Smart Growth Or Business-As-Usual? Contrasting Outcomes Of Different Approaches To Managing Growth Arthur C. Nelson, Ph.D., ASCE, FAICP
	Session:Designing Smart Growth Laws via Studies (March 11, 4:00pm)

INTRODUCTION “Smart growth” is a popular term these days. It conveys a new way of shaping the built environmental that is better than the old. It is pursued at many scales, from an individual house, to a subdivision, to a planned community, to a city, and to a region. At each higher level of geography, however, it would be safe to assume that smart growth is more difficult to achieve, and more difficult to measure. This manuscript addresses smart growth from the regional perspective. What is “smart growth” from a regional perspective and how can one know if it leads to better outcomes than the status quo or “business-as-usual?” This answers that question in four parts. The first part characterizes smart growth as a set of policies designed to achieve five goals: (1) preservation of public goods; (2) minimization of adverse land use interactions and maximization of positive ones; (3) minimization of public fiscal costs; (4) maximization of social equity; and (5), very broadly, maximization of quality of life. The second part compares how two metropolitan regions, Atlanta, Georgia and Portland, Oregon perform with respect to those five dimensions. Oregon and its principal metropolitan area, Portland, have been engaging in smart growth for about a quarter century while Georgia and its principal metropolitan region, Atlanta, have pursued a decidedly business-as-usual approach. If smart growth is to be effective in achieving its goals, objective measures of outcomes affecting Oregon and Portland should be positive with respect to Georgia and Atlanta. Measures developed along each dimension indicate that Portland performs consistent with smart growth expectations but Atlanta does not. The third part reviews recent literature evaluating social, job accessibility, metropolitan governance structure, presence of beltways that may stimulate urban sprawl, and economic development outcomes between different metropolitan areas representing varying degrees of smart growth approaches. Social connectiveness, otherwise known as neighborliness, improves with respect to the kind of density and reduced automobile dependency advanced by smart growth. The extent to which the public transportation network connects parts of metropolitan areas is shown to influence accessibility of jobs to low income households. Beltways disperse population so much that market thresholds for many retail and service operations are not achieved, thus robbing metropolitan areas of economic activity. In terms of the “bottom line,” recent research shows that metropolitan areas pursuing smart growth at least to some extent fare better than those that do not. Finally, metropolitan areas that include as an element of their overall governance structure mechanisms to address metropolitan-wide issues result in improved incomes to residents. The last part indicates that over the next generation, about 25 years, much will change in the built environment. Nationally, nearly half of the built environment existing in 2025 will be built between 2000 and then. This figure goes as high as 70 percent in such fast growing metropolitan areas as Orlando and Phoenix. Even in slow-growing or declining metropolitan areas, about one-third of their built environment in 2025 will be constructed between 2000 and then. There is, in effect, a real opportunity to reshape the nation’s built environment in ways that achieve smart growth goals. PART 1: SMART GROWTH “Smart growth” is sweeping the nation. It is a pejorative term that can mean anything to anyone. For example, providing “affordable” housing can be considered smart growth even if it is nothing more than conventional small-lot residential subdivisions. Zoning land for two-acre homesites can be considered smart growth because it preserves more trees per acre than conventional development (never mind the loss of functional open space). Reducing congestion through the construction of more highways is considered smart growth by the road building industry. What is smart growth, really? True smart growth rests on specific goals: Preserving public goods. A public good is something that we all need or use, no one can be excluded from using, but its overuse can damage everyone. Public goods include air, water, and culturally, historically, scientifically, or ecologically significant landscapes and habitats. Minimizing adverse interactions and maximizing positive ones. Certain land uses have adverse effects on others, such as spraying herbicides upwind from on petunias in neighboring yards. Some activities have synergistic effects on other land uses such as neighborhood schools on residential development. Smart growth should minimize if not prevent adverse impacts and maximize positive ones. An important consideration in this regard is the extent to which transportation systems provide access to land uses. Minimizing public fiscal costs. Smart growth should minimize cost per unit of development to provide public facilities and services. The result is more efficient use of scarce public resources. Maximizing social equity. Smart growth should improve conditions of low income households. This can include improving environmental quality, transportation accessibility, employment, income, and others. Maximizing quality of life. Finally, smart growth should lead to some maximization of quality of life, which is an admittedly elusive goal. Indicators of progress can be neighborhood quality, personal health, and housing affordability — the latter being among the most contentious discussions when directed to Portland’s urban containment efforts. Another indicator of quality of life is simply the vitality of central cities in the context of their metropolitan regions. For the past generation, several states have wrestled with the problem of accommodating growth in ways that achieve these goals. A few states, such as Oregon and Florida, have embarked on statewide smart growth efforts. In those states, local governments prepare plans meeting rigorous state criteria. Other states attempted, then abandoned smart growth efforts, such as Colorado and Maine. More recently, statewide smart growth efforts have been launched in Connecticut, Maryland, New Jersey, Rhode Island, Vermont, and Washington. Most states, however, take a business-as-usual attitude toward managing growth. This presents an opportunity for researchers to gauge the effectiveness of smart growth. Even if only partly successful, communities engaged