Although sprawl need not take priority over other issues, it is a problem. Its causes, however, are misunderstood, and anti‐sprawl policies are, as a result, often misguided. The best way to deal with sprawl is to address the NIMBY (Not in My Back Yard) syndrome that prevents developers from efficiently utilizing land in metropolitan areas. Unfortunately, most federal government land‐use laws of the past 30 years have exacerbated this syndrome by giving local anti‐development forces—the NIMBYs—more ammunition with which to stop reasonable development.
Highhways and Housing Subsidies
Many people argue that the federal government’s highway construction and tax policies contribute to sprawl by making people want to buy suburban housing. But neither of those is likely to be much of a contributor to the excessively low‐density housing patterns that characterize the suburbs of many of our metropolitan areas.
The federal government has indeed funded highway construction that makes it easier for cities to spread out. But it is not clear that the states would not have used the same gas‐tax revenues to do it themselves if they had had the opportunity. Regardless of who built the highways, however, most anti‐sprawl arguments are premised on highway congestion. Their implication is that curing sprawl will cure congestion. If people could be induced to live in high‐density areas, goes the argument, they would be eager to use newly built, fixed‐rail “mass transit” to get to work.
Traffic congestion is the result, not of too many highways, but of a reluctance on the part of the public everywhere to allow officials to charge congestion tolls for highway use. Until the public is willing to accept that urban highways should not be a free good, they will be subject to excessive congestion regardless of how many or how few people there are in urban areas. Putting people closer together by restricting suburban development would only make congestion worse.
The proposition that congestion can be overcome by building urban rail systems would be laughable if it were not so common and costly. None of the urban rail systems built in the last 30 years in the United States has made more than a trivial dent in traffic congestion in its metropolitan area, despite rider subsidies that dwarf those for automobiles. The reason is easy to see. Commuters like the idea of rail transit mainly because they suppose that the people in the cars in front of them will ride the rails. Almost nobody with a car actually wants to use the train himself.
The federal government subsidizes housing in many ways, but by far the largest is the income‐tax treatment of owner‐occupied housing. The income‐tax subsidy is both misunderstood and overstated as a cause of sprawl. The chief subsidy to housing is not, as is commonly asserted, the deduction of mortgage interest from taxable income. Most homeowners do not even itemize on their income‐tax forms, so deductibility means nothing to them. The benefit that all homeowners get is more subtle: They provide a service to themselves by managing their own property, and the government does not tax them on this income. They also do not have to pay capital gains on the sale of their homes. This makes an owner‐occupied home an untaxed asset, which gives it a huge advantage over almost all other accessible investments. To eliminate this feature, the U.S. government would have to imitate the Swiss government, which requires homeowners to pay taxes on what economists call “imputed rent,” or the money the homeowners would have paid to someone else to rent the home they occupy. A tax on imputed rent is pretty much like taxing home gardeners on the value of the vegetables they grow and consume. To state the nature of the tax is, I think, to show why it is unlikely to be adopted.
The homeowner tax subsidy, properly understood, does induce more homeownership and bigger homes. But its contribution to sprawl has to be small. That is because the extra acres on a multiacre lot, the kind that really does contribute to sprawl, account for such a small fraction of a home’s value. The real crime of excessively low‐density zoning is not the quarter‐acre lot. It is the multiacre zoning, which means that extra acres must be bought by homebuyers even though they value those acres very little. (Sometimes the multiacre aspect of such zoning is hidden by “cluster zoning,” which allows developers to group homes together. Clustering is just as wasteful, however, as regular multiacre zoning because the developer is usually required to keep the remainder of the parcel forever undeveloped.) Since the land that really contributes to sprawl contributes so little to home value, it hardly matters whether the federal government taxes its “imputed rent” or not.
All That Is Suburban Is Not Sprawl
The major contributor to low‐density development is the reluctance of suburban communities to accommodate higher‐density housing. Developers frustrated by NIMBY opponents in the areas close to jobs, amenities, and urban infrastructure are pushed to more rural areas. The result is lower‐than‐efficient housing density in the desirable areas and more housing in low‐density rural areas, which is what I would call sprawl.
