Picture a couple renting a two-bedroom unit in outer Melbourne. Both work full-time. They are not on any government payment and do not hold a concession card, and by most measures they are doing everything right. Between 2021 and 2025 their rent rose by several hundred dollars a month (approx.), their electricity bill climbed even after a government rebate landed on it, and the weekly shop that once filled the boot of the car now leaves gaps on the shelf at home. They are the household the phrase "cost-of-living crisis" was built to describe: full-time work, no financial buffer, and every essential bill moving in the same direction at once.
That is the defining feature of the squeeze that gripped Australia from 2021 onward. It did not spread evenly across everything households buy. A television is cheaper today, in real terms, than it was five years ago. A pair of jeans is barely dearer. But the costs almost no one can opt out of — a roof, power, food, insurance, a visit to the doctor — rose far faster than everything else, quarter after quarter, for the better part of four years. Understanding why means taking the Consumer Price Index (CPI, the Australian Bureau of Statistics' measure of average price change across the economy) apart, piece by piece, instead of treating "inflation" as one number that simply rises or falls.
What the government did
The government did not try to cut the price of housing, power, food, insurance or health care outright. Instead, across four successive Budgets from October 2022 to May 2025, it built a cost-of-living relief package around three main levers, each renewed or extended year to year rather than delivered once and left alone: direct energy rebates, cheaper medicines, and a higher rate of rent assistance.
Energy bill relief
The May 2023 Budget created the Energy Bill Relief Fund, delivered jointly with the states and territories, paying rebates of up to $500 to eligible pensioner, concession-card and low-income households and up to $650 to eligible small businesses (2023–24 Budget Paper No. 2, approx.). The May 2024 Budget widened this into a near-universal rebate: $300 credited to every household's electricity account and $325 to every eligible small business, paid as quarterly instalments through the financial year rather than a single cheque (2024–25 Budget Paper No. 2). A further, more targeted round followed in the 2025–26 Budget (approx.). None of these rebates changed the price a retailer charges per kilowatt-hour. They lowered the number printed on the bill the household actually pays, for as long as the funded rebate period lasted.
Cheaper medicines
From 1 January 2023, the government cut the maximum cost of a standard Pharmaceutical Benefits Scheme (PBS, the scheme that subsidises the cost of prescription medicines) prescription from $42.50 to $30 — the first reduction to the general co-payment in 75 years (Department of Health and Aged Care, Jan-2023). The change was estimated to save patients more than $280 million in its first year (approx.), and a further reduction toward $25 was flagged in a later Budget (approx.). Unlike the energy rebate, this was not a one-off payment. It permanently reset the ceiling price on subsidised medicines going forward.
Rent assistance
Commonwealth Rent Assistance (CRA), the payment that tops up the pensions and allowances of low-income renters, rose by 15% in the May 2023 Budget, effective September 2023, and by a further 10% in the May 2024 Budget, effective September 2024 — the first back-to-back increase to the payment's maximum rate in more than 30 years (Department of Social Services, Sep-2024). By late 2024 the maximum fortnightly rate stood at roughly $188 for a single renter and $177 for each member of a couple (approx., DSS payment rates, Sep-2024). As with the medicines change, this was a permanent increase to the base rate, not a temporary credit.
Around these three measures sat a wider package: a redesigned round of income tax cuts delivered to every taxpayer from July 2024 (examined in full elsewhere on this site), cheaper child care, fee-free TAFE places, and a cap on the indexation applied to outstanding student loans. Taken together, the government's central claim was consistent across every Budget: it could not set the price of an essential good, but it could soften the impact of that price on the households least able to absorb it, and it would keep doing so, Budget after Budget, for as long as the shock lasted.
The timing of that package matters for judging it fairly. The government took office in May 2022, three months after global energy prices spiked and as headline inflation was already accelerating toward its December-2022 peak. Its first Budget, in October 2022, was deliberately restrained — a small, targeted package rather than broad relief — on the explicit basis that large new spending would work against the RBA's rate rises rather than with them (2022–23 Budget Paper No. 1, approx.). The bigger, broader measures set out above did not arrive until May 2023 and May 2024, once the inflation peak had passed and the RBA's tightening cycle was closer to its end. That sequencing is itself part of the government's case: relief was scaled up as the risk of it adding to inflation fell, rather than delivered all at once regardless of the monetary-policy backdrop.
