The problem with housing is that we haven't found a way to offshore it to China. If we did, housing would probably have gotten cheaper, like everything else we've manufactured there.

There actually are companies that try to do this stuff in China. They will manufacture these modules. But again, to my point about, you know, you incur all these extra cost and all that, you know, they have to you have to break them into modules and put them on these boats and bring them over. So, it's, you know, it's a very sort of niche business.

Hello and welcome to another episode of the Odd Thoughts podcast. I'm Tracy Aloway.

And I'm Joe Weisenthal.

Joe, you're a homeowner now, right?

Yeah, it's really fun.

Uh, tell me tell me how fun it is. I know you've had a few issues that cropped up basically as soon as you bought your new place.

I don't want to talk about Let's go. Let's move on. No, it's

No, this is part of my carefully planned intro.

Yes, I know. Um, yeah, it's annoying and you have to deal with stuff and I may have mentioned on the show one time, actually I've been lucky by and large,

But yes, it was very funny. You know, I think literally the day or two days after we closed on my apartment in the East Village, you know, it's like when there was something wrong before, you just there's just a phone number of someone and you call. Yeah. We still the building still is a super, but you know, the landlord just deals with replacing something and then I think literally the day after we closed there was like a window leak and it's like wait is there no longer a number? Do I have to pay for this myself?

So who did you find to fix the window?

Okay, I don't remember that. I don't remember that probably there is still a super in the building and the super but we had to write a check. Okay, one of the things that has surprised me about being a homeowner is I'm thinking how to frame this, but like how low tech a lot of the housing construction and fixing actually is, right? So, you know, if there's a leak in your ceiling, in your roof or something, it's literally either you or some guy that you hire going up on a ladder and like putting down a few more shingles with a nail gun, right? It all seems very old.

What did you think was going to happen? I don't know. There's like going to be like some like unitry like robots that like crawled up there. Yeah. Drone.

No, seriously. It's the year 2025 and we're still putting houses together with like, you know,

Hammers and someone someone comes with a box and they look through the different screw. They're like, "No, this doesn't fit. This doesn't fit. Oh, I got it." And then they're like, "One second, I don't have the right screw it." And they go around to a, you know, in the city, you know, they go around to a construction store. All they'll be right back like, "Oh, I got this. I got Yeah. Yeah. It's still pretty probably the way it looked when we were kids."

Okay. Well, anyway, I've been thinking about it because a couple months ago, the Richmond Fed put out this paper and it was all about productivity in housing construction and the fact that construction labor productivity in housing has fallen by more than 30% from 1970 to 2020. So, you think about the arc of economic development. You know, normally there are productivity improvements as we invent new efficiencies and as we invent new technology and ways of making things. That hasn't happened in housing. It's still the guy, you know, showing up with a toolbox.

This strikes me as a really important element of the housing affordability conversation that I don't think gets enough attention, which is it's easy to talk about zoning and I'm sure there are issues there that need reform. It's easy, you know, permits and all that stuff. It's easy to talk about financing, etc. However, housing is a product and things get built at a cost and then sold at some profit margin above that. And if that underlying cost keeps going up, and it definitely has in say New York City and other areas, if that underlying cost keeps going up,

Then that is going to push up the cost of housing regardless of what you do on. Now, again, like maybe zoning can make help with productivity gains, etc. Nonetheless, it continues to get more expensive for the actual construction of a home, unlike many other built goods, right? I mean, the issue is that we don't build homes in China. Unfortunately, if we did, that would solve the problem. We don't. And maybe we could talk a little bit about why we don't, but maybe one day we will, but that would help a lot.

Well, on that note, I mean, there have been attempts through history to have prefab houses, right? And some of the old prefab houses that you order from like the Sears catalog in the 1920s or 1930s. Those are gorgeous, and I would happily live in one of those today. But we should talk about why it seems difficult to build houses or at least more expensive than some other things.

And also why it seems that manufacturing other products seems to get more efficient over time while housing seems to kind of stack. All right, so we do in fact have the perfect guest. I'm very happy to say we have Brian Potter who is the author of the construction physics newsletter on Substack. I subscribe to this and it's fantastic. He is also the author of the new book The Origins of Efficiency and a senior infrastructure fellow at IFP. So really the perfect guest. Brian, thank you so much for coming on the show. Thanks for having me.

Um, congrats on the book origins of efficiency. Are we right in thinking that housing hasn't become that much more efficient of a process?

You are correct. And you know the productivity metrics there's th those metrics are quite complicated and you can measure in different ways and different things show slightly different trends. Some of them just show flat productivity rather than falling productivity. But the general point is yet, yes, it has not improved unlike almost every sort of other like physical good has gotten, you know, cheaper and less expensive to make over time. Housing is not like that at all.

And you uh you worked for several years for a uh prefab housing company or something like that.

