Australia's 3.8% inflation figure is not your business's number

By Claire Bennett · 2026-08-04

Two numbers, neither of them yours

The Australian Bureau of Statistics published its June quarter CPI on 29 July, and the headline landed at 3.8% for the year to June 2026.^[1] That figure runs across every news ticker, turns up in supplier negotiations and gets quoted in staff requests for a pay review. The problem is that it measures what a representative Australian household pays for a basket of 87 goods and services. It was not designed to measure what your business pays to operate.

The Reserve Bank does not focus on that headline figure either. Its preferred guide is the trimmed mean, which came in at 3.6% for the same period.^[1] That gap of 0.2 percentage points matters more than it looks, because the two numbers answer different questions. Understanding which one does what work is the first step to pricing sensibly for the year ahead.

What trimmed mean actually measures

The trimmed mean is the weighted average of the middle 70% of CPI items after the largest positive and negative price changes have been stripped away.^[2] The RBA uses it because it filters out the noise of a single bad season for lettuce or a temporary spike in petrol prices, leaving a cleaner read on persistent, structural price pressure in the economy.

Monetary Policy Board member Ian Harper AO said trimmed mean inflation peaked in the December quarter of 2022.^[6] The trimmed mean has been decelerating since then, which is why the Bank has room to think about rate settings. Decelerating is not the same as low, though, and 3.6% is still well above the RBA's 2-3% target band midpoint.

The RBA has also flagged that, during the current transition to a complete monthly CPI, it will continue focusing on the quarterly trimmed mean series until sufficient monthly data is available.^[3] A monthly CPI release is something the Bank itself treats as provisional context rather than the decisive number.

The components that actually hit businesses

Inside the 3.8% headline, the categories that land hardest on operating costs are moving at very different speeds. Housing costs rose 6.8% over the year, food and non-alcoholic beverages rose 3.3%, and recreation and culture rose 3.3%, the three largest contributors to annual inflation.^[1] For a business with a physical premises in a capital city, the housing component is not a distant statistic about rental tenants. It is your next lease renewal.

Electricity pricing has been shaped by separate federal and state rebate programmes that distort the measured index up and down in ways that bear no relationship to your actual quarterly bill. The national average smooths over wildly different experiences across states, business sizes and tariff structures. None of that variation is visible in the 3.8% figure you might quote in a price letter.

ABS head of prices statistics Rachael McCririck said annual CPI inflation in May was 4.0 per cent, down from 4.2 per cent in the year to April.^[5] The June quarter data confirmed the descent is gradual. Businesses pricing for 2027 should not assume a softer headline number means their own input costs have moved at the same rate.

The wage cost the CPI does not count

The figure probably doing more damage to margins than anything in the CPI release is this: the Wage Price Index rose 3.3% over the twelve months to the March quarter 2026.^[4] Labour costs are tracked by a separate ABS index entirely. They do not appear in the CPI at all, because the CPI measures what households pay for goods and services, not what businesses pay to produce them.

For a service business where wages represent 50% or more of total costs, a 3.3% wage increase is captured nowhere in the 3.8% headline. A tradie whose cost base is 60% labour and 15% fuel faces a very different inflation environment from a manufacturer whose biggest line items are raw materials and energy. Both businesses would be poorly served by citing the national CPI in a pricing decision.

Building a basket that reflects your actual business

The practical step is to construct a weighted inflation basket from your own accounts, not from the ABS household basket. Pull your last full year of costs from your accounting software and sort them into broad categories: housing and occupancy, energy, food and consumables, insurance, transport and fuel, and labour. Calculate what share of total operating costs each category represents.

Then apply the relevant price change to each category. Where your cost base is heavily weighted towards housing or occupancy costs, the relevant rate is closer to 6.8% than to 3.8%. Where labour dominates, you are working with the 3.3% Wage Price Index figure, not the CPI at all.^[4] The weighted result is your business's actual inflation rate for the period.

This approach also gives you something defensible when a customer pushes back on a price increase. Quoting the national CPI is a weak position because a sophisticated buyer will point out that the trimmed mean is lower, that some components have eased and that the headline includes things your business never buys. A weighted cost basket tied to your own accounts is harder to argue with.

What CPI cannot tell you about customer tolerance

Even a perfectly calibrated weighted basket still answers only one question: what did it cost you to produce? Pricing power is a demand-side question, and the CPI is a supply-side measure. Your customers' own cost pressures, their alternatives and their perception of value relative to price all sit completely outside the index.

A business in hospitality or discretionary retail may find its own input inflation is 5% but that passing more than 2% through kills volume. A business supplying something genuinely constrained, or where switching costs are high, may have pricing power well above its actual cost increase. The CPI tells you neither of those things.

The ABS collects prices for thousands of items across 87 categories and 11 groups to arrive at the CPI.^[2] That breadth is exactly what makes it useful for monetary policy and exactly what makes it imprecise for a single firm setting its price list. The 3.8% figure is not wrong. It is just not yours.

Claire Bennett

Claire Bennett writes about work and workplace culture. She is interested in the gap between how organisations describe themselves and what it feels like to work inside them.