[*PG581]ABOLISHING EXCLUSIONARY ZONING:
A NATURAL POLICY ALLIANCE FOR ENVIRONMENTALISTS AND AFFORDABLE HOUSING ADVOCATES
Abstract: Exclusionary zoning limits residential development over large areas, and even entire cities or towns, to single-family housing on large lots. Exclusionary zoning is unfair to people and families of modest means (many of whom are members of racial or ethnic minorities) because it sharply limits where they can live and thus their access to jobs, education, and a good quality of life. For these reasons, exclusionary zoning was found to violate the New Jersey Constitution in the Mount Laurel case. But exclusionary zoning is also an environmental problem because it is a primary ingredient of the accelerating pace of urban and suburban sprawl. As a consequence, it is a major contributor to increased air and water pollution and habitat fragmentation. The Oregon planning program demonstrates how the abolition of exclusionary zoning promotes a more equitable range of housing choice in suburbs and simultaneously reduces environmental degradation associated with low-density urbanization.
The zoning in the suburb of the Township of Mount Laurel, which was challenged by the South Burlington County chapter of the NAACP in May, 1971, was typical of suburban zoning then and now.1 Most of the township’s land was zoned for single-family homes on half-acre lots; apartments, duplexes, and mobile homes were prohib[*PG582]ited.2 This kind of zoning practiced city-wide is known as “exclusionary zoning.”3
In the first two decades of the twentieth century, city zoning was used to separate incompatible uses—noisy and polluting industrial and commercial uses were to be kept out of residential areas.4 But from the beginning, many city planners treated apartments and other kinds of multifamily housing as equally noxious, a threat to property values and the public welfare.5
State supreme court decisions upholding exclusionary zoning featured paeans to the superiority of single-family detached houses as the best way to advance civic virtue and public health.6 In many cases, this zoning had explicitly racist origins.7
Only rarely did courts acknowledge the class- and race-based animus underlying residential zoning, and question its constitutionality:
And no gift of second sight is required to foresee that if this Kentucky [racial zoning] statute had been sustained [by the Supreme Court in Buchanan v. Warley], its provisions would have spread from city to city throughout the length and breadth of the land. And it is equally apparent that the next step in the exercise of this police power would be to apply similar restrictions for the purpose of segregating in like manner various groups of newly arrived immigrants. The blighting of property values and the congesting of population, whenever the colored races or certain foreign races invade a residential section, are so well known as to be within judicial cognizance.
. . . .
[*PG583] . . . The purpose to be accomplished is really to regulate the mode of living of persons who may hereafter inhabit it. In the last analysis, the result to be accomplished is to classify the population and segregate them according to their income or situation in life.8
The decision of the district court in Ambler Realty Co. v. Village of Euclid was overturned by the Supreme Court in 1926 in the decision confirming the constitutionality of municipal zoning.9 Unlike the district court, the Supreme Court shared the prevailing view of apartments as a kind of spreading blight on the public good and property values.10 In addition, the segregation of Americans by class and race fostered by exclusionary zoning was powerfully reinforced by the 1938 Federal Mortgage Insurance Underwriting Manual11 and other public and private guidance.12
[*PG584] In the 1960s and 1970s, the impacts of class-based low-density suburban zoning on access of lower-income Americans to affordable housing, jobs, and schools became the focus of judicial review at the federal and state level.13
Those cases show the continued growth in residential minimum lot sizes that allowed segregation of the highest income suburbs.14 Examples include: one-acre minimum lot size in Bilbar Construction Co. v. Board of Adjustment15 and Agins v. City of Tiburon;16 three-acre minimum lot size in Appeal of Kit-Mar Builders, Inc.;17 four-acre minimum lot size in National Land & Investment Co. v. Kohn;18 five-acre minimum lot size in Robert E. Kurzius Inc. v. Incorporated Village of Upper Brooksville;19 and ten-acre minimum lot size in In re Application of Wetherill.20
And the lot size minimums continue to grow, with ten-, twenty-, and even forty-acre lots in new ranchette subdivisions around the country.21 If these trends continue, then an increasing share of our urban regions will be off limits to Americans who cannot afford a big house on a big lot.
