The Australian economy is in recession – not yet depression! – and will be recessed for at least 2 years.

Monetary and fiscal expansion was too late to avoid this outcome, with a substantial loss of employment (jobs or hours worked) now in the pipeline.  The Rudd Government’s Industrial Relations (IR) changes will, if enacted, make it more costly for small business to hire people, which will further discourage jobs growth.

This article argues that a better understanding of the real state of Australia’s labour market would very likely have led to a different set of policies.
*  Monetary policy would not have been tightened as far as it was, and quite possibly the November 2007 rate hike that put the final nail in the coffin of the Howard Government would not have occurred
*  Fiscal expansion might well have been started by the Howard Government, adding to its chances of winning the 2007 Federal election.
*  The true value of WorkChoices might have been recognised more clearly and, even if Rudd’s Labor Party had won the 2007 election, Industrial Relations ‘Rollback’ might not now be on the agenda.

These points are controversial, or at least arguable.

But the facts of the labour market cannot be dismissed.

In 2006, Henry Thornton wrote about ‘Australia’s Missing Workers’.

Henry had for a long time been aware that the official definition of unemployment as used by the Australian Bureau of Statistics (ABS) was defective. 

He discovered that Roy Morgan used a broader definition.

It was broader mainly due to it’s inclusion of the disenchanted unemployed people who have not looked for work in the past four weeks, as well as those who are unemployed but are unable to begin work in the reference week. Because of it’s inclusion of these groups, it is a more realistic definition of unemployed persons.

However, that was not the end of the story.

There are two other ABS measures that need to be allowed for in any complete enumeration of wasted resources in the Australian economy.

Neither  Roy Morgan nor the ABS estimate includes those people who want to work but are unable due to the high cost of childcare, those disabled people who might be retrained to work in a new field or those people on disability or other pensions for whom the move to paid employment would involve such a high effective tax rate that they cannot be bothered. Unfortunately there are many of such people in Australia; they are forced (or encouraged) either not to work or to work as part of the cash economy!

Another ABS survey comes closer to revealing what these numbers may be.  From 2004/5, the Multi-Purpose Household Survey found that, of the 6.3 million people not employed or who worked only a few hours, around 1.8 million stated that they would like to work more hours, 977,300 of which were not in the labour force.  This substantial number could be deemed the ‘hidden unemployed’.

The ABS also calculates separate figures on those people working less than full-time who would prefer full-time work – the underemployed.  (NB: Look for the splendid diagram if you wish to see how complicated all this is).  As of September 2005, of the 2,839,900 part-time workers in Australia, 612,000 said they would prefer to work more hours.  Although the official ABS definition of underemployed siphons out a further 45,400 (for complicated statistical reasons) for a total of 566,600 underemployed people in Australia, it is clear that underemployment is as extensive as, if not more than, unemployment.

Therefore, adding unemployed, underemployed and the hidden unemployed, the total amount of labour capacity that remains unutilised or underutilised is closer to 15 % than 5 %.

The 2006 article included a table to support this conclusion.

In Henry’s experience, a time series is far more persuasive than estimates at a single moment in time.

The bottom two lines on the graph below compares the ABS and Roy Morgan measures of unemployed persons, expressed as a % of the relevant workforce measures. (The ‘trend’ measures in each case are 3 month moving averages.)

Both measures show a trend decline from January 1993.  The Roy Morgan measure drops faster in the recovery year of 1994, in 2002 and with the introduction of WorkChoices in 2007.

The main value of this measure, however, is the warning it gave about the deterioration in Australia’s Labour market.  This series began to rise, sharply, from October 2007.

Would the Reserve Bank have stayed its hand in November 2007 if it had had this data, which was not available until a December quarter 2007 release?  Only Glenn Stevens can answer this question with complete accuracy, and my guess is that he will not wish to do so.

(To be fair, in November evey other statistic was signaling that the boom was still accelerating, and even Henry was still saying rates needed to be raised further.)

