We’ve spent a lot of time examining the problem of construction productivity in the US — the fact that, across a variety of different metrics, construction never seems to get any more efficient (in terms of how much output you get for a given amount of input), or any cheaper.
We have found the industry extremely fragmented and competitive. We teach project teams how to apply Lean Construction approaches with shared risk & reward and how to work together rather than in a linear process. (Contractors become part of the design process - See Lean Construction Institute). Improvements vary but are usually around the 10 - 15% time & cost improvement. It should also be noted that construction processes based on various materials and tools has not really advanced technically. Modular and Industrial methods & materials have been explored for decades but still haven't crossed the chasm.
Is it possible that there is a global ceiling on productivity within the construction industry, and the US was just the first one to hit it?
The other nations with lower levels of productivity keep increasing productivity by copying technologies, skills, and techniques from other nations until they eventually bump up against the same ceiling.
By the way, I am not implying that new technologies or processes cannot break through that ceiling in the future. Just that it has not happened so far.
Possible, yes. Looking at the McKinsey data, I would expect that to put the US higher on the Y-axis, but ... We are clearly within measurement error of that being a reasonable explanation.
Belgian person working in regional planning here. You realise belgium had incredibly lax enforcement of zoning laws etc for decades, well into this millenium? People do nothing but complain that the netherlands are so much more well organised and much more “beautiful” as opposed to weird rurban flemish sprawl. But the upside is that house-prices are much cheaper than other western european regions with comparable population density. We Belgians don’t fully realise this yet. Also building has only recently become somewhat top down, the model is and especially was diy private home owners who could build wherever they could buy a plot of land (yes that includes protected natural parks and natura 2000 habitats) Very very different from the planned neighborhoods of the netherlands. You could read a book like ‘met voorbedachte rade, sluipmoord op de open ruimte’ - premeditated, assassination of open space (it is meant to be a very bad thing) in reverse to see why building anarchically works in Belgium. If this interests you I can email you a whole bunch of stuff on Belgian planning or rather, lack thereof
I've spent a lot of time in Peru and have 3 real estate developers in my immediate US-based family, so I'm familiar with construction quality and speed. You can really tell that Peru is ratcheting up in quality and speed from the ground, the build quality from past vintages of buildings is nowhere near the present build quality, and the speed is impressive compared to a US background (although it lags China's unmatchable construction speeds).
An evermore cumbersome regulatory climate is a huge factor driving the decline in construction labor productivity, but an underappreciated-and-maybe-bigger factor is Talent Pool Competence.
If you mount a hundred year long, society-wide campaign to discourage every ambitious, motivated, and competent young man from making a living by working with his hands, you're going to end up with industries that depend on manual labor becoming filled with the sorts of people who can't find work elsewhere. They will become filled with illegal immigrants, people who don't speak English, people with criminal records, people with drug problems, etc. These industries will become our employers of last resort.
Construction requires humans to do complicated things with their hands while collaborating with others IRL. In other words, it's extremely resistant to automation—much more so than manufacturing and agriculture—and so a less competent workforce will lead to plummeting productivity that can't be reversed by innovation.
Had we not spent the last century discouraging young men from entering the trades, I don't know that worker ingenuity necessarily would be able to offset our sclerotic regulatory burdens, but the situation would certainly not be as dire as it is.
Add up the fully-loaded cost of the productivity decline and you have a HUGE part of the explanation for why housing costs are so high. Yes, regulatory climate is a big part of it as well, but it's difficult to overstate the cost of our culture-wide, century-long war against the skilled trades.
Geoff, just want to give you a hat tip for your contribution here. Given the triggers in your post, I expected the conversation to be a mess, but your responses were well-articulated, empirically defended, and on point. Well done, sir.
My observations are informed by lots of time spent on construction sites over the previous 3.5 decades. While we have seen enormous innovations in tools and materials, we've also seen a precipitous decline in the quality of the workforce.
The pool from which the industry is able to hire has been shrinking and those who are in it are becoming more and more at-risk.
This is inescapably obvious to anyone who has worked in construction for a long time, but difficult to see for those who spend their workday in a climate controlled environment.
From a 2017 study: "The vast majority of young adults between the ages of 18 and 25 do not see themselves working in the construction trades... Most young adults who have yet to make up their minds on a career see very little chance they would join the trades even if the pay was high." It took about a hundred years of anti-marketing to convince young people that they should avoid construction *no matter the pay*. That's gonna have a big impact on the quality of the workforce.
What are your thoughts on "Lean Construction" as referenced by the commenter, Kathleen, above? First I've heard of it, but her note about contractors and builders partnering together to align and share incentives got my attention. I feel like there's a bifurcation between rough-and-cheap labor pool and the polished-and-expensive, "yellow pages" contractors. What builders need are win-win relationships with stable contractors who provide volume pricing, but homebuilding is so inconsistent that's it's hard for folks like that to survive the ebbs and flows of labor demand. The good ones at least supplement their work by getting retail clients, or they get out of the business. I'm rambling a bit here as I think out loud. I just know that those few contractors in my rolodex who are both reliable and priced well are the best part of my job as a builder. And getting stuff done where I don't have those guys is probably the hardest.
