In fact, as a new Brookings study shows, millennials are not moving en masse to metros with dense big cities, but away from them. According to demographer Bill Frey, the 2013–2017 American Community Survey shows that New York now suffers the largest net annual outmigration of post-college millennials (ages 25–34) of any metro area—some 38,000 annually—followed by Los Angeles, Chicago, and San Diego. New York’s losses are 75 percent higher than during the previous five-year period.
. . .The top 20 magnets include Midwest locales such as Minneapolis–St. Paul, Columbus, and Kansas City, all areas where average house prices, adjusted for incomes, are half or less than those in California, and at least one-third less than in New York.
2. I now have a review copy of Shut Out. On p. 5, he writes,
we did not have a housing bubble. We had a housing supply bust–first in the places where people want to live, in places where there is more economic opportunity. That supply bust caused prices to rise to extreme levels in those cities–most notably in New York City, Los Angeles, Boston, and San Francisco–metropolitan areas I call the Closed Access cities. After the turn of the century, millions of households flooded out of those cities because of a shortage of housing–so many that they overwhelmed cities in the main destinations for those households, such as inland California, Arizona, and Florida. Then we imposed a credit and monetary bust on the entire country in a misplaced attempt to alleviate the problem.
I have a difficult time wrapping my head around the idea of a “supply bust.” It suggests a supply curve rapidly shifting to the left. I find it more plausible to think of housing supply in the Closed Access cities as relatively fixed, with demand increasing rapidly. But how to reconcile “demand increasing rapidly” with net out-migration without resorting to a Yogi Berra theory?
One possibility is non-resident foreign buyers making up the difference. Another possibility, going back to the supply bust idea, is that there is a lot of reconstruction taking place with the existing housing stock, and while these houses are being fixed up, no one is living in them.
Overall, there are a lot of cross-currents here, and there are multiple housing markets even within a single metropolitan area. It’s a difficult picture to sort out.
Is it possible that the supply bust refers more to expectations of future supply suddenly becoming constrained? If initial price points reflected expectations of steadily growing supply rather than increasing restrictions on what people could build, I could see how that would effectively cause a leftward supply shock.
But how to reconcile “demand increasing rapidly” with net out-migration without resorting to a Yogi Berra theory?
Yes, in California net migration of US citizens has been mostly true since 1991 but the population still increase due to Immigration. I bet the same thing in New York city. (The last ten years has legal immigration btw.) Also California birth rate was one of the higher states until 2008 as minority citizens starting to drop birth rates.
Again, in terms housing supply there a coastal reality:
1) There is less brown to more easily build on.
2) It is expensive to build a large housing/apartment as a lot of long term residents don’t want to sell. (Yes regulation has impact.)
The first reference on millennial migration is more interesting. Millennials also decided they didn’t want Obamacare taxes, and don’t like higher interest rates. That doesn’t sound like socialist revolution, sounds like a generation trying to avoid the debt service of their parents. Even that feeble attempt at promoting MMT was a disguised hint at default. They are escaping the cost of their parents’ unpaid bills.
Another possibility is that the net available housing stock in the Closed Access cities is actually declining over time due to existing units being kept vacant or consolidated with others even at the high rents on offer. This in turn may be a spiraling reaction to increasing regulatory clampdowns on landlords, who are constantly villianized as “greedy” and thus at fault for high housing prices, and who may not wish to keep renting units under ever tighter restrictions.
“But how to reconcile “demand increasing rapidly” with net out-migration without resorting to a Yogi Berra theory?”
FWIW, some of the confusion might come from differences in trends for different groups. In the first part, Kotkin is talking about millennials ages 25-34. There could be relatively larger numbers of residential demand outside that age group. But to add to the confusion, the Brookings study that he cites also shows large net out-migration of seniors from the allegedly desireable tech hub cities.
Personally, I doubt that all the opportunity and demand is actually there and other factors account for housing cost increases. The best explanation that I have seen is that the number of occupants per residence has declined. The number of single occupant tenants has rapidly increased as a share of residents. An other trend is the tend for residences, even in cities, to grow in average size. Apartments may cost more simply because they are on average larger. Builders and developers can no doubt target the high end of the purchaser market with more amenities etc in much the same way car dealers peddle add-ons. Twenty years ago the density of in-building exercise facilities may have been less than what it is now. I suspect young married couples flee as fast as their feet will carry them while singletons may fit into the urban habitat more effectively.
