« From the Korean taco story | Main | Peter Leeson's new gypsies paper »
Why are so many homes unemployed?
Theories of unemployed labor are a subset of theories of unemployed resources.
The U.S. housing vacancy rate--an unemployment rate for homes--is at its highest level since at least 1965 (see figure). Why? Is it sticky prices? Lack of aggregate demand? Structural?
House prices may be sticky but they have fallen a lot--maybe not enough--but they have fallen a lot more than have wages. On the other hand, house prices rose a lot more than wages. Maybe house prices are sticky relative to the required variation in market clearing levels.
What about lack of aggregate demand? The homeownership rate was 67.2 in 2000 and today it's 66.9. Thus, we don't have too great a supply of houses in the aggregate so aggregate demand is likely a factor.
Is the problem structural? It does seem that we have too many houses in the South and the West where the boom was concentrated. If we think of the unemployment rate as a measure of where there are too many houses then the following figure shows that there is a positive correlation between the home vacancy rate and the unemployment rate. It's not as tight as one might expect, however. California, for example, has a high unemployment rate but a home vacancy rate slightly below the national average and many states such as Wisconsin have plenty of unemployment but a very low home vacancy rate.
My guesstimate is 50% AD, 25% sticky prices, 25% structural. Tyler would read it differently. I do think more progress could be made if greater attention were given to theories of unemployed resources and not just unemployed labor.
Posted by Alex Tabarrok on July 29, 2010 at 07:27 AM in Data Source, Economics | Permalink
Comments
I am very much in favor of this sort of exercise, but I don't completely follow your argument. For instance how are the "weak AD" and "sticky price" arguments conceptually distinct?
I don't so well follow your argument in this passage:
"What about lack of aggregate demand? The homeownership rate was 67.2 in 2000 and today it's 66.9. Thus, we don't have too great a supply of houses in the aggregate so aggregate demand is likely a factor."
I don't follow either why that shows "not too many houses" or the connection to AD.
On the empirics, should we aggregate state-by-state? Shouldn't the very populous California receive more weight in our assessment than say Rhode Island?
Posted by: Tyler Cowen at Jul 29, 2010 7:44:06 AM
Houses have increased their preference for leisure.
Posted by: matt wilbert at Jul 29, 2010 8:16:47 AM
I agree with Tyler, especially with regards to the supply of homes.
More importantly, you have not really considered the composition of homes, only the possibility of geographic bunching, about which I don't disagree. But, as I discus here [ http://www.liberalorder.com/2009/09/the-oversupply-of-larger-homes.html ] and here [ http://www.newsobserver.com/2010/02/09/328127/big-houses-big-inventory.html ], builders responded to the changing demand for the size of homes that were being demanded and there is now an oversupply of higher-end homes.
Posted by: Mark at Jul 29, 2010 8:17:06 AM
all the demand is coming from investors who intend to flip when housing recovers..
Posted by: rjs at Jul 29, 2010 8:26:04 AM
Prices have come down but are still too high.
Buyers may be waiting for prices to come down further.
Posted by: foosion at Jul 29, 2010 9:01:24 AM
How is the rate in the first chart over 10% and the second chart national rate between 2-3%. Are there different things being counted? Different scales?
Posted by: Eric at Jul 29, 2010 9:03:10 AM
Quick response to Tyler and Mark before I run off to a meeting.
On sticky prices v. weak AD the distinction I was thinking of was partial vs. general equilibrium. i.e. Sticky prices in the house market and everything else ok versus AD problem (perhaps caused by sticky prices more generally) reducing income which reduces the demand for homes. My judgment is that the latter is more important.
The actual number of homes is not that out-of-whack given the population--which is another way of saying we had price expansion more than quantity expansion. We could absorb the homes with our current pop.
I agree on weighting by pop.
Mark raises a good point about the type of homes--this could add to the structural or sticky price explanations. It would be interesting to know whether the homes that are currently being built in Las Vegas are different than the ones that are sitting empty.
