The 2026-27 Budget, against the data

Every claim in this finding is sourced. Where the arithmetic is the author's, the arithmetic is shown. Where the reading is interpretive rather than computed, it is named as such. Where the public record is not yet sufficient to settle a question, the question is left open.

The 2026-27 Federal Budget contains the most consequential set of changes to Australia's tax system in a quarter of a century. It also contains a number of moves whose direction in the appendix tables differs from the direction in the press kit. What follows is what survives a forensic reading, with the receipts attached.

The shape of the document

Real GDP growth slows to 1.75 per cent in 2026-27, from 2.25 per cent the year before, recovering to 2.5 per cent in 2027-28. Inflation runs at 5 per cent through the year to June 2026. Unemployment rises to 4.5 per cent. Net overseas migration falls from 245,000 in 2025-26 to a plateau of 225,000 by 2028-29. The underlying cash deficit holds at 1.0 per cent of GDP across the forwards, narrowing to 0.7 per cent in 2029-30. Cumulative deficits over the forwards total 150.5 billion dollars.¹

Treasury's commodity assumptions: iron ore declines to sixty US dollars a tonne over four quarters to end-March 2027; metallurgical coal to 140 US dollars; thermal coal to 70; LNG to ten US dollars per million British thermal units. Tapis crude averages 100 US dollars a barrel in the June quarter 2026 before declining to the mid-seventies. The Australian dollar is held at 70 US cents.¹

1. The five tax cuts raise individual income tax as a share of GDP

The cleanest finding in the document is in Statement 4 of Budget Paper 1. The chart data table for revenue as a percentage of gross domestic product reports the following projections.²

A rise of half a percentage point. At Treasury's projected nominal GDP in 2029-30 (roughly 3.5 trillion dollars, derived from the receipts share), half a percentage point is about 17 billion dollars per year in additional individual income tax revenue, after the five tax cuts have been counted.^a

The Treasurer's case-study worker on average earnings of 81,245 dollars receives a combined 2,816 dollars per year by 2027-28 from the five layered cuts.³ This is true for that worker. At the level of the whole tax base, the layered cuts return less than bracket creep adds. The structural drift through every unchanged threshold continues to lift income tax as a share of the economy.

The press kit explains the cut. The Statement 4 table explains the net direction. Both are accurate. They answer different questions.

2. The structural revenue shift is workers up, companies down

The same Statement 4 table records the offsetting shift on the company side.²

A 0.7 percentage point fall. In nominal terms, company tax receipts are projected at 151.3 billion dollars in 2025-26 and 153.9 billion in 2029-30, a real-terms contraction across the forwards. The driver is the assumed commodity-price path. Iron ore at sixty US dollars a tonne compresses resource-sector profits, and therefore receipts.

Net total taxation revenue holds at roughly 24 per cent of GDP across the forwards. The composition shifts. Workers' share rises by 0.5 percentage points. The company and resources share falls by 0.8 percentage points combined (company tax minus 0.7, resource rent tax minus 0.1). The Australian economy taxes itself at the same overall rate. It taxes workers more and companies less. This is the largest structural revenue movement in the document. It receives no mention in the press kit.

3. The negative gearing reform does not add a house

The Government models 75,000 additional owner-occupiers over the next decade from the negative gearing and capital gains tax changes. Treasury describes the mechanism honestly in its own text, on page 24 of the Budget Overview:⁴

"The small impact on housing supply from these tax measures will be more than offset by measures that increase housing supply."

The tax change rebalances ownership of the existing dwelling stock. Investors who would have purchased an established property in a post-reform tax setting purchase less, sell more, or defer. The marginal dwelling those investors did not buy is instead bought by an owner-occupier. The total dwelling count is unchanged by the reform itself.

