Thursday, 22 February 2018

Australian Construction Productivity in the 1990s

Construction Industry Productivity in Australia




In 1992 and 2003 I contributed to the research published on construction industry productivity by two government inquiries. The first was the Royal Commission into Productivity in the Building Industry in NSW (1991-1992), and the second was the Royal Commission into the Building and Construction Industry for the Commonwealth Government (2001-03). These posts have the relevant parts of those publications, which are reviews of the then current data and research.

The first post is from a paper that covered the 1980swas originally published as Productivity and the Australian Construction Industry by the Royal Commission into Productivity in the Building Industry in NSW. 

This second post covers the 1990s and was an Appendix in a Discussion Paper from the Royal Commission into the Building and Construction Industry for the Commonwealth Government.

A third paper in this series is a conference paper of mine from 1999 called Recalculation of Australian Construction Productivity which looked at changes in output per person in the special trades. The document can be downloaded from a Dropbox file here.


*

A3.1     Measures of productivity

Productivity refers to the relationship between the quantity of goods and services produced and the quantity of resources employed in producing those goods and services.  The economic definition of productivity is the ratio of output over input, and in very general terms can be described as the efficiency of production.  This definition focuses on the quantification of inputs and output.

Box A3.1           Measures of productivity
Within the economic definition there are three productivity measures used. 

The commonest measure is labour productivity, which refers to the output-to-labour ratio.  This is typically measured as output per person employed or per hour worked.  Labour productivity can rise through improvements or changes in the production process, technical knowledge, managerial expertise, greater use of equipment or machinery, and so on. 

Capital productivity is measured as output per unit of capital.  There are significant problems involved in determining the amount of physical capital (plant and equipment, buildings and structures) in use at any point in time because investment is creating new assets and wear and tear is depreciating existing assets.

The wider concept that combines labour and capital productivity measures is total factor productivity (TFP) or multi-factor productivity (MFP).  This combines the labour and capital inputs used in the production process by an industry and measures their respective contributions to output. 


Australian Data

In Australia, the ABS provides indexes of output per hour worked and per person employed by industry for the market sector.  The average increase in labour productivity in the market sector of the Australian economy between 1991-92 and 1999-2000 was 2% a year.  Over the same period labour productivity in the construction industry increased by 1% a year, with wide variation in the year-on-year rate.  As can be seen over the longer period 1986 to 2000 the construction industry managed to lift productivity by only 3%. These indexes are very sensitive to differences in the rate of change between output (measured as industry gross value added) and input (hours worked).  This difference explains the drop in construction labour productivity in 2000-2001, where the growth in hours worked was greater than growth in output.  The public sector (administration, defence and education) and property and business services have no estimates of output independent of inputs and productivity therefore cannot be estimated for these industries.  The rest of the economy is known as the market sector.

Table A3.1       Labour productivity:  indexes of gross product per hour worked, by industry

1986
2000
2001
Agriculture, forestry and fishing
66.9
100.0
99.7
Mining
44.2
100.0
107.6
Manufacturing
64.6
100.0
101.3
Electricity, gas and water supply
35.9
100.0
98.0
Construction
96.6
100.0
85.4
Wholesale trade
71.7
100.0
104.2
Retail trade
87.6
100.0
100.9
Accommodation, cafes and restaurants
109.1
100.0
93.8
Transport and storage
74.0
100.0
97.1
Communication services
34.0
100.0
104.8
Finance and insurance
57.0
100.0
102.0
Property and business services
0.0
0.0
0.0
Government administration and defence
0.0
0.0
0.0
Education
0.0
0.0
0.0
Health and community services
83.3
100.0
96.5
Cultural and recreational services
109.7
100.0
110.6
Personal and other services
0.0
0.0
0.0
All industries
80.2
100.0
100.1
Source: Australian System of National Accounts, ABS, 2000-01, Cat No. 5204.0

While the ABS does not produce industry level productivity indexes, the Productivity Commission has produced estimates using ABS data on output, hours worked and capital input by industry.  In a Productivity Commission research report on employment and productivity, Barnes et al. (1999) found that MFP growth for construction between 1978-79 and 1995-96 was negative, at -0.2% a year, compared to market sector MFP growth of 1.2% a year.  The Productivity Commission released revised MFP estimates for the market sector of the economy in December 1999.  The new release made significant revisions to estimates for earlier years, due to a new ABS methodology for estimating productivity based on an improved and more complex methodology for deriving capital inputs.  For the construction industry the revision lifted productivity growth from negative to positive, at 0.3% a year.  This was still well below the level of the market sector as a whole, at 1.0% a year growth.