I have more residential real estate developer experience and in-the-field-hammer-swinging experience than I have general contractor experience. I do still build my own buildings occasionally (and am doing so now), but others are way more qualified than I to answer that.
My intuition is that the improvements she describes are realistic. With regard to the materials and labor side of lean, I think she probably battles against the realities of each trade having to hand the building over to the subsequent trade, so there is a bit of over-engineering that is happening because of so much blurring of responsibility. eg A structural engineer over-engineers the building he will never see (much less pay for), but his design for the structure he is designing and building for himself is probably more efficient.
Our trades have become extremely specialized in the last several decades (My half-round gutter contractor is different than my metal roof contractor! Why?!), and we've passed the point where specialization is making us more efficient. Ivan Illich described this as "counter-productivity".
My gut tells me a single crew of six capable generalists plus some occasional specialists would build a bespoke home faster, of high quality, in fewer man hours, and with less waste than would different crews for excavation, foundation, framing, roofing, etc, etc. I think the permitting process pushes this to a degree. Also, because the overwhelming majority of homes are now built by extremely large builders, there is a lot of pull among smaller builders to follow the standard practices that work really well for the biggest builders who repeat the same few plans over and over.
So I guess was just imagining those twice daily underground lotteries for a gram of coke and night with a hooker up from Florida on that nuclear powerhouse job in 1981. Or the traditional 6 pack of beer at lunch the guys who worked the New York hi-rises in the 1970's used to talk about.
I've never known a construction worker that wouldn't trade for a job in a "climate controlled environment". You've obviously never worked a summer outdoors in Georgia.
Construction work tends to be a hot/cold, dirty, nasty, dangerous job. The men and women that do it, take pride in being there and then try everything they can to get out of it and find an air conditioned place to make a living in. Its the nature of the work.
I am from Georgia and that’s where I’ve mostly worked in the field, also in the SC Lowcountry. Much hotter summers than Georgia’s. Late 80s to mid 90s. Regardless, you are totally missing my point.
Women can do some of the work some of the time, but nearly every day's labor requires lifting, moving, and awkwardly holding very heavy objects. The physical demand of the work is also why it's rare to find a man over 60 on a crew.
The construction trade that's most accessible to women is interior painting; it demands a degree of fitness and endurance, but not great strength. There are many women on interior painting crews, but far fewer on exterior crews. Short ladders are light, but tall ladders are very heavy and most women can't unload one from the top of a van or constantly move one around a building.
Lack of employer willingness to pay high wages for competent and reliable workers is not the problem. All things being equal, the young man who enters the trades (instead of college) will earn considerably more—both in his 20s and over his lifetime—than the same young man who pursues an architecture degree. But the parents of those young men, their teachers, the messaging from society, government-backed low-interest tuition financing, etc, are all pushing that young man to pursue an air-conditioned job.
The median annual wage for architects in the U.S. is approximately $96,690, while for carpenters, it is about $59,310. This shows a significant difference in earnings between the two professions.
It shows a significant difference in earnings between the two profiles of people who pursue work in those fields—which in turn informs the average earnings of the professions.
Imagine the sort of young man who, at 17, has the background and attributes that secure him admission to architecture school, and—despite him hearing his entire lifetime from every corner of society that he should pursue a job in a climate controlled office—he begins working in construction. He will advance very quickly and, over his lifetime, probably outearn the version of himself that would have pursued an architecture degree.
Conversely, imagine the sort of young man who takes a job in construction today. He is generally not the sort of kid who could have chosen architecture. That path is not open to him for a variety of reasons: he might not have performed well in school, he might not speak english, he might have an arrest record, he might have a drug problem, he might struggle to show up on time or stick to things. That sort of person is over-represented in the talent pool from which construction hires, and that is a big part of the discrepency between average earnings in both professions.
How would that work on a population level? There are only but so many supervisory positions needed, only a limited number of the very highest paying jobs in the trades, and only but so much need for construction work (and so only a limited amount of opportunities to own your own business).
I think that in a healthy market, we'd have more and smaller firms, but I get your point. I think presently and for at least a decade to come, there is more opportunity for an ambitous and competent young man to become a prosperous owner of a small contracting business than for him to become a prosperous architect, but young men are discouraged from pursuing that path and that is harming them. I also think that lots of young men are put on a path that excludes construction, and even if they wouldn't ever own their business, they'd likely become a lead carpenter or project manager; as a result, their lifetime earnings and satisfaction suffer.
Until we change the way houses are built I don't see how productivity can change much. Many try to make the framing faster but that's the easy part generally (God I wish they'd quit hauling around those pre-fabs, they're a menace to other drivers). Foundations, grading, utilities, exterior and interior finishes slow it down. Building codes keep getting stricter, especially in earthquake or flood zones.
Building hasn’t really technically changed in two hundred years — so the lack of technology and labour intensity acts as a drag.
