Just how healthy is the Australian economy at the moment? In particular, how is our unemployment trajectory shaping up?

While the media might have switched to the view that the recovery is well under way and we’re only a couple of quarters away from another mining boom, the news on the job front is mixed — although rather better than a couple of months ago.

For a number of years Gary Morgan has been deeply concerned about the quality of ABS employment data and has waged a more or less lone battle to get a more comprehensive and accurate measurement of unemployment into the national debate.

Morgan has a range of concerns about the ABS unemployment numbers — youth and disability allowance classifications are used to hide people who want employment; the ABS criteria for unemployment is too stringent and don’t count those discouraged from looking for work, and, critically, underemployment isn’t measured appropriately, artificially enlarging the number of employed people based on people working even a small number of hours per week, despite wanting to work more or full-time.

The issue is important because underemployment as much as unemployment affects disposable income, consumer demand, and capacity to repay debt, especially mortgages.

Morgan put his money where his mouth is and has been spending money on underemployment surveys to supplement his unemployment surveys a couple of years ago. His estimate of the rate of unemployment (Morgan doesn’t seasonally-adjust his figures) is consistently 2-3% higher than that of the ABS, based on face-to-face interviews.

Morgan’s most recent data confirms that the great success of the Government’s stimulus — the lower-than-expected rise in unemployment — masks a significant rise in underemployment, particularly in June and July, although the deterioration pleasingly appears to have halted in August.

The ABS bought into the debate in a big way in August when it finally added an aggregate monthly hours worked measure — both trend and seasonally adjusted — to its labour force data. That finally gives us a direct indication of its estimate of how much the nation is working — although why the ABS held out on us for so long on what should be a crucial measure is anyone’s guess.

The ABS trend data for the last couple of years shows both the resilience of the Australian economy — we’re still operating at 2007 levels — and also a moderating, but still downward, trend in hours worked, suggesting we’ve got some months to go yet before we collectively start working more hours rather than less.

The longer term ABS data gives a nice snapshot of the Australian economy over recent decades.

The savage and prolonged impact of the early nineties recession is clearer than ever – it took several years for the country to start working as much as it was at the end of the 1980s. The pre-GST surge and the mini-recession afterward looks, in retrospect, more damaging than it felt at the time, and the huge increases in 2006 and 2007 show an economy running at capacity for the first time in decades.

Aggregate monthly hours worked also reflects population growth, of course, so I wondered what the hours worked per worker looked like. I couldn’t find a long-term working-age population number from the ABS but used the Australian population as an unsatisfactory proxy for it to provide a figure for the aggregate hours worked each month per Australian resident.

While flawed, the trend confirms the impact of the early nineties recession but also shows just how gangbusters the place was going in 2007 and into 2008 — and how precipitate the drop-off has been since the onset of the global recession.

We clearly haven’t seen the worst of unemployment yet — but we now set ourselves different standards. Even in the middle of a global recession we’re still running at levels that we could only have dreamed of in the 1990s.

*Listen to yesterday’s Canberra Calling, “Money, Della Bosca and stimulus“