Today the ABS released the Labour Force data for May 2010 which show that the unemployment rate has fallen by 0.2 percentage points ostensibly, if you believe the press reports and the comments from the bank economists, on the back of continued strong growth in full-time employment. The truth is different. While full-time employment growth was positive it is not accelerating and overall employment growth slowed in May 2010. More importantly, all the fall in unemployment was due to a further drop in the labour force participation rate. So employment growth remains sluggish and is barely keeping pace with the growth in the population. The good news is that aggregate hours worked continued to increase which is reducing underemployment a little. While the bank economists have hailed today’s figures as indicative of an economy “near full capacity”, the reality is that the data is consistent with a broad array of statistics showing the Australian economy is slowing as the effects of the fiscal stimulus dissipate and and private spending remains subdued. It is amazing how a few headlines can distort what is actually going on.
The summary ABS Labour Force results for May 2010 are (seasonally adjusted):
- Employment increased 26,900 (+0.2 per cent) with full-time employment increasing by 36,400 and being partially offset by a reduction of 9,400 in part-time employment.
- Unemployment decreased 25,400 (-4.1 per cent) to 600,900.
- The official unemployment rate fell 0.2 percentage points to 5.2 per cent.
- The participation rate fell by 0.2 percentage points remained at 65.1 per cent which helped bring the unemployment rate down. Participation is still well down from its most recent peak (April 2008) of 65.6 per cent. So the approximate number of workers that have dropped out of the labour force because of diminishing job prospects (that is, the rise in hidden unemployed) is 96 thousand persons.
- Aggregate monthly hours worked increased 43.9 million hours(+2.9 per cent) and have now exceeded the July 2008 peak from the last cycle.
- Total labour underutilisation (the sum of underemployment and unemployment) is down to 12.2 per cent from the February quarter value of 12.8 per cent. The drop is made largely due to a fall in underemployment (from 7.5 per cent to 7.0 per cent) which isn’t surprising given the recovery in working hours. The reality is that when you combine the participation rate effects (hidden unemployment) with the ABS broad labour underutilisation you still have around 13.2 per cent of workers without enough work. That is significant evidence of labour market slack despite the rhetoric that we are approaching full capacity.
This is how today’s data was reported on the ABC news – Full-time work leads unemployment fall. That sounds good. The report said:
Hiring not firing: Unemployment fell to 5.2 per cent in May … Australia’s labour market has surprised analysts again, with unemployment falling to 5.2 per cent in May.
A small fall in the proportion of people looking for work from 65.2 to 65.1 per cent combined with the increase in jobs to lead unemployment down from 5.4 per cent in April to 5.2 per cent in May.
As the analysis that follows will show the actual fact is that the growth in employment barely kept paced with population growth and the drop in unemployment (and its rate) was all down to the fall in the participation rate. Further, employment growth actually slowed this month.
When you combine those facts you will see how misleading the news headline (above) is.
ABS News also recorded the comments from the bank economists. So:
It’s a very strong report with that rise in employment solely driven by full-time employment … “We saw a big rise in the number of hours worked and a significant drop in the unemployment rate so all round a pretty positive report.[AND]
Not only are employees holding onto and finding new jobs, but existing workers have got back the bulk of the hours they lost in the GFC[AND]
We do think the RBA will sit on the sidelines for the time being, that said we are expecting another two rate hikes by year end.
So the bank economists are predicting further interest rate rises on the back of a weakening labour market! That about says it all.
The Sydney Morning Herald story on the data release carried the headlines Jobless rate falls , which again presents a positive take on things.
They quoted the Prime Minister who said “We now have about half the unemployment level of the United States, half the unemployment rate of many countries in Europe”, which is a fairly inaccurate statement. Not only does this assessment fail to take into account the participation effects I will discuss presently but the broad labour underutilisation rate (unemployment and underemployment) in Australia was reported today to be 12.2 per cent and if you add about 1 per cent for hidden unemployment you get 13.2 per cent. The comparable US figure is around 16.8 per cent. That more correct comparison puts things in different light.
Employment growth slowing but positive
The following graph shows the month by month growth in full-time (blue columns), part-time (grey columns) and total employment (green line) over the last 12 months to May 2010 ysing seasonally adjusted data The overall picture is mildly positive. The sample period covers the latter parts of the downturn (May 2009 to August 2009) as full time employment growth was negative and part-time growth was mostly positive, which kept a lid on the overall employment losses although the slack showed up in lost hours of work, which was a notable feature of the downturn.
