Home prices have recovered, but construction remains severely depressed. Why is that?
Let’s take Phoenix, which was the poster child for the 2006 housing bubble. Prices are not back at the peak, but they are well above 2003 levels, even in real terms:
Given the high prices, you might expect housing construction in Phoenix to be highly profitable. Indeed, just looking at that price data I’d expect construction to be well above 2003 levels, assuming there were no factors depressing supply. In fact, new construction in Phoenix is roughly half the level of 2003, at least for single-family homes:
Why do I focus on Phoenix? Recall that many articles pointed to Phoenix as the smoking gun behind the 2006 “housing bubble” hypothesis. It was generally conceded that the price spikes in New York, Boston and California might have partly reflected building restrictions, the so-called NIMBY phenomenon. But cities such as Phoenix and Las Vegas were seen as having almost unlimited land for development. The price spike seemed irrational, as it was assumed that the high prices would lead to a surge in new construction, eventually bringing prices down to much lower levels.
But what if supply is also constrained in Phoenix and Las Vegas? I don’t have any good explanation for what that might be so, but the data strongly suggests that there is some sort of supply problem. The high prices are back, but construction remains severely depressed. During the period from 1950 to 2006, the sort of prices we now see in Phoenix and Vegas would have led to a huge surge in housing construction. For some reason we are not seeing that surge. Thus the most powerful evidence in favor of a 2006 housing bubble—the anomalous rise in house prices in the inland Southwest—is no longer evidence for the existence of a housing bubble.
I have no idea why supply in these markets is so constrained. I’ve read articles that make vague references to the cost of land and labor, but no real explanation of why things are so different from 2003.
Kevin Erdmann wrote by far the best account of what really happened during the housing bubble:
READER COMMENTS
ee
Dec 18 2019 at 7:26pm
Maybe it has to do with the facts that home building gains productivity slowly and immigration is constrained
Scott Sumner
Dec 19 2019 at 1:12pm
Possibly. Do we have data on construction wages?
B King
Dec 20 2019 at 2:39pm
on a real basis, both wages and costs are up *some* but not enough to account for the decline in building.
https://fred.stlouisfed.org/graph/?g=pLBZ
john hare
Dec 18 2019 at 7:33pm
For my local area in central Florida, availability of skilled labor is a major bottleneck. At conferences, the problem seems to be widespread according to most of the people I discuss it with. The shortage could addressed if the unskilled would come in with the attitude of learning and moving up. Also, many of the smaller custom contractors are gone compared to a dozen years ago. The home building in my area is somewhat stratified with track builders on one end and high dollar custom building on the other.
May not be the problem you are writing about though.
Scott Sumner
Dec 19 2019 at 1:12pm
See my reply to the previous commenter.
john hare
Dec 19 2019 at 7:30pm
I don’t have documented data, just anecdotal. My regulars are making well above average for the county. Individual productivity is dropping some with age though machinery is better and offsets that. My prices are way above that of 10 years ago for concrete work.
Kevin Erdmann
Dec 18 2019 at 11:11pm
Thanks for the kind words Scott.
I would say supply in Phoenix was clearly constrained in 2005, but is clearly not constrained today.
Scott Sumner
Dec 19 2019 at 1:08pm
Then why so little building?
Phil H
Dec 19 2019 at 1:16am
This is a bit of an unformed idea, so I’m not sure if it’s coherent. In the UK, and I assume in the USA as well, the population is not growing particularly fast, nor is the housing stock deteriorating particularly fast. So what’s driving the need for *new* housing is the ongoing reduction in the size of households.
Quite how that affects demand is not yet clear to me, but it seems like it might be a bit different to other kinds of demand. People can choose to continue to live with other family members until certain levels of discomfort/financial freedom trigger the moves.
Then there’s issues like the one here: https://www.strongtowns.org/journal/2019/10/29/is-your-city-infrastructurally-obese where towns already have a lot of vacant real estate, though it may be in the form of extra rooms rather than completely empty properties.