However, buildings don’t get built if there’s nobody to build them or want them. And that’s what explains a lot of the data, in my view. For developed economies, If you split the construction workforce into technical and non technical, it could be the case there’s a shortage of the technical and increased hiring of non technical workers (regulatory, safety, project management). This reduces productivity.
Developing countries that don’t face declines in the technical labour force enjoy growing productivity as demand increases.
I’d like to tag @gerard de valence, who’s a construction academic in Australia and extensively writes on this topic.
Agreed. When I began working on crews in the late 1980s, nail guns were rare and battery powered tools didn't exist. Tools are WAY better and WAY more affordable. We've had so many innovations that ought to have improved the process, and yet, it takes longer and costs more.
I would love to see R&D investment from the government focused on new technologies that make construction work more comfortable and faster. I’m speaking out of school here but it seems far more effort has focused on modular, robotics, and 3D printing than on technology, like the nail gun, that augments human construction workers.
For sure reducing the cost of moving heavy objects short distances over awkward terrain would have a huge impact on jobsite safety, worker longevity, and overall working conditions. As it is, for things that weigh less than maybe a 150 pounds, one or two humans do that much more cost effectively, safely, effectively, etc than machines. On most construction sites, the robot that will be able to help with that will need to have legs rather than wheels, and that makes the technical and cost challenge considerably greater.
We already have jobsite security robots (patrolling "dogs"). I think the next really useful jobsite robots will be cleaning and sorting debris (garbage, nails, screws, sawdust, usable cuts, etc). Clean-up requires a lot of time and, given the meniality and the ROI, most usable debris goes to the landfill. A small robot could do that after hours without disrupting work. Maybe it could also inventory materials.
Yeah that sounds like hard, time consuming, and risky work.
I don’t know if it’s commercially viable yet, but I’ve heard about companies like German Bionic that are creating wearable robotic exoskeletons. They claim it provides 80lbs of weight compensation per lifting movement.
I wonder how much has to do with financing costs and pressures. In a capital market where steady moderate profit rates are not at all appealing to investors, firms are under pressure to replace skilled labor with lower wage unskilled labor and stint on automation.
I wonder if Ireland gives us a clue to what is going on. Around 2009 there is a massive spike in productivity which holds for a number of years. During the exact same period the Irish residential sector crashed while the non-residential (eg high value pharma) and infrastructure only went down slightly. The relative share of house building went down. This is even more interesting if you consider that Belgium is a pharma r&d hub. A quick look at the data there suggests the growth in Belgium productivity is related to this. So how influential is the denominator in all of this and would it be more interesting to disaggregate by sector before we theorize about the different numerators? See page 26 of this Irish study https://www.igbc.ie/wp-content/uploads/2022/10/WLC-UCD-IGBC_30.09.22_V4.0_MidRes.pdf
Construction is a business defined by its cost drivers: land, materials, and labor. As you wrote about in your book, applying technology to the process of construction has proven difficult and it remains a labor-intensive business. Perhaps it’s not reasonable to expect many gains in labor-intensive businesses, including construction, healthcare, and education (which you pointed out).
That said, another way to measure the productivity of construction is the time it takes to build some kind of standard unit, such as a type of building. The time it takes to build the average single family house has fallen a lot in the last 40 years, even though the delay and wait times have grown substantially.
Haven't read Brian's book but that's a great insight. Businesses that are intrinsically labor-intensive are incapable of the efficiency gains that technology-heavy enterprises can yield. Maybe we could take it even a step further back and say that enterprises which do highly customized things... can't be endlessly optimized as measured by marginal production costs. In a space like education, this feels obvious. No one is asking why schools can't function more like factories. In construction, the moral imperative for customization isn't so clear. And yet, while we all wonder why such an important product as housing hasn't succumbed to the cost-compressive powers of capitalism, we simultaneously thumb our noses at cookie-cutter housing. Maybe we just can't have it both ways. We'd like every construction work to be able to afford a house, just like every Ford factory worker could afford a Ford car. But if the price of that is, "You can pick any color as long as it's black," are we willing to accept it? Even assuming building and planning ordinances were made uniform nationwide, are we ok with a hundred million, exact replica, houses?
Here's a random hypothesis just looking at the definition of "productivity". If productivity is defined by the hours of labor worked in the denominator, is it possible that the quality of labor available to construction on average has decreased? This then offsets technological improvements, and on net looks like flattish productivity. Similarly maybe tech improvements in construction disproportionally look like "less skilled labor can perform the same work", which is a real improvement but does not necessarily increase labor productivity.
Why this might be plausible: over the last few decades, there have been higher and higher returns for skilled labor in service sectors and especially in easily-scaled areas (Finance, Tech, Doctor/Lawyer/Dentist). Maybe highly skilled labor is increasingly going to these areas instead of Construction, which manifests as a lower productivity value. This may also have explanatory power for why the US seems to have been uniquely bad for a while, as the US expanded these sectors faster and earlier.
