The Australian Bureau of Statistics (ABS) released data on construction work done over the September 2016 quarter earlier this week.  The data showed that the value of work done fell to its lowest level since December 2010.


Over the September 2016 quarter, $46.1 billion worth of construction was completed.  The $46.1 billion in work was -10.9% lower than over the previous quarter and -7.6% lower than the September 2015 quarter.  The -10.9% quarterly fall in work done was the greatest quarterly fall since the September 2000 quarter when work done fell by -13.5%.  The value of work done peaked at $56.9 billion over the December 2012 quarter (at the height of the mining led infrastructure boom) and the value of work done over the September 2016 quarter was -18.8% lower than this peak.


The $46.1 billion worth of work done over the quarter was split between $25.9 billion in building work and $20.3 billion in engineering construction.  The building work is further split into $17.6 billion in residential work and $8.3 billion in non-residential work.  Engineering construction work completed fell by -3.8% over the quarter and was -23.2% lower year-on-year.  The value of engineering construction work done is now at its lowest level since March 2010 and has fallen by -41.8% from its September 2012 peak.  Residential work was -3.1% lower over the quarter but has increased by 6.3% compared to September 2016.  Non-residential work was -10.9% lower over the quarter and -7.6% lower than a year ago.  Non-residential work done is -21.1% lower than its historic peak in September 2010.


Breaking this data into the proportion of total work done shows that 43.9% of the value of work done is engineering construction, 38.1% is residential work done and 18.0% is non-residential.  The chart highlights that the level of engineering construction activity experienced during the mining boom was like nothing seen previously.  At one point, 62.1% of all construction work done was engineering construction compared to its long run average of 37.9% of work.  As engineering construction activity has faded, residential work has lifted however, this lift has been nowhere near substantial enough to offset the sagging engineering construction work.  At its historic peak, residential accounted for 51.6% of work done and currently sits well below that level.  Residential work was never likely to rise to a substantial enough level to offset completely the drop in engineering work and as a result overall construction work done continues to trend lower.


Splitting residential construction work out further into new construction of houses and units shows the recent escalation in unit construction work.  Over the September 2016 quarter, $8.2 billion worth of new housing construction work was completed and $7.3 billion in new unit construction work was completed.  Housing work done fell by -4.8% over the quarter to be -2.5% lower year-on-year while unit work fell -1.7% over the quarter but is 18.9% higher year-on-year.  There remains a substantial pipeline of unit construction work underway so it is reasonable to expect that it will remain elevated for some time and may actually increase further over the coming quarters.

The ongoing slide in construction work done is concerning but it needs to be put in context.  The country has just been through a once in a lifetime mining boom which accompanied a substantial uplift in engineering construction as mining activity was expanded and infrastructure was built to support this substantial rise in resources related activity.  With this investment drying up, engineering construction is sagging.  Residential construction has ramped up, albeit not to the same magnitude as engineering construction did, however it too will be cyclical.  It is reasonable to expect construction work done to continue to fall however, a much needed increase in infrastructure expenditure could help to alleviate the inevitable slowdown in residential construction which will occur over the coming years.  Non-residential construction work done is typically much smaller than residential building and engineering construction and its recovery will be very much dependant on business confidence improvements or government stimulus in the form of infrastructure spending which would result in business hiring more staff and investing in their businesses.  There is yet to be any sustainable evidence of a turnaround in hiring and investment by businesses, so at least over the short-term we would expect non-residential construction to remain weak.