Yeah, my background is in as an engineer. I worked as a structural engineer for about uh 15 years. That's what my training is. And just that is basically just designing buildings. So I designed parking garages and water treatment plants and apartment buildings and and stuff like that. Um and I had yeah worked at different facets of the industry and as you say it it seemed extremely inefficient to me these you know we're putting up these buildings we're putting up these houses you know on site by hand with a guy with a hammer. You know it doesn't look that different than it did a hundred years ago. Uh we're you know on my side it was like we're designing these buildings from scratch every single time instead of designing it once and then making you know a million copies of it like you do with a car or something like that. So yeah, it seemed when I was, you know, much younger, it seemed like incredibly backward and and inefficient to me.

And then in about 2018, I had a chance to join an extremely well-funded construction prefab startup called Catera that would pay, you know, promised like, oh, we're going to come in, we've raised all this money, we're going to change all that, we're going to build buildings in factories, and it's going to be so much more efficient. We're going to be the Henry Ford of housing and we're just going to sweep away these old methods of of of building. And uh it uh long story short, it did not work out at all. They had raised a very very large amount of venture capital um about like two or three billion dollars in in VC and they burned through it in a few years and declared bankruptcy.

Oh dear.

Um and then the Yeah. And in the aftermath of that, I wanted to understand why had it all gone so wrong beyond just, you know, startups are hard, right?

Startups have a high, you know, a naturally high failure rate. Um,

Uh, you know, and whatever operational missteps they may have made because I, you know, I I kind of came to believe that there sort of fundamental thesis that if you build these things in factories like we build everything else, it will get cheaper. I came to believe that was either sort of not quite correct or at least missing a very big piece such we didn't know specifically what we needed to do in these in these factories to make the cost fall.

So why are houses I guess so customized and not standardized to your point?

Uh yeah so there's a lot of reasons. One is just, you know, to go back to your point about permitting, there's a very very large number of uh permitting jurisdictions in the US, something like 20,000 different jurisdictions which each might have their own, you know, slightly different versions of of building codes that sort of need to be enforced. Um there's just, you know, different building sites are different sizes and shapes and have different, you know, soils will require different uh different things, different environmental design factors, you know, different wind speeds. If you're in California, you need to design for earthquakes. If you need in Florida, you need to design for hurricanes. So, there's all these factors that kind of um conspire to uh make uh the sort of housing market very kind of fragmented and it's hard to build a very very large number of like a uniform product. The builders do try to do that. They have somewhat standardized designs that will plunk down in like different developments across the country, but it's um it's it's quite difficult to do in in really really large numbers.

Talk to us a little bit more about prefab housing specifically or housing in a uh a factory. What was the I mean you there have been multiple attempts and as Tracy mentioned it's existed for actually a long time and uh you used to be able to buy a home from Sears. Talk a little bit about the sort of recent trends in this area because you do over the last several years you've heard a lot of maybe the last decade you've heard a lot of ex excitement about this idea and as you mentioned you worked for a well-funded startup why now like what's the general thinking and then maybe we could get into the sort of the gap between the theory and then what happened in practice

Yeah so prefab is like this perennially popular

Idea it's always it's always the next decade

Yeah, yeah, it's always the next big thing that's going to come along and so every you decade or two, you will kind of see these these trends like some big, you know, company will come along. It's like, oh, I'm going to revolutionize the building with with the housing industry with uh, you know, prefab construction. So, and yeah, these these attempts, you know, stretch back a pretty long way. So, right after uh, you know, and I be I became something of a student of the of these companies that have have tried and and failed. So, right after World War II, you have this this company called the uh called the Lustron Home Company, which made these very interesting homes, many of which are still around. You can actually go um go look at them. Um but they're made these like steel panels covered in porcelain that turned out to be quite wellmade and quite durable. But the company went and they got government financing to sort of fund their operation. They had set up shop in this big uh what was once one time the biggest factory in the world. It was this repurposed aircraft engine factory.

Oh, they look pretty cool. Yeah, they're very they're very neat. Uh there's I live in Atlanta and there's some in Atlanta that you can go and look at. Um but uh they went bankrupt after quite a few years. In the 1960s, there was this company called Sterling Homex that said, "Oh, we're going to be the they uh we're going to be the next uh you know, Henry Ford of construction." They went bankrupt in a few years.

I'm sensing a pattern.

Yeah. They were a part of this government program called operation breakthrough which was trying to fund um you know just again fund the development of like mass production housing techniques. That program basically failed. None of their companies that they funded and there were quite a few of them ended up sort of succeeding in breaking through and and creating sort of these mass-produced housing uh housing uh products. Um, so you kind of see the same thing happening over and over and over again where people are trying to use uh prefab to like dramatically reduce the cost of construction. And there's ways that you can do that in kind of like a very limited sense. Um, and you can also like build a successful business off of prefab where you're not necessarily trying to compete on low cost but things like oh we can build much faster or or stuff like that. But nobody has managed to do what everybody is trying to do which is like you know be the Henry Ford of housing.