in smart growth should out-perform communities taking a business-as-usual attitude toward growth in key indicators. But do they? Background In their land-use planning efforts, states fall into two general categories, those that attempt to guide the location and timing of development which are now called smart growth states, and those that have a business-as-usual attitude towards managing growth.1 Smart growth states mandate locally prepared plans that are consistent with state criteria; this is usually facilitated by state agency review with enforcement powers. The first wave of statewide smart growth efforts included the states of Hawaii (1961), Vermont (1970), Florida (1972), Oregon (1973), and Colorado (1974) which later abandoned its effort.2 They faced rapid growth and high levels of stress of natural systems.3 The 1980s saw a second wave of statewide smart growth efforts principally to rationalize public facility and economic development resources. They included Florida (1984-86), New Jersey (1986), Vermont (1988), Maine (1988) which subsequently abandoned its efforts because of cost, Rhode Island (1988), Georgia (1989), Washington state (1990-91), and Maryland (1992).4 Other states have mandated substate smart growth efforts to address: coastal management (24 states); development of important natural areas such as the Lake Tahoe (California and Nevada) Regional Planning Agency, New York’s Adirondack Park Agency, New Jersey’s Pinelands Commission and Hackensack Meadowlands Commission, and Maryland’s Chesapeake Bay Critical Area Commission; and management of metropolitan area development such as the Minneapolis-St. Paul, Minnesota Twin Cities Metropolitan Council, and Oregon’s Portland Metropolitan Service District. Planners in smart growth states or substate regions must prepare and implement plans consistent with state guidelines. Many growth management scholars observe that such plans must: (a) contain specific elements such as housing, environmental protection, and public facilities provision; (b) provide infrastructure concurrent with development; (c) coordinate plans between adjacent and nearby jurisdictions and with state agencies; (d) contain urban development and preserve rural land for nonurban or resource uses; (e) protect natural resources; and (f) be consistent with state and/or regional growth management goals and objectives.5 Smart growth states typically rely on a state agency to advance the state’s interest in local growth management planning efforts. Many of them prepare rules, offer technical assistance, and review and comment on the extent to which locally prepared plans are consistent with state growth management interests.6 In some states such as Florida7 and Oregon,8 state funds can be withheld from communities that fail to prepare acceptable plans. Oregon’s Land Conservation and Development Commission can impose building permit moratoria on local governments until acceptable plans are prepared.9 Even if state agencies have no such power, the presence of legislation mandating growth management can empower interest groups to use litigation to force local governments into preparing plans consistent with state growth management interests. In contrast, business-as-usual states either do not have the development pressures necessary to concern many citizens about potential problems or a political culture that supports the use of public policy instruments to interfere with private real estate development decision-making. This applies to the majority of states. Lacking clear statewide planning mandates, oversight by state agencies or interest groups, and either the power or the will to coordinate with other jurisdictions or state agencies, local plans in business-as-usual states do not guide influence the timing, location, and dimensions of development. Communities in those states seek not to manage growth so much as to provide it with a steady supply buildable land. Many implement their plans with little (or sometimes no) zoning and subdivision regulations.10 Literature demonstrates reasonably well that in the absence of smart growth mandates, local governments are usually ineffective preserving natural resources, containing urban sprawl, and mitigating losses from hazardous events.11 The reason is basic to decision-making: individuals or firms or, in this case, local governments, will always act in their self interest. Unless all competitors react to the same set of restrictions, the voluntary actions by some communities to interfere with development may steer development to other communities. Mandates are viewed as a common leveling device to assure that all decisions everywhere are influenced by the same constraints.12 Does smart growth make a difference? More directly, are growth-management states effective in achieving certain outcomes relative to business-as-usual states? This is the subject of the next two parts. PART 2 TALE OF TWO METROPOLITAN AREAS In theory, growth-management states should generate more benefits to society than business-as-usual states. The measurement of those benefits is illusive because there is little knowledge of the counter-factual, that is, what conditions would have been in growth- management states in the absence of growth management policies (Knaap and Nelson 1992). Quasi-experimental (before-and-after) analysis is hindered by insufficient controls and the fact that only over time can growth management effects be detected. Longitudinal analysis, such as Knaap’s and Nelson’s analysis of Oregon (1992), may only be applicable to single states over time. Cross-sectional analysis is often problematic because of the difficulty in acquiring data comparable to many different locations. Nevertheless, general tendencies of growth management efforts may be made by comparing roughly comparable states with and without growth management efforts. This is the paired-set approach. I apply this approach to Oregon (with growth management) and Georgia (without growth management)13 which are reasonably comparable in terms of growth pressures (as will be shown below). They are also dominated by one major metropolitan area, Portland and Atlanta. Oregon and Portland can be considered the “experimental” case while Georgia and Atlanta can be considered the “control.” Of course, this presumes that before experimental treatment, both were reasonable comparable. In previous work I have shown this to be the case with respect to growth and urban spatial expansion between the early 1960s and mid 1970s, before Oregon’s statewide land use planning program was launched.14 While comparisons between any set of states and their major metropolitan