4 March 2005
Growth-management planning leads to rapid, artificial increases in the price of housing that can turn home buyers into speculators. Such speculation leads to housing bubbles that, when they burst, can impose huge costs on individual homeowners and the economy as a whole.
Such bubbles certainly exist in San Jose, Boston and other California and Northeast housing markets as well as those in a few other regions. Most regions suffering bubbles have used various forms of growth-management planning, while most regions that don't use growth-management planning appear safe unless there are other constraints on the land supply. While bubbles in Portland, Denver, and other smart-growth regions are not as severe as in San Jose or Boston, planners should be far more wary than they are of the likelihood that their policies will create such bubbles.
Real estate analysts are hotly debating whether or not the U.S. housing market is in a bubble. A bubble would mean people are paying unrealistically high prices and a fall in those prices could devastate our economy.
At one extreme is John Talbott, an investment banker whose 2004 book, The Coming Crash in the Housing Market, predicts that nationwide prices will fall by 20 percent or more. At the other extreme are investors such as Ed Wachenheim, who argues that housing is different from other markets and so won't suffer a huge decline in value.
In the middle are numerous analysts who think there is no national bubble but that prices may fall in "some communities." Home prices are too high, says one writer, in only about twenty urban areas, mostly located in eight states. "The bad news is that those areas contain roughly half the housing wealth of the country."
The regions that are most frequently mentioned as having overpriced housing markets -- such as San Jose and Boston -- are the ones that practice the strongest forms of smart growth or other growth management. These artificial constraints prevent builders from supplying the demand for housing, leading to rapid price increases and an expectation on the part of home buyers that such increases will continue.
Bubbles arise when prices are based more on speculation that such prices will go up than on the intrinsic value of goods. If speculators think that prices will continue to rise in the future, they will be willing to pay more for something today than it is worth based on today's values.
In a normal market, for example, the monthly mortgage on a home should not be much more than the cost of renting a similar home. But in many housing markets today, rents won't cover mortgages; in such cases, the Economist magazine suggests that people should rent rather than buy). But if people think that prices are rapidly rising, they will pay the mortgage premium figuring that they will make up for the extra costs of buying when they sell the house.
This is what happened to dot-com stocks in the 1990s. By standard measures of value, most dot-com companies were not worth a lot -- very few earned a profit. But their stock prices increased anyway, and people bought the stocks figuring they would increase forever. Anyone who mentioned "bubble" was told that the "new economy" was different. Of course, the bubble burst leaving many dot-com ex-millionaires.
Claims that there is little risk of falling prices because housing is "different" from other assets sound ominously similar to assurances given before the dot-com bubble burst. Supposedly, housing is special because people own their homes for many years, so any declines in value will be small.
In fact, during the early 1990s, housing prices in many Northeastern and California regions dropped by 20 percent or more. The premiums being paid by speculators for housing in those same markets today are much higher than they were before the early 1990s decline, so the next housing crash could be even worse.
Ironically, disillusionment with the stock market has contributed to the housing bubble as people put their money in real estate rather than other assets. A recent report by the National Association of Realtors found that more than a third of the homes sold last year were purchased by investors or for second homes, not for primary residences. This "astonished" the Realtor Association's chief economist as it suggests that speculation is much more significant than previously thought.
Another sign that buying is speculative is the fact that nearly a third of mortgages today have adjustable interest rates. Buyers ought to prefer fixed rates when rates are low and likely to rise -- unless they expect to sell soon. So much for housing being different because people will hold onto homes for many years.
Buyers are encouraged to speculate by increasingly easy money. Another National Association of Realtors study revealed that 42 percent of first-time homebuyers were putting zero percent down on their houses, and 69 percent were putting less than 10 percent down. Mortgage companies are also offering loans that require buyers to pay only the interest and even to make some payments that are less than the interest on the loan.
If any place is suffering a bubble, it is San Jose. Between 1995 and 2000, the dot-com boom doubled San Jose's already high housing prices. Then came the crash, and between 2000 and 2003 San Jose lost 190,500 jobs -- more than 18 percent of its workforce. Yet, sustained by low interest rates, housing prices in the same period rose another 6 percent. Despite continuing high unemployment, in the last year they have risen at least 12 percent more and realtors call it "a seller's market."
The National Association of Realtors says that the highest home prices in America are found in the San Francisco Bay Area, including San Jose. While the median price of homes nationwide was $184,000 in 2004, the median Bay Area home sold for more than $640,000. The Bay Area market is so hot that most homes sell for more, not less, than the price for which they are offered on the market. One weakness in sale price data is that it does not account for differences in home size and quality in various regions.
