Housing in 'Smart Growth' Cities: Is It Really Worth The Cost?

On August 12 the Wall Street Journal described a 350-square-foot former public toilet in south London that developers are turning into a "stylish apartment." They expect to sell it for around $200,000. "Believe me," a developer told the Journal, "there will be a lot of interest."

"In the past decade, the U.K. has been building fewer houses than at any time since World War II," says the Journal. The resulting housing shortage is reflected in the fact that people spend an average of just 18 minutes looking at a house before making an offer.

The Wall Street Journal attributes the housing shortage to "bureaucratic difficulties in getting planning permission -- especially in protected areas of greenery surrounding cities." As a result, "houses are so scarce that people will buy anything."

This seems likely to be the future of housing in Portland and other "smart-growth" cities. It is increasingly clear that housing affordability is strongly influenced by the level of government planning and regulation.

On August 9, USA Today printed a housing index developed by Coldwell Banker for scores of U.S. cities. The index is based on the median price of a mid-level, 2,200-square foot, four-bedroom, two-bath home. Table 1 below presents mid-level home prices for selected cities along with the growth rates of the city and urban area from 1990 to 2000.

A scan of the numbers suggests there is little correlation between home prices and growth rates. In the fastest growing urban area in America, Las Vegas, the mid-level home sells for $182,000. Despite slow growth and the dot-com collapse, housing prices in the San Francisco-Oakland and San Jose areas remain several times that amount.

On the other hand, there appears to be a strong correlation between land-use regulation and housing prices. Land-use regulations are strong in the San Francisco-Oakland and San Jose areas, in Oregon, Boulder, Massachusetts, and Maryland. Cities in these states and urban areas have the highest housing prices. Land-use rules are weak in Nevada, Arizona, Idaho, and Wyoming, and cities in these states have some of the lowest housing prices.

Of course, housing price is only part of the affordability equation. The other part is income. If incomes in Seattle are double those in Las Vegas, then Seattle housing (which costs slightly less than twice as much as in Las Vegas) may actually be the more affordable. Alas for Seattlites, Seattle household incomes are less than 50 percent greater than those in Las Vegas.

The National Association of Home Builders regularly compares median incomes with median home prices for nearly two hundred metropolitan areas. The "housing opportunity index" is the percentage of homes affordable to a family of median income in each metropolitan area. I compared the latest edition of the index (first quarter 2002) with the 1990s growth rates for those areas. These numbers are in Table 2.

The r-squared (a statistical measure of correlation) between the index and growth was less than 0.007, which is no better than random (i.e., two random number sets easily score r-squareds higher than 0.007). Thus, housing affordability has little relationship with growth. Instead, other factors such as land-use regulation are determining affordability.

According to the latest edition of this index, the nation's least affordable housing markets are almost all in California, Massachusetts, and Oregon, which are all heavily regulated states. Affordable fast-growing regions are in Arizona, Florida, Georgia, Nevada, and Texas. Except for Florida, all of these are lightly regulated.

Defenders of land-use planning argue that planning makes cities more livable, so naturally they would be more desirable and thus housing would be more expensive. But is Portland really more livable than Albuquerque? Or Oakland more livable than Las Vegas? San Francisco is a fun place to visit, but is it really four times more livable than Phoenix?

Attempts to make housing more affordable through "inclusionary zoning" -- an ordinance requiring developers to offer a certain percentage of their homes at prices affordable to low-income buyers -- will only make the problem worse. The "affordable housing" provided by this ordinance will make up a tiny percentage of the entire housing market. But developers will have to increase the cost of the other homes they build in order to cross-subsidize the affordable units. This will drive up overall market prices as resellers take advantage of higher new home costs.

I suspect the main beneficiaries of inclusionary zoning won't include many of the low-income people who are most hurt by housing regulation. Instead, recent college graduates, whose incomes are low enough to qualify for low-income housing but whose lifetime earnings are likely to be high, will probably snap up much of the low-income housing required by inclusionary rules.

Further research should develop an index of regulation that could be directly compared with, say, the NAHB housing opportunity index. The highest level of regulation might be found in cities such as Boulder or some parts of the San Francisco Bay Area that strictly limit the number of new building permits issued each year.

The most important thing homebuilders and realtors can do, however, is to put a human face on unaffordable housing. The South Carolina Landowners' Association is a coalition of realtors and low-income, often minority, landowners that is fighting land-use regulation in that state. This group provides a model that people in other regions should emulate.

Randal O'Toole is senior economist with the Thoreau Institute and author of The Vanishing Automobile and Other Urban Myths: How Smart Growth Will Harm American Cities.

APPENDIX

Table 1. Mid-Level Home Prices and City and Urban Area Growth

Table 2. First Quarter 2002 Housing Opportunity Index (HOI) and 1990-2000 Growth

HOI represents the percentage of homes in each region that is affordable to a family of median income in that region.
Growth is the population growth in percent between the 1990 and 2000 censuses.
MSA is metropolitan statistical area and PMSA is Partial MSA. MSAs and PMSAs are drawn along county boundaries.