What’s the Deal with Construction Productivity?

The data summarized here indicate that productivity in the US construction industry–both labor productivity and total factor productivity–has been much weaker than productivity for the US economy in general.  Indeed, productivity seems to have gotten worse:

To be clear, the raw BEA data suggest that the sector has become less productive over time. A lot less productive: value added per worker in the sector was about 40 percent lower in 2020 than it was in 1970.

Full paper here.

Your thoughts…

Do these findings seem correct?

If so, what are the causes of the poor and declining productivity?

What are the paths to improvement?

20 thoughts on “What’s the Deal with Construction Productivity?”

  1. I saw that article at Legal Insurrection today. California is too far gone by now. They are at the state of gas stove bans. It did not occur to any of CA politicians to build reservoirs during the drought to capture the heavy rain that has always followed drought in that state. I lived there 60 years and went through several such cycles. In 1961 my in-laws house burned down in the BelAir fire. In 1969 their newly rebuilt house almost slid off the mountain in the heavy rains that destroyed many of their neighbors’ whose houses had survived the fire. The cycles have continued but there are too many for whom history began last week. I finally gave up and moved to Arizona 6 years ago. Enough Californians have moved here to elect Democrats to begin the cycle here.

  2. Create millions of jobs for non-college-educated males? That is misogynist patriarchal doubleplusbad wrongspeak.

    Instead, we should be creating construction jobs for female college graduates — they have student loans to pay off, little education, and few marketable skills.

  3. LOL. I have worked in construction and there is a skill involved. I think that’s what the various pundits are missing. As well strength is involved in a lot of it. Creating construction jobs for female college grads, is I hope a joke.

    If the people who know what they are doing have retired, then its not at all hard to see why productivity has fallen off a cliff.

    A strong framing crew can put up a house in very little time, if they are used to working with each other and have the skills. A rag tag band of half trained people, using all power tools, will take a lot longer to throw up the frame.

  4. What drives increase in productivity? Answer that question and one will unravel the puzzle about it not increasing or even decreasing..

    Folks who have no or little connection with actual business decisions not only will not but, in my experience, cannot answer that first question. Productivity increase happens when someone chooses to figure out what might bring about that increase, then accomplishes the steps it takes to make that happen. These steps may involve making better choices such as who does what, how many are allocated to do what, removal of roadblocks by better arrangement of materials and task location, figuring out causes of delay in obtaining materials, scheduling tasks to avoid times of no productivity in absence of supplies, and a host of parallel accomplishments. But one or more of the steps will sooner or later include a need for investment in tools, even creation of new tools (think hammer vs nail gun, manually carrying shingles to a roof vs forklift, shovel vs trencher). All of these steps tend to result in an increase in efficiency of labor, in individuals being able to produce more during the same amount of time.

    But what if labor is cheap, if there are many willing to sell their time for less than it would cost to invest in tools? even less than it would cost to accomplish many/most of the other means of increasing efficiency? Then production increases happen when one adds more workers. But productivity, production per worker, does not. It will probably even decline.

    You think the open southern U.S. border makes no difference? Go to a construction site and listen. Ask, as I have in Spanish, if one needs to speak Spanish in order get a job as part of a work crew. Or, as I have while doing field service tech/engineering, go into a plant just S of Gary, Indiana, almost as far N as one can go without getting into a lake and leaving the U.S., and learn that one’s Spanish ability is needed to communicate with most of the plant’s workers.

    None of these observations lessen my admiration for the Hispanics I’ve met at work. They impress me with their willingness to work. They impress me with their family culture. I’ve watched lunch time arrive. While some of the Anglos pack their lunch, most go out for fast food. Meanwhile, Hispanic women (plural) arrive in a van, unpack it to set up tables and a prepared meal, their fathers, husbands, sons join them, and together they eat as an extended family…and their meal is less expensive than fast food.

  5. Since I am in a bar with time to kill while the missus is antiquing down the street, I’ll take up David’s offer and relive my social science academic days and offer my thoughts.

    Short story, I smell a rat both in the premise and in the article. A 40% drop in productivity over 50 years? In fact if you dig through some of the literature they cite, there is a McKinsey report that shows productivity at 1947 levels. In what way is this productivity acceptable acceptable? Yet it is otherwise we should be able to see industry rationalize and grind out the inefficiencies over the 50 years. If not industry rationalizing than certain large customers such as the U.S. Federal government.

    Second, if you review the article the methodology they claim to use that deal with potential measurement errors is in fact riven with measurement errors. I their sector focus on the residential housing industry they use as a metric sq. foot per worker as a measure of productivity without considering given the changes in housing and living standards whether that is a valid baseline statistic for a 50-year period. In my perusal of the article it doesn’t seem to be an apples-to-apples comparison.