Considerations against: It doesn't match up that cleanly against the data, why would the US eventually plateau and start increasing again? Why is manufacturing so much higher, maybe tech advances just have a different labor profile in manufacturing than construction? How does this explain Belgium ... uh I don't know.
This would be a more positive conclusion if true. Construction has low labor productivity growth because of selection effects and economic tradeoffs, not because of some inherent flaw.
Note: I have not thought about this enough nor know enough about construction to be remotely confident in this.
Coming from a manufacturing angle, I believe the reason we see productivity go way up (in manufacturing) is the proportion of value add to material input cost for a vast majority of products has gone way down—there is less touch time/labor hours (seconds) which are key cost drivers for most manufactured goods.
When I think about this for construction I can only imagine an aging workforce (net negative), increase in use of power tools (slight positive) and that’s about it.
On what tablet of stone is it written that labor must continually generate more output per [insert your favorite unit of measure]?
And at what cost does increased labor "productivity" come? #1: Quality. The one thing that can't be measured, so nobody ever talks about it. We're all just supposed to celebrate an inclined "labor productivity" chart while workmanship, aesthetics, durability and value are drained away from the discussion and the experience in the same way that inflation drains away your savings.
Stagnant or declining productivity means things aren’t getting cheaper and people have to spend the same or more time working for the same or less pay. How is that not obviously worse than the alternative, all else being equal? It’s fair to ask if productivity improvements come with trade offs in some specific cases of course, but trying to argue against the general case is just arguing everyone should have less time and less money
"Stagnant or declining productivity means things aren’t getting cheaper and people have to spend the same or more time working for the same or less pay."
If that's what they taught you, I'd ask for a refund. Read "The Road To Serfdom" in order to understand what you're trying to describe.
I would read the book itself, not the slander about its author. That way you can judge the thesis for yourself, rather than have somebody pre-judge it for you.
In theory, higher quality is worth more. You can argue about how well prices and quality actually correlate, but I don't think the problem is that housing has gotten too cheap. The productivity measure asks "_Given_ what we've decided to make, why can't we do it faster and cheaper?".
With rooftop/small scale solar power, the problem is that very little of the cost is the electronics; I think I saw recently that 90% of the cost is in the custom and high-touch portions of the permitting and installation process that can't be easily automated. You can argue that some of those processes *should* be automated or streamlined or eliminated, but other than site-specific installation requirements, that is beyond the power of any particular builder.
I suspect the same problem (most of the costs being either custom or relatively fixed regardless of project scale) happens with construction, and is even worse than the measurements show because of the difficulty in valuing anything but a new retail or office building, and *maybe* new housing.
I'm intrigued by the attempt to get in to this but feel a bit like it's difficult to compare the construction industry to others addressed by KLEMS. On the one hand, the VA to me is a bit questionable because it's not obvious to me if the value of inputs includes land rather than just materials? If not that would explain the 08 surge because the rising cost of construction land is capture by Value Added. Also, construction has a very strong link between cost of inputs and outputs. Sites are valued at the sale cost, minus of inputs, profit and financing. I would need a better look at the methodology of the KLEMS base data to really understand how accurately it is recording VA per hour worked in construction.
It certainly should, but unfortunately that adds to the measurement problem, since the value of the land depends heavily on what is built on (and near) it, and there isn't enough transparency to back out pre-project land values consistently.
Even when the same person profits by the same amount, there may be tax or other financial reasons to mis-attribute the gains to otherwise unexplained increases in land value rather than to the construction project.
Is there any sort of adjustment made for increasing standards and regulations, i.e. building codes? I think there is an argument to be made that new houses and residential construction in the US is built to a VASTLY higher standard than 50 years ago. Whether this is necessary or marginally helpful is another matter, but building codes have never RELAXED, I work in the trades and code compliance, especially in California, seems subjectively 100% more time and labor than say 30 years ago. Electrical, plumbing, seismic, framing, fire, roofing, cladding, security, ADA, all have much more stringent and expensive minimum standards than when I started.
Yes, fewer units are built with the same inputs, but the cost has doubled because the regulations have tripled.
Construction is an enabler of client productivity. Every year we are expected to include more complex specifications at the same price, and price is the proxy for value in these studies. The construction productivity study boundaries could be redrawn to include the added value of the construction deliverables for the clients. But even then the full value of construction would not be covered. No, the only productivity model which really counts is the whole economy model. And then there are the externalities to be added into any of the models; monetary price is a terrible proxy for value.
I believe it was Einstein who said that a thing should be simplified as far as possible, but no further. We simplified too far a long way back.
We have found the industry extremely fragmented and competitive. We teach project teams how to apply Lean Construction approaches with shared risk & reward and how to work together rather than in a linear process. (Contractors become part of the design process - See Lean Construction Institute). Improvements vary but are usually around the 10 - 15% time & cost improvement. It should also be noted that construction processes based on various materials and tools has not really advanced technically. Modular and Industrial methods & materials have been explored for decades but still haven't crossed the chasm.
