A budget about today, dressed in the language of tomorrow

Treasurer Jim Chalmers has now handed down five budgets. All of them, including this one, project deficits. The forward estimates show deficits stretching to 2029-30 and beyond. The budget itself does not return to balance until 2034-35, on Treasury's own modelling. Gross debt crosses \$1 trillion next year and reaches \$1.25 trillion within four years. Interest on that debt is now the second fastest growing expense in the budget, behind only the National Disability Insurance Scheme. At \$30 billion a year, interest alone exceeds the total cost of the Pharmaceutical Benefits Scheme.

The Treasurer calls this "responsible fiscal management." The budget is built around a claim of intergenerational fairness. Both descriptions sit awkwardly with the arithmetic, but the deeper problem is simpler. This is a budget about managing the present. It does very little to set the country up for the future: the productivity slump goes largely unaddressed, fiscal resilience continues to erode, and the housing supply problem is treated as a tax problem rather than a supply problem.

What a good budget should do

A good budget should do four things. It should respond to the immediate economic moment. It should repair structural problems that compound over time. It should distribute the burden fairly. It should avoid leaving large hidden bills for later.

These tests pull against each other, which is why budgeting is hard. A budget that aces the first by spending heavily can fail the fourth. A budget that aces the third by taxing accumulated wealth can fail the second if it discourages the investment that future prosperity depends on. The judgment is always about balance.

The second test is the one this budget most clearly fails. Productivity growth is stuck well below its long-run average. The structural deficit is normalising. The housing market is structurally short. The tax base is narrowing onto labour. None of these problems is new. None of them was created by this budget. But a budget judged by what it does for the future has to be judged by whether it addresses them, and on that test this budget mostly does not. It reforms what is politically saleable. It leaves the structural agenda largely untouched.

The pattern under the headline

Look at the underlying cash balance over the past fifteen years and a pattern emerges that cuts through the partisan debate. Every set of forward estimates, going back to 2010, slopes upward toward zero or surplus. Every actual outcome sits well below where those forward estimates promised. Treasury and successive treasurers have been forecasting a return to surplus for fifteen years. The return to surplus has happened twice, in 2022-23 and 2023-24, and not because the forecasts were right. It happened because revenue blew through expectations on the back of commodity prices and the post-pandemic income tax surge (in part thanks to bracket creep).

The structural position was never balanced. The cyclical lottery delivered the surplus. The moment the lottery stopped paying, the deficits returned, exactly as they always do. But Chalmers, having banked the two cyclical surpluses early in his tenure, has stopped presenting structural balance as a near-term objective. The new forward estimates normalise persistent deficits rather than treating them as temporary.

Australia has not had structural fiscal balance in living memory of most working-age voters. The government's revenue base does not cover its spending commitments. The gap is closed by debt. The debt is serviced by future revenue. Future revenue is the inheritance of the next generation.

A complication worth acknowledging. The post-2000 Australian state is structurally larger than the state Keating disciplined. Voters demand more services. The NDIS, expanded Medicare, larger defence commitments and aged care obligations reflect political choices that have been validated repeatedly at the ballot box. The question is not why spending isn't 1990s-sized. It is whether the tax base has been adjusted honestly to match the commitments, or whether the gap is being closed by debt and by silent tax increases that nobody had to vote for.

The structural pressures the Treasurer cites (NDIS, aged care, defence) are real, but they cannot fully explain the pattern. The deficits ran steadily through 2010 to 2019, before the NDIS reached scale and well before the current defence build. The pattern is partly political. Spending grows in good times and bad. Tax cuts are offered in good times and not reversed in bad times. The structural gap widens.

The silent tax on the young

Bracket creep is the most important number not mentioned in the Treasurer's speech.

Income tax brackets in Australia are set in dollar amounts that do not change with inflation. The 30% rate starts at \$45,000. The 37% rate starts at \$135,000. The 45% rate starts at \$190,000. These thresholds are adjusted occasionally through legislated tax cuts, but they are not automatically indexed. Wages meanwhile rise with inflation and productivity. So every year, more of every worker's income falls into higher brackets, and their average tax rate rises even when their real income has not.

This is the mechanism the budget relies on to close part of the structural deficit. Individual income tax revenue is forecast at \$390.9 billion for 2026-27, nearly half of total revenue. Through the forward estimates, it grows faster than wages and substantially faster than GDP. The arithmetic only works because the brackets are not indexed. The Treasurer has been asked directly whether he will index them. He has said no.