Okay. What's the sticking point then? I know you walk through some of the reasons that houses tend to be customized things like soil types and zoning and regulation but is is that just the story of why we can't have prefab houses that are like across America?

So there's yeah I kind of look at this in in two ways. I I think of it as terms of it's very hard to introduce uh some of these like improvements these sort of strategies you have for improving uh efficiency into the construction process without accumulating like more cost elsewhere in the process. And then so at the same time that that is you know these the gains that we've found in terms of like these strategies for efficiency improvement have historically been much more limited than you see in like if you're producing you know a car or a light bulb or something like that. And at the same time, we've continually added like more regulations and more difficulty, you know, more administrative overhead and getting stuff built, which is slowly like driving up of up the cost even as like uh we're getting, you know, we're not sort of in accumulating these productivity improvements.

What's an example of okay, this on paper looks like a productivity gain that we get in a factory, but then you pay for it somewhere else.

Sure. So, you know, a very simple example, you know, if you're building a car in a factory, you build the entire car, right? It rolls out of the factory line and the car is complete. Uh, a building is too big to come out of the factory all in one piece and because, you know, get something to sent to the job site or whatever. So, what happens is you have to sort of build your building in chunks. Uh, you know, that's why prefab construction is often called modular construction because you build it in these modules and there's different strategies for like, you know, how you break a building into modules but you're always like breaking it into these pieces and what that what happens when you break it into pieces is that you have to add a bunch of extra stuff that you wouldn't have had to add if you were just building it on site. So transporting it down the road, you have to sort of design that to be rigid enough to like not collapsed under like the load of like being bounced along on a truck. I used to work at a company that built um concrete parking garages which are built from pre-fabricated, you know, materials. They build these like big concrete pieces and then drive them down the uh road. And that was like the design that was like the controlling design factor is designing to survive the loads that would concur during transportation. And then once you have those pieces um driven, you know, driven down and gotten to the job site, you have to like stitch them together. And so you have all these extra connectors and this extra complexity and attaching them together that again doesn't exist in sort of the conventional construction process.

It's kind of crazy to think that whether we're talking about a house or a parking garage that's presumably designed to last for decades that a very significant factor in engineering is that few hours that it goes down the road. I mean that's what it sounds like. Yeah. Like that that's like that that's going to last forever but it has to be there just for that like that short transport period.

In the case of parking garages for sure.

I guess that's why we do have some prefab houses in the US and they're called trailers because you can actually move them,

Right? Um, okay. On this note, how much of this is also maybe a liability question where if you were producing prefab houses, presumably it's not the company that's putting them together on site. It's usually contractors or someone you hire to do it. And I imagine there's like a disconnect between the contractors on the ground who are putting the stuff together and the company that's building the parts.

There is to to some extent. I think that specific liability is not the is not necessarily the biggest controlling factor. What is a big controlling factor and I've I've written an essay about this is that uh the uh sort of construction cost like the cost of in developing and putting up a new building the sort of distribution of that is extremely fat tailed and right skewed which so basically what what that means is that a construction project that goes under budget that goes really really well maybe you're like 10% overbudget or under budget maybe you're 15% under budget you know you can't push down that cost very much But a project that goes over budget can be 100 or 200 or 300 100% overbudget. So what that means is that there's so the ri you know the risk involved in like introducing some new system that you're not sure really how it works. It just it's the the payoff is so asymmetrical. It's so much more likely to cause like a massive overrun than to sort of save a relatively small amount that these sort of builders are like rationally incentivized to not introduce some radically new technology that's going to dramatically change how things go because the potential costs are so extremely high

For the sort of maybe dovetailing with Tracy's question you know or maybe in the case of Catera or some of these others it's like okay they have some vision but because they're not doing the entire thing all around the country including presumably transport etc. How much is the issue like maybe scale on those ends experience with that? Like in theory if you had a really big experienced on the ground distribution and uh sort of final assembly uh workforce or whatever it is like how much would that help and how much is the issue that people who work on the ground etc just aren't doing this full time?

Yeah, that's a part of it is like you know this modular construction is kind of a somewhat yeah niche aspect of the construction industry. There are you know quite a few experts in it that know how to do it and they do it fairly regularly and you can usually find them in like any given market. I would say to kind of you know a different point that you kind of alluded to scale is like another definitely big issue where it's like it's not like a semiconductor where you can make a million of them in a factory in Taiwan and ship them all over the world right that's another big aspect that makes it hard to sort of you get gain efficiencies from uh from factory production and and goes to what I was talking about where it's hard to do that without introducing cost elsewhere in the system that the the cost of like moving these modules around is so large that it's typically not economic to ship it like more than a day's drive because then the costs just get so high. And so what you have instead of like one giant factory that can like produce, you know, at really really large large scales, these even even sort of these modular companies that are doing like large amounts of business, they will tend to be like done in factories like that are spread across the US.