areas is always risky, such comparisons can be viewed as indicative of central tendencies between policy regimes. Moreover, comparisons are used routinely to gauge differences in along such dimensions as quality of life, economic well-being, social stability, and so forth.15 In the end, this kind of paired-set comparison is not causative per se but based on correlations of outcomes with policy regimes. Causative assessments are a natural next step in research. Baseline Comparisons Atlanta and Portland will now be compared in many respects regarding smart growth. Let us begin by comparing population and employment growth, and income change. This is done in tables 1-3, respectively, for the most recent ten-year period for which data are available (1988 to 1998). For all intents and purposes, growth and income change is comparable between Portland and Atlanta. Neither metropolitan area has seen a ten year period with more rapid growth. It would seem that neither smart growth nor business-as-usual approaches stifle growth. -- Source: Regional Economic Information System, Bureau of Economic Analysis, for respective years compiled by the author. Table 1 POPULATION GROWTH  Area 1988 1998 Change Portland CMSA 1,714,516 2,150,320 25.4% Atlanta MSA 2,846,202 3,744,022 31.5% -- Source: Regional Economic Information System, Bureau of Economic Analysis, for respective years compiled by the author. Table 2 EMPLOYMENT GROWTH  Area 1988 1998 Change Portland CMSA 987,086 1,364,922 38.3% Atlanta MSA 1,839,022 2,522,671 37.2% -- Source: Regional Economic Information System, Bureau of Economic Analysis, for respective years compiled by the author. Table 3 GROWTH IN PER CAPITA PERSONAL INCOME [Nominal dollars.] Area 1988 1998 Change Portland CMSA $17,295 $28,453 +64.5% Atlanta MSA $19,052 $30,788 +61.6% -- Comparisons Based on Smart Growth Goals Let us move on to a series of comparisons based on smart growth goals. Preserving public goods. Given resources especially availability of comparable data, comparisons are made with respect to air quality, consumption of land, reliance on septic systems, and consumption of energy. Consider air quality, first. In the 1980s, both Atlanta and Portland experienced violations of federal air quality standards (in ozone). By the early 1990s, Atlanta’s air quality worsened but Portland’s improved (see table 4). Since then, for the period 1994 through 1999, Atlanta’s air quality has continued to worsen but Portland’s has been reduced below federal violation thresholds (see table 5). Ozone improvements are attributable primarily although not exclusively to automobile travel, which will be discussed in more detail later. Developed land area is another smart growth indicator. Generally speaking, the more land developed per resident the less likely an area is able to achieve smart growth goals particularly preservation of public goods, though it is an indirect indicator nonetheless. Table 6 reports the acreage, percent, and acres per new resident of non-federal developed land for Oregon and Georgia for the most recent ten years for which data are available (1987 to 1997). (Data for metropolitan areas is not immediately available). The results appear compelling. Both states grew at the same rate and indeed both states saw an increase in developed land. However, developed land per resident rose 125.6 percent in Georgia but remained constant in Oregon. In terms of land developed per new resident, Georgia comes in at 1.21 acres per person while Oregon comes in at 0.41 acres per person. Reliance on septic systems foreshadows the potential for land and water pollution. Failing systems often go undetected and pollute ground and surface waters, and alter the biological composition of soils. Table 7 shows that reliance on septic systems in metropolitan Atlanta is increasing but it is decreasing in metropolitan Portland. Finally, consider energy consumptions. One outcome of smart growth not often considered is the need for energy. Energy production usually impacts adversely on the environment such as mining and processing of coal into energy, damming rivers for hydroelectric power, and even harnessing the wind because of its visual clutter. In most metropolitan areas, most of the energy consumed is by households. Smart growth leads to development patterns that economize on energy needs. The greater the demand for energy by households the less energy is available for firms requiring its use. Table 8 compares Oregon and Georgia in terms of energy consumption for the latest longitudinal period for which data are immediately available (1979 to 1995). The general expectation should be that energy consumption per capita ought to decline because of improving fuel efficiency, and more sensitivity to energy conservation in building design. What we find is that energy consumption indeed fell per capita in Oregon but rose in Georgia. A primary contributor to energy consumption is automobile travel which is discussed in more detail later. Minimizing adverse interactions and maximizing positive ones. A major share of disputes in land use occur when land uses are incompatible with each other. A hog rendering plant next to a residential subdivision would be said to have an adverse impact on the residents. On the other hand, a park next to a residential area would be said to have a positive impact. At a broader level, however, positive interactions are reflected by accessibility of land uses to transportation systems. Measuring interaction effects of smart growth and business-as-usual on land use is difficult primarily because data are simply not collected with this purpose in mind. We can infer some relationships, however, using surveys of households on their neighborhood characteristics. This is provided in reasonably uniform fashion over time by the American Housing Survey. Table 9, for example, shows that the rate of change of households citing neighborhood problems is nearly double that over time in metropolitan Atlanta (1987-96) than in metropolitan Portland (1986-95), while problems with undesirable land uses rose by a substantial rate in metropolitan Atlanta but fell in metropolitan Portland. Mixing land uses appropriately can lead to desirable interactions. While even less measurable than adverse interactions, the American Housing Survey again sheds some light. Table 10 shows the change in land uses within 300 feet of households. In metropolitan Atlanta, residential land uses and residential parking are increasing over time while commercial, industrial, and institution land uses are falling. In metropolitan Portland, residential land uses are rising at a smaller rate but residential