Seventy-five thousand owner-occupiers over a decade is 7,500 net per year. As a sense of scale, the Australian Bureau of Statistics publishes first-home-buyer loan commitments at roughly 110,000 per year over recent quarters.⁵ A 7,500 per year net increase is meaningful at the margin (around 7 per cent of current first-home-buyer activity), but it is small against the 11 million Australian households the nation contains.

Treasury also models the price effect as approximately a 2 per cent fall over a couple of years, and rents rising by less than 2 dollars a week. A 2 per cent fall against a Sydney median dwelling value of around 1.15 million dollars (CoreLogic, recent months) is about 23,000 dollars. The marginal buyer who flips into ownership is the household for whom 23,000 dollars closed the gap. Households structurally locked out by a much larger deposit gap remain locked out. The reform is a fairness lever, not an affordability lever.

4. The tax-mix rebalance is back-loaded

The Government describes the new tax package as broadly revenue-neutral over the forward estimates. The neutrality holds because the action is concentrated in the last two years.

Budget Paper 2 reports the receipts impact of the negative gearing and capital gains tax reform as zero through 2027-28, 1.35 billion dollars in 2028-29, and 2.28 billion dollars in 2029-30.⁶ The 30 per cent minimum tax on discretionary trusts records the entire 4.5 billion dollar receipts impact in a single year, 4.47 billion dollars in 2029-30.⁷ The Working Australians Tax Offset shows zero receipts cost through 2027-28, then 3.02 billion dollars in 2028-29 and 3.36 billion dollars in 2029-30.⁸

The package's structural movements all materialise in 2028-29 and 2029-30. The first three years of the forwards see almost no fiscal action from the centrepiece reforms. The next federal election is due by mid-2028.

5. The fuel package's face value is not its deficit impact

The Strengthening Australia's Fuel Resilience Package totals 14.8 billion dollars by the Government's own count.⁹ That number is the sum of several fiscal instruments with different flow-throughs to the underlying cash balance.

The temporary fuel excise reduction is reported in Budget Paper 2 as a 3.8 billion dollar decrease in receipts and a 1.3 billion dollar decrease in payments over the five years from 2025-26, for a net underlying cash balance impact of approximately 2.5 billion dollars.¹⁰

The 7.5 billion dollar Fuel and Fertiliser Security Facility sits with Export Finance Australia and provides loans, equity, and guarantees in support of fuel and fertiliser supply.¹¹ EFA facilities are underwriting instruments rather than grants. The underlying cash impact registers as the concessionality cost of any below-market lending, not as the gross facility size. The 3.2 billion dollar Australian Fuel Security Reserve purchases physical diesel and jet fuel as a government-held asset, against the Minimum Stockholding Obligation increase to 50 days of supply. The 1 billion dollar National Reconstruction Fund Economic Resilience Program provides interest-free loans to manufacturing and logistics businesses; the underlying cash impact is the concessionality cost on those loans. The 1.1 billion dollar Cleaner Fuels Program is released against existing National Reconstruction Fund resources.¹²

Both numbers are correct. They measure different things. The 14.8 billion is balance-sheet capacity. The deficit hit, dominated by the 2.5 billion dollar excise reduction and the production-support program, is materially smaller. A precise underlying cash balance figure for the package as a whole is not published by Treasury and cannot be reconstructed exactly from the public papers, because the concessionality cost of the EFA facility and NRF loan program are not disclosed as separate line items in the available documents.

6. The fuel excise cut is universal in cents-per-litre and uneven in dollars

The 32 cent per litre excise reduction (52.6 cents to 20.6 cents) for three months from 1 April 2026 applies equally per litre to every motorist. The dollar benefit per household depends on litres purchased.

