Transit Cooperative Research Program Sponsored by the Federal Transit Administration RESEARCH RESULTS DIGEST June 1995--Number 7

Subject Areas: IA Planning and Administration and Responsible Senior Program Officer: Dianne S Schwager VI Public Transit An Evaluation of the Relationships Between Transit and Urban Form

This TCRP Digest summarizes the results of Phase I of TCRP Project H-I, "An Evaluation of the Relationships Between Transit and Urban Form." The objectives of this phase were to 1) review the existing literature on transit and urban form relations, 2) develop a framework to synthesize this knowledge, 3) identify gaps in current knowledge, and 4) develop the research plan for the balance of the project. This Digest, which brings together the results of more than 30 years of theoretical and practical examinations of transit and urban form relationships, provides a base of knowledge for future planning and decision making. The research plan will be implemented in Phase II. The Digest was prepared by Robert Cervero, University of California, Berkeley and Samuel Seskin, Parsons Brinckerhoff Quade & Douglas, Inc.

1.1INTRODUCTION of the densities and mix of uses that they must encourage to ensure system viability. The study of transit and urban form relationships is n Some transit agencies are rethinking joint commanding wide attention in the 1990s. Transit operators, development as a strategy for enhancing revenue and transportation and land-use planners, real estate developers, ridership. Staff need formulas and guidelines for the types, public officials, and concerned citizens should benefit from intensities, and characteristics of land uses that they ought the information included in this Digest as they face many to encourage. critical decisions in the future. The following challenges n Clean Air Act requirements are forcing public are increasing the need for information on transit and urban officials and planners to rethink ways to reduce vehicle form relationships. trips and shift travel to other modes. Land-use policy shows promise for furthering these goals. n The end of interstate highway construction, n The encouragement explicit in ISTEA to consider coupled with increased highway congestion, has led to the land use as a tool to manage demand has heightened need for new policies to manage travel demand. interest and awareness in land use without furnishing data n For 2 decades, demand-management programs, on what land use can contribute. such as ridesharing and preferential parking, have shown n Land development patterns continue to favor both their potential and their limitations for slowing the automobiles, reducing the prospect for transit's financial growth in vehicle trips. There is a need for adding new security and increasing the need to construct highway tools to the policy toolkit. improvements in an environment in which such n The inability of transit systems to enlarge market improvements take a decade or longer to complete. share--despite investment in new technologies, marketing n Real estate developers continue to look for techniques, and service enhancements--has created an guidance on formulas and guidelines for projects that will imperative for innovation. Land use is one promising lead to approval in an increasingly difficult regulatory policy. environment. However, these developers need evidence n Cities and regions planning new rail systems or that the promises made by proponents of neotraditional expansions of their bus systems need documentation plans can be met.

______

TRANSPORTATION RESEARCH BOARD NATIONAL RESEARCH COUNCIL 2

CONTENTS

1.1 INTRODUCTION ...... 1

Summary of Transit Impacts on Urban Form and Land Use...... 3 Summary of Urban Form and Land-Use Impacts on Transit Demand...... 5 Summary of Interactive Impacts of Transit and Urban Form...... 5 Summary of Research in Progress ...... 5 Conclusions and Implications for Future Research ...... 6

1.2 THE CHANGING URBAN FORM OF NORTH AMERICAN CITIES ...... 6

1.3 TRANSIT IMPACTS ON URBAN FORM AND LAND USE ...... 6

Macrolevel Research...... 6 Early Subways and Urban Development ...... 7 Studies of New-Generation Rail Investments ...... 7 Other Transit Technologies ...... 11 International Insights ...... 12 Intermediate-Scale (Corridor/Activity Center/Station Area) Level ...... 15 Downtown Impacts ...... 15 Impacts in Suburban Areas...... 15 Light Rail ...... 17 Impacts on Property Values and Rents...... 18 Institutional Issues of Station-Area Development...... 19 The State of Research ...... 19

1.4 URBAN FORM AND LAND-USE IMPACTS ON TRANSIT DEMAND ...... 20

Macrolevel Analysis ...... 20 International Studies ...... 23 Intermediate-Scale: Corridors and Activity Centers...... 25 Density and Travel Behavior ...... 25 Mixed-Use Developments and Travel Behavior ...... 27 Intermediate-Scale: Neighborhoods and Station Areas ...... 31 Transit Usage by Proximity to Stations ...... 31 Impacts of Traditional Designs...... 31 Studies on Pedestrian Access...... 35 Design Guidelines...... 38 Microlevel Analyses ...... 39 Conclusions...... 39

1.5 INTERACTIVE IMPACTS OF TRANSIT AND URBAN FORM ...... 39

1.6 RESEARCH IN PROGRESS ...... 41

Transit Impacts on Urban Form and Land Use ...... 41 Land-Use Impacts on Transit Demand...... 44 Interactive Impacts of Transit and Urban Form ...... 44

BIBLIOGRAPHY ...... 45

These Digests are issued in the interest of providing an early awareness of the research results emanating from projects in the TCRP. By making these results known as they are developed, it is hoped that the potential users of the research findings will be encouraged toward their early implementation. Persons wanting to pursue the project subject matter in greater depth may do so through contact with the Cooperative Research Programs Staff, Transportation Research Board, 2101 Constitution Ave., N W., Washington, DC 20418 3

n Citizens, planners, and While transit and urban form and development of cities and officials continue to look for strategies influence each other simultaneously, metropolitan areas since the late 1800s. to halt and reverse the economic almost all empirical investigations to n Large portions of our older decline of older communities. They date have focused on only one direction cities have been shaped by streetcar and look at rail transit as a policy to achieve of the relationship: either how transit subway lines. Transit no longer has the this goal, but they lack information on investments affect urban form and land ability to shape urban form the way it the types of policies and institutions use, or how densities, walking did in the streetcar and subway era that must accompany transit to lead to environments, and other characteristics when transit vastly increased the reinvestment. of cities affect transit demand and portions of regions accessible to n Land-use planners see the travel behavior. Accordingly, this downtowns. Nonetheless, today's rail need to incorporate transit planning literature review is organized transit investments can strengthen into their long-range plans. To do this principally around these two traditions downtowns while also encouraging work, they need information on the of transit and urban form research. In decentralization and multinucleation in ways in which transit affects and is addition, selections from the growing the suburbs. affected by land use. body of knowledge on the interactive n Urban rail transit investments n Transportation planners, in effects of transit and urban form are rarely "create" new growth, but more turn, acknowledge the role also included. typically redistribute growth that would transportation plays in shaping urban have taken place without the form but are uncertain about what to Although the most attention to investment. expect from transit or how to date has been given to heavy rail n In most metropolitan areas incorporate the feedback from transit, this review summarizes findings with heavy and light rail systems, the transportation to land use into their for the full spectrum of transit modes, greatest land-use changes have models and plans. including bus, light rail transit, heavy occurred downtown, in the form of rail, and commuter rail. Research has redeveloped land and new office, been conducted at the following scales: commercial, and institutional macro (city/regional), intermediate development. San Francisco, Much has been written on these (corridor/ activity center), and micro California; Toronto, Ontario; subjects, but there are only a few dozen (station area/ neighborhood/site). While Washington, DC; Buffalo, New York; empirical studies in the last 30 years on much of the literature cited in this San Diego, California; and Portland, which to base conclusions about transit review is drawn from a U.S. context, Oregon provide examples. The and urban form relationships. This some of the more important strengthening of downtowns stems in digest summarizes the significant international studies are discussed as part from the fact that downtowns are aspects of this empirical work. well. Finally, this digest includes a the hubs of all rail systems. discussion of key research in progress n New rail systems have also whose focus is closely related to transit been a force toward decentralization of The focus on key empirical work and urban form, and from which useful population and employment, rather leaves out a large body of literature that information is expected to result. than toward urban containment. Large focuses on these relationships from a subcenters or "edge cities," have theoretical perspective. While selected Past work on how transit affects formed around stations in a number of citations of this kind have been land use has generally been at a more North American rail cities--Washington included in the bibliography (Appendix macroscale, while investigations on Metro: Balston, Bethesda, Silver A), researchers do not intend this how urban form, densities, and urban Spring; Toronto: Scarborough, North literature review to be an exhaustive designs affect transit demand have York; San Francisco: Walnut Creek, examination of theoretical issues. generally been conducted at several Concord; Atlanta: Lenox Square, scales of analysis. Table 1 is a matrix Buckhead; Vancouver: Burnaby, New Further, there are a number of that cites the studies reviewed in this Westminster; Miami: South Dadeland. publications by local, state, and project that have been conducted on n There have been fewer national agencies that focus on transit transit and urban form relationships at changes in residential land uses than in and land-use relationships using different scales of analysis. (See commercial land uses as a result of rail descriptive statistics. Studies of this Appendix A for a complete investments. Some apartment kind include summaries of building bibliography of studies reviewed in this construction has occurred near permit data for sites proximate to rail project.) suburban rail stations in Washington, stations. While several of the more DC, Philadelphia, Toronto, San significant analyses of these kinds have Summary of Transit Impacts on Francisco, San Diego, and other cities. also been included in the bibliography Urban Form and Land Use There are a number of barriers to (Appendix A), citations included in this higher density residential development literature review are principally those Thirty years of case studies and near rail stations, including community that have used more rigorous analytic historical research have documented methods to deal with cause and effect the role that transit has played in the relationships examined in this project. growth TABLE 1 Selected literature: transit and urban form