The Coalition's analysis, which Treasury has not disputed, finds that a worker earning \$130,000 in 2028-29 will lose around \$7,000 of their decade of wage gains to bracket creep alone, on top of \$38,000 lost to inflation. Real after-tax income for that worker barely moves over ten years despite ten years of pay rises. The \$250 Working Australians Tax Offset, the budget's centrepiece relief for workers, is worth around \$5 a week. For most full-time workers, bracket creep takes more than \$5 a week, every week, silently.

The cohort that pays bracket creep most heavily is the cohort that earns predominantly from wages. Younger workers, on average, at the start of their careers, with few assets to shelter income. The cohort that bears bracket creep least is the cohort that earns disproportionately from accumulated assets. Older Australians, as a group, with capital gains now indexed for inflation, with superannuation taxed at 15% or zero, with franking credits, with the family home untaxed regardless of value, with negative gearing grandfathered on the properties they already own. The picture is not uniform. There are renters in their sixties and wealthy professionals in their thirties. But the aggregate pattern is real and the policy effect is large.

The Treasurer talks about rebalancing the system between assets and labour. The system the budget actually delivers leans further onto labour and away from assets every year, automatically, by the simple device of not indexing the brackets. The narrowing of the tax base onto younger wage earners is itself a structural problem the budget does nothing to address. It is, in fact, the budget's quiet plan for closing the deficit.

The housing reforms in plain English

The centrepiece is the housing and tax package, and it deserves careful explanation because the headlines have done real damage to clarity.

Negative gearing lets a property investor deduct the costs of owning a rental property (interest, maintenance, depreciation, rates) against their other income, including wages. If the property costs \$40,000 a year to hold and generates \$28,000 a year in rent, the \$12,000 loss reduces the investor's taxable income. At a marginal rate of 37%, that saves \$4,440 a year in tax. The system has existed in something like its current form since the 1980s.

The Capital Gains Tax discount, introduced by Peter Costello in 1999, halves the tax payable on capital gains from assets held longer than a year. So if you bought shares for \$100,000 and sold them for \$200,000 after two years, only \$50,000 of the \$100,000 gain is taxed, at your marginal rate. This replaced an earlier system under which the cost base of the asset was adjusted for inflation before tax was calculated.

The budget changes both. From 1 July 2027, the 50% CGT discount is replaced with inflation indexation, which is closer to the pre-1999 design. So tax is calculated on the real gain, after adjusting the purchase price for inflation. A minimum effective tax rate of 30% applies, regardless of the investor's normal marginal rate. Negative gearing on established residential property purchased after 12 May 2026 can no longer be deducted against wages or other non-property income, though losses can still be deducted against rental income from other properties and against capital gains when the property is eventually sold. Existing properties are grandfathered. New builds keep the old rules. Shares and commercial property are untouched.

This is a smaller change than the headlines suggest. Negative gearing has not been abolished. It has been narrowed and partially redesigned for one specific case. The CGT discount has been replaced rather than removed. The reforms target the asset class (established residential property) where the supply problem is most acute, and they do it through tax settings that have been recognised as distorting for decades.

This is also the one place where the budget does attempt structural reform. The CGT discount was too generous, indexation is the textbook-correct replacement, and redirecting negative gearing toward new supply is defensible. The instinct is right. The question is what the design actually delivers.

Where the implementation breaks down

The problems are in the design detail, and they matter because they invert the policy's claimed equity story.

The negative gearing change, as I read it, operates at portfolio level, not property level. This means an investor with several existing properties producing positive net rent can absorb the losses on a new purchase against that rent. The cost is silent. An investor with no existing portfolio has nothing to absorb the loss against. They pay the full cost.

Consider two investors after July 2027. The 35-year-old IT worker buying their first investment property, negatively geared by \$12,000 a year, cannot deduct that loss against their wages. They lose roughly \$4,000 of tax benefit every year until they either acquire another rental property or sell. The 60-year-old surgeon with four properties bought years ago and now positively geared by \$20,000 a year in aggregate buys a fifth property with a \$15,000 loss. The loss disappears into the existing rent. They lose nothing. The policy bites on the new entrant. The established holder is largely insulated.

The 30% minimum CGT rate has the same pattern. It only binds on investors whose marginal tax rate is below 30%, which means people on incomes under roughly \$45,000. High income investors continue to pay their marginal rate on the indexed gain. The floor never binds for them. So a wealthy retiree on \$250,000 a year selling a portfolio gain pays 45% on the real gain, as they would have anyway. A low-income retiree selling their one investment property to fund aged care pays 30% on the gain when their normal marginal rate would have been 19%. The government has carved out Age Pension recipients, which softens the worst case, but the structure of the rule remains regressive at the margin.

