On a ridge line outside Wagga Wagga, a construction crew spends its days hauling steel lattice towers into position and stringing high-voltage cable between them. It is one of dozens of transmission projects now under way or in planning around the country as part of Rewiring the Nation, the federal government's program to rebuild the electricity grid so it can carry power from new wind and solar farms to the cities that will use it. The steel, the concrete and the wages paid to that crew are real government-backed spending, committed by a Commonwealth decision and ultimately funded by Commonwealth borrowing.
None of it, though, will show up in the number the Treasurer reads out on budget night. When Australians hear that the underlying cash balance — the headline measure of whether the government is taking in more than it spends — has improved or worsened, tens of billions of dollars committed to transmission lines, social housing and manufacturing finance sit almost entirely outside that figure. In the language of the budget papers, this is off-budget spending: money the government commits, and typically borrows, structured as an investment in a fund or a corporation rather than a straight payment, so that it never crosses the ledger line used to define "the deficit" or "the surplus." The money is real. The debt behind it is real. The number most people are shown is not the whole story.
What the government did: four vehicles, one accounting structure
Since the 2022 election, the Albanese government has built or expanded four major investment vehicles that sit, wholly or substantially, off the Commonwealth's core budget ledger. Each was legislated separately, with its own Act of Parliament and its own governing board, and each channels money into a different part of the economy. Taken together, they represent the largest expansion of off-budget financing in at least two decades.
The government has generally described the four vehicles as parts of a single, broader strategy rather than four unrelated decisions: rebuilding a grid that can carry renewable power at the scale needed to meet Australia's legislated 2030 and 2050 emissions targets, rebuilding a housing and social-housing pipeline after a decade in which construction had not kept pace with population growth, and rebuilding sovereign manufacturing capability in sectors the government judged were strategically important amid intensifying global competition over critical minerals, clean-energy supply chains and advanced manufacturing — a framing that draws an explicit, and openly acknowledged, comparison to large industrial-policy programs launched around the same time in the United States, the European Union and elsewhere (Australian Government, Future Made in Australia policy statements and Budget Paper No. 2, May 2024-25, approx.). Whatever view one takes of that strategic case, examined on its industry-policy merits elsewhere on this site, the accounting question here is narrower and applies regardless of whether the underlying policy goal is judged a success: does the way this spending is recorded give the public an accurate, comparable picture of the government's overall financial position.
The first, and largest at the time, was Rewiring the Nation, announced in the October 2022 Budget — the government's first — as up to $20 billion (approx.) in concessional loans and equity to fast-track the transmission lines and interconnectors identified in the Australian Energy Market Operator's Integrated System Plan as necessary to connect new renewable generation to the grid. Rather than create a new entity, the government routed the money through the Clean Energy Finance Corporation (CEFC), a "green bank" established under the Clean Energy Finance Corporation Act 2012 by the Gillard Labor government, which already had a decade of experience investing Commonwealth capital in clean-energy projects on commercial or concessional terms.
The second, the National Reconstruction Fund (NRF), was legislated through the National Reconstruction Fund Corporation Act 2023, which passed the Parliament in March 2023 after months of negotiation with the Greens and crossbench senators over its investment mandate. It provides up to $15 billion in loans, equity and guarantees to businesses in priority areas of the economy including renewables and low-emissions technology, medical manufacturing, defence capability, and value-adding in agriculture and resources (National Reconstruction Fund Corporation Act 2023, approx.) — sectors the government judged were underserved by private capital markets on their own.
The third, the Housing Australia Future Fund (HAFF), was legislated through the Housing Australia Future Fund Act 2023, passed by the Senate in September 2023 after the government agreed to a package of concessions to the Greens, including a guaranteed minimum annual spend, to secure the numbers. It places $10 billion under the management of the Future Fund Board of Guardians — the same body that manages the Commonwealth's sovereign wealth fund — with investment returns, capped at $500 million a year (approx.) and guaranteed at roughly that level regardless of the fund's actual investment performance in a given year (Housing Australia Future Fund Act 2023, approx.), directed to social and affordable housing programs. Under the Act's funding mechanism, if the Future Fund Board's returns in a poor investment year fall short of the amount needed to meet the guaranteed distribution, the shortfall can be met by a direct appropriation from consolidated revenue (Housing Australia Future Fund Act 2023, approx.) — meaning part of a program designed to sit off-budget could, in a bad year, still show up as an on-budget cost.