By September 2009, the effects of the fiscal stimulus package introduced in February 2009 were now evident and employment growth started to pick up quickly with growth in full-time employment signalling renewed optimism. In the more recent period, full-time employment growth continues to be positive but is slowing as part-time employment continues to fall.
While employment growth remains positive it has slowed in the last month. It is clear that the impacts of the fiscal stimulus, which drove GDP growth in the March quarter (see Australia GDP growth flat-lining) are now dissipating and private spending is not yet strong enough to really push the labour market to the next level of recovery.
So the picture is far from rosy although the additional full-time work is pushing total working hours up.
The official data shows that unemployment fell by 25,400 (-4.1 per cent) to 600,900 and the official unemployment rate fell 0.2 percentage points to 5.2 per cent. This is being hailed as a wonderful result and indicative of the underlying strength of the Australian economy. Standby for politicians to start saying we are close to full employment as a result of this month’s data.
Well nothing could be further from the truth. The following Table shows you what is really going on. It calculates the impact on the labour force and unemployment of the drop in the participation rate (down 0.2 percentage points). First, I computed the civilian population (by dividing the labour force by the participation rate).
Second, I computed the Labour Force with April participation rate by multiplying the Civilian Population estimate in May 2010 by the higher April participation rate. So given growth in the underlying population, if the participation rate had not dropped the labour force would have been 27.8 thousand workers larger in May 2010 than the official estimate. Most of those extra workers entered the ranks of the hidden unemployed.
Third, I revised the unemployment rate estimate by adding the 27.8 thousand workers to the May pool of official unemployment (600.9 thousand) and expressed the new higher estimated pool as a percentage of the upwardly revised Labour Force. The revised estimate of the unemployment rate is now 5.4 per cent for May 2010 which is unchanged from the official April 2010 estimate.
Conclusion: Almost all the fall in official unemployment and the unemployment rate was due to the participation rate contraction. This is not an improvement at all. It is just shifting the unemployed from the official side of the line (in the Labour Force) to the unofficial (hidden) side of the line (Not in the Labour Force).
You just cannot conclude that the economy is robust when you simultaneously have slowing employment growth and declining participation.
So how much difference has these participation effects made over the course of the downturn? The peak participation rate in the recent period has been in April 2008 (65.6 per cent). The participation rate is currently at 65.1 per cent. I simulated what the unemployment rate would have been if the participation rate since April 2008 was constant at that peak value.
The following graph shows the results. The blue line is the participation rate-adjusted unemployment rate (%) and the green line is the official unemployment rate. The difference between the lines is the participation rate effect on the labour force (and hence unemployment) as the participation rate fell below its peak. It indicates the hidden unemployment rate since April 2008.
While the official unemployment rate is estimated to be 5.2 per cent in May 2010 and everyone is crowing happily about that, the participation rate-adjusted unemployment rate would be 5.9 per cent. Quite a different story indeed.
Broader labour underutilisation
The following graph shows the movement since February 1978 to May 2010 (quarterly data) in the ABS Broad Labour Underutilisation rate (dark blue line) and their measure of underemployment (light blue line). The difference between the lines is the unemployment rate.
First, you can see the steady rise in underemployment as the economy grew after the 1991 recession. The economy was increasingly reducing unemployment by the creation of part-time work which still rationed the hours available relative to the preferences of the workers (who wanted more). The fact that total underutilisation didn’t scale the heights reached at the peak of the 1991 recession is due to the relatively small rise in the unemployment rate this time.
As you can see underemployment rose more sharply in the current downturn than it did in the 1982 and 1991 recessions. Almost all the labour slack in the 1982 recession was associated with rising unemployment. Underemployment didn’t really become a major issue until the 1991 recession.
The following graph shows the movements in the ABS Broad Labour Underutilisation rate measure since the beginning of the downturn (February quarter 2008) for males (blue line), females (green line), and total (red line). Typically, females have been the victims of the hours rationing due to their over-representation in the service sector. A notable feature of the current downturn is that underemployment has broadened its impact to embrace males. You
The following graph my 3-recessions graph for broad labour underutilisation (as measured by the ABS). It compares how quickly the broad labour underutilisation rose in Australia in the 1991 recession and the current episode. The broad labour underutilisation was indexed at 100 at its lowest rate before the recession in each case (June 1981; November 1989; February 2008, respectively) and then indexed to that base for each of the quarters until it peaked. It provides a graphical depiction of the speed at which the recession unfolded (which tells you something about each episode) and the length of time that the labour market deteriorated (expressed in terms of the unemployment rate).