In addition to judicial review, the social and political effects of exclusionary zoning have also been the subject of scholarly and politi[*PG585]cal debate.22 Only recently, however, have the impacts of exclusionary zoning on the environment been examined.
Suburban zoning, founded on class-based, low-density zoning, has resulted in a massive expansion of America’s urban areas.23 As the following table shows, the rate of expansion far exceeds the rate of population growth.
| Changes in Population and Land Area for Select U.S. Metropolitan Areas 1950–9024 |
| Urbanized Area |
Population Growth | Urbanized Area Growth |
Ratio of Area Growth to Population Growth |
| Pittsburgh | 9.5% | 206.3% | 21.72 |
| Buffalo | 6.0% | 132.5% | 20.08 |
| Boston | 24.3% | 158.3% | 6.51 |
| Philadelphia | 44.5% | 273.1% | 6.14 |
| St. Louis | 39.0% | 219.0% | 5.62 |
| New York | 30.5% | 136.8% | 4.49 |
| Chicago | 38.0% | 123.9% | 3.99 |
| Minn./St. Paul | 110.7% | 360.2% | 3.25 |
| Atlanta | 325.4% | 972.6% | 2.99 |
| Washington | 161.3% | 430.9% | 2.67 |
Rapid expansion of an urbanized area can occur even in regions with no population growth or even declining population growth.25 For example, between 1982 and 1996, the urban region of Detroit, Michigan experienced a 1.1% decline in population even as its urbanized area grew by 19.6%.26 During the same period the Buffalo-Niagara urban area’s population was unchanged, but its urbanized area grew by 52%.27
[*PG586] If the ratio between population growth and the urbanization of land established between 1982 and 1997 continues into the future, then between 2000 and 2025 the United States will urbanize almost as much land in those twenty-five years as it did in the first 200 years of the nation’s existence—an area the size of Wyoming.28 This urbanization occurs at the expense of other land uses and resources.
Between 1982 and 1997, an estimated 10.3 million acres of forestland and 3.3 million acres of rangeland were developed, along with 11.3 million acres of crop and pasture land.29 In recent years the biggest cause of the destruction of wetlands has shifted from agriculture to development, which accounts for just over half of all wetland losses.30
Low-density residential development is associated with a rapidly expanding road network needed to serve dispersed housing.31 The hard surfaces of the road accelerate the speed of rainwater runoff, increase the temperature of the water, and carry a spectrum of dangerous automobile-related pollutants into streams and coastal waters.32 For example, research on development in the Tuckahoe Creek watershed in Virginia quantified the impact of the construction of about 20,000 large-lot home sites between 1958 and 1990, which were served by forty-two new stream crossings and a 155% increase in im[*PG587]pervious surfaces.33 During the same period, six species of indigenous fish became extinct, there was an 80% decline in the population of remaining indigenous fish species, and there was a significant decline in species diversity.34
Roads fragment ecosystems, which adversely affects wildlife.35 Low-density residential development patterns also fragment big game habitat. Each new house may have a negligible impact, but the cumulative effects can be significant.36 Because a disproportionate share of recent and future population growth will be in coastal regions, and because suburban development patterns magnify the impacts of this growth, suburban development will have far-reaching impacts on marine and estuarine natural resources.37
Falling population densities attributable to large-lot suburban zoning are linked to a significant increase in per capita driving.38 In the last half of the twentieth century, the population of cars and trucks in the nation more than quadrupled, while the human population has not even doubled.39 The average amount of driving per American more than doubled in twenty-five years, from 4587 miles per year in 1970 to 9567 miles per year in 1995.40 As a result, the total amount of driving is overwhelming the benefits of pollution control emission technology.41
Residential density plays an important role in determining how much driving is required; lower residential population densities (as well as the separation and dispersal of commercial and other uses) are an important factor in increased air pollution from cars and trucks.42 An illustration of the relationship between development density and [*PG588]driving is provided by western New York, where population decreased, but the total amount of driving increased.43 High capacity transit, which performs best in association with moderate densities within walking distance to stations, is not feasible in low-density residential suburbs.44