Australian Bureau of Statistics Household Expenditure data shows that fuel spending varies substantially by household income. Lower-income households spend less in absolute dollar terms, partly because car ownership rates are lower (around 8 per cent of Australian households reported owning no motor vehicle at the 2021 Census). Higher-income households drive more, in larger vehicles, over longer distances.¹³

The Treasurer's "Anjali" case-study saves around 170 dollars over the three months by filling a 40-litre tank weekly. A welfare-targeted alternative, such as a 200 dollar payment to Commonwealth Rent Assistance recipients or JobSeeker recipients, would have delivered the same dollar value to lower-income households at a fraction of the fiscal cost, and would have reached households without cars. The chosen instrument is universal because the political purpose was visible price relief at the pump, not welfare-targeted cost-of- living support. The distributional consequence follows from the instrument choice.

7. The NDIS savings cannot come mostly from fraud

Tonight's largest single fiscal line is 36.2 billion dollars in NDIS savings over five years (37.8 billion over four years in the Government's text).¹⁴ Averaged, that is around 7.2 billion dollars per year in slower Scheme growth.

The Government names four reform pillars: commissioning plan management and support coordination, standardised functional assessments for entry, tightening of plan reassessments with category budget resets, and fraud and provider oversight. The press kit weights the fraud line in the speech. The dollar weight cannot sit there.

The arithmetic is straightforward. The Scheme spends approximately 50 to 54 billion dollars per year in recent years, by reference to the National Disability Insurance Agency's Annual Financial Sustainability Reports.¹⁵ Publicly reported suspected fraud, identified through the NDIS Fraud Fusion Taskforce and related compliance activity, runs in the hundreds of millions of dollars annually. Even at a generous upper estimate of one billion dollars per year in eliminable fraud, fraud elimination accounts for roughly one seventh of the projected savings. The remainder, of the order of six billion dollars per year, must come from a combination of reduced participant numbers, lower average plan sizes, or migration of participants to alternative programs such as the new Thriving Kids initiative.

The Government does not publish a pillar-by-pillar decomposition of the savings. The arithmetic above does not allocate dollars to specific reforms. It does establish that the dominant share of the saving must arise from changes to eligibility, plan size, or service mix, rather than from fraud elimination alone.

Whether the projected saving materialises depends on whether the access and plan reforms hold politically. A previous attempt at Scheme tightening in 2023 produced participant outrage and a partial retreat from the original design. This package, which involves standardised functional assessments and reset category budgets for social, civic and community participation, is structurally larger. Treasury has booked the 36.2 billion dollar improvement. It is a forecast, not an outcome.

8. Iron ore at sixty US dollars is fiscal headroom

Treasury's commodity assumption (iron ore declines to sixty US dollars a tonne over four quarters from current levels) is the single most consequential assumption in the document for the revenue path.

Spot iron ore prices in early 2026 have run well above sixty US dollars a tonne. Treasury's Statement 8 (Forecasting Performance and Sensitivity Analysis) historically reports that company tax revenue is materially sensitive to sustained iron ore price differences from the assumed path. The direction is unambiguous: if prices remain elevated relative to the assumption, company tax receipts surprise on the upside. If they fall faster than assumed, the deficit projections worsen.

Standard Treasury practice is to assume commodity-price reversion over a few quarters. The practice is defensible and long-running. The consequence in this budget is that the deficit path published tonight is more likely to come in better than projected than worse, if global iron ore demand holds up. Watch the Mid-Year Economic and Fiscal Outlook in December 2026. If the commodity assumption proves conservative, the revenue upgrade tends to be banked toward the deficit. This is how Treasury's "improvement against MYEFO" lines are produced.

9. The lever they did not pull

Help to Buy is the Commonwealth shared-equity scheme for first-home buyers, legislated in 2024 with 40,000 places. It addresses the deposit gap directly by taking a Commonwealth equity stake in the purchased dwelling. The deposit gap is the most commonly cited binding constraint for would-be first-home buyers in Australian housing research.

Tonight's budget does not expand Help to Buy. The 5 per cent Deposit Scheme, which the Government credits with helping more than 250,000 Australians into ownership since May 2022, is unchanged in this document. The First Home Super Saver scheme is unchanged. The First Home Owner Grant remains a state responsibility.