A. The impact of transit on urban form

B. The impact of urban form on transit demand and travel behavior 5 opposition and weak markets for patronage in corridors leading to the metropolitan fringe, while in European multifamily housing. However, there is employment centers. This research cities, affluent residents often seek evidence that accessibility to rail focused principally on the New York amenities and more central locations. becomes capitalized into higher metropolitan area and was based on n At the neighborhood and residential land values. travel data that is 20 to 30 years old. station-area scale, transit has been n The urban form and land-use n Recently, regional planning shown to draw pedestrian patrons from impacts of light rail, busways, and bodies have used simulation models to up to 4,000 ft. Surveys in Washington conventional bus transit have generally assess the impact of various growth DC, San Francisco, and elsewhere been weaker than those of heavy rail scenarios on future travel behavior in indicate significant transit trip systems because the systems usually their regions. Most find that generation rates from residential confer less accessibility advantages, at concentrating jobs and housing where development proximate to rail stations, least relative to the main competition-- they can be served by transit increases especially for systems and regions in the highway system. transit mode shares and reduces vehicle which both housing and employment n In general, transit miles traveled, but these effects are are found adjacent to transit. investments and services are incapable diluted by the fact that two-thirds or n Pedestrian travel in both by themselves of bringing about more of the forecast-year development employment and residential areas can significant and lasting land-use and is already in place. be induced and pedestrian trips urban form changes without public n International work has lengthened by the provision of policies that leverage these investments documented strong relationships extensive and attractive pedestrian and the pressure of such forces as a between urban densities and energy amenities. rapidly expanding regional economy. consumption in metropolitan areas. In n Local jurisdictions and transit Experiences in Europe and Canada addition, European cities have agencies are increasingly developing underscore the importance of coupling settlement patterns that are design guidelines to support rail investments with reinforcing local substantially denser and more mixed in alternatives to the automobile, but policies such as up-zoning around character than American cities. specific features of successful transit- stations, supplemental acquisition, joint Europeans also ride transit, walk, and oriented site designs have yet to be development of station-area land, and bicycle more than Americans. demonstrated empirically. situating publicly provided housing n At an intermediate scale, near stations. dense office and residential activity centers generate larger numbers of Summary of Interactive Impacts of transit trips for work and nonwork Transit and Urban Form Summary of Urban Form and Land- purposes than do less dense, auto- Use Impacts on Transit Demand oriented suburban activity centers. Less While it is acknowledged that dense, less diverse suburban activity transit and urban form interact and Understanding how the densities, centers generate far higher numbers of influence each other simultaneously, settlement patterns, land-use vehicle trips and lower levels of auto these relationships are extremely compositions, and urban designs of occupancy, particularly when combined difficult to document without cities and neighborhoods influence with abundant, free parking. The interactive transportation and land-use transit usage is of vital importance to inclusion of retail and service activities models. Work done with an interactive transit planners and decision makers. in traditional suburban office model in Seattle suggests that regional Whether a future light rail extension developments can reduce auto land-use patterns organized around will be a cost-effective investment or dependency. multiple centers and supported by high- whether headways should be increased n Paired comparisons of pre- capacity transit will generate a on a conventional bus route hinges war transit-oriented communities that reduction in automobile dependence critically on whether the built feature in-neighborhood retail service and an increase in transit utilization. environment and the people living and and modified-grid streets with post-war Additional research is underway on this working there will support these auto-oriented planned subdivisions topic. changes with their patronage. suggest that transit-supportive environments can induce more walking Summary of Research in Progress n The key domestic study on and transit trips. the influence of urban form on transit n Residential density and Important research is in progress demand (Pushkarev and Zupan) design influence travel behavior on both the effects of transit on urban identified a set of relationships between directly, but in a less powerful way form and the effects of land-use residential densities in transit corridors than the socioeconomic characteristics patterns on transit demand. The impacts and levels of travel patronage. In of residents. Different types of of transit on urban form are focusing on addition, significant relationships were households live in dense and spacious the macroscale with updates of studies found between the size and areas within metropolitan regions. In on the impacts of the second generation extensiveness of employment centers American cities, affluent residents seek and transit space at the 6 of heavy rail systems on regional form. centers, each with its own catchment or identified between 6 and 54 subcenters Work on the impacts of land use on zone of influence. More popular (Gordon et al. 1986; Heikkila et al travel demand is going on at all scales accounts of metropolitan growth 1989; Giuliano and Small 1991). of development. One can expect trends--such as Garreau's Edge City increases in knowledge in this area in (1991), which chronicles the While large suburban downtowns the next several years. emergence of mini-cities on the and edge cities have gained recent metropolitan periphery--embrace the media attention, in many areas a far Conclusions and Implications for principle of "urban realms." more dispersed, less-structured form of Future Research suburban office development has taken It has been the polycentric or form. In a study of six large U.S. Several decades of research multicentered model of urban form that metropolitan regions, Pivo (1990) regarding the influence of transit on has gained the most attention in recent concluded that most office jobs were urban and metropolitan form have led years. Schneider, in Transit and the located in small- and moderate-sized, to the emergence of some consensus on Polycentric City (1981), was one of the low-intensity clusters along freeway the nature of this relationship. In first transportation analysts to study the corridors. Pivo has described America's contrast, research on the ways in which emergence of large-scale suburban suburban structure as "The Net of land use and urban form influence activity centers and their implications Mixed Beads," an analogy to convey travel behavior and transit patronage is for transit services. He advocated both the reality that office complexes in the a less settled question. Data on the an intervention into urban land markets suburbs come in all shapes and sizes, influence of land use at the corridor and to encourage more clustered, transit- some still true to the classic image of metropolitan level are relatively out of serviceable development and new low-density sprawl, some beginning to date. Research on the effects of the transit investments in the. forms of look more like compact, high-density built environment at the site level on center-focused schedules and cities (Chinitz 1993). For greater Los travel behavior is incomplete. paratransit services, outer city transit Angeles, Gordon et al. (1986) and terminals and internal circulation Giuliano and Small (1991) have found 1.2 THE CHANGING URBAN systems. Cervero (1986, 1989) that, except for several large FORM OF NORTH AMERICAN associated the trend toward suburban concentrations, small-scale clustering CITIES subcentering with increasing regional best characterizes its form of mobility problems, and called for a subcentering. These findings suggest Metropolitan areas have grown balance of land-use planning initiatives that the decentralization process in and changed substantially during the (e.g., transit-friendly site designs) and contemporary urban America is 30-year period of the studies reviewed. transit service strategies (e.g., timed- complex and spans a continuum Because it is important to keep these transfer networks) to adjust to these ranging from scatteration or dispersal, changes in mind, an overview of trends. Thomson (1978) has on one extreme, to more orderly monocentric and polycentric urban distinguished some of the world's polycentric forms on the other. forms is included herein. largest multicentered metropolises in terms of strong centers (e.g., London, With transit systems designed of The classical descriptive models Paris, Tokyo) or weak ones (e.g., Los necessity to serve concentrations of of urban form are 1) the well-known Angeles). employment and activities, the concentric zone model (Park et al. metropolitan decentralization process 1925) based on a solar system analogy; A number of empirical studies has weakened transit's ability to 2) the sector model (Hoyt 1939) where have documented the emergence of connect households with their work activities follow dominant subcenters in the United States. Using locations. As transit reaches transportation corridors; and 3) the minimum thresholds for office and proportionally fewer people, it's ability multinuclei model (Harris and Ullmann retail floor space and jobs, analysts to influence urban form has lessened as 1945) based on the premise that have identified 13 subcenters in greater well, as discussed in the next section. activities do not evolve around a single Washington, DC (Garreau 1987; 1991), core but around many nodes. None of 17 in greater Atlanta (Atlanta Regional 1.3 TRANSIT IMPACTS ON the three classical models is universally Commission 1985; Hartshorn and URBAN FORM AND LAND USE applicable, and all cities exhibit Muller 1986), and 22 in the Houston features of each model rather than one area (Rice Center 1987). In a national Macrolevel Research feature exclusively. More recently, study, Cervero (1989) found 57 large- urban geographers (Vance 1977, scale suburban employment centers Streetcar Development and Urban Muller 1981) have advanced an "urban located at least 5 radial mi from a Form realms" model based on the premise central business district (CBD) and that metropolitan areas are being containing over 2,000 full-time workers Historical case studies provide reorganized into a set of independent and over one million sq ft of office some of the richest insights into the space. Three separate studies of the Los ways Angeles area have 7 transit has shaped the structure and pre-automobile era, by the 1950s rail Studies of New-Generation Rail character of American metropolises. transit had negligible impacts because Investments Classic works by Warner (1962), of the dominance of the automobile- Vance (1964), and Fogelson (1967) highway system. In a study of 49 U.S. More recent macrolevel studies trace how the extension of electric metropolises over the 1920 to 1970 have focused on heavy rail transit, streetcar lines to suburbia around the period, for instance, Harrison and Kain since today this transit technology turn-of-the-century led to massive found increases in automobile provides the largest incremental decentralization in Boston, the San registrations had 3.5 times the effect on increase in regional accessibility and Francisco Bay Area, and Southern urban densities as increases in rail thus could be expected to have the most California. Streetcar suburbs not only transit mileage and 12 times the measurable land-use impacts. The land- defined the radial spines of large east influence as increases in bus in-service use impacts of new-generation systems coast and west coast metropolises, but mileage. built since 1960 in Atlanta, Montreal, also allowed for the physical separation San Francisco, Toronto, and of home from work and of social Early Subways and Urban Washington, DC have varied widely. classes (Schaeffer and Sclar 1980).. Development Overall, impact studies of Bay Area Middleton (1966, p. 44) has concluded Rapid Transit (BART) (Webber 1976; that Residential and employment Dyett et al. 1979), the Lindenwold line densities in New York City, Chicago, (Boyce et al. 1972, 1976), and ...more than any other development, Philadelphia, and Boston clearly reflect Washington Metrorail (Lerman et al. the electric streetcars contributed to the results of rail transit expansion 1978; Paget Donnelly 1982) found that, the growth of the metropolitan during the first half of the 20th century. consistent with location theory, suburbs. Population growth followed In all of these cities, rail transit regional rail systems have been a force car lines, and a new trolley line extension invariably increases land investments were followed by dramatic toward decentralization of both values. Not infrequently, real estate increases in downtown employment population and employment. Intercity syndicates built electric railways just densities and the clustering of comparisons with "control" cities to promote their land developments. residential subdivisions around without regional rail systems suggest suburban stations. Rail transit lines had these rail investments have probably As rail lines extended between their greatest impact around station had some "clustering" effects, leading 1880 and 1920, population levels in areas located farthest from the city to perhaps a more polycentric U.S. cities of 10,000 people or more center, which had been previously metropolitan form than would have increased from 11 million to nearly 45 undeveloped and unserved by public existed had any of these rail transit million, or almost one-half of the transit. Impacts were fewer in areas that systems not been built (Hilton 1968; national total (Smith 1984). Urban rail were already built up. Meyer and Gomez-Ibanez 1981; Smith ridership increased from 600 million to 1984). 15.5 billion trips annually. The The effects of metros, or subways, development patterns of urban cores on land development were studied as and nearby suburbs of many American early as 1930. In his review of this A recent examination of areas cities were irrevocably shaped by work, Boyce (1972) found that "the within ¼ mi of rail transit stations in streetcar lines over this period. Smerk subway reflects the condition of the the Washington DC area (Green and (1967) estimated that as much as one- area through which it passes...If the James 1993) suggests that heavy rail in quarter of the U.S. population resided district is growing rapidly, the subway Washington is responsible for at that time in urban and suburban areas accelerates such growth; where it is increasing the amount and scale of whose spatial organization was shaped stagnant, the values along the route development near rail stations. The by the streetcar. change little; where influences are such authors made use of an existing land- as to cause land values to drop, the use model of the metropolitan area and Based on a statistical analysis of subway fails to pull the area in question created a set of zones of ¼-mi radius 28 U.S. metropolitan areas from 1890 from the slump it is experiencing." around Washington Metropolitan Area to 1910, Harrison (1978) found that Transit Authority (WMATA) stations. each additional mile of streetcar line They examined historic development per capita was associated with a 3.2 The accessibility benefits of rail (changes in employment) in these percent increase in the share of single- transit may influence development long small-area zones in order to determine family housing for the regions. The after the initial investment in transit. whether they were similar to patterns of city-shaping impacts of streetcars were Recent office development along the employment change in the larger traffic found to be short lived, however. Hudson River waterfront in New Jersey analysis zones of which they were a Harrison and Kain (1974) found that has concentrated around pre- part. The larger zones were presumed while streetcars had significant impacts automobile era rail stations that provide to lack the kind of accessibility by on urban spatial structure during the access from many residential areas pedestrians from transit stations that (Zupan 1993). 8 would be required for the rail system to Dear 1975; Dingemans 1978). In the of transit usage than either the density influence overall zonal attractiveness. case of Toronto, the consensus is the or floorspace size of a center. This subway was not the "single cause" of suggests that "transit service may be The authors concluded that a observed land-use changes, but rather a the most important physical planning statistically significant difference variety of economic, social, and policy variable for encouraging transit existed between small-area zones and political factors combined to create a use" (Pivo 1993, p. 44). larger ones, which can be attributed to heavy and continuing demand for new the presence of transit. They attribute central-city office space and apartments At about the same time as their findings to a use of a finer grained (Figure 1). Toronto, the Greater Vancouver level of analysis than is typical of Regional District (GVRD) was station-area impact studies. They also Canadian cities like Toronto have planning for orderly multicentered note the importance of accessibility of been at the forefront of planning for growth, culminating in the 1975 plan station areas to high income residents urban form shaped by transit. In 1976, called The Livable Region. Three as being a significant cause of the Metropolitan Toronto published the important elements of the plan were (1) relatively higher level of employment Metroplan: Concept and Objectives, to promote a balance of jobs and growth. Lastly they found that those which explicitly called for the population in each subregion; (2) to station areas surrounded largely by development of a hierarchical, create regional Town Centres, resulting residential uses were likely to generate multicentered urban form. The in a polycentric metropolis; and (3) to less employment growth than station subcenters--downtown Toronto, two build a transit-oriented transportation areas whose environs were dominated major outlying centers, and 13 system linking residential by nonresidential uses. intermediate centers--would be neighborhoods, regional Town Centres, interconnected by various forms of and major work areas. Figure 3 shows a public transportation. The plan sought perspective drawing of how Toronto is often heralded as the to retain the preeminence of downtown Vancouver's polycentric form was best North American example of rail Toronto as the employment, intended to evolve from 1976 to 1986. transit's city-shaping abilities. A commercial, cultural, and political The lifeline that would eventually frequently cited statistic is that during center of the region. A subsequent plan connect the Town Centres was the the early 1960s following the opening revision in 1980, Official Plan for the SkyTrain that opened in time for the of Toronto's Yonge Street subway line, Urban Structure, changed the number 1986 World Expo, which had a theme around one-half of high-rise apartments of designated subcenters (adding four of advanced transportation systems. As and 90 percent of office construction in intermediate-level ones) but retained in Toronto, public control of land and the city of Toronto was within a 5-min the basic hierarchical subcenter goal. In zoning, as well as the siting of rail walking distance of a train station ensuing years, Metropolitan Toronto investments and other infrastructure, (Heenan 1968). The subway not only has adopted a number of strategies to became the primary means for brought about the development of implement the hierarchy plan, implementing the plan. Government vacant or under used areas (some including the targeting of infrastructure officials methodically put in place within a few miles from the city investments in subcenters and zoning various incentives and regulations that center), but it also recycled areas that incentives that encourage densification. encouraged the concentration of a were already built up. Stringent land- One strategy was to construct significant portion of growth in offices, use controls and various pro- institutional buildings in Scarborough service employment, and cultural/ development forces (e.g., regional Town Centre, one of the region's two entertainment facilities in a handful of governance that promoted coordinated second-tier centers, northeast of Town Centres, each slated for planning) were largely responsible for downtown Toronto. Along with density populations of 100,000 to 150,000. One intensive development around bonuses and the opening of advanced center, New Westminister, has Toronto's stations (Knight and Trygg light rail transit services, these public transformed from a declining, 1977). Besides complementary zoning investments helped spawn other moribund waterfront to a transit village and taxation policies, the consensus is activities in Scarborough Town Centre. with new apartment towers and retail that a number of other conditions are complexes. necessary for rail transit to exert a Using cluster analysis that strong influence on urban form and included variables representing urban Specific land-use planning tools land uses: a healthy regional economy, form, location, and accessibility, Pivo were used in Vancouver to capitalize the availability of land that is easily (1993) recently confirmed the existence on the SkyTrain investment and target assembled and developed, a hospitable of six types of office clusters in growth into Town Centres. Commercial physical setting (in terms of aesthetics, Metropolitan Toronto (Figure 2). areas in major centers have limited or ease of pedestrian access, etc.) and the Primary and Secondary transit clusters no setbacks, creating a pedestrian scale. existence of some automobile were located within walking distance of In many Town Centres, off-street constraints (such as parking a subway station. Pivo found proximity surface parking is not permitted, restrictions) (Knight and Trygg 1977; to a subway station to be a stronger allowing for more intensive use of land. predictor