Moreover, while browsing the Brookings website, I came accross an interesting article that suggests that desire to live in non-urban areas may be rebounding and at least is not as universally undesired as some would have it.
https://www.brookings.edu/blog/the-avenue/2018/03/26/us-population-disperses-to-suburbs-exurbs-rural-areas-and-middle-of-the-country-metros/
One wonders how much of the demand for housing in densely urban habitats comes from residents of other such areas. Movements between cities probably generates more residential turnover than kids from the sticks looking to move to the big city for “opportunity.”
On a related note, the Mercer list of top 30 livable cities worldwide came out today with zero US cities represented. I doubt people are moving to US tech hubs for the public schools and pleasant company. I think if any of the big tech companies offering jobs in urban centers ever needed to actually begin paying dividends, or was subjected to a VAT, their preferences for burning cash in huge bonfires might change.
That’s a reasonable reaction, Arnold. I would say that short term supply at the metro area level is like a tangent curve, with the left boundary equal to the existing supply and the right boundary equal to the maximum rate a metro area is capable of building. Many cities can build at 3% a year or more. Closed access cities are generally less than 1%. The right boundary has shifted left in those cities to a rate lower than any reasonable demand level would cross. Its lower than Detroit or St. Louis. Net growth is barely positive. Certainly a shift in demand is what will trigger short term equilibrium changes, and some of that was related to credit markets, etc. But I call it a supply bust because it’s an arbitrarily low rate of potential growth far below what functional cities in a free society should be able to accommodate.
The supply curve in cities like Phoenix is much wider. The flood of households out of California pushed demand during the bubble to the right boundary. So there was a classic bubble in those cities with inelastic short term supply and elastic long term supply. But the source of the new demand was the narrow supply curve in coastal California.
Wow, thanks for replying Kevin. I was just going to say my interpretation of your supply bust was that the housing market worked kind of like a water balloon with demand being water from the faucet. In an open access city the balloon grows as more water goes in, and so long as the water doesn’t enter faster than the growth rate it all stays there. In a closed access city, the balloon doesn’t expand much at all, so any inflow above a very tiny level results in water back flowing out and splashing all over. Some of that water is water that didn’t make it into the balloon, and some is water that was in the balloon but was displaced by new water. The overall volume of the balloon doesn’t change, as the inflow is roughly equal to the outflow of water, even though the make up of the water changes.
Maybe it was just the flooding metaphor, but is that pretty close to what you mean?
Yeah. That seems like a good analogy.
I find it hard to reconcile this description with the history of the bubble and post bubble years. Phoenix’s home prices grew faster than San Franciscos from 2000-2006, and when you count that Phoenix had a much faster supply growth during that span the gap is even wider.
As someone with a front row seat to the housing crisis, I think Joel Kotkins’ narrative does not come very close to describing what happened. Rather, events unfolded like this. By the way, I agree with commenter Kevin Erdmann regarding the different supply characteristics of core/urban markets versus tertiary markets:
Circa 2001 – 2004
1. Housing supply: Increases in housing supply was relatively moderate in the wake of the dot.com bubble/bust
2. Supply of mortgage credit: Supply of credit was pretty consistent with historical norms, although subprime and Alt-A mortgages were assuming a larger share
3. Demand for houses: Moderate increase in demand, following the dot.com bubble/burst
4. Net effect on home prices: Moderate increases in both primary urban (e.g., San Francisco) and tertiary markets.
Circa 2005 – 2007
1. Housing supply: Increasing home prices led to better economics for home builders, and thus acceleration in home production. However, increases most substantial in tertiary markets where houses could be built at scale (i.e., large 100+ tracks). Increasing housing supply not as pronounced in primary urban markets, although this varied from market to market. Barriers to entry are typically higher in core urban markets.