    Instead I’m going to go along the lines of what Roy and to an extent PenGun said and that is you need to understand the industry in order to define the analytical approach, let alone come up with solutions. Goolsbee and Syverson are what we used to call regression weenies, perhaps given they are from NBER, and while that approach has its good points, it also has its bad. In its quest to break down an object into pieces that can be analyzed, science runs the risk of not being to make a proper understanding of the item analyzed. Think of a methodological hammer searching for a social science nail; they use sq footage to measure value of output because it’s the stat to could come up with, not necessarily because it’s either a good or sufficient measure.

    I understand this is only a working paper and that Goolsbee and Syverson were probably just trying to push the debate along just a bit further down the road, but I think a better methodology would be a form of qualitative survey across various sectors of the industry to see why businesses there either are not pursuing increases in productivity or barriers to the productive increases they desire. That approach would not only provide some industr0relevant data that could be used as a jumping-off point for further analysis, but also starts with the satisfying assumption that people in an industry have some idea of how their business works, especially for 50+ years.

    I will note that the McKinsey report mentioned in the Economist article cited contradicts Goolsbee in that latter believes that capital investment, while lagging other sectors, is still high while the former states that their research shows the construction industry is reluctant to perform capital investment because of the cyclical nature of their business.

    We always think of business investing in capital in order to substitute for labor but there is no reason why it cannot work the other way if the numbers can be made to work. Over time, especially 50 years, the numbers will (barring government protection) like I have been dealing with in a K-12 start-up) grind out inefficiency so that what seem to be irrational actions of a business, is usually a cognitive problem on the part of the observer. Productivity has dropped in construction because there has been a disruptive influence and I believe that is cheap labor.

    Another issue is the amount of regulation involved in the actual construction (as opposed to planning) of various projects. We are all aware of the costs of regulation and those costs are simply added to what it takes to produce the same amount of output, thus lowering productivity.

    Note low productivity does not mean cost overruns. Major overruns for public and private sector construction projects a different subject.

    What can be done? Well, we can cut regulation through rationalization. We were joking the other month that process is going to happen anyway with the large rollout of all these windmills and solar renewable projects which have a very direct and detrimental impact on the environment. A second way is the improvement of construction techniques involving modular and 3-D printing. However those techniques have proven problematic and we’re probably a good 5 years away from getting even a good demonstration project built. A third, if productivity and not profits is a your goal, is to to make labor more costly and thus incentive capital investment. How? Shut down the border and enforce immigration laws in the national interior.

  6. Note that while higher labor costs incentivize capital investment, higher interest rates *de-incentivize* capital investment, and hence favor replacing capital…at least potential capital…with labor.

  7. First: What they’re talking about isn’t the construction sector, it’s the single family and small scale multifamily home building sector. The part that is building the huge warehouses for Amazon etc. are progressing nicely.

    Second: The answer’s easy. Houses are being built the same way they have been built here since Europeans landed. One hand built piece at a time. In the centuries since, the saw pit has given way to the saw mill in the 19th century. The hand saw has been replaced by the power saw in the 1920’s. The pneumatic nailer has taken over from the framing hammer. But you still start a house by unloading a big pile of lumber. Imagine building a car by handing out a set of drawings and ordering a load of steel. Productivity then comes down to the skill of the carpenters, as does quality.

    Home builders are purely financial entities. They certainly don’t have the resources to pursue technological innovation, or probably, any real incentive. They make their margin on the gross and the customer’s tend to pay just as much as the lenders will let them.

    What the study also fails to account for is increase in quality. Anybody that says they don’t make them like they used to is right. Anybody that’s had to deal with an old house will have no doubt that nearly all of the change has been for the better.

  8. MCS:

    I will beg to differ with your statement that, “Anybody that’s had to deal with an old house will have no doubt that nearly all of the change has been for the better.”

    My first house was built in 1901 as a farmhouse on a 160-acre site in what would be a major city. The lumber was entirely rough-cut from the trees on the land, and the dimensions were genuine…in other words, the 2×4’s were 2″ thick by 4″ wide, not the 1-1/2″ x 3-1/2″ size that’s now standard. That reduction to 75% can have a huge effect on the strength of the material, not to mention the fact that more than 15% of all dimensional lumber intended for use in construction today is rejected on-site because it’s so poorly cut from tiny trees.

    The roof was built with genuine size lumber, and had rough-cut boards laid across the rafters with huge spaces between them to provide “breathing” space for the cedar shingles. Those 2″ thick boards varied in width from a few inches to over 12″ wide, depending on what part of the tree from which they were cut, and were stronger after a 100 years than when they were put on. When I replaced the roof we had to put the plywood right over the top of those boards (removing the other two layers of roofing before hitting the cedar) and nailed into place so that the “modern” roofing materials would work. If a tornado hits that house the roof may come off, but it will do so in one piece and land intact somewhere.