I've been interested for a while in learning more about Lean Construction, could you send me an email? briancpotter@gmail.com
Is it possible that there is a global ceiling on productivity within the construction industry, and the US was just the first one to hit it?
The other nations with lower levels of productivity keep increasing productivity by copying technologies, skills, and techniques from other nations until they eventually bump up against the same ceiling.
By the way, I am not implying that new technologies or processes cannot break through that ceiling in the future. Just that it has not happened so far.
Possible, yes. Looking at the McKinsey data, I would expect that to put the US higher on the Y-axis, but ... We are clearly within measurement error of that being a reasonable explanation.
Belgian person working in regional planning here. You realise belgium had incredibly lax enforcement of zoning laws etc for decades, well into this millenium? People do nothing but complain that the netherlands are so much more well organised and much more “beautiful” as opposed to weird rurban flemish sprawl. But the upside is that house-prices are much cheaper than other western european regions with comparable population density. We Belgians don’t fully realise this yet. Also building has only recently become somewhat top down, the model is and especially was diy private home owners who could build wherever they could buy a plot of land (yes that includes protected natural parks and natura 2000 habitats) Very very different from the planned neighborhoods of the netherlands. You could read a book like ‘met voorbedachte rade, sluipmoord op de open ruimte’ - premeditated, assassination of open space (it is meant to be a very bad thing) in reverse to see why building anarchically works in Belgium. If this interests you I can email you a whole bunch of stuff on Belgian planning or rather, lack thereof
I've spent a lot of time in Peru and have 3 real estate developers in my immediate US-based family, so I'm familiar with construction quality and speed. You can really tell that Peru is ratcheting up in quality and speed from the ground, the build quality from past vintages of buildings is nowhere near the present build quality, and the speed is impressive compared to a US background (although it lags China's unmatchable construction speeds).
An evermore cumbersome regulatory climate is a huge factor driving the decline in construction labor productivity, but an underappreciated-and-maybe-bigger factor is Talent Pool Competence.
If you mount a hundred year long, society-wide campaign to discourage every ambitious, motivated, and competent young man from making a living by working with his hands, you're going to end up with industries that depend on manual labor becoming filled with the sorts of people who can't find work elsewhere. They will become filled with illegal immigrants, people who don't speak English, people with criminal records, people with drug problems, etc. These industries will become our employers of last resort.
Construction requires humans to do complicated things with their hands while collaborating with others IRL. In other words, it's extremely resistant to automation—much more so than manufacturing and agriculture—and so a less competent workforce will lead to plummeting productivity that can't be reversed by innovation.
Had we not spent the last century discouraging young men from entering the trades, I don't know that worker ingenuity necessarily would be able to offset our sclerotic regulatory burdens, but the situation would certainly not be as dire as it is.
Add up the fully-loaded cost of the productivity decline and you have a HUGE part of the explanation for why housing costs are so high. Yes, regulatory climate is a big part of it as well, but it's difficult to overstate the cost of our culture-wide, century-long war against the skilled trades.
What a disaster that war has been for America.
Geoff, just want to give you a hat tip for your contribution here. Given the triggers in your post, I expected the conversation to be a mess, but your responses were well-articulated, empirically defended, and on point. Well done, sir.
The data presented here doesn't show "plummeting productivity", it shows a global tendency of flattening of improving productivity.
But, yeah, go ahead and blame the workforce for a problem you have misidentified.
My observations are informed by lots of time spent on construction sites over the previous 3.5 decades. While we have seen enormous innovations in tools and materials, we've also seen a precipitous decline in the quality of the workforce.
The pool from which the industry is able to hire has been shrinking and those who are in it are becoming more and more at-risk.
This is inescapably obvious to anyone who has worked in construction for a long time, but difficult to see for those who spend their workday in a climate controlled environment.
From a 2017 study: "The vast majority of young adults between the ages of 18 and 25 do not see themselves working in the construction trades... Most young adults who have yet to make up their minds on a career see very little chance they would join the trades even if the pay was high." It took about a hundred years of anti-marketing to convince young people that they should avoid construction *no matter the pay*. That's gonna have a big impact on the quality of the workforce.
What are your thoughts on "Lean Construction" as referenced by the commenter, Kathleen, above? First I've heard of it, but her note about contractors and builders partnering together to align and share incentives got my attention. I feel like there's a bifurcation between rough-and-cheap labor pool and the polished-and-expensive, "yellow pages" contractors. What builders need are win-win relationships with stable contractors who provide volume pricing, but homebuilding is so inconsistent that's it's hard for folks like that to survive the ebbs and flows of labor demand. The good ones at least supplement their work by getting retail clients, or they get out of the business. I'm rambling a bit here as I think out loud. I just know that those few contractors in my rolodex who are both reliable and priced well are the best part of my job as a builder. And getting stuff done where I don't have those guys is probably the hardest.
I have more residential real estate developer experience and in-the-field-hammer-swinging experience than I have general contractor experience. I do still build my own buildings occasionally (and am doing so now), but others are way more qualified than I to answer that.
