Recent developments

[New page, May 2019]

Key points

The productivity estimates for the last six years show some positive signs, with the return of quite strong multifactor productivity (MFP) growth. Labour productivity (LP) growth, on the other hand, remains a little weak and by the last year (2017-18) has more or less disappeared.  

The acceleration in MFP growth has come about as input growth has been cutback more than the fall in output growth. This is in contrast to the MFP growth surge in the 1990s, when an acceleration in output growth was met with a smaller acceleration in input growth.

It appears there are two main things going on in recent times:
  • There has been lingering adjustment after the construction phase of the mining boom.
  • In the face of uncertainties and a slowing economy, businesses have slammed the brakes on further growth in capital and labour inputs. 
Signs of investments in new ways to lift productivity do not immediately stand out.

The recent decline in labour productivity is narrowly focused on the construction industry.

Without any substantial improvement in labour productivity growth to offset the income effects of the decline in the terms of trade, Australian living standards have stagnated.

The numbers

The ABS annual productivity estimates reported here include the 2017-18 year. The ABS establishes underlying trends in productivity as the average rate of growth between successive productivity peaks. However, the ABS has not yet declared a year to be a productivity peak after 2011-12. Although not a productivity cycle, the period 2011-12 to 2017-18 -- termed the 'recent period' -- can reasonably be taken to show recent underlying productivity trends.

The estimates presented on this page refer to the market sector of the economy, which encompasses 16 broad industry sectors. 

Figure 1 compares productivity growth in the recent period (2011-12 to 2003-04) with growth in the productivity slump period (2003-04 to 2011-12). 


Labour productivity (LP) growth of 1.6% a year in the recent period has been similar to the rate in the slump period. It is, however, still on the low side, being below the long-term average of 1.8% a year. (Long-term averages are calculated from 1994-95.)

Market-sector multifactor productivity (MFP) growth has increased from zero in the slump period to a solid 0.8% a year in the recent period. This is above the long-term average of 0.6% a year. 

The biggest turnaround has clearly been in capital productivity (KP) -- from deeply negative in the slump period to mildly negative in the recent period. It has been a 1.8 percentage point improvement in the average annual rate of growth. The recent period decline of -0.3% a year is milder than the long-term average.

The strong negative in KP growth over the slump was an aberration, however, brought on by the mining boom in particular. An improvement in KP is to be expected, given that the larger capital stock installed during the slump was meant to generate stronger growth in output going forward.

Year to year movements

These period averages, however, mask what has been happening year to year in the recent period. 

Rather starkly, annual LP growth has fallen away from 3% to a paltry 0.4% in 2017-18 (Figure 2). (In the 12 'selected industries' group, LP growth fared even worse, turning negative (-0.7%) in the last year.) 

The decline in KP growth over the recent period mostly took place in the first two years (Figure 2). There was either little change in capital productivity after 2013-14 or some small positive growth.

Annual MFP growth has been relatively steady, but has slowed since 2015-16. MFP growth can be viewed as a weighted average of capital growth and labour growth. And so MFP growth has risen as the improvement in capital productivity growth has offset the decline in LP growth.

In short,  MFP growth has re-emerged in strong fashion with substantial improvements in capital productivity. Labour productivity growth has not changed much over the entire recent period. However, while great weight should not be placed on one or two years' numbers, the very recent evaporation of labour productivity growth is of concern, given its link to growth in living standards and to sustainable growth in real wages. MFP growth has also slowed a little over recent years. 

Immediate contributors

Growth accounting relationships are used to investigate input and output contributors to change in productivity growth.
  • MFP growth = output growth - input growth
  • LP growth = output growth - labour growth
  • KP growth = output growth - capital growth

There has been no acceleration in output growth to push up the rate of productivity growth in the recent period. In fact, output growth has slowed since the slump to be below the long-term average (Figure 3). Productivity growth would have been about 0.5 of a percentage point higher (all other things being equal) if output growth had been at its long-term average.

MFP growth accelerated because input growth was reigned in further than the slide in output growth. Annual average input growth fell 1.4 percentage points, compared with a deceleration in output growth of 0.6 of a percentage point. And so MFP accelerated 0.8 of a percentage point (-0.6-(-1.4)=0.8).