parking is falling while commercial, industrial, and institution land uses are rising. -- Source: U.S. Environmental Protection Agency, National Air Pollutant Emission Trends, 1900-1995 Table 4 AIR QUALITY ATLANTA, GEORGIA AND PORTLAND, OREGON 1988-90 - 1993-95 Area Average Ozone Exceedance Days 1988-90 Average Ozone Exceedance Days 1993-95 Atlanta MSA 5.7 6.0 Portland CMSA 2.2 0.3 -- Source: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, AIRSData. Data for central cities only. Table 5 AIR QUALITY ATLANTA, GEORGIA AND PORTLAND, OREGON 1994 - 1999   Year Ozone Exceedance Days, Portland Ozone Exceedance Days, Atlanta 1994 1995 1996 1997 1998 1999 0 0 0 0 0 0 2 8 5 4 11 13 -- Source: Developed land data from Natural Resources Conservation Service, population data from Bureau of the Census. Table 6 LAND DEVELOPED 1987 - 1997   State  Year Developed Area Percent Developed  Population Developed Acres Per Resident Georgia 1987 1997 Change 2,698,200 4,238,100 1,539,900 7.8% 12.3% 57.1% 6,208,479 7,486,094 20.6% 0.43 0.97 125.6% Oregon 1987 1997 Change 1,071,300 1,295,500 224,200 3.6% 4.3% 20.9% 2,700,996 3,243,254 20.1% 0.40 0.40 0.0% -- Source: American Housing Survey for respective years and author. Table 7 SEPTIC SYSTEM RELIANCE 1986/87 - 1995/96 Category Portland 1986 & 1995 Atlanta 1987 & 1996 Septic system housing units, 1986/87 134,600 270,100 Septic system housing units, 1995/96 126,900 350,300 Percent change -5.7% +29.7% -- Source: Statistical Abstract of the United States for 1981 and 1998 as calculated by the author. Table 8 TOTAL ENERGY CONSUMED 1979 - 1995 [In British Thermal Units - BTUs.] Category Oregon Georgia BTUs/Capita, 1979 359,968,968 313,114,450 BTUs/Capita, 1995 332,900,000 348,500,000 Per Capita Change -7.5% +11.3% -- Source: American Housing Survey for respective years as calculated by author. "Undesirable land uses" includes categories for which problems were noted by respondent households in terms of "litter and deteriorating housing" and "commercial and industrial" uses. Table 9 CHANGE IN NEIGHBORHOOD PROBLEMS 1986/87 - 1995/96 Atlanta  Neighborhood Problems 1987 Housing Units 1996 Housing Units Percent Change All problems 389,000 466,700 20.0% Undesirable land uses 71,700 88,400 23.3% Portland  Neighborhood Problems 1986 Housing Units 1995 Housing Units Percent Change All Problems 282,600 318,600 12.7% Undesirable land uses 50,500 50,400 -0.2% -- Source: American Housing Survey for respective years as calculated by author. "Undesirable land uses" includes categories for which problems were noted by respondent households in terms of "litter and deteriorating housing" and "commercial and industrial" uses. Table 10 CHANGE IN LAND USES WITHIN 300 FEET OF HOUSEHOLDS 1986/87 - 1995/96  Atlanta  Uses Within 300 Feet 1987 Housing Units 1996 Housing Units Percent Change All housing types 887,400 1,199,100 35.1% Residential parking 56,100 172,700 207.8% Commercial, industrial, institutional 120,100 58,900 -51.0% Portland  Uses Within 300 Feet 1986 Housing Units 1995 Housing Units  Percent All housing types 604,900 699,100 15.6% Parking 104,100 65,900 -36.7% Commercial, industrial, institutional 65,700 119,300 81.6% -- These two tables offer an interesting study in contrasts. In metropolitan Atlanta, problems with neighborhoods are rising as are problems with undesirable land uses despite a falling rate of commercial, industrial and institutional land uses nearby, while in metropolitan Portland problems with neighborhoods falling as are falling with undesirable land uses despite a rising rate of commercial, industrial and institutional land uses nearby. Could it be that in addition to fostering more diverse land uses within its urban area, metropolitan Portland communities are able to improve design so that potential adverse interactions between land uses is softened? Let us now consider transportation. Local economies are dependent upon access between land uses and this is considered a positive element in smart growth. The more time and distance involved in trading, the higher the "transaction cost" and the less profitable a location becomes relative to competitors. Smart growth aims to improve accessibility first by shortening the distance between land uses and second by offering more ways to get around. Table 11 compares changes in vehicle-miles-traveled (VMT) per household in Oregon and Georgia between 1990 and 1995. Table 12 compares changes in use of different transportation modes in Portland and Atlanta during the same period. There are several interesting patterns that emerge. We know from data above that Oregon has pursued a policy of containment urban spatial expansion. One apparent outcome of this policy is that annual vehicle miles traveled per household rose more than ten times as fast in Georgia as for Oregon. Another apparent outcome is that dependency on automobiles for the commute to work is falling and that for transit and walking/bicycling is rising. Distance to work is falling although the time engaged in the commute is rising. For Atlanta, automobile dependency is rising and that for transit and walking/bicycling is falling. The distance to work is rising as is the time engaged in commuting. In all cases, outcomes for Portland are more favorable than those for Atlanta. -- Source: Nationwide Personal Transportation Study for 1990 and 1995 as calculated by the author. Table 11 VEHICLE MILES TRAVELED [VMT] PER HOUSEHOLD 1990 - 1995  Jurisdiction VMT/Household 1990 VMT/Household 1995 Percent Change Oregon 17,653 17,911 1.5% Atlanta 21,856 25,542 16.9% -- Source: Nationwide Personal Transportation Study 1995 for travel data and American Housing Survey for transit convenience among recent movers, as calculated by the author. Table 12 JOURNEY-TO-WORK 1990-1995 Region 1990 1995 Change Atlanta Automobile 93.2% 97.5% +4.6% Transit 4.5% 2.5% -45.4% Walk/Bike 2.3% 0.0% -100.0% Commute Distance (miles) 12.1 15.7 +29.8% Commute Time (minutes) 20.6 25.4 +23.3% 1987  1996 Convenience to Transit Among Recent Movers 1.4% 1.1% -23.1% Portland Automobile 95.7% 85.4% -10.8% Transit 3.4% 4.4% +28.1% Walk/Bike 0.0% 4.4% Gain Commute Distance (miles) 9.9 9.7 -2.0% Commute Time (minutes) 18.1 20.5 +13.3% 1986 1995 Convenience to Transit Among Recent Movers 0.8% 1.9% +151.5% -- Minimizing public fiscal costs. Let us now consider state and local revenues. One promise of smart growth is that the need for government revenues will not rise as fast. Table 13 compares changes in amounts of government revenues per capita for the period 1986-87 to 1996-97 for Oregon and Georgia. Oregon's total tax burden fell during this period while Georgia's rose. Of