Posted by: Alex Tabarrok at Jul 29, 2010 9:05:40 AM
It is not just homeowners as the vacancy rate includes investor owned homes and apartments.
If you compare housing( and multiunit) starts with the rate of household formations
they were in rough balance through the 1960s -- when babyboomers maturing generated a surge in household formations.
The level of government subsidies for housing has not changed drastically so why starting in the late 1990s did the supply of new homes start massively exceeding the fundamental demand created by household formations. Back in the 1960s-70s household formation justified housing starts of over 2 million annually. But in recent years when housing starts exceeded 2 million houshold formations only justified starts of some 1 to 1.5 million.
So if government subsidies did not change significantly why in the late 1990s did the private sector suddenly start building way too many homes.
Posted by: spencer at Jul 29, 2010 9:18:06 AM
Falling home prices give people less reason to buy, not more. Most purchases require people selling the homes they are already in and in many cases those homes are worth less than their mortgages. This holds back employment as well since people can't move as easily for work. People have become prisoners of their homes. They're under "house arrest," if you will.
Posted by: Ted Craig at Jul 29, 2010 9:20:13 AM
It seems to me that perhaps we can't accurately aggregate this picture. The two regions you mention as exceptions, California and Wisconsin, are areas I have personal knowledge of.
In California alone, you don't have an accurate aggregate picture. What happens in the central areas is impacted by the coastal zone. Prices, and occupation, in the coastal zones is so distorted by demand it is unlike any other region in our country. The central area only received a major push at the peak of demand in the bubble, and my guess is that the air in the bubble was almost completely restricted to the central areas, and that prices in the coastal zone will not have retreated much. Additionally, the coastal zones have occupancy patterns that differ significantly from the US norms. There is a lot of purchase and subsequent rental by people who only own one or two other homes. These rentals will subsequently often be occupied by the owner's relatives (children or other) when the relatives are able to do so. The home is essentially "parked" on the rental market.
Wisconsin, on the other hand, is significantly rural, and at the opposite end of the spectrum from California, with a great deal of affordable housing available.
As I said, these are anecdotal accounts - my personal experience, but I believe you will find value in them.
Recap - most California population is on the coast. Demand for housing there is so high, and building restrictions so great, that aggregate occupancy will be very sticky. Wisconsin, on the other hand, has a very large supply, and pricing is sufficiently low that demand is slow to be impacted.
Posted by: Hiero2 at Jul 29, 2010 9:24:06 AM
Shouldn't this graph also account for foreclosure rates so that it reflected how many homes there are and people that should be or were living in them (able to "afford" them)? If the aforementioned states in the South and West appear to have the excess supply of homes, did they also have the highest foreclosure rates? Had people been living in those homes, ever? Also, unemployment does not really reflect your inability to afford a home. Plenty of full-time employed folks cannot do so. A better look at "structural" vacancies should probably look more into wages.
Posted by: Anastasia at Jul 29, 2010 9:25:35 AM
@spencer
Perhaps the baby boomers of the 60's paid off their mortgages (30 yrs) now have all this extra money for investment properties, 2nd homes to retire in, added the the existing demand.
Posted by: AL at Jul 29, 2010 9:36:50 AM
"Houses have increased their preference for leisure." --matt wilbert
This works on so many levels. It's only 9:45am, but I still bet this is the funniest thing I'll hear all day (yes--I'm an econnerd). Thanks, matt!
Posted by: libert at Jul 29, 2010 9:44:43 AM
People leave the job market after a while and aren't included in the unemployment rate. Houses might hang around for decades, although they're virtually uninhabitable. What's the equivalent human term for capital depreciation?
Posted by: jj at Jul 29, 2010 9:55:02 AM
I, like eric, have trouble finding any value in the associated graphs when the national rate of one graph is ~4 times the other.
At least I read someone saying that "Houses have increased their preference for leisure." Perfect to read after rolling my eyes through supposedly serious commentary debating the efficiency of gypsy beliefs that 'the lower half of the human body is invisibly polluted'.
