Saturday, May 16, 2009

We'll see more clearly

Australia will once again drive with the lights on.

The Statistician has reversed a decision to hobble the employment figures that has made it hard to tell what's real as Australia enters recession.

In July last year in the lead-up to the downturn the Bureau of Statistics cut the size of its monthly employment survey from around 54,400 people to 41,100. In Victoria the sample was cut from 1,500 people to 8,700.

The resulting figures were derided by market economists as "a waste of space, and "like believing in the tooth fairy"...

"Given the uncertain effects of the crazy cost-cutting shift to a new one-quarter-smaller survey-sample size, the data now are a waste of space for economy-watching purposes," wrote Macquarie economist Rory Robertson at the time.

"Anyone who believes that full-time jobs actually rose by 53,700 and that part-time jobs actually fell by 42,800 in July probably also believes in the tooth fairy," added CommSec economist Craig James.

Even more recently with the changes bedded down the monthly movements have raised eyebrows. This year's April figures released on the eve of the Budget had the number of Australians in jobs climbing 27,300.

In a warning attached to the figures the Bureau said it could only be 95 per cent confident that the true movement was somewhere between a drop of 33,300 and an increase of 87,900.

The Statistician Brian Pink announced yesterday that a result of winning an extra 15 million per annum in the federal Budget the employment survey would be restored to its former size from December.

Also restored from November will be the Bureau's quarterly job vacancies survey, which until it was suspended last May was regarded as more reliable than the private surveys which counted only jobs advertised.

That survey suggested that in in May 2008 there were 3 Victorians unemployed for each vacant Victorian job compared to a national average of 2.6. Since then there have been no statistics available with which to make the calculation.

In an associated cutback last year the Bureau wound back the accuracy of its retail spending survey, only to restore it in November ahead of the December "cash splash".

Amid concern that Australia was weakening its ability to read economic signals at the time they were needed the Department of Finance and the Treasury reviewed the Bureau's base funding in the lead-up to the Budget and recommended an increase.

But some cuts will remain. The Bureau's Statistical Yearbook will be issued only every two years, and the 2011 census will use unchanged questions from those asked in 2006, when a poorly-worded question about volunteering was added at the insistence of the previous Treasurer Peter