Lambasting the reliability of the Australian Bureau of Statistics (ABS) jobs report is a favourite pastime of financial markets. It’s incredibly volatile, defying those poor souls trying to forecast its wild monthly swings. It’s been called a lottery, and, as seen in the tweet below following a reported employment increase of 121,000 in August last year, sometimes far worse.
Absolute joke #ausbiz
— David Scutt (@David_Scutt) September 11, 2014
In my former life trying to make sense of the data, let alone trade it, was fraught with danger.
While the 121,000 increase was eventually revised lower, as the chart below reveals, there have been some big swings in recent years.
While the monthly swings have calmed a little recently — a welcome development given the importance of having reliable data on the economy to help guide fiscal and monetary policy — there are many who still don’t trust the figures presented each month.
Those who doubt the reliability of the data will be crowing from the treetops this morning following news that the former head of the ABS believes employment figures are “not worth the paper they’re written on.”
In an interview with the AFR, Bill McLennan, former head of the bureau between 1995 and 2000 and one of the people responsible for establishing the labour force survey in the 1960s, gave a damning indictment of the numbers reported by the ABS over the past six months.
“The results of the last six months aren’t worth the paper they’re written on, so why are we wasting millions of taxpayers’ money on the survey?,” he said.
McLennan also suggested that, given concerns over the reliability of the data, the survey be dumped immediately for up to five years.
He also slammed the the bureau’s decision to change its labour force questionnaire, something that was applied from the October 2014 survey onward following an independent technical review, without testing the changes on a parallel survey using previous methodology to gauge the impact of these changes on the data.
“My experience tells me that when you make changes like that, you get different responses,” McLennan said.
“What they’re doing now is breaking all the principles under which the survey was done.”
“The impact of the various seasonal breaks needs to be measured separately, but there don’t seem to be any plans to do so. The implication from this is that the ABS will continue to have difficulties producing believable seasonally adjusted estimates of the labour force data.”
In June 2014 the bureau acknowledged it had been told to reduce expenditure by about $50 million over three years, a result of government budget cutbacks to help address a mounting fiscal deficit.
According to McLennan, on top of cuts to the bureau’s budgets under previous governments, the reduced funding didn’t allow for testing of the new survey methodology.
In response to McLennan’s criticism, current ABS head David Kalisch told the AFR that he stood by the “overall quality of the numbers.”
Kalisch stated that the bureau had no intention of scrapping the supplementary labour force survey as McLennan suggests.
“We don’t believe there is a major problem with our monthly adjustments now, so why would we induce a problem for the next three to five years?,” he said.
The ABS will release its September labour force survey at 11.30am AEDT today. Economists expect employment growth of 7,100 with unemployment remaining steady at 6.2%. If there’s one thing that’s nearly certain, based on recent volatility in the data, it’s unlikely to come in around market consensus.
You can read more from the AFR here.