The fourth and most recent, Future Made in Australia (FMIA), was the centrepiece of the May 2024 Budget: a package costed at around $22.7 billion (approx.) over the decade (Budget Paper No. 2, May 2024-25, approx.), built around a "National Interest Framework" with two streams — a Net Zero Transformation stream and an Economic Resilience and Security stream. Unlike the other three, FMIA is a mixed structure. Refundable production tax incentives for hydrogen and processed critical minerals sit on-budget, hitting the underlying cash balance as they are paid out from the mid-to-late 2020s. But a substantial share of the package's capital — top-ups to the National Reconstruction Fund, and new CEFC-administered programs such as the Solar Sunshot and Battery Breakthrough Initiative — was structured as equity and concessional finance through the same off-budget mechanism as Rewiring the Nation and the NRF (Budget Paper No. 2, May 2024-25, approx.).
The strongest case for the policy
Investment vehicles that earn a return are a legitimate, long-standing tool
Put the government's defence at its best, and it rests on a distinction every accountant recognises. When the Commonwealth pays unemployment benefits, buys hospital equipment or funds a school, it is spending money it will never get back; that is a straightforward expense, and it belongs above the line in the budget's bottom line. When it instead buys equity in a business, lends money on commercial or near-commercial terms, or invests capital expecting a financial return, it has swapped one asset — cash — for another: a shareholding, or a loan receivable. The transaction changes the composition of the government's balance sheet; it does not, by itself, make the government poorer the way a welfare payment does. Public-sector accounting standards, and the Government Finance Statistics framework the Australian Bureau of Statistics and Treasury both use, have drawn this distinction for decades, and Australia is not unusual in doing so — most advanced economies separate financing transactions in financial assets from recurrent expenses in their own headline fiscal measures.
Nor is the mechanism new, or invented to flatter this government's numbers. The Commonwealth has run off-budget investment vehicles since well before 2022: NBN Co, established from 2009, built the National Broadband Network with tens of billions of dollars of Commonwealth equity structured exactly this way; the original CEFC, from 2012, has invested billions in clean-energy projects on the same off-budget basis for over a decade; Export Finance Australia and the Australian Business Growth Fund operate on comparable principles. Commonwealth governments of both major parties have used vehicles like these when the policy problem looks less like "pay for a service" and more like "provide patient, long-term capital to a market that private lenders and investors are unwilling to fund at the scale or speed the country needs" — precisely the description that fits transmission infrastructure, early-stage advanced manufacturing and social-housing development, all capital-intensive, long-horizon undertakings that conventional finance has historically underserved.
Crucially, proponents can also point out that none of this is actually hidden from the official record. Every one of the four vehicles was created by its own Act of Parliament, debated and voted on. Every dollar appears in Budget Paper No. 2 when the measure is announced, in the relevant fund's annual report, and in the Australian Office of Financial Management's published data on Australian Government Securities on issue — the bonds the Commonwealth sells to fund itself, including the equity injected into these funds. Most importantly, the borrowing required to fund the equity injections shows up in gross debt, the headline measure of everything the Commonwealth owes, regardless of what that money was used for. A government trying to conceal its borrowing would not choose a mechanism that still shows up, in full, on the debt side of the ledger — it would simply borrow less and spend less. On this argument, "off-budget" is a description of an accounting classification for a well-understood category of transaction, not a description of concealment.
The evidence: what the numbers show
The steel-man is correct about the mechanics; the question is what those mechanics do to what the public actually sees and is able to compare, year to year. Two Budget Paper No. 1 measures matter here, and the difference between them is the whole of the accounting question this topic turns on. The underlying cash balance — receipts minus payments, with a specific and long-standing exclusion for "net cash flows from investments in financial assets for policy purposes" — is the figure reported as "the deficit" or "the surplus" in almost every news bulletin and almost every political speech. The headline cash balance is the underlying cash balance plus that excluded item: it captures the actual cash going out the door to buy equity in the HAFF, the NRF and the CEFC's expanded mandate, alongside older items like NBN Co. Both measures are published, side by side, in the same Budget Paper No. 1 tables every year (Budget Paper No. 1, Statement 3, various editions). Only one of them is ever quoted.
How "off-budget" actually works
The easiest way to see the distinction is to compare two one-dollar decisions. If the Commonwealth pays a dollar in unemployment benefits, that dollar is gone: no asset comes back, and the underlying cash balance falls by a dollar, full stop. If the Commonwealth instead spends a dollar buying a share in a company, or lending a dollar it expects to be repaid with interest, its cash goes down by a dollar but its holdings of financial assets — the share, or the loan owed to it — go up by a dollar at the same time. Because the balance sheet has not shrunk, only changed shape, the budget convention treats that second dollar as an investment in a financial asset for policy purposes and excludes it from the underlying cash balance. That exclusion is not new, and it is not, on its own, evidence of anything untoward — it is broadly the same logic that lets a household treat paying down a mortgage differently from paying a power bill, because one builds an asset back up and the other does not.
