$15

Executive Summary

More than 100 cities and counties have adopted some form of a growth boundary—a limit on land development beyond a politically designated area—to curb sprawl, protect open space, or encourage the redevelopment of inner-city neighborhoods. Statewide mandates for growth boundaries exist in Oregon, Tennessee, and Washington.

Urban-growth boundaries, however, have potentially negative, if unintended, side effects. By reducing the supply of developable land, for example, housing and land prices could increase, reducing housing affordability and production. Local policymakers and citizens need to understand the nature of these tradeoffs and impacts before they adopt growth boundaries.

Growth boundaries, in fact, have not achieved many of their supporters’ objectives. Their effectiveness has been constrained by:

• Persistent preferences for single-family, detached homes by prospective home buyers;
• Poor coordination among local public agencies;
• Housing-price increases; and
• Political manipulation by antigrowth interest groups.

This study explores the experiences of four cities and regions to more fully elucidate the intended and unintended effects of growth boundaries.

• Portland’s growth boundary was adopted in 1979 as part of Oregon’s statewide growth-management law. The growth boundary, encompassing 24 cities and three counties, has hemmed in development as Metro, the elected regional government, attempts to foster higher density development and redevelop urban areas. Meanwhile, housing prices have increased dramatically. In less than a decade, Portland has transformed itself from one of the most affordable to one of the least affordable housing markets on the West Coast. While inflation as measured by the Consumer Price Index increased by 52.5 percent from 1990 to 1995, lot prices in Portland more than doubled.
• Boulder County, Colorado has had urban-service areas in place since 1978. Combined with the county’s strict growth controls to protect open space, an effective urban-growth boundary exists. While housing prices in Boulder County are 13.2 percent higher than nearby Denver and 23.9 percent higher than Fort Collins, price increases have not been as rapid as in Portland. This appears to be a result of aggressive city-level annexations that bring new land inside the service areas and mitigate housing inflation. If the growth boundary becomes an effective limit line, however, the county will face a deficit of over 20,000 housing units by 2010 at current rates, and the boundary would place significant upward pressure on housing prices.
• Lancaster County, Pennsylvania has one of the nation’s most aggressive county-level growth- management programs. Fueled by a desire to protect the rural character of the county and the flavor of the local Amish culture, local policymakers enacted farmland protection programs (in the 1980s) and growth boundaries (in 1990) to hem in urban development. Nevertheless, almost 60 percent of the county’s land development still occurs outside growth boundaries. Despite a recommended density of five units per acre (one-fifth acre lots), actual densities average three units per acre (one-third acre lots) inside the boundary and less than one unit per acre (more than one acre) outside the boundary.
• Northern California’s experience illustrates how growth-management tools such as urban-growth boundaries can be used by political interest groups to stop new development. In Sacramento County, an urban-service area is being converted into a growth boundary via political resistance to new development even though infrastructure is provided privately. A proposed 5,000 home development in the Tassajara Valley in Contra Costa County in the San Francisco Bay Area was withdrawn as a result of local resistance despite a projected regional housing deficit of 34,000 homes by 2020. County elected officials are considering reducing existing growth boundaries to prevent further development.

Urban-growth boundaries are not the only approach available to create the growth patterns that Smart Growth advocates champion. Market-oriented approaches can produce many of these elements with greater efficiency and with fewer negative consequences. Recommendations for harnessing the incentives and power of the real-estate market to achieve these include:

• Relaxing density restrictions in zoning codes to allow for market-determined densities;
• Purchasing development rights to private land with private funds to protect open space in strategic areas of the city or metropolitan area; and
• Pricing public infrastructure at its full marginal cost to ensure new development pays its way without subsidizing new or existing residents.

Local policymakers need to recognize the political, economic, and social tradeoffs implicit in adopting restrictive land-use policies, including urban-growth boundaries. While local policymakers should avoid subsidizing low-density development, they must also avoid subsidizing high-density development. A market-driven approach is more likely to achieve broad land-use and housing goals than establishing arbitrary limits on land development through urban-growth boundaries.

Introduction?*

One of the hottest planning tools in the nation may be the urban-growth boundary, or urban-limit line. More than 100 cities and counties have adopted some form of an urban-growth boundary—a limit on new growth beyond a politically designated boundary—as a way to curb suburban development or protect open space. Oregon, Tennessee, and Washington have adopted statewide policies that mandate urban-growth boundaries for local governments. In Oregon, a regional-limit line hems in 24 cities and three counties in the Portland metropolitan area. Many cities and counties outside these states have adopted them unilaterally.

Packaged as part of a menu of growth-management policies (sometimes called "Smart Growth" initiatives), limit lines are pitched as a "common sense" approach to managing land development and economic growth. They have become particularly popular as potential ways to preserve farmland and open space: if land development is not permitted beyond a certain point, the theory goes, open space and farmland will be preserved, while existing urbanized areas will experience higher levels of investment and development.

But the consequences of adopting an urban-growth boundary extend beyond the intended goals of preserving open space and farmland. By restricting land development, an effective urban-growth boundary will almost certainly impact the pattern, density, and cost of housing and development. Adopting an urban-growth boundary implies tradeoffs about how land will be used and who can use it. In some cases, such as Portland, these tradeoffs are explicit: an elected regional planning agency, Metro, is consciously trying to shape the metropolitan area from the top down. In Napa, California, a plebiscite in 1973 led the city council to adopt a rural-urban limit line in 1975 to cap the city population at 75,000 by the year 2000.

Urban-growth boundaries, however, may have unintended, negative side effects. Reducing the supply of land for housing, for example, may reduce the amount of housing developed to meet market demand, increasing prices and reducing housing affordability. Using growth boundaries to alter regional land-use patterns also risks eroding the authority of local governments and home rule. In addition, growth boundaries and growth management tools are often implemented in an unrealistic political environment. Public support for growth management is often support for growth control; rather than directing development, grassroots opposition attempts to stop new development completely. Thus, the local debate often focuses on population growth and overall new housing built, not whether the people or homes are in the "right" place at the "right" density.

Despite their apparent popularity, growth boundaries often fail to achieve many of their most basic goals and objectives. Their effectiveness has been constrained by:

• Persistent preferences for single-family, detached homes by prospective home buyers, reducing the ability of growth controls and boundaries to slow the pace of suburban development;
• Poor coordination among local public agencies, creating mismatches between planning goals and actual investments in infrastructure such as roads, sewer, and water;
• Housing-price increases; and
• Political manipulation by antigrowth interest groups, converting a tool intended to manage growth into a vehicle for stopping growth.

Local policymakers and citizens should understand the nature of the tradeoffs and impacts implicit in using urban-growth boundaries as growth-management tools before they adopt them. This policy study examines urban-growth boundaries in four cities, counties, and regions, seeking insight into the diverse nature of growth boundaries and their effects—intended and unintended:

• Portland’s (OR) growth boundary is an example of regional land-use planning and is used to help reshape the metropolitan area into a higher density, more compact, transit-oriented city.
• Lancaster County’s (PA) growth boundaries are tied to an aggressive countywide effort to preserve farmland and a unique local culture.
• Boulder County (CO) is attempting to use city-level growth boundaries and highly restrictive county growth controls to slow development while bolstering its reputation as a high-income satellite community of Denver.
• Napa County (CA) and the San Francisco Bay Area illustrate how growth boundaries and growth-management initiatives fare in a highly charged political environment where land-use issues have been prominent for decades.

These case studies provide important lessons about the effectiveness and limitations of growth boundaries as growth-management tools, including their potential impacts on housing production and affordability. They also highlight important market constraints on the ability to use growth boundaries to alter development patterns given the realities of consumer preferences and local-government decisionmaking. The final section of this study provides a brief review of alternatives to growth boundaries, many of which achieve the same goals with fewer unintended and negative side effects. 

