The argument for lifting the cash rate on Tuesday is straightforward: trimmed-mean inflation is still above where the Reserve Bank of Australia wants it, and 37 of the 41 economists in Finder's pre-meeting panel expect the board to move the cash rate from 4.35 per cent to 4.60 per cent when it announces at 2:30pm AEST. That case is credible. What is missing is the other half of the decision — a labour-market plan. In August, full-time employment fell by 6,300 while the unemployment rate rose to 4.6 per cent and participation held at 67.1 per cent. If the board hikes into that print without saying how it will read the next one, it is asking households to carry a tightening whose employment costs it has never defined.
What the August data actually says
The headline unemployment number is doing a lot of hiding. A jobless rate that rises while participation stays at 67.1 per cent is not the same animal as a rate that rises because people have given up looking. On its face, the August release is the friendlier version: more Australians inside the labour force, more of them counted as actively searching. That is why some economists will tell you the labour market is still tight and the month is noise.
Then look at composition. Full-time employment fell by 6,300 over the month — small against a workforce of more than 14 million, and real in the way it lands on budgets. Full-time and part-time work are not interchangeable for a household carrying a mortgage. The first pays the bills; the second is the buffer. When the buffer does the hiring, average hours and average pay packets move in the wrong direction at exactly the moment the RBA is trying to slow demand.
The best objection — and its limit
The strongest objection is that the Reserve Bank does not target employment. It targets inflation, and a single monthly full-time print is noisy, revised, and far less informative than the trimmed mean, the participation rate or the broader suite of capacity measures. Governor Michele Bullock's public remarks have given no sign the board has lost its nerve on that mandate, and the major banks have already shifted their calls to a September hike. On this reading, Tuesday is close to a done deal and the employment line is a distraction.
Fair, as far as it goes. But the objection assumes the labour market can be read off the unemployment rate alone. It cannot. Employment composition is part of the inflation outlook, because full-time earnings are what service mortgages and what feed back into services inflation through wages. If full-time work keeps shrinking while headline unemployment looks benign, the board is tightening into a labour market quieter than the aggregate suggests — and it will only find out after the fact. Two hikes at $540 a month on a typical mortgage is also a spending shock that arrives with a lag, which is precisely why a hike now needs a defined off-ramp.
What the board should commit to
If it lifts on Tuesday, the board should say plainly what would stop the next move. Not a vague assurance that it will watch the data, but a stated trigger: for instance, that a second consecutive fall in full-time employment, alongside a rising unemployment rate, would be treated as evidence of spare capacity and would argue against a further increase. Publishing the full-time and part-time split in the post-meeting statement, and explaining how the board weighs it against the trimmed mean, would cost nothing and would settle a question markets are already pricing.
The other actors matter here too. The federal government should resist adding demand-side stimulus between now and the November board meeting; fiscal policy runs through the same channel as the cash rate. Lenders should be ready to show borrowers the arithmetic of another 50 basis points rather than have them discover it in a press release. And the board should remember that its credibility on inflation was earned partly by explaining the trade-offs, not just the target.
A hike on Tuesday may well be the right call. A hike that treats full-time employment as someone else's problem is a call the board has not yet justified.