The complication in practice is that the "asset" side of this equation is often far less certain than a bank balance. A share in an early-stage manufacturer, a loan to a transmission company building a project years from completion, or a capped, government-guaranteed return from a housing fund are all financial assets in the technical sense, but their eventual value depends on decisions, markets and construction timelines that are not resolved on the day the investment is made. The accounting convention books the investment at cost, or close to it, on day one, while the risk that the asset turns out to be worth less than that — a risk inherent in exactly the kind of underserved, capital-intensive markets these vehicles were designed to fund — is carried silently until, if ever, it crystallises.
The four vehicles, added up
No single line in any Budget paper adds these four vehicles together. Each was announced in a different year, described in a different chapter of a different edition of Budget Paper No. 2, and legislated under a different Act with its own long title and its own second-reading speech. To see the combined scale, a reader has to do the arithmetic themselves: $10 billion for the HAFF, $15 billion for the NRF, up to $20 billion for Rewiring the Nation, and a further package costed at around $22.7 billion (approx.) for Future Made in Australia, only part of which is off-budget. Add the off-budget shares together and the running total, in legislated or announced capital, is well past $60 billion (approx.) since October 2022 — before counting the off-budget stock the government inherited, including NBN Co's tens of billions of dollars of Commonwealth equity and the original, smaller CEFC mandate that predates Rewiring the Nation entirely.
That total is a stock, not a flow: it is the size of the capital envelope legislated or committed, not money all spent in a single year, and much of it will be drawn down, invested and in some cases returned over a decade or more. But it is also new borrowing capacity the Commonwealth has authorised itself to use, on top of whatever the underlying cash balance says about the ordinary business of running the government day to day. The distinction between a stock of committed capital and a flow of annual spending is real, but it is not a distinction the underlying cash balance headline communicates to anyone who is not already reading the fine print.
Scaled to the population, the figure is easier to hold in your head. Roughly $60 billion (approx.) works out to somewhere in the order of $2,200 for every person in Australia (approx.; on an estimated resident population of around 27 million, ABS National, State and Territory Population, approx.), or close to $5,600 for every household (approx.). Nobody receives a bill for that amount, and nobody should: these are financial assets the Commonwealth expects, at least in part, to recover, not spending that disappears the way a Medicare rebate or a JobSeeker payment does. But it is a useful sense check on the order of magnitude of new Commonwealth financial commitment created off-budget in under two years, committed through a mechanism that never touches the headline deficit number at all, and worth holding in mind the next time that number is reported as the whole of the government's financial position.
Debt that shows up, on a page most people never turn to
The steel-man's strongest point is also verifiable: this borrowing does show up in gross debt, the face value of Australian Government Securities on issue, published regularly by the Australian Office of Financial Management. Gross debt does not care what the money was used for; every dollar borrowed to fund an equity injection into the HAFF, the NRF or the CEFC's expanded book adds to it in exactly the same way a dollar borrowed to fund a pension payment would. Anyone checking gross debt figures across the life of this government would see the cumulative effect of these four vehicles reflected in the total, even in years when the underlying cash balance was in surplus (Australian Office of Financial Management, Australian Government Securities data, various, 2022–2025).
Net debt — gross debt minus a narrower set of the government's financial assets, mainly cash and deposits — treats the equity in some of these vehicles differently again, netting some of it off and not other parts, a technical distinction that depends on how liquid or programmatic Treasury and the Australian Bureau of Statistics judge each asset to be. The practical result is that Australia now runs at least four distinct headline debt and balance measures — underlying cash balance, headline cash balance, gross debt and net debt — that can each tell a different story about the same set of decisions in the same year, and only one of the four is what tends to get repeated on the nightly news.
There is one more piece of the mechanics worth spelling out, because it cuts against any suggestion that off-budget spending is somehow free. The capital sum injected into the HAFF, the NRF or the CEFC's expanded book sits outside the underlying cash balance, but the interest the Commonwealth pays on the bonds it issues to fund that capital does not: interest on Australian Government Securities is itself an on-budget expense, recorded in the underlying cash balance as it accrues, year after year, for as long as the debt is outstanding (Budget Paper No. 1, Statement 6, payments tables, various editions, approx.). In other words, off-budget treatment defers the visible fiscal cost of these vehicles; it does not eliminate it. The equity injection disappears from the headline number on day one, but a smaller, recurring version of that same borrowing cost reappears every year afterwards as an ordinary interest expense, indistinguishable in the budget papers from interest paid on debt raised to fund a hospital or a pension.