Growth-management initiatives are gathering unprecedented public support nationwide. Two hundred and forty growth-management and "Smart Growth" initiatives were on ballots during the November 1998 election. Seventy-two percent passed. Urban-growth boundaries have been an important part of this movement. In California, for example, most ballot initiatives with growth boundaries passed (see the discussion in Section 6). "I’ve seen what happened in Los Angeles, the San Fernando Valley, and Orange County," observed one city council woman in Contra Costa County, California, across the bay from San Francisco. "When I moved here 20 years ago, I saw the potential for the same thing to happen here." The growth boundary has become popular as a way to stop the spread of low-density residential development in suburban and rural areas.

While some local activists prefer to prohibit development altogether, this is both legally difficult and expensive. Zoning implies a legal right to develop property according to standards established by the local government. Permanent prohibitions would likely constitute a Fifth Amendment taking and require compensation to landowners, an expensive way to preserve open space. Moreover, the freedom to migrate and choose where one wishes to live is an ingrained part of the nation’s political and social culture.

While prohibiting development is not legally or financially feasible, courts give local governments wide latitude to determine development standards and performance levels as long as all the economic value of the land is not eliminated. Thus, state and local governments are permitted to engage in extensive regulation of land development to protect and promote the general welfare. In this context, urban-growth boundaries are increasingly seen as the growth-management tool of choice by local communities. 

An urban-growth boundary is a "line in the land" drawn around an urban area outside of which development is prevented or highly discouraged. Urban-growth boundaries are usually considered long-term growth-management tools, often established for 15- or 20-year periods. Proponents suggest they can accomplish at least six objectives:

• Preserve open space and farmland;
• Minimize the use of land generally by reducing lot sizes and increasing residential densities;
• Reduce infrastructure costs by encouraging urban revitalization, infill, and compact development;
• Clearly separate urban and rural uses;
• Ensure the orderly transition of land from rural to urban uses; and
• Promote a sense of unified community.

The first objective is straightforward and, in principle, achievable. By limiting development outside the growth boundary, open space is preserved outside the urban area. Inside the urban area, the loss of open space accelerates. In fact, if the growth boundary is effective, infill will accelerate along with increasing densities. In Portland, Oregon, the regional-planning agency is deliberately attempting to eliminate farmland inside the growth boundary since it is inconsistent with the goal of separating urban from rural uses and encouraging higher-density residential neighborhoods. Growth boundaries, however, have not always succeeded in achieving their goals and have created unintended side effects, as the cases of Portland and Lancaster County will show.

The second and third goals are more problematic. Proponents of growth boundaries often argue that reducing the amount of land available for development and reducing infrastructure costs is "efficient." Efficiency is narrowly defined by many proponents as "less" in absolute terms. A more precise (and meaningful) definition of efficiency would consider costs tied to producing a particular product of a certain quality. For example, the efficiency of automobile manufacturing is not determined solely by the amount of metal (or plastic) used in production. A midsize family car such as the Ford Taurus requires more metal than a subcompact car such as the Escort. Even among cars of the similar size and class, efficiency needs to be evaluated based on the quality of the product. A Toyota Tercel may use less metal, but the Escort may need more metal to provide certain kinds of amenities or ensure the proper level of engine or body reliability. Efficiency, then, depends on the product and the consumer market.

If the goal of housing is defined narrowly as providing shelter, reducing the amount of land could potentially reduce costs and increase the efficiency of housing production. Growth boundaries would be efficient because they minimize a cost of shelter (without using land). If, on the other hand, the goal of land development is to provide a high-quality living environment, minimizing land might not be efficient. Land may well be an important characteristic of housing that consumers want and are willing to pay for. People may value open space enough to buy it and preserve it as a yard. Tradeoffs between land and other values (such as shelter, parks, shopping, or access to employment) would be necessary.

A limit line ignores these tradeoffs. Instead, it sets up a static perimeter for development independent of what consumers want. If the limit line artificially reduced the supply of housing, it could boost home prices and reduce affordability without a compensating increase in benefits. The urban-growth boundary could well be an anticonsumer growth-management strategy if homeowners preferred larger lots and were willing to pay for them.

The fourth goal addresses more clearly the tradeoffs between land development and other community characteristics. The growth boundary, proponents suggest, is a way to shape communities to produce higher densities and a more compact urban form. The boundary, notes the Greenbelt Alliance in San Francisco, "is a pro-active growth management tool that seeks to contain, control, direct or phase growth in order to promote more compact, continuous urban development." Often this is tied to the belief that compact development promotes "community" and a sense of place (goal six from above). Preventing development beyond existing urban areas connects housing subdivisions physically through roads and other urban services. Physically connecting communities is expected to create a unified sense of place or community. Growth boundaries are thus intended to change the physical characteristics of cities, most often by attempting to transform them into higher density, transit-friendly cities with a distinct urban edge. By limiting development on the urban periphery, growth boundaries are expected to channel growth back into existing urbanized areas. Whether growth boundaries have achieved this policy goal is an empirical question explored later in this report. Whether compact development creates a greater sense of community is more difficult to assess and beyond the scope of this report.

The fifth goal—ensuring an orderly transition between rural and urban uses—recognizes the dynamic nature of land development. This objective was explicitly acknowledged as part of Oregon’s growth-management plan to meet the goals established in its statewide-planning law. In practice, growth boundaries serve many political purposes independent of the transition among land uses. For example, limit lines are used to change urban form (into more high-density neighborhoods) or preserve farmland and open space. Thus, rather than accommodating the dynamic nature of growth, boundaries may be transformed into growth-limiting policies or can evolve into permanent limit lines.

Many advocates would consider growth boundaries successful if they increased residential densities, redirected growth into central cities, reduced the pace of farmland and open-space development, lowered infrastructure costs, and provided for the orderly development of land. These outcomes would have to be attributed to the growth boundary, not ongoing trends in the local economy or other public policies. But these effects, even if they were achievable, do not capture the full range of potential impacts. Other impacts—intended or unintended—must also be considered as part of the decision to limit land development through planning.

For example, urban-growth boundaries require a sophisticated understanding of the real-estate market to avoid increases in housing prices. Researchers at the University of Illinois at Urbana-Champaign refer to it as an "inventory control" problem. The problem for planners and elected officials goes like this: if growth boundaries are successful, they must constrain land supplies sufficiently to encourage higher densities; reducing land, an important cost of new housing, can increase housing prices (therefore reducing affordability) as builders and developers bid up land prices for an increasingly scarce resource; the key is to expand the supply (inventory) of land before land prices increase enough to impact housing prices.

Knowing when to expand the supply of land and by how much is theoretically simple. In the practical world of land development, the decision is rife with uncertainty and imprecision. Despite their popularity, note the University of Illinois researchers, the "optimal" size of the growth boundary and expansions "depend on parameters and relationships about which very little is known."

These parameters are not purely technical. They are inherently uncertain. Real-estate markets change to suit different income groups and tastes. Thus, land-development needs at one point in time may not be the same a few years later, particularly in fast-growing areas undergoing significant economic changes (e.g., moving from a manufacturing economic base to white-collar services). Whether local and regional planning authorities can time the expansions of the growth boundary is both a technical and empirical question that the case studies in the following sections can shed light on.

Another salient issue is whether growth boundaries can accommodate the dynamics of community change. The Portland metropolitan area, as the next section discusses, is going through an economic metamorphosis. Traditionally, the region’s growth has been rooted in natural resources such as the timber industry. Increasingly, local economic growth is driven by high-tech and specialized manufacturing. These new jobs are attracting residents with higher levels of education and income. As the region evolves, the tastes of local residents will also change. Local planning policies need the flexibility to accommodate the needs and demands of new residents as well as existing residents.