Figure 3. Vancouver urban form, 1976 and 1986. (Source: Greater Vancouver Region District (1975) The Livable Region) 11

The cultural and institutional Other Transit Technologies In Pittsburgh, planners and transit differences between the United States agency representatives have been and other countries were featured in a unable to identify any specific new matched pair comparison of San Diego, Few macrolevel impact studies of downtown development attributable to California and Vancouver, British other transit technologies have been the trolleys. However, a major negative Columbia (Wilson and Anderson conducted to date. No substantial work impact of the Pittsburgh trolley has 1993). The authors examined these two on the effects of bus transit on urban been identified (Neuwirth 1990). cities to identify differences in planning form could be found. The modern Gimble's department store, which had and implementation of transit-oriented motor bus is a flexible technology. It been directly located on the above- development and concluded that there need not precede development because ground trolley line, went out of were four preconditions for such it can immediately serve developments business. Store management claims the successful development to occur. They of all types in all locations. Once a rail fact that the underground LRT line included coordinated policy support for line is installed, the likelihood of its bypassed the store is partially clustering development and transit, permanence or long life is great. Fixed- responsible for the store's loss of effective implementation tools, a pre- guideway investments signal a business and eventual closure. existing urban form and transportation permanent infrastructure addition to the system that supported transit, and an development community, thus Cervero's (1984) study of light rail alignment that enhanced development prompting competition to exploit the systems in the U.S. predicted modest opportunities. accessibility gains provided. land-use impacts because most LRT Competition for sites with good lines follow abandoned rail rights of The authors attributed the accessibility leads to densification and way with minimal development differences in transit oriented potentially agglomeration benefits. potential and also rely heavily on park- development patterns that followed the Without similar competition for sites and-ride access. For the most part, introduction of rail systems in each of served by bus transit, it follows that studies show recent light rail these metropolitan areas to "underlying impacts of a new bus line on existing investments to have little land-use public attitudes about planning and corridor or metropolitan development impacts outside of downtown areas government." They contrasted the will be negligible. (San Diego Association of Canadian political and social context, Governments 1984; Cervero 1985; that they assert is supportive of transit- The prospect of inducing real Barney and Worth 1993). oriented development to the U.S. economic growth is often used to context, which is less supportive of justify new light rail investments in In the case of San Diego's trolley, government intervention into markets. particular. Claims for LRT investments an evaluation 3 years after the system's They encourage U.S. planners to have included stimulating community 1981 opening concluded that it was "assess the context in their revitalization, job creation, economic "not a major locational determinant" communities as they consider the development along specific corridors, and that the "development and market transferability of Canadian TOD and maintaining and sustaining dense forces at work in Centre City and the strategies." urban centers whose vitality are being typical intense scale of development sapped by auto-oriented development. tend to overpower the trolley's role as a Past work also suggests that rail Both Buffalo and Pittsburgh built LRT factor in development decision" (San transit investments do not stimulate real systems in the 1980s in large part to Diego Association of Governments economic growth; rather they only help rejuvenate their respective 1984, p. 44). The most significant influence where already-committed downtown cores. A survey of recent trolley-related land-use changes growth takes place. (All rail developers in Buffalo conducted by the in San Diego outside of downtown investments, of course, induce Niagara Frontier Transit Authority have been the clustering of residential construction-related employment found that the transit system was and mixed-use development near growth which, in the case of Buffalo considered a positive influence for several stations in La Mesa on the line and other areas with fairly stagnant $650 million of new development in to El Cajon. Ridership surveys reveal regional economies, can be significant the downtown (Neuwirth 1990). that around 12 percent of work trips [Paaswell and Berechman 1981]). However, this development was also made by the residents of these projects Conventional wisdom holds, then, that affected by other public policy are by transit, nearly three times the all development impacts of rail transit decisions and public investment. The 1990 San Diego regional average for are distributive--e.g., in favor of one major new downtown development is journeys-to-work (Cervero 1993). radial suburban corridor instead of located at a transit stop but would not another. There is less evidence, by have occurred without the expenditure The nonrail transit technology that contrast, that transit investments cause of federal dollars to assemble and has been examined most closely in shifts in population and employment purchase the land for the project. terms of land-use impacts is dedicated between downtowns and suburbs (Knight and Trygg 1977). 12 busways. Knight and Trygg (1977) region's urban form was strongly projects and where there were studied the land-use impacts of the shaped by a combination of strategic perceptible gains in accessibility. Shirley Highway Express Bus Service, planning and regional rail transit Interestingly, densities were found to the San Bernardino "El Monte investment. Following World War II, decline from up to 300 m from CBD Busway," the Blue Streak in Seattle, Stockholm County government, which stations and then rise again; major and the Blue Dash is Dade County, owned over 70 percent of the region's corporations that place a high premium Florida. They concluded, tentatively, land, embarked on an urban spillover on spaciousness have tended to that "so far, the evidence available plan, seeking to direct future population congregate around CBD stations indicates no land-use impacts and industrial growth to new towns leading to this peculiar U-shaped attributable to busway systems, constructed around the same time as a density gradient (Meakin 1990). This including some which compare new regional rail network. The aim was finding likely reflects the influences of favorably in patronage to many to avoid a dormitory town environment cultural and institutional factors on rent commuter rail lines" (Knight and Trygg and to make satellites as self-contained gradients around rail transit nodes, not 1977, p. 177). In greater Washington, and balanced (both socially and in all of which are necessarily transferable D.C., one study suggests that the terms of jobs and housing) as possible to other countries. Shirley Highway HOV/Busway has (City of Stockholm 1992). allowed many Washington employees Interestingly, Stockholm's new towns Overall, studies show that metros to reside farther away than they would are far from balanced or self-contained- in Santiago, Chile; Mexico City, have without the dedicated lane (Meyer -the majority of Stockholm's new town Mexico; and Sao Paulo, Brazil like and Gomez-Ibanez 1980). While residents work out of town and most their North American and European several suburban stations on Ottawa's new town workers reside elsewhere. city counterparts, have encouraged dedicated busway are surrounded by With high levels of external commuting regional decentralization (Institution of mid-rise apartments and offices, and large concentrations of housing and Civil Engineers 1990). In the case of interviews with developers found that workplaces near rail stations, Santiago, Figueroa (1990) found that the growth would have occurred Stockholm's new towns are natural the metro relocated many poor to the regardless and that the busway merely havens for rail commuting (Cervero metropolitan periphery while accelerated the timing of development 1993). Stockholm's built form--a modernizing the inner city. This pattern (Bonsall 1985; Cervero 1986). In a strong, preeminent regional core of settlement, with the affluent located study of Houston's bus transitway, orbited by rail-served satellite centers-- nearest the city center and the least Mullins et al. (1989) found relatively largely accounts for low automobile affluent at the periphery, typifies few impacts--developers stated it had dependency. From a mobility and European cities as well. American no influence on their location choices, environmental standpoint, this has cities, however, tend toward the and before-and-after studies at park- more than compensated for the opposite, furnishing additional and-ride lots near the transitway found tendency of Stockholmers to live and evidence of the role of cultural factors few land-use conversions. And work in separate communities (Figure on development patterns. although Biehler (1989, p. 96) contends 4). that "there is no reason to think that One Brazilian city has attractiveness to development is Two other strong-centered rail successfully influenced urban form inherent in a specific mode," whether metropolises are Hong Kong and with an all-bus system. During the exclusive busway or light rail transit, Singapore. Both are products of strong 1970s, planners in Curitiba began to he fails to provide accounts of local planning controls. In both city- implement an urban design structure development near any of Pittsburgh's states, housing was developed over the that emphasized linear growth along exclusive busways. past 20 years with assumptions structural axes. Using various zoning contained in master plans that Metro tools and other land-use incentives, International Insights corridors would be a reality. Thus, urban growth has been encouraged many Hong Kong and Singapore along five main axes (see Figure 5). Thomson's (1977) study of traffic residential densities were committed Each axis was designed as a "trinary" in the world's largest cities found that before subway construction. road system. The central road has two rail services were most successful in restricted bus lanes in the middle large national centers, like Paris and Within a few years of the 1980 flanked by two one-way local roads Tokyo, that have strong central cores. opening of Hong Kong's regional (Figure 6). On land sites along the Weak centered metropolises, like subway, the 42-km system was turning structural axes, buildings with total greater Los Angeles and Houston, were a profit. Revenues from real estate floor areas of up to six times the plot found to be far more auto dependent. development have helped keep the size are permitted. This coefficient system in the black. Studies found the decreases the farther a site is from Several world cities stand out as Metro generally enhanced land values, public transport. This has encouraged strongly rail oriented. One is especially for residential and mixed-use new commercial developments outside Stockholm. The greater Stockholm Figure 4. Evolution of the Stockholm region. (Source: City of Stockholm)

