LONGER VIEW PIECE FOR THE J. OF THE AM. PLNG. ASSOC.

Robert A. Johnston,  Univ. of Calif., Davis

Sept. 1, 1999

SPRAWL IS A SYMPTOM, NOT A CAUSE

The contrasting articles on sprawl by Gordon and Richardson and by Ewing in the Journal (Winter, 1997) and an analysis of these and other publications by Myers and Kitsuse (1999) stimulated this Viewpoint article. I want to engage in a thought experiment, to see if it helps to sort out the various issues involved in the analysis of urban density. As there are many values involved, I will attempt to address them with simple, long-range thinking. I'll define sprawl as low-density or non-contiguous development, usually with well-separated land use types.

I have gleaned a series of stipulated problems from the above publications and elsewhere and will go through the issues, one by one, suggesting conventional solutions for each problem. I take a very long-run view (100 years +) and assume population growth of 1-2% per year. I also assume a continuation of the recent trends of increasing household income disparity, or at least that the current very wide income distribution maintains. I also assume continuing rising average household incomes. I use forseeable transportation technologies and guess at their costs, when necessary. This list starts with the easier problems.

1. Urban aesthetics. Density doesn't affect aesthetics. There are ugly places at all densities and beautiful places at all densities. Contiguous growth, though, could help to visually define communties. If some people don't like endless urbanization, a la Southern California, then we could adopt greenbelts around cities or around unincorporated urban areas. Developer fees could pay for the greenbelts, perhaps combined with local taxes. The greenbelts could be used for scenic relief, recreation, wildlife, and other uses. Such (narrow) greenbelts will have little economic effect in the land market and will also not significantly affect travel and emissions. Some communities have adopted greenbelts. It is too late in already urbanized regions, but some narrow greenways and greenbelts can still be built, depending on local land use patterns. Mainstreets can be fixed up with trees, wider sidewalks, etc. Signs and billboards can be regulated. We know how to do all of this.

2. Open space and habitat. Many of the opponents of sprawl fear the loss of habitat and of other open spaces. By open space, here I mean large land areas for habitat and recreation beyond the major urbanized areas, or along streams or in wetlands anywhere. The solution advocated for this problem is often higher densities and contiguous growth. Neither solution will work, in many urban regions, if we assume continuing growth for a hundred years or more. Doubling density just buys time, but eventually urban growth covers the same territory. Contiguous, not leapfrog, growth merely delays the time period in which any habitat is fragmented or destroyed, by a decade or two. The solution to habitat needs is to develop regional and statewide habitat plans and place permanent development easements on these lands. Regions may also wish to develop regional recreation plans, if the habitat lands do not serve all of the forecast recreation needs. The President's Council on Sustainable Development recommended that local governments collaborate in such regional planning.

3. Auto travel and mobile emissions. Another argument for compact growth, urban growth boundaries, and mixed land uses is travel reduction and mode shifting to transit, both intended to reduce emissions. Increases in travel are due primarily to increasing incomes and falling auto operation costs. Transit mode shares are projected to fall in virtually all U.S. urban regions and in most European ones, due mainly to increases in auto ownership, caused by rising incomes. European transit shares are higher, due to lower absolute incomes in most countries, higher densities in older cities, and much higher vehicle and fuel taxes. We can reduce travel and emissions by only about 10-20% by doubling or tripling gross densities, increasing land use mix, and by doubling or tripling transit service, according to cross-sectional data and to many simulations. We, however, can double or triple bike and walk mode shares with good street design at crossings, separated bike paths, traffic calming, bike priority at intersections, and other related measures. So transit goes to 1% and walk and bike to 10%, in terms of person-miles of travel, perhaps 1% and 30%, respectively, in terms of trips made. However, many of the walk and bike trips will be new trips, not replacing auto trips. Such changes are significant, in terms of emissions, but are dwarfed by the emission reductions coming in the next 20 years from tailpipe and fuel standards already adopted. It would seem to be much more efficient to simply require SUVs to meet passenger car standards, for example. One can argue for higher fuel and vehicle taxes on economic efficiency grounds, but they are not efficient methods of reducing emissions.

