Hacker News Reader

Comment by yjftsjthsd-h | Hacker News Reader

I'm open to believing you, but how? Why would building new homes increase housing costs anywhere?

It doesn't unless it is outpaced by the increase in demand for housing. This person makes it sound like they have a ton of brand new houses in Austin with nobody living in them because they're too expensive. I can tell you as a former Austinite, that couldn't be further from the truth.

Induced demand, the same mechanism by which building more and wider roads doesn’t always mean less traffic. There’s still only so much land and space in a given area, and even in high density areas there are more and less desirable spots (e.g. a penthouse suite vs a basement apartment next to the furnace). As more people move into an area, more competition for those desirable spots has an increased pressure on the prices for those spots, which in turn increases prices for other less lucrative spots because the whole market window is shifted.

that sounds vaguely plausible, but is there any evidence for it other than sounding vaguely plausible?

it seems like the other explanation there is simply that most cities aren't building more housing fast enough to meet demand. there's been a longstanding population shift to urban centres that's consistent enough across all major cities that i don't think it can be easily explained as induced demand from homebuilding. especially since most cities aren't building all that much new residential.

Well let’s talk about what it would mean for prices to go down. We’re in agreement that not building enough to meet demand will drive prices up. So why is that so? Well we can think about what demand is, it’s people already living in the area that want to move (perhaps because they’re renters looking to buy), it’s people outside the area looking to move into the area and it’s investors/rental landlords looking to buy because they anticipate an increased demand that they will be able to profit from by selling or renting to future demanders from one of the first two categories.

So in order for prices to go down, you need to have a 0 or negative expected future demand to eliminate the incentive for the investors. But if you have negative or 0 expected future demand, then you also have no reason to build new houses at all. No one is going to spend tens or hundreds of thousands of dollars to just sit and watch the building rot away. Either they expect to sell it or rent it after it’s built, and you can only do that by either having demand outpacing supply or anticipating demand in the future that is higher than the current demand. If we had a magical perfect equilibrium between demand and new supply, at best prices would remain steady/only increase with inflation. The only mechanism by which new building can reduce the costs of housing is to build more housing than both current and all anticipated future demand.

The article includes data showing that induced demand is not significant in the housing markets studied.

It's not inconceivable. There's obviously some agglomeration advantage. People are valuable to other people. For instance, I think if SF built sufficient new homes it would have a lot more young people and they'll make the place nicer and that will make it more valuable.

More people = more opportunities = more people attracted, and so on.

Reply on news.ycombinator.com