THE STRENGH OF LOOSE TIES IN CONSTRUCTION
I enjoyed Anna Dubois and Lars-Erik Gadde's paper on "The construction industry as a loosely coupled system," which brilliantly sheds light on the hidden forces shaping productivity and innovation in the construction sector—or, the lack of it.
In Australia, the construction sector is responsible for a million full-time equivalent workers (directly) and creates about $450 billion in value throughout the supply chain. Given its significant impact on the economy, it's worth taking a closer look at why the industry remains fragmented and resistant to modernisation or advancements that have transformed other sectors, such as mining, agriculture and manufacturing.
Dubois and Gadde analyse the construction industry through the lens of "a community of practice"—a network where companies develop shared, unspoken norms. This approach reveals the underlying reasons behind the industry's fragmentation and resistance to change, offering valuable insights into the complex dynamics at play.
One telling sign is costs.
While industries like manufacturing have seen steady declines, construction costs (both hard and soft costs) continue their climb, seemingly immune to the forces of progress or technological changes.
The paper’s authors trace this stagnation to ephemeral or loose couplings [1] among firms within each project. These fleeting partnerships enable short-term efficiency on a project but forestall long-term learning, innovation, and integration. Integration [2] that’s desperately needed to unlock the productivity gains that tend to push costs downwards.
And I think this is an important point outsiders miss when they analyse the construction sector, or try to make a play in it (think VC-funded tech start-ups). We see this when well-funded companies come in hot, with not-so-humble visions of revolutionising the sector.
We saw the rise and fall of Katerra (after raising US$2 billion in funding).
We saw the rise and fall of WeWork (valued at US$47 billion at its peak in 2019).
We saw the rise and fall of Zillow Offers.
And we saw the rise and fall of Google’s US$1.3 billion Sidewalk Labs Toronto precinct. Although this one had a host of privacy concerns.
Here’s the paradox enveloping the construction sector:
The sheer complexity of each building can promote coordination while at the same time, hinder transformation.
For context: a typical house contains about 3,000 parts, an apartment building upwards of 30,000. These parts need input from the hands of trades people, building material companies, builders, surveyors, engineers, architects, planners, developers, and financiers.
All separate companies.
All with varying macro goals.
But all share micro objectives—at a project level.
Project activities are also sequential.
Decisions and actions depend on each other. This, along with other risks such as abnormal weather, unforeseen site conditions, input price volatility, material availability, equipment breakdown and much more, add another layer of complexity.
In a way, it is impressive anything gets built at all.
Take a (hypothetical) new residential development, for example. The construction project involves multiple stakeholders, including architects, planners, consultants, engineers, contractors, and suppliers. Each stakeholder operates independently, with their own goals and processes. The architects design the building, the engineers ensure structural integrity, the contractors manage the construction, and the suppliers provide materials (bricks, timber, insulation, tiles, taps, pipes, appliances, etc).
During the project, the site manager coordinates the activities of these stakeholders. His primary job is to ensure each task is completed on time and within budget. The site manager's role is crucial in maintaining tight couplings [3] at the project level, ensuring the work of each stakeholder is integrated seamlessly.
However, once the project is completed, the temporary coalition of stakeholders dissolves. The knowledge and experience gained during the project are not systematically captured or shared with other projects. This lack of long-term integration and knowledge sharing hampers innovation and learning across the industry.
But, but, but: this only occurs at a macro level.
Manufacturing something equally complex, like a car with 30,000 components (fewer for electric vehicles), achieves far greater efficiency through tighter integration.
In construction, a "community of practice" emerges to manage the chaos, but it does so by simultaneously maintaining the status quo.
Dubois and Gadde trace this dynamic from design to final assembly, whereby we develop a pattern:
Knowledge flows freely within each project but struggles to percolate further. Incentives pull firms together for each job, only to dissolve those bonds upon completion. Leading back to inefficient procurement, inefficient bespoke contracts, and inefficient risk allocation.
This, in turn, brings us back to a super conservative decision-making — suspicious of any novel or innovative concepts or techniques.
And so it goes, the industry cycling endlessly without necessarily advancing.
Or as they put it:
“The particular pattern of couplings favours productivity in projects, while innovation suffers. These characteristics have made the industry as a whole lag behind other industries in terms of traditional performance measures. … The strong project focus makes coordination in other dimensions difficult or even pointless. Each project is considered to have a life of its own – without either history or future”
We can therefore establish that:
New construction technologies struggle to go mainstream not because they are ineffective per se, but because of the underlying permanent network of loosely coupled firms.
It’s not all bad either.
While the couplings slow down the absorption of new technologies, they have five key benefits.
