Australian Construction Productivity

Is the industry’s productivity as bad as claimed?

 

The Australian Bureau of Statistics publishes productivity measures for the whole economy, the Market Sector, and for the 16 industries that make up the Market Sector. Productivity is the ratio of output and inputs and is affected by innovation, research and development, education and training, the quality and age of the capital stock (of machinery, plant and equipment, buildings and structures), the rate of technological change and adoption of new technologies.  The effects of all these factors takes time, so productivity is a long-run measure that changes gradually. 

 

The post compares Construction productivity to the performance of the Market Sector. The data used is from the annual ABS Productivity Statistics release, which has data from 1994-95 to 2023-24 (the most recent release was February 2025). The ABS productivity indexes are based on 100 in 2022-23, however for this analysis they have been first rebased to 100 in 1994-95 to compare the long-run growth of Construction and Market Sector productivity, and then rebased at 100 in 2015-16 for comparing productivity in the short-run. 

 

Comparisons are made for labour productivity and multi-factor productivity (MFP) using both the hours worked and quality adjusted labour input measures. The quality adjusted labour input indexes take into account characteristics of the workforce like years of education, levels of training, industry of employment, age and sex. These quality adjusted measures reflect changes in the composition and skills of the workforce, and typically have a lower rate of growth than the hours worked measure. Capital productivity is also shown. 

 

As well as comparing the different measures of productivity for Construction and the Market Sector, there is data for the individual industries that shows Construction is in no way the worst performing industry, although it is far from the level of growth seen in the best performing industries.

 

 

Productivity Since 1995

 

The long-run performance of Construction includes a sharp rise during the mining boom between 2012 and 2015, followed by a gradual decline over the next few years as these major resource projects completed [1]. At the end of the mining boom productivity had fallen to around the level it was before the boom. This pattern was due to the large increase in Construction output during the mining boom because output included plant and equipment like the offshore drilling platforms and gas liquefaction plants, none of which involved much construction work and most of which was imported. Productivity increased because this statistical quirk increased output much more than employment and hours worked [1]. 

 

Labour Productivity

 

Starting with labour productivity over the long run since 1994-95, the difference between growth in the Market Sector and the lower productivity growth of the Construction industry is apparent in Figure 1. However, despite claims made that there has been no growth in Construction labour productivity, there has been an increase. Construction labour productivity has increased by 17% on an hours worked basis and 24% on the quality adjusted labour input basis which, although less than the Market Sector’s 64% and 41% respectively, is not nothing. 

 

Figure 1. Market Sector industries labour productivity

 

Source: ABS 5260. Gross value added per hour worked. Quali is the quality adjusted labour input measure. 

 

As Table 1 shows, since 1995 the three leading industries for hours worked labour productivity growth have been Agriculture, forestry and fishing 210%, Information media and telecommunications 228%, and Financial and insurance services 123%. The two industries with lower growth than Construction were Mining 6%, Electricity, gas, water 2%, and Administrative and support services had negative growth of -13%. 

 

For quality adjusted labour productivity, Construction had better growth than Rental, hiring and real estate services 4%, and there were three industries with negative growth: Mining -2%, Electricity, gas, water and waste -9%, and Administrative and support services -23%.

 

Table 1. Market Sector industries labour productivity change

 

Multi-factor Productivity 

 

The ratio of output to input of combined labour and capital is multi-factor productivity (MFP). For MFP the story is not as good as for labour productivity, because there has been only 1% growth in Construction hours worked MFP and a 3% fall in the quality adjusted measure.  Market Sector growth on the hours worked basis was 23% and on the quality adjusted labour input basis was 13%. After MFP rose and fell during the mining boom, instead of returning to the preboom level there was collapse in Construction MFP after 2015-16.

 

Figure 2. Market Sector industries multi-factor productivity

 

Source: ABS 5260. Gross value added per hour worked. Quali is the quality adjusted labour input measure. 

 

The 1% increase in Construction hours worked MFP is very small, but not the decline often claimed for the industry. Table 2 shows four industries had a fall in hours worked MFP since 1995:  Mining -28%, Electricity, gas, water -30%, Rental, hiring and real estate services -32%, and Administrative and support services -16%. The three high growth industries were: Agriculture, forestry and fishing 182%, Information media and telecommunications 64%, and Financial and insurance services 63%. 

 

Construction, however, was one of five industries with negative quality adjusted labour input MFP growth, although at -3% it had a much smaller decline than the other industries of Mining -31%, Electricity, gas, water and waste -33%, Rental, hiring and real estate services -36%, and Administrative and support services -25%. This raises the question of why Construction is singled out as the problem industry. 

