Paul Krugman ponders the meaning of the collapse of banks in Georgia and draws the predictable lesson - we need more regulation and more consumer protection.
In the course of this non-surprise he cites for support a column he wrote back in 2005 explaining the divergence in prices in the US real estate market. His conclusion is that what doesn't go up can't come down, but he has chosen to forget some basic economics to reach that conclusion. Here we go, from 2005:
When it comes to housing, however, the United States is really two countries, Flatland and the Zoned Zone.
In Flatland, which occupies the middle of the country, it's easy to build houses. When the demand for houses rises, Flatland metropolitan areas, which don't really have traditional downtowns, just sprawl some more. As a result, housing prices are basically determined by the cost of construction. In Flatland, a housing bubble can't even get started.
But in the Zoned Zone, which lies along the coasts, a combination of high population density and land-use restrictions - hence "zoned" - makes it hard to build new houses. So when people become willing to spend more on houses, say because of a fall in mortgage rates, some houses get built, but the prices of existing houses also go up. And if people think that prices will continue to rise, they become willing to spend even more, driving prices still higher, and so on. In other words, the Zoned Zone is prone to housing bubbles.
And Zoned Zone housing prices, which have risen much faster than the national average, clearly point to a bubble.
Krugman summarizes his notion in a subsequent column and restates the belief that parts of the country are immune to real estate bubbles:
Last summer I suggested that when discussing housing, we should think of America as two countries, Flatland and the Zoned Zone.
In Flatland, there's plenty of room to build houses, so house prices mainly reflect the cost of construction. As a result, Flatland is pretty much immune to housing bubbles. And in Flatland, houses have, if anything, become easier to afford since 2000 because of falling interest rates.
In the Zoned Zone, by contrast, buildable lots are scarce, and house prices mainly reflect the price of these lots rather than the cost of construction. As a result, house prices in the Zoned Zone are much less tied down by economic fundamentals than prices in Flatland.
So - in Flatland, because supply easily increases to meet new demand, we cannot see price bubbles. Krugman stops there, but let me ask some follow-ups - can we contemplate "supply bubbles", where we see an unwarranted (viewed ex post) increase in supply rather than price? And what does Krugman suppose will happen in Flatland if demand retreats following an economic boom? Will developers rewind the videotape and un-build all the new construction that was called forth during the surge in demand? Or will prices plunge while real estate brokers toss around words like "glut" and "overhang"?
Let me present a textbook supply-demand picture for housing in Flatland and the Zoned Zone as imagined by Paul Krugman. Demand in both regions slopes downward. As we enter Boom Time the demand curve shifts to the right; when the economy goes bust, the demand curve shifts back to its starting point.
At the outset, both zones are in equilibrium (at P1) with a housing price of 5 units (which I chose because it happened to scale easily).
In the Zoned Zone, supply is very inelastic. When demand surges prices rise a great deal and very little new housing arrives on the market, just as in Krugman's story. The result is a price increase to 7 (shown at Z2) with just a small increase in supply.
And in Flatland there are plenty of building opportunities, so when demand surges prices stay flat and supply expands to meet it; the result is a new price still at 5, but with a notably larger housing stock, as shown at F2.
And that is where Krugman's story ends, with no bubble in Flatland. However, we are going to soldier on and see if anything pops!
Suppose demand now fades back to its initial level in both the Zoned Zone and Flatland. Shouldn't prices simply work their way back to their initial levels? Good question! In the long run they ought to, but how long will that take, and what will happen while we wait for houses to fall down? In the short run very few houses were built in the Zoned Zone and a lot were built in Flatland. What is going to happen to them?
Unless people rev up the bulldozers and take supply off the market, in the short run prices in Flatland will fall from F2 to F3 (Roughly 2.5 units), which represents the price on the (initial and bust) demand curve that will account for all of the available homes. Clearly, that price is well below the long-term equilibrium, so we would expect new construction to plummet. In effect, the supply curve in Flatland is always kinked - adding new homes is easy but subtracting them is deeply problematic. That said, eventually, the housing stock will shrink to meet the reduced demand, and prices will get back to 5.
