It's interesting to compare this to the software industry, which seems to share a lot of characteristics with construction (high dollar value, each project has unique conditions, many sequential dependencies, mistakes are often difficult to fix though maybe not as difficult as construction). My impression is that people come up with new/better ways of building software much more frequently.
I think this is good evidence for your idea that a less-fragmented construction industry might be more innovative: a lot of new software tools were pioneered inside one of the few largest tech companies.
(A large part of this is also obviously that the construction industry has existed for >100x longer than the software industry, so there's much more low-hanging fruit in software. Another relevant difference might be that it's easier to experiment with new software tools in a low-stakes way via side projects; I'd guess that the minimum practical size for a software project is much smaller than for a construction project?)
"Maybe not as difficult as construction"???? Not close. Enormous life expectancy differences - working in a 50-year old building is not uncommon, working on 50 year old software is not common. I could do a long list of differences. I have experience in both domains.
Building software and buildings are entirely different beasts. In software you are typically delivering something that is intangible, e.g. productivity, user engagement, etc. and can be easily modified, so can rapidly iterate/ modify the processes and products delivered. In addition a product is never really done, whereas a building has to match spec, has a clear spec, is highly regulated and has a clear finish. Unfortunately, companies try applying project-management processes and metrics designed for building physical objects, and as a result fail even more spectacularly.
Robert Gordon finds that the construction industry recorded negative productivity growth from 1995-2005 in both the US and EU-10, owing entirely to declining multi-factor productivity, i.e. the residual that's supposed to capture output from new ideas and innovation.
Construction productivity is notoriously hard to measure, but stipulate that these estimates are at least directionally correct. What drove the decline in construction MFP? Would you consider doing a post on why and how construction innovation is not just hard, but may even be somehow regressing?
To achieve significant change (and risk reduction!) you need to vertically integrate and internalize. SpaceX is my favorite example of getting a stagnant industry into innovation mode. I think the problems of the Space Industry before SpaceX are similar to Construction now. And it's also massive, bold and complicated undertaking to establish a successful "BuildX". But the upside should be even more massive than in Space.
You might enjoy https://philpapers.org/rec/SCHGEA, which looks at a very similar set of constraints and processes in evolutionary biology. Under that model, I would anticipate the most innovative areas of the construction industry to either be, as you said, ones that improve something without changing other parts, or to be "finishing touches": how plumbing gets installed should change rarely, but the quality of tile in bathrooms could improve at a reasonable clip, because it is harder for failures at the end of a process to cascade substantially.
I LOVE your newsletter!!! Please keep it coming... It so much helps to think of system-level changes needed to drive this great industry forward to productivity and sustainability <3
It seems like the nature of cost overruns and their distribution is maybe related to the bidding process as well? Like if you see something that might decrease your costs, you assume everyone else sees it as well, so you lower your price accordingly. On the other hand, if you see something that might increase your costs, you're gonna look real hard at how likely that thing is to be an issue, and you might bid optimistically, hoping it's not a problem (knowing that it's not the worst thing in the world to come in over budget for "unexpected" reasons).
And I don't know exactly how all this works (like how they calculate cost exactly), but if a project ends up being cheaper than expected, the contract price doesn't usually get lowered, does it? So we'd expect the distribution to be skewed to the right no matter what, just because construction companies generally don't give price breaks if it turns out, after the fact, that their costs were low.
Very important detail! missed that. And also the final client has no interest in publicizing how the project was completed for much lower than they contracted for. Possibly there actually are projects that make double digit margins - but few outside the construction company would know.
I work for a large General Contractor and YES, much of this data is likely linked to the bidding process. It seems like this data is mostly being pulled from public projects. In the US this usually means the lowest qualified bidder gets the job and they only bid what is shown on the drawings or in the other provided documents. Also, these are usually lump sum, meaning if there are savings, the GC gets to keep them. It also means that if they missed something, they pay for it. Anything not shown or unknown at bid time would be an added cost to the client.
I think that is good insight, again it comes back to risk internalization, forecasting and being stuck in a sort of bad equilibrium.
There is also the possibility too that either culturally, cost benefits wise or myriad of other reasons, that people involved in the process prefer it that way. For unspecified reasons, they might not even understand.
On the cost side of things, there is something similar that happens with low-cost air travel. Whereby you have many stated preferences for more comfort, ease of access, better flying experience etc. But in actual actions, people will majority tend to choose the default low-cost option.
Now I can't necessarily say people are wrong to have this default in all situations, honestly, most people don't fly often enough to really evaluate the claims for the benefits of a higher price plane ticket. And it does seem the default low-cost option, in a situation where you suffer from a lack of information, either on the outcome, the claims of an airline or your own preferences might make sense. It's the most salient and clear piece of information.
