Why Can't We Build an Affordable House?

The housing market is in tatters, and house prices continue to fall precipitously in many parts of the country, so it might seem a strange time to bring up the subject of housing affordability. But one of the reasons we are in this mess is that people bought houses they couldn’t really afford. At some point in the future, consumer confidence will be restored and people will start buying houses again. ­Pent-­up demand, and the inevitable delays in ­re­starting an industry that has seen the withdrawal of many home builders, will likely produce a spike in prices, and once again the affordability issue will come to the ­fore.

The term “affordable housing” has come to be associated with social programs and government subsidies, but it once meant commercially built houses that ordinary working people could afford. A pioneer of affordability was the builder Levitt and Sons, whose famous “Levittowns” were the first postwar examples of large, ­master-­planned communities. The story is ­well ­known. After World War II, as GIs came home and the peacetime economy gathered steam, the demand for housing grew dramatically. Levitt, an established Long Island builder, set its sights on this new market. William Levitt, the eldest son, applied his wartime experience building barracks with the Navy Seabees to traditional ­wood-­frame construction. He organized the building site like an assembly line. Teams of workers performed repetitive tasks, one team laying floor slabs, another erecting framing, another applying siding, and so on. No one had ever built housing that way ­before.

The first Levittown was on Long Island, the second in Bucks County, Pennsylvania, and the third in New Jersey. The Long Island project, because it was the ­first—­and the closest to New York ­City—­is the best known, but the Bucks County development, which began in 1951, was larger and more comprehensively planned and designed. At that site, the more than 17,000 homes on nearly 6,000 acres were intended chiefly for workers employed at a nearby steel plant. The largest and most expensive of the six model homes, the Country Clubber, was for supervisors and executives, but the ­three-­bedroom Levittowner was the workhorse of the development. It sold for $9,900, which would equal $82,000 today.

The design of the Levittowner, like the planning of the community, was the responsibility of William’s younger brother, Alfred. Though William Levitt went on to have a long and ­well-­publicized career as a developer and builder, Alfred, who died in 1966 (at only 54), is less remembered. He was a ­self-­taught architect who had spent an entire summer observing the construction of one of Frank Lloyd Wright’s ­so-­called Usonian houses, in Great Neck Estates, Long Island. Many of the Levittowner’s ­cost-­saving features were influenced by this experience: the efficient ­one-­story plan that combined an ­eat-­in kitchen with the living room; the concrete floor slab without a basement; the ­under-­floor heating; the low, spreading roof with no attic; and the carport instead of a garage. (The Usonians, Wright’s answer to affordability, were beautiful, but since they were built one at a time, they were ­expensive—­the Rebhuhn Residence, the one Alfred studied, cost a whopping $35,000 to build in 1937, the equivalent of more than half a million dollars today.)

Many of the design innovations of the Levittowner were Alfred’s own ideas. A folding basswood screen that slid on a metal track separated a ­so-­called study-bedroom from the living room, allowing the space to be open or closed. Thermopane (insulated glass) covered a large section of the ­living-­room wall overlooking the garden. The kitchen had a large window facing the ­street—­an early example of a “picture window.” High window sills in the bedrooms provided ­privacy—­and reduced cost. Locating the bathroom and the kitchen on the street side reduced the length of piping to the street mains. There was no mechanical room; instead, a specially designed furnace fit under the kitchen counter, its warm top doubling as a hot plate. The Levitts were careful to give ­penny-­pinching buyers of the Levittowner touches of luxury: the purchase price included a kitchen exhaust fan, an electric range, a GE refrigerator, and a Bendix washing machine. The Country Clubber added a clothes ­dryer.

A ­two-­way fireplace was located between the kitchen and the living room. Two-way fireplaces were a standard Usonian feature, but while the Levittowner had a low, spreading roof and clean lines, no one would mistake it for a Frank Lloyd Wright house. Yet, although Alfred Levitt’s design looks unremarkable today, in fact this early example of the ­so-­called ranch house represented a revolution in domestic design. ­One-­story living was new to most Americans, as was the open plan combining kitchen, eating space, and living room. The undecorated exterior was unabashedly modern. Picture windows had no precedents in traditional homes; neither did carports. Instead of brick or wood, the exterior walls of the Levittowner were covered with striated sheets of Colorbestos (asbestos cement), which had been developed especially for the Levitts by the Johns Manville Corporation. With integral color that didn’t require painting, this was an early example of ­low-­maintenance ­siding.

We don’t use asbestos cement anymore, and some of the other novelties, such as ­under-­floor heating, proved troublesome (as they did in the Usonians), but the Levitt brothers’ achievement remains impressive. They introduced the American public to modern production building and proved that standardization, mass production, and technical innovation could be ­successfully—­and ­profitably—­used by commercial builders to produce houses for a large market. Moreover, unlike many architectural experiments that have been dealt with harshly by the passage of ­time—­the ­high-­rise public-housing projects of the 1960s come to mind—Levittowns have remained desirable places to live. Even the names of the house models have survived. “Fabulous expanded Levittowner,” reads a recent Internet real estate ad for a house in the Bucks County community, “3 bedrooms, one bath, custom ­eat-­in kitchen.” It’s listed as ­sold.

