Friday, June 8, 2012

Those growth figures - they'll be revised down

Wednesday's stunningly high estimate of 1.3 per cent growth in the March quarter is likely to be revised down in future, if past experience is any guide.

Bureau of Statistics data shows that of the past 100 quarterly GDP figures, 18 originally reported growth as being 1.3 per cent or higher. But 12 of those 18 figures were later revised down, most substantially so. Five were revised up, by small amounts, while just one of the 18 ended up as the figure first reported.

The bureau is constantly revising past GDP figures, often changing them dramatically as new data comes to light. But once the spotlight has moved on to the next figures, the public, politicians and economists take no interest. A spectacular example was the 1990-91 recession. The bureau originally estimated that GDP fell 3.2 per cent in the six quarters from April 1990 to September 1991. But now it estimates that GDP fell just 1 per cent in that time. What was said to be the worst recession since the Depression was revised down to be our smallest recession in loss of output - even though it saw the biggest loss of jobs.

The same is true at the top end. Initial estimates showing big GDP growth are usually revised down, on average by 0.4 percentage points.

Four of the past 100 quarterly estimates showed GDP growing by 1.3 per cent. All have since been revised down, on average to 0.9 per cent.

If that happened this time, it would still be a good result, given the strains on the economy revealed by wide-ranging job losses, falling house prices, and weak business and consumer confidence.

But the figures show most of our growth was in two areas - consumer spending and engineering construction - and both rest on very low estimates of inflation.

On these figures, consumer prices rose just 0.1 per cent in the six months to March. In the frenetically busy engineering sector, construction costs rose just 0.1 per cent in the March quarter, and 2.2 per cent in the year.

Despite the anecdotes and mining companies' complaints about the high cost of construction in Australia, the bureau says that in the past 3? years, engineering construction costs have risen just 1.8 per cent or 0.5 per cent a year. That's some contrast to the previous 3 years, when they rose 24.8 per cent.

If correct, that is remarkable, and it sheds a very different light on what a mining boom means for prices.

If it is wrong, then so are the GDP figures, since engineering construction made up almost half the growth in Australia's spending.