The Reserve Bank’s struggle over whether — and when — to cut interest rates has thrown up an interesting divide over jobs figures. The recent April employment data from the Australian Bureau of Statistics was leapt upon by the commentariat and economists to warn about a softening labour market, given the increase in the seasonally adjusted unemployment rate to 5.2% and the underemployment rate.
Except, the ABS prefers that people look at its trend series, not the seasonally adjusted series; that showed unemployment stable at a revised 5.1%. The trend estimate is what features on the ABS’s website dashboard of economic indicators and for some years it has been what the ABS emphasised in its media releases, ever since problems with its jobs sample began playing hell with the seasonally adjusted figures in mid-2014.
“The ABS produces trend estimates to provide a more reliable indicator of the underlying behaviour of the Labour Force series,” the ABS says. “Trend estimates were introduced into the Labour Force series in the mid 1980s and are available back to February 1978. Trend estimates are considered the best indicators of the underlying behaviour in the labour market.”
But even the Governor of the Reserve Bank, Philip Lowe, preferred to cite the rise in seasonally adjusted unemployment (to 5.2%) this week as he signalled a rate cut at the June 4 meeting of the RBA board, saying “some labour market indicators have softened a little: the unemployment rate ticked up to 5.2% in April; the underemployment rate has also moved a little higher as there are more part-time workers who are seeking additional hours; job advertisements have declined; and hiring intentions have come off their earlier highs.”
So, some tomato/tomayto statistical semantics? Except, the trend data for April were unequivocally strong — the strongest since last August in fact. The jobs market grew an annual rate of 2.5% in August last year, above the 2.0% average for the last 20 years. That slowed to growth of 2.3% in December and in February of this year, but by April annual growth was back up to 2.5% again.
There was another very solid bit of data as well — average hours worked rose 2.8% in the year to April against the long-term average of 1.7% and sharply up from the 1.5% growth in hours worked in December. The difference between trend and seasonally adjusted is significant given the Reserve Bank explicitly said it would be watching the jobs data.
From the RBA’s perspective, however, it may be that whether you use the stable trend data or the rising seasonally adjusted data, jobs growth isn’t strong enough to deliver falling unemployment, which is problematic when inflation is falling as well. And there are some concerns in the labour market: the March quarter construction data that came out on Wednesday showed construction work falling, no matter which way you measure it.
March was the third quarter in a row the value of construction has fallen, driven by falling residential construction — and hidden in the data was a worrying fall in infrastructure spending, with engineering construction slumping nearly 4% in one quarter.
Westpac senior economist Andrew Hanlan said he was surprised by the drop in public works and private infrastructure given the respective works in the pipeline — both the large states and federal government have been pumping up infrastructure spending for a couple of years now.
Construction employs more than 1.1 million workers, and in the past couple of years has provided a private sector back-up to the strong jobs growth provided by health, social care and education. But it’s on the decline. What could be better to stimulate construction jobs than some juice for the housing market via an interest rate cut?
Only, what do we do if, with interest rates at just 1%, we need another hit?