All of which *could* make for some weird demand effects? That *could* affect new builds?
Thaomas
Dec 19 2019 at 8:44am
Perhaps this is an aspect our your explanation, but what about the crisis itself plus more years of demographic change led to a shift in the type of demand for housing, more dense and “urban” and for that supply IS restricted for the same NIMBY/regulatory problems seen in the coastal cities.
Scott Sumner
Dec 19 2019 at 1:11pm
But then why have single family house prices risen so strongly?
Kevin Erdmann
Dec 20 2019 at 1:54pm
I’m usually not keen on market segmentation explanations, so this may be a just-so story. But, it does seem like in some markets low tier prices have recently recovered enough that we might expect more new entry-level building, yet it doesn’t seem to have changed much. I wonder if there is a segmentation issue here. The entry level owner-occupier market has been sharply curtailed through lending. It may be that investors are willing to buy existing single family homes at low price points, but that for the renters market, multi-family units are so much more efficient to manage, that new units for the renters market are highly biased toward multi-unit. There are a few single family neighborhoods being built for rentals, but not many.
So, it is the case that in cities like Phoenix, multi-unit building is back at pre-crisis peaks while single family production lags. I suppose this leads to the question, then, is there a political supply constraint on multi-family in cities like Phoenix? Although there are some regulatory hurdles, of course, I don’t think there are multi-year queues of projects waiting for hearing after hearing, paying extortionary fees, etc.
I think the main problem is that there are natural marketplace frictions. The multi-unit market would have to triple to make up for the lost units in single family. That’s not a reasonable thing to expect to happen in the short term. And, developers are reasonably concerned about over-building a certain type of unit.
That’s why the arbitrary limits to the market that come out of suppressed lending are so distortionary. If the transition to renting was the result of a natural shift in demand, I think you’d see more supply response. But, there is a ton of pent up demand for single family that would bankrupt multi-family projects if lending regulations were loosened up to allow the market to settle at an equilibrium that reflected unencumbered demand, in which case, families would move out of multi-family into single family. So, multi-family supply response is somewhat timid.
https://fred.stlouisfed.org/graph/?g=pLB7
To allow multi-family to fill in the gap, the federal government would need to create some credible commitment to permanently enforcing an excluded class of non-owning tenants.
Or, it could be that homeownership was a large part of the draw to cities like Phoenix, and having that avenue diminished is part of the reason population growth has slowed there.
Scott Sumner
Dec 19 2019 at 1:10pm
The mystery is not demand, it’s supply.
P Burgos
Dec 20 2019 at 12:46am
The US population is shifting, and a lot of US housing is obsolete and substandard, and has been for some time. So there are a lot of new homes needed in the places that are growing and adding lots of jobs, while there isn’t enough demand in places like urban Detroit, Baltimore, and in lots of mid sized and small towns to keep houses from becoming dilapidated and uninhabitable.
OneEyedMan
Dec 19 2019 at 11:06am
You may know this, but it is unclear from the post. The Case Shiller indices are in nominal terms. I count CPI inflation of about 40% since July 2003. The index was 122 on that date, so the in line with CPI inflation would be 171. Given the rise is over 16.5 years, this just means that repeat sales house prices in Phoenix grew by slightly more than 1% faster than inflation (16.9 percent over 16.5 years). Is that a significant mystery requiring a structural or policy explanation? Maybe this is just a Baumol cost-disease phenomenon. House building is labor intensive and construction productivity grows slowly so perhaps that drives real costs up over time. That’s not probably not what is going on in San Francisco or NYC, but may be the only explanation required here.
Scott Sumner
Dec 19 2019 at 1:09pm
You’d still expect housing construction to be higher than in 2003, but it’s fallen in half.