 

Table 2. Market Sector industries multi-factor productivity change

 

Capital Productivity

 

Capital productivity has been falling for both the Market Sector and Construction since the early 2000s.  This is a complex measure, because estimating the stock of capital requires an estimate of annual capital investment and a depreciation rate to account for declining efficiency of the existing stock due to use and age. Although Construction capital productivity peaked in the mid 2000s and declined during the mining boom, the post-boom fall in MFP was due to the sharp decline in capital productivity, because since then labour productivity was more or less flat but capital productivity was falling. As Figure 3 shows the Market Sector also had declining capital productivity, but after 2015-16 the decline in Construction capital productivity was much worse. 

 

Figure 3. Market Sector industries capital productivity 

 

Source: ABS 5260. 

 

What these long run graphs show is that there was a downward shift in Construction productivity around 2015, when both MFP and capital productivity went into significant decline. Up until then Construction productivity had been similar to Market Sector productivity for MFP, but after 2015 the Market Sector and Construction industry measures diverged. The next section looks at productivity over the short run since that divergence.

 

Productivity Since 2015-16

 

Labour Productivity

 

Labour productivity in the short run since 2015-16 has a distinctive and interesting pattern. The hours worked measure has fallen 4% from 100 to 96 but the quality adjusted labour input measure has increased by 6% from 100 to 106, and was in fact higher then both Market Sector measures in 2023-24. The increase in the Quali index occurred in the 2019-20 year with a big jump from 95 to 104, and there has been a gradual increase in the years since. 

 

Figure 4. Market Sector industries labour productivity 

 

Source: ABS 5260. Gross value added per hour worked. Quali is the quality adjusted labour input measure. 

 

The increase in the Construction quality adjusted labour input measure index will be the result of changes in the composition of employment, with the combined share of Professionals and managers increasing from 15% to 18% between 2019 and 2020, and peaking at 19% in 2022. Figure 5 shows the share of Professionals increased from 4% to 6% in 2020, and for Managers the share rose rom 10% to 12% in 2020 and was 13% from 2021 to 2023. In 2024 Technicians and trades workers were 50% of Construction employment, and Machinery operators another 6%, and their combined shares in total Construction employment have decreased by 3% since 2016. The share of Clerical and administrative workers has also declined, by 0.6%. Therefore, since 2016 the overall makeup of Construction workforce has become more skilled and qualified, raising the quality adjusted labour input measures [2]. 

 

Figure 5. Share of total Construction employment

 

Source: ABS 6291

 

Between 2016 and 2024 there were large differences in the productivity performance of the 16 Market Sector industries. As Table 3 shows, on the labour productivity hours worked basis there were two industries with high growth: Agriculture, forestry and fishing 44%, and Information media and telecommunications 40%. Four industries had growth between 10 and 20%, and five had growth less than 10%. Construction -4% was one of five industries with negative growth, the others were Mining -15%, Manufacturing -4%, Electricity, gas, water and waste -15%, and Financial and insurance services -4%.

 

On a quality adjusted basis Construction was the only industry to improve on the hours worked measure, all other industries had slightly lower quality adjusted labour input growth than hours worked. The other four industries with negative hours worked labour productivity again had negative quality adjusted labour input labour productivity growth. There were only six industries with better quality adjusted labour productivity growth than Construction’s 6%: Agriculture, forestry and fishing 41%, Wholesale trade 7%, Accommodation and food services 8%, Information media and telecommunications 33%, Professional, technical and scientific services 14% and Administrative and support services 12%.

 

Table 3. Market Sector industries labour productivity change

 

Multi-factor Productivity 

 

The MFP indexes for Construction do not show the same pattern as labour productivity. Both the hours worked and the quality adjusted indexes have fallen since 2016 and have closely followed each other down, ending at 92 and 91 respectively in 2024. However, the Market Sector has not performed particularly well, with the quality index only increasing to 101 and the hours worked index increasing to 104. 

 

Figure 6. Multi-factor productivity

 

Source: ABS 5260. Gross value added per hour worked. Quali is the quality adjusted labour input measure. 

 

MFP growth since 2016 is similar to labour productivity with a couple of exceptions. Table 4 shows on the hours worked measure only Agriculture, forestry and fishing 44% had high growth, and there were three industries above 10%. Five industries had negative growth: Construction -8%, Mining -3%, Manufacturing -1%, Electricity, gas, water and waste -15%, and Arts and recreation services -1%. Again, the growth in the quality adjusted labour input measure was lower than for hours worked, with Construction -9% one of eight industries with declining productivity, including Mining -3%, Manufacturing -3%, Electricity, gas, water and waste -16%. Transport, postal and warehousing -3%, Rental, hiring and real estate services -1%, and Arts and recreation services -3%. 

