Zelman on housing

[JRM]

Many dual income households and falling interest rates from 1980 to today is certainly a nice tailwind.

 

Taxes and insurance will increase over time.  Maintenance expense will be required to keep up with replacement cost.  If enough maintenance is deferred then appreciation will entirely be based upon the increased desirability of the location.

 

The middle-class boomer retirement plan: always buy the biggest house you can afford and downsize when you retire.  

Edited October 27, 2021 by JRM

[KJP]

I have not read it, but I understand part of the thesis of this book is that second incomes and competition over housing in good schools pushed real house prices above inflation over the last several decades:  https://www.amazon.com/Two-Income-Trap-Middle-Class-Parents-Still/dp/0465097707

 

If the causal claim is correct, how are second incomes still causing rapid real housing price inflation when female labor force participation in the US peaked in 1990 and is down about 4 percentage points since then (60% - 56%)?  See:  https://fred.stlouisfed.org/series/LNS11300002

 

Has the essentially continuous fall in interest rates since 1990 done the work?  https://fred.stlouisfed.org/series/IRLTLT01USM156N

 

What I'm trying to get at here is whether the last 50 years or so have seen a series of presumably one-time events (e.g., female labor force participation doubling [can't double again] and nominal rates falling 600 bps from 1990 to today [tough to repeat with 10-year rate at 1.5%]) coming one after another, such that the change in real housing prices going forward will, on average, be zero, because these were just one-time changes that affected the level, but not the long-term first derivative of, prices.  Or has something fundamentally changed such at the trend of the change in average real housing prices going forward will be greater than zero? 

 

I also acknowledge that I'm using national aggregate data that may hide more important long-term trends, e.g.,  housing in NYC and similar places always rising faster than inflation but other areas rising slower such that averaging them isn't very illuminating.  But I couldn't find city-specific charts that were long enough (century+) to look at that issue, other than the Amsterdam chart, which isn't consistent with this theory.

[ERICOPOLY]

1 hour ago, KJP said:

If the causal claim is correct, how are second incomes still causing rapid real housing price inflation when female labor force participation in the US peaked in 1990 and is down about 4 percentage points since then

 

If you were to consider the gender pay gap you can still have tailwind effects as the gap closes over time. 

 

But then there are other factors:  Interest rates.  Credit availability.  Population growth.  New household formation rates. Building codes and land use restrictions.  And education levels:  I understand that software engineering pays better than picking lettuce.

 

A lot of factors.  

 

You could also measure the price of raw land when Manhattan was purchased versus today to see the effects of population density.  Or the Louisiana purchase and see how population density has effected land prices.  Etc...

Edited October 27, 2021 by ERICOPOLY

[ERICOPOLY]

Curious as to the pay of a software engineer in 1930.  I think that as the economy has evolved it has affected income levels.

[KJP]

I agree that all of the factors you mention could be relevant and overwhelm the two factors I mentioned. 

 

Overall, what surprises me isn't a rising trend in real housing prices over the last 50 or so years; rather, I'm surprised there apparently wasn't a similar trend over the prior 100 years.  Without looking at any historical data, I would think that as land where people actually want to live becomes scarce and as where you live becomes relevant to social status, schooling, and so on, people would at least spend a constant percentage of their real income on housing.  In that scenario, real housing costs would correlate with real GDP per capita (this ought to capture the software engineer vs lettuce picker example you mention).  And so long as GDP per capita has an upward slope, then I would thing that real housing prices ought to as well.  But that model doesn't appear to always be correct, because real GDP per capita has been going up for longer than real housing prices have been going up.  See:  https://fred.stlouisfed.org/series/A939RX0Q048SBEA/

 

Perhaps rather than asking about 1975 to today, I ought to ask about 1890-1975.  Based on the data here,  https://valuabl.substack.com/p/housing-market-part-8-a-very-long real housing prices in the US didn't change from 1890 - 1975.  But real GDP per capital went up multiples over that period.  Populations were also increasing throughout that period.  So, the US had rising population and rising real income per capita, but flat real housing prices.  Given the effects of increasing wealth and population that you posit, how did that happen? 

 

I don't know the answers to any of these questions.  It's just interesting that the majority view on this thread (housing prices rise faster than inflation) seems to have trouble with 1890-1975, while the minority view (view of one? -- housing prices rise only with inflation) seems to have trouble with 1975-2021.

 

 

 

 

 

 

Edited October 27, 2021 by KJP

[ERICOPOLY]

1 hour ago, KJP said:

people would at least spend a constant percentage of their real income on housing. 

 

I mean.. in a 2 income household the first income can spend the historical amount and still have the remainder of the income to cover food and diapers.  Same as always, right?  Let's call that 25% on housing.