Higher density may also result in higher housing prices, less private open space, and increased traffic congestion. As vacant land becomes more scarce, developers and builders will bid land prices up (see the discussion in Appendix A). To minimize costs, builders will use less land—put houses on smaller lots—even when households would prefer larger lots if they were available. If residents continue to use cars as their primary mode of transportation, congestion will increase as higher traffic volumes clog narrow roads in concentrated areas and road capacity does not expand. These potential outcomes should also be part of the growth-management debate as local officials consider adopting urban-growth boundaries.

Finally, local policymakers and planners must consider how establishing a growth boundary may change the dynamics of the local political system. By subjecting land development to a legislative process—where elected officials determine what land will be developed, when, and at what density—the inherently conservative decision-making process of representative democracy may stifle innovation and repress the dynamism of local real-estate markets. The effects may be even more pronounced as "winners" and "losers" are created and new special-interest groups form to intervene in the political process. Groups, particularly antigrowth interests, that previously had little opportunity to impose their beliefs on the larger community may find local growth-management laws uniquely convenient mechanisms for achieving their goals, irrespective of the social costs.

Unfortunately, growth-management and growth-control policies such as urban-growth boundaries are often adopted without a full consideration of these potential tradeoffs and consequences.

Local planning policies need the flexibility to accommodate the needs and demands of new residents as well as existing residents.

Concerns about the unintended consequences of legislative control over land development are not theoretical. Urban-growth boundary supporters generally fall into three groups: those that want to encourage more-efficient land-use patterns (where low-density development is considered inefficient by definition), those that want to preserve open space, and those that want to stop or significantly slow development.

The first group is proactive and prospective. Proponents believe current development patterns are inefficient, and real-estate markets do not lead to proper land use. Supporters of this view believe the tools of government should be used to promote a certain kind of city and neighborhood with specific characteristics, usually compact development, that distinguishes between rural and urban lifestyles. By implication, this group believes strongly in the ability of planning, particularly land-use planning, to accomplish these goals.

The second and third views are more reactionary. Proponents see the growth boundary as a way to protect farmland and open space by excluding most other uses. Often, the urban-growth boundary is seen as a permanent wall, outside of which development will not (or should not) be allowed. Environmental activists and other antigrowth groups, for example, are attempting to convert Sacramento County’s urban-service area into a growth boundary to prevent further land development. Opposition to a proposed 1,900-acre, 3,000-home, mixed-use development called Deer Creek Hills may potentially serve as the vehicle for this conversion. 

How Urban-growth Boundaries Differ from Urban-service Areas

Growth boundaries are distinct from urban-service areas.¹ Service areas are determined (theoretically) by objective information about a local government’s costs to extend roads, water and sewer lines, or other publicly provided services. Beyond some point, the county or local government determines that the extension of those services is not cost effective.² Urban-service areas apply to public infrastructure and utilities and reflect decisions about the cost-effectiveness of extending these services into new areas.³

Urban-growth boundaries are explicit attempts to channel growth for broader political purposes and goals. In principle, a service area can change if improved efficiencies in service provision allow for the expansion of the region it serves, or natural market forces increase densities to the level where their extension is cost effective. Service areas also do not foreclose alternative forms of development. Growth boundaries, in contrast, attempt to place some land off limits to development irrespective of public-service costs and efficiencies. Development will be discouraged or even prohibited beyond the boundary (if legally possible), even when services are privately provided.

Maryland’s recent Smart Growth initiatives provide an example of how state policy can impact land development through indirect effects on local-government decisionmaking. Maryland adopted a limited form of a growth boundary in 1996 when it created Priority Funding Areas (PFAs). PFAs act as growth boundaries but apply only to state infrastructure-financing decisions. In principle, local governments and the private sector could provide infrastructure and develop property in non-PFAs. In practice, this is unlikely. Private developers still need to obtain development permission and planning approval from county governments and local cities, which would likely view development outside PFAs as inconsistent with their comprehensive plans. These projects would likely be rejected. Thus, in practice, Maryland’s PFAs serve as growth boundaries for state and local governments (although PFAs are determined at the county level and approved by the State of Maryland).

Notes:

This distinction differs from other interpretations. Some planners have suggested that service areas and growth boundaries are virtually the same since public infrastructure spending determines the location and pattern of development. See the discussion in V. Gail Easley, Staying Inside the Line, Planning Advisory Service Report Number 440 (Chicago, Illinois: American Planning Association, 1992), pp. 2-6.

²In principle, this is really a "pricing" problem: extending infrastructure services is not cost effective because the public agency is unwilling to charge the full cost of extending the service. For a more complete discussion of this issue, see Samuel R. Staley, The Sprawling of America: In Defense of the Dynamic City, Policy Study No. 251 (Los Angeles: Reason Public Policy Institute, January 1999), pp. 29-42; Tara Ellman, Infill: The Cure for Sprawl? Arizona Issue Analysis 146 (Phoenix, Arizona: Goldwater Institute, August 1997), http://www.urbanfutures.org/p82897.html.

³The Minneapolis-St. Paul Metropolitan Council (Met Council) administers one of the nation’s longest-running urban-service areas. Recently, political pressures have increased to convert the urban-service area into an urban-growth boundary as a mechanism for achieving Portland-style regional planning. See Myron Orfield, Metropolitics: A Regional Agenda for Community and Stability, rev. ed. (Washington, D.C.: Brookings Institution, 1997), especially Chapter Nine and Appendix. 

Deer Creek Hills was explicitly designed to satisfy environmental-activist concerns about water usage and land preservation. The proposed project was a master-planned, age-restricted, golf-oriented retirement community that would have included 20-acres of medical, professional, and community services, and an 80,000 square-foot shopping center. Seventy percent of the land was dedicated to open space and recreational land. If Deer Creek Hills "were inside the urban service boundary," says Tom Hutchings, Director of Sacramento County’s Planning and Community Development Department, "everybody would be falling all over themselves to approve it. It would have been an expedited project and it would have been built by now." Instead, the future of Deer Creek Hills is uncertain. The Sacramento County Board of Supervisors refused to approve the project in part because it lies outside the service area (even though it would not have impacted infrastructure services in Sacramento County). By denying development permission, opponents are successfully converting the urban-service area into a de facto growth boundary. The lesson for the land developer is that an urban-growth boundary prevents any type of development, "no matter how well you plan, no matter how well you breathe life into the vision."

Not surprisingly, antidevelopment groups who support growth boundaries encourage the adoption of the most-restrictive types. "If you’re looking to lock in your [urban-grown boundary] for the longer term, [urban-growth boundary] by voter approval is the better way to go," notes the Greenbelt Alliance, an environmental advocacy group based in San Francisco. City councils can be changed when new members take office. Initiatives and referenda are far more permanent since changing the boundary requires another ballot drive. Ballot-box growth-management initiatives are also becoming more popular, growing by more than 50 percent between 1996 and 1998.

While growth boundaries are promoted as growth-management tools, they are often adopted as part of a more general antigrowth movement on the local level.

Tradeoffs among political goals and outcomes are rarely clear in the political process when voters decide whether or not to adopt growth boundaries. When they are adopted at the ballot box, the effects are unknown to voters—limits are placed on future residents and homeowners. In addition, enough land is usually included at the outset so that short-term impacts on housing prices are small. Urban-growth boundaries may serve as political vehicles for claiming to preserve the physical character of the community (an explicit goal in Lancaster County), rather than as a dynamic growth-management tool tied to a progressive, evolutionary vision of the community.