Greenhouse gas emissions is a related issue. Indeed, the USEPA believes habitat loss and climate change to be the most important (long-range and irreversible) environmental problems in the U.S. The emission reductions forecast for air pollutants due to higher density development, discussed above, also apply to greenhouse gases. In addition, vehicle weight, displacement, or mpg taxes would reduce fuel use and, therefore, CO2 emissions. The U.S. has 4% of the world's population, but contributes about 23% of greenhouse gases worldwide. The U.S. transportation sector contributes about a third of the U.S. load and this share will rise to about half over the next 20 years, so this is a potentially devastating problem. The most effective method for reducing greenhouse gases, however, is a carbon tax, which would work very rapidly. This focused tax will then lead to the adoption of compressed natural gas vehicles, leading to liquified natural gas vehicles, finally leading to liquified hydrogen vehicles, the only technology that does not contribute to global warming from vehicle emissions. Hybrid electric vehicles, with batteries supplemented by small, clean fossil fueled combustion engines may play a role, as we ramp up to hydrogen.

It is difficult to project the cost of hydrogen vehicles, but let us just assume they cost twice as much to purchase and four times as much to operate per mile. These costs give a blended ownership cost per mile that is about three times the current cost. This cost increase, then, would reduce auto travel in vehicle-miles by about half, accounting for more-efficient vehicles being purchased. This reduction in travel then will result in people working closer to home and making fewer auto trips per day. We would expect much better transit service and an approximate doubling of densities in urban regions, over 30-50 years. Even if the U.S. doesn't adopt such policies, in order to reduce greenhouse gases, it seems likely that petroleum prices will begin to rise early in the 21st Century (due to demand increases in developing countries) and so start us down this path anyway.

If we adopt all of these policies, we now have permanent habitat protection, aesthetically satisfying urban land uses, and acceptable levels of pollutants and greenhouse gases. What are the other problems with low-density, noncontiguous urban development?

4. Local service costs. Many studies show that low-density developments far from existing services cost more per person or employee to service. This issue requires better empirical studies covering a range of jurisdictions and geographies, but the overall pattern seems to hold. In most cases, however, the differences in service costs are not very large, on the order of 10%. Employment uses pay their way and are favored in most jurisdictions, and so are often permitted to go practically anywhere they want. So, we see big box and superstore centers increasingly located beyond the urban edge. Even though they would be less costly to service in more central locations, local governments are not about to discourage them.  Large-lot residential projects generally pay their way and so many cities and counties zone very liberally for them. This land use generally can be on wells and septic tanks and so the public service cost burden is modest. We are, then, left with medium- and high-density residential land uses to consider (above 0.5 – 0.2 units per acre, depending on soil type for septic tanks). It appears that for these land uses, which are the majority of (nonroad) land coverage in most urban regions, being contiguous, or fairly close, to existing services is fiscally significant. Also, higher densities and clustering will also reduce service costs.

We now address several more-difficult socioeconomic issues.

5. Equity. Many commentators decry the inequity of current growth processes. First, let us address the issue of residential access to all parts of the region by lower-income households. This is the housing question. We have fragmented local government, which contributes to competition for favored land uses (employment, especially retail). This is not necessarily economically efficient, as there are both costs and benefits from this statutory approach. State laws that also make local governments dependent on local property taxes, returned to each jurisdiction, also increase the fiscal motives underlying local planning decisions. This legal situation then leads to further competition for rateables and also to exclusionary zoning, where many suburbs underzone for apartments or don't permit them at all. Jurisdictions also overzone for large-lot residential projects, which pay their way. Compact growth and contiguous growth are not related to these problems. Exclusionary, competitive local government behavior occurs at all densities. What is required is statutory revision, including amending the annexation and incorporation laws and the local taxation laws in each state. Regional revenue-sharing would be of great benefit, to remedy existing inequities. Regional decisionmaking for infrastructure investments would also reduce the subsidy to suburban areas, in terms of roads and sewers. Compact growth could actually make land prices rise and so hurt lower-income households, unless compensating policies are focused on encouraging affordable housing near to transit services. This could be done with density bonuses for only affordable housing projects, through the use of floating zones or other methods that make the bonuses not increase general land prices.