15 the central city, but along each Downtown Impacts picture. Overall, transit's share of total structural axis, and high-density daily trips is less than 3 percent in the residential development near public Within downtowns, rail transit Portland region, and the share transport services. Today, Curitiba has investments have stimulated continued to decline throughout the over 50 km of exclusive bus lanes. The redevelopment and brought life to once 1980s. Among work trips by system averages 1.2 million passengers moribund commercial districts. BART suburbanites, there was actually a 30 per day, or around 430 transit trips per is credited with focusing much of San percent decline in transit ridership in capita annually, one of the highest rates Francisco's downtown office the Portland region during the 1980s. in the world (Rabinovitch 1993; Lerner construction south of Market Street and As a stimulus to downtown 1994). rejuvenating inner-city Oakland. redevelopment and as a complement to Outside of downtown, however, Portland's very pedestrian-friendly In summary, noted sociologist BART's land-use influences to date downtown, the light rail system has Homer Hoyt (1939) observed over a have been inconsequential--except for been an unequivocal success. As a half a century ago that "...urban form is several stations in the East Bay mobility factor within the region, its largely a product of the dominant suburbs--because of such factors as role has been minor. transportation technology during a local opposition to growth, city's prevailing period of growth." The downzoning, and the siting of stations cumulative body of research over the in freeway medians. Webber (1976) has Studies also show that light rail past 50 years seems to support strongly argued that, more fundamentally, the and transit-mall investments cannot this supposition. However, evidence is reason for BART's negligible land-use overcome the effects of a weak regional persuasive that fixed-guideway transit impacts outside of downtowns stems economy. During periods of economic investments can shape urban form from its poor performance relative to downturns, studies found no under the right conditions. These the automobile--it paralleled corridors redevelopment impacts in Edmonton conditions fundamentally include an with excellent highways and provided and Calgary (Gomez-Ibanez 1985; City integrated approach to transit no real increases in regional of Calgary 1983), Denver or Portland investment and development. Outside accessibility for a number of years after during the late 1970s. After the transit- of the United States, cities that have completion. mall opening, retail sales in Denver's successfully integrated transit and land- mall area dropped as a share of regional use planning do not create a dichotomy sales while in Portland, off-mall between the influences of transit by Some of the more significant locations saw larger increases in itself on urban form and the influence downtown redevelopment impacts have commercial rents than on-mall of transit in combination with public been recorded with light rail and bus- locations (Gladstone Associates 1982; policy. The two activities coincide to mall systems. In downtown Buffalo, a Dueker et al. 1982). shape development as far as possible in number of building restoration projects support of transit investment. Debate in have been completed near the light rail the United States hinges on whether line (Callow 1992) and the trolley was Impacts in Suburban Areas transit can influence development, as if credited with partially attracting over transit itself were not a public $600 million in investments downtown. While some observers feel that investment on which an adequate return In Portland, the downtown transit (bus) transit adds too little accessibility were required. When viewed as such an mall is viewed as the centerpiece of outside CBDs to influence urban form, investment, the imperative to integrate downtown redevelopment, helping to others find that suburban areas have land-use planning and development is trigger such new investments as changed because of transit. In some clear, as are the benefits (in the form of Pioneer Place, a four-square-block areas, commercial and multifamily higher levels of patronage). The mixed-use complex. Arrington (1992, development has clustered near transit sections that follow elaborate on these p. 19) speculates that light rail might stations. Generally, transit agencies and points. have greater redevelopment impacts, local jurisdictions must work with the dollar-for-dollar, than heavy rail private sector for this type of because LRT "...operates at the surface development to occur because there are Intermediate-Scale Corridor/Activity and offers visibility, penetrates the a number of barriers to transit- Center/Station Area) Level community and is not separated from it supportive development. These barriers like heavy rail, which is down in a hole include community opposition to Studies conducted at the or up in the air, and is part of the urban higher densities or redevelopment, subregional (e.g., corridors and activity experience--an amenity, a signature for lender reluctance to finance new types centers) and station-area levels have the area." Eager (1993) notes, however, of projects, high costs of land shed light on rail transit's impacts on that based on 1990 journey-to-work assembly, and weak markets for urban densities, property values, and statistics, the Portland Banfield rail has multifamily housing (Cervero et al. land-use composition. made a fairly minor contribution to the 1994). city's travel 16