The brakes were mainly applied to capital growth (Figure 3). Of the 1.3 percentage point fall in the rate of growth in combined inputs, slower capital growth contributed over 0.9 of a percentage point (about 70%) and labour contributed around 0.4 of a percentage point (30%).

Capital growth in the recent period has been well below the historical average. Labour use has also decelerated from the previous period to fall below the historical average.

The noted similarity in LP growth rates in the two periods has come about because output growth and labour input growth have slowed to roughly the same extent.

Capital productivity growth improved because the fall in capital growth was greater than the fall in output growth. 

Another LP growth decomposition

We can look at the contributors to LP growth from another angle. Table 1 shows that LP growth can be decomposed into contributions from MFP growth and capital deepening (essentially increasing the ratio of capital to labour). [LP growth = MFP growth + capital deepening]

The table reveals a switch in the composition of labour productivity growth between the two periods. In the slump, all LP growth came from capital deepening, as the mining sector in particular invested heavily in mining and infrastructure development and so capital growth outpaced labour growth. There was no MFP contribution to LP growth at all. In the recent period, however, the return of MFP growth has contributed about a half of the LP growth. A slower rate of capital deepening, stemming from the stronger cutback in capital growth, has contributed the other half.

Year to year movements

The course of change through the recent period is worth noting.

Figure 4 shows a slide in output growth from a little under 3% to 2.2% in 2016-17. It then strengthened to 2.7% in 2017-18. 

Input growth fell further -- from about 2.5% to below 1.5% -- in 2015-16. It then strengthened more than output growth in the last two years, but more noticeably in the last year. The pick-up in output growth in 2017-18 was met with a strong pick-up in input growth.

Figure 5 shows that labour provided the uplift in combined input growth in the last year.

Annual growth in capital has fallen through the recent period, from 4.9% at the beginning to 2.0% in 2015-16 and beyond 
(Figure 5). Growth in the most-recent years is well below the long-term average of 4.2%. At the same time, annual labour input growth has increased from very low rates to over 2% (against the long-term average of 1.3%pa).

This combination of slowing capital growth and increasing labour growth means the rate of capital deepening has slowed throughout the period. It slipped below the long-term average (1.3% a year) in 2014-15. It went on to fall into negative territory -- that is, capital shallowing -- in 2017-18. (The pattern for the selected-industries group is similar to that shown in Figure 5, but with a more pronounced fall into capital shallowing.)

Capital shallowing is very unusual. In fact, it has not shown up in any other year on record for the market sector, going back to 1994-95 (or for the selected industries group going back to 1973-74).

Capital deepening is usually a major source of improvement in labour productivity. Capital deepening comes about by providing labour with more and better equipment, more-advanced technologies, increased mechanisation of routine processes and so on. It increases the amount of output that is produced per labour-hour -- that is labour productivity.

In short, restraining growth in input use has played the main role in lifting MFP growth in the recent period. Output growth has been weaker.  Businesses have scaled back more on growth in their use of capital than of labour, leading to a slower rate of capital deepening.  In  the last year or two, output growth has relied on increased use of labour. As a result, the usual march of capital deepening and labour productivity growth has dwindled and disappeared.

Underlying reasons

Looking at productivity trends for industries within the market sector is often a good way to get a sense of what is going on. Table 2 displays industry LP and MFP growth in the slump and the recent periods and the acceleration or deceleration between the two periods.

The mining sector stands out as having high LP and MFP growth in the recent period, which was a very strong turnaround from the slump period. The utilities (electricity, gas,water & waste), and rental, hiring & real estate (RHRE) have also had strong accelerations in productivity growth. 

On the other hand, the construction industry, agriculture and transport show relatively strong decelerations. 

Industry contributions to MFP acceleration

Figure 6 shows industry contributions to the acceleration in market-sector MFP growth from the slump to the recent periods. The contributions take into account the relative importance of the industries as well as the size of their productivity accelerations.

Figure 6 should be viewed in the context of a 0.8 of a percentage point acceleration in market sector MFP growth. And so, for example, agriculture worked against an acceleration in market sector MFP growth (its contribution was negative), while mining contributed between 0.2 and 0.3 of the 0.8 percentage point acceleration. All the industry contributions, taking into account the negatives as well as the positives, add up to 0.8 of a percentage point.