In addition, growth boundaries are rarely used in isolation. They are often accompanied by strong planning provisions that limit development through state or local ordinances. Growth-boundary proposals are increasingly accompanied by ballot-box initiatives that restrict the discretion of local governments to modify or expand the boundary. In California, for example, several cities in the San Francisco Bay Area are proposing that new developments that include ten or more housing units be subject to voter approval. Thus, while growth boundaries are promoted as growth-management tools, they are often adopted as part of a more general antigrowth movement on the local level. The following case studies illustrate these points in more detail.

While many growth-management advocates believe urban-growth boundaries can be used to channel and redirect new development, antigrowth interest groups are increasingly finding them convenient ways to stop any development beyond existing city boundaries. 

The most sophisticated and comprehensive attempt to implement urban-growth boundaries is in Oregon. The Portland effort, in particular, provides an excellent case study for examining the intended and unintended effects of growth boundaries. Growth boundaries in Oregon were intended to be the linchpins of regional-growth management and explicitly addressed the six main objectives listed in Part 2 of this study. The limit line has emerged as a key component of a regional strategy intended to increase housing density, encourage compact development, and redirect investment into the urban core. Its comprehensive design and implementation provides a unique opportunity to examine its impacts on a regional scale.

While considered by many as one of the most successful examples of growth boundary implementation, the Portland case also reveals many of the pitfalls and unintended consequences that can result from their application. More specifically, the Portland experience provides a useful perspective on how growth boundaries may start with one purpose and then be transformed into a vehicle for achieving new and sometimes unanticipated objectives. The potential impact of a growth boundary on housing prices is also becoming clear in Portland. 

In 1973, Oregon became the second state in the nation to adopt a statewide growth-management law. Growth boundaries were added in the mid-1970s as a mandate for all cities and metropolitan areas. Growth boundaries were considered the "cornerstone" of the state’s growth-management initiative, both as a way to ensure the orderly transition from rural to urban uses and to protect Oregon’s farmland. By 1986, all 242 cities and 36 counties had comprehensive plans and urban-growth boundaries in place, conforming to 19 state goals established by Oregon’s growth-management law. The Portland-area growth boundary (established in 1979) covers three counties and 24 cities and has been used to aggressively promote infill and increase housing density. Hundreds of policymakers and planners from across the nation have visited Portland to see how growth boundaries and regional planning work. "On the whole," notes the Greenbelt Alliance, echoing plaudits from numerous other sources, "the [urban-growth boundary] has been a huge success."

Evaluations of Oregon’s growth boundaries are hampered by the fact that the Oregon state-planning system did not establish a program for collecting data to systematically evaluate whether the programs were meeting state goals until the 1990s. The problems this oversight poses are particularly evident in examinations of the effects of growth boundaries on farmland protection (see box on page 13). A task force established by the Oregon Department of Transportation initially attempted to gather data on ten cities to evaluate land-use patterns inside urban-growth boundaries, but settled on three, significantly limiting the ability of researchers to generalize their results beyond the specific case studies. Despite these data limitations, a substantial body of research has emerged that attempts to evaluate the effectiveness of growth boundaries and other land-use planning techniques in Oregon.

Ironically, despite its perceived success outside the state, Oregon’s land-use planning system fell sufficiently short of its goals that many planners and policymakers argued for "mid-course corrections" as early as the mid-1980s. Actual densities inside the growth boundaries ranged from 33 percent to 75 percent of the level permitted by local comprehensive plans. In Portland, which had the highest-allowable densities, actual densities were one-third lower than those permitted by local land-use plans. While just 5 percent of residential housing units were built outside Portland’s growth boundary from 1985 to 1989, the proportion ranged from 24 percent to 57 percent for other cities and regions in Oregon. 

Despite increases in density, the amount of vacant land available for new development inside the Portland regional-growth boundary has fallen from 75,000 acres in 1985 to less than 55,000 today. Almost 40 percent of the land in the boundary was vacant in 1980. By 1997, the amount of vacant land represented just 19.8 percent of the land (Figure 1). At current trends, without an expansion of the boundary, the Portland metropolitan area will experience a 42,060 housing unit deficit by the year 2017. Even if densities increase to achieve those recommended in the Metro 2040 Plan, the housing deficit would be 8,590 units. 

Metro and local governments in the region have planned for adjustments in the growth boundary to account for future growth through the "2040 Process," a strategic transportation and land-use planning initiative intended to accommodate and allocate growth among the region’s cities and urban areas. Oddly, metro planners and officials are hesitant to view the decision to expand the growth boundary in terms of housing markets and prices. Rather, they are using the narrow legal view that the boundary should include a 20-year supply of land and should preserve farm and forest land irrespective of impacts on housing affordability.

Local politics is also making expansion of the growth boundary difficult. The Metro Executive Officer's recommendation for 7,000-acre expansion was rejected by a coalition of five of the seven council members in 1997. Metro voted to expand the boundary in December 1998 by 3,500 acres and passed a resolution to add 1,800 additional acres in the future. These votes were a compromise between environmental activists and zero-expansion advocates and prodevelopment groups who favored an expansion of at least 10,000 acres. The 5,300-acre anticipated expansion of the boundary accommodates less than two years of the average land take-up through development. Efforts to improve housing affordability through strategies such as fair-share housing or inclusionary zoning are unlikely to be effective given current growth trends (see box on page 22).

As Figure 2 suggests, as less new land has become available for development, the increased competition for existing developable land has translated into higher land prices. For a brief period (1988 to 1990), during the regional recession, housing prices stabilized along with the proportion of vacant land available for development. After 1990, prices rose significantly as the difference between the median price for a one-family home in Portland went from 71.7 percent of the national median in 1988 to 100.7 percent in 1994.

These effects are evident in suburban Washington County, the Portland metropolitan area’s second largest county. Lot prices for single-family houses in Washington County lagged inflation, as measured by the CPI from 1985 to 1990, the years the Portland area experienced a housing recession (Figure 3). After 1990, housing prices increased significantly. By 1994, home prices were one and one-half times greater than in 1985 (a 140 percent increase). Lot prices more than doubled in five years while the Consumer Price Index (CPI) increased by 52.5 percent. These trends significantly surpassed Metro’s housing-price forecast, which predicted land prices would rise by 20 percent in real terms from 1995 to 2000.

The housing price data in Figures 1, 2, and 3 are not adjusted for the quality or size of homes. Densities increased on average from five homes per acre (one-fifth of an acre) to eight homes per acre (one-eighth of an acre) from 1994 to 1997. Land absorption has been declining steadily as multifamily housing units have increased from 25 percent of all building permits in 1992 to 49 percent in 1997. Thus, to the extent Portland families prefer detached single-family homes with private open space in the form of yards, housing quality may well be declining as housing prices per unit increase.

These pressures are unlikely to moderate in the near future. Metro’s policy is to encourage redevelopment of existing land at higher densities and encourage the development of vacant land in existing urban areas. These goals may only be possible if home prices increase significantly. Moreover, higher "refill" rates may not be able to compensate for a reduction in overall housing production (see Appendix A). 

A glimpse of the growth boundary’s potential impacts on land and housing prices was evident soon after the boundary was established. In 1980, an analysis of 455 purchases of vacant lots for single-family homes found that land prices inside the boundary were significantly higher than those outside the boundary. The changes in market price were tied to expectations by builders and developers about the likelihood the land could be developed for residential purposes. Thus, land prices varied by how much local governments restricted development. Rural land values outside the boundary fell as developers recognized its availability for urban development was limited. Land values inside the boundary increased as developers recognized its potential for development. Thus, strict regulatory adherence to the growth boundary resulted in the largest differences in price.