2. Supply of mortgage credit: Mortgage underwriting standards were loosened considerably as Subprime and Alt-A mortgages became a mainstream product, and because delinquencies were very low. Mortgage backed security and collateralized debt security markets grew and became more liquid, causing increased prices for these securities, and consequently lower yields. Don’t recall exactly what the Fed and interest rate market was doing during this period, but I recall that rates were pretty low by historical standards.
3. Demand for houses: Tail wind generated by positive price appreciation, plus loose credit cause increasing housing demand that outstrips supply coming online. New and inexperienced buyer pool with speculative bent piles on.
4. Net effect on home prices: Although housing supply was increasing, it did not keep pace with increases in housing demand, plus greater availability of mortgage credit. All conspire for very rapid home price appreciation.
Circa 2008 – 2012
1. Housing supply: Because of the lag time in papering lots to final production and sales, production was still very high in the 2008 – 2009 period. Although “new” supply then starts to drop dramatically, homes in foreclosure and distress begin to create a substantial new substantial source of supply
2. Supply of mortgage credit: With the exception of conforming product (i.e., conventional 20% down Fannie Mae and Freddie Mack loans), the loan market collapses, once it become evident that the fundamentals of many mortgages were based on incorrect assumptions regarding default rates, and due to other issues (i.e., collapsing liquidity in these markets, etc.)
3. Demand for houses: Demand collapses as credit dries up and homeowners become aware of changing fortunes in the market. Inexperienced speculative investors flee market
4. Net effect on home prices: Home prices decline substantially in most markets, with the declines tending to be greater in the tertiary markets as compared to the core urban markets. The preponderance of construction related economic activities in the tertiary markets makes matters worse as laid off construction workers (and folks in the related fields of manufacturing, and real estate) comprise a larger share of the population in tertiary markets.
See some sites like
https://www.thinkglink.com/2005/11/30/real-estate-trends-2005-new-construction-sales-up-existing-homes-down/
Housing peaked in 2006. VERY very different market after the peak. Up thru the end of 2005, early 2006, there were speculators – house flippers – buying houses and selling them in 3 or 4 months. Netting 1% in 3 months on $200,000 loan, where the bank puts up 100%, is something “quite a few” did. Hundreds, maybe thousands. Not tens of thousands (buying then selling more than 5 houses in a year in their name).
As the market was rising (house prices never really drop!), more short term speculators were joining the market — all were making money.
Once the peak was reached, lots of speculators tried to cash out; all stopped buying, demand dried up.
Once the speculators stopped buying, unsold but built, or almost built houses started to drop in price. (2006-2007, depending on market).
When a housing market starts to drop, there is big pressure to “wait”, until there is a bottom.
New construction dried up. Millions of illegals stopped getting hired.
Millions.
Illegals, not on the books, not in the statistics, “invisible” to all professional economists — yet they had been getting paid and had been building houses. Now they weren’t. Huge reduction in demand, tho not visible, because the reported retail numbers were still rising. Their status means that they are not part of any “unemployment” numbers. So their unemployment didn’t count in the statistics.
The millions of illegals, directly and indirectly dependent and influenced by housing construction are part of the reason professional economists can’t quite agree on what caused what in the housing bubble, construction crash, price crash, Mortgage Backed Security crash (house prices never go down! oh wait… oh nooooo!), and financial crises. Big lag after bubble pop.
Glad Kevin is here, but I continue to disagree with his defining away the reality of a housing bubble and a pop. (Tho his site is excellent).
Arnold thinks about the supply bust as being more
plausible to think of housing supply in the Closed Access cities as relatively fixed,
I’m sure Arnold is more correct here. A true “bust” would mean a decrease in housing.
The closed access cities had an increase in demand, but little or no increase in supply — that’s why the marginal sold houses went for so much more.
Can’t help but wish I’d somehow kept that Palo Alto home I had in 1980. $150k then, now $2-3 mln. Very closed access, huge demand increase. Not a “supply bust”.
Various and Tom,
Those seem like generally good observations. Points I might interject would be:
1) I dont think you’ll find much speculative building in the bubble cities, at least from the builders’ perspective. They couldn’t build homes fast enough for the buyers they had in 2004-2005. Builders seem to have reacted pretty quickly to changing conditions. But by 2006, they were facing high cancellation rates. Building was negligible by 2008.
