Did the Productivity Commission miss an opportunity to change the debate?
The Australian Productivity Commission has released a research paper on the residential building industry, with new data on its declining productivity and recommendations to improve it. The Productivity Commission (PC) is an advisory body that provides research and analysis to the Treasurer and, as an independent agency mainly staffed by economists, has a long history of advocacy for market-based solutions. It plays an important role in developing policies through inquiries done at the request of the Australian Government on policy or regulatory issues, does its own research across seven work streams, and has benchmarking performance monitoring responsibilities.
The research paper is titled Housing Construction Productivity: Can We Fix It? A team of eight people worked on it, there were three roundtables with the Urban Development Institute of Australia, the Housing Industry Association and the Australian Local Government Association, and over 40 organisations and individuals participated in the research. There are new estimates for a set of proxy measures for dwelling construction productivity that ‘encompass the entire homebuilding process, from site preparation and project management to the installation of fixtures and fittings’.
This post looks at the key outcomes of the research. These are the five issues that have contributed to falling productivity and the seven ‘reform directions’ the paper has identified. How significant are these outcomes? How should the reforms be implemented? What other industry issues or characteristics have not been considered relevant or important?
Productivity Performance
Because there are no ABS estimates of productivity for Australian dwelling construction the PC developed a set of labour productivity proxies. This involved estimating outputs and labour inputs from ABS construction statistics across the two industries of residential building construction (builders and developers) and construction services (tradespeople that build the dwelling). The claim is these ‘productivity proxies capture the broad underlying changes in productivity in the sector’.
Figure 1 shows the PC’s measures of productivity in the dwelling construction industry declined between 1994‑95 and 2022‑23 as:
· The number of new dwellings built per hour worked declined 53%;
· Dwelling construction gross value added per hour worked declined 12%; while
· Labour productivity in the broader economy increased 49%.
Figure 1. Housing productivity
The average time taken to complete new housing has increased significantly. ‘In 2023‑24, the average time to complete a single detached house was about 10.4 months, up from about 6.4 months a decade earlier. The average time taken to complete new townhouses rose from about 9.4 months to 12.9 months, and for new apartments, from about 18.5 months to 27.8 months (though this only covers apartments in New South Wales, Victoria and Queensland)’. Higher density construction performed better than houses according to the PC, in Figure 2.
Figure 2. Houses compared to higher density
Methodological issues with these proxy measures include: indexation and the choice of 1994-95 as the base year; lack of data on annual growth rates; how accurately construction services are accounted for in hours worked; and the use by the PC of their own deflator. The ABS productivity measures for the whole of construction show no growth but nothing like the decline the PC has found. In a 2022 post on construction productivity for the ABS sectors of Building, Engineering and Construction services I found no significant fall in GVA per person employed between 2007 and 2021, in Figure 3.
Figure 3. Australian construction productivity by industry sector
Factors Driving Poor productivity
The PC discuses five issues that contribute to falling productivity in housing construction. The PC’s issues are:
1. Complexity and slow approvals –the development and construction approval process is complex, with major development projects such as new housing estates and apartment complexes taking ten years or more. After approval delays can continue as projects need construction certificates and infrastructure connections. Because construction is highly sequential, delays and disruptions can create ‘cascading failures’, which push up costs.
2. Fragmented - construction is done by small firms and individual subcontractors and is one of the least concentrated industries in Australia. The combined market share of the largest four firms is the lowest of any sector, at 12% in 2017, and the average residential building construction firm employs less than 2 people. ‘To a degree this is a function of the development and building process and subcontracting’, however:
· Regulations differ across geographic areas, which makes it hard to replicate a development or construction process across the country and for successful firms to scale up;
· Building regulation tends to be project‑based rather than firm‑based, requiring new regulatory approval for each new project. In other industries, regulations are a fixed cost that encourages firms to scale up to spread those costs over a larger revenue base. In construction, if a builder scales up, regulatory costs scale up proportionally, reducing or even negating the benefits of expansion;
· Project risk management and a regulatory focus on the builder as head contractor leads to highly disaggregated specialisation and outsourcing of work.
3. Lack of innovation – innovation activity and spending in construction is low compared to other sectors. Only 35% of all construction firms are ‘innovation‑active’ and the sector has been slow to adopt digital technologies and new processes like prefabrication. Estimates are less than 5% of total construction is prefabricated. ‘Low levels of innovation are due to fragmentation, industry culture, lack of direct benefits to firms from innovation and the ‘chilling effect’ of frequent regulatory changes’ [1].