The problem with endorsing the attention given to sprawl is that I seem to be in league with some unrealistic policies. Many planners, especially those who inhabit the academy, seem to regard almost all types of suburban housing developments as evidence of sprawl. The typical visual example of sprawl has long been an aerial photo of single‐family homes in Daly City, California, the San Francisco suburb that was the inspiration for Malvina Reynolds’s satirical song, “Little Boxes.” But Daly City’s residential development in 1970 (when the song was popular) was about two‐thirds the density of that of San Francisco, a city that most Americans properly think of as a high‐density place. If we lived in the homes that are scorned as sprawl, we would already be living at nearly the density of San Francisco.
Single‐family homes placed at four, six, or eight per acre do not take up much land. Inducing suburban developers to build at higher densities than this range both is unnecessary to prevent excessive land use and would frustrate the clear preference that Americans have for the freestanding, single‐family home. To convince yourself that rural land “loss” is not a national problem, follow some round‐number math: The United States contains about 2000 million (2 billion) acres. There are about 100 million households in the United States. If each household took up a full acre of land, that would amount to 100/2000, or 5 percent, of the nation’s area. But urban gross densities are in fact only about one‐half acre per household, even allowing for all the nonresidential uses such as shopping centers and parks. We don’t have to live in warrens to contain sprawl.
The free‐market side of the debate worries about sprawl because suburban governments are said to undercharge developers for their use of infrastructure. Property‐tax revenues generated by the development usually do not fully cover the costs of the new sewers, sidewalks, and utility lines, and so other taxpayers are in effect subsidizing sprawl. But this overlooks the fundraising capacity of land‐use regulators. Every municipality has the authority to insist that the developer actually cover those costs. The payments may be in kind or monetary exactions and impact fees. Indeed, the more realistic complaint, most often heard in “regulatory taking” court cases, is that municipal authorities demand too much. The problem with sprawl is not that developers fail to pay their way; it’s that they aren’t even allowed to start negotiations about higher‐density development in the face of NIMBY objections.
To contain sprawl, public policy should reduce the impact of NIMBYism on land development. One policy is to make the local governments that respond to unreasonable anti‐development measures pay for their preferences. The “takings clauses” of the U.S. and state constitutions empower judges to award monetary damages for unreasonably restrictive land‐use regulations. By making local governments financially liable for excessively restrictive zoning changes, the court system, especially if assisted by state legislation, could discourage unreasonably low‐density housing. Environmentalists, who are usually opposed to the compensation principle, should reconsider their stand in the case of urban and suburban land‐use issues. Compensation requirements would improve the environment by making NIMBYs put their money where their mouths are. Even a small prospect of having to pay would make them consider the costs as well as the benefits of opposing development in their neighborhoods.
A kinder, gentler way to discourage NIMBYism would attempt to assuage the financial anxieties of nearby homeowners. Homeowners adjacent to a proposed development often adopt seemingly irrational NIMBYism because of unlikely scenarios. The problem here is that a home represents such a large fraction of most owners’ assets, and devaluations from neighborhood change are not now insured. Home‐value insurance could be offered by developers (with solvency guaranteed by public authorities) to assure people immediately adjacent to new developments that they will not suffer financial losses. This would be similar in principle to the Home Equity Assurance programs in Chicago, whose intention is to discourage panic selling in neighborhoods undergoing racial integration. I have anecdotal evidence that the same program makes homeowners less anxious about other types of neighborhood change, such as apartment developments. It is an experiment worth looking into.
Federal Land‐Use Policies Promote Sprawl
The federal government’s role in dealing with NIMBYism should be to contract rather than expand its influence on land use. Most federal government regulations that affect local land use make it easier for suburban NIMBYs to keep densities in their communities low. Opponents of suburban development invoke federal endangered‐species legislation, wetlands regulations, and historic‐district designations to retard development. Those can frustrate efficient developments that the majority of residents of a community is otherwise satisfied with. Moreover, the application of federal land‐use regulations cannot be tempered by arguments that development is especially appropriate in the proposed areas. There is no “urban exception” permitted by such regulations, even though more than 95 percent of the land area of the United States is not in any sense urbanized.