Second, let us address equity in transportation. Here compact and contiguous growth could help create the demand conditions for better transit service. A bus ride is subsidized about one dollar and so the $500 annual subsidy for worktrips is much less than that required to give a car to the worker in the car-less or one-car household. If we throw in the nonwork trips, we get a subsidy of about $2,000 per year, still below  the cost of a basic car. If, however, we go to rail transit, the subsidies go up three to five times, making the publically provided, or subsidized, car a viable alternative. So, a rail-based compact growth scenario has high costs for serving new lower-income riders. A bus-based plan is more financially sound.

Economic segregation is another variety of equity problem in many people's minds. I don't see why compact growth would reduce segregation of households by income. State statutes simply permit, and even encourage, this kind of local planning and zoning (regulating residential density) and the U.S. Supreme Court smiled upon it in Euclid v. Ambler, in 1928. Also, private deed restrictions on lot size, house size, etc. are in widespread use. We could pass state laws requiring fair share housing in all local jurisdictions. Also, state laws could be passed requiring density minimums in zoning, to the extent necessary to get affordable housing in every urban area. This latter law would make ineffective deed restrictions intended to prevent affordable housing. Another area where reform is needed is federal and state banking regulation. Rules for increasing investment in properties in older and minority communities are needed.

 6. Obsolescence of central city economic activities. A major complaint of many critics is that rapid outward urban growth increases the premature obsolescence of inner-area housing and firms. It seems, however, that this process is caused not by low-density development, but by the fiscal competition described above, that is due to the local government legal structure in each state and to the local taxation and spending laws. Compact growth will not reduce the development of new retail centers and superstores in outer parts of the region. Only regionwide constraints on overall land availability or on the zoning for such activities would increase the reuse and redevelopment of the older sites. Again, as with the equity issues, the compact growth advocates need to look toward regional planning, taxation, and spending to attack this problem.

We now come to the thorniest of the problems, the ones that are primarily social problems.

7. Poverty, race, and crime.  Major stimuli for families to locate in the suburbs are the high concentrations of povery, nonwhites, and crime in central cities. This cycle seems to be due to racism in employment and policing, the culture of poverty, and trickle-down of the building stock. The latter refers to older buildings being occupied by lower-income households, and so much of the concentration of poverty is due simply to the outward waves of newer buildings becoming available to households of moderate and high incomes, over time, and the vacant older units going to those of lower incomes. Rehabilitation/gentrification often  occurs in some of  the inner areas and most urban regions are more of a mixed picture, with some well-off neighborhoods in the central cities. It is not clear why compact growth would change these behaviors. State statutes requiring fair share housing in all localities would help a lot. The reform of state and federal laws regarding victimless crimes and regarding discrimination in arrest, defense, sentencing, etc. would also help greatly. Changes in state and especially federal income tax laws would also help lower-income households access a greater number of housing units and locate nearer to their jobs. 

 8. Sense of community and socialization of children. Our sense of community has dissolved into a faceless suburbia, according to many writers. Some critics have said that many middle-class Americans are so fearful of crime and of nonwhites, poor people, and homeless people that they literally want suburban locations where they can be surrounded by people of exactly their income, be separated by large lawns, go into the garage with an electric opener, and have very little pedestrian activity on their streets. In other words, Americans are in a defensive posture, far away from crime and from people in general, located on the plains so to speak, so predators can be spotted far away. If we could reduce crime rates so that they are the same everywhere, which would be possible with a reallocation of resources based on crime outcomes, then this lifestyle would no longer be necessary. We could have vibrant streets and have our back yards at somewhat higher than current new densities and with through street patterns. A greater mixing of land uses would also help to provide for more stimulation. We would also have to stop watching TV news shows, of course, which greatly contribute to our paranoia. Some observers believe that mixing middle-income households throughout urban neighborhoods would reduce crime rates by increasing demands for police services. As noted above, this solution does not depend on compact growth.

This leaves us with the socialization of children. Some advocates of compact growth argue that mixing  housing types and land uses makes for much more interesting places for children. The children get to see people of differing ages and incomes, get to watch people at work, and generally have more aspects of life accessible to them. It seems undeniable that higher densities, walkable streets, and mixed land uses would provide this benefit. It also seems like we would help children a lot more by creating a national health care benefit for them, by mandating that employers permit workers with small children to go to 50% or 75% time, and by encouraging people to limit or eliminate television watching by their kids.

