Since house is the only (consumer) goods in the model, house price would become the numeraire. Indeed wage as defined would equilibrate between city A and B (assume no transaction cost for a min) since labor is allowed to move. Do you agree with this?
To say house is more expensive in city B post regulation, is to assume the existence of at least some other goods, beside house and labor. Then house price can be measured by a basket of goods.
I think I need you to clarify what you're saying. I don't know if this relates to your point, but once City B stops allowing new housing, housing in City B is really a bundle that consists of shelter + admittance to City B.
I don’t think your thought experiment fits with spatial equilibrium.
Consider that in the “before restrictions” scenario that household could move freely between City A and B but only do it when the housing price is identical. If the price of City A increased for any reason, people would move to City B until such time as relative price levels returned to the equilibrium. The same must be true in the “after restrictions” case. Why are people suddenly willing to pay more to live in the restricted City rather than move?
It’s not really a thought experiment as much as it is an interpretation of the data. I think it’s probably helpful to think of it in terms of sorting according to idiosyncratic housing demand elasticity. The most expensive American cities have shed millions of resident over the past 20-30 years. In general, the only ones left are the ones that didn’t choose to be displaced as costs rose. I think it’s similar to what happens in economically dying regions where residents with the deepest local ties stay in spite of low prospects. In both cases, they accept lower incomes in exchange for retaining the endowments of place. In the housing deprived cities, the gross incomes aren’t necessarily low, but incomes after housing costs are.
If I I understand the essay correctly, it boils down to the idea that a city’s housing restrictions will, over time, drive poorer people out, which means that the average and median incomes will rise even if those who remain in the city don’t earn any more money. The increased average incomes implies an increase in productivity even if, again, no productivity increases were made.
In addition, the population as a whole - that is, the people who remain plus those who are forced out by higher home prices - is made worse off.
This, in part, explains why those who tout progressive policies can point to statistics of blue cities and states to “prove” that their policies work. If you drive enough poor people away (in the name of helping the poor, of course), your averages are going to rise even if nothing gets better and maybe even if things get worse.
So, for example, I can make my school district’s average test scores better by driving out poorer families, which is far easier than hiring better teachers, enforcing discipline, and using proven teaching methods.
You lost me about halfway through the article - beginning with the discussion of Figures 4 and 5. Based on the labeling, I interpret Figure 4 to show that inflation, excluding housing, has been negative since 1960. This seems like a very startling claim, so I guess I'm misinterpreting the graph. Can you explain what each line represents, and how it was calculated? It looks like each line is a graph of some value over time, with 1960 set to 100. But what are those values, and where does the data come from. And how are the values related to your explanatory paragraph?
WRT Figure 5, you are saying that rent has increased as a percent of GDP. Is this only residential rent, or does it include commercial rent? Does it include imputed value of resident-owned housing? Does it control for changes in homeownership over the years? You imply that it shows consumers buying less house for higher prices, but house size has been increasing significantly over this time. Do you account for this?
You posit a very simple model that can explain some of the economic trends in the US. It's not obviously stronger or weaker than other models, but you don't offer a lot of support for it. Maybe that was your intent - merely to show that very different assumptions can lead to similar macro numbers, casting doubt on all the models.
Red: (cpi without shelter)/(cpi including shelter)
Black: (real personal consumption expenditures on housing)/(aggregate personal consumption expenditures)
In Figure 5, rent refers to the rental value of all occupied residential real estate, as estimated in pce statistics.
One way to think about consumption trends is that, for the past 40 years, for each 1% increase in real incomes, there has been less than a 1% increase in real housing consumption (real gross rental value), but for each 1% increase in nominal incomes there has been more than a 1% increase in nominal housing consumption (nominal gross rental value).
The “research” tab at the top of the substack page is probably the best introduction to some of the background.
Very nice writeup - thanks for taking the time. This mirrors what I've found in many social science studies and writings - you can find (and interpret) data that will go along with just about any "just so" story you like. Most academics seem to have a worldview that takes it as axiomatic that:
* Greater formal education leads to higher incomes
* Smart people with high degrees are capable of designing optimal policies that will overcome all the bad effects witnessed in the world
* Greater formal education leads to understanding and approving these more optimal policies
* Their own views on essentially all subjects are objectively superior to their opposites
They have no trouble creating mathematical models to support these assumptions, and studies that confirm the models. When they see other academics create such models, they take this as (further) confirmation that their worldview is correct. The academics, and the journalists who report the studies, generally don't even consider that the variables they consider may merely be proxies for genuine causes, or fundamentally irrelevant. Or that the causation may go in the opposite direction from the models.
"I am not the person to settle the discussion on the construction of the PhD level models. But, that isn’t going to keep me off my high horse."--KE
As a fellow internet-warrior, I salute you!
It sure looks like high housing costs are putting major dents in the US standard of living.
Interesting true story: Office space rents in downtown Los Angeles have barely budged in 40 years. Even in soggy markets, there was always institutional capital to build new towers, and always city say-so. Every tower builder planned in stealing tenants from other towers, locally or globally.
Thanks for diving into this Kevin. I had never heard the term "spherical cow" before, but I don't get out much these days. Agglomeration studies probably yield better data when they focus on more discrete examples, like a cluster of hotels near a convention center or an industrial park with a half dozen plastics manufacturing companies. Comparisons of cities can be brutally complex once you factor in long timelines that may include wars, cultural changes, environmental changes, etc... I happen to prefer your model because it aligns with our current reality of bad housing policy that has created a negative agglomeration effect across an increasing number of metro regions.