P Burgos
Dec 20 2019 at 12:52am
This is a variation on a theme throughout the comments, but here it goes:
If, in real terms, dwelling prices in Phoenix have increased only a little bit, and construction wages have increased by quite a bit more in real terms, building homes in Phoenix should be less profitable now than it was 15 years ago. If that is so, it shouldn’t be a mystery as to why fewer homes are being built there.
Just an anecdote (and not in Phoenix), but I had siding replaced on our home last spring, and in talking to contractors, they seemed to be expecting their labor costs to increase by 5% a year for the next few years, which is definitely more than inflation.
Mark Brophy
Dec 21 2019 at 1:56pm
The CPI understates inflation, it’s closer to 5% than to 2%, maybe 7% is the best estimate, although asset inflation is much higher, the stock market is up 28% in the last year.
Kevin Erdmann
Dec 19 2019 at 12:03pm
Here is a comment with a couple of graphs, which I have posted at my blog, which I hope you might find of interest.
https://www.idiosyncraticwhisk.com/2019/12/housing-supply-isnt-constrained-in.html
MJ
Dec 19 2019 at 5:40pm
Maybe hysteresis? Like there was so much unemployment in housing construction during the financial crisis that a lot of those workers switched careers.
Also, maybe there’s decreasing returns on developing, like the easy pickings highly profitable land has already been developed.
Alan Goldhammer
Dec 20 2019 at 8:52am
Bill McBride’s blog, ‘Calculated Risk,’ does a nice job of covering real estate both generally and in many key areas. He does track Phoenix and posted THIS on Wednesday. Trying to come up with simple explanations for the lack or increase in new home building is hard to do. Two things come to mind for me, as McBride notes there was a lot of distressed property purchased during and after the Great Recession by investors. A lot of the homes became rental properties (in our neighborhood of Bethesda, MD about 20% of the homes are rental as investors have been buying them up rather than prospective homeowners) as tax preferences continue to make real estate an attractive investment. Secondly, home location is playing a more important role as workers don’t want long commutes. Perhaps all the desirable locations in Phoenix are built out and existing open land is not attractive to developers.
Warren Platts
Dec 20 2019 at 12:45pm
Um, you all realize that places like Phoenix and Las Vegas are in, like, deserts? And that they have been in a drought since 2001?
This from phoenix.gov:
Anyways, annualized housing starts for November are up 18% year over year and up 36% from 2014. Is that the sign of a crisis?
Scott Sumner
Dec 21 2019 at 12:22pm
Warren, You said:
“Is that the sign of a crisis?”
Yes, for the reasons I explained in my post. Do you have any comment on my analysis?
Warren Platts
Dec 21 2019 at 4:16pm
Scott, the statistics I cited are for nationwide housing starts–and they seem to be increasing faster than GDP or population growth. So if there is a crisis, that is confined to some local areas–like Phoenix, that you confine your analysis to. If you had instead analyzed where I live (western Pennsylvania), you would find reasonable prices and no housing shortage. Of course it snows here though.
As for Phoenix, as I said, the one obvious factor you did not take into account is WATER. You saw the quote above from phoenix.gov: they have enough water for the people that already live there, but there is no extra.
Therefore, if a developer proposes a new construction project in Phoenix, the developer is expected to acquire water rights and, if necessary, pay for the construction of new aquaducts to get that water to Phoenix. And the days when you could go to a courthouse in the Rockies or Sierras and just claim some water rights are long gone. In a place like Phoenix, you will get in trouble for collecting the rainwater that comes off your own roof! It doesn’t belong to you.
That entails you’ve got to find a farmer somewhere with a nice center-pivot irrigation system who’s making a minimum of $100,000 of profit per year and willing to turn her lush corn field into an arid pasture that’s only suitable for grazing a few goats on. Such deals exist, but it’s going to cost you a few million at a minimum.
I submit that also explains the multi-family versus single-family dichotomy Kevin Erdman identified above: presumably, the H2O required per person living in a multi-family unit is less than that in single-family dwellings, and thus the cost of providing a water supply eats less into the developer’s profit margin.