Sophisticated computer models that relate land uses, residential densities, and air pollution can quantify these relationships.45 For example, Metro, the regional government in the Portland, Oregon metropolitan area46 analyzed several different regional development patterns for the Oregon portion of the metropolitan region.47 Alternative A, a lower density alternative, would have a gross residential density of 9.8 people per acre, 26% of its housing in multifamily types, and 57% of its land zoned for single-family residential use.48 Alternative B, a higher density alternative, would have a gross density of 12.4 people per acre, 40% of its housing in multifamily housing types, and 46.5% of its land in single-family residential zoning.49
Alternative B, the higher density alternative, was projected to produce about 29,000 kilograms less of carbon monoxide and about 7000 kilograms less of nitrous oxides each summer day than Alternative A.50 This reduced air pollution would, at least in part, be a reflection of a lower amount of driving per capita under Alternative B [*PG589](10.86 miles per day compared with 12.48 miles in Alternative A) and 50% higher transit use (although still at a modest level).51
Suburban zoning causes these environmental effects, as well as many others.52 Although not all of the environmental impacts of low-density, suburban-style urban development are caused by class-based exclusionary zoning, given the large share of urbanization devoted to single-family residential uses,53 large-lot suburban zoning may be the single biggest contributing cause of these impacts.
Conversely, efforts that reduce regulatory barriers to higher-density urban residential development will allow for both an increase in lower-cost housing types54 and a decrease in environmental impacts.
In 1973, the State of Oregon adopted a statewide land use planning program that addressed a wide variety of topics, but one of its greatest achievements remains largely unknown: its sweeping reform of the practice of low-density, class-based exclusionary zoning.55
Republican Governor Tom McCall teamed up with Republican State Senator Hector Macpherson and Democratic State Senator Ted Hallock, to win passage of Senate Bill 100, a comprehensive land use planning statute.56
[*PG590] Although Senate Bill 100 had many components, its most important feature was the establishment of mandatory state land use planning policies (Goals), which all local governments57 were required to implement through binding, not advisory, comprehensive land use plans.58 Those city and county plans were subsequently executed by local zoning and other land use regulations.59 The task of adopting the Goals,60 as well as reviewing city and county land use plans and implementing ordinances to determine if they complied with the Goals, was given to a new state board composed of volunteers appointed by the governor, the Land Conservation and Development Commission (LCDC),61 and aided by its staff in the Department of Land Conservation and Development.62
Overall, Oregon’s Goals reflect a mixture of development and conservation objectives, from preserving natural resources and farmland to promoting economic diversification and the orderly provision of public facilities and services.63 Probably the best-known element of the Oregon planning program is Goal 14, “Urbanization,” which requires that every city, regardless of size, have an urban growth boundary.64 Urban uses are to be developed inside the boundary land; urban development is prohibited outside the urban growth boundary,65 even on land neither suitable nor zoned for farming, ranching, or forestry.66
From the beginning, housing affordability was destined to be one of the State’s more important Goals, reflecting in part the important role played by home builder associations in state politics, including a role in the drafting of Senate Bill 100.67
[*PG591] The objective of Goal 10, “Housing,” is to “provide for the housing needs of citizens of the state.”68 Local governments’ land use plans must “encourage the availability of adequate numbers of housing units at price ranges and rent levels which are commensurate with the financial capabilities of Oregon households and allow for flexibility of housing location type and density.”69 The “planning guidelines” for Goal 10 require that:
In addition to inventories of buildable lands, housing elements of a comprehensive plan should, at a minimum, include: (1) a comparison of the distribution of the existing population by income with the distribution of available housing units by cost; (2) a determination of vacancy rates, both overall and at varying rent ranges and cost levels; (3) a determination of expected housing demand at varying rent ranges and cost levels; (4) allowance for a variety of densities and types of residences in each community; and (5) an inventory of sound housing in urban areas including units capable of being rehabilitated.70
Although this directive sounds admirable, most practitioners understand the gap between practice and theory is greater in practice than in theory. Whether Goal 10 would successfully reform exclusionary zoning would not be known until LCDC tried to apply it to the review of land use plans and regulations being prepared by Oregon’s cities and counties.
