Although I pretty much agree on everything here and think this is an abhorrently underrated insight, just want to point something out RE greenfield:
It’s completely possible for you to be correct about greenfield *supply* being on a flat part of the hysteresis curve, and for me to be correct about greenfield *demand* having hit a relatively vertical part of the hysteresis curve due to BOTH the mortgage crackdown AND underlying antipreferences for the suburban development style.
Basically, Millennials grew up in stroad-infested hellholes and don’t want to live in either the unaffordable ones they grew up in nor their exurban apotheoses; the mortgage crackdown simply ensured they didn’t have an attractively cheap escape hatch to just default to exurban greenfield *anyways*.
I don't think I ever said your hypothesis was impossible. I said that it fails to meet the facts.
edit: I should be more clear. I think the thing you are talking about exists. I think there are some city center residential developments that reflect new demand from younger adults to live in the city center and to have less dependence on cars. It just has very little to do with the macro level price and quantity trends I track here. Rent inflation in the exurbs has been very high.
I enjoyed the article! You mentioned how these complications make it possible to publish papers flawed reasoning, were you thinking, for example, of the recent GCPI paper (McCrae et al.)?
I haven't taken the time to look at that paper closely, but it definitely fits the profile. They found a lot of interesting and useful empirical evidence and they don't know how to interpret it. And, unfortunately, even economists with their hearts in the right place have huge blind spots that make it hard to address the mistakes head on.
It's wild that they just aren't aware of the mortgage crackdown, and it may be the case that nobody within their field of vision will address that.
All that said, though, other than their dismissal of the ability of market rate housing to create filtering, I don't see policy suggestions that would generally be harmful. At least they weren't moved to make conditions worse.
>But, the way that plays out at the metropolitan area level is that rents inflate, and home prices inflate, motivating builders to pay more for land and materials. This pushes the flat part of the supply curve within every metropolitan area higher.
I don't follow. Is this a shift left in S, or a movement along the supply curve?
It's a shift up in the supply curve. We can complete x homes this year. But there is demand for x+y homes. If demand for the x+1 home is in Austin, the rental value of homes in Austin rises, and builders in Austin are willing to pay more for land and inputs in Austin to build it. The market settles out with higher land values in Austin and higher input costs in Austin and Cleveland. The flat part of the supply curve in every city rises until total quantity demanded is x.
I didn't say there was a shift left. I said there was a shift up. Also, there is a change in what developers are willing to produce at a given price because the vertical national supply curve leads them to bid up the price of inputs. Because of that, as local demand curves shift to the right, the flat part of the local supply curve moves up until local quantities move back to the left enough that total quantity is at the vertical right end of the national supply curve.
Unexpected shifts in population and migration have created some localized volatility in prices, but in general, for the past decade, differences in demand are responsible for differences in local quantities. Prices have risen relatively uniformly across the country.
I think it's mostly materials, because its the price of shared inputs that would fluctuate to redirect inputs to high demand locations, and labor isn't nearly as mobile as materials. Though, I think in the high demand locations, labor and land get bid up, also.
Although I pretty much agree on everything here and think this is an abhorrently underrated insight, just want to point something out RE greenfield:
It’s completely possible for you to be correct about greenfield *supply* being on a flat part of the hysteresis curve, and for me to be correct about greenfield *demand* having hit a relatively vertical part of the hysteresis curve due to BOTH the mortgage crackdown AND underlying antipreferences for the suburban development style.
Basically, Millennials grew up in stroad-infested hellholes and don’t want to live in either the unaffordable ones they grew up in nor their exurban apotheoses; the mortgage crackdown simply ensured they didn’t have an attractively cheap escape hatch to just default to exurban greenfield *anyways*.
I don't think I ever said your hypothesis was impossible. I said that it fails to meet the facts.
edit: I should be more clear. I think the thing you are talking about exists. I think there are some city center residential developments that reflect new demand from younger adults to live in the city center and to have less dependence on cars. It just has very little to do with the macro level price and quantity trends I track here. Rent inflation in the exurbs has been very high.
I enjoyed the article! You mentioned how these complications make it possible to publish papers flawed reasoning, were you thinking, for example, of the recent GCPI paper (McCrae et al.)?
I haven't taken the time to look at that paper closely, but it definitely fits the profile. They found a lot of interesting and useful empirical evidence and they don't know how to interpret it. And, unfortunately, even economists with their hearts in the right place have huge blind spots that make it hard to address the mistakes head on.
It's wild that they just aren't aware of the mortgage crackdown, and it may be the case that nobody within their field of vision will address that.
All that said, though, other than their dismissal of the ability of market rate housing to create filtering, I don't see policy suggestions that would generally be harmful. At least they weren't moved to make conditions worse.
>But, the way that plays out at the metropolitan area level is that rents inflate, and home prices inflate, motivating builders to pay more for land and materials. This pushes the flat part of the supply curve within every metropolitan area higher.
I don't follow. Is this a shift left in S, or a movement along the supply curve?
It's a shift up in the supply curve. We can complete x homes this year. But there is demand for x+y homes. If demand for the x+1 home is in Austin, the rental value of homes in Austin rises, and builders in Austin are willing to pay more for land and inputs in Austin to build it. The market settles out with higher land values in Austin and higher input costs in Austin and Cleveland. The flat part of the supply curve in every city rises until total quantity demanded is x.
Why is that a shift up/left in supply? There's no change in how much developers are willing to produce (at a given price).
I didn't say there was a shift left. I said there was a shift up. Also, there is a change in what developers are willing to produce at a given price because the vertical national supply curve leads them to bid up the price of inputs. Because of that, as local demand curves shift to the right, the flat part of the local supply curve moves up until local quantities move back to the left enough that total quantity is at the vertical right end of the national supply curve.
Unexpected shifts in population and migration have created some localized volatility in prices, but in general, for the past decade, differences in demand are responsible for differences in local quantities. Prices have risen relatively uniformly across the country.
Shifting up and shifting left are equivalent, no?
Okay, changing the price of inputs would do it. Are you thinking of materials or labor?
I think it's mostly materials, because its the price of shared inputs that would fluctuate to redirect inputs to high demand locations, and labor isn't nearly as mobile as materials. Though, I think in the high demand locations, labor and land get bid up, also.
Here's a new post with more discussion on the topic.
https://kevinerdmann.substack.com/p/follow-up-to-multiple-binding-constraints