Rental, hiring & real estate (RHRE), just on its own, contributed about half the market sector MFP acceleration. Mining and the utilities (electricity, gas, water and waste services) together contributed nearly 0.5 of a percentage point.

On the other hand, construction was a major detractor, nearly offsetting the positive contribution of RHRE.

Mining was the major positive contributor (around 0.5 of a percentage point) to an increase in LP growth (Figure 7). Construction, agriculture and manufacturing were the main detractors.

Industry contributions to the deceleration in output and inputs growth

As noted above, the MFP growth acceleration was associated with a stronger cutback in input growth than in output growth. Figures 8 and 9 show industry contributions to the slowdown in market-sector output growth and input growth respectively. The context is a slowdown in output growth of 0.6 of a percentage point and in combined input growth of 1.4 percentage points.
While many industries contributed to the slowdown in output growth, the fall was very much concentrated on the construction industry. It contributed over 0.4 of a percentage point to the 0.6 percentage point slowdown. Mining and RHRE, which experienced stronger rather than weaker output growth, each contributed (positively) more than 0.1 of a percentage point.

Contributions to the fall in input growth were widespread (Figure 9). The largest negative contributions came from RHRE and the utilities (EGWWS). Mining, wholesale trade, transport, finance and professional services also each took off 0.1 of a percentage point or more.

Figures 8 and 9 also give some insight into the industry contributions to the MFP acceleration shown in Figure 6. 

Mining and RHRE made large contributions to the MFP acceleration (Figure 6) because they combined positive contributions to output growth (Figure 8) with negative contributions to combined input growth (Figure 9). For the utilities, its contribution to the MFP acceleration came overwhelmingly from reigning in input growth.

Construction's negative contribution to the MFP acceleration came about because the fall in its input growth was less than the fall in its output growth.

Industry contributions to deceleration in capital and labour growth

We can look further into industry sources of the fall in input growth and in the switch from capital to labour. Figures 10 and 11 show the industry contributions to the decline in market sector growth in capital and labour from the slump to the recent period. The annual average rate of capital growth in the market sector has fallen 2.4 percentage points and the rate for labour has fallen 0.7 of a percentage point.

The brake on greater use of capital has been widespread (Figure 10). All industries, aside from mining, have contributed to the slowdown in use of capital in the recent period. RHRE stands out, taking out more than half a percentage point from the rate of market-sector growth in capital. Transport, utilities and manufacturing have also been major contributors to slower capital growth.

The pattern with labour is more mixed (Figure 11), with some industries making a positive contribution as well as those making negative contributions. The mining sector made, by far, the strongest contribution (nearly 0.4 of a percentage point) to the labour growth slowdown. There have also been relatively strong contributions from the utilities, wholesale trade, the finance sector and professional services. 

Types of capital

I will provide the supporting evidence soon, but the restraint in capital growth came from less investment in computers, electrical & electronic equipment and industrial machinery & equipment. Together these asset types accounted for about two-thirds of the fall in capital growth. Lower research & development growth also contributed.

All these asset types are generally considered to be platforms for technological advance and productivity improvement.

Industry contributions to change in market sector LP growth

Figures 8 and 11 provide some explanation of the industry contributions to market sector LP growth shown back in Figure 7. 

The strong positive contribution (about 0.5 of a percentage point) from mining to a change in the rate of LP growth, for example, came from a combination of a stronger output contribution (about 0.15pp) and a lower labour contribution (about 0.35pp).

Constructions's negative contribution came entirely from weaker output growth. The weaker contributions from agriculture and manufacturing came from stronger labour contributions, despite having weaker output contributions.

Year to year and most-recent movements 

Capital growth has been much slower on average in the recent period. As Figure 5 showed above, that average consisted of strong growth at the beginning of the period and very weak growth at the end of the period. It turns out one industry -- mining -- can account for the year-to-year decline in market sector capital growth.
While the growth in mining's use of capital accelerated over the entire recent period, the year-to-year movements show a different trajectory. Capital growth was very high -- over 12% -- in each of the first few years, but drifted down to around 2% a year over the last two years (Figure 12). The contribution of mining to market sector growth in capital fell from 4 percentage points to about half a percentage point. 

