Scott Alexander posted this last month, arguing that building more in an area doesn't necessarily decrease local prices because it can cause more people to move in.
I see 3 levels of argument here, based on 1st, 2nd, and 3rd-order effects:
- More building means more supply, which by supply-demand curves means prices decrease. (YIMBY is here)
- Local demand isn't fixed: prices are determined by an equilibrium between people coming and leaving. Building more changes the character of the area (which probably makes it suit current residents less) but the equilibrium of quality vs cost is unchanged. (Scott is here)
- Building more in expensive areas causes people to move which causes jobs to move, which causes more housing to be unused in low-cost areas, which means housing costs don't decrease significantly on a national level either. (I'm here)
Position (3) has long been my view, and I thought I'd write a post explaining it a bit.
The problem in America isn't a lack of aggregate housing, it's houses not being where people want to live. Building more in those areas might seem like an obvious solution, but it wrongly assumes "where people want to live" is constant. If we instead consider that it could change, we need to consider what makes people move from an area with cheap housing to an area where it's expensive.
You can just ask people that, and they'll tell you: jobs. OK, so why are there better jobs in the expensive areas?
If you ask an economist, they'll probably say something about improved economic efficiency from people being closer together. I don't buy it, because it contradicts what I observe directly:
- Many people can work as well remotely.
- Programmers don't become more productive when they move to Silicon Valley or Seattle or NYC.
- Factories in America aren't built in the middle of cities. Boeing doesn't make aircraft in the middle of NYC, that would be stupid.
Yes, people in NYC and Silicon Valley get paid more on average, but I reject the assertion that wages reflect productivity or competence. If wages were high simply because aggregation improved efficiency, then we'd see a stronger correlation between density and wages, but the correlation of wages with housing prices and with wealth is stronger. In my experience and the experience of people I know, the dominant factor is how close you are to money. Wages in NYC are high because there are wealthy investors and corporate executives there.
Why, then, are those rich people in NYC? Because:
- It has luxury services. (Amazon/etc made this less important.)
- Other rich people are there, which is good for networking.
- The nice parts are expensive enough to keep poor people out. (More applicable in expensive suburban areas than in NYC, where exclusion involves more private schools and secure apartments.)
In this model, building more housing in Silicon Valley or NYC just leads to more competition for about the same total income from "good jobs", with that number determined by the amount of money there. That makes things better for rich people buying luxury services in expensive cities, and worse for other people. If you make housing cheap enough for many poor people to move in, the rich people might even leave because of that, decreasing the number of good jobs.
The solution, then, is not to build more in high-demand areas - it's to force demand to be spread out more.
There are inaccuracies in the article, period. It would be embarassing for an engineer to make the mistake of conflating Celsius and Kelvin when comparing boiling point ratios, as in the claim that sodium's boiling point is 8x higher than that of water. Bill Gates' audience is going to have a number of technically savvy people in it, he knows it, and this alone is a college freshman/high school-level mistake. There are others.
My update on reading the article is, in fact, to downgrade my perception of Bill Gates' technical expertise beyond the world of computer software and hardware, and to trust his ability to communicate science less.
That said, nobody needs to be an expert in every subject, and it might be that Gates' wealth and diverse interests and fame simply put him in a position to try and interpret areas of science he's not able to understand adequately. He's unusual for a billionaire founder/CEO figure, and I personally wouldn't update too much on his mistake here as evidence about the ability of other CEOs to understand their company's specific technology to a level of depth adequate to run the business well. But I would put some probability mass into "CEOs are, in general, shockingly bad at understanding the technologies and products their company sells and they also don't have the ability to tell who in their company does understand what their company is selling."