The structural deficit behind the temporary surpluses

In the Pilbara, the difference between a good year and a bad one is written on a screen in Shanghai. When the benchmark iron ore price is high, the road trains run around the clock, machinery contractors add a second shift, and the mess halls in Port Hedland and Karratha fill with fly-in fly-out workers earning six-figure wages to drive trucks. When the price falls, the contracts thin out fast. A haulage subcontractor who scales up his fleet in the good years knows the boom is not really his to keep — it belongs, in the end, to whatever a steel mill in China is willing to pay for ore that quarter.

Canberra's finances ran on the same clock between 2022 and 2024. Twice, the federal government announced something no treasurer had delivered in almost two decades: a Commonwealth budget surplus. Commentary reached for words like disciplined and responsible. But like the haulage contractor's ledger, the turnaround was written mostly in commodity prices and a labour market running hotter than anyone had forecast — not in a permanent change to what Canberra takes in versus what it spends. The bill for that becomes visible the moment the resources cycle turns, and by the government's own budget papers, it already has.

What the government did: two surpluses after fifteen years of deficits

The number Australian budget coverage usually leads with is the underlying cash balance — the difference between what the Commonwealth collects and what it pays out in a given year, adjusted for a few technical items but otherwise the plain measure of whether the government is running ahead or behind. When Anthony Albanese's government handed down its first budget in October 2022, five months after the election, that number was deeply negative: the pandemic-era deficit had peaked at around $134 billion (approx.) in 2020-21, the largest in Commonwealth history (Australian Government, Budget Paper No. 1 and Final Budget Outcome, various years, approx.). But the same global reopening that had driven the pandemic deficit was also driving an extraordinary run in iron ore and coal prices and an unemployment rate falling toward the lowest sustained level in fifty years. Tax receipts were coming in far ahead of what had been pencilled in only months earlier, at the pre-election fiscal outlook (PEFO) Treasury and Finance published under caretaker conventions in April 2022.

The government's stated fiscal strategy, set out in Budget Paper No. 1's fiscal strategy statement across its first term, was to bank the bulk of that revenue upgrade — to let the extra receipts improve the bottom line, rather than fund an equivalent amount of new permanent spending. In the October 2022 Budget, Treasurer Jim Chalmers said the government had banked almost all of the improvement in receipts since PEFO, a figure reported at the time at around 99 per cent (approx.) of the upgrade (Budget Speech, 25 October 2022; Budget Paper No. 1, Statement 3, Fiscal Strategy and Outlook, October 2022-23). The May 2023 Budget continued the general approach, though by the government's own account a smaller share of the further improvement was banked than in the first budget, with more of it directed to new spending, including cost-of-living relief such as energy bill rebates and cheaper medicines under the PBS.

The result showed up in the outcome documents. The 2022-23 Final Budget Outcome, released by the Treasurer and Finance Minister in September 2023, recorded an underlying cash surplus of $22.1 billion (approx.) — around 0.9 per cent of GDP (approx.) — the first Commonwealth surplus since 2007-08, ending a run of fifteen consecutive annual deficits. A year later, the 2023-24 Final Budget Outcome, released in September 2024, recorded a second consecutive surplus of $15.8 billion (approx.), or around 0.6 per cent of GDP (approx.) — the first back-to-back surpluses since 2006-07 and 2007-08, under the Howard government. Net debt came in lower than the pre-election forecasts had projected, and gross debt issuance was lower than planned.

That run ended with the 2024-25 Budget, handed down in May 2024, which forecast a return to deficit of around $28.3 billion (approx.), or about 1.0 per cent of GDP (approx.), for 2024-25. The 2025-26 Budget, brought down in March 2025 ahead of that year's federal election, projected the deficit continuing across the four-year forward estimates, and the ten-year medium-term projections published alongside it showed no surplus year in the published trajectory (Budget Paper No. 1, Statement 3 and medium-term projections, March 2025, approx.). The Mid-Year Economic and Fiscal Outlook updates released in December 2023 and December 2024 confirmed the same shape at each check-in: two good years, bracketed by deficits before and forecast deficits after.