Table A3.2       Growth in multifactor productivity by industry sector, percent per year
Industry
1985-86 to 1998-99
Agriculture
2.7
Mining
1.5
Manufacturing
1.3
Electricity, gas and water
3.0
Construction
0.3
Wholesale trade
1.9
Retail trade
-0.1
Accommodation, cafes and restaurants
-1.6
Transport and storage
1.2
Communication services
5.3
Finance and insurance
1.8
Cultural and recreational services
-2.1
Market sector
1.0
Source: Productivity Commission, 2000.  Update of Productivity Estimates: Industry Sector Productivity, 1998-99.

The Productivity Commission also recalculated the capital-labour ratio in producing these estimates, and found that construction had a ratio of 2.1 (two units of capital to each worker) compared to a ratio of 3.5 for the market sector as a whole.  Other industries with low capital intensity (ratio under 2) were agriculture and transport.  The high capital intensity industries (ratio over 5) were mining, manufacturing, utilities and retailing.

Recent data at the industry level for the United States is similar to Australia.  Gullickson and Harper (1999) give estimates of MFP by industry for the US.  The table shows the estimates of multi-factor productivity growth rates for the industries in the US private business sector.  The two sets of multi-factor productivity estimates are based on industry output series from the BEA and from the BLS.  These trends differ by more than 0.5% for only one industry (utilities).  There are negative multi-factor productivity trends for construction and for oil and gas.


Table A3.3       Estimates of US multi-factor productivity trends in industries, percent per year.
Industry
Bureau of Labor Statistics
Bureau of Economic Analysis
1963-77
1977-92
1977-92
Farms
0.8
1.8
1.7
Mining
-1.1
-1.2
-1.5
Construction
-0.7
-0.4
-0.9
Manufacturing
0.6
0.5
0.7
Transportation
1.9
0.4
0.2
Communications
2.4
0.4
0.9
Electric, Gas, and Sanitary Services
0.4
-0.3
-1.1
Wholesale Trade
2.6
2.1
1.0
Retail Trade
1.7
0.3
1.2
Finance, Insurance, Real Estate
0.7
-1.2
-1.3
Hotels and Other Lodging Places
1.5
-3.5
-0.1
Personal Services
2.3
0.7
0.0
Amusement, Recreation Services
-0.1
-0.3
1.7
Source: Gullickson and Harper 1999: 56.


International Comparison of Productivity

Despite the difficulties in the measurement of productivity in general, and for the construction industry in particular, some studies have found that the level of productivity in the Australian construction industry compares well with overseas industries.  Significantly, both these studies are for periods (1990 and 1995-96) when residential building was at a low point in the cycle.  Because residential building is the most labour intensive part of the industry, this lifts the overall level of productivity by improving the output to hours worked ratio.  An OECD study by Pilat (1996) found that Australian construction labour productivity (output per person based on National Accounts data converted with 1990 purchasing power parities) is above that of the US, Japan and all European nations, only Canada had higher construction output per person.

Table A3.4       Index of construction output per person, 1990 (USA=100)
Country
Labour Productivity
United States
100.0
Japan
79.7
Germany
75.5
France
83.3
Italy
84.2
United Kingdom
62.0
Canada
148.2
Australia
103.0
Austria
99.4
Belgium
90.1
Denmark
65.8
Finland
94.4
Iceland
69.1
Luxembourg
61.6
Netherlands
69.9
Norway
68.5
Portugal
38.5
Spain
86.8
Sweden
75.2
Turkey
49.1
Source: Pilat 1996: 21.

The McKinsey study of Australian construction was also positive in its conclusions (Lewis et al 1996).  The study found the Australian construction industry had made a number of positive changes since the 1980s, including the development of more harmonious working relationships and the introduction of multi-skilling.  The effect had been dramatic: average productivity is now close to the world’s best, and employment performance has been just as impressive.  Labour productivity, at 95% of best practice, was only just behind the United States, and on a par with Germany’s.  In addition, construction costs were low, slightly less than those of the United States and France and far below those of Germany, Sweden, and Japan.  The main drivers of Australia’s high productivity were intense competition in the industry, which encouraged the widespread transfer of innovative production processes, and improved industrial relations that increased labour flexibility and minimized time lost through industrial disputes.  Despite that, many more firms need to adopt international best practices in construction techniques and processes.  McKinsey found Australia owed its success to greater demand for housing and infrastructure and lower prices.  To raise productivity further and stimulate growth, McKinsey recommended the industry maintain or increase the rate at which it develops and transfers innovations, and continue to explore opportunities to expand into overseas markets.

Table A3.5       Construction compared internationally
Country
Productivity
Employment
United States
100
-0.1
Australia
95
-3.5
Germany
95
-9.9
France
85
-12.9
Japan
71
na
Sweden
67
-7.6
Note: Productivity is value-added per hour worked, index US = 100.  Employment is jobs created per thousand working age population.
Source: Lewis et al 1996: 99.

McKinsey concluded construction firms should seek collaborations with their large customers, especially industrial firms, to reduce capital expenditure and improve project economics in such fields as mining and new factory development.  Close working relationships of this kind have the potential to boost both capital productivity and profitability.