Philadelphia--Lindenwold prohibited high-density housing in spite public buildings. The most significant of BART's presence. office cluster in the suburbs has been Case studies of two communities around the Walnut Creek Station, near the Lindenwold High-Speed A shortcoming of the original which added 2.5 million sq ft of office Commuter Rail line--one with BART Impact Study, conducted several space between 1975 and 1992. substantial apartment development and years after the system's opening, was However, this amount pales in the other with almost none--found these that it was premature to expect BART comparison to the 22 million sq ft of outcomes were not influenced by to have exerted meaningful land-use office space added to the I-680 corridor proximity to the rail system, but rather changes in such a short period of time. in the suburbs of Alameda and Contra factors like land availability, zoning, A "BART at 20" update study is Costa County. Overall, 35 million sq ft and local attitudes to growth (Boyce et currently under way (Deakin, Cervero, of office space was built in areas al. 1972). Similar findings were and Landis 1994). It is important to unserved by BART since BART's 1972 observed for office and commercial revisit BART's impacts on the built opening, compared to only 9 million sq development (Gannon and Dear 1972). environment because a premise of the ft within 1/2 mi of an East Bay BART Studies showed proximity to the High- entire project was that it would station. The kinds of office functions Speed line was capitalized into higher eventually lead to mini-communities that are most attracted to BART land values, with the largest gain mushrooming around suburban rail stations are concentrated in the FIRE accruing to residential sites farthest stations, thus helping to create a more (finance-insurance-real estate) and from downtown Philadelphia (Boyce et multicentric, and thus ostensibly more nonbusiness service sectors. al. 1972). A subsequent study found sustainable, settlement pattern. Indeed, these gains held over time, with the $1 billion (1967 currency) property- properties within the station-area tax bond issue that was sold to the Bay Boston Red Line catchment increasing in value by about Area public was based partly on the 7 percent (Allen et al. 1986). argument that BART would materially The metropolitan area in Boston/ enhance quality of life in the region. Cambridge, Massachusetts, furnishes additional evidence of the role of San Francisco BART Between 1970 and 1990, community attitudes and policies in residential population grew, on shaping development near rail systems. Smaller land value benefits have average, 65 percent faster in Bay Area The extension of the region's Red Line been recorded for BART, though most corridors not served by BART (I-680 through the city of Cambridge in the BART studies were carried out too from Walnut Creek to Pleasanton, 1970s led to a review of station-area soon after opening to gauge any Highway 1 Marin corridor) than those zoning in several Cambridge statistically significant impacts. The served by BART. Additionally, neighborhoods. The decision of local largest increases in land values employment growth rates were 92 officials and citizens not to permit occurred prior to BART's operations, percent higher outside of suburban higher densities near transit ensured the result of station-area land BART corridors than along the that new rail stations would serve speculation; once services started, suburban Concord and Fremont BART existing structures and bring about little increases in land values stabilized and lines. Moreover, density gradients change in station-area land use. for parcels beyond one mile of most along the Fremont and Richmond lines stations fell to nearly zero (Dyett et al. were slightly flatter in 1990 than they At the line's terminus, the Alewife 1979). were in 1980. Overall, BART appears station experienced substantial office to have done little to channel suburban development. The area nearest the rail Substantial office clustering has population and employment growth station was devoted to a large parking occurred at BART's Pleasant Hill, over its first 20 years of operation. structure for inbound park-and-ride Walnut Creek, and Concord stations in commuters. While the office recent years, and efforts are underway As was found in the original developments do generate transit to build apartments on park-and-ride BART Impact Study, BART's major ridership, the majority of work trips lots (to be replaced by parking influence on office construction has have origins beyond the end of the structures) at several other stations been in downtown San Francisco, transit system and are largely served by through joint development where around 40 million sq ft of office autos. arrangements (Bernick and Carroll inventory was added within 1/4 mi of 1991). To date, however, few BART from 1975 to 1992. (This There is some evidence that recent residential projects have been built near compares to an addition of only 12 rail extensions in the Boston region suburban BART stations that could be million sq ft of office space elsewhere have encouraged dispersal of originally considered high density even though at in San Francisco over the same period.) downtown functions and accelerated least nine station areas have been Downtown Oakland added 4 million sq the pace at which back-office jobs have specifically zoned for high or medium- ft of office space between 1975 and density residential development. Some 1992; most of this has been for communities have explicitly 17 decentralized. For example, the studies have estimated that commercial and planned development within the extension of the Orange Line north to land prices near Metrorail stations submarket areas. Within these areas, Malden and Charlestown and the Red increased by around 100 percent other commercial projects are being Line south to Quincy and Braintree has several years after services began and built with less obvious ties to MARTA. supported new office and other as much as 400 percent in some locales In Atlanta, rail transit seems to be construction in these areas. Downtown (Damm et al. 1980; Rice Center for exerting its biggest influence in insurance companies and banking and Urban Mobility Research 1987). In transitional areas. In stagnant markets finance firms have relocated many of Arlington, Virginia, residential values near rail stations, MARTA seems their back-office and clerical functions per sq ft were around $10 more for inconsequential. On the other hand, in these suburbs. Although the units across the street from Metrorail some locations are so attractive, such as relocation of back-office functions is stations than for units elsewhere prestigious downtown areas, that transit part of a general trend in many (Rybeck 1981). A recent study of joint- is not a driving factor in development industries, the opening of new transit development mixed-use projects at decisions. It is the areas in between, lines has played a crucial role in job Bethesda and Ballston also found that where some elements are in place but dispersal in the Boston region they commanded a $2 to $4 per sq ft some encouragement is needed, that (Neuwirth 1990). rent premium relative to comparable MARTA seems to be exerting its projects a block away from stations greatest land-use influence. (Cervero et al. 1992). These findings Washington, DC Metrorail are consistent with those of Green and MARTA has benefited in part James (1993), previously cited. because of Atlanta's buoyant economic Over the past decade, the most growth during the 1980s and into the substantial rail-induced land-use 1990s. As noted previously, a healthy impacts outside of a U.S. downtown Atlanta MARTA regional economy is perhaps the most district have been experienced around important prerequisite to the transit- Washington Metrorail stations. High- While Atlanta's MARTA has been induced development impacts--that is, rise mixed-use joint development credited with stimulating office- there has to be growth for transit projects have sprung up at stations in commercial growth in the Arts Center, investments to channel. In the Atlanta Bethesda, Silver Spring, and Friendship Buckhead, and Lenox Square areas, region, MARTA has been cited as a Heights in Maryland, and Ballston, there has been far more office factor of growing importance in the Pentagon City, and Crystal City in construction and value added to land locational decisions of regional northern Virginia. At Ballston, the along the I-285 freeway corridor (e.g., headquarters of major corporations as extension of the Orange Line freed up Perimeter Center) (Davis and Holmes well as back offices that staff large land used previously for a bus transfer 1985). Still, rents at Resurgens Plaza, a numbers of clerical and technical facility. The station became the 27-story office tower directly adjacent workers. A significant part of Atlanta's centerpiece of the Ballston to MARTA's Lenox Station, were $3 to recent growth has resulted from redevelopment program, culminated by $5 higher in 1989 than at other offices decisions to locate such facilities in the the Ballston Metro Center, a 28-story of comparable leasing quality a block region, given MARTA's ability to tower that includes office and retail away (Cervero et al. 1992). move workers to and from downtown space, condominiums, and a hotel. and the airport. With its new terminal WMATA, the regional transit agency, Some of the most significant land-uses within the airport itself, MARTA now also receives air-rights lease income of changes in Atlanta occurred prior to the offers 20-min trips to downtown and over $200,000 per year for the site opening of the MARTA system. These midtown locations. (Cervero et al. 1992). changes were mainly in the form of high-rise office, residential, and mixed- Between 1970-1976 (prior to use building downtown and on the Light Rail Metrorail), 17 percent of high-density north line parallel to Atlanta's premier residential permits issued in the thoroughfare, Peachtree Street (Potter In Baltimore, one study estimated Washington region were located in 1979). In the case of Atlanta's North some property value increases within station areas; between 1977-1980 after Park development, MARTA's 2,000 ft of the light rail line, with the Metrorail opened, 23 percent of the proximity allowed an increase in the bulk of development activity occurring high-density units permitted were scale of the project, making higher at both the downtown and suburban located in station areas. Even after densities possible. At Lenox Park, the terminuses (Rice Center for Urban Metrorail, most of the higher-density proximity of transit appears to have Mobility Research 1987). Miami's units permitted in the region were affected the mix of uses (Neuwirth Metrorail has done little to induce authorized for construction in locations 1990). However, in both cases the inner-city redevelopment. The most outside the influence of Metro stations developments, although large, are only significant change has been the (MWCOG 1982). Based on interviews a fraction of the total amount of addition with real estate brokers and appraisers, existing several 18 of the Datrans office-retail-hotel project Metrorail found it had a negligible destinations, California's LRT lines, on at the Dadelands South station, which impact on residential property values, the other hand, have little impact on yields over $500,000 annually in lease and impacts did not vary significantly value. The modest land-value impacts revenues to the Metropolitan Dade by distance from a station (Gatzlaff and of California's LRT contrast with the Transit Authority (Cervero et al. 1992). Smith 1993). findings of Al-Mosaind et al. (1993) that residences within walking distance San Diego's Trolley is commonly The potential negative effects of of a Portland LRT station commanded recognized as an important factor in proximity to rail transit have likewise an 10.6 percent rent premium. Clearly, maintaining a viable downtown, been studied. An opinion survey impacts appear very localized, turning however virtually no significant land- conducted by Baldassare et al. (1979) heavily on the influences of other local use impacts have been recorded on the found less preference for homes near factors and perceptions. southline to the Mexican border. elevated BART stations. Burkhardt Recently, several large-scale apartment (1976) and Dornbush (1975) also note Voith (1993) has provided recent and mixed-use projects have been built value decreases around BART because evidence that CBD-oriented train near LRT stops in La Mesa. Based on a of such nuisances as noise and services provide a housing value matched-pair comparison of properties vibration, increased automobile traffic, premium. Using data for suburban served and unserved by the Trolley, a and the perceived accessibility of Philadelphia housing values for the recent study concluded that San Diego's different social classes and ethnic 1979-88 period, he found the value LRT had no measurable impacts on groups to otherwise homogenous premium associated with SEPTA rail land values (VNI Rainbow Appraisal neighborhoods. Collectively, these services increased dramatically during Services 1993). In a recently issued studies suggest the land-value impact the 1980s, despite the rapid growth in report in Portland, Oregon (Barney and of rail investments to be highly suburban employment during this Worth, et al. 1993), the author's note localized. period. The changes in the magnitude that there has been significant of the premium correlated with changes development along Portland's Banfield In a recent analysis, done as part in lagged employment growth in the LRT line, but it has been concentrated of the "BART at 20" study, the city of Philadelphia, but not with near downtown. Assessed values of capitalization effects of proximity of changes in suburban employment, station-area properties have risen more single-family homes to BART as well suggesting that the economic quickly than countywide averages, but as several other California rail systems conditions in the central city are an the intensity of development around were studied over a long-term time important determinant of the value of neighborhood station areas remains horizon. (The original BART impact suburban housing with commuter rail low. study was criticized for being service. premature in its evaluation of land- value impacts only a few years after the Impacts on Property Values and system's opening.) Landis et al. (1994) Likewise, Armstrong (1994) Rents used a hedonic price model to show a reports an increase in single-family If transit investments confer selling price premium, which depends residential values of about 6.7 percent benefits to surrounding properties, in on location within the East Bay, for in the Boston area if the residence is in theory this benefit should become every meter a home is closer to the a community with a commuter rail capitalized into higher land values and nearest BART station. station. Distance to the station did not rent premiums. One study concluded influence housing prices, although that San Francisco's BART "had a Similar models for single-family housing within 200 ft of the line had small but significant positive effect on homes near Sacramento and San Diego reduced values. This value reduction the price of single-family dwellings" LRT found no relationship between may be related to freight service on the (Blayney Associates 1978). This study proximity to transit and market values. lines. found a positive effect on housing In the case of the San Jose LRT, the prices at 1,000 ft from BART stations hedonic model showed that transit In terms of office rents, recent of between 0 and 4 percent, which actually takes away value from evidence from the "BART at 20" diminished rapidly with increasing properties that are located within easy update study indicates that office distance from the station. In no case did reach of a station. Landis et al. buildings closest to downtown San the BART effect extend beyond 5,000 speculate that the type of rail Francisco stations commend higher ft. Similar studies of Atlanta's MARTA technology and extensiveness of the office rents at the Embarcadero and system also concluded that transit system have some bearing on home Powell Street stations. In the case of the station proximity is beneficial to values. BART, as an expansive system Montgomery Street station, which residential values when stations are that operates modern trains that serve directly serves the Financial District, designed with sensitivity to major urban centers, seems to exert rents actually increase with distance to surrounding neighborhoods (Nelson positive influences on nearby single- BART. and McClesky 1990). A recent study of family homes. As mainly single lines Miami's that serve a limited number of 19