 

Table 4. Market Sector industries multi-factor productivity change

 

Capital Productivity

 

The performance of capital productivity has been particularly poor for construction, falling from 100 to 85 between 2016 and 2024, while the market sector index barely increased and ended at 103.

 

Figure 7. Capital productivity

 

Source: ABS 5260. 

 

Misunderstanding Productivity

 

There are two common misunderstandings about Construction productivity. One is that increasing offsite manufacturing and use of modern methods of construction like prefabrication and modular buildings will increase measured Construction productivity. It will not, because that work will be included by the ABS in the Manufacturing industry subdivisions of Prefabricated steel and timber buildings, Concrete products, and Structural steel. In fact, one reason for the lack of growth in measured Construction productivity has been the gradual but continual shift to more prefabrication and offsite manufacture. 

 

A second misconception is that improving Construction productivity will somehow decrease the cost and increase the number of dwellings being built. This mistakes new construction for the market for housing, where in the short-run price is determined by the interplay of demand and an inelastic supply of new dwellings due to limited industry capacity to build and lengthy approval times. Increasing onsite productivity might decrease the time to complete a build but will have a marginal effect on the total cost of delivery, and the number of dwellings built is determined by project feasibility (i.e. the profitability of development) at any one time. Improving Construction productivity might help, but on its own cannot and will not solve the housing crisis. 

 

Conclusion

 

That the Construction industry has had no or negative productivity growth for the last few decades has been repeated so many times by so many commentators it has become an accepted fact about the industry. There are, however, four different measures of productivity, and commentators can focus on those that support their claims, and productivity growth rates vary considerably over different time periods, allowing selective choosing of comparisons. 

 

The four productivity measures are labour productivity on an hours worked basis or quality adjusted labour input basis, and multi-factor productivity (MFP includes the capital stock) also on an hours worked basis or quality adjusted labour input basis. The ABS productivity statistics for the Market sector go back to 1994-95, and this analysis has been for two periods, the long-run from1994-95 to 2023-24 (the latest data) and the short-run period of 2015-16 to 2023-24, chosen because 2015-16 was the end of the rapid rise and fall in Construction productivity during the mining boom. 

 

When the productivity of Construction is compared to the Market sector, despite claims that there has been no growth in Construction labour productivity, there has been an increase. Since 1994-95 Construction labour productivity has increased by 17% on an hours worked basis and 24% on the quality adjusted labour input basis which, although less than the Market Sector’s 64% and 41% respectively, is not nothing. Construction is in no way the worst performing industry, although it is far from the level of growth seen in the best performing industries.

 

In Australia there is a wide difference between a group of high productivity growth industries and a group of low or negative productivity growth industries. On the hours worked measure for labour productivity, since 1994-95 there were three high growth industries, and ehree industries with lower growth than Construction. For quality adjusted labour productivity, Construction had better growth than Rental, hiring and real estate services’ 4%, and there were three industries with negative growth. 

 

For MFP the story is not as good, because since 1994-95 there was only 1% growth in Construction hours worked MFP. That 1% increase in Construction hours worked MFP is very small, but not a decline. Market Sector growth on the hours worked basis was 23%, and on the quality adjusted measure Market Sector growth was 13%. On the hours worked basis there were three high growth industries, and four industries had a decline. Construction was one of five industries with negative quality adjusted labour MFP growth, although at -3% it had a much smaller decline than Mining -31%, Electricity, gas, water and waste -33%, Rental, hiring and real estate services -36%, and Administrative and support services -25%. This raises the question of why Construction is singled out as the problem industry. 

 

Construction capital productivity peaked in the mid 2000s and falling MFP was due to this decline in capital productivity. The Market Sector also had declining capital productivity, but there was a downward shift in Construction productivity around 2015, when both MFP and capital productivity went into significant decline and the Market Sector and Construction industry measures diverged.

 

There is a notable difference between the quality adjusted labour input measures and the hours worked measures for Construction labour productivity since 2015-16, because the hours worked measure has fallen 4% but the quality adjusted labour input measure has increased by 6%. The increase in the Construction quality adjusted labour input measure index will mainly be the result of changes in the composition of employment, with the combined share of Professionals and Managers increasing from 15% to 19% in 2022. The Construction workforce has become more skilled and qualified, raising the quality adjusted labour input measures.

 

Between 2016 and 2024 on the labour productivity hours worked basis there were two high growth industries, four industries had growth between 10 and 20%, and five with growth less than 10%. Construction -4% was one of five industries with negative growth, the others were Mining -15%, Manufacturing -4%, Electricity, gas, water and waste -15%, and Financial and insurance services -4%. On a quality adjusted basis Construction was the only industry to improve on the hours worked measure, and there were only six industries with better quality adjusted labour productivity growth than Construction’s 6%.