Grandfathering compounds the problem. Every property purchased before 12 May 2026 continues under the old rules forever. So the older investor with four properties accumulated over the past twenty years keeps full negative gearing on all of them, keeps the 50% CGT discount on gains accrued before July 2027, and faces no change in their position. The reform applies only to future purchases, which means future investors.

The cumulative effect is a policy that protects existing investor wealth almost entirely while filtering new entrants at the door. The one structural reform the budget does attempt ends up shifting cost from established holders to aspirants, which is the opposite of the stated intent.

What the reforms will actually do

Treasury's own modelling expects 75,000 additional first home buyers over a decade. That is roughly 7,500 a year against an annual purchase volume around 500,000. House prices grow 2% slower than they otherwise would, according to Treasury. Independent forecasters put the price effect at 2 to 5% over a year or two. The Housing Industry Association says supply falls by 35,000 homes over the decade before the \$2 billion infrastructure spend adds 65,000 back. Net effect: a small positive on home ownership, a small negative on prices, a small positive on new supply, all spread thinly over ten years.

The rental market is the casualty almost nobody is discussing. Australia is running net overseas migration of around 300,000 a year (which the budget predicts will drop to around 250,000 a year), almost all of whom arrive as renters. Completions are running at 170,000 dwellings a year against a target of 240,000. The structural shortage is severe and getting worse. Housing supply is not purely a population arithmetic problem. It is also a planning problem, a labour problem, a construction productivity problem and a financing cost problem. But the migration intake at current levels into a market with these supply constraints makes adjustment materially harder, not easier. Layer on a policy that reduces investor demand for established stock immediately while the supply response from new builds takes five to ten years to flow through, and the rental market gets worse before it gets better.

Treasury forecasts \$2 a week of additional rent. Most independent forecasters expect rent rises materially larger than this, with two-year forecasts ranging from a few percent to double digits depending on assumptions about investor response, migration intake and supply elasticities. The Treasury number is at the optimistic end of a wide range; nobody else operating in the housing forecasting space is publishing numbers that low.

Rents are set at the margin where the last renter is willing to pay. The marginal renter is the recently arrived migrant or the working poor facing a thin market. The hit falls on them, not on the established homeowners the rhetoric implies are the policy's target.

The housing reform is the budget's one serious attempt at structural change, and even there the structural effect is modest. The supply problem barely moves. The rental market gets worse before it gets better. The benefit accrues slowly to a small number of marginal first home buyers. This is what the structural reform looks like. The reforms that aren't attempted at all are bigger.

The reform that is missing

Beneath both the housing problem and the fiscal trajectory sits a deeper structural problem: productivity growth.

Productivity is the mechanism by which each generation makes the next one better off. Real wages cannot rise sustainably without it. The tax base that funds rising service expectations depends on it. The capacity to grow into the debt this budget accumulates depends on it. Treasury's own forecasts show productivity growth running around 1.2%, against a long-run average closer to 1.7%. Treasury describes the trajectory as a "profound and protracted slowdown." Productivity compounds, and small differences in the annual rate become large differences in a working life. The difference between 1.2% and 1.7% over a working career is the difference between an Australia that is recognisably wealthier than today and one that has fallen behind its peers.

The budget's productivity response is thin against this diagnosis. \$10.2 billion of claimed regulatory burden reduction (a number budget papers have historically used loosely). \$1.5 billion for research institutions. Some welcome but modest changes to R&D tax incentives. Expanded venture capital settings. Faster environmental approvals. A "Single National Market" initiative whose claimed \$13 billion annual GDP benefit rests on assumptions about state cooperation that are not yet evident. These are not nothing. They are also not a productivity strategy.

A productivity strategy would address the binding constraints. Skills and education at every level. Competition policy that disciplines the oligopolies in banking, supermarkets, energy retail and aviation that extract rents without delivering productivity. Energy policy that delivers reliable, cheap power, the absence of which has driven manufacturing offshore and is now driving data centres there too.

The intergenerational fairness rhetoric is loudest on housing, where the policy effect is modest. It is quietest on productivity, where the effect could be transformational.

Fiscal resilience is the inheritance

The deepest cost of the current trajectory is not the debt itself. Australia is not approaching a sovereign debt problem. By international standards Australia's net debt remains low, its sovereign credibility is intact, its interest burden as a share of GDP is manageable, and its taxing capacity is substantial. The question is not sustainability. It is resilience.