    Those are just a few isolated samples of the quality of materials (e.g. plaster on lath) that were used in the construction of that house.

    I watched a tornado go through a brand-new neighborhood from the roof of building a mile away, and watched all of the ticky-tacky “insulboard”, insulation, siding, and “OSB” go sailing through the air, even though the tornado didn’t touch down in the development. Afterwards we drove through the area to see if anyone had been injured. On a slight rise was the old four-square house that had been the original house in the farm fields; it had no visible damage. In every direction around, 360°, all of the McMansions (built to current code) had been destroyed. This is an perfect example of how poorly houses are built today as compared to former times.

    Yes, the plumbing and electrical were wonky in the old house. As part of our maintenance and improvement, we eventually ripped out and/or disconnected all of the old “knob and tube” electrical wires, and replaced all of the old steel pipe plumbing with copper. But that’s normal upkeep on an old house.

    Not only has there ben no increase in quality, the quality of currently-built homes is TERRIBLE.

  9. Comparing all of today’s homes to the best of a century ago is an apples to oranges comparison. The crappiest homes built back then are no longer around.

  10. Of the many many horrors the Nazis perpetrated on the world, I would here spotlight the use, abuse, and ruination of the term “labor camp”.

    FDR’s CCC and other job programs were not the best of all possible programs, but perhaps they were the most politically palatable at the time. Taking young people away from depressed areas, teaching them to work together, teaching them basic skills in measuring, leveling, moving heavy things and stacking solid individual things up so that the resulting column or wall or steps will never ever fall over… Anyhow, a lot of the veterans of those camps went to war and survived, and others went to industry and survived, and eventually when the nation restructured the economy those skills were — surprise! — transferable.

    That’s a labor camp.

    A place where the frail and disfavored are set to building bombs in a sweat shop before being marched into a death chamber under the arch reading “Work Makes You Free” — that is NOT a labor camp.

    But now I fear we can’t talk about programs to bring together diverse groups in a common cause for permanent infrastructure developments in remote areas — without the term labor camp being introduced and mis-applied.

  11. The crappiest homes built back then are no longer around.

    True statement. But I have to wonder just how many of the houses built today will exist in a hundred years. My guess is not that many.

    A few years ago I toured the model of a new development of fancy houses. I was not impressed. The basement cement already had multiple cracks- I quit counting at a dozen- and the floor joists were made of this glued together kapok looking substance that reminded of what filled old life jackets.

    How many cracks will that concrete develop in the next fifty years? How about that glued together flooring? And what else could go wrong that I failed to even notice? And this was the model that the builder was showing to potential clients. The model. I rather doubt the rest of the development was built any better.

    The house I used to own was built in the mid-sixties. The floor was real wood and it had a genuine steel beam to support that floor. It had its issues- the insulation was nearly non-existent when I bought it- but the build quality was vastly better than what I saw at that development. It’s already made it to half a century and I bet it could make it all the way.

    New houses?

    Well, I just have my doubts about the longevity of the glue…

  12. A comment at Twitter:

    “Started reading the Trader Joes book and an interesting point he makes is that a big part of why America worked well in the 50s is that most institutions were staffed by overqualified men with a ton of life experience from the great depression + war

    Maybe a factor?

  13. It is a fact that almost everything done wrong, in any simple thing like most construction, is no easier than doing it right.

    If there is no one who knows how to hang a window, and can explain how to do that properly, then many windows will never be right. It is easy to screw up concrete if you are DIY when you should just get the air entrained goodness from whoever delivers concrete in your area.

    I could go on, but I learned from people who knew what they were doing and were willing to take the time to not only teach me, but yell at me when I screwed up. ;)

  14. Comments about houses depressing – we are (theoretically if they ever do it) buying a new house so our later years will be doable alone. We’ll see. I am sure it will be much better for someone in their eighties than most of the houses pre-1950, though people did seem to get along better than by themselves as they aged.

    We wanted a modular house – I’m pretty sold that in a lot of ways that can be a more productive use of construction labor than on-site, but I may well be wrong. If workers are employed in continuous tasks, protected from the elements, deadlines would (it would seem) be met more often. Given some earlier comments on my thinking by people who seemed to know more (not that hard given my knowledge level)I’m missing some points. Certainly some don’t seem aimed at a century (or two) of life, the kind Mark describes above.

    On the other hand, at one time, new cars seemed more “sheerable” by an old metal one; however, that 251 foot drop of the Tesla that left the dysfunctional (well, at least the father) unharmed is almost unbelievably better than my great grandfather’s experience, when his car was hit by a sudden burst of prairie wind and turned over, killing him in 1921.