The deficit falls, or rises less, on the same day the government borrows the money — the transaction is simply booked on a page most people never turn to.
What independent analysts have flagged
The Parliamentary Budget Office (PBO) — the independent statutory office that costs policy and audits the budget for Parliament, distinct from Treasury and the government of the day — has published analysis and commentary noting that the growing use of off-budget financing complicates like-for-like comparisons of the government's fiscal position over time, because a headline deficit or surplus figure from a budget with heavy off-budget activity is not measuring the same thing as the equivalent figure from a period with little or none (Parliamentary Budget Office, budget and fiscal risk analysis, 2023–2025, approx.). Independent commentators, including analysts associated with the Grattan Institute and other budget-focused research bodies, have made a related point in public commentary: that the growth of off-budget spending since 2022 risks flattering the headline number relative to the government's true net financial position, even where every individual transaction is properly disclosed (Grattan Institute and independent budget commentary, various, 2023–2025, approx.).
Neither body has alleged that disclosure itself is inadequate in the narrow, legal sense — the Acts, the budget papers and the funds' own annual reports are all public documents anyone can read. Their concern is about salience: whether the number that gets asked about at press conferences, debated in Question Time and repeated in headlines is still doing the job of summarising the government's overall financial position, when a growing share of new Commonwealth financial commitment sits outside it by design. A convention that was manageable when one long-standing green bank ran a modest annual book becomes a materially different proposition once four large vehicles are added inside two budget cycles.
The critique: a shadow budget with no single ledger
Three problems sit inside an otherwise defensible accounting convention. The first is scale. Off-budget vehicles have existed in Australia for well over a decade, but never before at this combined size, committed this quickly, across this many new entities at once. Moving from one long-standing green bank with a modest annual mandate to four separate vehicles worth a combined figure well past $60 billion (approx.) inside two budget cycles is a change large enough in degree that it starts to function as a change in kind: a genuinely material share of new Commonwealth financial commitment since 2022 now sits, by design, outside the number that headline political debate treats as the whole picture.
The second is governance. Ordinary Commonwealth spending is appropriated by Parliament roughly annually and is subject to Senate estimates hearings, Auditor-General scrutiny of individual programs, and line-by-line reporting in portfolio budget statements. Off-budget vehicles are voted on once, at the point of legislation, and then hand investment decisions to a board — the Future Fund Board of Guardians for the HAFF, the NRF's own investment committee, and the CEFC board for Rewiring the Nation and the FMIA top-ups — operating under a broad legislated mandate with considerably more discretion over individual deals than a department has over individual line items in an appropriation bill. That is by design, and it mirrors how the sovereign Future Fund itself has long operated. All four bodies are established as corporate Commonwealth entities under the Public Governance, Performance and Accountability Act 2013, which gives them their own board, their own annual report and their own set of statutory duties — a genuine governance structure, but a different one to a government department, whose spending decisions are reviewed line by line every year rather than assessed periodically against a broad legislated mandate. The result is that the day-to-day allocation of tens of billions of dollars of Commonwealth capital is now further removed from the kind of item-by-item parliamentary scrutiny that applies to a hospital grant or a welfare payment, at exactly the moment the total sums involved have grown fastest.
The third is risk. An equity stake or a concessional loan is not a guaranteed asset of equivalent value to the cash it replaced. Transmission projects can run over budget or behind schedule; manufacturing ventures in nascent industries like green hydrogen and critical-minerals processing carry genuine commercial risk, which is precisely why private capital has been reluctant to fund them at scale on its own; social-housing developments can face construction-sector capacity constraints and cost pressures of the kind examined elsewhere on this site. If a meaningful share of the loans or equity in these vehicles needs to be written down over the coming decade — a normal, expected outcome for a portfolio of genuinely risky, market-gap investments, not a scandal in itself — that loss will eventually show up as an expense, most likely well after the political cycle in which the original investment was announced and celebrated. The gap between "committed" and "returned in full" is where the fiscal risk actually lives, and it is a gap the underlying cash balance, by construction, does not price in at the point the money goes out the door.
Put the three together and the practical effect is a shadow budget: a growing pool of Commonwealth financial commitment that is legally and technically disclosed, but that a reader cannot see, size or compare over time without independently reconciling four separate Acts, several editions of Budget Paper No. 2, a handful of fund-specific annual reports, and the underlying-versus-headline cash balance reconciliation tables buried inside Budget Paper No. 1. That is a considerably higher bar than reading a single number off a press release, and it means the ordinary version of the question most people think they are asking — is the government's financial position improving or worsening this year — now gets a systematically incomplete answer from the figure built to answer it.