 

Okay, that second income?  Well, the first income covered everything.  Let's say the second income can go 100% to housing in that case and they still have food and diapers?  Why not?

[KJP]

1 minute ago, ERICOPOLY said:

 

I mean.. in a 2 income household the first income can spend the historical amount and still have the remainder of the income to cover food and diapers.  Same as always, right?  Let's call that 25% on housing.

 

Okay, that second income?  Well, the first income covered everything.  Let's say the second income can go 100% to housing in that case and they still have food and diapers?  Why not?

 

Yes, your points (and the other factors you noted earlier) make logical sense to me.  As I mentioned, your points appear to be consistent with the data from 1975 - today. 

 

But the same points appear to be contradicted by the data from 1890-1975.  Women's labor force participation increased throughout that period as did GDP per capita (the former presumably related to the latter), yet real housing prices did not increase.  Why not? 

 

Earlier in the thread I asked GMthebeau to explain 1975-2021, which appear to strongly contradict his/her claim that housing prices only rise with inflation.  As far as I could tell, he/she has not taken a crack at that question.

 

Your theory (and I think the near unanimous view on this thread) is that GMthebeau is wrong and housing prices ought to rise faster than inflation, due to, among other things, rising per capita wealth, rising population, and women increasingly entering the workforce.  That theory could explain the time period that GMthebeau cannot (or at least has not yet).  But that theory appears to be inconsistent with 1890-1975, when those three things were also true (rising population, rising income per capita, and increasing female labor participation), that's is why I'm asking you and the others who believe that housing prices increase faster than inflation specifically about what happened from 1890-1975.

[ERICOPOLY]

9 minutes ago, KJP said:

Your theory (and I think the near unanimous view on this thread) is that GMthebeau is wrong and housing prices ought to rise faster than inflation, due to, among other things, rising per capita wealth, rising population, and women increasingly entering the workforce.  

 

I'm saying it's a bidding war and desirable areas are driven by increasing population density -- the rest of it is fuel.  Check out this picture:

 

https://www.etsy.com/listing/1012638388/ansel-adams-orchard-santa-clara?gpla=1&gao=1&&utm_source=google&utm_medium=cpc&utm_campaign=shopping_us_ps-a-home_and_living-home_decor-wall_decor-other&utm_custom1=_k_Cj0KCQjw8eOLBhC1ARIsAOzx5cGKUcih8Dwpj4AHSAZ3B1p73hd71ZaI7Y8_q5TPYPL7FDmTmB100P0aArGSEALw_wcB_k_&utm_content=go_12567673668_122422048791_507253757890_aud-459688891595:pla-307501513391_c__1012638388_498908644&utm_custom2=12567673668&gclid=Cj0KCQjw8eOLBhC1ARIsAOzx5cGKUcih8Dwpj4AHSAZ3B1p73hd71ZaI7Y8_q5TPYPL7FDmTmB100P0aArGSEALw_wcB

Edited October 27, 2021 by ERICOPOLY

[ERICOPOLY]

That picture is Silicon Valley in the 1950s.

 

The prices of lots exploded as land ran out.

 

There were still open fields all around my parents’ home when they purchased.  I can recall watching the houses being built as all of those fields disappeared.

 

My parents’ home is all land value.  Most would tear it down because it isn’t impressive enough.

[ERICOPOLY]

So really, talking about building costs is missing the picture.  My parent's home is priced assuming a teardown.  I've seen so many homes like theirs that were well maintained and just get torn down.  People who want a big custom home built seek out houses like this so they "don't have to overpay" for what they plan to tear down.  LOL.

[ERICOPOLY]

Another factor not yet mentioned is that Bay Area tech workers (and likely others) simply make more money.

 

Businesses will literally pay people more to compensate for the higher cost of living.  And that drives it higher still.

 

 

 

 

[KJP]

Between 1890 and 1975, the US population increased from 70 million to 215 million or so.  (https://usafacts.org/data/topics/people-society/population-and-demographics/population-data/population/)

 

That big population increase didn't lead to any increase in real housing prices.  So, if the driver is population density, that appears only to be a real constraint starting around 1975 or starting at around 215 million people, depending on how you want to look at it.  Why is that? 

 

From a microeconomic perspective, you have to be right that prices starting rising faster than inflation in certain areas because demand for those houses from people with the ability to pay sharply outstripped supply.  But from a macroeconomic perspective, it's hard to understand why 1975/215 million people was some kind of population density tipping point.  