 

For MFP growth since 2016 on the hours worked measure only Agriculture, forestry and fishing had high growth, and there were three industries above 10%. Five industries had negative growth: Construction -8%, Mining -3%, Manufacturing -1%, and Electricity, gas, water and waste -15%. Again, the growth in the quality adjusted labour input measure was lower than for hours worked, with Construction -9% one of eight industries with declining productivity, including Mining -3%, Manufacturing -3%, and Electricity, gas, water and waste -16%.

 

Clearly, Construction is far from the worst performing industry, which raises the question of why it is so often singled out for low productivity growth. There were only six industries with better quality adjusted labour productivity growth than Construction. And are industries that have had declining productivity like Mining or Electricity, gas, water and waste not important? Should their productivity performance not be scrutinised? 

 

Maybe Construction could do better, but there have only been a few high growth industries in Australia over recent decades. Construction is one of a group of low growth industries, and compared to those industries its performance has been much better in both the long and the short-run. Instead of complaining about low productivity growth, attention should be focused on addressing the issues that have negatively affected Construction productivity, such as the number of micro and small firms, lack of standardisation of structural elements, the low level of investment in software and capital stock, state based occupational licensing and building codes, procurement methods, financing and project management, and education and training systems [3].

 

 

[1] See The long cycle in Australian construction productivity

 

[2] See The changing composition of construction employment

 

[3] See Housing productivity report a missed opportunity

Australian Productivity and Performance

  National and industry statistics from the ABS 

 

Australia is fortunate in having high quality data from the Australian Bureau of Statistics, one of the world’s leading national statistical agencies. The ABS regularly publishes a range of productivity measures for tracking and understanding economic performance. The ABS notes productivity measures should be used for analysis of long-term growth, and short-term fluctuations are not reliable indicators of performance: ‘Caution needs to be exercised in interpreting productivity measures, which are derived as a 'residual', and are therefore subject to any errors in the output and input measures.’

 

Productivity is a measure of the efficiency of production, how much output is delivered from the inputs used. That ratio between output and inputs is affected by economy-wide factors, such as innovation, research and development, education and training, technology and investment, management and industrial relations. There are also specific factors that affect worker productivity, such as the quality and age of the capital stock (of machinery, plant and equipment, and buildings and structures), the rate of technological change and adoption of new technologies [1]. 

 

Although productivity growth many countries in Europe and Asia has been low over the last decade, for the last two years Australian productivity growth has been very low and was negative in 2022-23. Nevertheless, Australia’s productivity performance has not been as poor as some commentators make out (see here and here). As Figure 1 shows, using Purchasing Power Parity (PPP) estimates rather than converting to US dollars, in 2019 Australian productivity was in the middle of the range for developed economies, and the growth rate between 1990 and 2019 was better than some countries (e.g. Canada, Italy) but worse than others (e.g. Denmark, Germany). Estimates using PPPs for 2022 are similar. 

 

Figure 1. Productivity compared

Source: Our World in data. Based on Purchasing Power Parity not currencies. 

 

The Treasurer Jim Chalmers has said that productivity will be a focus of the newly re-elected government. Late last year he started five Productivity Commission inquiries, on economic dynamism and resilience, workforce issues, data and digital technology, a more efficient care economy, and investment in the clean energy transformation. The recommendations from those inquiries will join, and probably be very similar to, the 29 reform directives based on 71 recommendations in the Productivity Commission’s 9 volume 2023 report Advancing Prosperity.

 

This highlights the problem: there is a shopping list of solutions to low productivity, and a basket of recommendations from industry associations, unions, think tanks and commentators with their favoured reforms. While most are worthy, like improving education and training or more R&D and innovation, some are self-serving, like tax reform and regulations. Another problem is that productivity is a long-run measure, and many of these recommended policies or actions would take a long time to implement or have an effect. What is needed is a more focused agenda with one or two measurable targets that can be achieved in the short to medium term, say five to ten years. 

 

Productivity Statistics

 

The ABS publishes productivity measures for the whole economy, the market sector, and for each of the 16 industries in the market sector. The number and variety of these measures can make it difficult to identify and find the most appropriate or relevant measure. This overview explains the different measures and provides data on each for comparison, with some comments on characteristics and performance.   

 

The most widely used productivity measure is labour productivity, which is gross domestic product per hour worked for the whole economy or gross value added per hour worked for the market sector. This measure is relatively straightforward and does not require assumptions to be made about the inputs and outputs, although there can be issues in the measurement of their quantities. 

