"Price/income ratios are high for one and only one reason" -- I assume you are exaggerating just to make a point there. It's transparently not for just one reason. Nor is to for just one reason to a first approximation.
Not exaggerating. The only places price/income patterns like that develop is where basic housing needs of a normally slowly growing population aren’t met and the high prices are associated purely with the self-selection of the resident population into those that are willing to move to avoid it and those that are willing to be poor to resist being displaced.
I mean, I suppose we could pump poisonous sulphur into the air over LA and you could then say that demand was the difference between cheap and expensive LA.
I don’t think I should need to explain myself on this when all of the famously expensive cities are declining in population and have deeply negative net domestic migration flows.
As Carl Sagan said, "Extraordinary claims require extraordinary evidence." So "I don't need to explain myself" won't suffice for such a strong statement. I didn't say anything about migration, I argued about your P/E ratio statement. There are very obviously multiple reasons for disparities in P/Es, just as there are multiple reasons for a variety of stock market P/Es.
You can read the papers under the research tab to get an idea of what my model is. But here’s one way I come at it.
In 2005, home prices spiked in New Orleans and population dropped by 25%. We can all agree that was 100% a supply issue.
LA population has dropped about 0.5% annually for about 6 years. I would say that any population growth rate less than about 0.5% positive growth is effectively a 100% supply issue because the city is forcing displacement by not accommodating normal births-deaths growth, and it leads to the price patterns I describe.
So I think LA is 1% below a level where new demand is necessary to create outmigration and price inflation.
There must be a number between 0.5% and -25% where we can agree that it’s 100% a supply issue. What’s that number, if it’s below -0.5%?
I agree with you that artificial supply barriers have contributed to high prices/rents and out-migration - my debate was around the statement that it's the "one and only" determinant of higher P/Es. Other major factors include differences in metro GDP and income per capita growth (which were materially higher on average in expensive metros and mathematically lead to higher P/Es), as well as geographic barriers that aren't found in many "cheaper" markets like Texas, but disproportionately found in expensive coastal markets.
By the way, it might be surprising, but if you look at factors contributing to the sharp slowdown in migration post-pandemic 2022+, a big one is the lack of Sunbelt, not Coastal, affordability. Renters are more mobile (esp. now, with lock-in) and metro rent-to-income ratios converged sharply from 2018-2022 across the entire income distribution, with many Sunbelt markets becoming less affordable relative to their metro-incomes than "expensive" markets.
Yes. I write about that a lot. The mortgage crackdown created a supply problem everywhere, and so the pattern in local housing costs that can be attributed to inadequate supply spread everywhere.
Also, the main reason that the expensive coastal cities have higher local incomes is the supply shortage. The outmigrants have lower than average incomes, and so the mass displacement raises the average income of the remainers.
I had a few more posts on this topic after this one. A couple of them address that.
I'll repeat my question. If a perennial population loss of -0.5% doesn't qualify as a "one and only" supply issue what rate of population loss would qualify? Can you give me a number?
Keep in mind, the mortgage crackdown caused new home construction in rust belt cities like Detroit and Cleveland to decline to levels similar to LA and NYC, and their local housing costs took on the same pattern. Regressive rent and price appreciation. Does Detroit have a demand problem?
Over the past decade, according to Case-Shiller, Detroit and LA homes have appreciated at about the same rate as the US on average. Detroit permitted fewer homes than average before 2008, but it permitted many more homes per capita than LA. Since 2008, both Detroit and LA have permitted about the same rate of new homes, below average.
Phoenix, on the other hand, permitted a lot more homes than average before 2008. You could say before 2008 the Phoenix price spike was entirely a demand problem (for some reason people will agree easily to that). Then, Phoenix permits declined sharply after 2008, but they are still above average. Even permitting at triple the rate of LA & Detroit, Phoenix prices have risen more than LA and Detroit.
I would say Phoenix today is partly a demand partly a supply issue.
Also, as with all the other points I'm pushing back against the geographic barriers point is plausible but wrong. Boston and New York City are only geographically constrained because they arbitrarily limit infill growth and have been surrounded by townships that have low limits on housing density.
