State That Grows Without a Vote

A FIVE-PART SERIES

Part Three of Five

Part One examined the ideas; Part Two followed them through political and institutional networks; Part Three asks what the accumulated direction looks like when measured in public servants, regulation, taxation, expenditure and debt. Part Two examined Fabian gradualism inside the Labor Party and the union movement. Part Three turns to the numbers: the size of the administrative state, the true tax burden, and a debt that grows quietly enough that no election is ever fought over it.

Has the Australian state become bloated?

No parliament will ever vote for it, and no election will be fought over it: a permanent, ongoing tax rise that arrives simply because incomes grow while tax brackets do not. That is bracket creep, examined in full later in this instalment, and it illustrates the kind of incrementalism central to the Fabian method, the state growing not through a single decision anyone could challenge, but through dozens of measures too small, on their own, to be worth a fight.

Together, they can be measured in the plainest terms available: how large government has become, what it costs, and what Australians receive in return. The argument that follows is not that these measures bypass Parliament, nearly all of them pass through it in the ordinary way, but that their cumulative effect on the size and responsibilities of the state is rarely presented to voters as a single choice.

The direction was stated plainly before any of it happened. In a 2023 essay in The Monthly, Treasurer Jim Chalmers outlined what he called “values-based capitalism”: government and business as active participants in “shaping a better society”, built through public-private co-investment and the renovation of markets and institutions, rather than government simply regulating from outside them. The Future Made in Australia agenda, industrial subsidies and local-content requirements followed.

Whether or not any minister has read a Fabian tract, an economy in which government increasingly co-invests, co-owns and directs, and in which the Treasurer himself argues openly for that role, is close to the practical outcome the Fabian method was built to produce: not one decisive break with market economics, but a patient, program-by-program shift in which government becomes not only the referee of economic life, but an increasingly large part of the game itself.

What follows is what that stated direction looks like once it reaches the ground. The older Fabian idea of permeation is relevant here too: change need not abolish existing institutions when government can progressively work through them, reshape their incentives and purposes, and embed new responsibilities within structures that outwardly remain much as they were.

There is a reasonable case that the Australian administrative state has become larger, more expensive and more intrusive. Using the Australian Public Service Commission’s current historical series, APS headcount rose from 159,179 in June 2022 to 198,529 in June 2025, an increase of approximately 24.7 per cent. That period largely coincides with the present federal Labor government and warrants scrutiny.

Yet the figures need context. The APS is the federal public service. It is not a count of every Commonwealth employee, nor does it include state, territory and local government workforces. In June 2025, the APS represented 1.36 per cent of the Australian labour force, almost identical to its 1.35 per cent share in 2005. Service delivery was its largest identifiable job family.

So, raw headcount alone does not prove waste. A stable share of the labour force, though, does not mean the absolute growth is costless. Just over 39,000 additional public servants still draw just over 39,000 additional salaries, still require internal coordination across more layers of policy, communications and reporting, and still raise the question of what proportion of that growth reaches the frontline rather than the offices that manage it.

A growing population may require more public servants. Bringing outsourced functions back into government can increase headcount while reducing consultancy costs. New programs create genuine service demands. The stronger questions concern function, duplication and outcome.

Are Australians receiving better services? Are decisions being made more quickly? Have consultants genuinely bee n replaced, or has one layer of expenditure simply been added to another? And how much productive time do businesses, doctors, farmers, schools, charities and community organisations lose satisfying compliance requirements?

The Productivity Commission’s own findings begin to answer some of these questions, and not encouragingly. In the 2026-27 Budget, the government committed to reducing regulatory burden by an estimated $10.2 billion a year, while separate analysis commissioned by the Australian Institute of Company Directors put the current cost of federal regulatory compliance at $160 billion a year, up from $65 billion in 2013.