So no gigafactory.

Yeah. So, like you you talked about like trailers, like manufactured homes, which is one area that you can actually see some pretty substantial cost reductions due to various things. Partly it's like they're targeting a lower end of the market and stuff like that. So, like the finishes and stuff aren't quite as nice. But if you look at like the way that their business is structured, they have like more than a dozen factories spread all across the US and then each one is serving like a day drive catchment area. So, they're not building like a, you know, 50,000 houses or whatever in one factory. It's more it's closer to like 500 or a thousand or something in in one factory. And so the economies of scale that you can capture in like a given operation are just fairly limited.

Joe, I have to mention here we didn't tell Brian that we were going to be talking about housing construction and trailers and all of that. This is all just off the top of his head. So I know he sat down and we were like, "By the way, can we talk about this?" And so yeah.

Um Okay. You kind of alluded to it in your previous answer, but if you were going to choose a manufacturing process that was the polar opposite of housing, what would it be?

Oh, interesting.

Uh probably something like chemical manufacturing you know or like you know raw material manufacturing something like uh you know steel or something like that because chemical manufacturing you it's you know the sort of ultimate form of like a manufacturing process which I I kind of talk about in the book is what's called like a continuous flow process where it just like continuously transforms your your group of inputs into the output what of what you're doing just one you know continuously without any sort of waiting interrupt interrupting or batching or or anything like that. It just all flows and is continuously transformed during the process. And that's what like chemical manufacturing is is you know very efficient chemical manufacturing is is done. And most other like really efficient manufacturing processes they ultimately try to sort of converge on something that looks like that. So like the assembly line and like mass production is kind of a way to like get something that approaches like a continuous flow process in um these like very complex manufactured goods with like thousands of parts. That was why it was such a huge that's a huge transformative thing is because now you could have this continuous flow process which existed before for you know much simpler things but now you could adapt that to like these really really complex goods like cars and that's why it was such a transformative thing when it was sort of invented in the early 20th century.

Um I'm curious does the do the economics or the diseconomics of building houses um in this way do they inform your perspective at all on the prospect for small modular nuclear reactors? I mean there sounds like there are some similarities especially when it comes to like learning curves etc or the soil of where you're building.

Do you have any views on that industry having come do you think there's any parallels? So yeah,

Because this is also something that is always the next decade away, right?

Yeah. Yeah. Yeah. So yeah. So so I talk about economies of of scale in the book and we're talking about like nuclear stuff. There's like kind of two related economies of scale that in some ways kind of pull in different directions. One is called and you see this with like other sort of processing equipment which is I call geometric scaling which is basically as you build a piece of equipment physically larger it costs less per unit of whatever it is in that you're producing basically because of like area volume relationships as you double the volume of something you less than double the amount of like surface area required to enclose it and be that like tanks or pipes or stuff like that. So like chemical processing equipment or like power plant equipment uh tends to get cheaper per unit of what you're producing up to a a certain point and eventually you run it gets you run into difficulties and those scaling laws stop working. Um and this is one reason why power plants you know and electricity got cheaper over sort of most of the course of the 20th century is because we're building these power plants physically larger.

Um but with nuclear you run into these very nasty uh disconomies of scale around the same thing which because as a nuclear plant gets bigger the accidents they could have get more dangerous for like a very small reactor you can't you can make it so you basically can't have a meltdown. There's just not enough heat in there to like melt through your container or or whatever. Um but for a very large one the heat gets so large of that you know it um it ends up being the the accidents become the risk becomes much much larger and so you have to have these like very involved safety systems that maybe you don't need to have on like a smaller reactor. So that's go you know that is something that maybe points to um points to a reason why these maybe small modular reactors were better. Plus, as you say, now you're getting it's sort of the benefits of like getting Yeah. series production uh and you know, repeating it over and over and over again. Um but there is that trade-off is that you're not getting those geometric effects that traditionally like a lot of this like processing type equipment has relied on to get cheaper.

This actually reminds me of something I wanted to ask, not necessarily about nuclear, but just manufacturing in general. When I think about efficiencies, I think about stuff like, you know, reducing costs as much as as possible, maybe just in time inventory management and stuff like that. The problem is that if something goes wrong, there's very very little cushion for protection. And I'm curious when it comes to efficiency, how do people balance streamlining production with, I guess, building in redundancies in case the worst happens?

Yeah, that's that's really interesting. And um I the Toyota production system slash lean which is like a descendant of it which I talk a lot about in in the book um is kind of an interesting balance of of these two things because everybody thinks you know lean or to production system it's like oh yeah just in time get rid of inventories you don't have all the costs you know your factories can be smaller blah blah blah.