In 1977, the City of Durham, a small, prosperous white suburb of Portland, adopted an ordinance that reduced the allowable density in its A-1 residential zone from ten to five units per acre, increasing the minimum lot size from roughly 4000 to 8000 square feet.71 Using the [*PG592]appeal provisions in effect at the time, nineteen property owners appealed the City of Durham’s amendment to LCDC.72
In its decision, LCDC explained the anti-exclusionary intent of Goal 10, citing Southern Burlington County NAACP v. Township of Mount Laurel,73 as well as various treatises and publications:
The housing goal clearly says that municipalities are not going to be able to do what they have done in metropolitan areas in the rest of the country. They are not going to be able to pass the housing buck to their neighbors on the assumption that some other community will open wide its doors and take in the teachers, police, firemen, clerks, secretaries, and other ordinary folk who can’t afford homes in the towns where they work.74
LCDC reviewed the record regarding the type and acreage of zoning for lower- and middle-income apartments in its analysis of the ordinance on appeal. Its findings and order did not mince words:
Nothing in the record suggests that Durham, in amending its plan, gave any consideration to low-cost housing needs of its own residents and workers, much less those of the region. The record clearly shows a contrary intent, namely to decrease the diversity of housing types and prices. Such planning runs directly contrary to the purposes of Goal 10.
. . . .
The real spirit and intent of Ordinance 61-77 is, in the Commission’s opinion, embodied in the remarks of its principal proponent, Planning Director John E. Sattler, also a member of the City Council. He stated that one of the main purposes of the change was to discourage “transient” types . . . .
[*PG593]. . . The Commission finds that the Durham City Ordinance 61-77 violates Goal 10 for the reasons set forth herein. The ordinance is therefore invalid.75
By 1981, LCDC’s quasi-judicial decisions and its staff’s established practice in reviewing local plans and regulations76 had been codified into two administrative rules laying out the meaning of Goal 10 and the methods for achieving compliance with the Goal—one rule, which applies statewide,77 and another rule, called the Metropolitan Housing Rule, which applies only to the Portland region.78 The Metropolitan Housing Rule required nineteen of the twenty-four cities, as well as the unincorporated parts of the three counties inside the metropolitan urban growth boundary, to meet combined density targets (namely densities resulting from averaging single-family and multifamily housing) for new construction of six, eight, or ten units per acre.79 The five cities not required to meet a target density for new residential development were exempted because they had populations of less than 2500.80
Shortly thereafter, LCDC’s Goal 10 and the related administrative rules became the basis for statutes adopted by the Oregon legislature. Like Goal 10 and the statewide Goal 10 Administrative Rule, the statutes mandate planning and zoning for multifamily housing, farmworker housing, and manufactured housing.81 The statutes specifically disallow using home rule charters as the basis for excluding multifamily, rental, manufactured, or government-assisted housing from all of a [*PG594]city’s residential zones,82 and gave special attention to both manufactured83 and farmworker housing.84
Implementing Goal 10 to break city-wide exclusionary zoning and the other goals was controversial, but LCDC was given, and used, the power to enforce compliance.85
In the 1990s, LCDC’s efforts to reform exclusionary zoning at the state level were supplemented in the Oregon part of the Portland metropolitan region by Metro, the regional government.86 Metro is governed by an elected council and president87 and has been given broad authority over regional land use planning by the legislature88 (including the power to require local governments to amend their [*PG595]plans and regulations to adhere to Metro’s own planning mandates89) and through its voter-approved home rule charter.90
Metro’s regional framework plan and urban growth management functional plan establish a wide array of housing objectives (including fair-share housing targets for each local government91) backed up by mandated changes to local plans and zoning.92
The implementation of Goal 10 was quickly translated into a fundamental change in residential zoning in the Portland metropolitan region. Between 1978, when the draft regional urban growth boundary was first drawn, and 1982, when Goal 10 had been largely implemented, the total amount of residentially-zoned land had in[*PG596]creased by 10%.93 However, the maximum number of dwellings that could be built under the revised zoning increased from 129,000 to over 301,000 units.94