The different behaviour in the current downturn is now starkly contrasted to the way the last two major recessions unfolded. You can clearly appreciate how harsh the protracted meltdown in 1991 actually was.
Hours worked – the good news
While total hours worked in April fell by 8.3 million hours (-0.5 per cent) which was on top of a fall in March of 10 million hours (-0.6 per cent), there was a sharp rebound in May – Aggregate monthly hours worked increased 43.9 million hours(+2.9 per cent).If the May figure was weak I was prepared to conclude that the trend would be weakening but the positive trend is now well-defined which is good news.
The following graph is taken from the the ABS data and shows the trend and seasonally adjusted aggregate hours worked indexed to 100 at the peak in February 2008 (which was the low-point unemployment rate in the previous cycle).
You can see a very flat V-shaped recovery with a positive trend – the national economy overall has now gone past the peak of July 2008 which is good news and will drive down underemployment.
State by State
Last month I considered the claim that had started to appear in the media as to whether the recovery phase was defining a two-speed economy where the “mining” regions (Western Australia, Queensland and Northern Territory) were driving growth and the old manufacturing areas (NSW and Victoria) were stagnating.
This is also relevant in light of the current political fiasco where the mining companies are resisting the introduction of a modest resource rent tax (stupidly terms a super profits tax by the Government) and spending millions on misleading advertising. The mining lobby has somehow managed to convince people that the industry is huge (it is not), that is saved us from the global financial crisis (it did not at all – it contracted more than most industries) and the tax will turn us into a communist state [if only! (-:]
Anyway, in the analysis last month there was no evidence to support two-speed hypothesis. The states with significant exposure to mining were not recording as strong employment growth.
I am just monitoring these trends at the moment. The following graph shows the percentage employment growth for the states and territories for each of the last two years (to May). There is nothing special about the periods chosen – just to correspond with the latest observation. The choice doesn’t really change the message.
You can see that the strongest employment growth is in the Northern Territory (although it is a tiny labour market). The old manufacturing stronghold of Victoria (VIC) and the Australian Capital Territory (ACT) are next.
The strong employment growth in the ACT, given it is a public sector economy (seat of government and main government departments are there) reflects the benefits of the fiscal stimulus and the modest expansion of government. Remember not to get tricked by scale. The ACT is a much smaller labour market in absolute terms than NSW and Victoria.
In the most populous states (NSW, VIC and QLD). NSW and Victoria are the manufacturing strongholds and have very little exposure to the mining industry. Queensland has some exposure to the mining industry clearly but also is probably benefiting from domestic-sourced tourism as our exchange rate appreciation makes holidaying abroad more expensive.
Importantly, states with significant exposure to mining like Western Australia are not recording strong employment growth,
So overall, while this analysis is crude, the data does not indicate a two-speed economy is emerging and doesn’t suggest any primacy in employment growth in the mining states. More detailed industry analysis will be available next week when the ABS publishes the detailed labour force data for May 2010.
While the business economists are claiming that the labour market is strong the facts are somewhat different. There is some growth especially in full-time employment and that is a good sign because it is contributing to the sharp increase in aggregate hours worked. This impact, in turn, is helping bring down underemployment.
But employment growth slowed overall and is barely keeping pace with population growth. Further, the usual signs of a strong recovery (rising participation) are absent. In fact, in the last month, the participation rate fell.
The combination of a slowing employment growth and falling participation do not usually augur a dynamic and fast growing economy. Taken together with the other data we are seeing on housing etc, the tentative conclusion is that growth overall is very weak. The National Accounts data for the March quarter clearly showed that without the fiscal stimulus we would have been in recession. That stimulus is being progressively withdrawn now and there is no sign that private spending is really ready to step up to the plate.
The other thing to note is that the fall in the unemployment rate (and unemployment) was almost all due to the falling participation rate. So we have substituted hidden unemployment for official unemployment. Given both cohorts would accept a job if one was offered to them, the overall wastage of productive labour remains the same.
You cannot escape the conclusion that the boost provided by the fiscal stimulus is waning.
Given today’s data and related data releases over the last few weeks, I am still of the view that a further fiscal expansion is required – and should be directly targeted at public sector job creation and the provision of skills development within a paid-work context. That would be a great boost to low inflation growth.
That is enough for today!