Ah, so much to debate, I guess. On geographic barriers there has been good academic work, including the Saiz QJE paper. New York City is surrounded by water on all sides, while Dallas is uninterrupted plains for hundreds of miles. In one you have to go up to get more density in the other you definitely don't. It's not impossible to get more density in NYC (see "City of Yes"), but I still don't think it's correct to say it's "only" because of artificial limits. Those limits are endogenous policy responses to the challenges of high density - not always good ones, but also not completely arbitrary.
Regarding your question - there is no number that implies that supply is the "one and only reason" - in fact the worse the issue the more likely it is that there are multiple contributing factors, it works the other way...
Another recent factor to throw in with CA is that if your locked in at 3% and your property tax basis is twenty years old, then you live in a different LA then someone thinking of moving there or that doesn’t own a house yet.
If mortgage rates were 3% again we would see 2020-2021 level outflows.
1) another thing to consider is that immigration has kept demand up in places like LA even as natives have emptied out.
Network effects are particularly important for immigrants. Several people I work with don’t want to live there but that is where their ethnic community and entire extended family is.
And whatever LAs problems, it’s still better than living in the third world or whatever.
2) there are some industries that just are in CA and aren’t anywhere else. My nephew and his fiancé are in the Bay Area because that’s where their lab to do their work is, not because they want to live there.
They are changing careers specifically to get a job that can be done anywhere so they can move someplace where they can afford a house.
Yes. Both of these things are true of a lot of places, and those places approve a moderate amount of new housing and so none of it seems worthy of discussion.
I've only ever had two economics courses, 50 years ago, but I have been saying for a while, my neighborhood in Boston isn't being gentrified. It's the same worn, struggling place it's been for decades, it's just getting tremendously more expensive. Which I think is the gist of this post. And it explains a lot, thank you.
I wonder if the orthodox macroeconomics profession gets this yet.
What reads like general inflation is actually a continuously ratcheting up of house prices, due to choked off supply and continuously tightening supply.
The entire West Coast, not just L.A., has become continuously more expensive for decades.
George Selgin has commented that a shortage, such as crop bust, is not alleviated by tighter money. The commodity in question will cost more, and costs more to produce marginal supplies, and maybe around the margins you can cool off that particular commodity through tighter money, but likely only by suffocating the general economy.
In the case of housing, we are talking about a very large commodity and one that feeds into other costs, such as labor bills. I assume labor costs more in L.A., and the West Coast, for obvious reasons. Who can pay the rent?
Japan had (and still has) only marginal increases in housing costs, and had near deflation for decades, despite negative interest rates, and rates on 10-year JGBs held at 0% by BoJ QE. This only modestly changed with the C19 situation, but they seem to be returning to a low inflation environment.
Trying to fight inflation in the US, with a rising population and rigid housing supply...well, I am not sure monetary policy is the best tool.
I also think inter-regional variation is often lost in these conversations. As someone living in LA, I strongly suspect that the coast has such a high amenity value that people will continue to pay sky high prices to live there even when building restrictions are lifted.
But within the county, there are plenty of places with pretty mediocre amenities where houses cost a million dollars, and that is mostly supply restrictions. I live between a highway and Amazon warehouse, and houses are just under a million
This is my point. Nice neighborhoods are expensive everywhere, regardless of supply conditions. Amenities sell for what they are worth. The thing that makes cities with a shortage more expensive on average is that the houses with the least amenity value get more expensive. That’s what my model here tracks. Houses on the coast were expensive in 1985 when Los Angeles built plenty of houses.
Since we broke housing in the entire country in 2008, this has been the shift everywhere. In Phoenix, it's the 60 year old unimproved apartments in high crime areas where the rents have doubled since then.
I agree with Matt Yglesias, and I don't think he's just reasoning from a price. LA county has 10 million people, twice as many as any other county. So that's Q. That's a lot of people willing to pay extremely high prices to live in coastal California.
I also agree with your point about counterproductive restrictions on building, as I'm certain Yglesias does as well. With more freedom to build, LA's population would be even higher and prices would come down. That would be great. But I think the combination of LA's very high price and its very high quantity suggest a demand curve that has shifted far to the right, relative to other cities. It really is a popular place to live.