I tend to travel relatively often, I through experience I found that choosing the more expensive flight that either leaves not too early or arrives in a destination not late, in order to either navigate to a hotel or get setup is worth paying extra for.
But honestly, old me didn't appreciate that benefit, only after much pain of paying 40%-30% less for a flight that leaves at 5 am did I realise for me the extra sleepiness and loss of productivity that subsequent day was just not worth it. And I didn't mind either the potentially extra wait or cost in the airport if I could leave at say 9.00 am instead. So from my own personal experience, my own cost-benefit calculations were essentially unknown to me, until I basically ran a bunch of experiments on myself.
The airline anecdote makes me wonder if something similar is happening too within the construction industry. Basically lots of plausible claims of benefits or possible alternatives, but nobody really knows their own needs, benefits or costs & there is a lack of real-life repeated experiments to see the differences.
Basically, we would need to run a myriad of experimental trial runs of builds and rebuilds under multiple scenarios for say residential housing in multiple areas to figure out the varied or common issues. Because my sense is many of the claims are all kind of plausible-sounding, but we just don't have enough purposefully done trials to know. Then repeat that for every area. Realise doing multiple builds and rebuilds of bridges might be absurd, but it does seem the only way to capture plausibly the potential fixes that would work.
Once you have all this practical information, you can use it to start a business, advocate for policy changes, develop a culture of prestige and status around your methods etc.
What we need is Construction Physics Build Laboratory :-)
To some extent, I believe insurance companies that issues builder's risk policies would fill some of this role. They receive data from general contractors of things that go wrong, new means/methods beings used, etc. for each project they insure. In many cases there will be projects similar in nature in locations all over the country. Internally, they have databases about what went wrong/right (maybe more so wrong). If only that information was widely available.
That is true, makes me very interested to know if such data exists currently.
Could easily imagine it's not linked up or siloed best on the policy product offered between companies.
Can imagine getting all that information and being able to create a clearer model of the issues, pitfalls and problems.
There is a start-up company here just bringing things together. And spinning out product ideas from that data. Or possibly open sourcing it and letting others build on it
I've been in custom residential construction since the early 80's. Homes have become much more energy efficient, structurally efficient and design conscious. In the old days it was a GC and a 4 page set of plans, with maybe 5 inspections. Today everyone is connected via smartphones, and it's a team of architects, engineers, contractors and inspectors building a new custom. Productivity does not take into account this huge leap in quality - It like the difference in a 6 cylinder F-150 & a Tesla - night & day...but they both count as 1 vehicle. So productivity doesn't pick up quality.
Perhaps qualitative differences could be expressed in the new structures extended lifespan..Perhaps the homes of the 80's were only good for 20 year before they need complete gut & rehab. Where as today's wonders are good for 40 - 50 years. Could this count as twice the house, and therefore twice as productive?
Isn't it true that manufacturing and agriculture also had innovations that were incremental and slow? After all, the processes of agriculture also basically stayed the same for hundreds of years. Before mechanized vehicles came into use. A great book on the topic of innovation itself is the excellent book How Innovation Works, by Matt Ridley. Using dozens of examples, Ridley shows how in most cases innovation is incremental and slow. The major exceptions being the rapid pace of technological evolution in the past 50 years or so. So I think that the construction industry having slow and incremental innovation isn't a bad thing. As long as innovation is there and continues to be there.
I think that it is difficult to get certified and achieve a reputation with other builders. For example, now there are technologies such as drone photogrammetry https://www.propelleraero.com/use-cases/drone-photogrammetry/ but they did not immediately take root, although they showed their effectiveness
Would appreciate if you could look into this startup, they are making some impressive claims and getting lots of funding but do not disclose much in detail about their product: https://www.nexii.com/
Its partially true. Specially the industry has changed gears during pandemic and many clients and contracting companies are now investing in digalitization. Digital Transformation in construction sector is the next big bet. Recent construction tech billion dollar acquisitions is testimony for the same.
Interesting article. My guess is that it reflects modest experience/involvement with actually buildings. Typically software deals with abstractions; physical reality is much more complex. Re: Revit adoption. as an example. Keep in mind that the economics of designers and contractors are totally different. If you want a detailed critique of the article, contact me.
It's interesting to compare this to the software industry, which seems to share a lot of characteristics with construction (high dollar value, each project has unique conditions, many sequential dependencies, mistakes are often difficult to fix though maybe not as difficult as construction). My impression is that people come up with new/better ways of building software much more frequently.
