frontlinethoughts.com
The Real Culprit in High Home Prices
Let's begin with a brand new study done for the National Bureau of Economic
Research (NBER) by Edward Glaeser. (You can see this fascinating study at
nber.org
In a normal economic world, the greater the demand for something the more
entrepreneurs who work to supply it. The higher initial demand spurs higher
prices which increase the number of businesses trying to make the product
which eventually causes a glut on the market and prices fall. It's basic
supply and demand.
The same should hold true for housing. If there is high demand, you would
think entrepreneurs would be building more homes. And it does happen in
many places, but not in all locales. Not surprisingly, the areas where
normal supply-demand does not exist are also the places of high prices.
"They aren't making any more land," our parents told us. Which is true. But
Glaeser suggests available land is not the prime culprit. Let's look at his
paper.
"...evidence on construction suggests that demand alone cannot provide the
answer [for higher housing prices]. For example, in Manhattan, before 1975,
housing price growth was modest, and there was abundant new construction.
Since 1980, housing prices have soared and there have been few new units.
The physical character of Manhattan has not changed between 1960 and today.
If the rise in housing prices during the 1990s were the result of demand
pushing along a stable supply curve, then surely we would see an explosion
in new construction as we did in the past. The increasingly common
combination of rising prices and tiny amounts of construction pushes us to
focus on housing supply.
"Differences across regions confirm the need for supply-side analysis. High
housing prices are not ubiquitous. The median housing value in the median
county in America in the 2000 census is $75,300. More than 95 percent of
counties have median housing values below $160,000. Soaring home prices are
primarily coastal phenomena that have left the growing states of the
American interior untouched.
"If the heterogeneity [def: - the quality of being diverse and not
comparable in kind] in price growth with the United States were the result
of different patterns of demand, then we would expect to see quantities and
prices move together. Places with high price growth would be places with
new construction."
Glaeser then shows us a very interesting graph. He compares the rise in
housing prices with the rise in new housing starts. In an Alice in
Wonderland economic world, what we find is that there is a 50% negative
correlation between housing prices and home construction. In general, the
areas with the highest prices have the least amount of new home
construction, which is the opposite of what they taught us should happen in
Economics 101.
"The places that are building have little housing price appreciation and
the places that have housing price appreciation are not building. Demand
alone can't explain the difference in housing price growth between New
England and the South Atlantic. Florida alone permitted almost as many
homes in 2002 as all of New England did over the entire five-year period.
If we want to understand why housing is so expensive, then we must
understand why housing supply in New England, the Middle Atlantic States,
and California has become so inelastic. Housing supply research is also
necessary because regional growth rates depend on the rate at which these
regions build homes."
Housing costs are roughly $80 per square foot and a 2,000 square foot home
can be bought for $160,000 in much of the US. That is because in most of
the US, cheap land is available. Raw land costs for most of America's homes
are not all that much.
(Although I am always amazed at how developers can turn raw farmland into
"exclusive" home lots costing 30 times as much as the raw land surrounding
the development. Only in America.)
Now they make a very important point as an aside to their paper.
"Of course, even where housing supply is perfectly elastic with respect to
positive shocks, housing supply is inelastic with respect to sufficiently
negative shocks. Because housing is fixed and durable, a major drop in
housing demand can always cause prices to fall. This explains why cities
decline so slowly and why declines show up in falling housing prices long
before they show up in falling population levels. Indeed, the growth in
housing prices in New England has been so spectacular in part because 20
years ago New England was declining and housing cost less than the physical
costs of replacing the buildings."
(By the term "elastic" they mean "able to adjust readily to different
conditions." Positive shocks, such as a large and quick rise in local
employment should ultimately produce a rise in supply in homes and would be
"perfectly elastic." Falling demand would not reduce the supply of homes,
thus home supply is "inelastic" in a negative shock such as a large drop in
employment at the local factory. In short, you do not destroy existing
homes when demand drops in order to maintain value of the local housing
stock.)
Next week, and in our conclusion, we are going to visit this theme again
and again. It is the "major drop in housing demand" and not last year's
prices, whether regionally or historically high or not, that may be the
driver for falling home prices. Can we predict falling demand and thus
lower prices? We will see, gentle reader. But first, let's look at some
more data.
So, what's the problem with areas which have so little new home
construction? Is it lack of land or possible (drum roll, please) government
regulation?
You can read their work for yourself, but their study shows that the
availability of land is not the driving factor in high cost areas. If local
laws limit new construction, then higher prices result as an area grows.
Even in areas where there is new construction, it might not be keeping up
with demand. California grew by 600,000 people last year, and 500,000 per
year for the last four years. 2,000,000 new people mean around 800,000 new
residences (homes and apartments) are needed just to keep up with
population growth.
Think about the above mentioned fact: Florida permitted as many new homes
in one year as all of New England did in five years. Is it any wonder that
home prices in New England soared well above the national average?