Despite the increases in land prices, early evaluations of Portland’s growth boundary found little initial impact on housing prices. Housing prices in the 1980s, noted one planner, followed national trends (although prices inside growth boundaries tended to be higher than those outside). The Portland housing market, in fact, was depressed in the 1980s, hit by the national recession and low timber prices. Thus, low demand offset upward pressure on housing prices from the higher cost of land induced by the growth boundary. By the late 1980s and early 1990s, the economy had picked up, and housing prices began to rise primarily in higher-income communities such as Lake Oswego, West and Northwest Portland, Tigard, and Wilsonville (Table 1). Population growth rates have tripled relative to the early 1980s, and land prices have doubled since 1990. While nominal housing prices have risen by more than 60 percent regionwide since 1990, inner-city Portland experienced little investment and housing-price appreciation until the 1990s.

From 1990 to 1995, inner-city neighborhoods in Portland experienced a housing renaissance. The average home price among these cities increased from $97,684 to $152,700.

From 1990 to 1995, inner-city neighborhoods in Portland experienced a housing renaissance: the North, Southeast, and Northeast Portland areas saw the greatest rate of housing-price appreciation in the region (although they started from a lower base price). Home prices in North Portland doubled, rising from $41,300 in 1990 to $83,800 in 1995 (in noninflation-adjusted dollars). The average home price among these cities increased from $97,684 to $152,700.

For those favoring growth boundaries in Portland, these trends are an indicator of success. As land becomes more scarce on the fringe inside the boundary, housing demand is pushed inward. Increased land prices, in turn, create incentives for higher-density development in suburbs and inner-city neighborhoods. As suburban neighborhoods achieve higher densities and become less distinguishable from inner-city neighborhoods in physical form, higher-income households compete with lower-income households for inner-city locations. The urban-growth boundary has pushed investment inward and promoted inner-city development, a key goal of urban-growth boundary proponents.

But, housing-price appreciation is a double-edged sword. On the one hand, higher prices may reflect greater demand for a scarce commodity (e.g., quality inner-city housing). On the other hand, higher housing prices may reflect a constraint on supply. In the latter case, higher housing prices reduce the overall quality of life for residents, since they must pay more for a home that may provide fewer benefits (e.g., smaller lots, more dense living).

While a number of factors contribute to rising home prices in Portland (see box on page 20), the growth boundary is increasingly recognized as an important supplyside constraint in the local housing market. Unlike the early- and mid-1980s, low densities and abundant land inside the boundary are not softening the impact of the limit line on housing prices. By restricting the supply of vacant land and forcing development into higher cost, inner-city locations, the boundary is contributing to Portland’s housing-price appreciation and potentially reducing overall housing quality by making homes less affordable. In the current boom, the growth boundary has become a binding constraint even though it is one of the few tools local policymakers can use to mitigate housing-price effects (by expanding the boundary to include cheaper land).

Homebuyers Trade off Lot and Building Size Based on Cost

Consumers make numerous decisions about the quantity and quality of housing when they decide when to buy a house. This is a dynamic process: consumers are continually evaluating different characteristics of a home such as the number of bedrooms, square footage, lot size, and other attributes while suppliers (builders) are continually trying to develop property to meet these needs. The lot size/building size tradeoff illustrates this process and the dynamism of real-estate markets.

Land is an important beneficial characteristic of a home. Sometimes this land is privately owned (e.g., back and front yards), and sometimes it is public open space (e.g., a neighborhood park). Not surprisingly, when land is abundant, people prefer homes with larger lots. As housing prices increase, homebuyers recalculate the benefits of different housing characteristics. Since houses usually provide important, high-priority benefits—bedrooms for personal privacy, shelter from the elements, kitchens, etc.—homebuyers will often choose smaller lots when home prices are high.

Portland, Oregon provides an interesting case. When Metro, Portland’s regional planning agency, began to implement its land-use plan, maximum allowable densities increased. In other words, developers were allowed to build more houses per acre of land. During the 1990s, Portland experienced rapidly increasing housing prices as immigration and rising incomes fueled demand for housing, and supply failed to keep pace (see the discussion in Part 3). Researchers at Metro examined almost 8,000 home sales from January 1996 through June 1997 to determine how households responded to higher housing prices and whether the lot size/building size tradeoff existed.¹ Portland-area homebuyers, it turned out, bought larger homes on smaller lots as prices increased.² They did not buy larger lots as home prices increased for houses of the same size. For homes of different sizes, researchers also found consumers trade off lot size for home size.³

Notes:

Sonny Conder and Karen Larson, Residential Lot Values and the Capital-Land Substitution Parameter—Some Recent Results from the Portland Metro Area, Growth Management Services Division, Metro, unpublished paper, May 1998. The sample consisted of 5,500 records from Clackamas, Multnomah, and Washington Counties in Oregon and 2,200 records from Clark County, Washington.

²Ibid., p. 6. Conder and Larson speculate that this may be the result of smaller household sizes and the increase in multiple-income households.

³Ibid., pp. 6-7. The estimates ranged from 0.642 to 0.8, where 1 represents a 1:1 willingness to trade off lot size for home size and 0 represents no substitution. In other words, a ratio of 1 would imply that, as home prices increase, a 10 percent increase in home size would be equivalent to a 10 percent reduction in lot size. 

Metro is also experiencing significant resistance among grassroots groups to higher-density residential development and the proposed transportation plan. For example,

• Portland residents rejected a referendum to fund the region’s light-rail extension in November 1998; and
• Residents of the suburban community of Milwaukee recalled all city council members that voted to accept Metro’s high-density zoning mandate to accommodate future population growth. 

Population Growth, Migration, and Housing Prices

A number of factors in addition to the growth boundary are probably contributing to the boom in housing prices, including: the national decline in interest rates since 1989; the precipitous decline in property taxes in Oregon begun through a 1990 ballot-box initiative; Portland’s rapid employment growth; investment in new high-tech businesses; and rising household incomes.

Many have argued that rapid migration into Oregon during the 1990s created substantial housing demand with which the local building industry has simply been unable to keep pace. On average, 35,000 people per year have moved into the Portland area in the 1990s, triple the rate of pre-1988 levels.¹ As much as two-thirds of Portland’s population growth may be attributed to migration from other states, particularly California.² Since a substantial portion of this migration was a result of job creation in high-tech industries, higher-than-average incomes may be fueling housing demand as well.³ Most regional economies will experience difficulties when facing an unexpected increase in population and housing demand.

These effects, however, should be temporary as the building industry adjusts to higher levels of demand and new housing supply comes on line and prices moderate. For the Portland metropolitan area, population is increasing between 2.0 percent and 2.5 percent per year.⁴ While larger than the national average, this is less than other metropolitan areas such as Las Vegas (about 6.5 percent per year), Phoenix (about 3.0 percent), Riverside (about 2.7 percent), and Orlando (about 2.7 percent). These fast-growing cities did not experience significant housing price appreciation during the same period. While Portland’s housing prices increased by 61.5 percent from 1990 to 1995, housing prices in Las Vegas—the fastest growing city among the cities listed—increased by about one-third as fast (22.0 percent during this period).

Notes:

Portland Metro, Urban Growth Report Addendum, August 1998, Figure 3, p. 7.

²Portland Metro, Urban Growth Report, December 1997, p. 42.

³For a discussion of projected trends in household income, see Portland Metro, Urban Growth Report, pp. 51-52.

⁴Portland Metro, Urban Growth Report, p. 42. 

These events and others suggest that anti-Portland and anti-Metro votes exist and may be mobilized.

Moreover, the coalition of interest groups supporting and opposing regional planning is shifting. Initially, low-income households supported statewide planning because the law included a plank supporting the housing needs of all residents in the state. Many believed a properly planned region would yield affordable housing.