course, Oregon's tax burden remains higher than Georgia, which is a pattern typical among western and southeastern states. Table 14 makes a more direct comparison in tax burden by reporting effective property tax and sales tax rates between Portland and Atlanta. This table shows that property tax rates in Portland are declining in Oregon while those in Georgia are rising. In many respects, property taxes are a significant factor in the location decision of capital-intensive firms, which suggests that Portland may have an advantage over Atlanta in this regard. Sales taxes in Atlanta are also rising rapidly, although Portland has no sales taxes. Both tables indicate that tax burdens in Atlanta are rising but those in Portland are falling. -- Source: U.S. Bureau of the Census, annual state government finance reports, for respective years compiled by the author. Table 13 CHANGE IN STATE AND LOCAL TAXES PER CAPITA [1997 dollars.] Area 1987 1997 Change Oregon $7,298 $6,906 -5.4% Georgia $4,667 $5,169 +10.8% -- Source: American Housing Survey for respective years and author. Table 14 CHANGE IN LOCAL PROPERTY AND SALES TAXES Region 1985/86 1995/96 Change Atlanta Property Tax Rate $7.96 $9.71 +22.0% Average Property Tax [1997 dollars] $932 $1,045 +12.1% Sales Tax Rate $0.04 $0.07 +75.0% Portland Property Tax Rate $19.32 $13.81 -28.5% Average Property Tax [1997 dollars] $1,813 $1,806 -0.4% Sales Tax Rate $0.00 $0.00 na -- Maximizing social equity. Social equity touches on many things such as environmental quality, transportation accessibility, employment, income, and others. Environmental quality and transportation accessibility are addressed indirectly above although without benefit of knowing how those elements affect low income households particularly. Employment and income can be more easily measured (if just with less inconclusiveness). Because central cities have traditionally been home to much of any given region's low income and unemployed and households, we look at employment and income trends among central cities over time here. Naturally, if central cities are a large share of the metropolitan area, such as San Antonio, job capture will be high. In this case, both Atlanta and Portland are roughly the same size in population and land area (see below). Table 15 compares job change in both central cities relative to their suburbs for the period 1992 to 1997. This table shows that Portland added jobs at a rate fifty percent higher than Atlanta while suburban job growth in both cases were about the same. Table 16 shows the change in families below the poverty level. The percent of households under poverty fell slightly in Portland but rose substantially in Atlanta. -- Source: Developed land data from Natural Resources Conservation Service, population data from Bureau of the Census. Table 15 JOB CHANGE BETWEEN CENTRAL CITIES AND SUBURBS 1992 - 1997  City Share of Metro Jobs Atlanta Central City 14.5% Suburbs 30.6% Portland Central City 21.4% Suburbs 29.1% -- Source: American Housing Survey for respective years as calculated by the author. Table 16 CHANGE IN HOUSEHOLDS WITH INCOME BELOW POVERTY 1986/87 - 1995/96 Attainment Portland 1986,1995 Atlanta 1987,1996 Percent below poverty level, 1986, 1987 20.2% 31.3% Percent below poverty level, 1995, 1996 19.8%  36.2% Percent change -2.0% 15.7% -- Improving quality of life. Quality of life can mean different things to different people. No attempt is made to define it exhaustively. In this context, quality of life focuses on central city development vitality, health, neighborhood quality, housing quality, and housing affordability. One clear objective of smart growth is to stimulate development in central cities. The argument is that central city vitality improves regional vitality in the long run. Table 17 compares residential building permit activity in the central cities of Portland and Atlanta over the most recent ten years for which data are available (1988 to 1998). This table shows that Portland's share of its metropolitan area's new housing units has risen steadily, from around seven percent to nearly 20 percent, while Atlanta's has fallen, from about eight percent to less than four percent. One outcome is that Portland's population has increased 15 percent while Atlanta's has remained nearly unchanged. One reason for Portland's growth is annexation but during the study period it essentially added sufficient land to be roughly equal to Atlanta. In effect, based on comparable land bases in 1994, Portland was 17 percent more dense than Atlanta. -- Source: U.S. Bureau of the Census, Building Permit annual reports. Table 17 SHARE OF NEW METROPOLITAN HOUSING UNITS BUILT IN CENTRAL CITY, 1988 - 1998  Year  Population Land Area New Unit Share of Region  Population Land Area New Unit Share of Region   Atlanta Portland 1988 400,218 132 7.9% 429,188 120 6.5% 1989 397,118 132 6.1% 433,254 123 6.6% 1990 394,017 132 9.1% 437,319 125 8.2% 1991 395,332 132 3.0% 444,570 127 10.7% 1992 396,647 132 2.1% 451,821 129 9.1% 1993 397,962 132 2.4% 459,072 131 8.9% 1994 399,277 132 2.6% 466,322 133 7.5% 1995 400,592 132 3.0% 473,573 135 9.7% 1996 401,907 132 6.7% 480,824 138 14.2% 1997 403,222 132 3.4% 488,075 140 14.4% 1998 404,537 132 3.9% 495,326 142 18.1% -- Another element of quality of life is community health. Many manufacturing firms cannot locate in some parts of the nation because of air quality limitations. A major source of air pollution is automobiles. If effective, smart growth should create land use patterns that improve air quality primarily by reducing distance between land uses and affording more transportation alternatives. The effect should also be an improvement in air quality. We know from above that air quality in Atlanta is deteriorating while that in Portland is improving. In areas of the country where air quality is low, people are advised to remain indoors. This reduces opportunities for exercise. Moreover, if people need to spend more time in vehicles getting from place to place, this also reduces time available for exercise. One outcome may be increasing girth. Key results of a study recently reported in the Journal of the American Medical Association on change in obesity rates among states are reported in table 18.16 This table shows that Georgia has gone from among the lowest states in obesity rank to 29 in just seven years while Oregon has remained essentially constant in rank. Moreover, Georgia ranked first in the rate of change in percent of its population considered obese and, based on current trends, will have the nation's highest level of obesity. The apparent result is, in part, improved air quality in Portland but