*Update 2.20pm: The ABS has provided a response to this article which has been added below.
There is an astonishing qualifier in today’s employment data from the Australian Bureau of Statistics.
In the commentary, the ABS says:
The incoming rotation group in Queensland for September 2016 was considerably different to the rest of the Queensland sample and its influence has been temporarily reduced as part of the estimation process. The data will be further reviewed when October data are available. For more information, please refer to Insights from the original data.
In those three sentences you have the government’s official statistics arm admitting that its sample is terrible and therefore the accuracy of the data cannot be guaranteed. In the field of statistics, where representative samples are the foundation of quality work, this is a remarkable admission of inadequacy, and part of a long-term pattern of the ABS retreating from its confidence in the data.
In fact, it almost reads like a cry for help.
The employment data is the most vital indicator of the health of the nation’s economy. Today the jobs data for September showed:
- The economy lost almost 10,000 jobs last month
- More than 50,000 full-time positions were lost
- At the same time, monthly hours worked in all jobs increased by four million hours to 1.7 billion, even though we lost all of those full-time positions.
Without wanting to overstate the problem here, if this data was to be repeated over several months, the Australian economy would be heading into a crisis. Projections for the welfare bill would be climbing, budget revenue from income taxes would be going backwards, the Reserve Bank would be putting rate cuts firmly on the agenda, and there would be significant political implications.
But the ABS itself is putting asterisks on the data, saying, well, we’re not really sure about this. So everyone has to sit around on their hands for another four weeks to see if there’s some kind of pattern emerging. And then next month there might be more asterisks anyway.
In an age where data is vital to decision making, where prime minister Malcolm Turnbull has put “government should be an exemplar” as a pillar of the government’s innovation agenda, this flimsy, heavily-qualified statistical reporting is a joke.
If the government wants to “coax capital out of its cave”, as Treasurer Scott Morrison said in one of his headland speeches this year, then the uncertainty about what’s really happening in the labour market is running directly against its policy agenda.
Sort it out.
UPDATE: Jacqui Jones, Program Manager at the Labour and Income Branch of the Australian Bureau of Statistics, provided the following comment in response to this article: “To ensure the quality of the latest Labour Force figures, released on 20 October 2016, the ABS made the decision to reduce the influence of the incoming Queensland rotation group as part of the standard estimation method. Data for the incoming Queensland rotation group was considerably different in its labour force characteristics compared with the rest of the Queensland sample, reflecting a higher degree of sampling variability in September 2016. This reduced the influence of 580 households of the 4,600 Queensland sample; or around 2% of the total Labour Force sample of 26,000 households. The ABS will review this one Queensland rotation group when October data are collected and analysed next month.”