 

As well as labour, production requires intellectual and physical capital, like R&D, tools, equipment and buildings. The ratio of a unit of output produced by a unit of the combination of labour and capital is multi-factor productivity (MFP). This is a more complex measure because estimating the stock of capital requires an estimate of annual capital investment and a depreciation rate to account for the declining efficiency due to use and age of the existing stock. Capital productivity has been the weak point in Australia, due to the low level of industry investment, particularly in IT capital equipment and software. 

 

There are also quality adjusted labour input indexes, which take into account characteristics of the workforce like years of education, levels of training, industry of employment, age and sex. These quality adjusted measures can be used for both labour productivity and MFP, and reflect changes in the composition and skills of the workforce.

 

Finally, there is the KLEMS productivity measure, the most comprehensive but also the most complex of the measures available. This measure has five input categories for capital (K), labour (L), energy (E), materials (M), and services (S), for market sector industries. The KLEMS measure identifies industry cost structures and the contribution to output growth of each of those inputs.

 

In this post these productivity statistics are reviewed and the most recent data presented. Economic and industry performance are discussed, starting with the high level whole economy data and then moving to the industry level measures for labour productivity, multi-factor productivity and the KLEMS measures. Construction industry productivity is discussed and the performance of the industry assessed. The importance of investment in the capital stock is emphasised. 

 

Labour Productivity

 

Labour productivity is the most straightforward measure, and is output per unit of labour input as persons employed or hours worked. The ABS publishes estimates of labour productivity as gross domestic product (GDP) per hour worked for the economy as a whole and for the 16 industries in the market sector as gross value added (GVA) per hour worked in the annual System of National Accounts. The long run data from 1995-96 shows growth in the 1990s that slowed down in the 2000s, and has been flat for the last five years. Figure 2 shows GDP per hour worked for the whole economy. The index has increased from 70 in 1995-96 to 100 in 2023-24, or 44%. 

 

Figure 2. Labour productivity, whole economy

Source: ABS 5204

 

Market sector productivity does not include estimates for three industries considered to be non-market because they lack meaningful prices and it is impossible to calculate volume measures of output (a cost-of-service delivery method is used, i.e. wages and salaries). The industries are Public Administration and Safety, Education and Training, and Health Care and Social Assistance.  Because productivity cannot be accurately measured in the non-market sector, comparisons should be made using market sector statistics. Also, although efficiency and productivity in the non-market sector should be improved as much as possible, the cost-of-service measure used means that will have little effect on labour productivity growth for the whole economy. 

 

Figure 3 shows gross value added (GVA) per hour worked for the 16 industries in the market sector. In 2022-23 market sector labour productivity fell 3.5%, the largest fall since the series began, due to an increase in hours worked of 6.9% compared to GVA growth of 3.8%. The index for GVA per hour worked has increased from 63 in 1995-06 to 101 in 2023=24, or 61%. The difference between the two measures is significant, for the whole economy as GDP per hour worked the growth in productivity has been less than for the market sector. Unsurprisingly, the public sector has lower productivity growth than the market sector, and as its share of the economy has grown in the last few years productivity growth for the whole economy has slowed. 

 

Figure 3. Labour productivity, market sector

 

Source: ABS 5204

 

 

Quarterly indexes of GDP and GVA per hour worked are published in Australian National Accounts: National Income, Expenditure and Product. The annual estimates smooth the quarterly labour productivity measures, which are revised for up to three years as more data becomes available. In the revisions June is often revised down and September often revised up, larger estimates have larger revisions, positive estimates are more likely to be revised down, while negative estimates are more likely to be revised up. Revisions are largest in the September quarter, when the ABS benchmarks their estimates. 

 

Figure 4. Labour productivity, whole economy, quarterly

Source: ABS 5204

 

The quarterly labour productivity measures very considerably from quarter to quarter, and clearly show the whole economy estimates do not vary as much from quarter to quarter as the market sector estimates. Market sector productivity growth has been stronger but more volatile. These quarterly estimates should be regarded as indicative of trends but not definitive.

 

Figure 5. Labour productivity, market sector, quarterly

Source: ABS 5204

 

Construction Industry

 

Construction labour productivity has barely changed over the last 20 years. However, during the mining boom 2011-2016 labour productivity rose and then fell, following the increase and decrease in the volume of work done. As Figure 6 shows, since 2020 productivity has been around the same level as it was before the boom [2].