The strongest case for the policy
Put the government's defence at its best
The government delivered targeted relief — energy rebates, cheaper medicines and rent assistance increases — aimed squarely at the households under the most pressure, and did so while an independent central bank was still raising interest rates to bring inflation down. Broad-based, untargeted spending into that tightening cycle would have fought the Reserve Bank of Australia (RBA) directly, arguably forcing the cash rate even higher for even longer. Instead, Treasury designed the relief to be, in its own language, "responsible": sized and targeted so it would not add materially to inflation, while still reaching pensioners, concession-card holders and low-income renters first. The PBS and rent assistance changes were not gestures. Both permanently reset a price or a payment rate, and both were the largest moves of their kind in decades. A government that inherited a global inflation shock it did not cause, and that used the tools actually available to it — the Budget, not the price mechanism — to blunt that shock for the households with the least room to move, has a reasonable claim to have done what a responsible government facing an external crisis should do.
The evidence: what actually drove the squeeze
Scale matters before the components do. For the decade before the pandemic, Australian headline inflation averaged a little under 2% a year — inside, or below, the RBA's 2–3% target band almost throughout (RBA, approx.). Nothing in the 2021–2025 period looked like that. The jump to a 7.8% peak was the sharpest break from that baseline in a generation, which is why the relief measures examined above were built, and should be judged, against a genuine shock rather than routine price growth.
Headline inflation versus the price level underneath it
Headline CPI peaked at 7.8% in the year to the December quarter of 2022 — the fastest annual price growth in more than three decades (ABS, Consumer Price Index, Australia, Dec-2022). From there it fell steadily: to around 4.1% by December 2023 and around 2.4% by December 2024 (approx., ABS CPI, Dec-2023 and Dec-2024). On the surface, that trajectory looks like the crisis simply ending on its own. It was not that simple.
Two things complicate the headline number. First, the RBA and the ABS both track a second measure, the trimmed mean (a "core" inflation gauge that strips out the largest one-off price swings each quarter to reveal the underlying trend). The trimmed mean stayed higher than the headline rate through 2024, running at 3.2% over the year to December 2024 even as headline CPI read 2.4% for the same period (ABS, Dec-2024). That gap matters. It means the underlying rate of price growth was still running above the RBA's 2–3% inflation target band even as the number reported on the evening news kept falling.
Part of why the trimmed mean stayed sticky is that services prices — haircuts, insurance, health appointments, other person-to-person services — are heavily driven by labour costs, and labour costs kept rising even as goods prices, much of it imported, cooled. Services inflation is harder to bring down with a rate rise or a rebate, because it reflects wages paid to Australian workers rather than the price of an imported container of goods. The RBA repeatedly flagged this as the stickier half of inflation through 2024 (RBA statements, 2024, approx.), and it connects directly to the productivity story examined in full under The productivity collapse: when wages rise faster than output per worker, the extra cost has to show up somewhere, and services prices are often where it shows up first.
Second, and more directly relevant to this topic, the ABS said plainly, release after release through 2023 and 2024, that government electricity rebates were suppressing the measured price of power in the CPI basket (ABS CPI releases, 2023–2024, approx.). A rebate that lowers the dollar figure on a household's power bill lowers the price the ABS records for electricity, even though the retailer's underlying tariff has not fallen by the same amount, or at all. The headline number, for those years, was in part a policy artefact: an accurate record of what households actually paid, but not a reliable measure of what the underlying cost of essentials was doing underneath the rebate.
A simple test: does it change the price, or just who pays it
The distinction this topic turns on is not complicated, and it is worth stating plainly before working through the evidence. A policy fixes a structural cost problem if it changes the price the market charges — more homes built lowers rent growth; more firmed generation capacity lowers wholesale power prices; a functioning competitive grocery market lowers retail margins. A policy provides temporary relief if it changes who pays an unchanged price — a rebate that covers part of a power bill, a subsidy that covers part of a medicine's cost, a payment top-up that offsets part of a rent rise. Both kinds of policy are legitimate and can coexist. But only the first kind survives the day the Budget line item is not renewed. The second kind, by construction, cannot outlast its own funding.
Inside the basket: five essentials, five different stories
Averages hide as much as they reveal. The single "inflation rate" reported each quarter is a weighted average of price changes across roughly eleven CPI groups. Breaking it into the five essentials at the centre of this topic — housing, energy, food, insurance and health — shows a squeeze concentrated overwhelmingly in the things a household cannot defer or substitute away from.