 

I can imagine some theories why that might be some tipping point.  First, to take your example, your parents were willing to move to a place that was relatively undeveloped, presumably because there was opportunity for them there and it was affordable to live there.  I suspect that was historically true throughout 1890-1970.  Yes, there were a lot of people around Boston, New York, etc., but there was also a cheaper "frontier" (particularly out West -- Arizona didn't even become a state until 1912) that kept housing prices everywhere in check by reducing demand in expensive places (people left).  According to this theory, this "frontier" was used up by 1975, and thus could no longer keep housing prices in check.  But an obvious rejoinder to this is that the US has plenty of undeveloped land.  Why hasn't that land performed the safety valve function that California provided in 1950?  

 

Second, 1890-1975 included a transportation revolution -- horse to car to better car/bus to interstate highways to regional mass transit -- that actually created additional close substitutes to desirable urban land.  By 1975, that land was largely filled up, and we haven't had another transportation revolution that would make land further away from cities equivalent substitutes.

 

Third, the aggregate data are misleading me.  A real housing price index for Manhattan, San Francisco, etc. would show real housing prices rising faster than inflation throughout the whole period, with additional places starting to rise faster than inflation as they fill up.  But that's obscured by other places that are stagnating or losing people and where real housing prices are actually falling.  Thus, even thought the aggregate numbers show no real housing price increases from 1890-1975, that's less instructive than looking at places where population has been consistently rising.  I haven't found long enough real housing price indices for US cities to test this, but it is inconsistent with the Amsterdam data.

 

Fourth, the US went off the gold standard in the early 1970's.  Inflation measures after that point systematically understate inflation, so real housing prices from 1975-2021 haven't increased as much as they appear to have when using CPI to calculate real housing prices. 

 

Fifth, there was always upward pressure on real housing prices, but it was counterbalanced by increasing construction labor productivity.  Construction labor productivity has gone backwards from 1975, so it's now also pushing housing prices up, rather than counterbalancing other factors and holding them down.  [This overlaps with the regulation point you mentioned earlier.]

 

I can't tell which (if any) of these theories are the main drivers.

 

 

[ERICOPOLY]

1 minute ago, KJP said:

That big population increase didn't lead to any increase in real housing prices.  So, if the driver is population density

 

If you look around the US there is a lot of open space and lots are dramatically cheaper in areas where there is availability versus in areas where there is not.

 

How much does a lot cost in Santa Monica versus in Bakersfield?

 

Desirable areas where there is no available land and there is a lot of density leads to high lot prices.   

[ERICOPOLY]

It is supply and demand economics.  The US in the aggregate has a lot of supply.  But in desirable areas not so much.

[LearningMachine]

17 minutes ago, KJP said:

we haven't had another transportation revolution that would make land further away from cities equivalent substitutes.

 

Until recently.  However, it needed Covid to prove that it would work for highly paid knowledge workers.  Still need to wait for supply and inventory to catch up for it to show its real impact.

 

https://fred.stlouisfed.org/series/HOUST

https://fred.stlouisfed.org/series/MSACSR

Edited October 27, 2021 by LearningMachine

[ERICOPOLY]

Montecito real estate is a good one.  As the country exploded in population size and the ranks of the wealthy grew, the desirable areas face price competition.  You have wealthy people from Chicago owning second homes in Montecito to get away from the harsh climate.

Edited October 27, 2021 by ERICOPOLY

[LearningMachine]

3 hours ago, ERICOPOLY said:

software engineering pays better than picking lettuce

 

Well said.  Now, software engineering can pay better not only in certain areas with restricted supply, but anywhere with plenty of desirable land with water views, acreage, etc. 

Edited October 27, 2021 by LearningMachine

[Spekulatius]

I think the relentless downtrend in interest rates from 1980 until know has a lot of do with higher housing prices. If this reverses, I think prices will have to come down.

The second factor is land use  / building permit restrictions. In many US cities, the population density is quite low compared to Asian or European cities, partly due to zoning laws. If these cities grow in population, they need to spread out rather than grow in density.

Edited October 27, 2021 by Spekulatius

[LearningMachine]

23 hours ago, ERICOPOLY said:

 

I don't recall the LTV or income percentage.  I believe it was maximum loan they qualified for and I believe they had down payment help from family. 

 

 

 

Thanks, the reason I asked was to confirm that there is a limit to what percentage of their yearly income year-after-year people are willing to pay for something in restricted supply they want or need badly.

 

The uppermost limit is 100% of income.   Practically, it would have to be somewhat less. 

 

If mortgage payments double or triple, people can still only pay a certain percent of income.  10% increase in income due to inflation in first year wouldn't be able to afford doubling or tripling of mortgage payment for any buyers out there in such an environment, without a price drop.  I understand folks holding with 30-year mortgages don't have to sell, but some have to always sell for reasons beyond their control. 