My intuition is that the improvements she describes are realistic. With regard to the materials and labor side of lean, I think she probably battles against the realities of each trade having to hand the building over to the subsequent trade, so there is a bit of over-engineering that is happening because of so much blurring of responsibility. eg A structural engineer over-engineers the building he will never see (much less pay for), but his design for the structure he is designing and building for himself is probably more efficient.
Our trades have become extremely specialized in the last several decades (My half-round gutter contractor is different than my metal roof contractor! Why?!), and we've passed the point where specialization is making us more efficient. Ivan Illich described this as "counter-productivity".
My gut tells me a single crew of six capable generalists plus some occasional specialists would build a bespoke home faster, of high quality, in fewer man hours, and with less waste than would different crews for excavation, foundation, framing, roofing, etc, etc. I think the permitting process pushes this to a degree. Also, because the overwhelming majority of homes are now built by extremely large builders, there is a lot of pull among smaller builders to follow the standard practices that work really well for the biggest builders who repeat the same few plans over and over.
So I guess was just imagining those twice daily underground lotteries for a gram of coke and night with a hooker up from Florida on that nuclear powerhouse job in 1981. Or the traditional 6 pack of beer at lunch the guys who worked the New York hi-rises in the 1970's used to talk about.
I've never known a construction worker that wouldn't trade for a job in a "climate controlled environment". You've obviously never worked a summer outdoors in Georgia.
Construction work tends to be a hot/cold, dirty, nasty, dangerous job. The men and women that do it, take pride in being there and then try everything they can to get out of it and find an air conditioned place to make a living in. Its the nature of the work.
I am from Georgia and that’s where I’ve mostly worked in the field, also in the SC Lowcountry. Much hotter summers than Georgia’s. Late 80s to mid 90s. Regardless, you are totally missing my point.
Can only men do construction work?
I think the problem is reluctance to PAY people for skilled work and strenuous efforts to replace skilled workers with low pay unskilled labor.
Women can do some of the work some of the time, but nearly every day's labor requires lifting, moving, and awkwardly holding very heavy objects. The physical demand of the work is also why it's rare to find a man over 60 on a crew.
The construction trade that's most accessible to women is interior painting; it demands a degree of fitness and endurance, but not great strength. There are many women on interior painting crews, but far fewer on exterior crews. Short ladders are light, but tall ladders are very heavy and most women can't unload one from the top of a van or constantly move one around a building.
Lack of employer willingness to pay high wages for competent and reliable workers is not the problem. All things being equal, the young man who enters the trades (instead of college) will earn considerably more—both in his 20s and over his lifetime—than the same young man who pursues an architecture degree. But the parents of those young men, their teachers, the messaging from society, government-backed low-interest tuition financing, etc, are all pushing that young man to pursue an air-conditioned job.
The median annual wage for architects in the U.S. is approximately $96,690, while for carpenters, it is about $59,310. This shows a significant difference in earnings between the two professions.
It shows a significant difference in earnings between the two profiles of people who pursue work in those fields—which in turn informs the average earnings of the professions.
Imagine the sort of young man who, at 17, has the background and attributes that secure him admission to architecture school, and—despite him hearing his entire lifetime from every corner of society that he should pursue a job in a climate controlled office—he begins working in construction. He will advance very quickly and, over his lifetime, probably outearn the version of himself that would have pursued an architecture degree.
Conversely, imagine the sort of young man who takes a job in construction today. He is generally not the sort of kid who could have chosen architecture. That path is not open to him for a variety of reasons: he might not have performed well in school, he might not speak english, he might have an arrest record, he might have a drug problem, he might struggle to show up on time or stick to things. That sort of person is over-represented in the talent pool from which construction hires, and that is a big part of the discrepency between average earnings in both professions.
How would that work on a population level? There are only but so many supervisory positions needed, only a limited number of the very highest paying jobs in the trades, and only but so much need for construction work (and so only a limited amount of opportunities to own your own business).
I think that in a healthy market, we'd have more and smaller firms, but I get your point. I think presently and for at least a decade to come, there is more opportunity for an ambitous and competent young man to become a prosperous owner of a small contracting business than for him to become a prosperous architect, but young men are discouraged from pursuing that path and that is harming them. I also think that lots of young men are put on a path that excludes construction, and even if they wouldn't ever own their business, they'd likely become a lead carpenter or project manager; as a result, their lifetime earnings and satisfaction suffer.
eye-roll.
Convincing rebuttal, and a valuable contribution to the discussion.
They become so predictable so fast.
Until we change the way houses are built I don't see how productivity can change much. Many try to make the framing faster but that's the easy part generally (God I wish they'd quit hauling around those pre-fabs, they're a menace to other drivers). Foundations, grading, utilities, exterior and interior finishes slow it down. Building codes keep getting stricter, especially in earthquake or flood zones.
Building hasn’t really technically changed in two hundred years — so the lack of technology and labour intensity acts as a drag.