One genuine policy decision sits inside this period and deserves to be named on its own terms. In January 2024 the government redesigned the already-legislated Stage 3 income tax cuts — a decision examined in full in a separate audit on this site — redirecting more of the benefit toward low- and middle-income earners ahead of the change taking effect on 1 July 2024. Treasury's own costings described the redesign as broadly cost-neutral against the original legislated package over the forward estimates (Budget Paper No. 1, Statement 3, May-2024, approx.), meaning it changed who received the tax cut far more than it changed the overall size of the structural gap. It is worth flagging here because it is one of the few genuine policy choices, rather than a parameter movement, made in this whole period — and, on the numbers, it is not the reason the deficits returned.

The evidence: how much of the turnaround was policy, and how much was luck

Budget documents distinguish between two very different reasons a bottom line improves. A parameter variation is a change nobody in government decided — a higher iron ore price, a lower unemployment rate, a shift in the exchange rate — that mechanically lifts or lowers receipts and payments. A policy decision is a choice the government of the day makes deliberately — a new spending program, a tax change, an eligibility rule. Budget Paper No. 1's reconciliation tables split every movement in the bottom line between the two categories. Reading them across 2022 to 2025 answers the central question about the two surpluses: were they built, or were they lucky? What follows works through the receipts upgrade the government chose to bank, the commodity prices and jobs market that produced it, and the independent structural-balance analysis that shows what remains once both are stripped away.

Banking the upgrade, budget by budget

Treasury's own account of the October 2022 Budget is the cleanest data point. Of the improvement in expected tax receipts since the pre-election fiscal outlook, the government said it had banked around 99 per cent (approx.) — meaning almost none of the extra money flowing in from higher commodity prices and stronger employment was matched by new permanent spending (Budget Speech, 25 October 2022; Budget Paper No. 1, Statement 3, October 2022-23). That was a genuinely disciplined first budget. Measured over the four-year forward estimates each budget carried, the cumulative tax receipts upgrade recorded across the government's first two budgets ran into the hundreds of billions of dollars (approx.) — among the largest such revisions in at least a decade of budget updates (Budget Paper No. 1, Statement 3, October 2022-23 and May 2023-24, approx.).

The pattern did not hold at the same intensity. By the government's own telling and independent commentary from bodies including the Parliamentary Budget Office and the Grattan Institute, the share of each successive upgrade that was banked, rather than spent, fell across the May 2023, May 2024 and March 2025 budgets, as new spending commitments — from public sector wage outcomes to industry and energy programs to the redesigned Stage 3 tax cuts — absorbed a growing share of the improving revenue base (Budget Paper No. 1, Statement 3, various years; independent budget commentary, 2023–2025, approx.). None of this means the later budgets were reckless; underlying cash balances still improved relative to the pre-election baseline. But the arithmetic that produced two surpluses depended on a large first-budget banking rate that was not repeated at the same scale, at exactly the point the underlying commodity windfall itself began to fade.

The commodity assumption that kept being wrong

Every Commonwealth budget for the past two decades has used a deliberately conservative convention for forecasting the iron ore price — Australia's single largest export earner. Budget Paper No. 1's revenue statement assumes the price will fall from wherever it sits at budget time to a much lower benchmark within one or two quarters, on the reasoning that commodity prices are volatile and a government should not build its numbers around a price spike lasting forever. Across the budgets between October 2022 and March 2025, that assumed floor itself crept upward — from a benchmark in the order of US$55–60 a tonne free-on-board in the earliest of these budgets to around US$75 a tonne (approx.) by the 2024-25 and 2025-26 editions — an implicit admission that the "conservative" assumption had been too low, two years running (Budget Paper No. 1, Statement 5, Revenue, various years, approx.).

The actual price told the story. The benchmark iron ore spot price traded well above US$100 a tonne for most of 2022 and 2023 (approx.; Reserve Bank of Australia, Index of Commodity Prices), consistently overshooting Treasury's built-in assumption. Treasury's own sensitivity tables illustrate why that matters: a sustained US$10-a-tonne change in the iron ore price is estimated to move Commonwealth tax receipts by billions of dollars a year (Budget Paper No. 1, Statement 5, sensitivity analysis, approx.). A price that stays $40–50 a tonne above the budget assumption for two straight years, as it did, mechanically produces exactly the kind of receipts upgrade the government then had the choice to bank or spend.