deteriorating air quality in Atlanta, and deteriorating health generally as measured by change in obesity levels. -- Source: Adapted from Ali H. Mokdad, et al., 1999. Table 18 CHANGES IN OBESITY IN GEORGIA AND OREGON, 1991 - 1998 Measure Georgia Oregon Obesity Rank, 1991 4 18 Obesity Rank, 1998 29 22 -- Smart growth promises to deliver neighborhoods of higher quality relative to business-as-usual. New movers typically seek neighborhoods that are better than their recent experience. It is thus interesting to note that neighborhood quality among recent movers has fallen in Atlanta but risen in Portland. Table 19 reports changes in neighborhood quality of life between Atlanta and Portland. This table indicates that among recent movers, neighborhood quality in Portland is rising while that in Atlanta is falling, although both were about the same in the mid-1990s. -- Source: American Housing Survey based on recent movers for respective years calculated by author. Table 19 NEIGHBORHOOD QUALITY AMONG RECENT MOVERS 1986/87 - 1995/96   Area Better Neighborhood 1986-87 Better Neighborhood 1995-96 Percent Change Atlanta 51.0% 43.3% -13.8% Portland 39.7% 43.5% +9.5% -- The relationship of smart growth to housing quality is a mixed bag. Smart growth probably leads to higher housing densities, smaller lots, and possibly smaller homes relative to business-as-usual. This is because under smart growth all housing consumers in theory internalize a larger share of their externalities while under business-as-usual new housing consumers push some externalities to others. Yet, as seen in table 20, housing quality in "smart growth" Portland appears to be rising relative to "business-as-usual" Atlanta. Consider that home ownership rates have risen faster in Portland than in Atlanta, and that persons per room (an indicator of overcrowding) has fallen in Portland but remained steady in Atlanta. On the other hand, Atlanta housing sizes remain higher than in Portland, and the average lot size in Atlanta, already very large in the 1980s, is getting larger while the average in Portland has remained constant. On the whole, it appears that households in Portland are on the whole as well housed if not better housed than households in Atlanta. Among the indicators of house and neighborhood quality, Portland households are enjoying higher rates of increases than Atlanta households. Much has been made in the national media about metropolitan Portland's housing price increases. Popular measures of housing affordability are inaccurate reflections of housing costs, however. Housing costs are more appropriately the combination of housing payments, taxes, utilities, repairs and maintenance, and transportation. These are often called "location costs." Transportation must be included because standard housing economics includes it as a necessary housing budget component. Table 21 compares location costs between Portland and Atlanta between 1992/93 and 1997/98. Three trends are apparent. First, incomes are rising more rapidly in Portland than in Atlanta. This is probably attributable to Portland's recovery from a serious recession during the early to mid 1980s. Second, transportation costs are rising rapidly in Atlanta but falling in Portland. Third, housing costs as a percent income are falling in Portland but rising in Atlanta. That Portland's housing prices have risen is unquestioned but whether prices are out of line with household's ability to pay considering total local costs appears unsubstantiated given objective evidence. Indeed, when considering all housing costs, housing appears to be more affordable in Portland than in Atlanta. -- Source: American Housing Survey as calculated by author. Table 20 HOUSING CONDITION COMPARISONS Atlanta and Portland Condition Atlanta Portland Home ownership, 1986/87 63.0% 60.0% Home ownership, 1995/96 63.7% 64.7% Rooms per house, 1986/87 5.7 5.5 Rooms per house, 1995/96 5.8 5.8 House size, 1986/87 1,813 1,598 House size, 1995/96 2,001 1,703 Lot size, 1986/87 33,106 9,583 Lot size, 1995/96 33,977 9,583 Persons per room, 1986/87 0.40 0.40 Persons per room, 1995/96 0.40 0.39 Change in opinion of house quality, all households +1.3% +2.2% Change in opinion of house quality, owners +0.1% +1.0% Change in opinion of neighborhood quality, all households +1.0% +3.6% Change in opinion of neighborhood quality, owners +0.6% +1.2% -- Source: Consumer Expenditure Survey, Bureau of Economic Analysis, for respective years calculated by author. Table 21 LOCATION COSTS HOUSING AND TRANSPORTATION 1992/93 - 1997/98 Metropolitan Area & Cost Component  1992-93  1997-98  Change Atlanta Income $43,903 $48,318 10.1% Housing 28.9% 27.9% -3.5% Transportation 14.5% 18.2% 25.4% Total 43.4% 46.1% 6.2% Portland Income $35,583 $45,806 28.7% Housing 29.5% 29.1% -1.4% Transportation 16.4% 15.9% -3.1% Total 45.8% 44.9% -2.0% -- PART 3 RECENT STUDIES Recent studies shed light on the relationship between regional smart growth approaches and outcomes in terms of social connectivity ("neighborliness"), job accessibility by low income households, metropolitan governance structure, and economic development. Consider first social connectivity. In an article in the Journal of the American Planning Association, Lance Freeman evaluated the relationship between automobile dependency and social ties within neighborhoods in Atlanta, Boston, and Los Angeles.17 Using cross-section statistical analysis of surveys in each city, he found that the odds of developing social ties within neighborhoods increased as automobile dependency decreased. Technically, a one unit decrease in automobile dependency (percent who drive to work alone) increases the number of neighborhood ties by 0.6 to 0.7 units. One desirable outcome of smart growth is making jobs accessible to more people. In a study of the connection between public transit and employment reported in the Journal of the American Planning Association, Thomas W. Sanchez found that access to public transit is a significant factor job accessibility throughout metropolitan areas.18 This writer's interpretation of Sanchez' work is that the higher percent of a metropolitan area that was accessible to public transit, the greater is use especially among nonwhite populations. Regional smart growth is probably effective only with a regional approach to addressing problems crossing all jurisdictions within the region. In an article appearing in the Journal of Urban Affairs, Arthur C. Nelson and Kathryn A. Foster evaluated the association between metropolitan governance structures and growth in per capita personal income among which in 287 of the largest metropolitan statistical areas for the