Unit labour costs combine wages and productivity into a single measure, and the RBA's September minutes say their growth is running too fast for the inflation target, for a reason that has little to do with big pay rises.

Ask most people what drives inflation and they will mention petrol, groceries or rent. Ask a central banker and the answer may well include a less familiar measure: unit labour costs. The minutes of the RBA's September 2026 Monetary Policy Board meeting state that unit labour cost growth is above what is consistent with the inflation target, and that this is driven mainly by subdued productivity growth rather than strong nominal wages.

That single observation helps explain why the Board raised the cash rate target to 4.60 per cent, and why there are limits to what interest rates alone can achieve.

What unit labour costs measure

Unit labour costs are the labour costs a business incurs to produce one unit of output. They bring together two things:

  • How much businesses pay for each hour of work, including wages and other labour costs.
  • How much output each hour of work produces, which is labour productivity.

As a rough rule of thumb, growth in unit labour costs is approximately equal to growth in labour costs per hour minus growth in productivity. If pay per hour rises and productivity rises by the same amount, the labour cost of each unit of output stays roughly unchanged. If pay rises and productivity does not, each unit of output costs more to produce.

Businesses facing rising unit labour costs have a choice: absorb them through lower margins, or pass them on through higher prices. When that happens across the economy, it feeds into inflation, particularly in labour-intensive services.

An illustration

The following illustration uses hypothetical numbers. It is not drawn from RBA data and is not an RBA estimate.

Suppose labour costs per hour rise by 3.5 per cent over a year.

  • If productivity grows by 1 per cent, unit labour costs rise by roughly 2.5 per cent.
  • If productivity is flat, unit labour costs rise by roughly 3.5 per cent.
  • If productivity falls by 0.5 per cent, unit labour costs rise by roughly 4 per cent.

The pay rise is identical in each case. The only difference is productivity, yet it moves the cost pressure from something broadly compatible with a 2–3 per cent inflation target to something well above it.

It is also worth being clear about what unit labour costs are not. They are not the same as wages, and a rise in unit labour costs does not by itself mean workers are being paid generously. A business could pay its staff more each year and still see unit labour costs fall, if each hour of work were producing enough extra output. Equally, unit labour costs can rise quickly even when pay increases are modest, if output per hour is stagnant or falling. That is why the measure is so useful to a central bank: it shows whether the combination of pay and productivity is putting upward pressure on prices.

Why the RBA is concerned

The RBA's August 2026 forecasts show labour productivity falling 0.5 per cent over the year to June 2026, and again by 0.5 per cent over the year to December 2026. Modest growth is forecast after that: 0.4 per cent by June 2027, 0.6 per cent by December 2027, and 0.7 per cent through 2028.

On the wages side, the forecasts show the Wage Price Index growing at 3.3 per cent through to mid-2027, then easing to 2.9 per cent by the end of 2028. Nominal average earnings per hour in the non-farm sector were 2.7 per cent in June 2026, forecast to rise to 3.7 per cent by mid-2027 before easing to 3.2 per cent by the end of 2028.

Put together, these figures are consistent with the minutes' conclusion. Wages are not surging, but with productivity going backwards, the labour cost of each unit of output is rising faster than is consistent with inflation returning to target.

Governor Michele Bullock put it in stark terms at her media conference on 29 September: "The bottom line is that productivity is doing nothing." She also said: "Productivity growth remains weak and the labour market remains a little tight."

The minutes suggest a link between these costs and the type of inflation the Board is seeing. July inflation data were stronger than expected, particularly in market services, and the share of items with prices rising by more than 2½ per cent was "well above average". Staff judged the upside surprise more likely reflected domestic capacity pressures than the pass-through of Middle East cost shocks. Headline CPI was 4.0 per cent over the year to August 2026, while the trimmed mean measure of underlying inflation was 3.6 per cent in the June quarter.

The limits of interest rates

Unit labour costs pose a particular challenge for monetary policy. Higher interest rates work mainly by cooling demand, which can ease pressure on wages and prices over time. But they do not directly raise output per hour.

Bullock acknowledged this plainly: "monetary policy can't do anything about productivity." In the meantime, the Board has acted on the demand side. In September it decided unanimously to raise the cash rate target by 25 basis points to 4.60 per cent. Bullock said: "Inflation is too high and has been driven by domestic capacity pressures."

The Board's statement said it "remains focused on ensuring that high inflation does not become embedded." For unit labour costs, that means the Board is watching both sides of the equation: wages, which it can influence indirectly through demand, and productivity, which it cannot.

The bottom line

Unit labour costs capture the interaction between pay and productivity. According to the RBA's September minutes, their growth is above what is consistent with the inflation target, mainly because productivity is weak rather than because wages are rising strongly. The August forecasts expect productivity growth to return from 2027, which would ease that pressure, but the Governor has said monetary policy "can't do anything about productivity." For anyone trying to follow the Board's reasoning on rates, unit labour costs are a measure worth understanding.