In 2023, I wrote that inflation and interest rates could remain higher for longer than most people expected. That view was based on 50 years of history showing that once inflation exceeds 8%, it usually takes a decade or more to settle back down.
Underlying inflation subsequently fell from around 4% in mid-2024 to a low of 2.7% in the June quarter of 2025. However, it has since climbed back to 3.6%. After cutting rates through 2025, the Reserve Bank has already raised them 3 times in 2026, possibly 4 by the time you read this. So far, that is broadly consistent with the research I cited in 2023.
However, revisiting that research alongside newer studies has changed my understanding of why inflation is proving so persistent. Several factors are reinforcing each other, a pattern also evident in previous periods of high inflation.
More importantly, I think there is another reason interest rates may need to work harder to slow spending, and it hasn’t received enough attention. A growing share of the population, including a growing share of households with mortgages, is less affected by rate rises than was the case 20 years ago.
The inflation rebound the RBA missed
The trimmed mean, the Reserve Bank’s preferred measure of underlying inflation, fell steadily from 2023 to a low of 2.7% in the June quarter of 2025. It then climbed to 2.9% in September 2025, 3.5% in March 2026 and 3.6% in June 2026, where it has since levelled off. Headline inflation followed a similar pattern, falling to 2.1% in mid-2025 before rising to 3.9% by June 2026. A Middle East conflict also caused a temporary spike in monthly headline inflation in March 2026 as fuel prices rose.
The Reserve Bank did not anticipate this rebound. In February 2024, it forecast inflation would return to the 2% to 3% target band during 2025 and reach the midpoint by 2026. Its November 2024 forecast still had trimmed mean inflation falling to 2.5% by late 2026. Even in August 2025, it expected underlying inflation to remain at 2.6% through to the end of 2026.
These forecasts consistently underestimated inflation. Rather than continuing to fall as the Bank expected, inflation has accelerated again. We are experiencing a second wave that the Bank’s own models failed to anticipate.
What history tells us about persistent inflation
The research I cited in 2023, by Rob Arnott and Omid Shakernia at Research Affiliates, examined 52 episodes since 1970 across 14 OECD economies where inflation exceeded 4%. Once inflation rose above 8%, as it did in Australia, it went on to exceed 10% around 70% of the time. The median time it took to fall back below 3% was nearly 11 years.
With better research tools now available, I wanted to test how robust those findings were. The headline figures hold up, but there are some important qualifications. Most of the 52 episodes occurred before central banks adopted inflation targets. The long recovery periods therefore partly reflect the policy settings at the time, rather than simply how difficult inflation is to control. Also, these economies experienced the same global oil shocks in 1974 and 1980. That means the sample captures several countries responding to common shocks, rather than 52 independent events.
More recent research by Ben Bernanke and Olivier Blanchard examined pandemic-era inflation across 11 economies. They found that supply shocks did not become entrenched as they had in the 1970s, largely because central banks now have credible inflation targets. This suggests inflation could be brought under control more quickly than the Research Affiliates’ median of nearly 11 years implies.
However, both studies point to the same broader conclusion: major inflation episodes have several causes that reinforce each other. In the 1970s, oil shocks combined with loose monetary policy, increased US government spending on the Vietnam War and Great Society programmes, and a cycle of rising wages and prices. Together, these forces made inflation harder to control.
Several factors are also contributing to inflation today. Understanding how they interact is essential to assessing how long inflation might persist.
The 3 forces keeping inflation elevated
Firstly, the Reserve Bank raised interest rates less aggressively than comparable central banks. Australia’s cash rate peaked at 4.35%, compared with 5.25% to 5.50% in the US, 5.50% in New Zealand and 5.25% in the UK. The Reserve Bank’s own model estimates put the neutral rate as a wide, uncertain range, with the current cash rate sitting at or just above the top of that range. On that basis, the current cash rate of 4.35% appears only modestly restrictive.
The International Monetary Fund’s staff, in their September 2026 Article IV mission, urged the Reserve Bank to maintain a tightening bias and “stand ready to hike rates as needed.”
