Today CofFEE released our latest quarterly labour market indicators (CLMI) which are hours-based measures (see below) that I have developed to more accurately measure the state of the labour market. The data shows that the impact of the global economic crisis is now manifesting in the Australian labour market with a marked deterioration in conditions in the February 2009 quarter. Total labour underutilisation in February has jumped to 11.2 per cent, up from 9.7 per cent in the November 2008 quarter. Things are heading south.
Summary for February 2009
- Total labour underutilisation has now jumped to 11.2 up from 9.7 in the November 2008 quarter. Over the last 12 months, total labour underutilisation has risen from 8.7 per cent to 11.2 per cent (seasonally adjusted).
- All components of the CLMI total labour underutilisation measure have risen. Unemployment for the February quarter was 5.1 per cent (up from 4.5 per cent in the November quarter); Hidden unemployment rose from 1.3 to 1.6 per cent in the February quarter; and Underemployment has jumped from 3.9 per cent in the November quarter to 4.6 per cent in the February quarter.
- Total underutilisation is now 9.5 per cent for males and 15.7 per cent for females.
- Underemployment in the February quarter (4.6 per cent) has already risen above the peak hit during the 1991 recession. As the economy struggled to resume growth in 1992, underemployment peaked in the November quarter 1992 at 4.4 per cent.
- Females have higher rates of underutilisation than males on all indicators, especially those measured in hours.
- Underemployment rates for females (7.1 per cent) are more than twice that for males (3.3 per cent). Both groups are now experiencing sharp rises in underemployment as firms make hours adjustments as sales drop.
- Total labour underutilisation is highest in South Australia (12.5 per cent) and lowest in Western Australia (8 per cent). NSW (11.9 per cent); Victoria (11.6 per cent) and Queensland (11.0 per cent) are next in order. This reflects relatively poorer performance across the three components of labour wastage: joblessness; reduced participation and underemployment.
- Underemployment is highest in South Australia (5.4 per cent) followed by Queensland (5.2 per cent). The deterioration in the South Australian and Queensland labour markets is most pronounced.
The following table shows the deterioration in the labour market over the last 24 months. U3 is total official unemployment rate; CU4 is U3 plus hidden unemployment; UE is underemployment; CU7 is U3 plus underemployment; and CU8 is the broad labour market underutilisation indicator which adds U3 plus hidden unemployment plus underemployment. You can see that the deterioration has been sharp in the last 3 months (the data is seasonally adjusted). If you want to learn more go to CofFEE Home Page and look around for the CLMI (currently the headline news story). You can download data and reports etc.
Why use broad indicators?
The labour force is the most commonly used measure of available labour resources in the economy. It counts all those over 15 who are either employed or unemployed. A person is defined as employed if they have at least one hours work per week. A person is defined as unemployed if they do not have work, but they are available for work and they are actively seeking work. The unemployment rate measures the number of persons unemployed as a percentage of the economically active population (the civilian labour force). The inference from the measure is that the economy is wasting resources and sacrificing income by not providing enough opportunities for work and underutilising labour.
However, labour underutilisation refers to all persons who are currently not working but who are willing and able to undertake work. Some of them may be classified as “in the labour force”, like the unemployed. Others may not be “in the labour force” but nevertheless have an attachment to it. These persons are referred to as marginally attached workers of which the hidden unemployment is a subset. The hidden unemployed or discouraged workers want to work and are available to start work in the reference week but are not actively looking for work. A major reason for their lack of search activity is that they believe that search is futile given the poor state of the labour market. They are discouraged from actively looking for work, and thus participating in the labour force, because of labour-market related constraints. The discouraged worker is thus more like the unemployed (in the labour force) worker than they are, for example, like a retired person or a child in full-time education, who are clearly not in the labour force. It is therefore appropriate to take into account their labour market attachment by including them in measures of underutilised labour.