Coal added a second, larger and less predictable shock. Thermal and coking coal prices spiked to record levels after Russia's invasion of Ukraine in February 2022 disrupted global energy markets — benchmark thermal coal briefly traded above US$400 a tonne in 2022 (approx.), several multiples of pre-pandemic levels, before easing back through 2023. Almost none of that price movement had anything to do with domestic policy; it flowed from a war reshaping global energy trade, and it fed federal company tax receipts, and state royalties, at a scale no forecast from early 2022 had anticipated.

Near-full employment did more of the work than any single policy

The second parameter behind the surpluses was the labour market. The unemployment rate fell to around 3.5 per cent in mid-2022 (approx.; Australian Bureau of Statistics, Labour Force, 2022 releases) — near the lowest sustained rate in fifty years — even as the Reserve Bank lifted the cash rate from a record low of 0.10 per cent in April 2022 to 4.35 per cent by November 2023 (Reserve Bank of Australia, cash rate target series), the fastest tightening cycle in a generation. More people in work and fewer people drawing unemployment benefits does two things to a budget at once: it lifts income tax receipts, and it lowers welfare payments. Both moved in the government's favour through 2022 and 2023.

That tightness in the labour market was itself mostly a global and cyclical phenomenon — a post-pandemic reopening dynamic playing out across most advanced economies, not the product of a specific new industrial-relations or employment policy enacted after May 2022. Some of the government's own workplace-relations changes may have affected wages and job security at the margin, but the headline unemployment rate that did the heavy lifting for the budget bottom line was already near historic lows before most of those changes had taken effect.

What the Parliamentary Budget Office found underneath the headline

The Parliamentary Budget Office (PBO) — an independent statutory office that costs policy and audits the budget for Parliament — publishes a separate measure called the structural budget balance. The idea is straightforward: strip out the temporary swings caused by the economic cycle and one-off price movements, and ask what the budget balance would be if commodity prices and employment were sitting at their long-run average levels instead of wherever they happen to be this year. What is left is the structural gap — the underlying difference between what government has committed to spend, on an ongoing basis, and what the tax system raises on an ongoing basis, independent of the cycle.

The PBO's structural budget balance analysis, published across 2023 and 2024, found that Australia's structural position remained in deficit throughout the period the headline cash balance was in surplus — commentary drawing on the PBO's work has put the gap at somewhere in the order of one to two per cent of GDP (approx.) even in 2022-23 and 2023-24 (Parliamentary Budget Office, Structural Budget Balance report and related analysis, 2023–2024, approx.). In plain terms: take away the iron ore and coal price windfall and the unusually tight labour market, and the underlying budget was still short of balance in the very years Canberra was announcing surpluses. The headline number and the structural number were telling two different stories at the same time, and only one of them was cyclical.

Company tax told as much of the story as personal tax

The composition of the receipts upgrade matters as much as its size. Company tax — paid on the profits of Australian businesses, and heavily weighted toward the resources sector when commodity prices spike — was among the largest single contributors to the upgrades recorded in the October 2022 and May 2023 budgets, reflecting the extraordinary profitability of iron ore, coal and gas exporters at prices well above their cost of production (Budget Paper No. 1, Statement 5, Revenue, October 2022-23 and 2023-24, approx.). Personal income tax also rose, but mostly for reasons connected to the same cycle: more people in work, and existing workers pushed into higher tax brackets by inflation without those brackets being adjusted — a phenomenon known as bracket creep, examined in a separate audit on this site. Both of the largest contributors to the receipts upgrade, in other words, were parameter effects tied to the commodity and employment cycle, not new taxes the government legislated.