Housing carries the largest weight of any group in the CPI basket — 21.4% of the average household's spending on the ABS's own 2025 weight update, ahead of food and non-alcoholic beverages at 17.4% (ABS, Annual Weight Update of the CPI and Living Cost Indexes, 2025, approx.). The group combines rents, the cost of building or renovating a home, and, for owner-occupiers with a mortgage, mortgage-interest charges. Advertised rents climbed sharply as vacancy rates fell toward multi-decade lows — a dynamic examined in full under Rents, mortgages and housing-cost stress — while mortgage-interest charges rose in step with the RBA's cash rate, which climbed from a record-low 0.10% to 4.35% across the tightening cycle that began in mid-2022 (RBA, approx.). Housing-group inflation was running at roughly 6.5% over the most recent full year available at time of writing (ABS CPI, approx., late 2025). Only part of that is a national story; state planning and zoning settings, examined under The States, shape how quickly new supply can respond.
One further wrinkle sits inside the housing number and is worth naming precisely: the CPI's mortgage-interest-charges line, which captures what owner-occupiers pay in interest, moved almost mechanically with the RBA's cash rate rather than with any housing-market fundamental. As the cash rate rose across thirteen separate increases between May 2022 and November 2023, this line surged for every mortgage holder whose rate reset, regardless of what was happening to rents or vacancy rates in their suburb (ABS CPI, approx.; RBA cash-rate decisions, 2022–2023). "Housing costs," in other words, actually bundles two different stories: a rental-market story driven by supply and vacancy, and a monetary-policy story driven by the RBA's own independent decisions. A rebate touches neither. Only more supply touches the first; only interest-rate cuts, set independently of any government, touch the second.
Energy — specifically electricity, the component the ABS tracks directly — is the clearest illustration of the temporary-versus-structural split this topic exists to draw out. Electricity prices measured in the CPI rose 21.5% in the twelve months to December 2025, and the ABS attributed this directly to state and federal rebates being drawn down or exhausted, no longer masking the underlying tariff households actually pay (ABS CPI, Dec-2025). In the years the rebates were fresh, electricity's measured price barely moved, or fell. The moment the credit ran low, the number households actually pay reasserted itself. The underlying driver — the cost of building, firming and transmitting an electricity system moving away from coal-fired generation, examined in full under Energy & Climate — did not go anywhere in the meantime. It was simply not being paid, for a while, by the household holding the bill.
Food rose more moderately than housing or energy but from a much larger, harder-to-avoid base — nobody opts out of eating. Food and non-alcoholic beverage prices rose roughly 3.3% over the most recent full year available, a slower pace than the 2022 peak but still a further increase stacked on top of the sharp rises already banked since 2021, and food carries the second-largest weight in the basket after housing (ABS CPI, approx.). The drivers were a mix of global input costs — fuel, fertiliser, shipping — and the market structure of Australian grocery retail. The government's main policy response was not a price control but an inquiry: the Australian Competition and Consumer Commission (ACCC) ran a supermarket pricing inquiry through 2024, examining the pricing practices and market power of the major chains, a process this site covers in full under Groceries, food inflation and market power.
Insurance produced one of the sharpest price rises of any CPI component: 14.0% in the year to the June quarter of 2024, driven by higher reinsurance costs (the cost insurers themselves pay to spread their own risk), more frequent and more costly natural-disaster claims, and rising vehicle and building repair costs (ABS, Jun-2024). Insurance is compulsory or close to it — home, contents and car cover, and for anyone with a mortgage, building insurance, are not optional line items in a household budget — which makes this one of the more regressive price rises in the basket. A fixed-dollar premium increase takes a much larger bite out of a low-income household's budget than a high-income one's, and no rebate examined in this topic was aimed at insurance specifically.
Health costs, chiefly medical and hospital services and private health insurance premiums, rose around 3.8% over the most recent full year available, with medical and hospital services the largest single driver, up roughly 5.0% (ABS CPI, approx.). This is the one essential category where a specific government policy — the PBS co-payment cut — worked directly against the broader trend, lowering the price of subsidised medicines even as specialist fees, private hospital costs and annually regulator-approved private health insurance premium rises kept pushing the group higher. Bulk-billing rates and out-of-pocket costs for a standard GP visit are examined in full under Medicare and bulk-billing.
Inflation did not fall evenly. It fell for the things Australians can defer, and kept rising for the things they cannot.