 

Figure 6. Construction labour productivity

Source: ABS 5206

 

Multi-factor Productivity

 

Labour and capital can be combined into multi-factor productivity (MFP), measured as GDP per unit of combined labour and capital. The ABS publishes annual indexes of industry level MFP estimates for the market sector, in Estimates of Industry Multifactor Productivity.

 

Figure 7. Multi-factor productivity

Source: ABS 5206

 

Construction MFP also rose and fell with the mining boom, and has followed a similar path as labour productivity with somewhat lower fluctuations in the year-to-year changes. What Figure 8 shows is that in many years up to  2014-15 there has been little or no increase in Construction MFP, except for years such as 2001-02 and 2011-12, when it jumped by 10% or more, and for the last decade has fallen in every year except two. 

 

Figure 8. Construction MFP

Source: ABS 5206

Industry Comparisons

 

Over the last five years there have been significant differences between industries in productivity growth. Table 1 has the 16 industries in the market sector, and shows between 2019 and 2024 three industries had high growth in output per hour worked of over 10%, but overall performance has been affected by five industries with negative growth, low growth in another seven industries, and no growth in construction. For most industries, MFP growth was worse than for labour productivity as hours worked. The highest performing industries were Agriculture, Forestry and Fishing and Information, Media and Telecommunications, the worst performing industries were Mining and Electricity, Gas, Water and Waste. 

 

 

Quality Adjusted Labour Input Measures 

 

Measuring labour input as hours worked assumes that the workforce is homogeneous. An alternative approach to human capital is quality adjusted labour inputs (QALI). These classify workfers by industry of employment, sex, educational attainment, and age group [3}. QALI indexes are published for the market sector in the Australian System of National Accounts, and for each market-sector industry in Estimates of Industry Multifactor Productivity. The difference between hours worked and quality adjusted labour input is a measure of composition change in the workforce due to a shift in the skill mix of labour input.

 

Figure 9. Quality adjusted labour input productivity

Source: ABS 5204

 

Figure 10. Quality adjusted labour input multi-factor productivity

Source: ABS 5204

 

Growth Cycles

 

Labour productivity growth in Australia could be better, recently averaging less than 1% a year growth for the market sector. The 20 year average growth rate was 1.7% in 2015-16, but by 2022-23 it had fallen to 0.9%. There is a strong relationship between economic conditions and productivity, because the rate of growth in GDP affects the rate of growth of productivity. The ABS uses peak-to-peak growth cycles in MFP to adjust for the effect of changes in GDP growth, and publishes growth cycles for the different productivity measures for the market sector, in Table 2. These show a slowdown in labour productivity growth in the recent cycle from 2017-18 to 2021-22, but a small increase in MFP growth.  

 

 

 

Industry Comparisons

 

Table 2 compares labour productivity across the 16 market sector industries for the long run 1995-06 to 2023-24 and the last five years 2019-2024, as GVA per hours worked and GVA with QUALI. There are wide differences in industry performance over both periods. The increase in the QUALI indexes for the long-run were generally 10 to 20% below the unadjusted labour productivity, but for the five years 2018-19 to 2023-24 changes in the QUALI indexes are typically close to the hours worked. 

 

Construction was the only industry where the QUALI indexes had a greater increase than the hours worked indexes. Several industries had worse productivity performance than Construction, such as Mining, Electricity, Gas, Water and Waste, and Administrative and Support services. Despite the common claim that there has been no growth in Construction productivity, over the five years to 2024 Construction was one of the best performing industries in quality adjusted labour productivity as hours worked. 

 

 

Table 4 compares MFP across the 16 market sector industries for the long run 1995-06 to 2023-24 and the five years 2019-2024. MFP increased by a lot less than labour productivity in both periods, and again the QUALI indexes increase less. In the 5 years to 2024 there was little difference between the two measure for many industries. Construction MFP had a 1% rise in the hours worked measure in the long run, but declined in the other periods. 

 

 

 

Capital Productivity

 

The two major inputs to productivity are labour and capital, and these have the two characteristics of quantity and quality. Labour is adjusted for years of education and experience in the QUALI measures above, while capital is adjusted for age through a depreciation rate that reflects wear and tear from use. This is not a straightforward calculation because of assumptions about the effect of age and use on the performance of plant, equipment and machinery and the durability of building and structures [4]. This makes estimates of the capital stock (the quantity of those assets) and capital productivity sensitive to the rate used. 

 

Capital productivity has fallen, particularly in the decade after 2004. This decline in capital productivity is the reason Australia’s productivity growth has been so low. Although the workforce is increasingly trained and skilled, the capital stock has not been replaced and updated due to the low level of business investment. As Figure 11 shows, in the majority of years since 1995-96, capital productivity has decreased. 

