I take exception to the claims in the Sydney Morning Herald that the Australian Bureau of Statistics has blundered on its jobless figures and that we have somehow got the official employment figures wrong (P5, Tuesday 3 July Bureau of Statistics admits blundering on jobless figures and P10 4 July 2012 Markets malfunction when fog closes in)
This is simply not the case. The ABS labour force survey is the best available, using sound statistical methodology to ensure we produce an accurate picture of employment and unemployment in Australia.
Benchmarking against the Australian population is an ongoing area of focus because of its importance in ensuring an accurate estimate. As population estimates can fluctuate, the ABS uses the best available data it has at the time to produce the most accurate and relevant information possible. Any revisions we might make, as a result of changes to the populations, are minor and unlikely to change the headline figure of unemployment. I reiterate that this is neither a blunder nor an error: this is the ABS using statistically sound methodology to provide decision makers with the most accurate, up-to-date and complete information.
Retrospective analysis postulating that there may have been 30,000 more jobs in the first part of 2011 may be of theoretical interest to some, but holds little value for decision makers twelve months later. The timeliness of the ABS labour force data is essential for most data users. In addition, the differences suggested by this type of analysis are well within the margins of statistical tolerance that we expect in conducting a sample survey.
Your articles have suggested that we will not revise history more frequently than five-yearly, as is our current practice. There is presently no compelling case for doing this, as there is unlikely to be any change of statistical significance, and it will make little difference to decisions that have already been taken based on the high quality statistics the ABS has already published.
The ABS is one of the most respected national statistical organisations in the world, providing key indicators on the economic, social and environmental health of Australia. If we get something wrong, we are open and transparent and correct the record as quickly as possible. In this case, it is the Sydney Morning Herald that has got it wrong, by suggesting our official employment figures are wrong.
Brian Pink
Australian Statistician
Australian Bureau of Statistics
Canberra ACT
ABS responds to "Bureau of Statistics admits blundering on jobless figures” (Sydney Morning Herald, 3 July 2012, page 5) - Additional letter to the Editor