What a rate rise buys

On Tuesday 29 September 2026 the Reserve Bank’s Monetary Policy Board raised the cash rate target by 25 basis points to 4.60 per cent, the fourth rise of 2026, and the decision was unanimous ¹. The four rises total 100 basis points from 3.60 per cent, and put the cash rate above the 2023 peak of 4.35 per cent and at its highest since late 2011 ².

A rate rise is a purchase. Borrowers pay for it now, and savers earn more on their savings ⁸; the Bank says the goods arrive later. This article sets out the Bank’s own price list: what its models say a one-point rise takes off inflation and output, what it adds to unemployment, and how long all of that takes. Every figure below is the Bank’s or the ABS’s, or is attributed to its source, unless it is marked as THE RORT’s calculation.

The headline is in the Bank’s April 2025 Bulletin. In most of its models, a 100 basis point rise has its peak effect after around one to two years; across them it lowers the level of GDP by ¼ to 1 per cent and year-ended inflation by ⅛ to ½ of a percentage point ³. In one of its two main macroeconomic models, MARTIN, inflation “does not decline until” the contraction in demand translates into a rise in the unemployment rate and a fall in input costs ⁴. And the Bank has published no estimate of what the 2026 rises will do to unemployment or inflation ⁵⁶.

01How the tool works, in the Bank’s words

The Bank’s own explainer puts the mechanism in two stages: “Changes to monetary policy affect interest rates in the economy. Changes to interest rates affect economic activity and inflation.” Expectations matter too: “if workers expect inflation to increase, they might ask for larger wage increases to keep up with the changes in inflation.” ⁷

The same explainer is candid about timing and confidence: “Some estimates suggest that it takes between one and two years for monetary policy to have its maximum effect. However, there is a large degree of uncertainty about these estimates because the structure of the economy changes over time, and economic conditions vary.” ⁷

Assistant Governor Christopher Kent listed five channels through which a rise works, in October 2023. Of the cash-flow channel, he said: “When interest rates go up, households pay more on their debt and earn more on their savings. Because the cash-flow channel is so noticeable, and felt so keenly by borrowers, it gets a lot of attention.” ⁸

The Bank’s models rate the channels differently from the attention they get: across two of them, MARTIN and DINGO, the exchange rate channel “tends to be very important”, especially for inflation, while the cash-flow channel “is less important in aggregate, despite it having an obvious effect on individual households through changes in mortgage repayments” ³.

02What 100 basis points buys

The Bank’s cross-model estimate is in its Bulletin of April 2025, by Mulqueeney, Ballantyne and Hambur. It used three RBA models that span the different classes, and included an external benchmark labelled “Murphy” ³. Its finding: “most models estimate the peak impact of policy occurs after around one to two years. But the estimates of the peak effect range from ¼ to 1 per cent for GDP, and ⅛ to ½ percentage points for inflation.” ³ The GDP figure is a per cent of the level of GDP, not a percentage point, and the inflation figure is a fall in the year-ended rate. The authors stress that “the entire future path of interest rates matters” ³.

MARTIN is the Bank’s macroeconometric model, one of the two the Bulletin calls its main macroeconomic models ³. In the 2019 paper that documented it, a 100 basis point rise lasting four quarters lowers the level of GDP by around 0.8 per cent six quarters after the shock. “The lower level of economic activity translates into an increase in the unemployment rate, which rises by 0.3 percentage points.” Inflation falls by a little less than 0.2 of a point after two years ⁴. The most interest-sensitive part of spending in MARTIN is dwelling investment, which declines by “slightly more than three per cent” ⁴.

In MARTIN, inflation does not decline until the contraction in demand translates into a rise in the unemployment rate and a fall in input costs.

A second Bank estimate, by Benjamin Beckers in a 2020 paper using data from 1994 to 2018, finds that “prices fall by around 0.7 per cent over the two years following a 100 basis point temporary increase in the cash rate, the unemployment rate peaks after six quarters at around one-third of a percentage point above its previous level and output falls by around 0.8 per cent over the course of one year” ⁹. Its inflation effect is about twice as large as MARTIN’s, while its output and unemployment effects are “closely in line” with MARTIN’s ⁹.

Read together, the Bank’s estimates for a 100 basis point rise are a fall in year-ended inflation of ⅛ to ½ of a point, a fall in the level of GDP of about ¼ to 1 per cent, and, in the two Bank papers that report it, a rise in unemployment of about 0.3 to one-third of a point, from models whose output fall of about 0.8 per cent sits near the top of the cross-model range (THE RORT’s comparison) ³⁴⁹. Of its timing estimates the Bank says there is “a large degree of uncertainty” ⁷.

03What that means in people

The Bank publishes rates, not headcounts. THE RORT’s calculation turns its unemployment estimates into people, on the August 2026 labour force of 15,559,500 (722,900 unemployed and 14,836,600 employed) ¹⁰. MARTIN’s 0.3 of a point is about 46,700 more unemployed people at the peak of a 100 basis point rise; Beckers’ one-third of a point is about 51,900 ⁴⁹¹⁰. The method is a straight multiplication of the labour force by each rate.