The Reserve Bank’s own explanation is that Australia has a lower neutral rate than these countries, and that Australian households are more exposed to variable-rate mortgages, so a smaller rate rise here has a similar effect to a larger one overseas. That reasoning anticipates the argument I make later in this article. If higher rates affect Australian households more directly, on average, than households in the US or UK, that points to a difference in how rate rises affect the economy, not simply how far the Reserve Bank chose to raise them.
Secondly, government spending is making the Reserve Bank’s job harder, although the extent matters. Commonwealth spending is around 26.6% of GDP, compared with a historical average of 24.5%, and the 2026–27 Budget has been described as neutral to mildly expansionary.
Australia’s underlying deficit is around 1% of GDP, substantially smaller than the US at around 6%, the UK at 4% to 5%, and the euro area at 3%. These governments are also spending in ways that work against their central banks’ efforts to control inflation. Australia is experiencing the same tension, albeit on a smaller scale.
Thirdly, the AI infrastructure boom is adding to inflationary pressure, particularly in the US and increasingly in Australia. In March 2026, former Federal Reserve Chair Jerome Powell said data-centre construction “is actually probably pushing inflation up at the margin”. Goldman Sachs has found that US electricity prices are rising by nearly 7% a year, compared with headline inflation below 3%, with data-centre demand a substantial contributor.
In 2024, Australia ranked second globally for data-centre investment, behind only the United States, according to Knight Frank’s Global Data Centres Report. CommBank projects around $150 billion of construction by 2030. These projects compete with government infrastructure projects for the same construction workers and materials. That illustrates how the pressures reinforce each other: government spending and private investment can combine to push up costs, making it difficult to isolate one dominant cause of inflation.
Why rate rises may have less impact
I think this next point matters most, and it is largely separate from the causes outlined above. Even if the Reserve Bank had raised rates as aggressively as its peers, there is good reason to believe each rate rise would have slowed the economy less than it would have 20 years ago.
A smaller share of the population is exposed to higher mortgage rates, while a growing share benefits from higher interest rates on savings. Australians aged 55 and over accounted for around 20% of the population in 1985. By 2025, that had risen to almost 29%. That is a substantial demographic shift.
The Reserve Bank’s own research helps explain why this matters. Households aged 65 and over hold just 18% as much debt as the average household, but 29% more deposits. By comparison, households aged 35 to 44 hold 73% more debt than the average household, but only 71% as much in deposits. Higher rates therefore affect these groups very differently.
The gap shows up in wealth accumulation too. Between 2004 and 2015–16, the average wealth of households aged 55 and over rose by around $630,000, compared with around $90,000 for households aged 15 to 34.
The Reserve Bank’s January 2025 Bulletin explains the difference clearly. Rate rises have the greatest impact on the cash flow of households aged 30 to 54. Older households with substantial savings and little debt can benefit from higher interest income.
The RBA’s January 2025 study estimated that a one percentage point increase in the cash rate reduces total household disposable income by around 0.2%. This is because the increase in borrowers’ interest payments outweighs the extra interest income savers receive.
The effect on spending also depends on who gains and who loses. Research cited by the RBA suggests borrowers spend roughly three times as much of each additional dollar of income on durable goods, such as cars and household appliances, as savers do. Consequently, the additional spending by savers is unlikely to offset the reduction in spending by borrowers.
CBA’s spending data illustrates this divide. Spending by people aged over 65 grew by 10.1% in the year to June 2026, the fastest of any age group. By comparison, spending among those aged 25 to 54, the age group most likely to have a mortgage, grew by barely 4%. Over roughly a decade, self-funded retirees have also increased from 30% to 44% of the population aged 65 and over.
A growing share of spending therefore comes from households that are less directly affected by higher mortgage rates. Some receive more income when rates rise. Lisa Denny at the University of Tasmania made a similar argument in December 2025.