    Anyway, all this aside, consider the CCC & all those dams & public buildings in the 30’s (and later interstates) next to the apparent 40 years without a dam built in California. California’s excess water following (and probably preceding) extraordinary drought, indicates some states don’t consider productivity – short run, long run, projects, consequences, etc. This flippancy toward reality (and disrespect of what man/engineers/laborers can do to make life longer, healthier, more productive) borders on evill: what other description for these choices when we consider the number of people employed and potentially fed by California fields?

  15. My new Solterra EV is just wonderful. Took it off road a couple of days ago and it was very concerned that I might hit something. It has Radar on the front and back corners and a pile of LIDAR cameras all over the car. It can sense anything close and will flash yellow on the laptop sized screen where that is, and brake to not hit it. I have to get into settings and turn that down a bit. ;)

    The Safety Videos are entertaining. It will not hit anything that jumps out in front of you, slams on the brakes. It has received the highest ratings available in both Europe and Asia for the crash tests.

    But really its so much fun. I don’t have to worry about short tripping the car, as I did with my Subaru BRZ. Its a good idea with a regular engine to get to it temperature and hold it there for a while, to get the best results. The EV is just a set of electric motors and that matters not at all. Its real strong, and instant acceleration is a bit different. A regular car needs a split second or so to get it together, and the EV is instant. No waiting at all and you can hit little holes in the traffic flow effortlessly. The regeneration available means the brakes don’t work that hard, and should last a very long time.

    Just a better car, and I am very pleased. I’m paying about 3 cents a kilometre out of the 120v socket on my fifth wheel. ;)

  16. Blackwing,
    I won’t argue the comparative quality lumber between 1901 and now. If they used site harvested trees, don’t forget to add about a year between felling and milling and about the same to let the milled lumber start to dry to your timeline. Then you throw out all the corkscrews and hope too much doesn’t pull apart or warp too far out of line as the green lumber finishes drying over the next few years. The quality would depend a lot on where you were and what sort of trees you started with. The biggest difference between how the 1901 model was built and now was that every individual board was cut to length by hand rather than with a power saw.

    There’s no question that the 1901 house was smaller, less insulated, if it was insulated at all, may have had one bathroom, maybe, may have had an electric light socket in every room, if it had any wiring (1901 is before power outlets were standardized), etc. The difference between a house built in 1950 and now would be only slightly less stark.

  17. MCS:

    My comment was primarily to illustrate the simple STURDINESS of older construction. My house was the first one in that full section, but it was quickly subsumed by the MegaCity which annexed and then platted the land. It was quickly built out since the streetcar line was then run to a point 4 blocks farther and stopped. Almost every other house in that area was then built between 1918 (the end of the war) and 1930 (when the Depression kicked in). They’re all still standing, some have been nicely re-habbed, all are in excellent shape.

    It was originally gas-lit, and the abandoned gas pipes were left in place when it was electrified. It had exactly zero insulation when built; whoever owned it back in the 1970’s had blown-in fire-proofed cellulose insulation put in the walls and ceiling; we added fiberglass insulation to the attic and did our best to mitigate the air leakage through the ceiling. No water, there was a well with a hand pump in what became the neighbor’s yard; the bathroom was added to the house at some later date, probably when they lifted the house to put in the full foundation.

    I contrast that inner-city neighborhood with a development that went in across the highway from my folk’s (what used to be) country place. It was a cornfield with a few pine trees, so they gave it a Swiss-sounding name, bulldozed it as flat as they could, and stuck ginormous McMansions on itty-bitty little lots. The houses were jammed so close together you could reach out your window and shake hands with your neighbor (only a slight exaggeration). They were so poorly built (but, by gosh up to code!) that the concrete floors were cracking and heaving, the foundation blocks being pushed out by the weight of the backfill, roofs were leaking, ice dams forming in badly built joints of wall/roof, and massive mold problems. People who had bought into them got the heck out as fast as they could, leaving the next bunch of suckers to deal with the problems.

    And you cannot tell me that this was an isolated problem with that one developer…I’ve seen this over and over again with current (but built to C*O*D*E!) housing. Gee whiz, fully inspected by city in$pectors, too.

    A major part of the problem is with the materials being used today. Almost all of the floor joists are now “composite” materials (i.e., glue and junk wood), almost all of the flooring, roofing, and sheathing is OSB (glue and junk wood), and all of them are horribly susceptible to moisture damage, both during construction and after. I strongly suggest that if you have an interest in this type of information that you take a look at some of the earlier ASHRAE articles by Dr. Joseph Lstiburek:
    P.S.: I am a retired BSME design engineer with an emphasis in heat transfer and thermogoddamics, so his subjects are close to my heart.

Comments are closed.