Coldwell Banker provides a bit more detail as it annually estimates the value of a standard "middle-manager" home: four bedrooms, two baths, two-car garage, 2,200-square feet, one fireplace, etc., in more than 300 cities. According to the realtor's latest report, such a home in San Jose currently sells for around $950,000. This is an unrealistic 50-percent jump from Coldwell Banker's 2003 price of $620,000. Yet it may not be too far off, as real estate listings on Realtor.com indicate the median-priced four-bedroom, two-bath home on the market in San Jose today is being offered for $800,000.
Even the 2003 price is nearly twice the national average and four times as much as someone would pay for the middle-manager home in such places as Ft. Worth, Knoxville, and Billings. Fourteen of the twenty most expensive cities on Coldwell Banker's list are in California and all but one of the rest are in the Northeast (the exception is in Hawaii).
San Jose homes typically sell for $300 to $400 a square foot, though some, especially newer homes, can go for $500 or more. This compares with a nationwide median cost of around $150 a square foot and a cost of less than $100 a square foot in such economically healthy regions as Raleigh, San Antonio, Nashville, and Tucson.
While some of Coldwell Banker's numbers may be a bit uncertain, the bureaucratically-titled Office of Federal Housing Enterprise Oversight uses repeat sales of identical homes to meticulously track the change in price of homes in 265 metropolitan areas. This index doesn't use actual prices, only the change in prices. But a graph of this index for selected regions shows that San Jose and Boston prices have far outpaced the rest of the country (http://americandreamcoalition.org/hpi.jpg).
A thirty-year mortgage on an $800,000 home would require payments of $4,300 a month. By comparison, San Jose newcomers can rent a four-bedroom, two-bath home in San Jose for $1,800 to $2,500 a month. In such a market, the only reason to buy rather than rent is an expectation that buying is a "good investment," that is, that housing prices will continue to rise.
RealFacts, a group that does research on the multi-family housing market, notes that home sale prices in the San Francisco Bay Area, including San Jose, grew by 15 percent last year while rental rates actually declined. In this, the Bay Area is not unusual: Home prices in fifteen western housing markets reviewed by RealFacts increased by an average of 23 percent last year, while rental rates increased by only 2 percent. "Clearly, there is a fundamental disconnect in real estate markets between home prices and apartment rents," says the research firm.
San Jose workers earn about 50 percent more than the national average, but this is not enough to make up for the region's high housing costs. San Jose housing prices are about eight times median incomes. By comparison, housing prices in many Midwestern regions are only two to three times median incomes.
As a result of their high home prices and the fact that they had the highest unemployment rates in the nation, in early 2004 the PMI Mortgage Insurance company ranked San Jose and Portland as the two riskiest housing markets. But in mid-2004 PMI added a measure of housing affordability to the way it calculates risk. Since Portland housing prices are still well below those in San Jose, PMI now estimates that the two riskiest markets are Boston and San Jose. In fact, PMI says it is more likely than not that housing prices in these two regions will decline in the next two years.
The Ten Riskiest Housing Markets: Likelihood that Prices Will Decline In Next Two Years Metropolitan Area Probability Boston 53.3% San Jose 53.0% San Francisco-Oakland 47.9% San Diego 43.3% Providence 39.7% Sacramento 36.9% New York 36.3% Los Angeles 35.9% Riverside-San Bernardino 31.6% Detroit 27.4% Source: PMI Mortgage Insurance
As the table shows, most of the riskiest housing markets are in California or Massachusetts (the Providence metropolitan area laps into Massachusetts). These regions are famous for their growth-management planning. Other risky housing markets on PMI's list include the Twin Cities, Denver, and, of course, Portland.
Among PMI's least risky regions (less than a 10-percent chance that prices will fall) are Las Vegas, San Antonio, Memphis, and Oklahoma City. These are all reasonably fast-growing regions that practice little if any growth management. While the correlation between growth management and risk is not perfect, it is very high.
Of course, realtors are often quick to deny that prices in their markets are too high. "I get calls every day from people looking to buy," says one Boston realtor in response to PMI's latest ranking. Considering that Coldwell Banker's standard house sells for more than a million dollars in Boston, these could turn out to be famous last words.
"Buyers are not price sensitive because they are leveraged and playing with other people's money," says John Talbott. The zero-down-payment loans now being offered by mortgage companies promote even more speculation. Nor do banks "care about prices paid because they don't hold the mortgages they create," Talbott adds, since they generally sell them "upstream" to other mortgage companies.