The ABS's own split between non-discretionary spending (things bought to meet a basic need, maintain a current living arrangement, or meet a legal obligation — rent, electricity, insurance, groceries, fuel, school fees) and discretionary spending (largely optional purchases such as holidays, alcohol and electronics) confirms the pattern from a different angle. For most of the period from 2021 through late 2024, non-discretionary prices rose faster than discretionary prices. The first time that gap meaningfully narrowed was the December 2024 quarter, when non-discretionary prices actually fell slightly quarter-on-quarter (down 0.5%) while still running 1.8% higher over the full year — the first time in nearly four years that non-discretionary inflation ran below discretionary inflation (ABS, Dec-2024, approx.). In plain terms: for roughly three years, the cost of things households cannot avoid buying rose faster than the cost of things they can skip, and the gap only began to close once temporary rebates were doing part of the work of holding the non-discretionary numbers down.
Why targeted relief could not fix a structural problem
The government is right that it inherited, rather than caused, the two biggest shocks behind this squeeze. A pandemic snarled global supply chains and pulled forward demand as every advanced economy, Australia included, ran large stimulus programs to keep their economies afloat through lockdowns. Then Russia's invasion of Ukraine in February 2022 drove global gas and coal prices to records, pushing up the cost of the fuel behind a large share of Australia's electricity generation at the time. Those were global events. No Australian government of either major party could have stopped either one landing here, and both plainly show up in the 2022 inflation peak that predates most of the relief measures examined in this topic.
But at least three structural drivers behind this squeeze sit closer to home, are not temporary, and are not addressed by a rebate, a co-payment cut or a rent-assistance increase — however welcome each has been to the households receiving it.
The first is housing supply. Dwelling completions have run below the rate needed to keep pace with population growth for several years running, a shortfall examined in full under The housing affordability emergency and The migration surge and the housing collision. A cash rebate on an electricity bill does nothing to a capital city's vacancy rate. Until enough homes are built relative to the number of households needing one, rents and housing-group CPI inflation have no structural reason to slow, regardless of which party holds office federally or in the states.
The second is climate-linked insurance and disaster cost. Reinsurers price premiums using models of the frequency and severity of floods, cyclones and bushfires; as those events become more frequent and more damaging, premiums rise to match, and the 14.0% jump in insurance CPI in the year to June 2024 reflects that repricing directly (ABS, Jun-2024). No relief payment lowers a reinsurer's model of flood risk on the Hawkesbury or bushfire risk in the Adelaide Hills — that is a function of climate and geography, not Budget policy.
The third is the energy transition's near-term capital cost. Building enough new transmission lines, storage capacity and firmed renewable generation to replace an ageing coal fleet — a program examined in full under the Energy & Climate tab — carries a large, unavoidable capital cost that ultimately flows through to network charges and retail tariffs, whichever government manages the transition. Rebates paid down the household's bill for a defined period. They did not pay down that underlying capital cost, which remains on the system and, in most retail pricing models, returns to the bill once the credit stops.
This is where the temporary-versus-structural distinction at the centre of this topic becomes the whole story. The Energy Bill Relief Fund and the follow-on $300 rebate were, by design and by the government's own Budget papers, time-limited measures funded a year or two at a time (2023–24 and 2024–25 Budget Papers, approx.). When a rebate ends and is not replaced at the same scale, the number the ABS records for electricity does not fall back to where it would have been without any shock at all — it snaps toward the level the underlying tariff had already reached while the rebate was in place. That is close to what the data show happened in the twelve months to December 2025: electricity CPI jumped 21.5%, not because power stations suddenly became far more expensive to run in that single year, but because households started paying, in fuller measure, a price that had already been rising underneath the rebate the whole time (ABS CPI, Dec-2025).
A rebate can smooth a bill. It cannot rebuild a grid, build a home, or slow a warming climate.
The same logic applies, with less drama, to food and health. A cheaper PBS co-payment is a genuine, permanent saving on the medicines it covers — but it does nothing to the specialist fee, the private hospital excess, or the regulator-approved health insurance premium rise sitting next to it in the same CPI group, which is a large part of why health-group inflation kept running above 3% even after the co-payment cut took full effect. A supermarket pricing inquiry can expose market concentration and recommend reform, and the ACCC's 2024 inquiry did both, but a recommendation is not a price cut; it cannot by itself lower the cost of diesel, fertiliser or shipping that flows into a loaf of bread. Relief measures, in short, worked on the margin the government directly controls — the price of a specific subsidised good, or the size of a specific payment — while the deeper cost drivers behind housing, energy, insurance and much of health sat almost entirely outside that margin.