However, buildings don’t get built if there’s nobody to build them or want them. And that’s what explains a lot of the data, in my view. For developed economies, If you split the construction workforce into technical and non technical, it could be the case there’s a shortage of the technical and increased hiring of non technical workers (regulatory, safety, project management). This reduces productivity.
Developing countries that don’t face declines in the technical labour force enjoy growing productivity as demand increases.
I’d like to tag @gerard de valence, who’s a construction academic in Australia and extensively writes on this topic.
Agreed. When I began working on crews in the late 1980s, nail guns were rare and battery powered tools didn't exist. Tools are WAY better and WAY more affordable. We've had so many innovations that ought to have improved the process, and yet, it takes longer and costs more.
I would love to see R&D investment from the government focused on new technologies that make construction work more comfortable and faster. I’m speaking out of school here but it seems far more effort has focused on modular, robotics, and 3D printing than on technology, like the nail gun, that augments human construction workers.
For sure reducing the cost of moving heavy objects short distances over awkward terrain would have a huge impact on jobsite safety, worker longevity, and overall working conditions. As it is, for things that weigh less than maybe a 150 pounds, one or two humans do that much more cost effectively, safely, effectively, etc than machines. On most construction sites, the robot that will be able to help with that will need to have legs rather than wheels, and that makes the technical and cost challenge considerably greater.
We already have jobsite security robots (patrolling "dogs"). I think the next really useful jobsite robots will be cleaning and sorting debris (garbage, nails, screws, sawdust, usable cuts, etc). Clean-up requires a lot of time and, given the meniality and the ROI, most usable debris goes to the landfill. A small robot could do that after hours without disrupting work. Maybe it could also inventory materials.
Awesome.
Yeah that sounds like hard, time consuming, and risky work.
I don’t know if it’s commercially viable yet, but I’ve heard about companies like German Bionic that are creating wearable robotic exoskeletons. They claim it provides 80lbs of weight compensation per lifting movement.
I wonder how much has to do with financing costs and pressures. In a capital market where steady moderate profit rates are not at all appealing to investors, firms are under pressure to replace skilled labor with lower wage unskilled labor and stint on automation.
I wonder if Ireland gives us a clue to what is going on. Around 2009 there is a massive spike in productivity which holds for a number of years. During the exact same period the Irish residential sector crashed while the non-residential (eg high value pharma) and infrastructure only went down slightly. The relative share of house building went down. This is even more interesting if you consider that Belgium is a pharma r&d hub. A quick look at the data there suggests the growth in Belgium productivity is related to this. So how influential is the denominator in all of this and would it be more interesting to disaggregate by sector before we theorize about the different numerators? See page 26 of this Irish study https://www.igbc.ie/wp-content/uploads/2022/10/WLC-UCD-IGBC_30.09.22_V4.0_MidRes.pdf
Construction is a business defined by its cost drivers: land, materials, and labor. As you wrote about in your book, applying technology to the process of construction has proven difficult and it remains a labor-intensive business. Perhaps it’s not reasonable to expect many gains in labor-intensive businesses, including construction, healthcare, and education (which you pointed out).
That said, another way to measure the productivity of construction is the time it takes to build some kind of standard unit, such as a type of building. The time it takes to build the average single family house has fallen a lot in the last 40 years, even though the delay and wait times have grown substantially.
Haven't read Brian's book but that's a great insight. Businesses that are intrinsically labor-intensive are incapable of the efficiency gains that technology-heavy enterprises can yield. Maybe we could take it even a step further back and say that enterprises which do highly customized things... can't be endlessly optimized as measured by marginal production costs. In a space like education, this feels obvious. No one is asking why schools can't function more like factories. In construction, the moral imperative for customization isn't so clear. And yet, while we all wonder why such an important product as housing hasn't succumbed to the cost-compressive powers of capitalism, we simultaneously thumb our noses at cookie-cutter housing. Maybe we just can't have it both ways. We'd like every construction work to be able to afford a house, just like every Ford factory worker could afford a Ford car. But if the price of that is, "You can pick any color as long as it's black," are we willing to accept it? Even assuming building and planning ordinances were made uniform nationwide, are we ok with a hundred million, exact replica, houses?
Good points, all. It is a conundrum. Just like the choice between cheaper housing and preserving property values.
Here's a random hypothesis just looking at the definition of "productivity". If productivity is defined by the hours of labor worked in the denominator, is it possible that the quality of labor available to construction on average has decreased? This then offsets technological improvements, and on net looks like flattish productivity. Similarly maybe tech improvements in construction disproportionally look like "less skilled labor can perform the same work", which is a real improvement but does not necessarily increase labor productivity.
Why this might be plausible: over the last few decades, there have been higher and higher returns for skilled labor in service sectors and especially in easily-scaled areas (Finance, Tech, Doctor/Lawyer/Dentist). Maybe highly skilled labor is increasingly going to these areas instead of Construction, which manifests as a lower productivity value. This may also have explanatory power for why the US seems to have been uniquely bad for a while, as the US expanded these sectors faster and earlier.