And what about the Zoned Zone? Well, they have a few more houses than in the pre-boom era, so their prices should fall a bit below 5 (roughly 4.5 on this chart) and then work their way back up to 5 as the housing stock shrinks slightly.
The key point is that Flatland prices can plunge like stones even though they never soared like eagles. Or, more prosaically, developer's units and foreclosed homes can sit unsold with no meaningful market price at all. Now, the actual decline will depend on the shape and shifts of the demand curve; we have not seen 50% declines in Atlanta, but we are certainly seeing something that looks like a post-bubble economy, despite Krugman's insistence that Flatland would be "pretty much immune" to bubbles.
By overlooking this key point Krugman has talked himself into a bit of a corner and is trying to draw grand lessons from the flawed conclusion that Flatland can't have a real estate bubble. Of course they can, as described by the Atlanta Fed:
The deeper and more rapid deterioration in the performance of Southeastern banks is likely the result of their greater exposure to the effects of the housing downturn. Given the region's historically strong demographic trends, banks headquartered in the Southeast have naturally played a more active role in financing the acquisition, construction, development, and sale of real estate. Indeed, the business plans of many banks chartered earlier in this decade called for specialization in this type of lending. This specialization is readily apparent in Southeastern banks' loan portfolios, where loans secured by real estate account for 71 percent of lending compared to 55 percent in the rest of the nation (see chart 2). (Lending secured by real estate includes construction and development; commercial, multifamily, and single-family mortgages; home equity; and farm loans.) Joseph R. Mason, a banking professor at Louisiana State University (LSU), said that high levels of exposure to real estate lending contributed to banks' failures, adding that Florida banks with limited branches were especially vulnerable to a downturn. "Many smaller, less diversified institutions therefore had debilitating local real estate exposures from the real estate bubble," he said. That seems pretty simple - if a lot of local business is geared to building homes and those businesses become redundant, local banks are likely to be pressured on their commercial loans as well as on loans to homeowners. Now even a casual reader might have wondered just where Flatland was located - this is from Krugman's current column:Real estate lending at the center of the tempest
Most of the post-bubble hangover is concentrated in states where home prices soared, then fell back to earth, leaving many homeowners with negative equity — houses worth less than their mortgages. It’s no accident that Florida, Nevada and Arizona lead the nation in both negative equity and mortgage delinquencies; prices more than doubled in Miami, Las Vegas and Phoenix, and have subsequently suffered some of the biggest declines.
But not all of Flatland has gotten off lightly. In particular, there’s a sharp contrast between the two biggest Flatland states, Texas — which avoided the worst — and Georgia, which didn’t.
Florida, Nevada and Arizona are in the Zoned Zone with limited land? Hmm, I could listen to a story about zoning restrictions and limited water in Nevada and Arizona but still - back in 2005, the Zoned Zone was described as lying "along the coasts". And let's pick on Las Vegas - in 2000, they had 559,799 housing units; by 2009 that had increased by 46% to 819,600. Does that sound like the Zoned Zone, where supply is inelastic? By way of contrast, during the same period New York City experienced a tremendous building boom, during which its housing stock rose from 3,200,912 to 3,350,926 (an increase of 4.7%). And Los Angeles? Their housing stock grew by 3.1% from 2000 to 2006-08, per the Census Bureau, from 3,270,909 units to 3,372,376.
If the concept of a "Zoned Zone" is meaningful then Las Vegas does not belong in it with New York City and Los Angeles. On the other hand, if Flatland is immune to bubbles, why did prices in Las Vegas rise so fast, and why are both Las Vegas and Atlanta suffering from the current collapse? [My bad - Krugman already explained that Las Vegas is in the Zoned Zone, although he does not mention the 46% increase in the housing stock. I suggest a revised definition of Zoned Zone to be "Any area where prices rose".]