The minimum lot size in single-family residential zones was reduced from 12,800 square feet—slightly more than one-quarter of an acre—to 8280 square feet.95 This translated into a land cost savings of $7000 for the average-sized single-family house lot in 1982 dollars.96 The amount of land zoned multifamily more than tripled from 8% to 27% of net buildable acreage.97
LCDC insisted on maximum densities in excess of the target densities required by the Metropolitan Housing Rule,98 because it anticipated that site conditions and neighborhood resistance would result in projects built below maximum densities. A decade after LCDC began implementing Goal 10, Portland-area homebuilders collaborated with 1000 Friends of Oregon to research whether the regional density targets were being met.99
The research found that the six cities and part of one county assigned a target density of ten units per net buildable acre for new residential development achieved an actual built density of 9.58 units per acre (81% of the maximum allowable density), whereas the six cities and parts of two counties assigned a target density of eight units per acre exceeded the goal, reaching a net density of 8.42 units per acre.100 The only small city in the study that had been assigned a target density of six units per net buildable acre fell far short, achieving only 3.09 units per acre.101
[*PG597] The size of lots for single-family homes has fallen again in recent years; the median lot size for a new single-family home was 6738 square feet in 1995 and 5132 square feet in 2001.102
How do Mount Laurel Township, New Jersey, and Durham, Oregon compare today, decades after the legal challenges to their exclusionary zoning?
According to 2000 Census data, Mount Laurel’s minority population was 14.3% of the total,103 which is less than half the 34% share for New Jersey.104 Durham’s minority population, at 14.2% of its total population,105 was close to the overall state percentage of 16.5%,106 and slightly more than two-thirds of the Portland regional107 average of 20.1%.108
Multifamily housing makes up 40% of Durham’s housing supply,109 compared with 31% in metropolitan Portland (Oregon portion only)110 and 23% in the state as a whole.111 Rental housing makes up 44% of Durham’s housing,112 which is higher than both the region [*PG598](39%)113 and the State of Oregon (31%).114 In contrast, the supply of rental housing (16%) in Mount Laurel115 is significantly smaller than in Burlington County (23%)116 or New Jersey as a whole (34%).117
Rents follow the same pattern; rents in Durham ($708/month)118 are only 5% higher than the regional median of $671119 and about 14% higher than the State of Oregon as a whole.120 Median rents in Mount Laurel ($939)121 are about 24% higher than in both Burlington County122 and the State of New Jersey.123
The mix of multifamily and affordable housing units in Durham is particularly interesting given that the median value of its single-family housing, which is $248,300,124 is about 40% higher than the regional average, which is $176,565.125
No discussion of Goal 10’s implementation would be complete without presenting data on home prices in the Portland metropolitan region. The National Association of Home Builders (NAHB) Housing Opportunity Index continues to rank Oregon cities as among the least affordable in the nation.126
Yet the NAHB data show that metropolitan Portland, and other Oregon cities, have single-family home prices that are comparable with or lower than the prices paid in similar western metropolitan areas and also have more favorable ratios of average family income to [*PG599]average home price.127 The median sale price of homes in metropolitan Portland, Oregon-Washington, in the first quarter of 2002 was $167,000—compared with $146,000 in Phoenix, $177,000 in Riverside-San Bernadino, $208,000 in Denver, $218,000 in Sacramento, $234,000 in Seattle, and $451,000 in San Jose.128 And the ratio of median family income to median sale price of single-family homes was better in Portland than all of those cities except Phoenix, even though the NAHB ranked Portland as less affordable than Riverside-San Bernadino, Denver, and Seattle.129
A study commissioned by Oregon homebuilder organizations noted that according to the NAHB Housing Opportunity Index, at the time of the study, only 35% of the houses for sale were affordable by the median household income in the Portland area.130 However, a “separate analysis conducted for [the Oregon Housing Cost Study] indicated that in 1998, households classified as having median income (as defined by HUD) could still afford the median house price in Portland . . . .”131