Why mortgage buffers keep growing
There is another factor to consider: the financial position of mortgage holders themselves. Mortgage offset balances have risen by 49% since rate increases began in 2022, reaching around $340 billion. This is a different pattern to household savings generally. The overall saving ratio spiked to a record 23.6% in June 2020, fell to a 16-year low of 0.9% in March 2024, and has since recovered to around 6% to 7%, roughly its pre-pandemic norm. The one-off pandemic savings buffer has largely been spent. Offset balances have not followed that pattern. They have kept rising throughout the entire tightening cycle. That might seem surprising given the pressure of higher repayments. However, rising total balances do not mean every borrower is saving more. Some households can accumulate savings while others draw theirs down.
Firstly, e61 research found that fewer than 7% of variable-rate borrowers were effectively living pay to pay in the second half of 2022. Most had at least some readily accessible savings to help absorb higher mortgage repayments. Its analysis of bank transactions found that borrowers funded around 70% of the additional repayments through savings in offset and redraw accounts, rather than by cutting their spending.
Secondly, wage growth has helped borrowers manage higher repayments, with wages rising by 3.2% in the year to June 2026. The revised income tax cuts introduced in July 2024 provided additional support, worth around $32 to $42 a week for someone earning $80,000 to $100,000. However, the renewed rise in inflation has eroded the purchasing power of these gains.
These wage figures only tell us part of the story. Judging on the increase in mortgage offset balances, clearly some borrowers have received substantially larger pay increases, while others also earn income from businesses, rent and investments. For some households, total income may therefore be growing comfortably faster than living costs and mortgage repayments, allowing them to keep adding to their offset accounts. This may help explain why total offset balances have continued to rise, even while other borrowers are under considerable financial pressure.
Thirdly, lending standards provide some protection. From late 2021, APRA increased the serviceability buffer for new bank lending from 2.5 to 3 percentage points above the loan rate. This requires banks to assess whether borrowers could meet repayments at a higher interest rate. Borrowers taking out loans after rates increased have also generally been assessed at higher rates than those borrowing during the pandemic.
These savings buffers can delay the impact of higher rates on spending. Their effectiveness depends on how much households have saved, how quickly they draw those savings down, and whether their incomes allow them to replenish them.
In my view, this helps explain why higher rates may take longer to slow spending. Many borrowers have savings to draw on before they need to cut back, while some older households receive more income when rates rise. The impact depends on both the debt households owe and the assets they hold.
How long could interest rates stay high?
For most investors, the most useful question is how long inflation and interest rates are likely to remain elevated. The ASX RBA Rate Tracker (as at 18 September 2026) implies a cash rate approaching 4.85% by March 2027.
I think we need to see two consecutive quarters of trimmed-mean inflation at or below 0.6% before becoming more confident that inflation is coming under control. Conversely, annual underlying inflation above 3.75%, or rising inflation expectations, would strengthen the case for further increases and likely push the cash rate above 4.60%.
There is another implication that I think deserves more attention. If changes in household finances have reduced the impact of rate rises, they could also reduce the impact of future cuts.
We saw evidence of this in 2025, when only around 10% of borrowers requested lower repayments after rates fell. Most kept their repayments unchanged and used the interest savings to rebuild their buffers.
For that reason, I would not expect rate cuts to produce an immediate rebound in spending. Many borrowers may continue saving the benefit, while the households currently driving spending growth have little or no mortgage debt to begin with.
Does your strategy rely on rate cuts?
I would not plan for a significant reduction in interest rates anytime soon. Based on the assumptions above, a more realistic base case is a cash rate that remains between 4% and 5% for an extended period. I would also allow for future rate cuts to be more gradual and modest than in previous cycles. Borrowers may use the interest savings to rebuild their financial buffers before increasing their spending, delaying the economic benefit.
If your investment strategy relies on rate cuts to make the numbers work, whether you are borrowing to invest in property or shares, I would stress-test it against this scenario. Make sure you can sustain the strategy if rates stay higher for longer and any reductions are smaller and slower than you expect.