Every existing direct first-home-buyer subsidy has been left at its 2024 settings. The chosen lever instead is a tax-system reform that raises revenue rather than spending it, and that helps 75,000 marginal households into ownership at the margin over the next decade. The choice between a direct subsidy that costs money and a structural change that raises money is not a technical accident. It is the constraint the budget was optimising for.

The pattern that emerges

Read across the eight findings above, a consistent pattern is visible. Where the Government had a choice between (a) a direct spending program that costs money and helps clearly identifiable people and (b) a structural change that raises or saves money and helps marginal people at the edge, this budget reaches for option (b). Negative gearing reform rather than Help to Buy expansion. NDIS eligibility tightening rather than new disability support spending. Universal fuel excise cut rather than welfare-targeted payments. Five layered tax cuts that return less than bracket creep adds, rather than a single structural threshold increase.

The pattern is consistent with a fiscal strategy that prioritises the return-to-balance forecast in 2034-35 over visible, immediate cost. The budget is not deceptive. It is selective in framing and consistent in instrument choice. Read against itself and against the available external data, what it does is more clearly visible than what it says it does.

What this finding does not establish

The reading above stops short of several claims that would require evidence not yet available.

It does not establish how the NDIS savings will be distributed between participant categories. The Government has not published a pillar-by-pillar dollar breakdown. The arithmetic in section 7 establishes that most savings cannot come from fraud. It does not establish whether they come from access tightening at entry, from smaller plans for existing participants, from category budget resets, or from migration of children to the Thriving Kids program.

It does not put a precise dollar figure on the underlying-cash impact of the full 14.8 billion dollar fuel package, because the concessionality cost of the EFA Fuel and Fertiliser Security Facility and the NRF Economic Resilience Program are not disclosed as line items in the public budget papers. The 2.5 billion dollar excise impact and the 1.1 billion dollar Cleaner Fuels program are explicit. The rest is balance-sheet capacity whose cash-flow registration depends on instrument-by-instrument concessionality assumptions.

It does not put a precise number on the fiscal headroom in the iron ore assumption, beyond noting the direction and Treasury's long-running practice. A precise figure would require running the Statement 8 sensitivity analysis against a forward path of spot prices, which depends on assumptions about Chinese steel demand, substitute supply, and the timing of any retreat from current levels. The number is real. The magnitude is an empirical question.

It does not establish that the budget's chosen instruments are worse policy than the alternatives. A welfare-targeted fuel rebate instead of a universal excise cut would have reached lower-income households more efficiently in dollar terms, but would have been less visible at the pump and may have produced different political outcomes. A direct Help to Buy expansion instead of negative gearing reform would have helped first-home buyers more directly, but would not have raised revenue or reformed a tax setting most economists consider distortionary. Trade-offs exist in both directions. The finding establishes what the choices were. It does not claim they were wrong.

This finding will be revised as the Portfolio Budget Statements, the Parliamentary Budget Office independent costings, and the first round of expert and media interrogation produce additional detail. Where revisions are made, the previous version of the finding will be retained and linked.