The other major source of labour wastage manifests as underemployment. This refers to employed workers who are constrained by the demand side of the labour market to work fewer hours than they desire. It can also reflect underutilisation of skills due to lack of opportunities. In terms of time-related underemployment, a part of an underemployed worker is employed and a part is unemployed, even though they are wholly classified among the employed (because they work more than one hour in the survey week). Given this, it is appropriate to also include this unsatisfied willingness-to-work in measures of unused labour resources.
Typically, if these groups are taken into account, it is based on the numbers of persons involved. A truer measure of underutilisation is gained by taking into account how many hours each of the unemployed and marginally attached workers would like to work. Similarly, underemployment can be measured by considering how many more hours each of the currently underemployed would like to work. I developed these hours-based indicators some years ago to address these issues.
The CofFEE CLMI series is the most current available for policy purposes. The ABS have an annual series for underemployment and some more infrequent estimates of marginal attachment but these are computed in persons not hours.
How bad are we tracking?
In late January when I computed the November quarter 2008 indicators (I do it every 3 months for data availability reasons), I wrote a blog called From boom to bust where I started to track the current deterioration in comparison with the last two major recessions: 1982 and 1991. To visualise how the official unemployment rate (UR) and total labour underutilisation (CU8) are tracking since the boom ended, I constructed a graph which allows us to understand what happens during a major economic downturn.
I have another quarter’s data now and so I reconstructed the graph. Remember it tracks the quarters since the last peak in total employment (most recently February 2008). Here is the graph:
The graph shows the official unemployment rate (UR) and the CofFEE Labour Market Indicators broad measure of labour underutlilisation (CU8) from their lowest respective values to their highest values during the 1982 and 1991 downturns compared to 2008. So the lower three lines are the UR and the upper three lines are the CU8 measures (which are more accurate indicators of how much labour is being wasted).
You can read my earlier analysis of the relationships in the From boom to bust blog.
The data is now showing that the official UR is rising at the same rate as it did in 1991 and faster than in 1982. But the average level is lower (courtesy of the long boom). How long it will be before UR starts steepening as it did in 1982 and 1991 is something we will find out about in the coming months. But you can see how quickly it deteriorated in 1982, for example, and step-jumped up to the 1991 levels.
But what is now different?
The big difference at present is the behaviour of underemployment (which is implied by the sharp spike upwards in CU8 in the last two quarters. You can see that total labour underutilisation is now higher at this stage of the downturn than it was in 1982 and getting closer to the 1991 levels (notwithstanding the lower level of official unemployment).
In the 1982 downturn, underemployment (UE) went from 1.4 per cent to 2.3 per cent. So while it was an issue the number of part-time workers who wanted to work more hours but couldn’t find the extra hours was relatively modest. The major wastage of labour was unemployment.
However, in the 1991 downturn, a fundamental shift occurred in the labour market as employers scrapped full-time jobs rapidly and so began the sharp jump in underemployment. Over the 12 quarters that the labour market deteriorated, UE jumped from 2.2 per cent to 4.3 per cent. Even in the recovery of the late 1990s, UE never really returned to its lower levels.
The stark fact in the February 2009 data is that underemployment is now above the peak hit during the 1991 recession. As the economy struggled to resume growth in 1992, underemployment peaked in the November quarter 1992 at 4.4 per cent.
I said in the previous blog in January:
You can see that while the UR hasn’t yet started its quick climb, CU8 is already tracking the 1982 levels and will go above them relatively soon. This is why I think UE will be a significant component of the labour underutilisation in the coming downturn.
That is now true. So if the downturn continues then a projection that the total labour underutilisation rate might climb above 20 per cent is still a reasonable one.
In late May, I will update the chart again and provide some interpretations.
Digression: EVI reaches India
Our Employment Vulnerability Index which we launched last Tuesday gained a lot of publicitly in Australia. Today someone sent me a news item from the Hindustan Times which is a large Indian newspaper with a readership base of 12.4 million people! They seem to like the EVI!
I leave for Manila tomorrow for two days of meetings at the Asian Development Bank. I currently have some research contracts with them which I will tell you all about in the coming months as the confidentiality clauses allow me. So tomorrow a blog from there!!