As of December 1998, all four of Oregon’s metropolitan areas ranked among the nation’s least affordable housing markets, according to the National Association of Home Builders, which tracked home prices in 186 metropolitan areas throughout the nation (Table 2). The homebuilders examined interest rates, median-family income, and home sales in each metropolitan area to determine what proportion of homes could be purchased by a family with the median income in the metropolitan area. In Eugene, the least-affordable housing market among Oregon’s metropolitan areas, a family earning the median income could only afford to buy 30.4 percent of the homes sold (in the last four months of 1998). When Oregon’s metropolitan areas are compared to 48 areas in western states, Eugene ranks among the bottom 3 percent in housing affordability, Portland and Medford among the bottom 10 percent least affordable, and Salem among the bottom 25 percent least affordable.

In the current environment of rapidly rising housing prices, political resistance to expanding the growth boundary, and efforts to increase residential densities in existing neighborhoods, low- and middle-income households may be less supportive of regional planning. Similarly, the emergence of hobby farmers as a significant voting block opposing expansion of the growth boundary can potentially expose inequities created by farmland-preservation efforts as noncommercial farms owned by wealthier households are subsidized by lower land values outside the growth boundary. Thus, regional planning has created new "winners" (e.g., hobby farmers) and "losers" (e.g., lower- and middle-income families), and how these and other groups impact the land-use planning process will determine the fate of Portland’s regional-planning system.

Oregon farmers may also become even more divided over the growth-boundary issue. Land-use planning was sold as a way to bolster the economic viability of farming by guaranteeing permanence and "protecting" farmers from urban encroachment. By prohibiting specific types of uses, such as residential and commercial development, land prices would fall for farmland (since demand would fall), making it easier for farmers to acquire and maintain land in agricultural activities with relatively low economic value added. While early results indicate some of these goals may have been achieved, a policy of minimum boundary expansion creates additional problems. To keep land-use regulation less costly for farmers, and to insure support from the agricultural community, farmers retained the right to develop their properties should their land be included in the growth boundary at a later point. In the future, the agricultural community is likely to divide farmers who own their land (and want to see their land reach its highest-market value) and farmers who lease their land and fear that development threatens their livelihood. The farming implements and supplier lobby will also resist development of farmland in order to preserve its customer base.

Growth versus Affordable Housing in Portland

Metro and local-government officials have been quiet in discussing the rise of housing prices in inner-city neighborhoods and their gentrification by higher-income groups. Housing prices in inner-city areas such as Southeast Portland, Northeast Portland, and North Portland increased by 85 percent, 78 percent, and 103 percent, respectively between 1990-95, while the suburban communities averaged 45 percent. Inner-city gentrification, while a laudable planning goal, creates a burden carried mainly by Portland’s poor. As they are displaced by higher-income families, poorer households are less able to find better housing on the urban fringe and in suburban areas.

To respond to affordable-housing concerns, advocates of zero-expansion, headed by 1,000 Friends of Oregon, a leading environmental interest group, formed the Coalition for a Liveable Future. The coalition sought to merge the efforts of environmental interests with low-income housing advocates. The coalition was seen as a vehicle by environmentalists to prevent low-income housing advocates from forming a coalition with developers that favored expanding the boundary to increase the supply of housing and mitigate housing-price appreciation. To attract the housing advocates, the coalition added planks to their agenda that linked local low-income housing trust funds to real-estate transfer taxes and land-sales taxes and supported an inclusionary zoning mandate on new-housing development. Both planks appealed to the environmental community by reducing the likelihood that the metropolitan area would expand. Developers on the urban fringe would be required to build at inner-city densities, pay for local-service expansion, build low-income housing, or pay funds into a low-income housing trust fund.

Inclusionary zoning is unlikely to significantly improve affordable housing. Portland has considered requiring developments over ten units to provide 20 percent of their units at a price affordable to households earning 80 percent of the region’s median income.¹ From 1990 to 1997, local real-estate market consultant Jerald Johnson reports that 6,450 single-family homes and 3,530 multifamily units were approved.² The inclusionary-zoning program, if enacted, would have required that 1,032 single-family homes and 565 multifamily apartment units be priced at "affordable" levels.³ Yet, from the second quarter 1995 through second quarter 1997, Johnson estimated that housing-price appreciation alone pushed 80,000 single-family homes over thresholds for affordability. In other words, 80,000 fewer units were considered "affordable" in 1997 compared to 1995. Thus, programs such as the one considered by Metro would make little headway in improving affordability in the Portland metropolitan area.

Nevertheless, the economic effect of these policies would be to reinforce housing-price appreciation in the Portland area. High-density housing is more expensive to build per unit for the same size and quality (see Appendix A),⁴ and the burden of impact fees and low-income housing subsidies would be shared by reduced returns on investment for land owners and higher rents and purchase prices for housing consumers. By raising the cost of building new homes, this policy would raise the market price of housing in the region and create additional burdens on low-income and middle-income households.⁵

Notes:

For homes, the affordable housing threshold was pegged at $125,000. For apartments, the threshold was $500 per month. Jerald Johnson, "Issues Associated with the Imposition of Inclusionary Zoning in the Portland Metropolitan Area," Hobson, Johnson, and Associates, Portland, Oregon, unpublished paper, December 1, 1997.

²Ibid.

³Ibid. When this report went to press, Portland Metro had not adopted this policy.

⁴Higher-density houses tend to be smaller, thus increasing the cost per square foot of living area (see Appendix A).

⁵For an alternative approach to housing affordability, see Howard Husock, Repairing the Ladder: Toward a New Housing Paradigm, Policy Study No. 207 (Los Angeles: Reason Foundation, July 1996). 

As Portland’s growth boundary evolves beyond its first generation of implementation, support for growth management is emerging as special-interest support for growth control. The urban-growth boundary in Portland appears to have transformed itself from a growth-management tool, where growth was guided or channeled, to a growth-control strategy, where Metro sets density requirements and actively directs growth through infrastructure planning and land-use regulation. Metro is increasingly moving toward an approach where the planning agency sets, even mandates, densities and infrastructure policies.

Growth management in the Portland region appears to be politically unstable, as new interest groups and coalitions have emerged to support specific aspects of the regional plan. The Portland region has experienced rapid population growth and a significant rise in housing demand. As land inside the growth boundary becomes increasingly scarce, rising housing demand has led to significant increases in housing prices. These effects are compounded by the fact that density targets have not been met within the boundary. Portland still retains many of the characteristics of suburban living that the growth boundary was intended to discourage. Ironically, achieving these density targets may only be possible through high land prices that make expensive and inefficient parcels of land profitable to develop.

Old and new special-interest groups are using the growth boundary to create a static line in the land around the metropolitan area.

The political and economic requirements for greater land supply were partially addressed by a minor expansion of the urban-growth boundary in 1997. However, the push to expand the region's growth boundary has been led more by the development community than those most affected by housing-price escalation. That expansion decision has been hamstrung by the requirements placed on the housing to be built. While some believe a substantial amount of land still exists within the urban-growth boundary for land development, Metro is predicting a housing deficit even if significantly higher densities are achieved on current land. In addition, the remaining land inside the boundary is generally less productive and more expensive to develop. Combined with a political climate supportive of zero growth, Metro will not likely expand the growth boundary significantly to moderate upward pressure on housing prices.

In sum, the Portland case illustrates several potential pitfalls for local policymakers considering growth boundaries. While farmland loss has moderated, and densities have clearly increased, housing affordability has suffered. Housing appreciation is destabilizing the local political environment as growth-management goals conflict and new interest groups and coalitions emerge to challenge the system of regional planning. While the boundary was initially intended to be a dynamic tool that would guide growth, old and new special-interest groups are using the growth boundary to create a static line in the land around the metropolitan area. These outcomes, albeit unforeseen and unintended, must also be considered when evaluating growth boundaries. 

