house prices

=construction =housing

 

 

Making a single-family home in the US took about 3500 man-hours in the 1960s, including production of materials. Let's assume that's still accurate. You might argue that better power tools and components should have reduced that, but construction labor productivity has declined slightly over the past few decades.

The average US employee cost $40.35/hour in 2021, including benefits. Multiply that out, and it's $141,225.

In 2019, land was 37% of the value of houses. Now we're up to $224,166 for a house.

The median US home sale price was $347,000 in 2021, a ratio of 1.55x the calculated labor value * land multiplier. Again, our calculated value uses the labor going into materials, while the "labor cost percent" of houses doesn't. Also, construction labor is cheaper than average - despite being generally less pleasant than the average job in the US - but let's consider a hypothetical world where everyone is paid the same per hour.

So, we have a 0.55x labor value going somewhere - presumably mostly to the owners of capital, such as construction equipment and factories. They're getting less than the workers in this calculation, but again, construction workers are cheaper than the average worker. There are also some fees going to government for permits and inspections.

The average US home construction cost is now about $248,000 (not including land) which with our average land multiplier is $393,000 - a bit more than the average home sale price, which makes sense.

 

 

A lot of pundits have post-scarcity lives, where if something they want isn't happening, it must be because of regulation, not because it's too expensive on a societal level. They then apply this internalized principle to everything. For example, if nuclear power isn't happening, in their minds, it must be because of regulation rather than costs.

Construction costs across the US vary widely. Why? They correlate fairly well with average wages, as do land prices. This is a difference in PPP GDP relative to nominal GDP, and if you look at US states by PPP GDP per capita rather than nominal, their ranking is completely different.

Differences between nominal and PPP GDP are driven by cross-border capital flows, not by internal economics. For nominal GDP, only things that cross borders matter: raw materials, transportable food, portable manufactured goods, software, investment, and rich people bringing their money with them. Some areas are better at exporting manufactured goods or attracting rich people to live there, but that has little to do with productivity when making houses.

The reason why Americans can't afford housing today is simple: it takes just as much work and capital to make a house as it used to, but more Americans are poor relative to the average. If somebody makes the average American wage, they can afford a house at current prices, but $40/hr (including benefits) is several times what many people are making.

"Why hasn't productivity of X matched the average for the economy" is a silly question; productivity increases in X require improvements relevant to it, and there's no reason to expect them to come equally for different things. But we might expect at least some improvement, which hasn't happened. Why is that? The answer is, the average US construction worker is now less skilled than several decades ago. The people who would have been skilled construction workers are often now doing something that pays better or are trying to get out of construction work instead of trying to get really good at it.

Yes, land cost is also relevant, but cheaper greenfield construction also causes people to move to areas with cheaper land.

 

 

So, the solution to the US housing affordability crisis isn't "build more". The only solutions are to:

A) pay Americans more equally
B) make construction more efficient with technological and process improvements
C) do construction without involving the highly-paid kinds of Americans

 

 

 

 

back to index