Fejarang (1994), however, found institutional and cultural factors redevelopment projects need some that commercial property values near contributing to this success. public assistance with financing to be planned Metro Rail corridors in the Los viable. Housing authorities may Angeles area appreciated faster than In Stockholm, Toronto, subsidize below market housing units. similar properties away from the Vancouver, and Curitiba, land-use and Transit agencies or local development corridors during the 1980s when the transportation decisions were made by agencies may issue tax exempt or rail system was being planned and regional planning bodies that had the taxable bonds and thereby share some developed. The market was apparently ability to use transit to shape regional of the risk in the project. The financial anticipating effects of the light rail, as development patterns. In the U.S., arrangements are often complex results were measured prior to transit service is the responsibility of (Ciocca 1994; Cervero et al. 1994). operation of the system. transit agencies, and land-use planning is done by local governments. Transit One proposal to overcome some The findings that accessibility agencies may want dense development of these institutional issues was advantages to rail transit get capitalized near stations, but the communities who introduced in the 1994 California into higher rents and land values control the zoning may not. In addition, legislative session. A Transit Village conflict with the general conclusion most transit-supportive development Development Act was proposed that that rail investments fail to induce real requires more than the provision of would have allowed the use of land economic growth. If business and transit service and supportive zoning. assembly and infrastructure financing housing costs are higher near transit Transit stations must be located in areas powers, similar to those for stations, those living and working there that are conducive to development. redevelopment agencies, to create must be earning more to cover the This is often not the priority when transportation-supportive development higher premium. It may be that transit alignments are chosen (Cervero 1984, at rail stations. The measure did not is a necessary input into maintaining a Lutz and Benz 1992). The transit pass (Cervero et al. 1994). viable downtown (e.g., agglomeration agency, municipality, and developer economies) and that without transit must also work together to produce a In Portland, Oregon, the transit services, investments in downtowns viable project. Public involvement in agency is engaged in successful would shrink. As Voith (1993) has land acquisition, the provision of cooperative planning with local argued, the fact that housing values infrastructure, and financing is often jurisdictions for light rail station areas. rose sharply for properties near CBD- needed for transit-supportive These efforts are aided by state oriented rail lines in suburban development to occur. regulations requiring local jurisdictions Philadelphia, even when many jobs to reduce auto use through land-use were heading to the King of Prussia One of the reasons development planning. A number of relatively dense area and other outlying job centers, has clustered around transit stations in transit-oriented developments are being suggests that transit is part of a larger cities like Stockholm and Toronto is planned. urban dynamic, and any impacts that that local governments bought are directly related to new transit considerable amounts of land that could services may be considerably later be leased or sold to developers. The State of Research outweighed by the indirect impacts U.S. transit agencies cannot legally associated with transit's contribution to acquire with federal funds land beyond The bulk of empirical research on the health (or decline) of cities. One of what is needed for the transit system. transit's impacts on urban form was rail transit's "unmeasurable" benefits There are some other mechanisms for conducted up to 20 years ago. Little might very well be to help stave off the land assembly, such as redevelopment new ground has been broken and few economic decline of some downtown agencies in "blighted" areas, but in new pieces have been published in the centers by providing an important general the only assembled land transit past decade--partly because federally lifeline for sustaining agglomeration agencies have to offer is their park-and- sponsored research on transit and land- economies. ride lots. use impacts ended in the early 1980s. It is perhaps also because of the The cost of providing needed consistent and widely embraced Institutional Issues of Station-Area infrastructure for development can also conclusions of the research, best Development be considerable. When an area is summarized by Knight and Trygg "blighted," a redevelopment agency (1977), regarding the necessary As noted in the section on may have tools like tax increment conditions for rail transit to influence macrolevel analysis, Canadian, financing to generate revenue, but in urban form and land uses. Updates of European, and South American cities many instances no special revenue impact studies in the San Francisco Bay have been more successful in using sources exist (Cervero et al. 1994). Area and Atlanta are currently probing transit as a tool to shape urban form the more recent impacts of these than U.S. cities have. There are a Finally, experience in California systems. Because the findings of these number of suggests that most large-scale updates are pending, 20 complimentary insights into the transit- studies of proposed rail projects. the American population," the authors' urban form dynamic can be gained by findings conform to those of Zupan and focusing on how land uses and the built In another cross-city comparison of Smith in terms of the range and environment shape transit demand and of six U.S. metropolises (ranging in thresholds of density that can be travel behavior. size from Springfield, Massachusetts to associated with higher utilization of the New York City region), Smith transit. (The density of 10,000 persons (1984) found that transit trips rose most per sq mi corresponds approximately to 1.4 URBAN FORM AND LAND- sharply when residential densities seven dwelling units per gross USE IMPACTS ON TRANSIT increased from around 7 dwelling units residential acre.) DEMAND to 16 dwelling units per acre. In the case of greater New York, for instance, A macrolevel study of American Macrolevel Analysis this residential density jump increased new towns examined differences in average weekday transit trips per vehicle miles traveled (VMT) per American Studies person from 0.2 to 0.6 At residential household. Part of the rationale for new densities of 100 dwelling units per acre, communities has been the possibility of In the seminal study Public he found that each resident in the New reducing travel by the planned Transportation and Land-Use Policy, York City region was averaging around juxtaposition of complementary land (Pushkarev and Zupan 1977) developed one mass transit trip per day. uses. A comparison of travel behavior a set of "land-use thresholds" that are in 15 new communities with fifteen necessary to justify financially different An earlier study concluded the "semi-planned" control suburbs showed types of transit investments, based on opposite about the relationship between no discernible reductions in VMT or intermodal comparisons of transit unit density and transit usage. In an transit usage from planned designs, costs and intercity comparisons of econometric analysis of 1973 except in the category of recreational transit trip generation rates (Figure 7). Nationwide Personal Transportation trips (Burby et al. 1974). They found the key land-use Survey (NPTS) data, Peat Marwick & determinants of transit demand to be Mitchell (1975) tested a number of A cross-sectional analysis of the the size of a downtown (in demand functions in an attempt to relationship between density, nonresidential floorspace), distance of a estimate per capita passenger miles for congestion, and mode choice in several site to downtown, and residential both bus and rail transit. The authors dozen American cities (Dunphy and densities. To justify a light rail line, for concluded that "...for both bus and rail Fisher 1994) identified a weak instance, the authors concluded that systems, the explanatory variables of relationship between macrolevel urban minimum residential densities of nine average square miles per capita (the area densities and several measures of dwelling units per acre were needed to inverse of average population density), travel behavior. The findings may be serve a downtown with at least 20 price, and headway were not sufficient explained by reviewing the way in million sq ft of nonresidential to explain very much of the variation which the density measure was defined. floorspace. Urban Rail in America by among urbanized areas in the demand Using overall MSA densities as an Pushkarev, with Zupan and Cumella for transit services." The study independent variable, the researchers (1980), the sequel to Public suggested that socioeconomic noted that the U.S. metropolitan area Transportation and Land-Use Policy, characteristics of residents explained with the highest level of gross density used the same concepts, but developed far more of the observed variation in (households per acre) was the Los six demand-based threshold criteria for modal split. Angeles metropolitan area. Given the fixed-guideway transit (rapid transit, finding that Los Angeles outranked light rail, automated people-movers). A paper developed for the New York and all other cities in urban Each criteria was linked to residential Montgomery County, Maryland, area densities, it is appropriate to densities and to downtown Planning Department (Levinson and rethink measures of urban densities in nonresidential floorspace. Kumar 1994) uses the 1990 NPTS to terms other than those used in their examine relationships between density paper (e.g., in terms of corridors) in The Pushkarev and Zupan findings and several indicators of travel order to understand the ways in which need to be reviewed today since most behavior, including mode choice, travel densities affect transit utilization. U.S. metropolitan areas are time, and vehicle ownership. Although multicentered, thus diminishing the critical of the thesis that density can As part of ongoing responsibilities importance of the size of the CBD. The explain travel behavior, the authors for regional transportation and land-use use of data from the New York City found that relationships between planning, and in part because of the region also limits the ability to density and mode choice "are found absence of adequate empirical data on generalize the findings. Still, this work only at densities greater than 10,000 the effects of land use on travel is the only systematic examination of persons per square mile." While behavior, a number of organizations these relationships in the United States. asserting that these densities are "home It is cited and used frequently in to about 6 percent of feasibility Figure 7. Transit modes suited to downtown size. (Source: Pushkarev and Zupan [1977]) 22 have undertaken studies using travel- modeling, they concluded that in the The Metropolitan Washington demand forecasting models to simulate long run, the pattern of development Council of Governments undertook a the effects of alternative land-use (clustered or dispersed), had more similar exercise in 1991. The Council patterns on highway and transit influence on congestion than the rate of developed two alternatives to the 2010 systems. These efforts differ in the growth or the job and housing "Base Case" forecasts for the region. ways they create scenarios to evaluate mixtures. Dispersed origins and The "Balanced Scenario" involved and in the set of attributes that vary. destinations, regardless of other adding households to areas where the Among the entities conducting such assumptions, produced more highest growth in employment was studies are the Montgomery County, congestion, than clustered origins and forecast for the period 1995-2010 and Maryland, Planning Department, the destinations. They also concluded that reducing households elsewhere. An 11 Southern California Association of they could not eliminate congestion percent shift in household locations Governments, Orange County induced by growth using only highway resulted in a 13 percent increase in California Planning Department, construction, but needed to reduce transit work trips and a 10 percent Metropolitan Transit Commission and dependence upon the automobile by decrease in VMT per household, with the Association of Bay Area using a combination of policies (e.g., an overall reduction in regional VMT Governments in California's San provision of alternative modes, pricing, of 4 percent. Francisco Bay Area (ABAG 1990, etc.) to break the automobile habit and 1991), the Metropolitan Area Planning develop patterns that support transit use The "Concentrated Employment Council in Boston Massachusetts, the (MNCPPC, Montgomery County Scenario" further increased job growth Puget Sound Council of Governments Planning Department 1989). in high transit use areas to double the (Seattle, Washington), the Baltimore level in the balanced scenario. Job Regional Council of Governments, the In Baltimore, the Regional growth elsewhere was correspondingly Metropolitan Washington COG, North Council of Governments used their reduced. The redistribution of 23 Central Texas COG, and the Denver existing forecasting model to simulate percent of regional job growth over the COG. the effects of three land-use scenarios. balanced scenario resulted in a further Each involved a different allocation of increase in transit work trips Among these efforts is a study residential growth within the region's regionwide of 5 percent, but VMT for completed in New Jersey (MSM counties and communities between the region did not change significantly. Regional Council 1991). For this 1990 and 2010. project the researchers, supported by Neither the Baltimore nor the the Federal Transit Administration of The "Concentrated Growth Washington, DC scenario involved the U.S. Department of Transportation, Scenario" assumed growth in all changing the transportation network. developed a sketch plan version of a counties and jurisdictions, including the By contrast, a set of simulations, in regional highway network on which City of Baltimore. The "Sprawl Portland, Oregon, focusing on the they estimated the regional Scenario" assumed a decline in largest suburban county in the region, consequences of alternative regional population in the City of Baltimore and involved changing both transportation land-use plans and site-specific urban relatively more growth in suburban and land use to the year 2010. In design improvements. counties and areas. The "Transit Access contrast to the "Base Case" Scenario" assumed growth would occur transportation and land-use plan, the only in those zones and locations "Bypass Alternative" involved the Significant effects on slowing the having transit accessibility, regardless construction of a major freeway. The growth of trip generation and VMT, of jurisdiction. "Land Use/Transportation/Air Quality and a deterioration of highway speeds (LUTRAQ) Alternative" involved the were documented. Incremental impacts construction of light rail transit lines of 30 percent on forecast growth in The simulations suggested that the and a different assignment of trips and 33 percent in forecast growth centralized and the transit scenarios households and jobs, focusing both of VMT were shown to result from the would result in fewer vehicle trips near the rail lines. implementation of urban design overall, and in an increase of measures in a context of suburban approximately 3 percent in nonauto Researchers developed a transit- activity centers, strategically located in trips throughout the region. Overall oriented land-use plan for a 150-sq-mi the three-county region. regional travel times and miles traveled portion of the suburbs of Portland, would decline by approximately the Oregon. The land-use plan involved the Montgomery County, Maryland, same amount. Forecast air quality was assignment of households and in the rapidly growing Washington, DC worst under the sprawl scenario. The employment to traffic analysis zones area, projected the impacts of different authors noted that impacts in specific proximate to transit services (existing rates of growth, job and housing zones and jurisdictions were more or proposed) at densities consistent mixtures, and spatial patterns of varied and significant. with regional development on congestion. From their 23 market conditions. The new land-use more delay and 1 to 2 percent slower widely in the degree to which they are plan emphasized transit and pedestrian- travel speeds than the base case. The sensitive to the demonstrated effects of friendly design measures at the site and Uncongested alternative resulted in 7 land-use density, mix, and amenities on neighborhood levels. The land-use plan percent less vehicle hours of travel trip length and mode choice, and in the included a package of travel-demand (VHT) and 5 percent higher travel extent to which they model mode management measures including a speeds than the base case. Data on choice at all. Thus comparisons across $3.00 per day parking charge on work VMT, trip length, mode split, and air regions are problematic, and results of trips and a provision of an employer- quality were not presented in the report. some simulations are unreliable. paid transit pass to employees in the Nevertheless, the simulations, as a study area. In Denver, the Denver Region group, may be indicative of regionwide Council of Governments compared a impacts of land-use alternatives on The LUTRAQ Alternative for the Corridor plan to the base case Regional travel demand. Further work on model year 2010 generated a home-based Transportation Plan, which includes a development will improve the work mode share of 20.2 percent for new 100-mi radial rail system and reliability of these regional forecasts in transit in transit-oriented development circumferential freeway. The Corridor coming years. areas. This compared favorably with a plan involved concentrating 7.7 percent average mode split for employment (but not households) transit work trips in the study area within 1 mi of rail lines. International Studies under the no-action alternative and an 8.8 percent mode split for the highway- The simulations suggested no Several notable studies with an oriented alternative (which did include difference in trips, trip length, VMT, or international focus have examined the some additional transit). Looking not air quality between the alternatives, impacts of urban form on travel only at the transit-oriented development though transit patronage increased 27 behavior. Using international sites, but also at the study area as a percent in the Corridor alternative (to comparisons of U.S., European, and whole, the LUTRAQ alternative mode 10 percent of work trips and 5 percent Asian cities, Newman and Kenworthy split for home-based work trips on of all trips). The authors noted that (1989) found that U.S. cities like transit was 12.8 percent. employment up to 1 mi from transit Phoenix and Houston averaged roughly was not located sufficiently to induce a four to five times as much fuel Portland is a region with both change in mode choice, and that consumption per capita as comparable western and eastern U.S. city features, shifting employment, but not size European cities (Figure 8). in terms of the patterns of development households precluded effects on trip Differences in petrol prices, incomes, and the transit orientation. Two other length and VMT. and vehicle efficiency explain only regions that have recently conducted about half of these differences. What is similar analyses are more clearly In summary, recent work to significant is urban structure: cities western, in the sense that the vast analyze the regional effects of land use with strong concentrations of central majority of development has occurred on travel behavior indicate, for the hubs, and accordingly a better in the last 50 years. The effect of this most part, that measurable changes in developed public transport system, orientation is evident in both the behavior can be associated with have much lower energy use than cities alternatives developed, the choice of different regional development where jobs are scattered. Newman and travel behavior to be analyzed, and the patterns, but that these effects are Kenworthy also found a strong results of the simulations. diluted by the magnitude of existing relationship between density and development relative to new energy consumption within The North Central Texas Council development. With between 65 percent metropolitan areas (Table 2). For the of Governments (Dallas/Fort Worth) and 83 percent (depending on the New York region, for instance, developed three alternatives. The region and the alternative) of land uses Manhattanites average 90 gallons of "Corridors" alternative focused growth in place for both the base year and the fuel consumption per capita annually, through 2010 into radial routes. The forecast year, the effects of alternative compared to 454 gallons per capita in "Centers" alternative focused growth patterns are muted. Further, with only a the outer suburbs. This work has been into and near existing employment fraction of the future development heavily criticized, however, notably centers. The "Uncongested" alternative being allocated in a different fashion over the lack of statistical controls that spread development to peripheral areas across regional alternatives, impacts are account for other factors influencing where highway and land capacity further diminished. fuel consumption, such as differences existed. in the fuel efficiencies of U.S. versus Of course, local conditions, land- foreign fleets (Gordon and Richardson The authors of the report noted use patterns, transportation systems and 1989; Gomez-Ibanez 1991). that the Corridors and the Centers models differ from region to region. In Regardless, the analysis has spurred alternatives resulted in 2 to 3 percent addition, the models themselves vary healthy debate within public policy