period 1976 to 1996.19 After controlling for several factors, they found that the presence governance structures capable of addressing regional issues increased per capita income. The reasons include their ability to marshall resources to address regional solutions (such as transportation, water and wastewater provision, and growth management), and offer a forum within which disputes among jurisdictions within a region may be resolved. One of the most influential factors in stimulating urban sprawl is the construction of beltways and perimeters highways. Research reported by Arthur C. Nelson and Mitchell Moody in the Journal of Urban Planning and Development evaluated the association between the number of beltways in the 44 largest metropolitan areas and metropolitan economic activity measured as retail and service sales per capita.20 They found that the presence of one beltway is associated with a present value reduction in retail and service sales of about $8 billion for each million population, and two beltways reduce sales by about $10 billion for each million population. The reason is that beltways disperse population thereby preventing market areas from achieving densities needed to support retail and service operations at the margin. At its heart, regional smart growth attempts to improve the ordering of development to improve outcomes relative to the status quo. To taxpayers, growth management promises more efficient delivery of public facilities and services which equates to lower costs per unit of delivery. To developers, growth management promises more certainty in where development will be accommodated and at what scale. To citizens activists, growth management promises resolution of development problems in advance, instead of on an ad hoc basis. These promises are heroic. Nonetheless, even if only partly successful, communities engaged in growth management should out-perform other communities in overall economic output. This is because lower and/or strategically expended taxes increases local investment income and certainty in development combined with resolution of potential problems in advance streamlines the investment decision-making process so that market opportunities are exploited sooner. In an article appearing in the Journal of Planning Education and Research, Arthur C. Nelson and David Peterman evaluated the economic performance of 182 metropolitan statistical areas (MSAs) with 1990 populations between 100,000 and 500,000 over the period 1972 to 1992 with respect to presence or absence of regional smart growth efforts such as urban growth boundaries, urban service limits, and state or regional oversight of local planning.21 They found a positive association between the presence of growth management and economic performance; communities engaged in growth management realized marginal improvements in economic performance relative to other communities, all things considered. On balance, the evidence suggests that regional smart growth initiatives generate important benefits relative to business-as-usual. PART 4 THE NEW CANVAS In the United States today, we have about 110 million housing units and, by my reckoning, about 22.2 billion square feet of office space, 16.5 billion square feet of retail space, 7.7 billion square feet of warehouse space, and 14.8 billion square feet of manufacturing space. Nearly half of all this development is more than 40 years old. Much of it will be replaced over the next generation. Residential areas will be replaced by commercial and mixed use development. Office buildings and factories will be torn down and replaced, or converted to other land uses. What few people realize is that over the next generation probably most existing development will be replaced or converted to other land uses. This is on top of building new structures to accommodate new growth. To amplify this point, consider the nation as a whole. Between 2000 and 2025, the nation's population will grow from 281 million to 340 million, and employment will grow from 166 million to 222 million. If we assume that two-thirds of the built environmental is devoted to residential land uses and one-third to employment-based land uses, roughly 25% of all development seen in 2025 will have been built between 2000 and 2025 (see Table 22).22 What about existing development? Existing development deteriorates and existing uses in such development become obsolete. There are no good figures on the extent to which existing development is torn down and replaced, though some assumptions are reasonable. Suppose we assume, to be conservative, that all land uses survive 100 years before being replaced (although not necessarily torn down). This may be high or low depending on the area but it seems a useful point of departure. Using this assumption, about 20% of the development seen in 2025 will be existing development in 2000 that is replaced (see Table 22). For the nation as a whole, nearly 50% of the built environment in 2025 will have been built (or reconfigured) between 2000 and 2025. For some large, rapidly growing metropolitan areas, this figure exceeds 60% such as Atlanta, Las Vegas, Orlando, Phoenix, and Portland (Oregon). Even the slowest growing metropolitan areas will see new development and conversion exceeding 25% such as Cleveland and Pittsburgh. Let us be somewhat more specific. Between 2000 and 2025, the United States will add, roughly: _ 60 million people. _ 24 million households. _ 56 million jobs of which half will be in business, professional, and personal services. In rough terms, this equates to: _ 45 million housing units (half of which replace existing units). _ 10.0 billion square feet of retail space of which 5.3 billion will be newly added. _ 25.0 billion square feet of office space of which 12.2 billion square feet will be newly added. _ 2.0 billion square feet of industrial space of which 1.0 billion will be newly added. Incomes will rise, opportunities expand, the role of technology broaden, and quality of life improve for most people; maybe everyone. For planners and analysts and the communities for which they work, this suggests that there exists a real opportunity to reshape development patterns over the next generation. We have set before us a new canvas. How shall we paint it? -- Source: Arthur C. Nelson (2001). Table 22 DEVELOPMENT NEEDS 2000 TO 2025     Area New Development Needed Conversion @ 100-Year Useful Life Total Development Needed United States 24.7% 20.1% 44.7% Atlanta 43.9% 17.4% 61.3% Baltimore 23.2% 20.3% 43.5% Boston 12.3% 22.3% 34.5% Chicago 16.4% 21.5% 37.9% Cleveland 5.0% 23.8% 28.8% Dallas 39.6% 17.9% 57.5% Denver 39.6% 17.9% 