The Productivity Commission’s own interim report, Creating a more dynamic and resilient economy, the first of five interim reports released under its 2025 Five Pillars of Productivity inquiry, did not dispute the direction of travel. It found that the average wait for a house-building planning decision in the ACT is nearly six months; that approving a windfarm in NSW takes an average of nine years; and that Brisbane City Council had to introduce a 31-step checklist jus t to help people open a café. As the Commission put it: “We have most of the tools and procedures we need to regulate well, but they are just not working.”

Red tape is therefore not merely a complaint made by people hostile to government. It is recognised as a serious economic and institutional problem by one of the Commonwealth’s principal advisory bodies.

The same pattern appears in the energy transition. Rewiring the Nation alone commits $20 billion in concessional finance to modernise the electricity grid, administered through the Clean Energy Finance Corporation in pursuit of an 82 per cent renewable-generation target for the National Electricity Market by 2030. Layered on top are cleaner-fuels programs, battery subsidies, and the institutional apparatus required to plan, finance and regulate a transition of that scale. Each increment creates additional public-sector roles and compliance obligations for businesses and households. None of it was put to voters as a single, explicit choice between the existing energy system and a permanently larger directing role for the Commonwealth.

The National Disability Insurance Scheme illustrates the same pattern at the level of a single program. Designed in 2013 to support around 400,000 Australians with permanent and significant disability, the Scheme supported 717,001 participants by 31 March 2025. The 2023-24 Annual Financial Sustainability Report projected Scheme expenses of $210.3 billion over the four years to June 2028; the 2024-25 AFSR, using data to June 2025, projected $225.3 billion over the four years to June 2029. Because the two figures cover different rolling four-year windows, they should not be read as a straightforward upward revision of one another. The Scheme had been on a trajectory of expenditure growth materially faster than the economy, prompting the sustainability reforms discussed below.

An independent review, co-chaired by Professor Bruce Bonyhady AM and Lisa Paul AO PSM and reporting in December 2023, was commissioned in part to restore what its own terms of reference called trust, confidence and pride in the Scheme, language that concedes those things had been lost. The NDIA’s own integrity chief, John Dardo, told Senate Estimates in June 2024 of a growing catalogue of misuse, and Opposition NDIS spokesperson Michael Sukkar has since characterised NDIA officials’ evidence as showing around $2 billion of the Scheme’s $45 billion annual budget being diverted from genuine participant needs, including misuse by organised crime.

In April 2025, the NSW Crime Commission, working with NSW Police and the NDIA, restrained more than $40 million in assets, 36 properties, two luxury vehicles, shares and cash, as part of an ongoing fraud investigation. Officials told Senate Estimates in June 2024 that courts could not keep up with the volume of exploitation cases, citing instances including a $20,000 holiday and a $73,000 car funded through fraudulent claims.

The government has responded: a Fraud Fusion Taskforce has opened more than 500 investigations, and cost growth has slowed under recent reforms. But none of this means the Scheme should not exist. It supports people who could not otherwise live with dignity and independence, and that purpose is not in question here. What it does show, again, is that a program can be well intentioned, genuinely necessary, and still grow beyond the state’s capacity to administer it honestly, for years, before the scrutiny that should have caught it earlier finally arrives.

A parallel expansion is under way in early childhood education and care. The activity test that once limited subsidised care to working families was removed in January 2026, guaranteeing at least 72 hours of subsidised care a fortnight, equivalent to at least three days per week, to every eligible child regardless of parental employment. A further $1 billion Building Early Education Fund is intended to support around 160 new or expanded services. Of this, $500 million is for capital grants, while a further $500 million has been set aside for potential Commonwealth investment in an owning-and-leasing model, still subject to development and business-case work.

Workforce wages have risen 15 per cent since 2024, with a further $3.6 billion committed in 2026 to extend the increase. What began as targeted subsidy support has become a structural shift of responsibility for the early years from the household toward the state, again without a discrete electoral mandate for that particular boundary.