2) Migration patterns are really important. The Contagion cities (bubble cities not closed access) had a large inflow in 2004-2005 and by 2008 flows were negligible. That’s what made it seem like they had overbuilt.
3) In the Closed Access cities, something like 8% of the housing stock is owned by households with a mortgage who earn 60% of US median income or less. And those households were streaming out of those cities in large numbers. It is implausible that lending to those households could have had a measurable role in pushing prices up by 100%+.
4) in cities like Phoenix, prices hadn’t even started to rise unusually until mid 2004 when the fed had started raising rates.
Following your own population graphs- https://www.idiosyncraticwhisk.com/2016/01/housing-series-part-101-effects-of.html
Phoenix had a lower population growth rate in 2004 than it did for all of 1995-2000 and 2005/06 were lower than 1995 and (eyeballing) slightly lower/even with 96/97/98.
So population growth for Phoenix between 2000 and 2008 was highest in 2005/2006, but housing permits for Phoenix/Mesa/Scotsdale (going by easily accessible data) and Arizona peaked in mid 2004.
https://fred.stlouisfed.org/series/PHOE004BP1FH
https://fred.stlouisfed.org/series/PHOE004BP1FH
The July 2004 peak for the P/M/S area was 40%+ higher than the July 2003. If we cram the Phoenix area home price index over top
https://fred.stlouisfed.org/graph/fredgraph.png?g=nhQs
We see that permits fall well off before the peak before price increases break to the upside. This indicates that builders were pulling back first especially when you consider that they were doing so with strong incentives to keep building (ie faster population growth and faster price increases).
Tom,
Good points. The speculation frenzy created a huge distortion in the perceived housing demand. That activity was a bubble that popped. When that happened the home building industry ground to a halt. Those jobs vanished, many of them never accounted for in the first place.
Also crushed in the bubble pop was the price of land, especially in the exurbs. Land that was valued at $20k an acre, undeveloped, crashed back to $5k an acre. Loans on this land went into default. So much credit associated with housing we t into default.
Now, ten years later, raw land is cheap but the cost of permits is high, the cost of development is high, and the cost of construction is high. The housing economy has not returned to what it was in the early 2000s.
Millennials aren’t that young anymore, and it kind of sounds like they’ve just gotten old enough that a house in the suburbs starts looking more attractive than a loft in the city.
P.S. We have housing supply busts here in Florida, usually after hurricanes. A town called Marianna was pretty much flattened one afternoon last year. Naturally, shelter in the area was at a premium for a while (probably still is).
Kevin Erdmann is right, and the root of the problem is property zoning. Cities along the West Coast cannot build housing, due to property zoning and the complete trampling of property development rights.
Perhaps Erdmann erred in the use of the words “supply bust.” Yes, it is more of a fixed supply along the West Coast. But this is trivial cavil and sniveling, to focus on that expression and ignore the larger reality.
What amazes me is that free-market types are so interested in Trump’s trade tariffs or possible minimum wage hikes, but then go mute on property zoning .
Many free-market types do propose easing property zoning and then get attacked by people who fear truck stops and other business next to them.
Also property zoning is not an absolute limit but a starting point for negotiations used by non free market types.
“Those organizations often use the carrot of increasing zoning heights when negotiating with developers in order to ask for higher rates of affordable housing. But if those heights are already set higher, their bargaining position may be weaker, Fujioka said.”
http://www.sfexaminer.com/new-version-housing-density-bill-calls-suburbs-share/
Nobody treats property zoning with the respect it deserves, and as such they run into resistance that is all too predictable. You need to view changing zoning as no different from changing lot lines, as a breach of contract and a violation of property rights. If you want someone to give up a property right fairly paid for, you need to offer something adequate in return. Not run roughshod over them.
Here are things you need to consider when asking to change zoning:
1) Is the local transport infrastructure capable of handling the new arrivals?
2) What will the demographics of the new arrivals look like?
3) How will the local schools be affected?
If you don’t have good answers to these questions, why should people change their zoning.