Puzzling. My suggested resolution - Krugman should scrap the whole "Zoned Zone" versus "Flatland" framework as being incomplete and misleading. That would spare us such howlers as this, from the current column:
To appreciate Georgia’s specialness, you need to realize that the housing bubble was a geographically uneven affair. Basically, prices rose sharply only where zoning restrictions and other factors limited the construction of new houses. In the rest of the country — what I once dubbed Flatland — permissive zoning and abundant land make it easy to increase the housing supply, a situation that prevented big price increases and therefore prevented a serious bubble.
Since Krugman has made up his mind that Georgia could not have experienced a bubble (such as a supply bubble) he is left to grasp for other explanations of their current banking problems. Here we go:
But not all of Flatland has gotten off lightly. In particular, there’s a sharp contrast between the two biggest Flatland states, Texas — which avoided the worst — and Georgia, which didn’t.
This contrast can’t be explained by the geography of the two states’ major cities. Like Dallas or Houston, Atlanta is a sprawling metropolis facing few limits on expansion. And like other Flatland cities, Atlanta never saw much of a housing price surge.
Yet Texas has managed to avoid severe stress to either its housing market or its banking system, while Georgia is suffering severe post-bubble trauma. The share of mortgages with delinquent payments is higher in Georgia than in California; the percentage of Georgia homeowners with negative equity is well above the national average. And Georgia leads the nation in bank failures.
Let's see - the AP had a long article about Georgia a year ago. Their take:
Eleven Georgia banks, most surrounding Atlanta, have been shuttered by regulators, followed by nine in California and four in Florida. Experts predict more could be closed in Georgia in the future. But what propelled Georgia to No. 1 in bank failures is complicated.
Experts say it's a combination of an antiquated state law that favored a plethora of smaller community banks over multi-branch giants; a population explosion in metro Atlanta that fueled massive suburban real estate development and a crush of new banks formed to cash in on the Atlanta boom shortly before the market tanked.
First, Georgia is home to a huge number of state and federally chartered banks. At the end of 2008, Georgia had 334 banks. That's more than California, which has nearly four times Georgia's population, or Florida, which has twice as many people. Only five states — Texas, Illinois, Minnesota, Iowa and Kansas — have more banks than Georgia, according to the FDIC.
What these states had in common, until the mid-1990s, was some of the nation's most restrictive laws on branch banking. Georgia, for example, prohibited banks from opening branches across county lines until 1996.
The law shielded local banks from worrying about competition from out-of-town rivals. It also guaranteed that Georgia, with a whopping 159 counties, would have a correspondingly large number of banks.
Lots of small banks growing without diversifying:
Metro Atlanta had three of the nation's 10 fastest growing counties of the 1990s. Because of that growth, about half the state's banks ended up clustered around Atlanta, said Joe Brannen, president and CEO of the Georgia Bankers Association.
"Georgia is a tad unique in that we don't have five or 10 big metropolitan areas. We've got one real big one," Brannen said. "We haven't enjoyed the statewide growth in population that Florida or California have."
And from a different source, the same story:
Georgia’s unwanted role as the nation’s current bank failure capital stems from a perfect storm of economic factors. The state’s rapid growth, coupled with the recession and collapse of the residential mortgage market contributed powerfully to the weakening of banks that were heavily invested in residential development and construction loans.
“We are in a 500-year flood in the world, and in Georgia it is especially significant. We are no different from other fast-growing states — Florida, Arizona, Nevada, California — states where people want to live,” says Joe Brannen, president and CEO of the Georgia Bankers Association in Atlanta.
So we have the Atlanta Fed and two news services telling us that a collapse in residential and commercial real estate wiped out a bunch of small Georgia banks that were too heavily involved in real estate lending. And we have Paul Krugman with a Flatland theory that defies both textbook economics and actual experience (unless Las Vegas is in the Zoned Zone despite a 46% increase in housing stock in eight years).
My guess as to the Texas-Georgia discrepancy that intrigues Krugman? Texas bankers went through the wringer in their real estate bust which followed the oil bust of the 1980's and learned something.