Given how woefully ineffective economics has proven in solving the problems it was created to address (accurately describing market conditions/cycle/stability, or making coherent economic policy recommendations)... you'd think economists might eschew the patently absurd -- efficient gypsy religions, leisure favoring unemployed, unemployed with a marginal product of 0.
Posted by: cynic at Jul 29, 2010 10:25:23 AM
@cynic: while you appear to be an educated person, I would humbly suggest that, based on the nature of your criticism, you're more than a little out of step here and do not understand what you are seeing. Your posts should probably be accompanied by a giant wooshing sound over their heads.
Posted by: BKarn at Jul 29, 2010 10:35:46 AM
In general, markets with high condo concentrations do worse. (Might explain Hawaii, Florida, and some others.)
Plus age of population may affect vacancy rate. In Florida the death of a older homeowner may leave a home vacant for a long period until it sells. However an older established community such as Michigan may have many people who still have positive equity in their homes (or even own their home clear). The rapid drop in home prices does not affect their monthly housing costs - unless they need to move to seek employment. You just see gradual disinvestment in the housing stock
Low interest rates spurred people to buy housing, the monthly payment was lower. The increased demand for housing increased prices. People saw the increased value of their homes and reacted by taking out home equity loans or buying up.
The acceleration in home values made foe easy profits for many, at the start. Lenders caught the same fever. A spike in oil prices leads the Fed to talk about the need for higher interests rates - at the same time a record number of ARM are resetting. Home prices would have fallen just from the increase in interest rates and a slowing economy. However structural problems in the auto industry, related to the oil spike and long term problems in the industry increase unemployment in the midwest. California is suffering from a failed political structure. Arizona, Florida, and Nevada have seen wild speculation in real estate at the same time oil prices have reduced tourism spending. The Carolinas are less business friendly and suffer from international competition. Politicians promising an activist government with increased government spending and higher taxes take a strong lead in public polls.
The Federal government start screaming fire but don't seem to have the ability or knowledge to fight the fire in the financial sector. Billions are still being spent on a war strategy without exit. Deferred taxes in the form of growing debt slows recovery. etc
Posted by: DanC at Jul 29, 2010 10:46:52 AM
curious about what part you find so clueless? paragraph one just reiterates that the two graphs in the post have incompatible data. two is just sharing amusement at the previous comments poking fun at employment of resources vs preference of leisure.
I guess it's paragraph three you don't like, where I tangent off into criticism of economics in general. I'm not sure whether you object to my general criticisms of economics not doing its job lately (failure to anticipate the recent crisis, lately failure to agree on any policy prescription, re fiscal stimulus, lately austerity, and so on)... or whether you just object to the specific economic explanations I find ridiculous...
Posted by: cynic at Jul 29, 2010 10:53:38 AM
Maybe we should fill these houses with sand.
Posted by: Andrew at Jul 29, 2010 10:56:27 AM
I think you have 3 trends here:
1) Supply not keeping up with demand due to migration (TX)
2) States that have never been able to draw building dollars even though there is demand. (LA, AK)
3) States that have little desirable room to build. (RI, NJ, NH)
I think #2 or #3 accounts for most of the bottom of the graph. Who puts investment dollars in Lousiana or Alaska no matter what the demand might be?
Posted by: Tim at Jul 29, 2010 11:01:35 AM
The rental housing vacancy rate is 9.9%, the homeowner housing rate is 2.8%.
http://www.census.gov/hhes/www/housing/hvs/qtr308/files/q308press.pdf
If you have an apartment building, you can't let a new tenant in at 20% less than everybody else. Sticky.
Posted by: TomHynes at Jul 29, 2010 11:02:59 AM
And fog....
Posted by: Bernard Guerrero at Jul 29, 2010 11:03:27 AM
It's interesting to see New Jersey as an outlier, since I think of New Jersey as having quite high property tax rates.