25 circles about the appropriate role of Orski 1985; Leinberger and Lockwood densities had the dominant influence on central planning versus market forces 1986; Giuliano and Small 1990). modal splits, followed by land-use in responding to pressing Several other studies have looked at the mixing and parking supplies. environmental and energy consumption impacts of various land-use and problems. physical design features of activity In another study, Cervero (1989) centers on travel behavior along a classified America's largest suburban In a 1988 article, Pucher compared number of dimensions, with a particular activity centers on the basis of the size, transit modal splits for 12 countries in focus given to impacts on transit usage. densities, land-use composition, and Western Europe and North America. site designs/amenities. The study found On average, European cities were In an analysis of suburban activity all of these factors to be significant found to be on the order of 50 percent centers in metropolitan Toronto, Pill predictors of transit modal choice, with denser with substantially more mixed- (1983) found dense office and densities being the dominant factor. use neighborhoods than their American residential subcenters like North York The incidence of ridesharing and transit counterparts. Pucher found the and Scarborough to be vital in usage was the highest in suburban work percentage of all trips made by maintaining multidirectional flows on settings with the largest retail automobile to be more than double that the regional rail transit network. These components. Earlier work of subcenters of the majority of Western European centers were found to have captured in the greater Houston area reached countries, most of which have per nearly three times as many transit trips similar conclusions about the capita incomes comparable to those in for work purposes and around twice as importance of mixed uses (Rice Center the United States. America's 3.4 many for shopping purposes as other for Urban Mobility Research 1987). percent national transit modal split for non-CBD locales in metropolitan all trips was also around half that found Toronto. Cervero (1986) documented A more recent study in the in European countries. Pucher the effects of rapid suburban office Washington, DC area found denser and attributed transit's success in Europe to growth during the 1980s on travel more mixed-use employment centers to supportive urban development and behavior, finding that most (low- be more transit-dependent. Among automobile taxation policies rather than density, single-use) campus-style office workers with similar incomes, 55 transit subsidies (Figure 9). parks with abundant free parking percent of those working in downtown averaged transit modal splits under 2 Washington commuted by mass transit, Based on an evaluation of new percent, a finding also confirmed by compared to 15 percent of those British towns, Potter (1984) presents Fulton (1986) in his analysis of working in a suburban downtown evidence that communities designed for intersuburban commuting in the United (Bethesda) and only 2 percent of those good transit access enjoy higher States. working in a suburban office park ridership and more efficient services. (Rock Springs Park) (Douglas 1992) Compared to two low-density, auto- Several recent studies have (Figure 11). oriented new towns (Milton Keynes enriched our understanding of how the and Washington) (Figure 10), two built environments of suburban activity transit-friendly communities (Runcorn centers influence travel behavior. Density and Travel Behavior and Redditch) average per capita transit Hooper's (1989) survey of six mixed- ridership levels that are nearly 30 use activity centers across the United Several studies have focused percent higher (Table 3). They also States found transit modal splits to be specifically on the relationship between enjoy far more frequent bus services at consistently below 10 percent, although the employment and commercial one-third the deficit per rider of their there was considerable variation across densities at activity centers on travel auto-oriented new town counterparts. individual properties within centers behavior. On balance, research (Tables 4A and 4B). In the case of consistently shows density to be one of Bellevue, Washington, for example, 37 the most important determinants of Intermediate-Scale: Corridors and percent of workers carpooled and 12 transit modal choice. Activity Centers percent rode bus transit at an office project where parking restrictions and Two recent studies of subregions Overview Studies pricing were in place. At a nearby in the San Francisco Bay Area building where parking was abundant underscore the importance of urban The emergence of suburban and free, only 11 percent of workers densities in influencing travel behavior. downtowns and edge cities over the either shared rides or patronized transit. Using 1981 superdistrict data in the past 2 decades has spawned a number Cervero's (1991) statistical analysis of Bay Area, Harvey (1990) found a of investigations into how these built travel characteristics to sites from the strong negative exponential environments influence travel behavior National Cooperative Highway relationship between residential (Baerwald 1982; Long Island Regional Research Program (NCHRP) suburban densities and the amount of vehicular Planning Board 1984; Cervero 1984 activity centers data set revealed that travel--a doubling of densities 1986; building Figure 9. National modal-split as percent of total trips. (Source: Pucher, Urban Travel Behavior as the Outcome of Public Policy, 1988)

(a) Optimal structure for private motorized transport. Uniformly low-density to reduce traffic intensity and random distribution of facilities to even-out loading on roads

(b) Optimal structure for public transport. Urban facilities located along corridors hence concentrating demand to maintain a high frequency service. Facilities located evenly along corridors to avoid peaks in loading. Increase in density towards public transport route to minimize distances

Figure 10. Optimal urban structures for public and private transport. (Source: Potter, Stephen, "The Transport Versus Land Use Dilemma," Transportation Research Record 964, 1984) 27

TABLE 3 Key characteristics of the selected new towns (Source: Potter, Stephen, "The Transport Versus Land Use Dilemma," Transportation Research Record 964, 1984)

results in a 30 percent decline in A study of travel behavior at five averaged between 3 to 5 percent more VMT/household (Figure 12). Holtzclaw regional shopping centers in California commute trips by walking, cycling, and (1990) found a similar relationship (JHK & Associates 1993) focuses on transit than centers without on-site across five Bay Area communities with land-use mix in the vicinity of retail housing. Nowlan and Stewart (1991) similar income profiles--using data centers, and the contribution that land- present evidence that reducing job and from smog-check odometer readings use mix makes to mode choice and trip housing imbalance can improve and trip logs, Holtzclaw found that generation at these destinations. The mobility along corridors to the central residents of a dense part of San researchers surveyed a large sample of city core. They found that although Francisco logged, on average, only shoppers at centers with different levels substantial new office construction one-third as many miles on their private of transit service and with different occurred in central Toronto between vehicles each year as residents of surrounding land uses. They concluded 1975 and 1988, much of its impact on Danville, an East Bay suburb (Figure that "a significant portion of the peak-hour work trips entering the area 13). Both authors concluded that every variation found in travel mode to was offset by accelerated housing doubling of residential densities regional shopping centers can be construction. Over half of downtown reduces annual VMT by 20 to 30 explained by the amount and regional Toronto housing additions were percent. coverage of public transit service and occupied by people working there, thus the density and proximity of the allowing mobility conditions to surrounding land uses." They stabilize while office floorspace nearly Mixed-Use Developments and Travel concluded that different packages of doubled (Figure 14). Behavior travel-reduction measures, focusing on land-use mix and transit service, could Taking the other side in the debate Cervero (1989) cited the land-use reduce trips by 5 to 7 percent to these over the merits of job and housing mix as an important factor in shaping destinations. The packages of measures balance as a tool in managing region employee commuting behavior at 57 did not include parking pricing, which and mobility, Giuliano (1991, 1992) large U.S. suburban activity centers. the researchers concluded could further analyzed the location of jobs and His analysis found that a substantial reduce vehicle trips by 10 percent, but housing in a number of metropolitan retail component increases transit and "may have negative economic impact." areas and concluded that, "the ridesharing by around three percentage relationship between job and housing points for every 10 percent increase in Job and housing balance has also balance and commuting holds only in floorspace devoted to retail and gained policy attention in recent years very general terms. While isolated commercial uses (Table 5). In a as a mixed-use development strategy examples of job and housing comprehensive study of mixed-use sites which could yield mobility dividends; mismatches have been identified at the in Colorado, the ITE Colorado Section however, evidence to date is scant. In community level, there is little Technical Committee on Trip his analysis of 57 U.S. activity centers, evidence suggesting that such Generation recommended reducing ITE Cervero (1989) found that suburban mismatches have significantly affected peak hour rates by 2.5 percent when activity centers with some on-site commuting patterns. Regulatory applied to mixed-use developments. housing policies aimed at improving TABLE 4A Characteristics of selected U.S. suburban activity centers (Source: Hooper [1989])

*The employment figures for the NCHRP centers include only workers associated with the office and retail space. **The Houston study did not focus directly on the travel characteristics of residents in the centers and so no counts of residential units were done. The figures given for Bellevue and Tysons Corner represent only those surveyed and not total units in the centers.