In the past 4 years for which the BEA has data, LA MSA has lost 3.3% of its population. NYC, San Francisco, Boston, and San Diego are all also negative.
Is there a rate of population loss below which I can question the high demand thesis?
I see a stock/flow distinction. I'd say the level of demand is very high, but slightly less high than 4 years ago--hence the declining population.
Also, hasn't the price risen sharply, and doesn't that lead to working families being replaced by childless families, reducing total population, but not reducing the number of occupied units? I recall you making that argument about the housing bubble years.
That’s part of it, but it probably explained more of the decline then than it does today, if only because the population decline has been steeper in the last few years than it was then.
So would you say there is more demand for living in Detroit than for living in San Diego? More demand for Philadelphia than San Francisco?
San Francisco is actually an interesting case and I’m not totally sure how to model it or rhetorically refer to it. It is losing population at about the same rate as LA. And the asymmetrical price trends still suggest the intra-metro compromises that inadequate supply creates. But prices have been dropping in San Francisco in a way that suggests a deep cyclical contraction - like high tier residents moving by choice are moving away in droves. It’s the worst I’ve seen in the 22 years of data that I track in any major metro area.
Again, I see that as a drop in demand, but not low demand. If demand goes from extremely high to very high, then it's a decrease. But it's still very high.
I think you need to look at total Q and P, not change in Q and change in P.
The concept I’ve been developing here is that when housing production gets low enough that the marginal price is set by leavers, it isn’t amenities that determine the marginal price, it’s the endowment value of the families being displaced. People develop endowment value everywhere. Every city can be as expensive as LA if they get new building low enough. Since 2008, even the rust belt has been moving that direction.
"Price/income ratios are high for one and only one reason" -- I assume you are exaggerating just to make a point there. It's transparently not for just one reason. Nor is to for just one reason to a first approximation.
Not exaggerating. The only places price/income patterns like that develop is where basic housing needs of a normally slowly growing population aren’t met and the high prices are associated purely with the self-selection of the resident population into those that are willing to move to avoid it and those that are willing to be poor to resist being displaced.
I mean, I suppose we could pump poisonous sulphur into the air over LA and you could then say that demand was the difference between cheap and expensive LA.
I don’t think I should need to explain myself on this when all of the famously expensive cities are declining in population and have deeply negative net domestic migration flows.
As Carl Sagan said, "Extraordinary claims require extraordinary evidence." So "I don't need to explain myself" won't suffice for such a strong statement. I didn't say anything about migration, I argued about your P/E ratio statement. There are very obviously multiple reasons for disparities in P/Es, just as there are multiple reasons for a variety of stock market P/Es.
You can read the papers under the research tab to get an idea of what my model is. But here’s one way I come at it.
In 2005, home prices spiked in New Orleans and population dropped by 25%. We can all agree that was 100% a supply issue.
LA population has dropped about 0.5% annually for about 6 years. I would say that any population growth rate less than about 0.5% positive growth is effectively a 100% supply issue because the city is forcing displacement by not accommodating normal births-deaths growth, and it leads to the price patterns I describe.
So I think LA is 1% below a level where new demand is necessary to create outmigration and price inflation.
There must be a number between 0.5% and -25% where we can agree that it’s 100% a supply issue. What’s that number, if it’s below -0.5%?
I agree with you that artificial supply barriers have contributed to high prices/rents and out-migration - my debate was around the statement that it's the "one and only" determinant of higher P/Es. Other major factors include differences in metro GDP and income per capita growth (which were materially higher on average in expensive metros and mathematically lead to higher P/Es), as well as geographic barriers that aren't found in many "cheaper" markets like Texas, but disproportionately found in expensive coastal markets.
By the way, it might be surprising, but if you look at factors contributing to the sharp slowdown in migration post-pandemic 2022+, a big one is the lack of Sunbelt, not Coastal, affordability. Renters are more mobile (esp. now, with lock-in) and metro rent-to-income ratios converged sharply from 2018-2022 across the entire income distribution, with many Sunbelt markets becoming less affordable relative to their metro-incomes than "expensive" markets.