Let’s look at them:
Localised adaptation: Loose couplings allow individual projects and firms to adapt to local conditions and unique requirements without impacting the entire sector.
This flexibility enables construction teams to respond quickly to site-specific challenges.
This also helps contain losses in small locations so that when one player falls or wobbles, it doesn’t lead to catastrophic ripple effects we see in the fintech space.
Here’s an example: A project team working on a historic building renovation can modify their approach to preserve the architectural heritage, deviating from standard practices
Buffering mechanism: Loose couplings help isolate problems, preventing them from spreading throughout the entire network.
If one project experiences delays or issues, it is less likely to disrupt other projects or firms.
Here’s an example: When a specific trade, like plumbing, encounters challenges on one project, it does not necessarily impact other ongoing projects by the same company.
Sensitivity to environment: With loose couplings, the construction industry can maintain a high degree of sensitivity to local environments and client needs.
Each project can tailor its approach to the specific context without being constrained by rigid, system-wide standards.
Here’s an example: A residential project can adapt its design and materials to meet the specific environmental or cultural preferences and needs of the local community.
Variation: Loose couplings allow for a diversity of approaches and solutions across different projects.
This variety can foster innovation, as firms experiment with new methods and technologies within the confines of individual projects.
Here’s an example: A firm working on a high-rise building can pilot a new prefabrication system for that specific project before considering wider adoption.
Autonomy and self-determination: Loose couplings provide firms and project teams with a high degree of autonomy and self-determination.
This, I think, is the most vital benefit.
This independence can boost motivation and enable firms to respond quickly to challenges without waiting for approval from a central authority.
Here’s an example: Subcontractors can independently determine the most efficient way to complete their scope of work, leveraging their specialised expertise.
In all five cases, the benefits are clear at the project level. Not so much when you zoom out.
While loose couplings offer these benefits, the authors also argue that they can hinder long-term learning, collaboration, and industry-wide transformation. The key challenge is to find ways to preserve the advantages of loose couplings while also enabling the deeper integration and knowledge sharing needed to drive sustained innovation and productivity growth.
Yes, but:
Critics might argue that tighter integration, as seen in manufacturing, could solve these issues. They might point to the automotive industry, where integration has led to significant efficiency gains.
However, this comparison overlooks the unique challenges of construction.
Unlike cars, buildings are not mass-produced in controlled environments. Each construction project is unique, influenced by local conditions, regulations, and client needs. The last factor is often overlooked, despite being a vital one. Therefore, while tighter integration might work in theory, in practice, it could stifle the very flexibility and adaptability that are crucial to the construction industry’s success. This, I suspect, is also why technologies such as modular construction struggle to take off, despite their obvious efficiencies, and productivity gains. Which handicaps programs like MMC (Modern Methods of Construction) even before they start.
Moreover, the push for integration often comes with significant upfront costs—and risks. The failures of Katerra, Zillow Offers, and Sidewalk Labs, for instance, highlight the dangers of imposing a one-size-fits-all solution on a complex, fragmented industry. Such companies tend to underestimate the importance of local knowledge and the intricate web of relationships that underpin successful construction projects.
A note from Makao
Once a project is completed, the temporary coalition of stakeholders dissolves.
The knowledge and experience gained during the project are not systematically captured or shared with other projects, at least not directly. This lack of long-term integration and knowledge sharing hampers learning across the industry.
This is what makes innovation in the construction sector so difficult, which then translates to rising costs over time. There are, of course, other contributing factors, but these hidden forces have a potency that is rarely discussed or addressed at a deeper level.
Again, I must repeat, while the construction industry’s loose couplings present challenges, they also offer unique advantages. The key is not to eliminate these couplings but to find ways to enhance collaboration and knowledge sharing within this framework.
By doing so, the industry can achieve the best of both worlds: the flexibility to adapt to local and on-site conditions and the efficiency gains from innovation.
In a future article, we’ll look at potential solutions.
Go deeper:
In the meantime, see ACA’s Nailing Construction Productivity report, for a more exhaustive list on how to improve productivity in the construction sector.
FOOTNOTES
[1] These ephemeral, or loose couplings, mean that companies within the sector are independent, flexible, and adaptable to change. The connections between them are weak, allowing for autonomy. Knowledge sharing might occur at a project level, but this fails to transfer to the broader industry level.
[2] Integration here refers to the process of coordinating and combining various stages of production, from raw materials to finished products, to streamline operations, improve efficiency, and reduce costs.
[3] Of note: The construction industry exhibits a duality where loose couplings exist at the industry level, and tight couplings occur at the individual project level. This duality allows for flexibility and adaptability at the macro level while ensuring coordination and efficiency at the micro level. Another complexity to add into the mix.
Comentários