Fundamental thing with the model:
Since house is the only (consumer) goods in the model, house price would become the numeraire. Indeed wage as defined would equilibrate between city A and B (assume no transaction cost for a min) since labor is allowed to move. Do you agree with this?
To say house is more expensive in city B post regulation, is to assume the existence of at least some other goods, beside house and labor. Then house price can be measured by a basket of goods.
I think I need you to clarify what you're saying. I don't know if this relates to your point, but once City B stops allowing new housing, housing in City B is really a bundle that consists of shelter + admittance to City B.
I don’t think your thought experiment fits with spatial equilibrium.
Consider that in the “before restrictions” scenario that household could move freely between City A and B but only do it when the housing price is identical. If the price of City A increased for any reason, people would move to City B until such time as relative price levels returned to the equilibrium. The same must be true in the “after restrictions” case. Why are people suddenly willing to pay more to live in the restricted City rather than move?
It’s not really a thought experiment as much as it is an interpretation of the data. I think it’s probably helpful to think of it in terms of sorting according to idiosyncratic housing demand elasticity. The most expensive American cities have shed millions of resident over the past 20-30 years. In general, the only ones left are the ones that didn’t choose to be displaced as costs rose. I think it’s similar to what happens in economically dying regions where residents with the deepest local ties stay in spite of low prospects. In both cases, they accept lower incomes in exchange for retaining the endowments of place. In the housing deprived cities, the gross incomes aren’t necessarily low, but incomes after housing costs are.
If I I understand the essay correctly, it boils down to the idea that a city’s housing restrictions will, over time, drive poorer people out, which means that the average and median incomes will rise even if those who remain in the city don’t earn any more money. The increased average incomes implies an increase in productivity even if, again, no productivity increases were made.
In addition, the population as a whole - that is, the people who remain plus those who are forced out by higher home prices - is made worse off.
This, in part, explains why those who tout progressive policies can point to statistics of blue cities and states to “prove” that their policies work. If you drive enough poor people away (in the name of helping the poor, of course), your averages are going to rise even if nothing gets better and maybe even if things get worse.
So, for example, I can make my school district’s average test scores better by driving out poorer families, which is far easier than hiring better teachers, enforcing discipline, and using proven teaching methods.
You lost me about halfway through the article - beginning with the discussion of Figures 4 and 5. Based on the labeling, I interpret Figure 4 to show that inflation, excluding housing, has been negative since 1960. This seems like a very startling claim, so I guess I'm misinterpreting the graph. Can you explain what each line represents, and how it was calculated? It looks like each line is a graph of some value over time, with 1960 set to 100. But what are those values, and where does the data come from. And how are the values related to your explanatory paragraph?
WRT Figure 5, you are saying that rent has increased as a percent of GDP. Is this only residential rent, or does it include commercial rent? Does it include imputed value of resident-owned housing? Does it control for changes in homeownership over the years? You imply that it shows consumers buying less house for higher prices, but house size has been increasing significantly over this time. Do you account for this?
You posit a very simple model that can explain some of the economic trends in the US. It's not obviously stronger or weaker than other models, but you don't offer a lot of support for it. Maybe that was your intent - merely to show that very different assumptions can lead to similar macro numbers, casting doubt on all the models.
Thanks for the questions!
Yes. Figs 4-5 are set to 100 at the start.
The measures in Figure 4 are
Blue: (cpi shelter)/(aggregate cpi)
Red: (cpi without shelter)/(cpi including shelter)
Black: (real personal consumption expenditures on housing)/(aggregate personal consumption expenditures)
In Figure 5, rent refers to the rental value of all occupied residential real estate, as estimated in pce statistics.
One way to think about consumption trends is that, for the past 40 years, for each 1% increase in real incomes, there has been less than a 1% increase in real housing consumption (real gross rental value), but for each 1% increase in nominal incomes there has been more than a 1% increase in nominal housing consumption (nominal gross rental value).
The “research” tab at the top of the substack page is probably the best introduction to some of the background.
Very nice writeup - thanks for taking the time. This mirrors what I've found in many social science studies and writings - you can find (and interpret) data that will go along with just about any "just so" story you like. Most academics seem to have a worldview that takes it as axiomatic that:
* Greater formal education leads to higher incomes
* Smart people with high degrees are capable of designing optimal policies that will overcome all the bad effects witnessed in the world
* Greater formal education leads to understanding and approving these more optimal policies
* Their own views on essentially all subjects are objectively superior to their opposites
They have no trouble creating mathematical models to support these assumptions, and studies that confirm the models. When they see other academics create such models, they take this as (further) confirmation that their worldview is correct. The academics, and the journalists who report the studies, generally don't even consider that the variables they consider may merely be proxies for genuine causes, or fundamentally irrelevant. Or that the causation may go in the opposite direction from the models.
"I am not the person to settle the discussion on the construction of the PhD level models. But, that isn’t going to keep me off my high horse."--KE
As a fellow internet-warrior, I salute you!
It sure looks like high housing costs are putting major dents in the US standard of living.
Interesting true story: Office space rents in downtown Los Angeles have barely budged in 40 years. Even in soggy markets, there was always institutional capital to build new towers, and always city say-so. Every tower builder planned in stealing tenants from other towers, locally or globally.
That could have been the story on housing too.
Thanks for diving into this Kevin. I had never heard the term "spherical cow" before, but I don't get out much these days. Agglomeration studies probably yield better data when they focus on more discrete examples, like a cluster of hotels near a convention center or an industrial park with a half dozen plastics manufacturing companies. Comparisons of cities can be brutally complex once you factor in long timelines that may include wars, cultural changes, environmental changes, etc... I happen to prefer your model because it aligns with our current reality of bad housing policy that has created a negative agglomeration effect across an increasing number of metro regions.