57.5% Detroit 10.7% 22.6% 33.3% Houston 37.7% 18.2% 55.9% Kansas City 25.7% 19.9% 45.6% Las Vegas 71.1% 14.6% 85.7% Los Angeles 10.1% 22.7% 32.8% Miami 23.8% 20.2% 44.0% Minneapolis-St. Paul 34.9% 18.5% 53.4% New York 3.4% 24.2% 27.6% Oakland 28.7% 19.4% 48.1% Orlando 55.7% 16.1% 71.7% Philadelphia 8.5% 23.0% 31.6% Pittsburgh 5.0% 23.8% 28.8% Phoenix 58.8% 15.7% 74.6% Portland 49.0% 16.8% 65.8% St. Louis 14.5% 21.8% 36.3% San Diego 39.9% 17.9% 57.7% San Francisco 13.8% 22.0% 35.7% San Jose 23.5% 20.2% 43.8% Seattle 32.8% 18.8% 51.7% Tampa-St. Petersburg 31.7% 19.0% 50.7% Washington, DC 28.7% 19.4% 48.1% -- SUMMARY OBSERVATIONS The analyses reported here are only preliminary. Very little work has been to quantity smart growth outcomes. There is a need for research that properly characterizes what smart growth is, how to measure its outcomes, and how to apply those measures to different areas pursuing different smart growth or business-as-usual approaches. Nevertheless, based on objective analyses, many having undergone peer review in refereed journals, on balance it would appear that smart growth efforts generate better outcomes than business-as-usual. The benefits of smart growth may take many years to see, however. Oregon, for example, began its program in 1973. It is only now, a full generation later, that measurable results can be seen. The extent to which politicians and, more directly, the general public are willing to sacrifice short term benefits of business-as-usual today for tangible smart growth benefits in 25 or more years is not clear. What may be clear, however, is if the short term sacrifice is made, the quality of life for our children will likely be improved. Notes They used to be called "growth management" states. For examples, see John M. DeGrove, Land Use Plans and Politics, Chicago, IL: American Planning Association (1984); Douglas R. Porter, Managing Growth in America's Communities, Washington, DC: Island Press (1997). See John M. DeGrove, Land Use Plans and Politics, Chicago, IL: American Planning Association (1984); Bollens (1992). See John M. DeGrove, Planning and Growth Management in the States, Cambridge, MA: Lincoln Institute of Land Policy (1992); DeGrove (1984); Arthur C. Nelson and James B. Duncan, Growth Management Principles and Practices, Chicago, IL: American Planning Association (1995). Bollens (1992); Nelson and Duncan (1995). See Gerrit J. Knaap and Arthur C. Nelson, The Regulated Landscape, Cambridge, MA: Lincoln Institute of Land Policy (1992); DeGrove (1992); Bollens (1992); Nelson and Duncan (1995); and Kaiser, Godschalk, and Chapin (1995). DeGrove (1992). DeGrove (1992). Knaap and Nelson (1992). H. Jeffrey Leonard, Managing Oregon's Growth, Washington, D.C.: Conservation Foundation (1983). Kaiser, Godschalk, and Chapin (1995). See Linda C. Dalton, and Raymond J. Burby, Plans or Planning? Building Local Commitment to Managing Urban Growth, New Orleans, LA: College of Urban and Public Affairs, University of New Orleans (1993); Steven P. French and Arthur C. Nelson, The Northridge Earthquake: Land Use Planning for Hazard Mitigation, Atlanta, GA: Georgia Institute of Technology (1996); Raymond J. Burby, Steven P. French and Arthur C. Nelson, "Plans, Code Enforcement, and Damage Reduction," Earthquake Spectra 14: 59-74 (1998); DeGrove (1984; 1992); Knaap and Nelson 1992; Nelson and Duncan (1995). See D. Mazmanian and P Sabatier, Implementation and Public Policy, New York, NY: Landham (1989); M. Goggin, Implementation Theory and Practice: Toward a Third Generation, Glenview, IL: Scott, Foresman (1990); Ronald C. Fisher, State and Local Public Finance, Homewood, IL: Scott, Foresman (1995). Arthur C. Nelson, "Growth Strategies: The New Planning Game in Georgia," Carolina Planning 16(1): 1-6 (1990). Arthur C. Nelson, "Comparing States With and Without Growth Management: Analysis Based on Indicators With Policy Implications," Land Use Policy 16: 121-127 (1999). When using objective measures, one can compare even the most different areas along common dimensions as is commonly done in such widely-used publications as Places Rated Almanac (David Savageau and Geoffrey Loftus, Macmillan Travel, 1997). Ali H. Mokdad, et al., "The Spread of the Obesity Epidemic in the United States, 1991-1998", Journal of the American Medical Association, October 27, 1999, Vol. 282, No. 16, pp 1519-1522. Lance Freeman, "The Effects of Sprawl on Neighborhood Social Ties," Journal of the American Planning Association, 67(1): 69-77 (2001). Thomas W. Sanchez, "The Connection Between Public Transit and Employment," Journal of the American Planning Association, 65(3): 284-296 (1999) (awarded "article of the year" by the APA). Arthur C. Nelson and Kathryn P. Foster, "Metropolitan Governance Structure and Economic Performance," Journal of Urban Affairs, 21(3): 309-324 (1999). Arthur C. Nelson and Mitchell Moody, "Effect of Beltways on Metropolitan Economic Activity." Journal of Urban Planning and Development 126(4): 189-196 (2000). Arthur C. Nelson and David R. Peterman, "Does Growth Management Matter?" Journal of Planning Education and Research. 19(3): 277-286 (2000). Arthur C. Nelson, A New Canvass: The Time Really is Now to Shape the Built Environment. Atlanta, GA: City and Regional Planning, Georgia Institute of Technology (2001). Author and Copyright Information Copyright 2001 by Author Arthur C. Nelson is a professor of City Planning and Public Policy at the Georgia Institute of Technology. He is a nationally known expert in growth management, urban containment, resource land preservation, infrastructure finance, and urban development policy. Dr. Nelson is currently serving as an expert consultant to the U.S. Department of Housing and Urban Development (HUD) on smart growth policy. His books on growth management and impact fees are standard texts. Recent sponsors of his research include the Fannie Mae Foundation, National Academy of Sciences, National Science Foundation, HUD, the U.S. Department of Transportation, the Lincoln Institute of Land Policy, and the American Planning Association. His awards include election to the College of Fellows of the American Institute of Certified Planners, teacher of the year, researcher of the year, professional educator of the year, and advisor to students winning the national AICP student project award. Dr. Nelson’s current research interests include methods to measure exclusionary zoning, rethinking the composition of land uses within neighborhoods and communities, evaluating alternative urban containment institutional structures, and measuring the effect of different growth management regimes on regional development patterns.