I have heard of the British Empire in Africa instituting taxes in order to force subdued populations to work. Is it possible that taxes can be a spur to productivity at the margin, given certain circumstances? Does this work for property taxes and real estate as well? Could it be that higher property taxes discourage unproductive property? There is some evidence for this in the Japanese real estate bubble of the 1980s (where near zero property tax rates meant that land speculation was lower risk and more profitable, leading to more speculation and a bubble).
Posted by: dolo at Jul 29, 2010 11:07:24 AM
I think the chart is crap. This isn't really something that translates well at the national view. While there is a national slump in housing after the bubble burst, each region or state or even metro area is going to have their own set of issues and reasons for problems. A national economic boom hid each states' problems, and the bust now brings them to light. (Sort of like an economic boom hides corporate problems, and a bust brings them to light.)
Arizona, Florida and Nevada seem to be skewing things nationally. They had very big bubbles in their housing markets that could have lead to too, too much over-development. They are victims of over-speculation.
All of the New England states are low on the housing vacancy level, but I believe that's because of the difficulty of building new houses in New England. (Zoning restriction, septic system restrictions, etc., etc.) But slow building also means much higher prices. Expect more home-owners to be staying because they are underwater on their mortgages.
Virginia, Maryland and Georgia would be my main focus. All three fall right on the national rate for home vacancies. All three are the poster children for suburban sprawl. (Atlanta and DC Metro areas, I'm looking at you).
I would have thought that Michigan, with all of the stories of the abandonment of Detroit, would have been much worse.
Posted by: Xmas at Jul 29, 2010 11:15:05 AM
The correlation looks pretty darn good to me considering how aggregated it is. But then I'm not an economist, although I usually don't brag.
Posted by: Andrew at Jul 29, 2010 11:25:53 AM
While it says "non-seasonal," I wonder how effectively the statistics take into account condos and houses that are rented in vacation/resort areas that can be utilized all year, but are not now because of people travelling less. I especially was interested in Colorado and Hawaii, which are year-round vacation spots, and seem to have disproportionately more vacancies than their unemployment rate would indicate.
Posted by: Joe Y at Jul 29, 2010 11:58:32 AM
I'd hazard a guess that we've built too many houses that aren't useful as primary residences for working age people (mostly due to poor location). It might take very large price declines to induce people to accept awful commutes or to get second-home buyers back into the market. Some subdivisions in weird locations have simply passed their tipping point from vacancies and have become ghost towns where few would venture to live at almost any price.
Posted by: Eric at Jul 29, 2010 11:58:48 AM
House prices are still way too high. People think that because they've dropped 30% that they're cheap, but they forget that for much of the nation, houses more than doubled in price.
Another 20-30% from here will make them more palatable. But by then, nobody will have a job. And there's the whole rising property tax rates in many areas to 'offset the decline in home prices'.
Posted by: Eddie at Jul 29, 2010 12:27:49 PM
Liquidity (or lackthereof) seems to be a big factor. It makes sense that a lot of houses would be vacant if a lot more potential buyers (who are often homeowners) are underwater on their current mortgages. Even as prices get lower, these potential home-buyers are even more underwater they are with their own mortgage.
Houses can also be seen as "toxic assets," where even if prices are low, liquidity preferences (for both current homeowners and non-homeowners) prevent people from buying them. We have already seen this in financial markets. I think it could be a plausible explanation too.
Posted by: Scott Wentland at Jul 29, 2010 12:48:13 PM
What exactly does it mean for vacancy rate to go down? Where do the former occupants go? Some leave the country (such as foreigners losing their jobs), some consolidate households (such as moving in with parents, or roommates). What other changes can contribute to the vacancy rate? And which of these are the biggest contributors?
