There are, unfortunately, several concerning trends coming out of the October 2016 labour market report. Employment on a National level is declining, annual full-time (FT) employment is declining, part-time (PT) employment growth is slowing and the size of the labour force is declining as a result of falling participation. On the surface it looks like the falling number of unemployed persons is a bright spot in the report. That would be great if it was so, but falling participation is likely masking a higher number of unemployed persons. We don’t know the exact reasons why people are leaving the labour force, but I’m more likely to take a negative view of this outcome due to the number of declining employed persons. The implications of a weak labour market for private sector growth and budget outcomes is important. Slowing private sector credit growth, rising mortgage interest rates and a weak labour market is a combination that doesn’t bode well for private sector growth in Australia.
As always, I use trend data to analyse the labour force. The data points will move around from month to month, but it’s the broader trends that are important to focus on.
Annual growth in employed persons has slowed from +290k in Oct 2015 to +108k in Oct 2016
What a difference a year makes.
The chart below shows how employment growth has been distributed over the last 36 months. The most negative change has occurred over the last 12 months (if you add the monthly change for each of the last 12 months, you’ll get the annual change in employed persons of +108k).
The monthly growth in employed persons has slowed from a peak of +33k in Sept 2015 to National employment declining by -1k persons in the latest month.
There are two important points about this chart.
Firstly, even though employment growth is declining, it is declining slower than the labour force. This means that, despite the point that employment is declining, the total number of unemployed persons is also declining. This is the labour market situation in a nutshell at the moment. Rather than be counted as ‘unemployed’, people are leaving the labour market and participation is falling. Declining unemployment isn’t a clear sign of an improving labour market when both employment and labour force participation are also declining.
There is a second important point. Declining National employment, even on a monthly basis, is a relatively rare event. Below is the monthly change in total employed persons going back to 1979 – there are only seven (7) periods where National employment declined on a monthly basis:-
Since 1979, the periods when the monthly change in employed persons was negative were – 81/82, 90/91, 2000, 2003, 2008, 2013 and now. While only two of these periods were officially ‘recessions’, it easy to link each of these periods back some more general economic weakness. The most recent period of ‘weakness’ during 2012/13 (ToT falls/end of mining investment phase, GFC fiscal stimulus was wearing out and the impact of the European sovereign debt crisis), prompted the RBA to cut rates twelve (12) times, or 325bps, between Nov 2011 and Aug 2016. This helped indebted households as well as residential construction and the real estate market. We no longer have twelve rate cuts available to deal with any economic weakness.
Clearly, this current episode of declining employment is not anywhere near the severity of previous periods. But the direction remains concerning.
It would be easy to argue that this is a ‘one month’ data point and is more likely to result in revision in the following month due to the nature of the trend data. But you can break down this National employment trend to FT and PT trends and further still, down to state trends, to see that this is not a recent phenomenon. Some of the larger states have recorded declining employment well before the most recent month – NSW since August 2016, QLD since Jan 2016 and WA since Apr 2016. Employment growth in ACT has just started to dip into negative territory.
The number of FT employed persons declined by 50k over the last year
This is a disturbing part of this current labour market situation – the persistent decline in FT employed persons throughout 2016. This has been the driver behind the decline in overall employment.
Growth in FT employed persons has been slowing since Sept 2015, turned negative in Jan 2016 and has stayed negative since. Until Sept 2016, the decline in FT employed persons was at least off-set by the growth in PT employed persons. The growth in PT employed persons peaked back in May 2016. Since then, growth in PT employed persons has been slowing such that the decline in FT employment equalled and exceeded any PT employment growth in the last two months. If there is a glimmer of hope, its that the cycle of decline in FT employed persons may have peaked in Sept looking at the chart above.
The state data shows how widespread the decline in FT employed persons has been. On an annual basis, only VIC, SA and ACT have seen any growth in FT employment. But in the latest quarter, the state picture worsens – only SA recorded any FT employment growth:-
Australia is becoming increasingly reliant on part-time employment. Of total employed persons, 32% are PT employed – this is the highest proportion of PT workers in the data history. The underemployment rate has reached a new high in the August 16 quarter of 8.6%, highlighting that an increasing proportion of the labour force are available for and want more hours of work. It’s a telling point that suggests the shift to PT is not entirely by choice.