Yes. I write about that a lot. The mortgage crackdown created a supply problem everywhere, and so the pattern in local housing costs that can be attributed to inadequate supply spread everywhere.
Also, the main reason that the expensive coastal cities have higher local incomes is the supply shortage. The outmigrants have lower than average incomes, and so the mass displacement raises the average income of the remainers.
I had a few more posts on this topic after this one. A couple of them address that.
Here: https://kevinerdmann.substack.com/p/expensive-cities-arent-attractive-6b9
and here: https://kevinerdmann.substack.com/p/expensive-cities-arent-attractive-784
I'll repeat my question. If a perennial population loss of -0.5% doesn't qualify as a "one and only" supply issue what rate of population loss would qualify? Can you give me a number?
Keep in mind, the mortgage crackdown caused new home construction in rust belt cities like Detroit and Cleveland to decline to levels similar to LA and NYC, and their local housing costs took on the same pattern. Regressive rent and price appreciation. Does Detroit have a demand problem?
Over the past decade, according to Case-Shiller, Detroit and LA homes have appreciated at about the same rate as the US on average. Detroit permitted fewer homes than average before 2008, but it permitted many more homes per capita than LA. Since 2008, both Detroit and LA have permitted about the same rate of new homes, below average.
Phoenix, on the other hand, permitted a lot more homes than average before 2008. You could say before 2008 the Phoenix price spike was entirely a demand problem (for some reason people will agree easily to that). Then, Phoenix permits declined sharply after 2008, but they are still above average. Even permitting at triple the rate of LA & Detroit, Phoenix prices have risen more than LA and Detroit.
I would say Phoenix today is partly a demand partly a supply issue.
https://fred.stlouisfed.org/graph/?g=1CnPu
Also, as with all the other points I'm pushing back against the geographic barriers point is plausible but wrong. Boston and New York City are only geographically constrained because they arbitrarily limit infill growth and have been surrounded by townships that have low limits on housing density.
Ah, so much to debate, I guess. On geographic barriers there has been good academic work, including the Saiz QJE paper. New York City is surrounded by water on all sides, while Dallas is uninterrupted plains for hundreds of miles. In one you have to go up to get more density in the other you definitely don't. It's not impossible to get more density in NYC (see "City of Yes"), but I still don't think it's correct to say it's "only" because of artificial limits. Those limits are endogenous policy responses to the challenges of high density - not always good ones, but also not completely arbitrary.
Regarding your question - there is no number that implies that supply is the "one and only reason" - in fact the worse the issue the more likely it is that there are multiple contributing factors, it works the other way...
Another recent factor to throw in with CA is that if your locked in at 3% and your property tax basis is twenty years old, then you live in a different LA then someone thinking of moving there or that doesn’t own a house yet.
If mortgage rates were 3% again we would see 2020-2021 level outflows.
1) another thing to consider is that immigration has kept demand up in places like LA even as natives have emptied out.
Network effects are particularly important for immigrants. Several people I work with don’t want to live there but that is where their ethnic community and entire extended family is.
And whatever LAs problems, it’s still better than living in the third world or whatever.
2) there are some industries that just are in CA and aren’t anywhere else. My nephew and his fiancé are in the Bay Area because that’s where their lab to do their work is, not because they want to live there.
They are changing careers specifically to get a job that can be done anywhere so they can move someplace where they can afford a house.
Yes. Both of these things are true of a lot of places, and those places approve a moderate amount of new housing and so none of it seems worthy of discussion.
I've only ever had two economics courses, 50 years ago, but I have been saying for a while, my neighborhood in Boston isn't being gentrified. It's the same worn, struggling place it's been for decades, it's just getting tremendously more expensive. Which I think is the gist of this post. And it explains a lot, thank you.
I wonder if the orthodox macroeconomics profession gets this yet.
What reads like general inflation is actually a continuously ratcheting up of house prices, due to choked off supply and continuously tightening supply.
The entire West Coast, not just L.A., has become continuously more expensive for decades.
George Selgin has commented that a shortage, such as crop bust, is not alleviated by tighter money. The commodity in question will cost more, and costs more to produce marginal supplies, and maybe around the margins you can cool off that particular commodity through tighter money, but likely only by suffocating the general economy.