UNRELENTING: I know (KNOW!) that in the course of some self-congratulatory remarks about how he called the housing bubble, Krugman singled out the Flatland column as a particular source of pride for its originality and insight. I would love to find that cite [and I have!], which I would guess is from after August 2008, but right now my knowledge is a bit of a faith-based initiative.
AS GOOD AS IT GETS: Here is Krugman reminiscing fondly:
Calculated Risk, in a discussion of home price declines, links to my three-year-old analysis, That Hissing Sound, which I think was one of the best pure-economic pieces I’ve done in my tenure at the Times.
He may be right!
And let's have a To-be-fair moment - if the key insight is that looking exclusively at prices won't tell you if you are in a bubble, well, OK. In Flatland, a bubble is preceded by above-trend levels of construction, not price increases.
On the other hand, other lessons simply don't follow. In his current column Krugman is locked in on the notion that since Georgia banks could not have gone through a bubble, their current travails must be due to something else. In my world, they financed what is clearly, in retrospect, a supply bubble which is now being dealt with by price declines.
Once again TM, thank you for reading Krugman so we don't have to!
Posted by: Mad Jack | April 12, 2010 at 09:06 PM
Your analysis is spot on. The idea of Flatland vs. Zoned Zone is empirically and theoretically correct, but it has nothing to do with bubbles the way Krugman seems to think.
The best popular article on the subject (pre-crash), in my only slightly prejudiced opinion:
http://www.theatlantic.com/magazine/archive/2007/11/a-tale-of-two-town-houses/6334/
(I'm using Safari, so I don't think all that LUN stuff will work for me.)
Posted by: srp | April 12, 2010 at 09:15 PM
Zzzzzzzzzzzzzzzzzzzzzzzzzzzz
Posted by: MarkO | April 12, 2010 at 09:33 PM
TM-
Might you compare the overbuild of twisted copper leads from the 1910's on out to the fiber optic, in ground, install going on today?
I think it's money well spent on a long vision project. Unless you have a faster way to provide data to end points than those laying the fiber, this is a fully recoverable system if impacted by an EMP. Not as fast as the old, rotary switch, self powered, twisted pair, copper system (that still exists, BTW, thanks NORAD!), but a few swap out modules and the lasers are up and running. Cheap lasers too.
Posted by: Melinda Romanoff | April 12, 2010 at 09:43 PM
with Bill Gross dumping treasuries for German debt and numerous hints of inflation. It's Katie bar the door. We'll be lucky if it is as moderate as during the Carter years.
Just as Obumble jawbones the Cinese into revaluing the RMB, we gonna get double/triple whammied. Krugman should be tarred and feathered for his nonsense.
Melinda, that overbuild is now being sucked up by streaming video and other bandwidth greedy media. It will get interesting when the Indians start getting internet disruptions.
"Is your computer turned on?" may get a lot more expensive.
Posted by: matt | April 12, 2010 at 09:53 PM
What Krugman misses is that some of the worst of the zoned zone has actually continued to appreciate. In San Francisco proper, real estate has gone UP since 2007.
On the other hand, developers constructed housing on former farmland in Galt and similar towns in the central valley. Those houses have dropped by two-thirds in some cases.
As I understand the issue in LA and San Diego, the steepest drops in prices happened in those areas far from the center* where building restrictions were lax. I wouldn't speak for CT, but in StL, we have seen only minimal drops in fully built-out areas (housing stock 100-150 years old), but we have seen your graph in action in what used to be productive pastures.
-------
* Yeah, I know. For LA let's just use Malibu and Venice Beach for their restrictive zoning.
Posted by: Walter | April 12, 2010 at 09:56 PM
Enough of this Krugman stuff. How are we going to entice him here to play poker with us?
Posted by: Clarice | April 12, 2010 at 10:03 PM
I tried to get through the article but I zoned out.
Posted by: Dave (in MA) | April 12, 2010 at 10:12 PM
matt-
The fiber "pipe" isn't even being tested to its simplest capabilities. The iterations are endless, the frequencies for the digital combinations might approach complication, but the pipeline is way bigger than you can imagine. Calcuble, but way bigger than estimated demand.