Posted by: maximizer at Jul 29, 2010 1:21:41 PM
Correct me if this got captured somewhere I didn't see, but wouldn't another part of the increase be due to increased housing turnover rates due to a much more mobile population? Specifically, when I look at my parent's (and grandparent's) generations they tended not to move much, either within or between geographic regions. I don't have data for this, but am assuming that the average number of years a given occupant spends in a house has been consistently going down over time. This would mean that any individual house is being sold more and more often over time. Even if the average amount of time a given home stays vacant at time of sale is constant, this effect alone could greatly increase vacancy rates.
Posted by: Dave at Jul 29, 2010 1:26:20 PM
Al -- that is exactly what many of the builders expected -- baby boomers buying the houses as second or retirement homes.
That is one of the main reasons they built such expensive homes and not homes better suited to what younger less affluent first time buyers needed.
Posted by: spencer at Jul 29, 2010 1:31:40 PM
All seem partially responsible. Structurally it is about energy prices making outlying areas uneconomic to commute from. The south is still growing long term so the excess will be absorbed over time, but for now many are in limbo or lenders hands that are slow to liquidate and won't rent, until speculators will take them over and hold onto many of these and rent them until the market tightens, so the stickiness may not about price so much as performance. AD is the largest culprit though since renewed lending standards have shown the income to afford these doesn't exist, consumers are overindebted, and unemployment is very high.
Posted by: Lord at Jul 29, 2010 1:37:17 PM
One reason California doesn't have a high vacancy rate like the other Sand States, Arizona, Florida, and Nevada, is because a lot of the laid-off illegal immigrants who constructed the surplus housing in those states have moved back to California. It would be cheaper for them to move back to Mexico, but then they might miss out on the Amnesty that Obama keeps promising.
Posted by: Steve Sailer at Jul 29, 2010 5:02:57 PM
This may be a factor: it takes a long time to get kicked out of your house in California during a foreclosure. I don't have figures in front of me, but I remember reading that it takes about a year to get booted out in CA, compared to a month in TX, for instance.
Posted by: Dave at Jul 29, 2010 6:30:31 PM
"The homeownership rate was 67.2 in 2000 and today it's 66.9"
Are vacant houses included in these numbers? After all, they are owned by people "who own their own home" as opposed to rented out by investors.
Furthermore, how would the Census Bureau (the source of the data in first chart) would know anything abaout vacant houses? Especially for 2010? There was no such question on the form...
Posted by: Albert Farangh at Jul 29, 2010 11:13:10 PM
Speaking of how long it takes to get kicked out in California, have any of you seen "Pacific Heights" featuring Michael Keaton as the villainous tenant ruining the dreams of innocent yuppies? Don't, it's a bad movie.
Posted by: TGGP at Jul 30, 2010 12:37:41 AM
Boy, it seems everyone thinks housing unit equals house!
From http://www.census.gov/hhes/www/housing/hvs/qtr210/q210def.html
Housing Unit. A housing unit is a house, an apartment, a group of rooms, or a single room occupied or intended for occupancy as separate living quarters. Se....
Homeowner Vacancy Rate. The homeowner vacancy rate is the proportion of the homeowner inventory which is vacant for sale....
The report just released states:
National vacancy rates in the second quarter 2010 were 10.6 percent for rental housing and 2.5 percent for homeowner housing, the Department of Commerce’s Census Bureau announced today.
Now as to the reason for the high rental vacancy rate; that was covered by Marketplace earlier this week, as I recall.
Unemployed and underemployed people have moved from their underwater homes or from college dorm or from their own apartment, to their parents, a crashed at friends, shared rent on an apartment or house.
In other words, the amount of money those who needed housing was too far below the lowest price on the supply curve that no landlord could cut the price low enough. If all landlords cut prices 10%, it would still be too high for those who can only afford 50%. Maybe if most cut prices to 70%, those who were doing ok with work might be able to afford to move out.
But for the people who have no income and who have moved back home, the rental market price would need to fall enough so the parents would decide to pay their kid's rent to get him out of their kitchen, which might be $200-300 a month if dad likes to keep beer in the fridge.