Sources

1. Commonwealth Treasury, Budget Overview 2026-27, Appendix A (Budget aggregates) and Appendix E (Detailed economic forecasts), 12 May 2026.
2. Commonwealth Treasury, Budget Paper No. 1: Budget Strategy and Outlook 2026-27, Statement 4, chart data table 4 ("Major categories of (accrual) revenue as a proportion of gross domestic product"), released as bp1-bs4.xlsx alongside the budget papers.
3. Commonwealth Treasury, Budget Overview 2026-27, "Tax reform for workers, businesses and future generations", page 20; New tax cuts for Australian workers factsheet, May 2026.
4. Commonwealth Treasury, Budget Overview 2026-27, "More homes and a fair go for first home buyers", page 24.
5. Australian Bureau of Statistics, Lending Indicators, recent quarters. First-home-buyer loan commitments by number, all dwellings. Specific quarterly figures vary; the approximately 110,000 per year figure in section 3 is a recent-period annual rate.
6. Commonwealth Treasury, Budget Paper No. 2: Budget Measures 2026-27, measure "Tax Reform – Boosting Home Ownership – reforming negative gearing and capital gains tax", page 21.
7. Commonwealth Treasury, Budget Paper No. 2, measure "Tax Reform – introducing a minimum tax on discretionary trusts", page 22.
8. Commonwealth Treasury, Budget Paper No. 2, measure "Tax Reform – cutting taxes with a Working Australians Tax Offset", page 16.
9. Commonwealth Treasury, Budget Overview 2026-27, "Responding to the global oil shock" (pages 12 to 16).
10. Commonwealth Treasury, Budget Paper No. 2, measure "Taking Pressure Off Australians – temporary reduction of fuel excise and heavy vehicle road user charge", page 16. The measure is estimated to decrease receipts by $3.8 billion and decrease payments by $1.3 billion over the five years from 2025-26, for a net underlying cash balance impact of approximately $2.5 billion.
11. Commonwealth Treasury, Budget Paper No. 2, Export Finance Australia fuel resilience facility measure. The Facility provides "up to USD5.0 billion (approximately AUD7.5 billion) in financial support including loans, equity, [and] guarantees".
12. Commonwealth Treasury, Budget Overview 2026-27, "Building resilience" (page 16). Reference to the National Reconstruction Fund's $5 billion Net Zero Fund and the released $1.1 billion Cleaner Fuels Program.
13. Australian Bureau of Statistics, Census of Population and Housing 2021, motor vehicle availability by household. Approximately 8 per cent of Australian households reported no motor vehicle. Household Expenditure Survey results inform the distributional point on fuel spending; the most recent fully released national HES microdata is the 2022-23 release.
14. Commonwealth Treasury, Budget Overview 2026-27, Appendix D (Major Budget improvements), line "Securing the National Disability Insurance Scheme for Future Generations". Text reports $37.8 billion over four years; the appendix table totals $36.2 billion over the five years from 2025-26.
15. National Disability Insurance Agency, Annual Financial Sustainability Report, recent editions, and the NDIA Quarterly Report to Disability Reform Ministers. Annual Scheme spending in the 50-to-54 billion dollar range in recent years; publicly reported suspected fraud through the NDIS Fraud Fusion Taskforce in the hundreds of millions of dollars annually.

Footnote (a): Nominal GDP for 2029-30 derived as receipts (894.8 billion dollars) divided by receipts share of GDP (25.5 per cent), equals approximately 3,509 billion dollars. Half a percentage point of this is approximately 17.5 billion dollars.

Methodology

Every Treasury figure in this finding is sourced to either the Budget Overview, Budget Paper 1, Budget Paper 2, the chart-data tables, or one of the factsheets released by Treasury on 12 May 2026 and available at budget.gov.au under Creative Commons Attribution 4.0 International.

External Australian data references (first-home-buyer activity, household fuel spending and vehicle ownership, NDIS scheme size and fraud reporting, Sydney median dwelling values) are named with their source. Precise current values may differ from the period-typical values cited here; the analytical points hold under reasonable revision of the external numbers, but specific figures in those sentences should be checked against the named source for anyone who wishes to rely on them precisely.

Where this finding's arithmetic is derivative (the nominal-GDP calculation in footnote (a), the implied bracket-creep figure in section 1, the share-of-savings analysis in section 7), the working is shown. Where the reading is interpretive, the relevant section names it as such.

First published 12 May 2026, within hours of the budget speech. Subject to revision as further detail (Portfolio Budget Statements, Parliamentary Budget Office costings, and the December 2026 Mid-Year Economic and Fiscal Outlook) becomes available. Revisions will be logged on this page.