4. Regulatory burden – to develop and construct housing requires compliance with layers of regulation on:
· Where housing can be built and state planning that affects which land is zoned for residential development;
· How housing should be built and the National Construction Code’s minimum standards on how buildings are constructed;
· What housing should look like, through local government planning requirements for setbacks and roof pitches;
· Other indirect regulations, such as environmental regulations; and
· The regulatory burden is increasing which ‘unambiguously increases the cost of development and construction, and ultimately the cost of housing for Australians’.
5. Workforce issues – the sector struggles to attract and retain some skilled workers because of ‘stagnating apprenticeship commencements and completions, restrictive and inflexible training pathways for trades, and competition for labour from public infrastructure projects in recent years (at least for the higher‑density housing sector)’. Regulatory settings contributing to low labour mobility are inconsistent occupational licensing accreditation across jurisdictions and limited pathways for migrants to join the construction workforce.
The PC uses the housing sector, housing construction and residential construction to describe the industry that the ABS calls residential building. In discussing the issues the paper frequently moves between total construction and residential building. This might be because data is often only available for the whole of construction, but it is not helpful as it obscures the points being made. Residential building is completely different from engineering construction and there are limited similarities with non-residential building, which together have a larger share of construction industry employment and output than residential building.
Reform Directions
To improve housing construction productivity, the PC identified seven reform directions:
1. Coordinated and transparent planning approvals and appropriately funded regulators – governments need to improve coordination in the planning approval process, and should consider establishing coordination bodies to speed up the process and address delays such as the Queensland State Assessment and Referral Agency. Governments need to adequately resource regulatory and service delivery agencies, ensure there is sufficient accountability, and set performance targets for planning approval decisions.
2. Review building regulations – governments should commission an independent review of the National Construction Code and its implementation by state and territory governments. A review should consider the NCC’s objectives, the regularity of code updates, consistency in implementation by governments’ in approvals, certification, compliance and enforcement, and impediments to innovation and local government rules that affect construction of dwellings.
3. State and territory governments should continue implementing ratings systems that improve information available to consumers about new and existing building quality to help create a price premium for higher quality builders and building.
4. There are opportunities to increase diffusion of technology across the housing construction sector, similar to the government‑funded extension services provided to agriculture.
5. Governments should consider the adequacy of public research and development funding for housing construction to improve productivity performance. And better information, including rating schemes, could improve building quality.
6. Governments should continue to reduce unnecessary regulatory impediments to greater uptake of modern methods of construction in housing construction, including prefabricated and modular construction.
7. Improve workforce mobility and flexibility – governments should continue removing occupational licensing exemptions and conditions and work towards national consistency, address barriers to migration to improve labour supply, and improve training pathways (including microcredentials) and support for apprentices.
Implementation is Missing
Discussion on implementation of the reform directions is not included in the paper, a serious oversight that leaves the recommended directions rudderless. For example:
· How to reform planning if state governments are not prepared to make difficult, unpopular decisions on allowing development while restricting appeals and reviews of planning decisions? How does the PC propose coordinating planning and approvals across the 537 local councils in Australia?
· Who will finance and who should undertake the independent review of the National Construction Code and its implementation by state and territory governments? Over the last decade the Commonwealth Government has actually been reducing funding for the Australian Building Codes Board, and the PC could have usefully recommended increasing funding;
· What about the relationship between building defects and the National Construction Code? The reason there is a Code is to set standards and measure compliance, and state variations typically reflect different conditions [2]. The states are also responsible for the regulators that are dealing with the plague of defects in apartments built over the last ten years, their programs to replace flammable cladding that was illegally installed, and they all have different licensing criteria for builders. How would ‘consistency in implementation by governments’ in approvals, certification, compliance and enforcement’ be achieved?
· How can funding for public research and extension services be increased when competing with priorities like defence, health and aged care?
· What are the ‘unnecessary regulatory impediments to greater uptake of modern methods of construction in housing construction’? if these are not identified how can they be reduced? Why didn’t the PC draw on the 2024 research on regulatory barriers by Swinburne University and the Housing Institute of Australia or the 2023 prefabAUS industry roadmap for example?