The Digital ID system, already examined earlier in this series, continues the same trajectory: additional budget, expanded regulatory capacity, and the progressive opening of the system to private-sector participation from December 2026. A separate regulatory stream, online safety, has moved on its own parallel track: the under-16 social media minimum age law took effect on 10 December 2025, and further eSafety Commissioner codes expanding platform obligations have continued to come into force as late as March 2026. Infrastructure and powers accumulate across both streams. Compulsion, where it exists, tends to arrive gradually rather than by a single vote.

The broader economic data adds a further dimension. Australia recorded seven consecutive quarters of falling real GDP per capita, a run that continued through the September quarter of 2024 and ended with a small rise in the December quarter, described by economist Gene Tunny of Adept Economics, writing in the edited volume Promise and Performance: Albanese’s First Term (Connor Court Publishing, 2025), as a “per capita recession.” By early 2025, real net national disposable income per capita had fallen 4.7 per cent from its mid-2022 level, among the largest declines in living standards recorded across the OECD over that period. More than 1.1 million jobs were created over the same term, yet Tunny’s analysis of ABS data shows that half of all new jobs during the government’s first term were in publicly funded sectors, healthcare, education and public administration, rather than the private, trade-exposed economy that ultimately funds them.

Even judged against its own stated achievements, the record is mixed rather than damning: the same government held the line on spending against real political pressure, delivered the first back-to-back budget surpluses in almost two decades, and brought inflation back within the Reserve Bank’s target band. Economists disagree, reasonably, about how much weight the weaker figures should carry against those genuine successes.

What about taxation?

The experience of taxation is not measured only by an international league table. Households encounter income tax, bracket creep, GST, rates, levies, fuel excise, registration costs and charges embedded in the price of regulated goods and services, and businesses also carry the cost of compliance.

According to the OECD’s headline measure, Australia’s tax-to-GDP ratio was 29.9 per cent in 2023, below the OECD average of 33.7 per cent that same year. This is the figure most often used to call Australia a low-tax country.

That comparison, on its own, doesn’t tell the whole story. The OECD measure excludes Australia’s compulsory Superannuation Guarantee; the Centre for Independent Studies argues that compulsory superannuation should be treated analogously to the compulsory social-security contributions counted as tax in many OECD countries, and separately notes that the OECD average is heavily weighted toward Europe’s high-tax, high-spend model.

The official ABS measure tells a related story on its own terms: total taxation revenue across all levels of government reached 30.2 per cent of GDP in 2024–25, close to the historical record. The free-market think tank Centre for Independent Studies has separately argued that once the compulsory Superannuation Guarantee is added, on the basis that it functions like the social-security contributions most other OECD members already count as tax, Australia’s ratio sits closer to 34.5 per cent, near the middle of the OECD range rather than well below it. A second CIS analysis, adjusting GDP itself to exclude government’s own operating surplus on the grounds that it is not part of the economy’s true taxable capacity, arrives at a similar but distinct figure of 34.6 per cent. The two adjusted figures rest on different methodologies rather than a single calculation, but they point the same way.

It would therefore be inaccurate to describe Australia as either one of the world’s most highly taxed countries, or as a clearly low-taxing one. The honest position sits closer to the middle of the developed world than either side of that debate usually admits.

A more defensible concern than the overall tax rate is what Australians receive in return for it, and where the strain is genuinely concentrated. Household disposable income and GDP per capita both sit comfortably above the OECD average. Housing is the exception, and the sharpest one.

OECD analysis describes Australian housing affordability as strained, with high housing costs in major cities and supply constrained in part by land-use restrictions, even as tax and regulatory settings continue to grow more complex around it. This is not a coincidence of timing. The same red tape already discussed, a six-month average wait for a single planning decision in the ACT, compliance obligations layered on developers project by project, is a direct contributor to why supply cannot meet demand. A state that can find $22.7 billion to direct industrial policy and $20 billion to rebuild the energy grid still takes six months to approve the construction of a single house. The one failure that touches nearly every Australian directly is the one the administrative state has struggled hardest to solve.