(Another very good study, for those who are interested, is a paper done on
regulation and the effect on affordable housing last year for the Federal
Reserve Bank of New York by C. Tsuriel Somerville and Christopher J. Mayer
at heartland.org
The Demand for Housing Will Grow
But with the Baby Boomer generation retiring, some argue we will need fewer
new homes. But the data does not support that conclusion. The Joint center
for Housing Studies at Harvard University issued a lengthy research paper
on the state of US Housing.
(http://www.jchs.harvard.edu/publications/markets/son2004.pdf)
"Residential construction has been on the rise for most of the last 12
years, adding significantly to the housing stocks in both metropolitan and
non-metropolitan areas-particularly in the South and West. While large
metros such as Atlanta and Las Vegas have seen spectacular increases in new
construction, the rate of growth in many small and medium metros exceeded
25 percent.
"Despite growing concern over the pace of development, housing construction
over the next 10 years is likely to exceed that over the last 10. The
Census Bureau's newly revised population estimates imply that household
growth from 2005-15 will be as much as 1.1-2.0 million more than the Joint
Center for Housing Studies previously projected. Add to that the growing
demand for second homes and replacements of units lost from the stock, and
the total number of homes built in 2005-15 could reach 18.5-19.5 million
units. This compares with 16.4 million homes added in the 1990s."
Part of this is the growth in immigration into the US. When looking at the
natural growth of minority households coupled with new immigration, the
minority percentage of households has risen from 17% in 1980 to about 26%
today and is expected to rise to 34% by 2020.
Without the rapid growth of minorities, the renter portion of the nation
would have decreased over the last decades. "Instead, the renter population
increased modestly and the minority share of renter households surged from
31 percent to 39 percent over the decade. In addition to keeping rental
demand from sagging, minorities also accounted for fully two out of every
five net new homeowners from 1994 to 2003. Despite these strong gains,
though, minority homeownership rates still lag those of whites by nearly 25
percentage points."
Long-term, then, the total aggregate demand for housing -whether owned or
rented - is going to rise. So, then can we conclude that housing prices are
also going to rise as well solely because of a growing population? The
short answer is no. We can point to several periods of rapidly growing
population where housing prices did not rise other than due to inflation.
There are other factors we need to look at.
The Housing Price to Earnings Ratio
One of my diligent readers (John R.) sent me a study he did, comparing the
median per capita income and average housing prices since 1965. He then
looked at returns for the next five and ten years.
It was his effort to find a type of Price (home price) to Earnings (income)
Ratio for housing. While there was no straight line correlation, overall
you can make some very interesting observations. You are better off buying
in periods of low "P/E" ratios than high ones. While his home-made housing
P/E ratio was in the mid-range only 5 years ago, today it is at the top.
Looking at historic norms, one would expect much lower returns, perhaps
even negative returns, in the next five years. You simply cannot get too
far from the mean before the gravity of the regression to the mean takes
hold and pulls price increases down.
I would have rather seen household income than per capita income, but I
doubt the actual conclusions would be different. Buying "value" is always a
good idea.
Income, Employment and Home Prices
Is there a correlation between income, employment and home prices? The
Harvard study group did another study of US housing in 2003, focusing on
income and housing prices, and they think they have found one.
"While surging home prices have sparked fears of a housing market collapse,
widespread price declines are unlikely because home prices in most areas
have increased in line with income growth. History demonstrates that few
localities experience the kind of concentrated job losses that precipitate
severe home price deflation. Over the past 15 years, fully 53 of the 100
largest metropolitan areas have not experienced a single year of declining
nominal home prices. Most areas, in contrast, have experienced slow
deflation in real home prices following prolonged run-ups. In fact, real
prices in 58 of the 100 largest metros have fallen ten percent or more at
least once since 1987. In about a third of these locations, however, the
real home values of owners who bought at least two years before the peak
exceeded their purchase prices even at the bottom of the real declines.
They do not expect a dramatic decline in home prices. "Furthermore, despite
the rise in both delinquencies and foreclosures in 2001 and 2002, problem
loans represent a very small fraction of active mortgages. Serious
delinquency rates on conventional loans are still under one-half of one
percent. Although delinquencies on government backed loans are higher, new
ways of working with borrowers promise to reduce the share of problem loans
that end in foreclosure."
As a side note, we have seen the number of Americans who are behind on
their mortgage payments begin to drop this year. The number of homes in
foreclosure fell to 1.27%, down from 1.43% this time last year. "The
delinquency rate for prime mortgages, those made to the most creditworthy
borrowers, fell to 2.26 percent from 2.40 percent in the fourth quarter.
Subprime-loan delinquencies also fell, to 11.19 percent from 11.59 percent;
FHA late loans fell to 11.68 percent from 12.23 percent and VA loan
delinquencies dropped to 7.37 percent from 7.99 percent." (CBS Marketwatch)
Harvard also has the following more sobering thoughts: "Despite this
remarkable buoyancy, housing market risks have intensified in recent years.