Three cautions travel with those numbers. They are per 100 basis points, at the peak, in models of a temporary rise, so they are not an estimate of what the four 2026 rises will do, which the Bank has not published ⁵. THE RORT rates them medium confidence as an order of magnitude. And the Bank warns that the whole path of rates matters, not one step ³.

Scaled down to a single 25 basis point step, the same arithmetic gives about 11,700 to 13,000 people, and the Bank’s GDP range works out at about 0.06 to 0.25 per cent of the level of GDP and its inflation range at about 0.03 to 0.125 of a percentage point ³¹⁰. That is a rough scaling of the Bank’s 100 basis point estimates, and THE RORT rates it low confidence: linear scaling is a simplification the Bank warns against, and its model shocks are not one-off 25 basis point steps.

Beside those numbers, and not as their cause, sits the Bank’s own forecast. In August it forecast unemployment rising from 4.4 per cent in June 2026 to 4.8 per cent by 2028 ⁶. At the August 2026 labour force, on THE RORT’s calculation, that is about 62,000 more unemployed people: each 0.1 of a point is about 15,600 people ¹⁰. From the August count, 722,900 people or 4.6 per cent, reaching 4.8 per cent would mean about 24,000 more (15,559,500 multiplied by 4.8 per cent is 746,856, less 722,900) ¹⁰. Employment is still forecast to grow, so the 24,000 is a floor; both figures are THE RORT’s arithmetic, not a Bank statement.

The forecast rise is not a headcount for the rises. The Bank does not split its forecast by cause ⁶.

04What the Bank has not published

The Bank has published no estimate of what the 29 September rise, or the 2026 cycle, does to unemployment or inflation ⁵⁶. Its May 2026 Statement on Monetary Policy ran adverse scenarios, but its own words are that “the assumed cash rate for Australia in the scenario is the same as the baseline assumption and implies a cumulative tightening of 60 basis points by mid-2028” ⁵. THE RORT searched the August 2026 Statement (PDF) for the word MARTIN and found it zero times ⁶.

What the August Statement does say is: “Monetary policy is judged to be somewhat restrictive, which is expected to help bring the economy into better balance. But that process will take some time, reflecting the normal lags in transmission.” ⁶

The August forecast table has unemployment rising from 4.4 per cent (June 2026) to 4.8 per cent by 2028, while trimmed mean inflation falls from 3.6 to 2.4 per cent and headline inflation from 3.9 to 2.4 per cent. “Inflation is expected to remain elevated in the near term and ease back to target only gradually, reaching 2½ per cent by early 2028.” The forecasts “assume less than one full cash rate increase by the end of 2026, before a small reduction in the cash rate further out” ⁶. Because the Bank does not split its outlook by cause, neither the 0.4 of a point on unemployment nor the fall in inflation can be read as the cost or the proceeds of the rises.

The next Board decision is on Tuesday 3 November 2026 at 2.30 pm, followed by the Governor’s media conference ¹¹. THE RORT will check then whether the Bank has published any such estimate.

05An old worked example

The Bulletin carries a worked example of the trade-off, and it is illustrative and 18 months old. It applies to the Bank’s February 2025 Statement, “and so does not incorporate any information received since then” ³. On the February 2025 market path, the February 2025 Statement had GDP growing by about 2¼ per cent a year on average from the beginning of 2025 to the end of 2026, and inflation was expected to stay above the midpoint of the target range, at about 2.7 per cent at the end of 2026. Under a “hold” path at 4.35 per cent, the models predicted GDP would grow by about 1½ per cent a year, and inflation would decrease to be “around or below the midpoint” of the target range by the end of 2026 ³.

THE RORT’s calculation from those two growth rates is about 0.75 of a percentage point a year of growth given up (2¼ less 1½) for inflation at or below, rather than above, the midpoint. The example compares two paths for the cash rate as they stood in February 2025, before any 2026 rise. It is not an estimate of the trade-off now.

06What the rises cannot buy

The Bank is plain that a rise does not touch the fuel-price shock. In prepared remarks at her 5 May 2026 media conference the Governor, Michele Bullock, said of the fuel and commodity price increases already feeding through to inflation: “The recent increases in interest rates will have no impact on this.” What the increases do, she said, “is to help to contain the domestic inflationary pressures after the inflation due to oil and related commodity prices eases.” ¹²

At the same conference she said the rises “are not going to do anything for inflation in the next six months. That’s done and dusted.” ¹²

The May Minutes say the same in the Board’s voice: monetary policy “could not prevent a near-term increase in the price level as higher fuel prices worked their way through to final prices”, and “could not alter the near-term trajectory of inflation and, additionally, that output growth would likely be lower than potential growth for some time” ¹³. What it could do was “limit the risk that this cost shock resulted in a broader and sustained lift in inflationary pressure, by bringing aggregate demand into closer alignment with aggregate supply and ensuring medium- to longer term inflation expectations remained anchored” ¹³.

