- The authors show that since roughly 1980, the relative supply of buildings has declined while the real price of buildings has risen.
by Vanguard Staff
A newly-published paper in the European Economic Review finds that construction — specifically the supply and price of buildings — plays a central and previously underappreciated role in determining workers’ wages and the share of income going to labor, with important implications for housing policy.
The paper, “Capital Composition and the Decline of the Labor Share: Why Buildings Matter,” analyzes more than 200 years of data from 16 advanced economies and concludes that the long-term decline in labor’s share of income is not primarily driven by the overall amount of capital in the economy, but by changes in the type of capital being built .
The authors show that since roughly 1980, the relative supply of buildings has declined while the real price of buildings has risen. Because workers and buildings function as complements in production, a shortage of investment in buildings reduces labor’s share of income. In contrast, machinery and equipment tend to substitute for labor, pushing wages down when investment shifts toward automation rather than toward physical space.
Using newly-constructed macroeconomic data stretching from 1820 to 2021, the study finds that rising costs and weak productivity growth in the construction sector led to a sustained decline in the ratio of building capital relative to overall output. That decline, the authors conclude, has directly contributed to falling wage shares across advanced economies, even as overall capital investment increased.
While the study focuses on non-residential buildings, its findings carry clear implications for housing. The authors identify rising construction costs and constrained building supply as structural forces that limit the productive capacity of labor. Those same forces also restrict housing supply, drive up housing costs and indirectly weaken workers’ bargaining power by increasing the share of income devoted to rent.
When housing construction fails to keep pace with population and employment growth, workers face higher living costs while also experiencing downward pressure on wages as building scarcity limits the environments in which work takes place. The study’s findings suggest that housing shortages and wage stagnation are not separate problems, but interconnected outcomes of prolonged underinvestment in the built environment.
The research challenges the idea that capital investment necessarily harms workers. Instead, it argues that what gets built matters. Investment in buildings — including housing, offices and industrial space — supports labor by expanding the physical environments required for productive work. When zoning, land constraints or regulatory barriers restrict building, the resulting scarcity benefits capital at labor’s expense.
The paper also distinguishes its conclusions from earlier research that emphasized residential land scarcity or automation alone. By isolating non-residential buildings, the authors show that underbuilding has economywide effects on income distribution, reinforcing inequality even when technological growth continues.
The findings suggest that policies that expand building supply — including housing — could strengthen workers’ economic position by lowering building costs, increasing real wages and raising labor’s share of income. Measures that accelerate permitting, reduce construction barriers and improve productivity in the building sector may therefore function not only as housing affordability reforms, but also as wage and inequality policies.
The paper was authored by Jacob Kerspien and Jakob Madsen of the University of Western Australia and Holger Strulik of the University of Göttingen and was accepted for publication in December 2025 .
Follow the Vanguard on Social Media – X, Instagram and Facebook. Subscribe the Vanguard News letters. To make a tax-deductible donation, please visit davisvanguard.org/donate or give directly through ActBlue. Your support will ensure that the vital work of the Vanguard continues.
Of all the absurd housing articles you’ve posted, this one from Europe win first prize.
Let’s look at a couple of claims, here:
“That decline, the authors conclude, has directly contributed to falling wage shares across advanced economies, even as overall capital investment increased.”
Yes – the same thing is occurring in regard to Amazon, driverless cars, artificial intelligence, etc. Does that mean we shouldn’t pursue those technologies, because someone is then out of work?
And when “capital investment” increases (as you’ve noted), does that mean that the money “disappears”, or is it just shifted (and increased) to companies that are implementing such technologies?
“Instead, it argues that what gets built matters. Investment in buildings — including housing, offices and industrial space — supports labor by expanding the physical environments required for productive work. When zoning, land constraints or regulatory barriers restrict building, the resulting scarcity benefits capital at labor’s expense.”
Apparently, the author is not familiar with telecommuting, nor is he/she familiar with the glut of office/retail space in this country, nor is he/she familiar with the housing markets that are crashing throughout the country.
The entire article reminds me of those who tout “construction jobs” which are TEMPORARY in nature, as a reason to build. And amazingly, they do so as if it’s a legitimate argument.
Sort of like building more housing to avoid laying off teachers or selling unneeded facilities.
These type of arguments are not in the OVERALL interest of the societies from which they emanate.
These are the type of arguments that suggest we should take steps to ensure that blacksmiths still have work making wagon wheels, etc.
Or the type of people “worried” about a lack of new car sales, as people keep their existing cars longer.
Perhaps you want to disprove the peer-reviewed article…. Start here:
You first – go ahead and analyze that formula and see what conclusions they’re drawing from it.
But you already know that conclusions are not the same thing as data, nor do they necessarily consider other factors (some of which I just mentioned).
In any case, I noticed that they (as with all of these other studies) don’t even bother to define “housing shortage” in the first place. Nor are they limiting their analysis to housing, since they discuss ALL types of construction.
Also, are they arguing that a housing shortage is curtailing (overall) GDP? I’m not seeing that in the article.
Actually, that’s a good question regarding America, as well – is anyone claiming that a lack of housing is causing the stock market to crash, for example? Or that it would be “double” what it currently is, if only the NIMBYs would get out of the way?
Isn’t that the “bigger” societal question, rather than focusing on individual industries?
(Of course, there might also be studies which analyze the impact of increased GDP on the natural environment – which, due to the limitations of studies themselves – would also not be “complete”.)