I think this is good evidence for your idea that a less-fragmented construction industry might be more innovative: a lot of new software tools were pioneered inside one of the few largest tech companies.
(A large part of this is also obviously that the construction industry has existed for >100x longer than the software industry, so there's much more low-hanging fruit in software. Another relevant difference might be that it's easier to experiment with new software tools in a low-stakes way via side projects; I'd guess that the minimum practical size for a software project is much smaller than for a construction project?)
"Maybe not as difficult as construction"???? Not close. Enormous life expectancy differences - working in a 50-year old building is not uncommon, working on 50 year old software is not common. I could do a long list of differences. I have experience in both domains.
Building software and buildings are entirely different beasts. In software you are typically delivering something that is intangible, e.g. productivity, user engagement, etc. and can be easily modified, so can rapidly iterate/ modify the processes and products delivered. In addition a product is never really done, whereas a building has to match spec, has a clear spec, is highly regulated and has a clear finish. Unfortunately, companies try applying project-management processes and metrics designed for building physical objects, and as a result fail even more spectacularly.
Great piece.
Robert Gordon finds that the construction industry recorded negative productivity growth from 1995-2005 in both the US and EU-10, owing entirely to declining multi-factor productivity, i.e. the residual that's supposed to capture output from new ideas and innovation.
See: "The Industry Anatomy of the Transatlantic Productivity Growth Slowdown" https://www.nber.org/papers/w25703
Construction productivity is notoriously hard to measure, but stipulate that these estimates are at least directionally correct. What drove the decline in construction MFP? Would you consider doing a post on why and how construction innovation is not just hard, but may even be somehow regressing?
I very much agree with your conclusion:
To achieve significant change (and risk reduction!) you need to vertically integrate and internalize. SpaceX is my favorite example of getting a stagnant industry into innovation mode. I think the problems of the Space Industry before SpaceX are similar to Construction now. And it's also massive, bold and complicated undertaking to establish a successful "BuildX". But the upside should be even more massive than in Space.
How were the problems of the Space Industry similar to the ones in Construction?
You might enjoy https://philpapers.org/rec/SCHGEA, which looks at a very similar set of constraints and processes in evolutionary biology. Under that model, I would anticipate the most innovative areas of the construction industry to either be, as you said, ones that improve something without changing other parts, or to be "finishing touches": how plumbing gets installed should change rarely, but the quality of tile in bathrooms could improve at a reasonable clip, because it is harder for failures at the end of a process to cascade substantially.
I LOVE your newsletter!!! Please keep it coming... It so much helps to think of system-level changes needed to drive this great industry forward to productivity and sustainability <3
Great article. I am always trying to think of non-normal distributions in real life and the nail example is such a good one!
It seems like the nature of cost overruns and their distribution is maybe related to the bidding process as well? Like if you see something that might decrease your costs, you assume everyone else sees it as well, so you lower your price accordingly. On the other hand, if you see something that might increase your costs, you're gonna look real hard at how likely that thing is to be an issue, and you might bid optimistically, hoping it's not a problem (knowing that it's not the worst thing in the world to come in over budget for "unexpected" reasons).
And I don't know exactly how all this works (like how they calculate cost exactly), but if a project ends up being cheaper than expected, the contract price doesn't usually get lowered, does it? So we'd expect the distribution to be skewed to the right no matter what, just because construction companies generally don't give price breaks if it turns out, after the fact, that their costs were low.
Very important detail! missed that. And also the final client has no interest in publicizing how the project was completed for much lower than they contracted for. Possibly there actually are projects that make double digit margins - but few outside the construction company would know.
I work for a large General Contractor and YES, much of this data is likely linked to the bidding process. It seems like this data is mostly being pulled from public projects. In the US this usually means the lowest qualified bidder gets the job and they only bid what is shown on the drawings or in the other provided documents. Also, these are usually lump sum, meaning if there are savings, the GC gets to keep them. It also means that if they missed something, they pay for it. Anything not shown or unknown at bid time would be an added cost to the client.
I think that is good insight, again it comes back to risk internalization, forecasting and being stuck in a sort of bad equilibrium.
There is also the possibility too that either culturally, cost benefits wise or myriad of other reasons, that people involved in the process prefer it that way. For unspecified reasons, they might not even understand.
On the cost side of things, there is something similar that happens with low-cost air travel. Whereby you have many stated preferences for more comfort, ease of access, better flying experience etc. But in actual actions, people will majority tend to choose the default low-cost option.