My own community is in the middle of a zoning battle. It’s a typical gentrifying neighborhood that started with some artist types moving into a run down area and fixing it up. The neighborhood is on the cusp of a transformation and the developer did the following:
1) Asked the city to put up parking meters on previously free street parking. Actually lied to the city about the residents wanting it, lied to us about telling the city about that, and got caught in the lie at a community meeting. They did this because they plan to built over the current parking (free) parking lots and then charge significant money to buy parking in their underground garages. If the free street parking is taken away that makes their paid spots more valuable.
2) Cancelled the leases of the founding artist studios that gentrified the area in the first place and told them to get out. Did this immediately without consultation, seems they think a Starbucks or such could draw in more rent.
3) Plans on building large low rent apartment/multi family buildings (the area was entirely owner owned and more on the high end before this). What do you think bringing a bunch of low income renters in is going to do to the community? I’ve already discussed the problems we have had with crime at length on this blog, what happens when those same people don’t even need to come in from another neighborhood? We are ten blocks from where the riots in Baltimore started, now that area is coming to us.
4) Along with above, any shot at the local school flipping to “acceptable” as the area gentrified goes out the window with those demographics moving in.
5) There is no plan for how the one single street that handles all incoming traffic is suppose to handle all this. It already gets backed up and can become intolerable when a truck is doing a delivery (also, one of the places out of the way trucks did park is going to be built over).
6) Oh and the new buildings will be way taller than existing structures and block all the sunlight and views.
So what exactly is evil about the NIMBY people that gentrified this land and made it valuable in the first place no wanting the property they poured their life savings into wrecked.
Seems like that is an unusual and unique case in terms of the low income housing. Most of the time that I hear about this, people are objecting to developers building expensive housing for people with high incomes in areas where land values are really high. Which impacts congestion, has variable impact on schools (a lot of those high income households may be DINKS replacing middle class families with kids, such that the local schools may have actually have a greater concentration of impoverished students afterwords).
Though I will say that it makes some sense that a story like this would come out of Baltimore, which is a basket case of a metro, and not someplace like Denver, Austin, Seattle, or Portland.
“According to demographer Bill Frey, the 2013–2017 American Community Survey shows that New York now suffers the largest net annual outmigration of post-college millennials (ages 25–34) of any metro area—some 38,000 annually—followed by Los Angeles, Chicago, and San Diego. New York’s losses are 75 percent higher than during the previous five-year period.”
I have to say that I don’t like this approach, there are a ton of issues selecting an age bracket and then defining net migration in raw numbers. You could have the opposite effect be true, if lots and lots of people moved to NYC at age 22-24 then it would be natural for NYC to see large decreases in age 25-34 year olds 3-10 years later as some of those people will have decided that their wants have changed or that NYC is not for them (I’m also unclear of post-college means has a college degree or just the age bracket).
It is also not particularly informative to compare to the previous 5 year period when that period covered most of the great recession and slow recovery.
There are lots of cross currents in Chicago. Chicago is a city of micro-climates, with overall net outmigration.
There is a $6b mixed use development with approx. 5,000 new residential units that will be built on previous industrial land. So all net additions. There are going to be a couple 600 ft buildings in an area with few buildings higher than 5 stories. Yet the politicians are all up in arms about crowding and affordable units. I imagine that millennials will be the primary occupiers of the units.
Meanwhile, in other areas of the city, some not that far from the new development, there are abandoned lots that the city gives away for $1.
Perhaps you would find it more believable if the words “housing cartel” were used in place of “supply bust.” There’s no denying that so-called urban planning is exactly such a cartel.
Sorry Arnold. I realize I should have responded directly to your comments on my earlier replies. I do address them briefly later in the book.
On international migration, most major metro areas in the US take in net foreign migration. The closed access cities arent particularly unusual in the scale of those flows. But since their building markets are sclerotic, it is a source of pressure on domestic outmigration.
It is possible that an unusual portion of closed access permits are replacement units that dont add to the net housing stock. I think a larger factor may be housing utilization. In the late 90s, closed access populations were rising faster than their housing stock. That was unsustainable. The credit boom was facilitating a decline in occupants per unit that was a rebound from 90s trends.