Data revisions have partly undercut the RBA’s narrative of lower aggregate supply than expected. The RBA will need to skate to where the labour market is going to go, not where it used to be.
Recent events have not materially changed our view about the outlook for the Australian economy. Recent communication from the RBA has, however, changed our minds about how the RBA is seeing things and how it will respond to the data. While it is still possible that the RBA Board will change its mind, RBNZ-style, and pivot sooner than our current base-case expectation, we suspect that fast backflips are not in the RBA’s breakdancing repertoire.
Recall our earlier observation that the RBA had concluded that the labour market was tighter than it had previously thought, even though all bar one of the indicators in its labour market dashboard eased between May and August.
Deputy Governor Hauser’s speech earlier in the week addressed this to some extent. The upshot of the speech is that, because inflation has recently overshot the RBA’s earlier forecasts, it must be that demand is stronger than it thought, or supply is weaker, or a combination of the two. As Westpac Economics colleague Senior Economist Pat Bustamante and I noted earlier this week (PDF 427KB), though, the net surprise on the June quarter 2024 trimmed mean inflation forecasts since the RBA’s November forecast round was in fact roughly zero. In some respects, the RBA is seeking to explain the forecast miss for the September quarter 2023. We are now halfway through the September quarter 2024.
We are reminded of the famous quote from ice hockey legend Wayne Gretsky, to ‘skate to where the puck is going, not to where it has been’. Inferring a lower level of aggregate supply currently from a forecast miss that mainly happened a year ago feels a bit like skating to where the puck used to be quite a while ago.
Yesterday’s labour force results highlighted how quickly that puck can move. As Westpac Economics colleague Economist Ryan Wells reported, the labour force participation rate in July was the highest recorded in more than a century. This represents a significant boost to labour supply.
In addition, following up on a key point of Pat’s and my note, the re-benchmarking of quarterly hours worked reported in the July labour force release also works against the RBA’s thesis of weak productivity dragging on supply. Recall that the RBA revised down its forecast for year-ended productivity growth in the June quarter by a full percentage point between its May and August forecast rounds. As far as we can tell, this was largely because of an outsized reading for June quarter hours worked.
The re-benchmarking changed the story considerably. Growth in total hours worked was previously reported as 0.4% over the year to the June quarter (0.2% for non-farm hours, which is more relevant to the measure of productivity in the RBA forecasts). Each of the quarters in that year has been revised down at least a little. Hours worked are now reported to have been essentially flat compared with a year previously (and –0.2% for non-farm).
This 0.4 percentage point revision in hours worked over the year eliminates a considerable fraction of the rationale for the RBA’s downward revision to its forecast for productivity growth. Even a small upside surprise on June quarter GDP growth, or revisions to recent history, would close the gap further.
Flat hours worked in the face of a strong rise in the number of people participating in the workforce does look a lot like growth in labour supply outstripping labour demand and building up some spare capacity.
It is no surprise that labour supply has been so strong. With the cost of living having risen so much, people need the extra work to earn more money. A separate (and more lagged) ABS release, the Labour Accounts, shows that the share of people with a second job has risen to new highs. This is an example of what economists call an ‘income effect’, where labour supply rises when real hourly wages fall. It stands in contrast to the ‘substitution effect’ of people working more when their hourly wage rises, because working is then more attractive relative to leisure.
If strong labour supply does indeed reflect people seeking more income, we might also see unusual labour market dynamics as this effect unwinds. Normally when labour demand softens, we see some combination of higher unemployment and higher underemployment. Over recent decades, the underemployment adjustment has become stronger relative to that in unemployment. Employers are increasingly using the ‘hours margin’ to adjust their labour demand, rather than laying people off entirely. The RBA knows this, and indeed the key paper describing this effect was co-authored by the newly appointed head of the RBA’s Economic Analysis department.
If the income effect has been the dominant driver of the recent rise in labour supply, then as inflation subsides and real incomes recover, we may see an ongoing softening in hours worked, with neither unemployment nor underemployment rising very much. If people no longer need the extra hours or the second job, they will not report themselves as seeking more hours, and so underemployed.
The implication is that it will be all too easy to misinterpret an easing labour market as still tight. In principle, the RBA’s full employment checklist should help guard against such misinterpretation. But seeing how one quarterly outcome for hours worked seems to have changed the RBA’s view on supply capacity, and so the inflation outlook, it is hard to be confident of this.
The RBA will need to skate to where the labour market is going to go, not where it used to be.