Now I can't necessarily say people are wrong to have this default in all situations, honestly, most people don't fly often enough to really evaluate the claims for the benefits of a higher price plane ticket. And it does seem the default low-cost option, in a situation where you suffer from a lack of information, either on the outcome, the claims of an airline or your own preferences might make sense. It's the most salient and clear piece of information.
I tend to travel relatively often, I through experience I found that choosing the more expensive flight that either leaves not too early or arrives in a destination not late, in order to either navigate to a hotel or get setup is worth paying extra for.
But honestly, old me didn't appreciate that benefit, only after much pain of paying 40%-30% less for a flight that leaves at 5 am did I realise for me the extra sleepiness and loss of productivity that subsequent day was just not worth it. And I didn't mind either the potentially extra wait or cost in the airport if I could leave at say 9.00 am instead. So from my own personal experience, my own cost-benefit calculations were essentially unknown to me, until I basically ran a bunch of experiments on myself.
The airline anecdote makes me wonder if something similar is happening too within the construction industry. Basically lots of plausible claims of benefits or possible alternatives, but nobody really knows their own needs, benefits or costs & there is a lack of real-life repeated experiments to see the differences.
Basically, we would need to run a myriad of experimental trial runs of builds and rebuilds under multiple scenarios for say residential housing in multiple areas to figure out the varied or common issues. Because my sense is many of the claims are all kind of plausible-sounding, but we just don't have enough purposefully done trials to know. Then repeat that for every area. Realise doing multiple builds and rebuilds of bridges might be absurd, but it does seem the only way to capture plausibly the potential fixes that would work.
Once you have all this practical information, you can use it to start a business, advocate for policy changes, develop a culture of prestige and status around your methods etc.
What we need is Construction Physics Build Laboratory :-)
To some extent, I believe insurance companies that issues builder's risk policies would fill some of this role. They receive data from general contractors of things that go wrong, new means/methods beings used, etc. for each project they insure. In many cases there will be projects similar in nature in locations all over the country. Internally, they have databases about what went wrong/right (maybe more so wrong). If only that information was widely available.
That is true, makes me very interested to know if such data exists currently.
Could easily imagine it's not linked up or siloed best on the policy product offered between companies.
Can imagine getting all that information and being able to create a clearer model of the issues, pitfalls and problems.
There is a start-up company here just bringing things together. And spinning out product ideas from that data. Or possibly open sourcing it and letting others build on it
I've been in custom residential construction since the early 80's. Homes have become much more energy efficient, structurally efficient and design conscious. In the old days it was a GC and a 4 page set of plans, with maybe 5 inspections. Today everyone is connected via smartphones, and it's a team of architects, engineers, contractors and inspectors building a new custom. Productivity does not take into account this huge leap in quality - It like the difference in a 6 cylinder F-150 & a Tesla - night & day...but they both count as 1 vehicle. So productivity doesn't pick up quality.
Perhaps qualitative differences could be expressed in the new structures extended lifespan..Perhaps the homes of the 80's were only good for 20 year before they need complete gut & rehab. Where as today's wonders are good for 40 - 50 years. Could this count as twice the house, and therefore twice as productive?
Isn't it true that manufacturing and agriculture also had innovations that were incremental and slow? After all, the processes of agriculture also basically stayed the same for hundreds of years. Before mechanized vehicles came into use. A great book on the topic of innovation itself is the excellent book How Innovation Works, by Matt Ridley. Using dozens of examples, Ridley shows how in most cases innovation is incremental and slow. The major exceptions being the rapid pace of technological evolution in the past 50 years or so. So I think that the construction industry having slow and incremental innovation isn't a bad thing. As long as innovation is there and continues to be there.
I think that it is difficult to get certified and achieve a reputation with other builders. For example, now there are technologies such as drone photogrammetry https://www.propelleraero.com/use-cases/drone-photogrammetry/ but they did not immediately take root, although they showed their effectiveness
Would appreciate if you could look into this startup, they are making some impressive claims and getting lots of funding but do not disclose much in detail about their product: https://www.nexii.com/
having built the first BIM systems, I can only agree
Its partially true. Specially the industry has changed gears during pandemic and many clients and contracting companies are now investing in digalitization. Digital Transformation in construction sector is the next big bet. Recent construction tech billion dollar acquisitions is testimony for the same.
I couldn't help but think about this in the context of Katerra
Interesting article. My guess is that it reflects modest experience/involvement with actually buildings. Typically software deals with abstractions; physical reality is much more complex. Re: Revit adoption. as an example. Keep in mind that the economics of designers and contractors are totally different. If you want a detailed critique of the article, contact me.
Nice piece!