The same tax base also funds the energy transition, the expanding early-childhood guarantee, and the administrative architecture of Digital ID and online safety, already examined above. Bracket creep and the overall tax take therefore underwrite not only existing services but the gradual assumption by the Commonwealth of new directing and co-ownership roles that were never presented to voters as a single package.

The issue, in other words, is not only how much government collects. It is what government has assumed responsibility for, how effectively it carries out those responsibilities, whether meaningful limits remain, and why the one area where Australians are genuinely squeezed relative to their peers has been allowed to become so.

What about debt and inflation?

Australia’s federal gross debt has exceeded $1 trillion, reaching $1.051 trillion, or about 34.0 per cent of GDP, by 30 June 2027, under the 2026-27 Budget. The Parliamentary Budget Office, Australia’s independent fiscal watchdog, puts the raw size in perspective: debt remains low relative to comparable economies. Its own sharper concern lies elsewhere, in the trajectory rather than the total.

The cost of servicing that debt is what should concern Australians most. Interest payments reached $24.4 billion in the most recent financial year and are projected to climb toward $38–40 billion by 2028–29. The PBO’s own words: interest is now “one of the fastest growing areas of spending” in the federal budget, ahead of most other categories of government activity. Some of that spending is already familiar. The $22.7 billion Future Made in Australia commitment discussed earlier is one identifiable contributor among many, a decade-long spending program legislated without a referendum, a plebiscite, or any mechanism beyond the ordinary passage of a budget.

A fair reader could ask what the same $22.7 billion might have done differently: lower compliance costs, faster approvals, a tax and regulatory environment in which every business, not a government-selected handful, had a genuine opportunity to compete and succeed. That case has real defenders, and it is not obviously wrong. Governments elsewhere are subsidising the same industries, and an economy that unilaterally declines to compete risks losing the sector to nations willing to underwrite it. What gives the choice a recognisably Fabian character in method is not the subsidy itself. It is how the authority to make it was acquired: gradually, funded by a tax base that grows quietly through bracket creep, and never once put to voters as an explicit choice between a state that directs the economy and a state that levels the field for everyone within it.

The same budgets that service rising interest also lock in liabilities well beyond any single term of government: the childcare guarantee, energy transition infrastructure, and 230,000 additional Commonwealth-funded university places under the Universities Accord, administered by a newly created Australian Tertiary Education Commission with the power to allocate places from a government-controlled pool. Each commitment is reasonable in isolation. Together, they grow the future claim on taxpayers without a fresh electoral contest ever being held over the cumulative size of the state’s role.

Bracket creep deserves particular attention, because it illustrates this series’ central argument in a domain that has nothing to do with ideology. Without a single piece of new legislation, the Parliamentary Budget Office’s latest outlook projects personal income tax revenue to rise from 12.6 per cent of GDP in 2026–27 to 14.7 per cent by 2036–37, and the average personal tax rate to climb from 24.9 per cent to a historical high of 28.6 per cent over the same period, in the absence of further announced policy.

No parliament need vote for each incremental increase. No election will be fought over it. Inflation and wage growth simply push incomes into tax brackets that were fixed years earlier, and government revenue rises accordingly. The Parliamentary Budget Office’s own history of the practice describes governments deliberately using bracket creep as their preferred instrument of budget repair, “returning” a portion through a legislated tax cut every several years while allowing it to rebuild quietly in between. This is gradualism in its purest form: a change in the real relationship between citizen and state that requires no one to notice it happening.

Inflation compounds the same problem from another direction. Headline CPI has fallen from its December 2022 peak of 7.8 per cent to 3.8 per cent in mid-2026, still above the Reserve Bank’s 2–3 per cent target band. The claim that official inflation understates the true rate by as much as double is not well supported; the methodology critiques that hold up under scrutiny have historically pointed the other way; and Australia’s headline rate has not approached 8–10 per cent outside the acute 2022 shock.