Other Industry Issues
A number of other industry issues or characteristics have not been considered relevant or important. Not discussed in the research paper are:
· The volatility of approvals, commencements and demand means the building cycle from year to year can vary by 10% or more – one solution is to allow builders a tax free sinking fund so in good years money can be saved to be used to finance projects during downturns;
· Industry contractual relationships – subcontracting is flexible and a method to manage costs and risk but direct employment has a smaller span of control and is more efficient, however, builders will only increase direct employment if there is continuity of work (see above);
· The cyclical boom-bust nature of residential building also makes increased modular and prefabricated housing and successful industrialisation is less likely than in manufacturing;
· Improving project management through training and education would significantly improve productivity, particularly for smaller firms. This should include access and training on digital platforms and AI enhanced systems;
· The reluctance of banks to finance modular and prefabricated houses - although this is starting to change the traditional method of financing and standard contracts do not work for prefab;
· The lack of standards and an industry quality assurance accreditation system for modular and prefabricated housing - there is work being done on these but nothing official so far;
· The lack of a building products compliance system – the exception is Queensland’s 2017 Non-Conforming Building Products law that creates a chain of responsibility and requires the supply chain to take responsibility for building products being compliant and fit for purpose;
· The potential of digital platforms and AI enhanced systems to increase productivity;
· Automated planning assessment and compliance with codes.
Vocational Education and Training
The PC focused on occupational licensing, but there is no discussion on vocational education and training (VET) beyond noting the need to improve training pathways. This is bizarre, because beyond attracting and retaining new entrants are many issues with access, curricula, articulation and the new skills needed with digitisation and automation. To be fair, the PC references their 2020 review of the National Agreement for Skills and Workforce Development and the 2024 Department of Employment and Workplace Relations report Skills For Tomorrow: Shaping The Future of Australian Apprenticeship, Strategic Review of The Australian Apprenticeship Incentive System. Nevertheless, in this paper the discussion is limited to a few points on occupational licensing, although there have been other reports with detailed recommendations on Australia’s VET sector that could have been built upon.
The 2019 review of Australia’s VET by Steven Joyce, a former New Zealand Minister for Tertiary Education, Skills and Employment made 71 recommendations. The key issues identified were:
· Variations in quality between providers and the relationship between the regulator and providers;
· A cumbersome qualifications system that is slow to respond to changes in industry skills needs;
· A complicated and inconsistent funding system that is not well matched to skills needs;
· A lack of clear and useful information on vocational careers for prospective new entrants;
· Unclear secondary school pathways into the VET sector and a dominance of university pathways;
· Access issues for Aboriginal and Torres Strait Islander Peoples and second chance learners seeking skills that will help them obtain and stay in meaningful work.
The Joyce review concluded ‘the sector needs to be more active in emerging skills areas in order to be seen as a more modern method of education’ and the essential role for VET is training new entrants and upgrading the skills of current workers as digitisation and automation spread through the economy, including support for second chance learners needing foundation language, literacy, numeracy and digital skills.
A more recent 2024 review of NSW VET made 21 recommendations, including better governance, funding and pricing, with industry compacts, regional skills plans, microcredentials, updated curricula, and apprenticeship roadmaps. Other recommendations were for the government to employ apprentices, particularly in regional areas, provide support to VET teachers, simplify the complex and highly prescriptive approach to VET qualifications, expand pathways to entry for teachers, and invest in facilities.
How Significant Are The Outcomes?
None of the productivity issues in the PC paper are new and can be found in many other reports on the industry, and the recommended reforms are peculiarly limited. For context, below are the recommendations of four comparable reports on construction productivity that show how constrained the PC’s reform directions are, two in 2024 from the NSW Productivity and Equality Commission and McKinsey, and two older reports from the CIOB on the U.K. and the McKinsey Global Institute on the U.S.
The 2024 NSW Productivity and Equality Commission report Review of Housing Supply Challenges and Policy Options for NSW found barriers to housing supply included high construction and borrowing costs, capacity constraints in the construction sector, and bottlenecks in the development process, with over half of the 32 recommendations on planning. It recommended reforming planning to streamline the development process and reduce approval times, and reviewing the Design and Place State Environmental Planning Policy because ‘prescriptive rules’ on land block innovation. Other recommendations included education and skills training, business regulations and tax, improving infrastructure and transport, replacing stamp duty with a land tax, establishing an Urban Development Program to report on the housing market and a housing supply council to advise on housing targets, and incentives for local government to meet targets. It argued for non-regulatory approaches wherever possible, and avoiding excessive regulation. The NSW report’s planning reforms were controversial because they would lower the quality of high density apartments, but the range of recommendations show a more ambitious agenda than the PC’s conventional and well-known seven reform directions is both possible and required.
A 2024 McKinsey Insights paper called Construction Productivity is No Longer Optional identified seven issues holding back measurable productivity gains and made five recommendations: adopt project steering with production rate metrics instead of conventional project management; nurture a supplier ecosystem across projects; upskill project staff; scale initiatives across project portfolios; and apply technologies such as generative AI to streamline and accelerate engineering, procurement, and construction.