Job losses have forced more mortgage borrowers into foreclosure, increased
the number of homeowners spending half or more of their incomes on housing,
and softened some rental markets. In addition, expansion of mortgage credit
to borrowers with past payment problems has elevated foreclosure risks.
"Finally, increased mortgage debt levels and growing shares of homebuyers
with high loan-to-value ratios have raised concerns about the amount of
debt carried.
"Two risks, however, could lead to more serious problems if the recovery
stalls. One is the growing number of loans to borrowers with weak credit
histories. Though serving many borrowers who just ten years ago were denied
access to credit, default rates on these loans are predictably higher than
on standard loans. Because these loans are highly concentrated in low-
income, primarily minority communities, a wave of foreclosures could put a
glut of homes on the market, lowering prices and threatening the stability
of entire neighborhoods.
"The second serious risk factor is the dramatic jump in the number of
homeowners spending more than half their incomes on housing. About 7.3
million homeowners reported cost burdens of this severity in 2001, up from
5.8 million in 1997. If more job losses occur, some homeowners in the
bottom two income quintiles-who make up 84 percent of these vulnerable
homeowners-could ultimately face the loss of their homes."
Notwithstanding their note about risk factors, there is a fly in the
ointment of their logic. They assume home prices rise with income. And it
is reasonable to think that as incomes rise 20%, people will be willing to
pay 20% more for a home. But why would they pay 20% more for the same home?
Other than inflation, why is rising income a driver for housing prices? As
Dean Baker of CEPR notes:
"The Home Price Index [HPI, which the Harvard study uses] tracks the resale
prices of the same homes. In other words, if the HPI rises by 5 percent in
a year, it means that, on average, every individual home has increased in
price by 5 percent compared with the price it sold for last year. There is
absolutely no reason whatsoever why it should be expected that individual
home prices would rise in step with family income - and it has never
happened before during a period when family income was rising.
"Unfortunately, the HPI only goes back to 1975, but prior to 1981, the
homeownership component of the consumer price index was constructed in a
manner that is similar to the current HPI, tracking the resale prices of
the same house. In the years from 1955 to 1973, the shelter index of the
CPI (which includes the homeownership component in addition to a rent
component) rose by 77.9 percent, or an average of 3.3 percent annually. By
contrast, median family income rose by 173 percent over this period, or at
an average annual rate of 5.7 percent."
Looking at a graph on page 8 of the 2004 Harvard study, you can easily see
that there have been "bubbles" in certain areas of the US housing markets,
not in the past few years, but in the 80's. They were all followed by
serious corrections. The recent annual run-ups have been tame by
comparison. What makes the recent experience so different is that the
growth in certain areas (the Pacific, the Northeast and the mid-Atlantic)
has been so steady over long periods. In the earlier decades, you would get
real spikes in housing prices.
The real link between income and housing prices? Employment. In the 80's we
had what was called a rolling recession in the US. Each region seemed to
experience its own recession while other parts of the country were booming.
Looking at that graph of housing prices, you could see which area was
suffering its own boom or recession.
The Real Growth in Housing
Finally, these quick facts, which illustrates the powerful fascination of
housing on Americans.
In 1950, the average new house was 983 square feet and cost $11,000. In
2000, the average new house was 2, 265 feet and cost $205,000. In 1950,
there were 3.37 people per household, and now there is but 2.6. In 1950,
only 6% of homes had a two-car garage. In 2000, 65% had two-car garages and
17% had three (or more) garage spaces.
In 1950, the average cost per square foot was $11. Today it is $91. Much of
that is inflation, as inflation alone would increase prices to $76. The
actual value of a square foot today is far more however. I grew up in one
of those 1950 homes. No air-conditioning, one bathroom, rudimentary
appliances, heating was a space heater in the main room. And three young
kids and two bedrooms. I truly enjoyed my youth, but I am not nostalgic for
the old homestead.
Things got better (and somewhat bigger) over time, not only for my family
but for much of America. We will go into the psychological factors in
housing prices next week, but homes play a central part in the American
psyche. We will also look at affordability.
What have we learned this week? The core thoughts are that in the short-
term home prices are more subject to regulation (which reduces supply) and
employment. A down-turn in employment is going to affect housing values,
perhaps significantly. In prior recessions, employment dropped along with
housing prices. There were no parts of the country immune to a drop on
housing prices. In the recent recession, we did not see a real spike (in
historical terms) in unemployment or drop in overall consumer demand, and
thus the negative affect on housing prices was muted. This was a true
historical anomaly.
Over the long-term, inflation and a growing population will kick in to
increase nominal home prices, but the long term may be a decade or more.
But that's a story for later, as this letter is long enough. |