Unemployment is declining…but so is labour force participation
In the latest month, total unemployed persons declined by a further 4.15k persons. In the last 12 months, the total number of unemployed persons declined by -45k persons.
This should be a great highlight of the labour force data – the falling number of unemployed persons. But we are in a situation where employment is declining at a slower rate than the labour force. Going back to the first chart in this post highlights this relationship:-
The difference between the blue line (employment) for Oct 16 (-1.05k) and the orange line (total labour force) for Oct 16 (-5.2k) equals the monthly change in unemployed persons of -4.15k. As mentioned, this declining unemployment isn’t a clear/consistent sign of an improving labour market when both employment and labour force participation are also declining.
Labour force participation falls to 64.5% – almost back to where it was ten years ago
The other way to measure the changes in the labour force is to estimate what population growth adds to the labour force plus changes in participation. This perspective highlights the severity of the current round of declines in the labour force participation:-
Source: ABS, The Macroeconomic Project
As mentioned in previous posts, its best to ignore the two most recent estimates for underlying population growth – they are always low for the most recent months.
In the last year, I’ve estimated that underlying population growth has added approx. 185k persons to the labour force and the decline in participation has resulted in -122k persons leaving the labour force. This equals the annual growth in the labour force of +63k persons. On a monthly basis, the labour force size has been declining for the last 3 months – driven by declines in participation.
As of Oct 2016, the labour force participation rate is 64.5% – almost back to where it was ten years ago. Since Dec 2015, participation has declined sharply and I estimate this resulted in -125k persons leaving the labour force since then. This has been driven by both male and female workers leaving the labour force, but mostly males (-78k males left the labour force over the last year).
The decline in participation is potentially masking the real rate of unemployment. If over the last year, participation had remained constant, and employment had continued to fall, it would mean that our unemployment rate could have been as high as 6.5% – not the 5.6% that is quoted. The important point here though is that we don’t know exactly WHY people are leaving the labour force. I’ve previously looked at the decomposition of participation declines by age and gender and by state to at least understand whether we are seeing ‘boomers’ retiring from the labour force or if there was a geographic element to the trends. It will be worthwhile revisiting this analysis once updated data is available.
Looking at the state distribution of participation rate changes shows that the bigger changes in participation have occurred in key mining states such as WA and QLD, but participation is down in all states except VIC on an annual basis:-
Part of the reason for the fall in participation in WA & QLD is likely to be the result of workers transitioning to other states or jobs as the more labour-intensive investment phase of the mining boom continues to wind down. The latest quarter data suggests some improvement with participation higher in VIC, TAS & NT.
Hours worked confirms the FT and PT employment growth
Hours worked continued to grow for PT employed persons and continued to fall for FT employed persons:-
The trend in hours worked still looks lacklustre. Even though there is growth in PT employed persons, the growth in PT hours worked is only just above the longer term average. The year on year change (decline) in FT hours worked is well below the longer term average growth in FT hours worked.
The implications on spending and taxation are large.
We are a few weeks away from understanding the impact of the weakening labour market on tax receipts at the MYEFO. We are well into the first half of the 2016/17 budget year and wages are growing below growth assumptions, participation was forecast to remain at 65% – it’s now fallen to 64.5% and employment was forecast to grow at 1.75% and so far, on a seasonally adjusted basis, employment has fallen by -0.21%. It’s going to take quite a shift in activity to see these trends reverse and accelerate higher by the end of the financial year.
The household budget/income seems vulnerable right now:-
- Employment has started declining
- The ongoing shift from FT to PT employment will likely result in lower household income
- Continued slowing wage growth, where real wages are not increasing fast enough (on aggregate) to sustain the same level of disposable income
The other important source of spending growth in the economy is credit – and private sector credit (driven by business) is not accelerating. This suggests lower private sector growth and employment growth in the near term.
None of this would be a problem, except that we are more indebted than ever before – and we are now also looking down the barrel of rising interest rates. Rising mortgage interest rates will impact those households with variable rate mortgages. If deposit rates also rise, then it will help those with some interest income.