In the case of housing, we are talking about a very large commodity and one that feeds into other costs, such as labor bills. I assume labor costs more in L.A., and the West Coast, for obvious reasons. Who can pay the rent?
Japan had (and still has) only marginal increases in housing costs, and had near deflation for decades, despite negative interest rates, and rates on 10-year JGBs held at 0% by BoJ QE. This only modestly changed with the C19 situation, but they seem to be returning to a low inflation environment.
Trying to fight inflation in the US, with a rising population and rigid housing supply...well, I am not sure monetary policy is the best tool.
I also think inter-regional variation is often lost in these conversations. As someone living in LA, I strongly suspect that the coast has such a high amenity value that people will continue to pay sky high prices to live there even when building restrictions are lifted.
But within the county, there are plenty of places with pretty mediocre amenities where houses cost a million dollars, and that is mostly supply restrictions. I live between a highway and Amazon warehouse, and houses are just under a million
This is my point. Nice neighborhoods are expensive everywhere, regardless of supply conditions. Amenities sell for what they are worth. The thing that makes cities with a shortage more expensive on average is that the houses with the least amenity value get more expensive. That’s what my model here tracks. Houses on the coast were expensive in 1985 when Los Angeles built plenty of houses.
What's interesting about southern California too is that there are almost no cheap exburbs left. Here is a 1300ft home in Coachella 3 hours from the coast, and it's still over 400k https://www.zillow.com/homedetails/52335-Michelle-Dr-Coachella-CA-92236/95863273_zpid/?utm_source=txtshare
Since we broke housing in the entire country in 2008, this has been the shift everywhere. In Phoenix, it's the 60 year old unimproved apartments in high crime areas where the rents have doubled since then.
I agree with Matt Yglesias, and I don't think he's just reasoning from a price. LA county has 10 million people, twice as many as any other county. So that's Q. That's a lot of people willing to pay extremely high prices to live in coastal California.
I also agree with your point about counterproductive restrictions on building, as I'm certain Yglesias does as well. With more freedom to build, LA's population would be even higher and prices would come down. That would be great. But I think the combination of LA's very high price and its very high quantity suggest a demand curve that has shifted far to the right, relative to other cities. It really is a popular place to live.
In the past 4 years for which the BEA has data, LA MSA has lost 3.3% of its population. NYC, San Francisco, Boston, and San Diego are all also negative.
Is there a rate of population loss below which I can question the high demand thesis?
I see a stock/flow distinction. I'd say the level of demand is very high, but slightly less high than 4 years ago--hence the declining population.
Also, hasn't the price risen sharply, and doesn't that lead to working families being replaced by childless families, reducing total population, but not reducing the number of occupied units? I recall you making that argument about the housing bubble years.
That’s part of it, but it probably explained more of the decline then than it does today, if only because the population decline has been steeper in the last few years than it was then.
So would you say there is more demand for living in Detroit than for living in San Diego? More demand for Philadelphia than San Francisco?
San Francisco is actually an interesting case and I’m not totally sure how to model it or rhetorically refer to it. It is losing population at about the same rate as LA. And the asymmetrical price trends still suggest the intra-metro compromises that inadequate supply creates. But prices have been dropping in San Francisco in a way that suggests a deep cyclical contraction - like high tier residents moving by choice are moving away in droves. It’s the worst I’ve seen in the 22 years of data that I track in any major metro area.
Again, I see that as a drop in demand, but not low demand. If demand goes from extremely high to very high, then it's a decrease. But it's still very high.
I think you need to look at total Q and P, not change in Q and change in P.
The concept I’ve been developing here is that when housing production gets low enough that the marginal price is set by leavers, it isn’t amenities that determine the marginal price, it’s the endowment value of the families being displaced. People develop endowment value everywhere. Every city can be as expensive as LA if they get new building low enough. Since 2008, even the rust belt has been moving that direction.
"Every city can be as expensive as LA if they get new building low enough."
I disagree. I don't think many people wish to live in Detroit even if homebuilding falls to zero. Demand is low. Prices would remain low.