 

One thing I haven't been able to fully explain is why did the nominal home prices didn't fall even though inflation-adjusted prices fell, when interest rates shot up in 1970s and early 80s.  Could it be that LTVs were low in general and housing wasn't really a financial asset equivalent dependent on interest rates like it is now?   

 

If LTV was equally high then, and percentage of income people were already paying was very high also, how did buyers find the money to be able to pay mortgage payments above a certain percent of their income when interest rates shot up?  Maybe lenders didn't restrict that debt service payments had to be below certain percent of income? 

 

https://www.multpl.com/case-shiller-home-price-index

https://www.multpl.com/case-shiller-home-price-index-inflation-adjusted

https://www.multpl.com/10-year-treasury-rate

 

Edited October 27, 2021 by LearningMachine

[Gregmal]

16 minutes ago, LearningMachine said:

 

Well said.  Now, software engineering can pay better not only in certain areas with restricted supply, but anywhere with plenty of desirable land with water views, acreage, etc. 

Tangentially my father grew up in various parts of NJ and PA. Settled most of his childhood in Ridgewood/Ho Ho Kus area. Similar to what Eric described, in this area, in the 60/70s...its was fields and open land. Yes, Ridgewood and especially Ho Ho Kus, and even myself, I recall Mahwah NJ being barren in the 90s. Todays its packed and its all expensive suburban homes and a lot of big mansions. Saddle River too, especially so. So having left the area and moved to Tampa a decade ago, my parents are asking me to assist them in finding ANYTHING on a lake with some land with an open mind to pretty much ANYTHING in NJ/NY/PA...and guess what Ive noticed? Basically ANYTHING with a good parcel, waterfront, is now $500k+...even remote stuff in upstate NY or Nowheresville, PA. Crazy.

[rohitc99]

Maybe its also that after a certain level of income, you will spend more on discretionary stuff like cars, 5$ lattes and better housing ? ofcourse does not explain why this happened only after 71. ofcourse cars dont have limited supply (unless its a ferrari), whereas premium locations have finite supply

[rohitc99]

There is also a level of keeping up with the joneses and status to it. How do you show that you are successful in life ? i have looked at housing as functional but realized that a lot of friends have bought into communities near to mine which are atleast 50% more expensive to show that they are successful. my neigbour moved out from a good 3000 sqft house to 6000 sqft after he became an empty nester. why would you need 5 bedrooms for a couple. he did it because his friends did the same and he could afford it

 

housing also is like a luxury good like a 5000$ hand bag or shoes

[LearningMachine]

14 minutes ago, Gregmal said:

Tangentially my father grew up in various parts of NJ and PA. Settled most of his childhood in Ridgewood/Ho Ho Kus area. Similar to what Eric described, in this area, in the 60/70s...its was fields and open land. Yes, Ridgewood and especially Ho Ho Kus, and even myself, I recall Mahwah NJ being barren in the 90s. Todays its packed and its all expensive suburban homes and a lot of big mansions. Saddle River too, especially so. So having left the area and moved to Tampa a decade ago, my parents are asking me to assist them in finding ANYTHING on a lake with some land with an open mind to pretty much ANYTHING in NJ/NY/PA...and guess what Ive noticed? Basically ANYTHING with a good parcel, waterfront, is now $500k+...even remote stuff in upstate NY or Nowheresville, PA. Crazy.

 

For some knowledge workers freed from geographical constraints, $500K waterfront is still cheaper and more desirable than $3-10M shack with no water views and issues created by density in some tech hubs. 

Edited October 27, 2021 by LearningMachine

[ERICOPOLY]

Obviously rates matter.  

 

500,000 loan at 6% for 30 yr term:    2999 per month payment

500,000 loan at 3% for 30 yr term:    2108 per month payment

 

[LearningMachine]

11 minutes ago, rohitc99 said:

There is also a level of keeping up with the joneses and status to it. How do you show that you are successful in life ? i have looked at housing as functional but realized that a lot of friends have bought into communities near to mine which are atleast 50% more expensive to show that they are successful. my neigbour moved out from a good 3000 sqft house to 6000 sqft after he became an empty nester. why would you need 5 bedrooms for a couple. he did it because his friends did the same and he could afford it

 

housing also is like a luxury good like a 5000$ hand bag or shoes

Another way to show status for geographically freed folks could be that $500K waterfront lot instead.  It is starting to happen.  You will have to give it some time for supply to catch up and for people to hear others flaunting new construction, water views, etc. for others to follow to move out.

Edited October 27, 2021 by LearningMachine