Where does this leave us? From this discussion of the complaints against sprawl, we have induced a list of conventional policy reforms suggested by experts, most of which have nothing to do with compact growth. Resource scarcity will soon begin to bring about higher densities, such as the rising cost of petroleum. The seemingly inevitable adoption of carbon taxes will augment this trend. Most of the policies that I believe are needed to combat the problems being addressed by the compact growth advocates have to do with regional planning, taxation, fair share housing, and  habitat protection. Fairer health care, penal, and income tax laws would also help. These are all policies that have been advocated for decades, but are much more difficult to understand and implement than better urban design.

With the increased disparity in household incomes that we have seen over the last 3 decades, since Reagan made greed acceptable and many of us have taken his lead, it seems we have come to a new place in America. We have a developed country and an undeveloped country, both existing in most of our urban regions. Household incomes for the bottom third have fallen 40% in California over the last 30 years, for example. Most basically, what we need is urban regions where everyone can afford to live. We need fairer housing and tax laws. We need vital downtowns and small towns for those who want to live in such places. We also need shiny new suburbs for those who want to live in clean, safe, predictable places. To attain higher densities and land use mix in our central cities and small towns, we will need to overcome neighborhood opposition, at least in some of these neighborhoods, by people who like their neighborhoods as they are now. This can be done, in many instances, through the negotiation of improvements to the area, including traffic calming. Most of the benefits of higher density can be achieved by requiring fairly contiguous development at the urbanizing edge and leaving project densities alone. Better design of buildings, projects, and streets will come about through market demand, if local governments will deregulate density and mix, street widths, and other elements of current land use controls.

It seems to me that the problems in our urban regions that are attributed to sprawl are really problems of governance and political philosophy. We have a planning (and penal and tax) system that favors the middle- and upper-class majority and screws the poor. We favor short-term personal economic gain at the expense of: the poor families among us, the very existence of many of the nonhuman animals and plants, and any sense of pride in our governments and our cities. Perhaps we need to address these issues. The sustainable development advocates seem to be on to these issues, so maybe we can generate a national discussion. Perhaps discussions of compact growth can lead citizens into these broader evaluations of the  underlying legal, economic, and political issues. The visual preference survey method used in community design often succeeds in leading local citizens committees to address some of these more fundamental problems.

Congress could pass a law funding long-range regional planning. Section 701 of the Housing Act of 1954 funded the first county plans in much of the U.S. in the 50s and 60s. Forty years later we need to advance land use and transportation planning to the regional level. In several metro regions where such scenario exercises have been done, the participating citizens and interest groups chose compact growth and transit, rather than more freeways and low-density suburbs. They wanted shopping streets in walking neighborhoods and traditional gridded towns. Such regional planning exercises could possibly overcome the influence of the raw land owners, who seem to be able to keep local officials supporting new freeways and competitive overzoning beyond the urban edge. A few states now require regional planning, with the adopted plans generally implemented by the localities.

I hope that the energies of the compact growth/urban village scholars and interest groups can at least partly be directed to these more fundamental problems.

_______________

Epilogue.

A few years ago I was flying back to Chicago from Zurich, thinking to myself how beautiful the Swiss landscape was with its sharp urban edges, low vacant land rate, and generally clean looking land uses. I recalled how convenient the underground from the airport to the city was and how great the trams were also. On the plane, I started talking with three fellows who worked for a U.S. company based in a Chicago suburb with an office near Zurich in a village. I remarked about how I guessed that many of the U.S. employees probably tried to get assigned to the office near Zurich, because of its evidently superior urban environment. They all said, no, that it was the Swiss employees who tried to get assigned to the Chicago office. I asked why this was and they said it was because they could not afford to buy a home in the village near to Zurich and that in the U.S. suburb they could own a nice house with a big yard. From this experience and from national homeowership data, I conclude that we need to carefully evaluate the effects of compact growth policies on shelter costs. Surely we are smart enough to get more compact growth and keep our high home ownership rate.