Communities often implement growth-management plans or growth controls when they face historically high levels of growth and development. Boulder County, Colorado experienced significant growth in the 1970s, prompting local community leaders to adopt a comprehensive plan that now provides an example of a hybrid urban-growth boundary. In contrast to Portland, Boulder County has not experienced a rapid escalation of housing prices. This may well be a function of the localized and flexible (and broadly market driven) nature of growth-management policies implemented at the city, rather than regional, levels. Ironically, the boundaries may avoid the pitfalls of growth boundaries precisely because they do not function as true growth boundaries. Decisions about boundary expansion are locally driven by cities, which provides an unintended relief valve for the local housing market, and fail to truly constrain development on the fringe. In fact, significant regional efficiencies may be emerging because land use is politically determined, and economic diversity and balance are not important local political priorities. 

Local officials began to examine growth-management tools to help preserve open space and the rural character of the county in the 1970s when the county’s population began to expand rapidly. Boulder County’s population has grown dramatically and steadily since the 1960s (Figure 4), a result of its proximity to the diverse work force in the Denver metropolitan area and the growth of the University of Colorado-Boulder (see box on page 25). In the 1960s, the county’s population grew 77.6 percent to 131,889 people. Growth rates began to slow in subsequent decades, dropping to 43.8 percent growth in the 1970s and 18.8 percent in the 1980s. Since 1990, however, the county has added another 47,161 people (20.9 percent), increasing the county population to 272,500. By 2010, the Boulder County planning department expects the county’s population to expand by another 29.4 percent to 352,605.

Public officials and planners in Boulder County use community-service areas—boundaries within which infrastructure such as roads, sewers, and water will be publicly provided—and restrictive county ordinances and regulations to ostensibly protect open space, increase urban densities, preserve the quality of life, ensure the efficient provision of infrastructure, and solidify a sense of place and community. A growth boundary was created through the simultaneous implementation of community-service areas at the city level (most notably around the City of Boulder) and highly restrictive land-use and permitting policies at the county level.

Boulder County’s Access to Denver Drives Growth

Boulder County is located northwest of Denver. The City of Boulder, the county seat and largest municipality, is just 26 miles from Denver with easy access via a limited-access highway (U.S. Route 36). Most of the county’s population lives on the eastern side, in the cities of Boulder, Longmont, Broomfield, Lafayette, and Louisville. The bulk of the county consists of Roosevelt National Forest and the Arapahoe National Recreation Area. About 45 miles northeast of Boulder is Fort Collins, another fast-growing suburban community with easy access to Denver via Interstate 25 (which is 60 miles south of Fort Collins).

 

In 1978, Boulder County officials adopted a comprehensive plan that established community-service areas for each of its municipalities. Technically, the community service areas are not growth boundaries, although they were adopted to achieve "the basic planning goal of encouraging all future urban development to occur in and around existing urban centers." Infrastructure is provided at the city level in Boulder County. Thus, the county does not play a significant role in providing water, sewer, or other urban services.

The community-service areas and county-comprehensive plan were adopted to explicitly address the issues of rapid urbanization, suburban development, and fragmented local government. Since community-service areas are city based, the county comprehensive plan recognizes that "municipalities and not the county are best able to provide, at realistic, fair costs, the types of public services an urban society requires." This principle is reinforced by the explicit acknowledgement in the plan’s goals that existing communities should be allowed to grow at "whatever rate they consider desirable." Thus, unlike cities in California, which must receive approval from a regional council of governments to annex land, cities in Boulder County can expand their service areas unilaterally through annexation. This policy has inadvertently weakened the ability of the community-service areas to act as true urban-growth boundaries, as the following section will show.

Boulder County’s comprehensive plan explicitly discourages development outside community-service areas. In Colorado, property owners can develop property without planning review if the density is one housing unit per 35 acres, and Boulder County strictly enforces this rule. Thus, by adhering to the one dwelling unit to 35-acres rule, the county attempts to preserve the "agricultural, environmental, visual, and cultural characteristics that make up the rural character of the county’s unincorporated areas."

Municipalities and not the county are best able to provide, at realistic, fair costs, the types of public services an urban society requires.

Housing prices in the Boulder metropolitan area are higher than in nearby Denver and Fort Collins. Boulder County’s home prices may be influenced by higher incomes, a product of a thriving local economy and the University of Colorado-Boulder. Higher incomes can fuel higher housing prices, a claim often made in Portland (see the discussion in Part 3). With a median-family income of $63,100, incomes in Boulder are 13.2 percent higher than in the Denver metropolitan area ($55,700) and 23.9 percent higher than in Fort Collins ($50,900). According to the National Association of Homebuilders, 63.4 percent of the new and existing homes in the Boulder metropolitan area could be purchased by a family earning the median income, the same proportion as in Fort Collins but lower than Denver (67.9 percent). While the average sales price of a home was nearly the same in Denver and Fort Collins (about $150,000), 1998 home prices in Boulder County averaged $180,000, about 20.0 percent higher.

An offsetting factor in Boulder County may be aggressive annexation policies by other cities within the county that have enabled homes to remain affordable to nearly two-thirds of the families in the county. Annexation increases the supply of buildable land on the fringe, which usually costs less than land closer to the cities (see Appendix A). Thus, cities are able to keep land costs from placing significant pressure on housing prices, facilitating low-cost building on vacant land.

The City of Boulder, for example, annexed 3,776 acres, or 24.1 percent of its total land area, between 1990 and 1997 (Table 3). The City of Longmont, the county’s second-largest city, annexed 3,712 acres (one-third of its total land area) during this period. In fact, most cities in the county have aggressively annexed land, increasing the supply of land for developable purposes. Overall, cities in Boulder County brought 11,598 acres of land (22.9 percent of the total) inside their borders, suggesting a substantial willingness by local communities to accommodate future development and growth. Without these annexations, restrictive development controls in the county would shift housing demand to cities outside the county. Thus, negative housing impacts of restrictive growth policies are mitigated by more market-driven policies at the city level within and outside the county. Rather than hem in development, local cities annex new land and, by extension, expand the community-service area.

Since real-estate prices are driven in part by builder expectations about future land availability, many builders recognize that annexation will likely result in significantly more housing units added to the growth areas to accommodate housing demand. Overall, local city annexations will help accommodate 24,000 new single-family and multifamily residential units that can house another 58,100 residents.

The likely impacts on housing supply become apparent when the number of potential housing units is compared to projected population growth. Planners estimate that Boulder County will grow to 352,605 people by 2010, a 56 percent increase since 1990 (Table 4). Without the annexations, most of this growth would either be accommodated through "refill," the redevelopment of existing sites at higher densities and development of vacant land in current urban areas, or the exportation of housing demand to other communities. These land-supply constraints would inevitably put upward pressure on land and home prices since redevelopment and infill tend to be more costly than building on vacant land, opposition to higher density development often reduces the overall level of housing production, and more developers would bid on fewer parcels of land. To the extent land-supply constraints reduce the beneficial characteristics of a home by reducing lot sizes, these effects will be even greater. Annexation serves as a de facto expansion of the growth boundary, mitigating land-supply constraints as a factor in housing production and price appreciation. With current annexations, however, Boulder County would still have a deficit of 20,196 residential units by 2010 even if existing land is fully developed ("built out") at current densities.