Nor can a rise fix productivity. In her 28 July 2026 speech the Governor said: “One thing monetary policy can’t do, however, is address the economy’s slow productivity growth. While this persists, the ability of the economy to grow without generating inflation is constrained, and Australians will continue to experience limited growth in real wages.” ¹⁴ That is a statement about productivity, not an admission of a cost of the rate stance.

None of this is a charge THE RORT makes against the Bank. These are the Bank’s own statements of what a rise cannot reach.

07The Board’s case: the cost of waiting

The Bank’s answer is that the alternative costs more. In November 2022 Governor Lowe said that “bringing inflation back down again after it becomes ingrained in people’s expectations is very costly and almost certainly involves a recession”, and that in Australia and the United States in the 1970s and 1980s it required “a rise in the unemployment rate of at least 5 percentage points” ¹⁵.

Governor Bullock made the same argument on 28 July 2026, in her account of the 1970s oil shocks. Supply shocks, she said, have become more frequent, and credible targets reduce the risk that they persist ¹⁴. She also said that “a given increase in oil prices has a less direct and less pervasive effect on inflation today than it would have in the 1970s” ¹⁴.

Putting off a period of tight monetary policy today can mean higher rates and higher unemployment down the track.

The Bank for International Settlements makes a related argument. Its Annual Economic Report of 28 June 2026 says policymakers “must prioritise price stability” ¹⁶. BIS staff research published on 5 August 2026, which states the authors’ views rather than the BIS’s, finds that “When expectations are above target, the inflationary impact of oil supply shocks can be more than twice as large as when they are well anchored”, and that prompt responses cost less where activity is resilient ¹⁷.

In March the Board’s majority judged that developments in the Middle East “would add to global and domestic inflation under a wide range of scenarios”, that conditions were “not sufficiently restrictive”, and that if medium- and long-term inflation expectations rose, it “would ultimately require significantly more contractionary monetary policy” ¹⁸. The same majority conceded that the war’s effect on demand “remains uncertain, given Australia’s position as a net energy exporter and households’ generally healthy balance sheets” ¹⁸.

The Governor has also defended gradualism. On 11 August 2026 she said that “if we were just focusing on inflation and we weren’t focusing on employment and the economy more broadly, then we could possibly raise interest rates very substantially and induce a very big impact but get inflation down very quickly”. The Bank’s more flexible mandate exists, she said, “to allow us reasonable time to get inflation back down and hopefully avoid some of those costs, particularly on the employment side” ¹⁹.

08The other side of that case

One member of the Board has said, in his own name, that he sees no evidence of a wage-price spiral now. Iain Ross said on 22 September 2026: “The overall thesis is that there is no evidence of the emergence of a wage-price spiral in the present circumstances and recent data suggest such an outcome is unlikely.” He said the 1970s mechanisms, comparative wage justice and quarterly indexation, are gone, and that enterprise agreements lock wages in until they expire. He added: “I am expressing my own views. I am not purporting to speak on behalf of the Monetary Policy Board or the RBA.” ²⁰ The May Minutes frame the risk as “a broader and sustained lift in inflationary pressure”, and the response as keeping medium- to longer term inflation expectations anchored ¹³.

Inside the Board, the March decision to raise to 4.10 per cent was 5 to 4. The minority said inflation was too high and that “a further tightening in monetary policy would probably be required”, but judged there was “merit in delaying any tightening of monetary policy until the potential effects of the current conflict in the Middle East become clearer”, and placed more weight on weaker-than-expected consumption and slowing growth in unit labour costs ¹⁸.

In May the decision to raise to 4.35 per cent was 8 to 1. The lone dissenter judged that capacity pressures before the conflict were “somewhat less than the staff had assessed”, saw a higher risk of a prolonged conflict that sapped demand, expected inflation to return to target without more tightening, and preferred to hold “while awaiting additional evidence on how the Australian economy would respond to the conflict”, noting that this was consistent with other central banks ¹³. The majority said underlying inflation was projected to be above target “for an extended period” across a range of scenarios, and that an increase “would best balance the Board’s two objectives, accepting that the shorter term trade-off between these had worsened” ¹³.

Both dissents, in March and May, were about timing, not the tool: the dissenters argued for waiting, not for a different instrument ¹⁸¹³. The Minutes attribute no votes to individual members, and this article does not say how any member voted.

So what does a rise buy, on the Bank’s own numbers? A fraction of a point off inflation, a year or two on, at the price of a lower level of output and more unemployed people. The Bank’s case is that not paying would cost more later. What it has not published is the sum for the 2026 rises. That is the question this article leaves open.

If it’s a rort, we cover it.