Bottom line: Prices in Phoenix are high, but they are not high enough to justify wholesale purchases of the water supplies that would be necessary to sustain a major housing boom in Phoenix.
Michael Pettengill
Dec 22 2019 at 4:03am
Easily solved by regulations, eg, no lawns, and tax hikes to fund water and sewer and water treatment to turn waste water into potable water. Very old tech. I remember when they tried to get president nixon to drink water flowing out of a new western “sewage” treatment plant.
Warren Platts
Dec 22 2019 at 10:23am
Easy to say, but fact is that the people who already live in Phoenix do not want to curb their water consumption even more than they already have in order to provide water for newcomers.
Hence the regulation that developers are responsible for securing–and paying for–new water supplies. The necessity of securing costly water supplies vastly changes the cost of construction equation and hence the ability to earn a profit on investment. Relative to the total cost of construction–including purchases of new water rights–housing prices in Phoenix are probably pretty cheap tbqh.
Kevin Erdmann
Dec 23 2019 at 10:29pm
Warren, I don’t think you’re on to something here, for two reasons. The first is that your description of the Phoenix water situation doesn’t match my impression of it very well. Phoenix doesn’t really use any more water than it did 60 years ago, because every acre of irrigated farmland that is replaced with residential building saves water. Water is certainly a concern, but it’s not a significant drag on housing supply. Clearly it wasn’t in 2005. Also, Phoenix isn’t really an outlier on this topic. Most cities in the US have “mysteriously” low rates of building since the crisis.
Henri Hein
Dec 21 2019 at 12:42pm
Couldn’t it be a capital issue? The other big industry that was severely hit in the Great Recession was the financial sector. House construction requires capital. Anecdotally, I know it’s harder to get a loan now than it was in 2006. Is it not likely the same is true for housing, so that it is harder to get projects financed?
Michael Pettengill
Dec 22 2019 at 3:56am
Guessing that areas like Phoenix are not welcoming builders like during 1995-2005. Either, bonding for roads, water and sewer, etc was supported by business to boost economic growth with new customers, with all revenues growing far beyond the debt service. Then the endless recession which made the debt service a big burden.
Or, developers paid for all infrastructure and then turned it/forced it on government. Corners were cut in construction which by the middle of the recession put burdens on government to fix the problems in response to angry taxpayers expecting the government deliver in exchange for property high taxes.
New Hampshire went through both kinds of such problems in the 70s downturn and late 80s downturn, so builders find developments hard and costly, thus eliminating the starter home tracts, a la Levittown.
I read about the government issues in newspapers, had friends in local government, and builders, so I heard the stories, and the “solutions”, ie stricter planning board oversight. Eg, regulations, standards. For those who were not around when the tax hikes had to be done during recessions, the rules seem arbitrary.
It’s all about taxes. And since especially 1980, hiking taxes is evil, so the people who must fix infrastructure never want to need to demand tax hikes to do their jobs.
Michael Rulle
Dec 24 2019 at 12:17pm
While secondary housing sales are small part of GDP, I believe it was the secondary market still was the primary cause of the tripping point.
One aspect of the housing bubble was the mark to market problem. Not the securities mark to market, but the actual housing mark to market problem. California was front and center—-and perhaps even Countrywide and a few others itself were somewhere at the head of the pack. The housing mark to market problem played out through the refinancing market—-which was at least half the increase in the amount of mortgage securities outstanding. There may have been a housing shortage, but there was no financing shortage.
Further, before the housing crash was being addressed in DC by the freaky deaky no nothings of the Bush administration, the actual secondary housing market was “clearing’ in places like Stockton and Phoenix as sales volume had began to rapidly rise prior to Sept 2008, at way lower than peak prices. The losses had already been taken——and for all practical purposes the crisis was already over.
I like Scott’s normal view on Fed tightness as causal for the crisis. Also, to repeat, any story which excludes refinancing is incomplete.
Comments are closed.