At least these devices won't be "taxed-upon-entry" like new medical devices will.
Posted by: Melinda Romanoff | April 12, 2010 at 10:23 PM
matt-
One other point, and I hope you have forty-nine minutes for the most interesting debate between the two most diametrically opposed economic thinkers today, Jim Grant and David Rosenberg.
Pay particular attention to David's assertions on deflationary aspects of the current economy.
You will not be disappointed. And you won't by Jim Grant either, by the way.
Posted by: Melinda Romanoff | April 12, 2010 at 10:40 PM
"At least these devices won't be "taxed-upon-entry" like new medical devices will."
They will *now*, Mel; you've spilled the beans.
Posted by: Frau Steueramt | April 12, 2010 at 10:47 PM
Walter,
You have it in one. The Riverside County explosion is the SoCal equivalent of Galt. There is also a non-neglible variable that we might call Mozilla Fraud, having to do with pumping unqualified buyers into houses they couldn't afford in order to feed the Wall Street MBS pimping operation.
A close study of the mortgage broker, realtor, appraiser fraud network and its very tight ties to the larger Wall Street hogs would generate a much clearer picture of how to blow [up] a bubble than toying with Krugmans ZonedZone/Flatlander cartoon. That one is strictly for consumption by North-easterners cramped in their cocoons.
Posted by: Rick Ballard | April 12, 2010 at 11:38 PM
my Wall Street buddies are struck by the same conumdrum, Melinda, but Gross is more often correct. When we have so many factors at odds, it is a very uncertain outlook. The debt bombs plus printing presses running wild are uncharted territory. It's a perfect environment for sharks such as Goldman.
As to the interweb pipe, the key factor seems to be the ability to multiplex, but nevertheless with the huge demands of gaming, streaming video, and other apps, I still stand by my statement. There may be a Moore's Law, but I tend to think that we will hit a wall sooner rather than later.
Posted by: matt | April 13, 2010 at 12:28 AM
From Krugman re Atlanta:
Basically, prices rose sharply only where zoning restrictions and other factors limited the construction of new houses.
What a crock of shite!
The median value of prices in the ATL soared over the last ten years due to Blue hell flight. These folks came in and instead of paying cash for a $300k 3,500 sf home (cash which they had in abundance due to the difference in home prices in blue hells v the south and the attendant LTV cash out upon selling their blue hellholes) and instead constructed nifty new $1.1 million 4,000 sf homes with a new 80% mortgage (all cash from sale reinvested or used to furnish said bluehellhole).
Now those suckers are carrying debt on homes that are not marketable and for which they can no longer afford the payments thanks to the downsizing of their jobs, balloon payments and ARMS.
The neighborhoods in crisis in ATL are at two very different ends of the real estate ladder - those that were "first home" neighborhoods that were flooded with new homeowners thanks to Fred/Fan's no money down, no doc mortgages to illegals, toe-holders and former renters who had no business buying homes (speculators were rampant in these hoods, too, BTW) which was endemic in all areas of the country AND the blue prog spec home neighborhoods built to accomodate the "nouveau riche" blue flighters who didn't have the common sense God gave to Moses to move into existing neighborhoods of upper middle class homes and have little to no mortgage and instead wanted a showplace and a honking big mtge to go with it. That additional 1K SF and a few fancy upgrades came at the expense of doubling the purchase price, but purchase they did.
The neighborhoods that are in fine shape in the ATL are the established neighborhoods (which are still building out) with residents that didn't feel the need to move into "BLING" city to establish their cred. These areas experienced a 40 to 60% increase in appreciation thanks to their proximity to the blue hellholes and have only fallen back by about 15%. Foreclosures are minimal and the houses still sell at 30 days or less. Many of these home were upgraded but within reason so the homes have the bling but not the size of the bluehellholes and are usually on much larger lots.