Um, but really what it is it's sort of a it's sort of a balance of like for every factor that you can control try to like have like the you know the flow be like as smooth as possible and a minute you know eliminate all this like you know slack and waste in the system but for the things that like you can't control you should have your system adaptable to sort of you know accommodate that and be have sort of a flexibility built into it. So like one example is um uh after the sort of you know to bring it back to nuclear after I believe the sort of Fukushima nuclear accident in Japan, Toyota had a uh it ran into like a shortage of like automotive semiconductors for a while and they realized that this was like a big like potential like just bottleneck in their process. So after that they even though they're like so focused on you know inventory minimization and whatever after that accident they um basically decided they were going to like stockpile certain critical things that maybe you know had diff you know difficult lead times or or something like that. And so they stockpiled like ser automotive semiconductors. Um, and so during COVID when all of a sudden these like you know there was all of a sudden a huge like shortage of like automotive semiconductors, Toyota for a while was like much less affected by some of these other these other car manufacturers because they had learned the lesson that oh yeah in sort of you know logistics trenches these semiconductors can be quite difficult to get. And so, you know, basically people were saying, "Hey, I thought you were all about like inventory minimization or whatever." But then somebody, you know, some spokesman basically responded like, "No, this is basically a standard lean solution where it's um, you know, you try to deal, you try to, you know, minimize the things, you know, the sort of waste in the sort of areas that you can control, but in the stuff that you can't control, you need to sort of adapt your system to accommodate it."

America seems to have gotten because I'm trying to understand sort of big picture the trajectories in this country for various things. You know I joked in the beginning the problem with housing is that we haven't found a way to offshore it to China. If we did then probably housing would have gotten cheaper like everything else that we uh have manufactured there. But I'm curious um do you have like a theory because I have heard a satisfactory answer why the US seems to have gone backwards in plane construction. like why does Boeing seem to be worse at the job of building an airplane than it was say 20 years ago?

Oh yeah, that's interesting. One point about your your uh housing construction in China. There actually are companies that like try to do this stuff in China. They will like manufacture these modules, but again to my point about you know you incur all these extra cost and all that, you know, they have to you have to break them into modules and put them on these boats and bring them over. So it's, you know, it's a very sort of niche business. Uh in terms of Boeing, I mean the standard, you know, story there is that uh you know, they went into this MC merger with McDonald Douglas, the finance people. Uh

I don't like the standard story. And here's why. Because I love it so much, right? And so it's like, oh, this is such a good story. It's too perfect. They brought in the finance people. They care about the shareholders too much. Looking at the Boeing stock chart, you would not think this is a company that cares a lot about shareholders, right? They care about shareholders too much. If they had just stuck with the engineers. I love that story and because it feeds all of my biases, I'm like skeptical of that.

Yeah. And I mean, you know, a big part of it is just manufacturing commercial aircraft is like extremely hard. There's like a reason that uh only like right now two maybe three maybe four uh companies in the in the entire world do it. Boeing, Airbus, Embberer in Brazil of all places and then Comarmac is looks like it's going to be the big fourth one. you know, Japan tried for many years to sort of break into the commercial aircraft manufacturing uh market and they were basic, you know, after spending like billions of dollars, they were basically gave up just because it was so difficult. There used to be many more competitors in the 1970s and they sort of got winnowed down one by one, you know, one or two pro, you know, plane projects that don't pan out and you're basically out of out of business. That's essentially what happened to Lockheed. Um and so it's just a very very difficult business to sort of be in. And so Boeing is having a lot of struggles right now. Um but you know earlier in the 20th 21st century you have Airbus having a lot of struggles. They had all sorts of problems. Um I think on their uh A380 that they were rolling out they had like a lot of delays and stuff they had to fix on it as well. So a big part of it is just it's a very difficult business to to to be in. there's very few people that can kind of really do it effectively.

Here's a question that I've wondered about and maybe it relates to planes and maybe it relates to housing, maybe it relates to other things. Like if when you see us going backwards or stalling out like here's my question. Could we have some sort of I don't know if Dutch disease is the right answer where it's like maybe in the middle of the 20th century there were people who were really you know working on the floor or working in management at Boeing in a plant somewhere in Washington who are really on the ball and really um whatever and now in that same person in 2025 could make 10 times as much money working for a startup in Silicon Valley or Wall Street etc that there's some sort of like levels of you awareness or skill or agenticness or whatever, etc. And that because we have these other areas where there's so much money that there's like a type of individual who used to maybe work in a sort of construction related field that left that and then it's changed the talent distribution with

Yeah, I think that's definitely part of it. There's like a you know theory in certain circles that we've just we've gotten like in a sense too good at like allocating talent efficiently or whatever. So like somebody extremely like talented and smart it's like okay we're going to send you to be a software developer or an AI programmer or work in finance or whatever and you're you know there's going to be you're the value that you can contribute there is going to be very highly rewarded but what that means is like there's you know there's fewer people going into sort of these other um other sort of you know engineering disciplines. I worked as a structural engineer and which and it which is you know it plays reasonably well but doesn't you know not you know Silicon Valley software developer salary and I used to sort of tell very young engineers or engineers thinking about uh you know in in college or whatever I used to say you know if you're smart enough to be a structural engineer you're smart enough to you know switch to like software software development and make a lot more money and have a much less stressful job.