A comprehensive review of academic research on the effect of urban growth boundaries on housing prices was commissioned by the Brookings Institution and prepared by academics from three universities.132 They found little evidence that the regional urban growth boundary increased housing prices, in part because of the other policies reducing zoning barriers to building more houses on the same amount of land, such as Goal 10.133
The impact of Oregon’s reform of low-density exclusionary zoning on urban densities and sprawl is now becoming evident through statistical research and geographic information systems (GIS) analysis.134 The Northwest Environment Watch of Seattle, Washington commissioned research of the three largest metropolitan regions along the north Pacific coast: Portland, Oregon-Washington,135 Seattle-Everett-Tacoma, Washington,136 and Vancouver, British Columbia.137 The research was based on satellite imagery and analyses of 1990 and 2000 U.S. census, and 1991 and 2001 Canadian census data.138
The Northwest Environment Watch study noted that:
Greater Portland sits astride the border between two states, each with its own approach to growth management. . . . As this analysis of US Census data and satellite imagery details, few places in North America illustrate more clearly the consequences of different planning regimes. During the 1990s:
� The population of greater Portland—which includes Multnomah, Washington, and Clackamas Counties, Oregon, and Clark County, Washington . . . add[ed] 376,000 new residents during the decade for a total of just under 1.8 million . . . .
� In the Oregon counties, total population increased by 270,000, and the number of people living in compact neighborhoods (defined as 12 or more people per acre) increased by 141,000 . . . .
[*PG601]� In Washington’s Clark County, population grew by 106,000, and the number of residents in low-density, sprawling areas increased by 78,000. Per capita, Clark County converted about 40 percent more land from rural to suburban population densities than did the Oregon counties. And by the end of the period, Clark County’s residential areas had partially or fully paved over 23 percent more land per resident than the Oregon counties.
� If the Oregon counties had grown in the pattern of Clark County, suburban development would have overtaken an extra 14 square miles of farmland and open space—an area roughly twice that of Forest Park.
� The major difference between Clark County’s sprawl and Oregon’s smart growth was Portland’s growth management policies, which protect open space and foster compact communities.139
The increases in residential density fostered by Goal 10 were almost certainly a contributing factor to the shift to higher transit use in the Portland metropolitan region. Between 1990 and 2000 total passenger miles on the regional transit system, which serves the Oregon portion of the region, increased by more than 57%, more than the 37% increase in the total hours of transit service, and far more than the Portland metro area population growth rate of 23%.140 It was also faster than the 26% increase in vehicle miles traveled on state highways in the region.141 The consequences of this shift from cars to pub[*PG602]lic transportation may include the maintenance of existing good air quality.142
GIS comparisons between metropolitan Portland and other urban areas of similar size, like San Antonio, Texas and Columbus, Ohio, show that Portland’s compact urbanization saved substantial amounts of rural land from development.143 An analysis of past and future development patterns in Oregon’s Willamette Valley, where two-thirds of the State’s population and growth are located, revealed that in the twenty-five years before Oregon passed its comprehensive land use laws, the Willamette Valley’s population grew by 570,000, and that during the same period about 900,000 acres of farmland were lost.144 In the twenty-five years since the adoption of the land use planning goals, the Willamette Valley’s population grew by 670,000 people, but only 105,000 acres of farmland were converted.145
Research conducted by the Pacific Northwest Forest and Range Experiment Station and Oregon State University revealed that the rate of conversion of private farm and forest lands to low- and high-density development in western Oregon slowed dramatically in the 1990s, even though the pace of growth was far greater in absolute numbers than in the 1980s.146
More research is needed to quantify the environmental benefits, and any offsetting detriments, of the reform of exclusionary zoning, within the framework of Oregon’s planning program. But enough evidence at the national level is already available to oblige advocates for affordable housing and advocates for environmental protection to re-examine their attitudes toward each other. In the arena of land use planning, zoning, and permitting, they share a common foe—low-density exclusionary zoning. Replacing that development pattern with something better, for people and the environment, is a natural alliance.147