Despite the county’s commitment to comprehensive planning, it relies heavily on its proximity to more affordable housing in neighboring counties. Boulder County is facing rapid employment growth relative to its population. While its labor force participation rate—the proportion of working-age people with a job or actively seeking a job—is 62 percent, the county is creating 0.67 jobs for every resident according to a recent study from the University of Colorado at Boulder. Thus, Boulder County is "importing" workers from other parts of the state. Moreover, the population/job ratio is expected to increase to 0.85 as large employers such as Sun Microsystems, Iomega, Level 3, and other high-tech companies move into the county and the small-business sector continues to grow. To accommodate the housing needs of the work force these companies rely on, Boulder County will have to increase residential development or significantly upgrade transportation infrastructure to enable commuting from outside the county. The jobs-housing imbalance is likely to worsen as local communities adopt strategies that favor commercial development, including national retailers such as Home Depot, Target, and CompUSA, rather than residential development.

Congestion and the jobs-housing imbalance within the county will persist if growth-management policies attempt to channel housing development into more densely packed patterns than local workers prefer.

These trends toward more restrictive growth policies may have significant regional consequences. By promoting commercial development, the county has significantly boosted job growth without a corresponding increase in residential development. Thus, workers increasingly commute from outside the county, adding congestion and demands on transportation infrastructure. Congestion and the jobs-housing imbalance within the county will persist if growth-management policies attempt to channel housing development into more densely packed patterns than local workers prefer. Indeed, Boulder County’s commercial development may depend on neighboring communities pursuing more relaxed growth-control policies, allowing affordable housing to develop within commuting distance of jobs in the county. Ironically, Boulder County’s "success" may well depend on the willingness of other counties to house Boulder County workers and regional commitments to upgrade infrastructure to accommodate increased traffic volumes from low- and moderate-income commuters.

In sum, even though Boulder County has some of the highest-cost housing in Colorado, further significant housing-price appreciation may have been forestalled because community-service areas "breathed"—they expanded to include more vacant land through annexation. This accommodating growth policy is unlikely to continue as the county and local cities restrict the supply of land through aggressive open-space protection programs (e.g., conservation easements, purchases of future development rights, and further restrictions on land development). If the growth boundaries are converted into static lines in the land, the county is likely to face the density/housing price tradeoff already evident in Portland within the next decade.

Farmland protection is a major goal of many states and counties across the nation. Nowhere is this more evident than in Lancaster County, Pennsylvania, home to one of the nation’s largest communities of Old Order Amish and Mennonites. As in Oregon, local policymakers in Pennsylvania see urban-growth boundaries as an important policy tool for protecting the local agricultural economy, particularly farming. As a result, Lancaster County has one of the nation’s most ambitious programs geared toward protecting farmland. Lancaster County’s experiences are particularly relevant for states and regions that have convened agricultural and farmland-preservation task forces, such as the Central Valley near Sacramento and states such as Michigan, Indiana, and Ohio.

Once again, Lancaster County’s experience must be considered within the broader context of all the impacts and the realities of plan implementation. Despite the laudable goals of the county and optimistic projections of local planners, urban-growth boundaries in Lancaster County have provided few tangible benefits to local residents. In fact, growth boundaries appear to have a negligible impact on population growth, development patterns, and farmland preservation. 

Lancaster County pursues a two-pronged approach to farmland and open-space preservation to protect a rural and semi-rural lifestyle. The first strategy, begun in the early 1980s, purchases future-development rights to land as a way of permanently preventing development or conversion of farmland to other uses. To date, the county has preserved almost 30,000 acres through this program.

The second strategy, implemented when the county adopted its comprehensive plan in 1990, establishes urban-growth boundaries to limit development beyond existing urbanized areas. The growth boundaries were "promoted as the key growth management tool" in the county’s comprehensive plan that would direct, through the power of regulation, growth to occur in "appropriate places" based on its regional land-use planning model. More specifically, local officials argued the growth boundaries would:

• induce more-compact development close to existing urban areas;
• create a definable urban/rural edge to protect scenic vistas;
• preserve farmland to protect open space;
• protect the local agricultural industry from encroachment by nonrural land uses;
• promote infill in already developed areas and foster the revitalization of the county’s major city; and
• provide for efficient public services near already urbanized areas.

Nineteen of the county’s 26 cities and villages have established urban-growth boundaries in an attempt to contain land development and preserve farmland. The county’s comprehensive plan is expected to evolve as a "living plan" that addresses new issues and challenges as they emerge in the county. The plan is reviewed every five years to "maintain its vitality."

Lancaster County’s aggressive growth-management and farmland-preservation initiatives are an outgrowth of local desires to preserve its unique cultural mix and small-farm feel. Lancaster County is also one of the larger agricultural counties in Pennsylvania. About 4,700 family farms, many owned by Amish families, currently operate on 64.5 percent of Lancaster County’s land area. Farming is considered a critical component of the local economy (although most Amish businesses in the county are no longer linked to agriculture).

Lancaster County Is Easily Accessible from Major Urban Areas

Lancaster County’s location in southeastern Pennsylvania places it directly in the path of development. Lancaster County is easily accessible from a number of metropolitan areas: an hour and a half from Philadelphia (69 miles) and Baltimore (67 miles) and less than an hour from the state capital of Harrisburg (36 miles). The county is serviced by two four-lane state highways (U.S. 222 and S.R. 283) and hosts a major interchange for the Pennsylvania Turnpike (I-76) and I-176 (link to Reading, Pennsylvania) at Morgantown. This accessibility, combined with the aesthetic qualities of rural life, helped push Lancaster County’s population from 362,346 in 1980 to 454,063 in 1996, an increase of 25.3 percent. More than one-third of these new residents have moved in since 1990, and the county is adding about 5,200 residents each year. Most of the county’s growth has occurred outside its major city, Lancaster, and in dozens of smaller boroughs and villages. In fact, the City of Lancaster has been losing population steadily, falling 3.5 percent in the 1990s to 53,597.¹ Most of the county’s growth has occurred in low-density rural and suburban settings.

Note:

Estimates of the Population of Places: Annual Time Series, July 1, 1991 to July 1, 1996, Population Estimates Program, Population Division, U.S. Bureau of the Census, Washington, D.C., http://www.census.gov/population/estimates/metro-city/scts96/sc96t_PA.txt.

Formal farmland-protection efforts in Lancaster County date to April 1980, when county commissioners appointed a nine-member Agricultural Preserve Board to develop and administer a conservation-easement program. Three years later, the commissioners established the board as a separate department within county government for the sole purpose of acquiring open space and farmland to prevent development. The county has since preserved 314 farms and more than 26,200 acres of farmland. The county spent $2.2 million for agricultural land preservation in 1998 and allocated another $1.5 million in 1999. In addition to farmland preservation, the county allocated almost $2 million for program grants to municipalities for community parks in 1998 and allocated another $1.7 million for this program in 1999. Combined with additional parkland acquisition and facilities improvements, these open-space initiatives accounted for 15.6 percent and 18.9 percent of the county’s capital budgets in 1998 and 1999 respectively. While a private land trust is also active in securing conservation easements, the county’s program is responsible for acquiring about 80 percent of all the land currently protected from development.

In 1990, Lancaster County responded to ongoing public concerns about development’s perceived threat to the cultural and aesthetic characteristics of their county by adopting a comprehensive plan with strong growth-management provisions. The plan called for urban-growth boundaries for significant urbanized areas in the county and recommended that townships adopt strong pro-agricultural zoning. To date, 39 of the county’s 41 townships have adopted agricultural-zoning provisions to discourage farmland conversion to other uses such as housing or businesses, and three-quarters of the cities and villages expected to adopt growth boundaries have implemented them. Growth boundaries in Lancaster County were expected to include enough developable land for as much as 20 years at a density of five dwellings per acre (one-fifth of an acre, or 8,800 square-foot lots) for cities and boroughs, and 2.5-dwellings per acre for rural villages and