Considerations against: It doesn't match up that cleanly against the data, why would the US eventually plateau and start increasing again? Why is manufacturing so much higher, maybe tech advances just have a different labor profile in manufacturing than construction? How does this explain Belgium ... uh I don't know.
This would be a more positive conclusion if true. Construction has low labor productivity growth because of selection effects and economic tradeoffs, not because of some inherent flaw.
Note: I have not thought about this enough nor know enough about construction to be remotely confident in this.
Coming from a manufacturing angle, I believe the reason we see productivity go way up (in manufacturing) is the proportion of value add to material input cost for a vast majority of products has gone way down—there is less touch time/labor hours (seconds) which are key cost drivers for most manufactured goods.
When I think about this for construction I can only imagine an aging workforce (net negative), increase in use of power tools (slight positive) and that’s about it.
On what tablet of stone is it written that labor must continually generate more output per [insert your favorite unit of measure]?
And at what cost does increased labor "productivity" come? #1: Quality. The one thing that can't be measured, so nobody ever talks about it. We're all just supposed to celebrate an inclined "labor productivity" chart while workmanship, aesthetics, durability and value are drained away from the discussion and the experience in the same way that inflation drains away your savings.
Sum ting wong with this picture.
The whole article is about pointing out that construction productivity isn’t like other sectors’ productivity, which tends to increase.
Also you might have noticed that the rent is too damn high.
The implication is that lacking an increase in "labor productivity" is somehow undesirable; and your comment indicates you agree with that idea.
Bonus: the rent is not to damn high because the slaves aren't producing enough bricks.
Stagnant or declining productivity means things aren’t getting cheaper and people have to spend the same or more time working for the same or less pay. How is that not obviously worse than the alternative, all else being equal? It’s fair to ask if productivity improvements come with trade offs in some specific cases of course, but trying to argue against the general case is just arguing everyone should have less time and less money
"Stagnant or declining productivity means things aren’t getting cheaper and people have to spend the same or more time working for the same or less pay."
If that's what they taught you, I'd ask for a refund. Read "The Road To Serfdom" in order to understand what you're trying to describe.
https://krebscycle99.net/2013/08/27/the-road-to-serfdom/
I would read the book itself, not the slander about its author. That way you can judge the thesis for yourself, rather than have somebody pre-judge it for you.
In theory, higher quality is worth more. You can argue about how well prices and quality actually correlate, but I don't think the problem is that housing has gotten too cheap. The productivity measure asks "_Given_ what we've decided to make, why can't we do it faster and cheaper?".
With rooftop/small scale solar power, the problem is that very little of the cost is the electronics; I think I saw recently that 90% of the cost is in the custom and high-touch portions of the permitting and installation process that can't be easily automated. You can argue that some of those processes *should* be automated or streamlined or eliminated, but other than site-specific installation requirements, that is beyond the power of any particular builder.
I suspect the same problem (most of the costs being either custom or relatively fixed regardless of project scale) happens with construction, and is even worse than the measurements show because of the difficulty in valuing anything but a new retail or office building, and *maybe* new housing.
I'm intrigued by the attempt to get in to this but feel a bit like it's difficult to compare the construction industry to others addressed by KLEMS. On the one hand, the VA to me is a bit questionable because it's not obvious to me if the value of inputs includes land rather than just materials? If not that would explain the 08 surge because the rising cost of construction land is capture by Value Added. Also, construction has a very strong link between cost of inputs and outputs. Sites are valued at the sale cost, minus of inputs, profit and financing. I would need a better look at the methodology of the KLEMS base data to really understand how accurately it is recording VA per hour worked in construction.
I believe that gross value add excludes the value of land.
It certainly should, but unfortunately that adds to the measurement problem, since the value of the land depends heavily on what is built on (and near) it, and there isn't enough transparency to back out pre-project land values consistently.
Even when the same person profits by the same amount, there may be tax or other financial reasons to mis-attribute the gains to otherwise unexplained increases in land value rather than to the construction project.
Is there any sort of adjustment made for increasing standards and regulations, i.e. building codes? I think there is an argument to be made that new houses and residential construction in the US is built to a VASTLY higher standard than 50 years ago. Whether this is necessary or marginally helpful is another matter, but building codes have never RELAXED, I work in the trades and code compliance, especially in California, seems subjectively 100% more time and labor than say 30 years ago. Electrical, plumbing, seismic, framing, fire, roofing, cladding, security, ADA, all have much more stringent and expensive minimum standards than when I started.
Yes, fewer units are built with the same inputs, but the cost has doubled because the regulations have tripled.
Construction is an enabler of client productivity. Every year we are expected to include more complex specifications at the same price, and price is the proxy for value in these studies. The construction productivity study boundaries could be redrawn to include the added value of the construction deliverables for the clients. But even then the full value of construction would not be covered. No, the only productivity model which really counts is the whole economy model. And then there are the externalities to be added into any of the models; monetary price is a terrible proxy for value.
I believe it was Einstein who said that a thing should be simplified as far as possible, but no further. We simplified too far a long way back.