Yeah. Um I've I there's like there's like this is I think one part of the reason um why uh you know Bell Labs the very famous AT&T R&D lab that had all you know the US invention factory for most of the 20th century um why it would be very hard to have a sort of organization like that today is because the quality of talent that you could a place like that demanded would be so much more reward

Somewhere else. Just quickly on aircraft, it's not like the US doesn't have a history of producing like largecale amounts of planes. And the example that everyone tends to turn to is in World War II, right? We produced, I think, hundreds and thousands of aircraft very very quickly. What were the conditions that existed back then that allowed us to ramp up that kind of production so fast?

Yeah, I've I've written about uh yeah, aircraft manufacturing during World War II. The US actually built more airplanes during World War II than have actually been built uh during for commercial purposes for the entire history of aviation.

Stunning.

Um I think a big part of it is just the US was just such a huge manufacturing power at like the mid 20th century. You know, we weren't building very large amounts of airplanes, but we were building, you know, we were building more cars than anybody else. We were building more machine tools than anybody else. We're making more steel than anybody else. The US was like, you know, the China of today. It was like the by far the largest manufacturing country in the world. And so we just had the ability to sort of redirect this um manufacturing capability and talent into sort of these these new industries. I think that was a very large part of it. You know, people think about it's maybe since we're talking about war, we're talking about the US is the China of the 20th century, which is kind of

It sounds very Eric Adams, doesn't it? The US is the China of the 20th century.

The China could there be more than one global manufacturing hub? Because when you think about economies of scale and you think about network effects, you know, it's like how many other cities in the US have tried to be the San Francisco or the Silicon Valley and they've never you really never replicate it. When you think about sort of elomeration and network effects, etc., And you think about what it takes for manufacturing and you think about okay now in the 21st century China is the China of the 21st century and you think about the efforts at US sort of re-industrialization because there's all this angst about whatever we can't build advanced things it's like well can we have advanced things if we don't also do the simple things like washing machines etc like can there be more than one in the world at any given time or does market logic sort of dictate that one will be the powerhouse

Yeah that's a really interesting question and so my my first answer is that I don't know and and the the second answer is

I think historically kind of what you see is that like transportation costs have kind of been a you know something of a barrier to uh preventing one uh country from like just dominating production of a certain thing. So even like when you know in the in the 80s and 90s when Japanese like car manufacturers were like trouncing the US, it it was still so expensive to like ship them from Japan or whatever that you you know they still it was still valuable to like set up local manufacturing operations um in in the US. They still wanted to build their plants in in in the US to sort of make it not so expensive to sort of transport these things. Um, so I wonder if like as transport, you know, transportation has gotten more and more inexpensive, I wonder if that has like shifted the logic of this and you can have this sort of like concentration of production in a way that maybe it wasn't quite didn't quite make sense to do uh before when like these transportation costs bit a little bit harder. I don't know. But it's an interesting way of interesting question.

Uh, just going back to Boeing for a second. So, we talked about the classic story of what went wrong at Boeing and it's that tension between I guess the manufacturers on the ground and the executives in the boardroom who are trying to cut costs and boost the share price and things like that. Is that is that like a trend or a thing that you tend to see across manufacturing where the manufacturers are constantly under pressure from the finance execs on the top or are there companies out there who have like cracked that relationship where finance and manufacturers engineers kind of work together?

Yeah. Um I think you know the the pressure definitely seems to be you know constant pushing to find uh you know lower sort you know lower costs. Um and which is one reason why you know the trend in the second half of the 20th century was like just outsourcing finding new places to sort of make this stuff that is uh you know cheaper. I talk about a Nike in the book and there's sort of an interesting path that Nike takes where it's like constantly hunting the the cheapest sources of labor where they start out they start out manufacturing somewhere cheap, right? They start out manufacturing their shoes in Japan because it's much cheaper to make them there than in uh than in the US. But then as Japanese labor starts getting more expensive, they move their manufacturing to uh Taiwan and then from Taiwan to China. And I think now it's almost all done in like a Vietnam and Indonesia. So it's just constant hunt for the new sources of labor and you see that just broadly across industries like um where New York which used to be sort of a uh you know a garment manufacturing center but as New York labor got more and more expensive they moved you know AC first you know out of state into New Jersey and then to the south and then over o overseas. So there's this constant pressure. But then there's also companies that have like I think bucked that trend and been you know more like we think the returns from like investing in our workforce and its capabilities um is going to be valuable enough that we're going to do it. I think you know a company like Newor which is a steel company in the US which is like very sort of famous in terms of like how it invests in its workforce and things like that and they've been extremely successful in doing that. They grew from nothing to being the the largest steel manufacturer in the US I believe.