I find it odd that economists ignore everything they should have learned in 101 microeconomics.
And as for homeowner vacancy rates, those have fallen as prices have fallen, from 2.9% at the peak in 2008, down to 2.5% today. That is well above the more normal 1.7%.
Rental vacancy peaked a year ago at 11.1% and has fallen to 10.6%, reflecting the improvement in employment. But rental vacancies have run higher since the second year of the Bush term over 9%, relative to 8% in the late 90s.
Posted by: mulp at Jul 30, 2010 4:00:55 AM
I don't think houses should be thought of as a resource - apart from when people work from home, they don't contribute towards production. They are a hybrid good/asset.
And I don't quite understand what point is being made about the correlation between unemployment and vacancies? Alex says that 50 per cent of the vacancy rate is caused by a lack of AD - so that will also be creating the unemployment. The relationship could,I guess, me used to measure how much of the vacancy/unemployment is down to a lack of AD and how much is down to structural factors/sticky prices. Outlier observations would suggest that structural/price factors are particularly prevalent
Posted by: Toby Kenward at Jul 30, 2010 5:23:46 AM
This unemployment is the result of poor home policies that are taking its toll on the people. It is the moral duty of the govt. to provide job opportunities to its citizens.
Posted by: Revolution E-Cigarette at Jul 30, 2010 6:01:18 AM
Furthermore, how would the Census Bureau (the source of the data in first chart) would know anything abaout vacant houses? Especially for 2010? There was no such question on the form...
The Census Bureau has cut the number of questions and eliminated the "long form" by moving all the detailed questions to quarterly samplings. (The decennial survey did ask if the housing unit you live in was: owned by you, owned by someone else in the household, rented, or homeless, which helps calibrate the quarterly sampling. The vacant housing data falls out as a consequence of finding no one living in what appears to be a housing unit.)
Posted by: mulp at Jul 30, 2010 11:59:22 AM
Bubbles attract excesses driven by speculation.
The housing bubble was created, like so many bubbles, by politics, not by natural economic forces.
Housing is disjointed from from what it should be related to in a free market? What a surprise.
Posted by: Cowboy at Jul 31, 2010 11:44:48 AM
Actually, the graph almost showcases the biggest suburban growth areas since the 1960's. I'd argue there is more to the story as it relates to over-extended infrastructure and higher gas prices as they relate to commuting times.
Posted by: JP at Jul 31, 2010 9:10:11 PM
That makes comparative analysis of different minority populations difficult - if there are differences in life expectancies between Romani and Uighurs is that due to their social organization or due to the different political situations that they face? You can't really control for such a thing in a comparison.
Posted by: Converse all star at Aug 2, 2010 5:43:18 AM
I agree with most of the views posted here that unemployment is growing due to the poor policies of the govt. If it is not controlled in time, then the govt. will have to face its dire consequences.
Posted by: Pro Cleanse Gold at Aug 2, 2010 6:35:29 AM
Also, Christians of today have always struck me as one of the religions least likely to take offense of criticism, mockery, or denunciation.
Posted by: rolex submariner at Aug 4, 2010 5:01:53 AM
I agree with most of the views posted here that unemployment is growing due to the poor policies of the govt. If it is not controlled in time, then the govt. will have to face its dire consequences.
Posted by: coach handbags at Aug 4, 2010 10:51:38 PM
Also, Christians of today have always struck me as one of the religions least likely to take offense of criticism, mockery, or denunciation.
Posted by: mbt at Aug 4, 2010 10:52:41 PM
The thing that is missing in govt. policies is that proper attention has not been paid to the youth problems these days. They need to be given full attention.
Posted by: Xtreme NO at Aug 9, 2010 7:19:04 AM
i think it ultimately starts at banks making mistakes and people fooling others for profit. originally, bank would give loans to people who were economically unstable and when banks would lose money if the people couldn't pay back on the loans.
Posted by: Acai Optimum at Aug 12, 2010 4:30:36 AM