Let me give an example from my 10 years’ service on a local zoning board in New Hampshire. Hanover, my hometown, has a compactly developed area in its southwest corner, surrounding Dartmouth College. There are also some outlying small villages and, increasingly, scattered developments on three‐acre lots (the minimum required by our zoning) in the rural part of town. The town’s master plan sensibly calls for most development to be encouraged near the compact area, where public water and sewerage are available and from which people can walk and bicycle (and sometimes ski) into the central area.
The owners of a large tract of land within half a mile of our “downtown” decided to develop it with single‐family homes. Those owners had previously developed about half of their holdings in the 1960s, also with single‐family homes. The remaining land was across a brook and adjacent to an extensive wetland. (They had previously donated much of the wetland for conservation purposes.) The owners proposed rezoning the land for moderate‐density homes (quarter‐acre to one‐acre lots) and building a bridge across the brook, which was about 15 feet wide, to provide access to their land.
To undertake their new project, the owner‐developers had to get their plans approved by the town, the state, and the federal government. At each of these phases, several nearby neighbors of the project, who enjoyed seeing the land in its natural state, sought to block the plans. Wetlands issues were the primary hurdle, but endangered species and Native American history (the latter two entirely unfounded) were also thrown up as reasons to stop the development. Regulators themselves were not hostile to the project, but the process became so drawn out and the outcome so uncertain that the landowners finally gave up. The land remains undeveloped.
Now this might seem to be a conservation success, and I must concede that I enjoy the open space that is available on this tract. (The owners have always permitted public access.) But I cannot help thinking that the houses that would have been built there will now pop up on the three‐acre lots in the more rural parts of town. By foreclosing development of land that was close in, we have probably induced development in areas less environmentally sensitive but in a pattern that can only be described as rural sprawl. The buyers of homes in those rural places will have more land around their homes, but, if the market is any test, they attach very little value to the extra acres. A landowner in the rural area of my town who owns a one‐acre lot (established before the three‐acre minimum lot size became the rule) can sell it for almost the same price as a three‐acre lot that cannot be subdivided further. Really.
One might ask, what’s the matter with such patterns? If road congestion and farmland acreage are not much affected by sprawl, why not let it happen? Here I advert to the more thoughtful anti‐sprawl arguments. People spread out in low‐density areas remain strangers to their neighbors. It is harder to get much sense of community if one lives in a spacious version of Ted Kaczynski’s Montana cabin. It’s next to impossible for kids to form a spontaneous baseball game (if their parents have to chauffeur them, the game’s not spontaneous). Moreover, excessively low‐density metropolitan areas lose much of the “agglomeration” economies that make cities productive places in which to work. Telecommuting is fine for some, but face‐to‐face contact is what makes the world go round. Growth Boundaries Are Not the Answer.
Now, the diehard “smart growth” enthusiasts will say, why not foreclose development in the rural areas, too? Establish an “urban growth boundary” around the compact area and eliminate building outside it. Even if our Constitution did not protect at least some property rights from public expropriation, this would be problematic. To foreclose development in the rural areas would cartelize the housing market. Buyers would be limited to bidding for the stock already available inside the boundary. Housing prices would rise as people moving to the area had to bid for a fixed stock. In the short run, existing homeowners would get a capital gain from such a policy. (I sometimes suspect that is what makes urban growth boundaries politically popular.) In the longer run, cities would suffer as employers found they could not hire people because the housing prices were too high. Anyone who claims that housing prices are not a problem for employers has not been involved in hiring people in a high‐priced area.
Of course, the “smart growth” advocates have a solution to that one, too. Instead of building more single‐family homes of the type most people aspire to own, make the suburbs build high‐density apartment houses to accommodate the growth. The immigrants can live in those downsized places, goes the argument, and we can prevent sprawl.
The appeal of this argument is pretty much like the appeal of urban rail transit. Just as people who drive to work hope that the folks in front of them will take the train, so the homeowners who vote for high‐density infill hope that newcomers will accept the high‐density, lawnless urban highrises in which they themselves would not dream of living. In fact, I think this accounts for the appeal of the “smart growth” movement in general. It’s a great idea, as long as we can get someone else to do it.
This article originally appeared in the September/October 1999 edition of Cato Policy Report.