The continuing popularity of the Levittowner after more than half a century does not mean that the demands of home buyers haven’t changed over time. Builders found out long ago that buyers would pay the small extra cost for the additional space provided by a basement. ­One-­story houses are still popular, especially with older owners, but ­two-­story houses have come back into vogue. So have traditional features such as porches, dormers, shutters, and bay windows. (The spare look of Alfred Levitt’s design would be a hard sell today.) Finally, buyers of the Levittowner were not given any choices; although Colorbestos came in seven colors, and the precise location of the carport varied from one house to another, these alternatives were predetermined by Alfred Levitt to create variety on the street. But modern buyers expect to personalize their homes. In response, while today’s builders still sell ­pre­designed models, they also offer scores of options: alternative façades, different materials, a variety of interior finishes, and “extras” such as upgraded kitchens, higher ceilings, and ­add-­on sun ­rooms.

Would it be possible to build a modern version of the affordable Levittowner? It would probably be a small house, closer to the 1,000 square feet of Alfred Levitt’s design than the 2,469 square feet that is today’s national average for new houses. Building smaller houses not only reduces construction costs, it is also good for the environment, saving materials and ­energy—­and land. The house would still have three bedrooms, but it would also have at least ­one ­and ­a ­half bathrooms, since people have come to expect a powder room, even in small houses. Closets would be bigger, and there would be more of them. There would probably not be a living room, but the house would include a family room facing the backyard. The kitchen would be larger, the ­hot-­plate furnace would be replaced by a conventional model, and the fireplace would be ­optional.

What would such a house sell for? In 1951, the price of the original Levittowner ($9,900) was three times the national average annual wage ($3,300). In 2008, with an estimated national average wage of $40,500, a similarly affordable house should have a sticker price of $121,500. Yet according to the Census Bureau, even in the current declining market the median price for a new ­single-­family house in the first quarter of 2008 approached twice that: $234,100. So, the price of a modern Levittowner would have to be nearly 50 percent cheaper than that of today’s average new house. Easy, you say, just make the house 50 percent smaller, about 1,200 instead of 2,469 square feet. But it’s not that simple. In most metropolitan areas, the selling price of such a house would still be more than $200,000, considerably more than $121,500.

So what’s keeping housing prices high? It’s not the size, and it’s not the construction costs, either. The Levittowner cost $4–$5 per square foot to build in 1951, equivalent to $30–$40 per square foot in 2008. That is approximately what an efficient, ­large-­scale production builder spends today. Home builders have followed the Levitts’ lead in streamlining construction, introducing ­labor-­saving techniques, and using industrialized materials. Plans are rationalized to reduce waste. Components arrive on the building site precut and preassembled so that the entire construction process for a typical house takes as little as three months. Perhaps the most important change in home building concerns scale. Since the 1980s, the industry has come to be dominated by a dozen national builders. These publicly owned companies, the largest of which produces as many as 50,000 houses a year, are able to take advantage of economies of scale that the Levitts could only dream about. Large, efficient enterprises buy materials in bulk, optimize ­mass ­production of building components such as windows and doors, and operate their own prefabrication factories. This keeps construction costs ­low.

What’s driving the high cost of houses today is not increased construction costs or higher profits (the Levitts made $1,000 on the sale of each house), but the cost of serviced land, which is much greater than in 1951. There are two reasons for this increase. The first is Proposition 13, the 1978 California ballot initiative that required local governments to reduce property taxes and limit future increases, and sparked similar ­taxpayer-­driven initiatives in other states. Henceforth, municipalities were unable to finance the ­up-­front costs of infrastructure in new communities, as they had previously done, and instead required developers to pay for roads and sewers, and often for parks and other public amenities as well. These costs were passed on to home buyers, drastically increasing the selling price of a house.

The other reason that serviced lots cost more is that there are fewer of them than the market demands. This is a result of widespread resistance to growth, the infamous not-in-my-backyard phenomenon, which is strongest in the Northeast, California, and the Northwest. Communities in growing metropolitan areas contend with increased urbanization, encroachment on open space, more neighbors, more traffic, and more ­school-­age children. Roads have to be widened, traffic lights added, and schools expanded, all of which lead to higher taxes. Voters commonly respond to these ill effects of growth by demanding restrictions on the number of new houses that can be built. Usually this is achieved by tightening zoning, invoking environmental constraints, and generally ­drawing ­out and complicating the permit process. It is no coincidence that house prices are highest in the Northeast, California, and the Northwest. According to the research of economists Edward Glaeser of Harvard and Joseph Gyourko of the Wharton School, since 1970 the difficulty of getting regulatory approval to build new homes is the chief cause of increases in new house prices. In other words, while demand for new houses has been growing, the number of new houses that can actually be built has been ­shrinking.

The most common tactic communities use to restrict development is to zone for large lots. In many parts of the country, the median size of new lots now exceeds one acre; by contrast, the 70-by-100-foot Levittowner lot covered less than ­one-­sixth of an acre. For the neighbors, requiring large lots has two advantages: It limits the numbers of houses that can be built and, since large lots are more expensive, it ensures that new houses will cost more, which drives up surrounding property values. But reducing development has another, less happy effect: It pushes growth even farther out, thus increasing sprawl. While ­large-­lot zoning is often done in the name of preserving open space and fighting sprawl, in fact it has the opposite ­effect.

It is a vicious circle. Smaller houses on smaller lots are the logical solution to the problem of affordability, yet ­density—­and less affluent ­neighbors—­are precisely what most communities fear most. In the name of fighting sprawl, local zoning boards enact regulations that either require larger lots or restrict development, or both. These strategies decrease the ­supply—­hence, increase the ­cost—­of developable land. Since builders pass the cost of lots on to buyers, they justify the higher land prices by building larger and more expensive houses—McMansions. This produces more community resistance, and calls for yet more restrictive regulations. In the process, housing affordability becomes an even more distant ­chimera.