Fiscal resilience is the room a government has to respond to crises. The GFC stimulus was \$50 billion. The pandemic response was over \$300 billion. Each cushioned a severe economic shock. Each was possible because previous governments had preserved fiscal flexibility in the good years. The deficits flowed in the bad years and were partially closed in the recoveries. The cycle worked.

Deficits that fund productive investment can leave future generations better off; the concern here is persistent deficits funding recurrent consumption during favourable conditions. A government running \$65 billion headline deficits in the good years has less flexibility when the next bad year arrives. Either it accelerates the debt trajectory more steeply than it otherwise would, or it lets the recession bite harder than it needs to. Both choices are worse than they would have been with more room. Preserving that room requires building it in advance. The current budget runs the room down rather than preserving it.

It does so during a period of historically favourable conditions. Unemployment is at 4.2%. Commodity prices are elevated. Population growth is at record levels. Revenue is near a 30-year high. These are the conditions under which fiscal flexibility should be increasing. It is not. The structural budget position is what it would have been historically during a mild recession.

Ken Henry and John Fraser, both former Treasury secretaries, made this point bluntly in the days after the budget. Henry: with low unemployment and revenue near record highs, the budget should be in surplus. Fraser: Keating and Costello imposed binding spending limits, and Chalmers has not. Real payments growth at 1.5% per year sounds restrained until you remember that Keating cut real spending outright for three years to deal with a similar set of pressures.

The Treasurer's response is that times have changed and structural pressures are different now. This is true and it is partial. Every treasurer who declines fiscal discipline says the times are different. The discipline is what builds the room for the next crisis.

Whoever inherits the next downturn will start from a worse position than this government did. The fiscal flexibility Albanese and Chalmers inherited from Frydenberg, who inherited it from Morrison, who inherited it from Howard and Costello, is being eroded. The next government, whether Labor or Coalition, will have less to work with. The generation that pays for this is the same generation the budget claims to champion.

What a budget for the future would have looked like

The framework also suggests the alternative. A budget that took the structural agenda seriously, judged by the four tests, would have looked different.

On the immediate moment: the cost-of-living measures and fuel relief are reasonable. Leave those.

On structural repair: address the actual problems. The housing constraint is supply, not investor advantage. Ten times the announced infrastructure spend, tied to state-level planning reform, with migration moderation until completions catch up. A productivity strategy at the centre of the budget rather than the periphery: skills investment at scale, competition policy that disciplines incumbents, energy policy that delivers reliable cheap power. The Single National Market initiative is the right idea at one-tenth the necessary ambition.

On distribution: target the assets where wealth is actually concentrated. The family home and superannuation tax concessions are far larger leakages than discretionary trusts. A broad-based land tax, phased in over ten years and offsetting stamp duty, is the textbook reform that every serious review has reached. None of these were touched. The reforms hit the assets where political resistance was lowest, not the assets where the distributional issue was largest.

On hidden bills: index the tax brackets, so the cost of bracket creep is visible rather than silent. Build a path to surplus during good times. Cap real spending growth at zero for three years. Keating cut spending; sustaining it would be modest by historical standards. Accept that fiscal flexibility is an inheritance and treat it as one.

This is a more progressive budget than the one delivered, not less. It taxes accumulated wealth more (land tax, family home cap above a threshold, super reform). It protects future generations more (fiscal consolidation, indexed brackets, productivity reform). It addresses the housing problem more (supply at scale, migration moderation). It is also harder to deliver, which is why it did not happen.

Verdict

The 2026 budget is reformist in symbolism and conservative in substance. The one place it does attempt structural reform is housing, where the instinct is right but the implementation protects existing wealth and shifts costs onto new entrants. Everywhere else, the structural problems that determine the country's future are largely untouched. The productivity slump that determines living standards. The fiscal drift that erodes crisis capacity. The narrowing tax base that funds spending growth silently through bracket creep. None of these is addressed seriously.

The budget manages the present competently. The cost-of-living measures are reasonable. The fuel relief is reasonable. The health and aged care commitments are reasonable. The political positioning, particularly to under-45 voters on housing, is shrewd. But a budget judged on whether it sets the country up for the future has to be judged on whether it addresses the structural agenda, and on that test this budget mostly does not. It is a budget about today, dressed in the language of tomorrow.

This is the fifth Labor budget. The deficits are now structural. The debt is approaching \$1.25 trillion. The fiscal flexibility that funded the GFC response and the pandemic response is being eroded. The productivity slump that determines future living standards is being treated as a chapter heading rather than a crisis. Whoever holds office when the next crisis arrives will have less to work with, and the country they inherit will be growing more slowly than it could be. That is the real cost, and it is the one the budget's rhetoric leaves unaddressed.