In a major about turn, the Australian Bureau of Statistics will today restate the jobless data for July and August when it reveals the September employment data. And the September figures will be presented unadjusted for seasonal factors, moves that have stunned economists in government and in the private sector.
The Bureau said yesterday it will make is changes to employment figures for the past three months that, on the face of it, would show far fewer jobs created in August than the shock 121,000 part time jobs that were reported.
The ABS said yesterday that it had "concluded that the seasonal pattern previously evident for the July, August and September labour force estimates is not apparent in 2014. This assessment was made while preparing labour force estimates for September 2014 and relates to all seasonally adjusted labour force estimates other than the aggregate monthly hours worked series.
"As there is little evidence of seasonality in the July, August and September months for 2014, the ABS has decided that for these months the seasonal factors will be set to one (reflecting no seasonality).
"This means the seasonally adjusted estimates (other than for the aggregate monthly hours worked series) for these months will be the same as the original series and this will result in revisions to the previously published July and August seasonally adjusted estimates,” the Bureau said in the statement.
The surprise announcement has confused economists and markets. The monthly jobs data figures are the most important of all the Bureau’s releases because they are the most timely official look at the way the economy and the important jobs market is travelling.
The figures are produced within two weeks of the end of the previous month. All other statistics are up to five to six weeks, or a quarter or more behind, depending on their frequency.
And this is not some academic argument over statistics – unemployment is the slow burning fuse of the economy. It’s been slowly rising now for more than a year and is forecast to peak around 6.25%, or perhaps a bit higher. Billions of dollars of spending, tax revenues and other money is riding on the accuracy of this series of data.
Accurate jobs figures are especially vital now with the Federal Government and its departments preparing the important mid year budget review, due for release in late November or early December.
If the ABS figures are unreliable, no one will really understand the labour market until well after the jobless rate peaks. That could cost governments and business huge amounts of wasted spending and other costs. It could mean tax receipts are either overestimated, or underestimated, leading to strains with the budget.
Or the data could understate the problems in the jobs market and lead to a longer and more sustained period of welfare payments for tens of thousands of people.
ABS confesses to problems with jobs data
.png)
Economists had been looking for a substantial readjustment in the employment figures for September after the surprise rise in August. Some estimates had put the size of the fall forecast by the market up 30,000 – others had claimed there would be a much smaller drop, but a reduction nevertheless because of the too high nature of the August figure.
The ABS said in yesterday’s statement that that while compiling the September data it found that the normal seasonal pattern evident from July to September was not apparent, so it had decided to report the raw numbers instead.
The original series showed employment rose by 32,100 in August, rather than by 121,000. For July, original employment fell by 11,900 while the adjusted number dipped 4,100.
“The ABS will commission a review with independent external input to develop an appropriate method for seasonally adjusting October 2014 and following months’ estimates,” Australian Statistician Jonathan Palmer said in the statement. "The report on the results of this review will be presented in due course,” he said.
The changes should also see major revisions to the unemployment rate. The seasonally adjusted series showed the jobless rate jumped to 6.4% in July from 6.1% in June, only to drop back to 6.1% in August.
But the original numbers put the jobless rate at 6% for both July and August, so no real change.
In the statement the Bureau said it had “not made this decision lightly and believes this approach will result in a more meaningful set of seasonally adjusted estimates.
“The ABS will continue to produce trend estimates and, as always, encourages users to use the trend estimates to help understand underlying movements in the labour force series,” the Bureau said.
But economists say the reworking of the July and August figures and those to be released today for September, will mean a whole quarter’s data will be rendered unusable and unreliable, along with the data for the 12 months ending July, August and September, unless the ABS again restates them to produce new seasonally adjusted estimated figures after the promised review.
This will not endear the ABS to the Reserve Bank, Federal Treasury and a host of private companies and organisations because it has (and not for the first time) introduced a degree of unreliability into the ABS’s figures.