Nothing here is hidden. It is simply scattered across four Acts, four Budget papers and a reconciliation table almost nobody reads — which, for most practical purposes, has much the same effect.
What the government got right
Credit belongs in a few specific places. First, on transparency in the narrow sense: none of the four vehicles was created by regulation or administrative decision alone; each went through its own Bill, its own parliamentary debate, and in the case of the NRF and the HAFF, months of public negotiation with the Greens and crossbench that generated extensive public commentary on exactly what each fund would and would not do. A government trying to genuinely hide tens of billions of dollars of commitment would not choose a path that puts each vehicle through a separate, high-profile legislative fight.
Second, the policy logic behind using investment vehicles rather than grants is defensible on its own terms for at least some of this activity. Transmission infrastructure, manufacturing scale-up and housing development are exactly the kind of long-horizon, capital-intensive undertakings where a return-seeking loan or equity stake, rather than a one-off grant, can crowd in private co-investment and give the Commonwealth some prospect of getting its capital back — a genuinely different economic proposition to a subsidy that is spent once and gone. The Housing Australia Future Fund, in particular, was explicitly designed to give social and affordable housing programs a durable funding stream that does not have to be re-argued and re-appropriated in every annual budget cycle, a point examined in more detail in this site's dedicated analysis of the HAFF and Help to Buy.
Third, scrutiny of these vehicles has not been zero, even if it is lighter than for ordinary appropriated spending. Executives from the CEFC, the NRF and Housing Australia have appeared before Senate estimates hearings, where committees can and do question investment decisions, portfolio performance and governance arrangements; each entity's annual report is tabled in Parliament and open to scrutiny by any member, journalist or member of the public; and the Auditor-General retains the power to conduct a performance audit of any of them under the Public Governance, Performance and Accountability Act 2013, as it periodically does across the Commonwealth's corporate entities. This is real, if less frequent, oversight — a genuinely different thing to no oversight at all, and it is a distinction the critique above should not be read as erasing.
Fourth, it is fair to note this government inherited, rather than invented, the accounting convention itself. The distinction between underlying and headline cash balance, and the practice of running some Commonwealth investment off-budget, predates this government by well over a decade; NBN Co is the largest and longest-running example, built under a Labor government from 2009 and carried through Coalition governments from 2013 to 2022 without either side of politics fundamentally challenging the accounting treatment itself. What has changed under this government is the pace and combined scale of new off-budget commitment, not the existence of the mechanism — a distinction worth holding onto even while making the case, as this analysis does, that the pace and scale now warrant closer scrutiny than the convention has typically received.
The risk to the future
A shadow budget that keeps growing while the headline number stays flattered
The forward-looking risk is not that off-budget investment vehicles exist — they have existed, largely uncontroversially, for over a decade. It is that the mechanism is now being used at a scale and pace that lets a government, of either party, keep committing tens of billions of dollars without that commitment ever becoming visible in the single number the public has been trained to treat as the report card on the nation's finances. Each individual vehicle can be defended on its own merits, as this analysis has tried to do fairly. But nothing in the current structure caps how many more such vehicles get created, or requires any government to publish a single consolidated figure for the Commonwealth's total off-budget exposure, updated each budget, the way the underlying cash balance itself is.
That matters for the next generation in two concrete ways. First, if the investments underperform — and a portfolio of loans and equity in early-stage manufacturing, long-lead transmission infrastructure and a capped-return housing fund will not all perform equally well — the eventual write-downs land as expenses on a future government's watch, years after the original announcement, at a point when today's political credit for "off-budget" spending has long since been banked and today's decision-makers have moved on. Second, and more structurally, a headline fiscal debate that no longer captures a growing share of real Commonwealth financial commitment makes it harder for voters, the Parliament and financial markets to judge, in real time, whether the country's finances are actually improving or merely being reported as though they are — a distinction that matters most in precisely the moment a government most wants to avoid scrutiny of it.
None of this requires assuming bad faith. It requires only that the next government facing a difficult budget, of whichever party, will face the same incentive this one had: a legitimate accounting convention that happens to make the headline number look better than the whole picture, used more, not less, over time. Left unaddressed — for instance, by requiring Budget Paper No. 1 to carry the same combined, off-budget aggregate this analysis had to construct by hand — the gap between what the deficit figure says and what the country actually owes and has committed is more likely to widen than to close.