Now the Roy Morgan measure is 8 %, compared with the official ABS number of 5.2 %.  The missing 2.8 % are people who are not counted as unemployed by the ABS, but are no doubt deeply distressed by more urgent problems.

The ABS survey is based on the USA unemployment survey.  This was designed when there were few part-time workers, few contractors, no people working from home using the internet and very few people working the telephone from their home, car or a table in a coffee shop.

Everyone knows that the ‘world of work’ has changed dramatically.  It is time that the ABS changed its measuring rod, but unless and until it does, Roy Morgan’s monthly survey will fill the gap.

More general measures of wasted resources.

To get a better handle on the overall picture, Julian McCrann of Roy Morgan interpolated the annual ABS measures of underemployed persons and discouraged workers and added them to the Roy Morgan measure of unemployed persons.

This produced the dotted lines on the graph, after smoothing by use of a 6 month moving average to indicate the imperfect nature of the data.

Henry then spoke to Gary Morgan, Executive Chairman of Roy Morgan Research.  After some spirited exchanges, he agreed to ask a new question in his regular household survey.

He asked of people who described themselves a ‘Working part time’  “Are you now looking for a full-time job or additional hours?’

This provides a sharper measure of people who are underemployed, and that is shown in the second highest series on the right of the ‘break point’ when we are able to switch to using the new data.

The measure of underemployed persons (as % of relevant workforce) is fairly stable, so the total behaves more or less as the Roy Morgan measure of unemployment.

But the key point is this.  While unemployment in February was 8 % (compared with the ABS measure of 5.2%), the Roy Morgan measure of Unemployment + Underemployment is a staggering 14.3 % of the relevant workforce.

Henry notes in passing that the large number of people unemployed, or working far less than they wish, is a reason, perhaps the reason, that there was no wage breakout during the resource boom.

If we add the (still interpolated) ABS discouraged workers, as in the final line on the graph we have the Henry Thornton Underemployed Labour (including discouraged workers).  In February, this group was, as best we can measure, a shocking 18 % of the relevant workforce.

As noted in the 2006 article, the broader measures are more speculative, and no doubt include some double counting and other inaccuracies.

Concluding comments

The main points I believe will survive any amount of statistical refinement or nit-picking.

1. Unemployment + underemployment in Australia is a far bigger problem than generally believed.
2. The more accurate measures developed by Henry Thornton and Roy Morgan warned of the seriousness of Australia’s economic slump far earlier than the official statistics.
3. Australia’s economic policy, perhaps even its government, would be far different if the more accurate data had been taken seriously by Australia’s power brokers.

Ed, 6/7: We are not alone.  Ambrose Evans-Pritchard says:

'The Centre for Labour Market Studies (CLMS) in Boston says US unemployment is now 18.2pc, counting the old-fashioned way. The reason why this does not "feel" like the 1930s is that we tend to compress the chronology of the Depression. It takes time for people to deplete their savings and sink into destitution. Perhaps our greater cushion of wealth today will prevent another Grapes of Wrath, but 20m US homeowners are already in negative equity (zillow.com data). Evictions are running at a terrifying pace.'

Pollster Gary Langer adds to the debate as US unemployment reaches 9.8 %, 2/10.

'The employment numbers released today underscore what an ugly time it is for the American workforce – a reality that, as our polling shows, resonates beyond the economy to the health care debate, politics and public health alike'. 
 
'September’s unemployment, 9.8 percent, is at a level unseen since the 1981-82 recession; it peaked at 10.8 percent in November ’82. We haven’t seen these levels previously in a federal data series extending to 1948, and thus very likely since the Great Depression.

'Add in discouraged ex-workers – those no longer looking for a job – and it’s worse still: That jobless rate is 17 percent, up from 11.2 percent a year ago and the highest in available data back to 1994.

'Workers who still have their jobs are getting knocked around another way: Less work. The government says non-supervisory workers had an average of 33 hours of weekly work last month (this includes part-timers), tying June for the fewest in data back to 1964. The lowest annual average was 33.6 hours last year. This year it's running lower still'.

Read on here.