How the surplus compared with the rest of the world

Australia's position also has to be read against what resource-exporting economies typically do in a commodity upswing, and what nearly every other advanced economy was doing at the same time. Most OECD governments continued running budget deficits through 2022 and 2023, still working through pandemic-era spending and a slower post-pandemic revenue recovery than a commodity exporter like Australia enjoyed (OECD and IMF Fiscal Monitor comparisons, various, approx.). Australia's surplus was therefore unusual by international standards — but resource-exporting economies have a long history of headline fiscal positions swinging sharply with commodity prices, from sovereign-wealth-linked oil exporters to the mining-royalty booms that lifted Australian state and federal budgets in the mid-2000s and again in the early 2010s. What makes a fiscal position structural, rather than a favourable point in that well-worn cycle, is whether the improvement survives a fall in the commodity price back to its long-run average. On that test — the one the PBO's structural balance measure is built to apply — the 2022-23 and 2023-24 surpluses do not qualify, however creditable the spending restraint that helped produce them.

The surpluses were real. So was the fact that almost none of what produced them was a decision anyone in Canberra made.

The critique: the deficits are back, and the structural gap is still there

The clearest evidence for the structural argument is what happened next. The 2024-25 Budget, handed down in May 2024, forecast a return to deficit of around $28.3 billion (approx.), about 1.0 per cent of GDP (approx.), ending the run at two consecutive surpluses. The March 2025 Budget — brought forward from its usual May slot ahead of that year's federal election — projected deficits continuing across the full four-year forward estimates, and the medium-term projections published alongside it (covering the decade beyond the forward estimates, on the same broad methodology the Intergenerational Report uses) showed no return to surplus in the published trajectory (Budget Paper No. 1, Statement 3 and medium-term projections, March 2025, approx.). The December 2023 and December 2024 Mid-Year Economic and Fiscal Outlook updates told the same story at each check-in: deficits ahead, not surpluses.

This is not, on the government's own numbers, a surprise reversal. Every one of the budget papers that announced a surplus also carried, in its risk statement and sensitivity tables, an explicit warning that the commodity prices doing most of the work were assumed to fall, and that a large share of the improvement was cyclical rather than structural. Australia's terms of trade — the ratio of export prices to import prices, a broad gauge of how much the country earns per unit of what it ships overseas — reached a record high around 2022 and have been easing since (Australian Bureau of Statistics, Balance of Payments and National Accounts, various releases, approx.). That easing is precisely the mechanism the structural deficit was always going to reassert itself through once it ran its course, and by 2024-25 it had.

Underneath the cyclical story sits a base of spending that keeps growing faster than the economy regardless of the commodity cycle. The National Disability Insurance Scheme, aged care, health, defence and the interest bill on accumulated debt are among the fastest-growing components of Commonwealth payments, and none of them shrinks when iron ore prices fall. When the parameter windfall that funded two years of restraint-plus-surplus faded, those structural pressures did not fade with it. The interest bill in particular is now itself one of the budget's largest and fastest-growing line items, a direct and mechanical cost of the debt built up through the pandemic and not repaid by two years of temporary surplus (Budget Paper No. 1, Statement 3 and payments tables, various years, approx.).

The scale of the swing is itself instructive. Between the 2022-23 surplus and the 2024-25 forecast deficit, the underlying cash balance moved by an amount in the order of $50 billion (approx.) — a shift of roughly two percentage points of GDP (approx.) in two years, almost entirely explained by the fading of the same commodity and employment parameters that had produced the surplus, rather than by any comparably sized new spending or tax decision (Budget Paper No. 1 and Final Budget Outcome documents, 2023–2025, approx.). A budget position that can swing by that much without a single major new policy decision driving it is, by definition, one still dominated by the cycle rather than by structural settings government has locked in.

State governments captured a parallel windfall through mining royalties, which sit outside the Commonwealth's own budget but interact with it through the GST distribution system administered by the Commonwealth Grants Commission. Western Australia in particular recorded its own run of large state surpluses over the same period, driven by the same iron ore prices (Western Australian state budget papers, various years, approx.). The point is not that state royalties inflate the Commonwealth's own numbers directly — mostly they do not — but that the same commodity cycle was doing similar work across every level of government at once, further evidence that the story is a national terms-of-trade event rather than a Commonwealth-specific policy success.

The government was entitled to celebrate the achievement, and it did — the back-to-back surpluses were marketed as proof that Labor could be trusted with the budget after years of Coalition attacks on its economic-management credentials. But the same budget papers that recorded the surplus also carried the parameter-versus-policy tables showing how little of it reflected a change in the underlying settings. A reader who took the political messaging at face value, without checking the primary documents, could reasonably have concluded the structural position had improved. The Parliamentary Budget Office's structural-balance work exists precisely to correct that kind of reasonable but mistaken inference, and on the record, it did.