But the perception that inflation has run far hotter than the headline number is not simply a misunderstanding. It reflects something real: essential, largely unavoidable categories have run far above the average for extended periods. Electricity prices rose 21.5 per cent in the twelve months to December 2025 alone; by August 2023, insurance prices were 14.7 per cent higher over the year and rents 7.8 per cent higher. The ABS’s own Living Cost Indexes show that household groups whose spending concentrates in these categories experience cost growth materially different from headline CPI, because their expenditure patterns differ from the basket the headline figure averages across.

The 21.5 per cent figure needs its own caveat before it can carry any argument. The ABS is explicit that the increase was primarily driven by state government electricity rebates in Queensland and Western Australia being used up by households over the year, not by underlying retail price movements. Excluding the effect of Commonwealth and state rebates, the ABS puts the actual rise in electricity prices at 4.6 per cent over the same twelve months, still above general inflation, but nowhere near the headline figure, and not evidence, on its own, of transition-driven cost growth. Any separate case that grid-rebuild spending, renewable subsidies or transition compliance costs are pushing up the underlying price of electricity needs to rest on that 4.6 per cent figure and other network- and generation-cost evidence, not on the rebate-inflated headline number.

Small businesses feel this compounding pressure with particular sharpness. COSBOA’s 2026 Small Business Perspectives Report, the peak body’s fifth annual study and its first focused exclusively on regional, rural and remote Australia, surveyed 572 regional small business owners and found 87 per cent had seen operating expenses rise over the past year, 73 per cent had seen profit fall, and 77 per cent had experienced business-related stress or anxiety as a result. A small business absorbing an electricity increase, an insurance premium or a new compliance cost has no equivalent to the scale a large corporation can spread the same cost across. The same rise in the cost of doing business lands as a rounding error on one balance sheet and an existential threat on another. These businesses were never asked to vote on the cumulative settlement whose costs they now carry. The costs they now absorb are not simply high. They are, like so much else in this account, arriving without anyone having asked.

None of this is hidden. All of it is published, quarterly, by agencies whose entire purpose is to report it accurately. But a rising debt serviced by rising interest payments, a tax system that quietly raises itself through bracket creep, and cost pressures that fall unevenly across household types, together describe a state whose real burden grows steadily larger without ever requiring a single democratic decision that says so out loud.

None of this settles the deeper question this series is actually asking. The question underneath the figures is therefore the same one with which this series began: are institutions and economic systems ultimately ordered toward the human person, or is the person progressively being ordered around the requirements of the system? A society can argue endlessly about the right size of government and still never ask why the vulnerable matter in the first place. That is where the Christian tradition has something to say that no budget line can capture.

This is a very old Christian instinct, not a modern political slogan: responsibility belongs first to the household, then the parish, then the town, and only then, and only where those cannot reach, to the province or the Commonwealth. A household that no longer manages its own economic margin has not been relieved of a burden. It has been quietly relieved of a responsibility.

Where does this leave us? A tax system that quietly raises itself, a bureaucracy that grows without a single dramatic vote, and interest payments that outpace most other spending all change the relationship between citizen and state without ever asking the question out loud. What kind of limits should a Christian still insist upon, even while rendering what is due?

“For the one in authority is God’s servant for your good... Give to everyone what you owe them: if taxes, then taxes; if revenue, then revenue; if respect, then respect; if honor, then honor.” — Romans 13:4–7 (NIV)

Paul’s point is not that the state is owed unlimited submission, but that it is owed its due, and that determining what is due requires discernment, not passivity.

Part Four turns to that Christian tradition directly, not as a slogan against government, but as an entirely different account of why the poor, the vulnerable, and the institutions built to serve them matter in the first place.

Thank you for reading

God Bless

Mark