 

Figure 11. Capital productivity

Source: ABS 5204

 

Figure 12 shows Total private gross fixed capital formation (GFCF), which is private sector investment in the capital stock, and capital productivity. The increases during the fiscal response to the 2008-09 global financial crisis and the expenditure on plant and equipment during the mining boom from 2012 is apparent. What is significant is the decline in expenditure on software, which is a separate item in the capital expenditure statistics, which will have contributed to the already established decline in capital productivity (so not the only causal factor), and between 2018 and 2021 capital productivity and the shares of GDP of both total private GFCF and private software GFCF were all flat, before an increase in GFCF in 2023 and 2024 lifted capital productivity. Australian productivity overall can be improved by increased investment in the capital stock, and in particular in IT and software capital [5]. 

 

Figure 12. Capital productivity and private gross fixed capital formation

Source: ABS 5206. Total private GFCF and Private GFCF on Software as percent of GDP in current dollars. 

 

Industry KLEMS MFP

 

KLEMS MFP is only available for individual market sector industries, the ABS does not publish aggregate KLEMS indexes for the market sector. The KLEMS framework has five input categories: capital (K), labour (L), energy (E), materials (M), and services (S), with the contributions to output growth of each of those inputs, which allows analysis of changes in the input mix, such as labour hours worked, capital services or intermediate inputs. The intermediate inputs are energy (E), materials (M) and services (S), which includes hiring, renting and out-sourcing between industries. For example, if a construction company leases a crane from the rental and hiring industry, it is recorded as a service component in the intermediate inputs of the lessee and as capital services by the lessor in the rental and hiring industry. 

 

Industry cost structures are diverse and vary from industry to industry and over time. The variation from year to year in the contribution of each input to output growth reflects factors like wage and price changes, interest rates, and fluctuating demand. Therefore, averages are conventionally used, and Figures 13 A and B show the long-run averages for each industry for the period data is available, from 1995-06 to 2022-23.. 

 

Clearly, for most Australian industries the main contributor to growth has been services provided by other industries, including Construction. Non-IT capital intensive industries are Mining, Electricity, Gas, Water and Waste Services, Transport, Postal and Warehousing Services, and Rental, Hiring and Real Estate Services. Agriculture, Forestry and Fishing, Manufacturing and Construction have a high level of materials as inputs.

 

 Financial and Insurance Services is the only industry with a high level of IT capital services. The low level of IT capital services in many industries is a major problem at a time when IT related innovations like generative artificial intelligence, cloud computing and reality capture are advancing rapidly, and points to the most important challenge for Australian industry, which is increasing the level of investment in new physical and intellectual capital, and IT and software in particular.

 

Figure 13A. Industry averages, market sector

Source: ABS 5206

Figure 13B. Industry averages, market sector

Source: ABS 5206

 

 

The main features of these averages for each industry are:

 

Agriculture, Forestry and Fishing – hours worked and non-IT capital services detracted from growth, services and materials were the major contributors

Mining – non-IT capital services was the main contributor (reflecting the use of heavy machinery and equipment), and in 2022-23 were 59% of input costs. 

Manufacturing – hours worked and energy detracted from growth, hours worked and materials were the main contributors, and in 2022-23 materials were 48%) of total input costs.

Electricity, Gas, Water and Waste Services - hours worked and non-IT capital services were the main contributors. In 2022-23 the services cost share was 58%.

Construction - services, hours worked and materials were the main contributors

Wholesale Trade – services and hours worked were the main contributors

Retail Trade - services, hours worked and both IT and non-IT capital services were the main contributors

Accommodation and Food Services - services and hours worked were the main contributors

Transport, Postal and Warehousing Services – services and non-IT capital services were the main contributors

Information, Media and Telecommunication Services – services, IT and non-IT capital services were the main contributors. In 2022-23 the services cost share was 54%.

Financial and Insurance Services – services and IT capital services were the main contributors

Rental, Hiring and Real Estate Services – services and non-IT capital services were the main contributors

Professional, Scientific and Technical Services – services and hours worked were the main contributors

Administrative and Support Services - hours worked was the main contributor. As a labour-intensive industry labour services accounted for 59% of total input costs in 2022-23.

Arts and Recreation Services - services and non-IT capital services were the main contributors

Other Services – services, materials and non-IT capital services were major contributors

 

Conclusion

 

The ABS publishes productivity measures for the whole economy, the market sector, and each of the 16 industries in the market sector. This post reviews these measures and provides their data, discusses their purpose and application, and recent economic and industry performance. 