We do know that climate-changing emissions were drastically reduced during the pandemic, due to lack of activity.
There seems to be little doubt that technology is replacing jobs on a broad level, but not that it’s hurting the overall economy.
Me first, you’re the one evaluating claims without looking at the paper, data, or methods.
“ We construct data over the period 1820-2021 for a panel of 16 OECD countries: Australia, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Japan, the Netherlands, New Zealand, Norway, Spain, the United Kingdom, and the United States”
That’s a 200 year study over 16 countries including the US, but you can debunk it without even looking at the data, that’s very impressive.
I’m addressing what you wrote in your article. Unlike you, I’m not running a business where I’m subscribing to sources which house studies.
The fact that you expect your readers to “research” what you write about is a reflection of your reporting (and unwarranted expectations of your readers), not my personal reading comprehension skills.
And what, exactly is their claim anyway? There’s more than one described in your article – it’s all over the place.
Note that I’m “agreeing” with much of what they’re stating, since some of it is obvious. Yes – technology is causing a shift in wealth/ capital investment, and is also eliminating some jobs.
Again, your article literally states that capital investment is INCREASING, not decreasing as a result of this.
The same thing is likely happening in this country, as there are a few “stars” in the stock market propping up the entire thing.
You’re not addrssing what I wrote (since I didn’t write it).
You could ask the authors for a copy of their dataset, they would probably let you see it. Then you could enroll in a few econometrics courses at UC Davis, and then you would be able to evaluate their study. Personally I would limit it to the US and you could probably reduce it to the last 100 years or since the depression. That would give you a good baseline to be able to respond to their findings.
Let’s face it – we’re at a point in time where those with a little bit of money to invest (e.g., via 401Ks) are going to do better than the barista down at the coffee shop (assuming that they’re not also investing).
Then again, Bitcoin has been taking a beating, lately. My advice regarding that is to buy low, and sell high. (You’re welcome.)
But seriously, it does seem like the stock market has been defying predictions for quite some time. Perhaps it’s because our system essentially “forces” many workers to invest in it by incentivizing that arrangement.
Then there’s the vast sums being pumped into AI, by those with more funds. The reason I get to see dead celebrities coming back to life, apparently.
Just saw your other comment. You seem to be assuming that I don’t have training and understanding of how studies are conducted, as well as their inherent weaknesses – depending on type. You are mistaken, if that’s what you’re assuming. I did pretty well in those courses, though it was some time ago at this point.
But again, I’m addressing your article. And some of what I wrote already address the type of questions that would arise, when examining a study.
For that matter, I’m not actually seeing YOU analyze the study – you’re likely just repeating the conclusion, and possibly not even doing so fully.
You’re not going to “win” this argument in this manner.
And again, what are we even arguing about, here? I already noted that I agree that capital investment is shifting, and is likely impacting jobs.
So perhaps what I’m “arguing” is that this change is inevitable, is not related to “shortages” (that aren’t even defined in the first place).
Capital investment flows to places where people believe it can provide the highest return.
As technology progresses, there will be less people working (and more machines doing the work). That’s been occurring since the Industrial revolution (and perhaps since the invention of the wheel).
I don’t know what that means, when we’re all sitting around eating bon-bons all day, scrolling blogs and social media (and taking GLP-1 medications to counteract that). I suspect that it means we’ll all become useless, at some point.
When even software developers are losing their jobs to artificial intelligence . . .
Perhaps the only “safe” fields to go into right now are blue collar jobs (e.g., house repair, etc.).
Ron O
You’re missing the point of the article: WHY has capital intensive technology innovation become so dominant? The answer from this article is because a key input–land and structures–has become more expensive relative to others and because labor and land are complementary production inputs (unlike capital which is a substitute for labor), capital has become relatively less expensive. That’s what the equation that David pasted in says. (I could post my equally complex set of equations from my doctoral dissertation, but I don’t know how to post images in comments here in the new format.) BTW, the authors use a Cobb-Douglas production function which is a fairly common modeling framework.
Your critique focuses only what has changed since 2020–the article highlights the trend change since 1970, so your points are not valid. Here’s a selection of other articles, including by economists at the Federal Reserve Bank, that come to the same conclusion that housing and building productivity has lagged the rest of the economy.
https://www.richmondfed.org/publications/research/economic_brief/2025/eb_25-31
https://www.nber.org/papers/w30845
https://www.nber.org/digest/202502/stagnation-us-construction-productivity?page=1&perPage=50
https://www.sciencedirect.com/science/article/abs/pii/S0166046225000249
The obvious conclusion is that housing prices have risen in part due to this lag in productivity and that has created the housing availability crisis that we face.
Not sure what you mean by “But you already know that conclusions are not the same thing as data”. The cited article uses data to arrive at its conclusions. So do the other articles that I’ve referenced. When presented with facts that contradict your stated position, you consistently fail to yield to the reality that has been presented.
Here’s the link to the article:
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4927970
Abstract. In this paper, we present a model showing that it is not the overall quantity but the composition of capital that has contributed to the decline in the labor share. Constructing annual macro data for 16 advanced countries over two centuries, we show that, since 1980, increasing real prices of buildings and the associated relative decline in buildings capital have reduced the labor share because buildings and labor are complements. The decline in the labor share has been reinforced by declining real prices of machinery capital and the associated relative increase in machinery capital because machinery capital and labor are substitutes. We show that these concurrent developments can explain a significant proportion of the evolution of the labor share.