What the government got right — and the honest limits
Two of the three flagship measures examined here are not, in fact, vulnerable to the "temporary relief" critique in the same way the energy rebate is. The PBS co-payment cut and the Commonwealth Rent Assistance increases were permanent resets to a price and a payment rate, not one-off credits due to expire on a Budget cycle. Both will still be in place long after the last energy rebate instalment clears, and both delivered real, checkable savings to the households that received them — more than $280 million a year in medicine costs alone on Treasury's own estimate (approx.), and a genuine, roughly one-quarter cumulative lift to the safety net under low-income renters across the two increases combined (approx.), the first such back-to-back lift in more than three decades (DSS, Sep-2024). On distributional targeting, the package broadly hit its mark: concession-card holders, pensioners and renters — the groups with the least capacity to absorb a price shock — received the largest, most durable increases of anyone.
It is also fair to flag a design critique that came from economists rather than political opponents: extending the 2024–25 rebate to every household, including high-income households with no particular cost-of-living pressure, diluted the distributional precision of the earlier, means-tested 2023–24 round. A rebate paid to a high-income household suppresses measured electricity CPI by the same amount as one paid to a pensioner, but it does far less for genuine hardship, and it costs the Budget more for a given amount of targeted relief actually delivered. Universality bought political simplicity and administrative ease — no means test to apply, no eligible household missed out — at some cost to the distributional logic the steel-man case above rests on.
It is also fair to note the fiscal-monetary constraint the government was working under for most of this period. With the RBA raising the cash rate through 2022 and 2023 specifically to cool demand and bring inflation back inside its target band, a government spending heavily and indiscriminately into that fight would have made the RBA's task harder, not easier, and risked an even higher peak in the cash rate or a longer wait before rate cuts began. Treasury's own language — describing the relief as "responsible" and modelled to be broadly disinflationary rather than inflationary in its near-term effect — reflects a real design constraint, and the RBA's own commentary through the period broadly accepted that the targeted rebates were, at minimum, not working against it (RBA statements, 2023–2024, approx.).
The honest limit is that targeting and permanence do not, on their own, add up to sufficiency. Even a roughly one-quarter cumulative lift to rent assistance across two Budgets (approx.) has not kept pace with advertised rent growth over the same years, which this site's Rents, mortgages and housing-cost stress topic puts at roughly 30–40% nationally from the pandemic trough (CoreLogic; ABS rents series, approx.). Cheaper medicines do not touch the much larger and faster-growing medical and hospital services bill sitting beside them in the same CPI group. The measures were real, they were targeted, and in two of three cases they were durable rather than temporary — but they were relief calibrated to smooth a shock, not a redesign of the systems generating the underlying cost growth in housing, energy, insurance and health.
The risk to the future
A second cost shock, with less patience left to meet it
The risk is not that the relief measures were wrong to introduce. It is that a public conditioned to watch headline inflation fall may not distinguish between "prices are falling" and "the price rises are being paid by the Budget instead of the household, for now." Every rebate funded from the Budget is a rebate that must eventually be renewed, replaced, or withdrawn — and withdrawal, as the December-2025 electricity read shows, produces a sharp, visible price jump that lands on households at the exact moment the structural drivers underneath it — a housing shortfall, a repricing climate, an ageing population's rising health bill, an energy system still mid-transition — are still compounding. Two or three consecutive Budgets funding rebates to hold headline inflation down is not a stable long-run equilibrium; it is a bet that supply-side fixes — more homes built, more firmed generation connected, a more competitive grocery sector, faster productivity growth to fund a growing health and aged-care bill — arrive before the fiscal room to keep extending relief runs out. If that bet is wrong, the next generation inherits both the unresolved structural cost growth and a Budget with less capacity left to cushion it, at a point when public trust in government's ability to manage the cost of living has already been drawn down once.
Back in outer Melbourne, the couple from the opening of this topic will keep doing the same fortnightly sum. If the next rebate is not renewed at the same scale, their power bill will jump in a single quarter for reasons that have nothing to do with anything they did. Their rent will keep climbing until enough homes exist to house everyone who needs one. Their insurance premium will keep tracking a climate model, not a Budget line item. Relief bought them, and millions of households like them, real time and real dollars. It did not buy them a lower cost of living once the funding runs out — only the government's continued willingness, Budget after Budget, to keep paying part of the bill on their behalf.