Of course, these things are true everywhere. What makes places such as San Jose special are the constraints on home builders imposed by urban-growth boundaries and other land-use regulations.
Any expectation that San Jose prices will never fall seems utterly unrealistic in the face of San Jose's high rates of unemployment and office vacancies. Between 2000 and 2003, commercial vacancies increased from less than 3 percent to 30 percent and office rents fell by half (see table 2 of this report on "the meltdown" of the Bay Area economy). Recently, I visited entire San Jose industrial parks and downtown office buildings that are completely or almost completely vacant.
This is because employers are more sensitive to housing prices than home buyers. Employers have to pay their employees enough to cover their costs of housing. All else being equal, if you are considering locating an office or factory in one town where your employees will have to pay $800,000 for a house vs. another where they can get the same house for $200,000 or $300,000, which would you pick? The answer helps explain why the populations of such high-tech regions as Atlanta, Austin, and Raleigh all grew around 5 percent per year in the 1990s while San Jose's population grew by less than 1 percent per year. As Joel Kotkin says, major cities are losing their traditional advantage over smaller towns and less populated regions.
Coldwell Banker's numbers suggest that seven of the nation's ten most expensive housing markets are in California cities that have urban-growth boundaries or other forms of growth management planning. Five of the ten most expensive markets on the National Association of Realtors' list are in California. Both lists place the highest housing prices on the coasts, with housing in the Midwest and South remaining very affordable. Other expensive housing markets on both lists are in Boston, Denver, and Seattle, all cities that practice various forms of growth management. Coldwell-Banker's list also includes Canadian cities, and the most expensive is Vancouver, that nation's smart-growth leader.
This graph of house price index for three managed-growth cities (Boston, Los Angeles, and San Jose) and three relatively unregulated cities (Atlanta, Austin, and Las Vegas) suggests that growth management can make prices more volatile, both up and down.
Growth management is not the only cause of high housing prices. Despite being the fastest-growing major urban area in America for well over a decade, Las Vegas enjoyed very affordable housing through 2002. But the last two years have seen a rapid increase in prices for the simple reason that the region has run out of private land. According to one report, Las Vegas home prices leaped by more than 50 percent in the last year alone. In past years, the federal government has sold land to developers to keep up with housing demand, but land sales have declined due to environmental concerns. Of course, this is simply another form of an urban-growth boundary.
The Las Vegas experience shows that it doesn't take long for a natural or artificial land shortage to lead to a rapid rise in housing prices. Many growth-management techniques will create such artificial shortages, including:
By causing rapid increases in home prices, growth management contributes to housing bubbles because people who might otherwise invest their incomes in stocks, bonds, or money-market accounts will be tempted to put more of their incomes in real estate instead. Thus, San Jose and Portland can have hot real estate sales markets even as their rental markets remain cool and priced well below mortgage costs. Speculators in these markets are betting that government policies will not be relaxed so that the artificially high prices will continue to increase.
In May, 2003, the Economist predicted that the housing bubble would burst "in the next year or so." While that didn't happen, the Census Bureau announced at the end of February that seasonally-adjusted new home sales and sale prices both declined significantly in January, 2005. Meanwhile, real estate stocks lost 15 percent of their market value in February, suggesting that Wall Street is beginning to worry about falling prices.
"A housing bubble doesn't burst, but only deflates," suggests UCLA economist Edward Leamer (see also his earlier report. Far from being reassuring, Leamer says this can mean a "long agonizing decline in prices." The best case is that high-priced markets could enjoy a soft landing, meaning homes in those markets slow their appreciation until homes in other markets catch up.
The worst case is a Tokyo-like crash that would force large numbers of people to default on their loans, which in turn could jeopardize the health of the nation's mortgage insurance programs, Fannie Mae and Freddie Mac. With the nation's housing stock worth trillions of dollars, a taxpayer rescue could conceivably cost far more than the savings-and-loan bailout of the 1980s.
Housing in many Massachusetts and California markets seems far too expensive for a soft landing. Such housing markets, suggests Leamer, are more likely to experience declines similar to (or worse than) those in the early 1990s. From 1989 through 1995, Los Angeles housing prices fell by 20 percent. Prices in the San Francisco Bay Area, San Diego, and Boston fell by 10 to 15 percent during the same time period, but most other housing markets did not lose at all.
I can't predict the future, but I suspect Bay Area, Boston, and other California and Northeast markets will suffer a severe correction in the next few years. Provided the contraction in those markets doesn't have huge economic impacts on the rest of the country, most other regions can hope for a soft landing. Still, this should provide a caution to those planners who think they can "manage growth" without potentially severe impacts on housing and other markets.