The builders have wised up and now the $1.1 million neighborhoods are being built out (after 12-18 mos of no new starts) with $400K homes, and those are just now starting to sell. The existing $1.1 behemoths, not so much. And the new "Mutt and Jeff" neighborhood aesthetics are amusing...
BTW, ATL real estate agents call this bluehellhole phenomenon the "second coming of the carpetbaggers."
Posted by: Stephanie says Obama sux | April 13, 2010 at 01:20 AM
Florida, Nevada and Arizona are in the Zoned Zone with limited land? .. And let's pick on Las Vegas - in 2000, they had 559,799 housing units; by 2009 that had increased by 46% to 819,600. Does that sound like the Zoned Zone, where supply is inelastic?
You can't compare just absolute levels of supply. It's supply relative to demand. Population and housing demand soared even faster than housing in Las Vegas and Nevada.
What makes Nevada unusual is that so much of the land is owned by the Federal Government, and government land sales slowed.
Some areas with extremely limited housing supply have house prices that have continued to go up, because they're simply targeting themselves at the wealthiest in the country.
Ed Glaeser's papers and recent book, as well as this study suggests that Florida and Arizona did adopt growth-management methods that slowed down house building.
It's not just the amount of housing that's built. It's also the amount of added supply relative to demand, and how long it takes for the permitting and planning process to proceed. If it takes years to get a house built, a bubble will proceed while everyone's waiting for the housing to get built.
Posted by: John Thacker | April 13, 2010 at 11:38 AM
Of all your great writing, I love the Krugman take-downs best of all (by the way, Andrew Ross Sorkin may be calling you later today, looking for a little help). Anyway, if Krugman thinks more regulation will help in containing market bubbles he must mean "other than in real estate".
Posted by: MTF | April 13, 2010 at 02:57 PM
Sorry, bad html. Here is the attack on Sorkin.
Posted by: MTF | April 13, 2010 at 02:59 PM
Krugman has unexpectedly backed over himself. In other news, job growth was unexpectedly low last week and Iran unexpectedly moved forward in it's effort to develop a nuclear weapon.
Posted by: EBJ | April 13, 2010 at 05:49 PM
You needed to follow up on the restrictive branch banking law differences between the various states and how it negatively impacted the banking solvency. You mentioned that 5 other states had more banks than Georgia - several had more population, but at least 2 did not: Minnesota & Iowa.
America's banking system has had a fragility since the days of Andrew Jackson when he went to war against the bank of the U.S. We've had restrictions on branch banks within states - thanks to the political clout of bankers. We never followed the Constitution and its forbidding of states' blocking interstate commerce, which banking seems to me to be.
In fact, we've had runs on banks, bank panics, etc. in hard times while our neighbor to the north, Canada, has not. That's because they didn't have the tangle of restrictive banking laws we did and were able to develop large, well-balanced national banks. They did not even have FDIC like insurance until the 1960s - but no runs on their banks in the 1930s like we did.
Posted by: Mark Michael | April 13, 2010 at 09:49 PM
I was going to share my side-splitting laughter at Krugman's assertion that Atlanta didn't have a real estate price bubble... but Stephanie did it for me. Nice job.
Posted by: Will Collier | April 15, 2010 at 08:58 AM
I live in the middle of nowhere, NY. Farm country. (Yes, NY has plenty of it.) My 3,000 sq ft house (on 8 acres) ran me $90,000 in 1997. Less than the then price of building it. I could sell it today for- maybe $90,000 if I got lucky. Population near here is going down- houses sit vacant and abandoned.
No housing bubble here, but still, prices-especially adjusted for inflation, are going down.
In most real rural areas, far from a city, prices are stable or down, because population is stable or down. If you build a house in the middle of nowhere, you build what you want, and count on living in it. You're not going to recover your cost on a sale.
Posted by: Harold | April 15, 2010 at 12:53 PM
Harold,
I get a few national real estate classifieds and am always amazed at the low prices in upstate New York, even lakeside stuff.
Comparable to scrubby Great Basin desert land.
Posted by: Ignatz | April 15, 2010 at 01:28 PM