That just reminded me something that I really want an answer to and you might actually have it. Why are communists obsessed with steel?

Oh good question. I just think, you know, it was I think in the, you know, sort of the early 20th century, late 19th century, um these were the things that like represented like industrial might and like successful country or whatever. Um and then they never kind of like got out of that uh sort of framework or whatever. So like one criticism of I you know the Soviet Union um was that you know they were really maybe good at sort of building these like industrial things but they never managed to build like a successful computer or anything like that. They couldn't get out of their sort of mindset. It's it's sort of maybe easy to measure like success in terms of like output or whatever. It's this like relatively relatively simple commodity that you can you know

Right this physical thing you can point to look at our stock pile of steel.

The more tons you have the better you've done. um with more complex things maybe it doesn't uh you know it's not maybe so amendable to sort of this like top down management style. I don't know but uh those are some of my theories.

Let's get back to housing because when we started and Tracy was talking about the fact that even setting aside the prefab stuff and you're talking about it the actual on the ground oneoff etc home building it's arguably uh uh by some measures productivity has gone backwards or maybe it's flat. setting aside like prefab, you mentioned zoning regulations, etc. What is your view for why just the normal sort of style of housing construction hasn't gotten any more efficient over the years?

Yeah. And so I I I talk about the kind of this in the book. This goes to sort of the this main thesis of the book where I sort of break down there's a specific points of intervention that you can do in any process to to make it cheaper. I've talked about these a little bit. Um you know, economies of scale, you can make more of of what you're doing. Um, you can introduce some like fundamental new technology. You know, an example of this to talk about steel, like the best summer process for making steel, which like fundamentally was simpler and took less inputs than the processes that that came before it.

Um, so I kind of break down these different factors in the book of like what you need to do um to make some process more efficient. And if you can do those things, your process can get cheaper and you can sell it for less. Um, and if you can't do those things, you're if you know, you can't make the process any cheaper. And all those basically paths uh to improvement are either like blocked or like very difficult to do in construction. So like we talked about economies of scale uh you know that's a big one. We talked a little bit about risk aversion. It's hard to kind of introduce like fundamentally new technology. Uh the cost of your inputs is like a big one. you know, the the the materials and sort of energy and and and labor that you're using uh in the process. Those costs are hard to sort of reduce in construction. Sort of it's very hard to make building materials any cheaper than they already are and kind of in like you know dollars per pound or dollars per like cubic foot terms. Uh they're basically as cheap as anything that civilization produces. It's not easy to make those things cheaper than they already are. It's also hard to use less of them in a way that doesn't make the house worse. And so all these paths are that you have, you know, that a process can take advantage of to sort of get less expensive. Uh they're all very uh difficult to do in construction.

I just remembered another question that I always want to ask people and I feel like you will also know the answer to this one which is why does America build houses out of uh wood frames and timber versus the rest of the world where they tend to use a lot more cement and concrete.

Yeah. So the wood you know uh US has historically built a lot of stuff out of wood because we had such huge supplies of it. Uh in places like Europe, a lot of those uh you know large forests got cut down way earlier. And so they just didn't have the huge huge supplies of uh of wood that uh that the US had. And so that kind of you know shaped how the technology developed because wood is very very inexpensive. You can use it. Um it's a very inexpensive way to sort of uh build a house. Um they do actually do sort of still timber frame construction in uh in other parts of the world. It's maybe not quite so dominant, but it's like used in uh you know the UK, Germany uses it. Germany actually they have some like prefab construction startups that are made using like timber frame timber panel construction. It's used in a lot of the Nordic countries uh that have a decent amount of supply of wood. So it's not like unheard of. It's it's most dominant in like the US and in Canada, but in other places that have a decent amount of wood, they tend to use it as well.

Should we completely give up on uh prefab or major efficiency gains in housing in the US or like are you still open to is it just there's not a way to do this or is there still out there some magic bullet that's waiting to be found?

Yeah, you should tell us what happened to the company you were working for as well.