Just on its own, this would have more than accounted for all of the year-to-year decline in capital growth shown for the market sector in Figure 5.

The utilities also showed a similar pattern, except that its capital growth and contribution rose in the last year.
We can also look at the surge in output growth and labour growth in the last year. Figure 13 shows all industries, except agriculture, made a positive contribution to market sector output growth in 2017-18. The largest contributions came from construction and then finance, professional services and mining. The largest contribution to labour growth also came from construction and a number of industries that did not make strong output contributions. 

Construction took about half a percentage point of LP growth in the last year because its labour growth was so much stronger than output growth. Agriculture also took off about the same amount because it took on more labour, even though its output declined.

It appears the end of the construction phase of the mining boom has had a big influence on the improvement in productivity. 
  • The mining sector has combined stronger output growth with weaker input growth to make strong contributions to improved market sector MFP growth and, especially, to LP growth. 
  • While it did not contribute to improved productivity, the construction industry had a marked decline in output growth, consistent with the tailing off in mining construction. 
  • The rental, hiring and real estate industry, which made the strongest contribution to the MFP acceleration, made the strongest contribution to weaker capital growth. This could also be consistent, at least in part, with less growth in mining construction.
The utilities (electricity, gas, water & waste services) were also a major contributor to improved market sector MFP growth. This came about through much less growth in capital and labour.

The clearest pattern in the recent period has been the widespread cutback in capital growth across industries. This does not appear to be generally related to weaker output growth or to stronger labour growth.

The recent uptick in labour use has come predominantly from the construction industry. Construction experienced some additional output growth, but even more labour growth, leading to a decline in labour productivity. This took a full percentage point off market sector LP growth. Construction has acted as a break on MFP and LP growth generally since 2011-12, due to some input 'stickiness'.

In conclusion, it appears the productivity acceleration (MFP) has most to do with the end of the construction phase of the mining boom. There are broader signs of slower capital growth as well, which may be due to market uncertainties and a slowing in growth. The MFP acceleration has not come predominantly from investment in new technologies or other efficiency-enhancing capital (as happened in the 1990s surge), as these kinds of investments feature heavily in the cutback in capital growth. The weakness in LP growth is concentrated on the construction industry. 

There is some evidence that capital deepening has weakened in Australia. However, it does not appear to be strong enough and widespread enough to say that Australia has clearly joined other OECD countries in experiencing widespread falls in labour productivity growth due to a switch away from capital toward labour as a source of growth in output (OECD, OECD Compendium of Productivity Statistics, 2019). 

At this level of analysis, the signs of better productivity performance appear to have more to do with circumstance than widespread productivity-enhancing investment and effort.  

Implications for living standards

Australia's living standards have stagnated since 2011-12. They actually fell through to 2015-16, but have recovered in the last two years.

Gross Domestic Income (GDI) per person is used as the indicator of average living standards (Figure 15). GDI per person is GDP per person, adjusted to reflect the income effects of shifts in the terms of trade. GDP per person is often used as an indicator of living standards. But it is important to consider GDI per person because of the major shifts in the terms of trade over the past 25 years.

Figure 15 shows how GDI per person rose more rapidly than GDP per person through to 2011-12, as the terms of trade boom associated with the mining boom added further income gains. The opposite has been true of the recent period. A decline in the terms of trade has detracted from improvements in average living standards.

Figure 16 provides a decomposition of growth in GDI per person in the two periods. The average annual rate of growth in living standards was 2.7% in the productivity slump, of which the rising terms of trade contributed about half. The other half came from productivity (as measured by GDP per hour worked) and labour utilisation (the number of hours worked per head of population). (A rise in utilisation usually stems from an increase in labour force participation or a decline in unemployment.) 

The rate of growth in living standards in the recent period has been well down, at 0.3% a year. Productivity growth has picked up a little, but nowhere near enough to offset the fall in the terms of trade and lower labour utilisation.

That is why the signs of slower growth in LP and, in fact, its recent disappearance are of concern.