 

The most widely used productivity measure is labour productivity as gross domestic product per hour worked. This is the measure commonly used for international comparisons, which are often done by converting countries’ currencies to US dollars but should be done using purchasing power parity to adjust for costs. Using PPPs, the level of Australian labour productivity is in the middle of other OECD countries. The index has increased from 70 in 1995-96 to 100 in 2023-24, or 44%.

 

Labour productivity is also available as gross value added per hour worked for the 16 industries in the market sector. Market sector productivity does not include estimates for three industries that lack meaningful prices and volume measures of output. The industries are Public Administration and Safety, Education and Training, and Health Care and Social Assistance.  Because productivity cannot be accurately measured in the non-market sector, comparisons should be made using market sector statistics, and improvement in non-market productivity will have little effect on labour productivity growth for the whole economy.

 

The index for GVA per hour worked in the market sector has increased from 63 in 1995-06 to 101 in 2023=24, or 61%. The difference between the two measures is significant, for the whole economy as GDP per hour worked the growth in productivity has been less than for the market sector because the non-market sector has lower productivity growth than the market sector, and as its share of the economy has grown in the last few years productivity growth for the whole economy has slowed. Construction labour productivity has barely changed over the last 20 years. However, during the mining boom 2011-2016 labour productivity rose and then fell, following the increase and decrease in the volume of work done.

 

As well as labour, production requires capital, for example tools, equipment and buildings, and the combined labour and capital measure is known as multi-factor productivity (MFP). This is a more complex measure because it uses a depreciation rate to account for use and age of the existing stock of capital. There has been a slowdown in labour productivity growth between 2017-18 to 2021-22, but a small increase in MFP growth over that period.  

Capital productivity has been falling, in the majority of years since 1995-96 capital productivity decreased, and declining capital productivity is the fundamental reason Australia’s productivity growth has been so low. The capital stock has not been replaced and updated due to the low level of business investment, particularly in IT and software. 

 

Quality adjusted labour input indexes take into account years of education, levels of training, industry of employment, age and sex. These quality adjusted measures are provided for the 16 market sector industries, and for both labour productivity and MFP. In the five years 2018-19 to 2023-24, changes in the QUALI indexes were close to the hours worked index. Despite the common claim that there has been no growth in Construction productivity, over the five years to 2024 Construction was one of the best performing industries in quality adjusted labour productivity as hours worked.

 

Finally, there is the KLEMS productivity measure for market sector industries, with five input categories for capital (K), labour (L), energy (E), materials (M), and services (S). The KLEMS measure identifies the cost structure and the contribution of each input to an industry’s output growth. For most Australian industries the main contributor to growth has been services provided by other industries. Non-IT capital intensive industries are Mining, Electricity, Gas, Water and Waste Services, Transport, Postal and Warehousing Services, and Rental, Hiring and Real Estate Services. Agriculture, Forestry and Fishing, Manufacturing and Construction are industries with a high level of materials as inputs.  Financial and Insurance Services is the only industry with a high level of IT capital services. 

 

The low level of IT capital services in Australian industry is the major problem at a time of IT related innovations like artificial intelligence, cloud computing and reality capture. To improve the productivity of Australian industry, the single most important requirement is increasing the level of investment in new physical and intellectual capital, and IT and software in particular. Industry policies should be targeting a significant increase in capital investment in the short-term, while policies targeting longer-term issues like innovation and R&D are developed. 

 

 

 

[1] The ABS methodology is in Chapter 19 of the Australian System of National Accounts: Concepts, Sources and Methods is on productivity measurement. 

 

[2] A 2023 post analysed construction productivity during the mining boom here https://gerarddevalence.substack.com/p/the-long-cycle-in-australianhtml?r=wtchb

 

[3] The ABS interpolates workforce compositional changes from the five yearly census data. Workers are grouped by education, age, and sex. For education there are four categories: Unqualified, Skilled Labour, Bachelor Degree, and Higher Degree. For age there are five categories: 15 to 24 years, 25 to 34 years, 35 to 44 years, 45 to 54 years, and 55 to 64 years. For sex there are two categories: Male and Female. The skills and education of the Construction workforce were detailed in

https://gerarddevalence.substack.com/p/the-changing-composition-of-construction?r=wtchb 

[4] Measuring the capital stock and estimating capital productivity is complex. It relies on a ‘rental price’ for capital, a long-term discount rate used to estimate capital services as an input to production by weighing different types of productive capital stock adjusted for efficiency loss and retirement. The ABS uses a discount rate of 4% based on long-term bond rates adjusted for CPI. Estimates of capital services growth and MFP are very sensitive to the rate of return used.

 

[5] The intellectual and physical capital stock of Australian industry in 2022-23 was analysed in https://gerarddevalence.substack.com/p/investment-in-physical-and-intellectual?r=wtchb