Yeah, so they Yeah, they raised a huge amount of money. Uh yeah, and they basically after raising like several billion dollars and building all these factories to like f you know, massproduce housing, uh they basically ended up uh going bankrupt part, you know, partly because they were never able to, you know, sell enough of these, you know, buildings that they were trying to sell. It's very hard to like pivot your operations when you're like invested in all this like physical infrastructure. Um and then COVID, you know, was kind of the final nail in the coffin. Uh I'm not totally, you know, I'm not completely pessimistic. I mostly just think if you're gonna sort of try to change the industry with with with prefab, you need to have a very strong thesis as to what specifically this time is different and you know why it didn't work previously and what is now has changed what constraints have been relaxed um that uh will make it possible now and I don't have an you know any number of things could change that you know certain materials get cheap enough uh it gets like inexpensive enough to build a factory you know any number of things uh I personally Personally, I think that uh very very good, you know, automation and robotics have a good chance of of changing this just because, you know, even if nothing else work, you know, I'm not the most creative person in the world. You know, I can't necessarily come up with like some amazing new way of putting a house together. Um, but if even if nothing else works, if you have like a robot that costs $5,000 and like basically duplicate, you know, 90 95% of what a person can do, that would effectively solve your problem, right? you drop in these robots, um, and you build, you know, whatever it is that you're trying to build using this robotic, much less expensive labor. Um, and I have a sort of a vignette in the in the conclusion of the book as to what that might look like in the future.

So, we could get the uh the drones that go up on the roof and fix leaks and stuff like that.

All right. Anything's possible.

All right, Brian Potter, thank you so much for coming on OddLots. Really appreciate it.

Thank you. It was great to be here.

Yeah, that was a lot of fun. Thank you so much,

Joe. That was a really fun conversation. Yeah, that's really fun. I love Brian's uh stuff and I feel like one of those conversations where both of us could just ask a bunch of questions we've been meaning to ask for a long time.

Um I did read apparently his new book, The Origins of Efficiency, has a bibliography that's like 600 items long. Amazing.

Which is incredible um to think about him pouring over like 600 different reference sources. We didn't get into it in this conversation, but he's written about his process before of like, okay, I want to learn about some new industry like fighter jets or whatever it is and how he goes about um diving into it, etc. And yes, it seems like

People who talk about stuff I should probably read that.

Yeah, totally.

Um, speaking of which, I am very proud of us for not mentioning the word deflator at all in a conversation about housing productivity, housing construction productivity. I should just mention that Richmond Fed paper that I talked about in the intro. So, their conclusion was sort of an amalgamation of all the factors that we talked about um in this conversation. So, they basically said

They think that smaller construction firms have reduced incentives to invest in technological innovation because you don't get that economy of scale. And in fact, if you look at areas with uh stricter zoning laws and things like that, you can see that they actually have even less construction productivity than other areas again because like the incentives and the economies of scale just shrink and shrink and shrink. So that's kind of interesting.

It would be interesting. Um something I was sort of curious about during that is I wonder if if you just look at one market where it's very flat and dry. I was thinking about Arizona or something like that. Like have they seen productivity gains there? Especially thinking about some of our conversations with J say Chase Emerson where okay there's there's one master plan and a bunch of houses we all driven by them that all look exactly the same. Do you get those learning curves at least on some limited effect? I would be curious about that.

I feel like to some extent that is the story of the sunb belt states right you have cheaper houses there because you can have these big developments

But it's scale it's scale way down. And so I like that point that he made which is that you can identify the sources of efficiency gains. Scale is one, a scientific breakthrough such as the best of our process in steel is another one, etc. But they're not going to they're not magic, right? They're not just going to emerge. you know, you probably get some some efficiency gain just from, you know, repeatable process or a sales labor force that gets better over time, but by and large, you're not just going to the silver bullet isn't going to come out of nowhere. And especially the fact that there are all these sort of very fundamental reasons why you can't have scale or even the factories that build the prefab homes cannot be a gigafactory due to distribution cost. You see why that you run into that constraint.

That's what I was thinking. It's like the tyranny of of physics basically. Like even though we have all these technological breakthroughs, transportation has gotten cheaper over time. You're still limited by geography and the realities of, you know, real world objects. And I guess that's why you can have gigafactories for something like chips, which are really easy to send abroad. They're tiny, but you can't do it for something large. And I really am I I curious about the sort of the point about talent distribution and whether there you know in the 1950s

The the for a certain type of uh talent the the Boeing factory was the best place they could have gone the cutting edge and that was the cutting edge and now the Boeing factory still more or less resembles the Boeing factory of the 1950s with maybe similar pay scales rel it's and so forth. I'm sure the nominal pay has gone up a lot, but there are just so many great opportunities to make 10 times or much more money elsewhere. And so that the sort of remaining talent, whether there's some, you know, whether that has an effect on how well it can operate, I'm very interested in.

You know what Boeing needs?

Go on.

They need like a Madmen style TV series that glamorizes the aircraft manufacturing process and gets people interested in it again.

Yeah. No, that's a really good idea.

All right. That's a free idea for Boeing. Shall we leave it there?

Let's leave it there.

All right. This has been another episode of the All Thoughts podcast. I'm Tracy Aloway. You can follow me at Tracy Aloway.

And I'm Jill Weisenthal. You can follow me at the stalwart.

And if you enjoyed this conversation, then please like, comment, or subscribe to the channel.

Thanks for watching.