That report followed up on The McKinsey Global Institute’s 2017 Reinventing Construction Through a Productivity Revolution, which was particularly blunt: ‘while other sectors from retail to manufacturing have transformed their efficiency, boosted their productivity, and embraced the digital age, construction appears to be stuck in a time warp … The industry is extensively regulated, very dependent on public-sector demand, and highly cyclical. Informality and sometimes corruption distort the market. Construction is highly fragmented. Contracts have mismatches in risk allocations and rewards, and often inexperienced owners and buyers find it hard to navigate an opaque marketplace. The result is poor project management and execution, insufficient skills, inadequate design processes, and underinvestment in skills development, R&D, and innovation’. Changes are needed in seven ‘key areas’: regulations; contractual framework; design and engineering; procurement and supply chain management; onsite execution; digital technology; and upskilling the workforce. Changes in these areas would also benefit Australian housing productivity.
A 2016 CIOB report Productivity in Construction had the subtitle Creating a Framework for the Industry to Thrive. This overview of the issues in the UK says a number of times the problems are well known, as are solutions that have been proposed. Some of the recommendations that would also apply to Australian housing were:
· Create construction innovation and excellence hubs and promote, through incentives, clusters of construction-related businesses in key regions;
· Improve leadership and behavioural understanding and shift the emphasis in construction thinking and policy making from industry processes to behavioural aspects of construction;
· Develop new business models and financial models because the business models currently used inhibit progress on productivity;
· Get better data and measures of construction productivity, and compile built environment satellite accounts, similar to those produced for tourism, to capture inputs from related professions, materials, manufacturing, plant and machinery suppliers [3].
· Invest more heavily in attracting new entrants to the industry and improve the skills of the existing workforce. Investing in management professionals has great potential to improve construction productivity.
Conclusion
By the middle of the twentieth century residential building had developed the materials, methods and processes largely in use now. While there have been changes like more electrical appliances and offsite manufacturing of trusses, windows, doors and cabinetry, the structure and services like electricity, water and plumbing in a 1960s dwelling are those found in a new build today. Houses are often larger and apartments smaller than a few decades ago, but how they are procured and built, and what they are made of, has not substantially changed in decades. Fundamentally, that is also why the level of productivity has not changed [4].
The PC has focused on regulation and planning as the main issues, but these are just two of the factors that affect onsite productivity, and arguably skills, technology and project management are more important. Also, planning and zoning decisions have no effect on other important supply side issues such as the construction and finance costs for new housing, constructing the infrastructure needed for new developments, and the rate of conversion of approvals into commencements by developers.
A major problem is not what is in the PC’s research paper, but what has been left out. Without question the five issues driving productivity are important, however these are long-standing and well-known, and the paper says nothing new about them. Similarly, the seven reform directions have been suggested many times by many different industry analysts and organisations. Missing is any discussion on implementing these reforms, or how barriers to their implementation might be addressed, which undermines both the usefulness of the research and the relevance of the recommended reforms.
There are several other industry issues not discussed, like the volatility of the building cycle, the lack of standards for offsite manufacturing, and product compliance laws. It is disturbing that the PC does not appear to understand why there is a construction code and how it works. Improving productivity through better project management and reform of the VET system are also overlooked. There is no discussion of digitisation and automation in residential building. Apparently, as far as the PC is concerned, digital tools and platforms, AI enhanced systems and automated planning and code compliance checks are not relevant to productivity.
The PC’s recommended reforms are peculiarly constrained and limited, and the paper is a missed opportunity for introducing new ideas and a failure of imagination. Although an agency like the PC is not expected to be imaginative, residential building is stuck with a well-established system of production that will only change when and if there is a clearly superior method of delivering quality housing that is also profitable. Tinkering with regulation, the NCC, planning and approvals processes, and occupational licensing might make a difference at the margin, but will not deliver the big improvement in housing productivity that is required.
*
[1] Innovation in Australian construction was discussed in a 2024 post, incremental innovation in this post, and innovation and procurement in this post.
[2] The PC does not seem to understand how the construction code works. For example, in Far North Queensland houses have to withstand a category 5 cyclone with roof tie downs, window shutters and a safe room, but in rural Victoria bushfires are the main issue and houses have to be of non-combustible material, have cavity barriers and provision for escape. See also https://theconversation.com/better-building-standards-are-good-for-the-climate-your-health-and-your-wallet-heres-what-the-national-construction-code-could-do-better-166669
[3] I have argued for a built environment satellite account for many years, including the post here and a 2019 journal article here.
[4] With current technology, tools and processes, construction may be close to the efficiency frontier, see Is Productivity Growth in Construction Possible?