Treasury's own long-run modelling reinforces the same conclusion from a different angle. The 2023 Intergenerational Report — Treasury's periodic forty-year projection of the budget's trajectory, produced roughly every five years — projected that an ageing population and rising health, aged care and NDIS costs would keep structural pressure on Commonwealth spending for decades, regardless of where commodity prices or the unemployment rate happened to sit in any given year (Australian Government, Intergenerational Report, Treasury, 2023). That report was published in the same window as the two surplus years, and its central message was not that the budget's underlying position had been repaired; it was that the underlying pressures were still building. Reading the Final Budget Outcome and the Intergenerational Report side by side captures the whole picture in miniature: a genuinely good two years on the headline number, sitting on top of a forty-year structural trend running the other way.

None of this means the government caused the commodity cycle, and it would be unfair to suggest otherwise — nobody in Canberra sets the Shanghai iron ore price or the war in Ukraine's effect on coal markets. What the government is responsible for is what it chose to do with the windfall, and how it described what had happened. Banking most of the first upgrade was a genuine policy choice, and a good one. Allowing the political narrative to describe a temporary, price-driven alignment as evidence of a durably repaired budget was a choice too — and the return to forecast deficits within a year of the second surplus is the primary record correcting that impression.

A structural deficit does not become a structural surplus because the tide comes in for two years.

What the government got right

Credit is due in three specific places. First, the restraint itself: in the very first budget after taking office, with a large unexpected revenue upgrade sitting in front of it, the government chose to bank the overwhelming majority of it rather than fund an equivalent amount of new permanent spending — the harder and less popular of the two available paths. Second, the surpluses were not achieved by cutting services in a downturn; they coincided with unemployment near fifty-year lows and with targeted cost-of-living relief still being delivered, which shows fiscal repair and short-term household support were not, in this period, mutually exclusive. Third, lower net debt than the pre-election baseline projected means a lower ongoing interest bill than the counterfactual, a real and durable benefit even after the deficits returned.

International credit rating agencies took note. Australia retained its top-tier AAA sovereign credit rating throughout this period, with S&P Global Ratings and Moody's both citing the improved fiscal trajectory in their rating actions (S&P Global Ratings and Moody's, sovereign rating commentary, 2022–2024, approx.) — one of a small number of major economies to hold that rating through the pandemic and its aftermath. A credit rating is, in effect, a market-tested judgement on a government's capacity to service its debt, and by that measure the two surpluses were not purely cosmetic; independent credit assessors read them as a genuine, if partial, improvement in the government's fiscal position.

It is also fair to note that the structural deficit is not something this government invented. Independent analysis, including from the Parliamentary Budget Office and the Grattan Institute, has argued that Australia's Commonwealth budget has carried a structural gap — spending commitments growing faster than the revenue base, for demographic and program-design reasons that predate any one government — for the better part of two decades, through Coalition and Labor budgets alike (Parliamentary Budget Office and Grattan Institute budget analysis, various years, approx.). The NDIS was legislated in 2013 under a different government; an ageing population and a deteriorating strategic environment driving up defence spending are structural facts, not partisan choices. What this government did was inherit a pre-existing structural gap, use a temporary and largely unearned commodity-and-jobs windfall to paper over it for two budget cycles, and then, as the windfall faded, arrive back close to the position it started in. That is a fair distinction from claiming Labor created the structural deficit from nothing — but the fairness of the distinction cuts both ways. It means the surpluses were less a Labor achievement than the political framing around them suggested, and it also means the return to deficit is less a Labor failure than a straightforward description of an inherited structural problem the windfall never actually solved.

It is worth adding that the Stage 3 redesign, whatever one concludes about the broken 2022 election promise it involved, was on Treasury's own costings not itself a driver of the return to deficit; the change was designed to be close to cost-neutral relative to the original legislated cuts. The 2024-25 deficit's return is better explained by the fading of the commodity and employment parameters described above than by any single new spending or tax decision taken in this period.