Key Highlights

  • The August forecasts assume labour productivity climbs from −0.5% to 0.7% by 2028.
  • The RBA blames weak productivity, not strong pay, for elevated labour costs.
  • The forecasts were finalised before the September rise and will be updated; Bullock’s verdict was that productivity is doing nothing.

The RBA's August forecasts pencil in labour productivity swinging from a 0.5% decline to 0.7% growth by 2028 — a shift that matters because weak productivity, not strong pay, is what the Bank blames for elevated labour costs.

Minus 0.5 per cent. That is the year-ended growth rate for labour productivity that the Reserve Bank of Australia (RBA) showed for June 2026 and forecast again for December 2026 in its August Statement on Monetary Policy (SMP). It means output per hour worked was going backwards.

The same forecast table then shows a turnaround: productivity growth of 0.4% by June 2027, 0.6% by December 2027 and 0.7% in both June and December 2028. Those are modest numbers, but they carry a lot of weight in the RBA's outlook for inflation.

Governor Michele Bullock left no doubt about her view of recent performance when she spoke to the media on 29 September 2026, the day the Monetary Policy Board raised the cash rate target to 4.60%. "The bottom line is that productivity is doing nothing," she said.

Why output per hour feeds into prices

The link between productivity and inflation runs through what economists call unit labour costs: the cost of labour for each unit of output produced. If hourly labour costs rise while output per hour stays flat or falls, each unit of output becomes more expensive to produce. If productivity is growing, some or all of a pay rise can be absorbed without lifting the cost per unit.

The minutes of the 28–29 September meeting spell out where Australia currently sits. Unit labour cost growth was above the level consistent with the inflation target, and this was driven mainly by weak productivity rather than strong wages.

That is a significant distinction. It means the RBA is not pointing the finger at workers' pay claims. Bullock stated directly that "there is no wage price spiral going on in Australia", and in another remark said: "They're seeing real wage cuts." The pressure on costs instead comes from the other side of the equation — the economy has not been getting more output from each hour of work.

What the turnaround looks like on paper

Setting the productivity forecasts beside the RBA's pay forecasts shows why the projected recovery matters. All figures below are year-ended forecasts from the August SMP, which were finalised on 5 August 2026, before the September rate rise.

  • Labour productivity: –0.5% (June 2026), –0.5% (December 2026), 0.4% (June 2027), 0.6% (December 2027), 0.7% (June 2028), 0.7% (December 2028)
  • Wage Price Index: 3.3%, 3.3%, 3.3%, 3.1%, 3.0%, 2.9%
  • Average earnings per hour (non-farm): 2.7%, 3.0%, 3.7%, 3.6%, 3.4%, 3.2%

In 2026, pay measures are growing while productivity is shrinking — the combination that pushes up unit labour costs. By 2028, productivity growth has turned positive while both pay measures have eased from their peaks. On these forecasts, the gap between pay growth and productivity growth narrows over the period.

The broader economy forms the backdrop. The same table has GDP growth at 1.9% in June 2026, slowing to 1.4% in December 2026 and then recovering gradually to 1.8% by December 2028, while population growth eases from 1.4% to 1.2%. The September minutes added that June-quarter GDP was a little stronger than expected, led by private business investment, which was up more than 10% over the year.

This is a simple reading of the published table rather than an RBA calculation of future unit labour costs, and the forecasts themselves are subject to revision.

The inflation path that sits alongside it

The Assumed Productivity Recovery is part of a broader set of forecasts in which inflation returns to the RBA's 2–3% target range. The trimmed mean, the RBA's preferred underlying measure, was 3.6% over the year to the June quarter of 2026. The August forecasts had it at 3.3% by December 2026, 3.0% by June 2027, 2.6% by December 2027 and 2.4% through 2028.

The SMP text said underlying inflation was expected to stay at 3.0% in June 2027 and decline to "the midpoint of the 2–3 per cent target range in 2028". Headline inflation, which was 4.0% over the year to August 2026 on the monthly CPI, was forecast to return to the target range by Mid-2027.

The timing is worth noting: in the forecast table, the return of productivity growth to positive territory in mid-2027 broadly coincides with the trimmed mean dropping to 3.0%. The forecasts present these as parts of one outlook rather than as a stated cause and effect, but the minutes' emphasis on productivity explains why the recovery is relevant to the inflation story.

Why the September decision still went ahead

If a productivity recovery was already built into the forecasts, why did the Board raise rates again in September? The answer in the RBA's statements lies in what happened after the forecasts were prepared.

The post-meeting statement said "some of the upside risks flagged in August are materialising." The minutes record that July's monthly CPI was stronger than expected, particularly for market services, and that developments "pointed to an increase in the risk of persistently elevated inflation." Staff judged the underlying inflation surprise more likely reflected domestic capacity pressures than the pass-through of Middle East-related costs.

Bullock described underlying inflation as running at "about 3.5 per cent for the past six months" — "a percentage point above our target". She also said about 0.6% quarterly inflation is consistent with roughly 2.5% a year, compared with recent quarterly readings of 0.8 and 1.

Against that backdrop, a forecast productivity recovery is not something the Board could rely on to do the work alone. "Will it be enough? I don't know," Bullock said of the September increase.

The bottom line

The RBA's August forecasts assume labour productivity climbs from –0.5% to 0.7% by 2028, a recovery that would ease the unit labour cost pressure the minutes attribute mainly to weak productivity rather than strong wages. Those forecasts were finalised before the September rise and will be updated. For now, Bullock's verdict — "productivity is doing nothing" — describes the starting point. The next CPI data are scheduled for release on 28 October, ahead of the Board's next decision, scheduled for 3 November.

The 0.7% assumption represents a projected medium-term productivity trend, not evidence that the recovery has already occurred.