TABLE 4B Work trip modal split--selected U.S. suburban activity centers (Source: Hooper [1989])

Note: Modal statistics were gathered for all of the centers through the administration of travel surveys However, the Houston surveys obtained only information on mode split for all trips, not just work trips. Therefore the information presented here for the Houston centers is taken from 1980 Census journey-to-work data. Figure 11. Modal shares for work trips, selected activity centers, Washington DC area (landscape). (Source: Douglas [1993])

Figure 12. Relation of residential density to VMT per resident San Francisco Bay Area. (Source: Harvey [1990]) This graph plots 1990 drive-alone commuting shares versus residential density of 33 superdistricts in the nine- county Bay Area. (The downtown San Francisco superdistrict was omitted. At 130 dwellings per acre, it had the lowest share of work trips by solo- commuters in 1990--only 20 percent.) The negative exponential function implied by the graph confirms what others have found--every doubling of density is associated with a 25-30 percent decline in drive-alone commuting.

Figure 13. Influence of residential densities on drive-alone commuting San Francisco Bay Area, 1990. (Source: Cervero [1993]) 31

job and housing balance are thus for radial trips that paralleled the rail are exacted. In the Bay Area, 92 unlikely to have any measurable impact system (JHK and Associates 1986, percent of those living within 1/4-mi of on commuting behavior, and therefore 1989). For example, around 25 percent a BART station and heading to a job in cannot be justified as a traffic of those working at the Silver Spring San Francisco where parking costs over mitigation strategy." Metro Center (near the Silver Spring $2 per day commute via rail transit. If station) patronized transit for work the workplace is in East Bay's major While the debate over the merits trips. Modal shares varied significantly employment centers in Oakland, of job and housing balance has thus by place of origin, however. If the Berkeley, Walnut Creek, or Pleasant been inconclusive, it should be noted worker was coming from Washington, Hill (all served by BART) where that the concept may be applied at DC, the transit modal share was 52 parking fees are exacted, the odds of scales as small as individual percent whereas if the trip originated in station-area residents commuting by developments and as large as entire Montgomery County the transit split BART is 45 percent. For virtually any metropolitan areas. Smaller scales was only 10 percent. The study also other Bay Area workplace location become synonymous with the concept found a number of housing projects where parking is free, fewer than 2 of mixed-use development, at which near suburban Metrorail stations where percent of station-area residents scale the Cervero (1989) research the transit modal splits exceeded 50 commute via BART. Clearly, if transit- findings apply. At the metropolitan percent, though in all cases this was based housing is to reap mobility and scale, the merits of job and housing only for work trips headed to environment dividends, it must be balance continue to be debated. Washington, DC or other places on the matched by transit-based office Metrorail line. Overall, the share of development and commercial Intermediate-Scale: Neighborhoods trips by rail or bus transit declined by clustering (Figure 15) and by pricing and Station Areas around 0.65 percent for every 100-ft strategies. increase in distance of a residential site To date, three lines of research from a Metrorail station portal. Figure 16 summarizes research have been conducted at a neighborhood done on ridership by proximity to scale on how land uses influence transit Both the Washington and transit stations. It appears that, all else trip-making: (1) studies of transit Canadian studies found that transit being equal, rail ridership potential is modal shares and ridership gradients modal splits for offices located near clearly related to station proximity. around station areas; (2) the impacts of suburban rail stations were traditional neighborhood developments considerably lower than those of Impacts of Traditional Designs and transit-oriented developments on residences near the same stations, ridership; and (3) determinants of perhaps reflecting the availability of The second line of neighborhood- pedestrian walking distances. In sufficient parking at the suburban level research has sought to empirically addition, transit agencies and local businesses surveyed. For developments measure the extent to which traditional jurisdictions have been developing near rail stations, JHK and Associates and neotraditional neighborhood design guidelines to encourage transit- (1987, p. 1) concluded that "the most designs influence travel behavior. oriented development. significant factors affecting the percent These are typically neighborhoods that of trips by transit are (1)the location of either grew around a streetcar or Transit Usage by Proximity to the site within the urban area and on the commuter line system, or in the case of Stations rail system; and (2) the proximity of the newer communities, are designed to building to a Metrorail station function like older transit-based In a study of ridership among entrance." The origin and destination neighborhoods. The central idea is to housing and commercial developments patterns of trips were found to be build suburban places that are less near four rail stations in Edmonton and crucial--"poor transit accessibility at dependent on the automobile and that Toronto, Stringham (1982) found either end of the trip results in poor are attractive environments for transit modal splits to be about 30 transit ridership between those pairs (p. walking, ridesharing, and using transit. percent higher for apartments than for 1)." The neotraditional designs of Andres single-family units. He also found the Duany and Elizabeth Plater-Zyberk "walking impact zone" to be as far as A recent examination of housing borrow many of the successful 4,000 ft from a station, a distance that and office developments near rail elements of traditional communities can accommodate around 1,200 acres stations in California has confirmed like Princeton, New Jersey and of development, sufficient to create and extended these earlier findings Annapolis, Maryland. Peter Calthorpe's strong transit-oriented communities. (Cervero et al. 1993). For housing near Pedestrian Pocket scheme adopts many rail stations, the principal determinants of these same principles, though the A study of ridership levels for of whether station-area residents will centerpiece of these projects is a rail office, residential, and hotel structures commute by rail transit were found to transit station. Among the hallmarks of near Washington Metrorail stations be the size (office and commercial these transit-friendly environments are found surprisingly high transit modal square footage) of the destination and a commercial core shares whether parking fees 32

TABLE 5 Comparison of workforce travel characteristics and areawide traffic volumes among suburban activity center groups (Source: Cervero [1989])

Variable Definitions SPEED Average travel speed for work trip in m.p.h VANSHARE Percentage of work trips in vanpools WALKSHARE Percentage of work trips by walking DRIVEDIFF Drive alone work trip percentage minus regional drive alone percentage BUSRIDE Average weekday ridership of all bus runs serving SEC ARRIVALTIME Most frequently occurring time of arrival, a m peak DEPARTIME Most frequently occurring time of departure, p m peak ADT Average daily directional traffic volume on main freeway or roadway serving SEC ______within walking distance of a majority theoretically how grid network designs densities using1980 travel diary data of residents, a well-connected can result in more direct routing of (Fehr and Peers Associates 1992) (typically grid) street pattern, narrow vehicles in traditional suburban (Figure 18). In traditional streets with curbside parking and back- subdivisions--a comparison of two neighborhoods, 23 percent of trips were lot alleys, mixed uses, and varying contrasting neighborhoods showed made on foot and 22 percent were by densities of housing (Lerner-Lam 1992; VMT could be reduced by 43 percent transit. In comparison, suburban Bookout 1992; Beimborn and with rectilinear street layouts. More residents made only 9 percent of trips Rabinowitz 1991). recent simulations by Stone and by foot and 3 percent by transit. A Several empirical investigations Johnson (1992) and McNally and Ryan follow-up study of suburban village have sought to measure the degree to (1993) confirmed that grid networks centers proposed for Stockton, CA which traditional communities affect can reduce VMT and average trip estimated there would be 25 percent travel behavior, however these efforts lengths, though they estimated fewer daily automobile trips and 33 have been hampered by the fact that reductions in the 10 to 15 percent percent less VMT in a community most neotraditional communities are range. utilizing the suburban village center still under construction or being concept. Another study, however, planned. Thus work to date has focused A study of San Francisco Bay found no significant difference in the mainly on comparing travel behavior Area travel found a dramatic difference share of walking trips to retail centers between long-established traditional in mode choice between standard among neotraditional versus communities and nearby 1960s-style suburban developments and traditional, conventional suburban neighborhoods suburban neighborhoods. Kulash et al. pre-World War II neighborhoods with in Santa Clara County, CA (Handy (1990) (Figure 17) demonstrated mixed uses and moderate to high 1992). Figure 14. Inbound trips and intra-cordon employment cumulative growth since 1975 Toronto CBD. (Source: Nowlan and Stewart [1991])

Figure 15. Ridership and station access selected U.S. cities (landscape). (Sources: As noted) Figure 16. Empirical evidence on ridership by distance. (Source: Cervero [1993])

Figure 17. Comparison of internal VMT in alternative residential development. (Source: Walter Kulash, Traditional Neighborhood Development: Will the Traffic Work?, 1990) 35

White Mt. Survey Co. (1991) such as income, age, workers per were supplemented by density and found evidence that trip generation household, persons per household or land-use mixture variables for census rates of traditional developments are automobile ownership. This constitutes tracts where survey respondents substantially below the norm. Using a serious flaw in all of the empirical resided. Researchers found mixed-use trip data compiled for two traditional research conducted to date on this neighborhoods were most strongly neighborhoods in Portsmouth, New important subject. correlated with walking trips to work, Hampshire, the authors found the but rather surprisingly mixed-use average daily traffic (ADT) generated For residential neighborhoods, environments had no influence on by these neighborhoods to be about 50 some research suggests that the mode choice for shopping trips. The percent lower than the ADT predicted availability of local retail centers can use of census tract-level data to by the Fifth Edition of the ITE Trip encourage more internal (i.e., within represent land-use compositions in Generation Manual (Figure 19). neighborhood) walking trips for neighborhoods and the entropy-based shopping purposes. In a comparison of definition of mixed uses might explain A recent study in Montgomery shopping trips made in traditional why the mixed-use variable had no County, Maryland, provides the best neighborhoods versus those in more statistical significance. insights to date on the travel auto-oriented ones in the San Francisco characteristics of traditional Bay Area, Handy (1992) found that the Interest in the influence of land- neighborhoods that are served directly traditional neighborhoods had higher use mixes is also finding its way into by rail transit (MNCPPC 1992). The rates of walking trips. Residents of large-scale regional modeling authors compared transit modal splits traditional neighborhoods averaged programs. Work for 1000 Friends of between three transit-oriented similar rates of auto travel to regional Oregon and Portland Metro has focused traditional neighborhoods (served by shopping malls, however, suggesting on identifying land-use variables for the B&O commuter railroad or a trolley that walking trips might not have inclusion in the region's travel demand line) and three nearby newer replaced auto trips but rather have models (1000 Friends of Oregon 1991). neighborhoods with a branching system supplemented them. Handy observed of streets designed for auto access. The that the destination choice for shopping Model development work for study found that residents of the transit- was affected by both the number of Portland Metro has led to the oriented communities patronized transit possible destinations and their confirmation of a statistically between 10 percent and 45 percent as proximity, as might be predicted by a significant relationship between land- much as residents of nearby auto- conventional gravity model. This use mix at the neighborhood level and oriented neighborhoods (Figure 20). suggests that a more rigorous analytical automobile ownership. The measure of framework, which accounts for the land-use mix that was tested was the Another recent study has further numbers, types, and locations of retail number of retail jobs within 1 mi of the confirmed these research findings. In establishments, is vital toward centroid of traffic analysis zones in the the San Francisco Bay Area, residents understanding the influence of land-use regional travel-demand forecasting of traditional, pre-World War II mix on mode choice in neighborhoods. model. The inclusion of this measure neighborhoods (with moderate-to-high reduced the difference between densities and grid streets) traveled by A recent study by Ewing et al. observed automobile ownership rates transit 22 percent of the time, compared (1994) employed a gravity-like and forecast automobile ownership to only 3 percent for those residing in expression of accessibility in rates at the zonal level in statistically nearby 1960s style suburban tract comparing work and nonwork travel significant ways. developments (Cervero 1993). behavior in six communities of Palm Beach County, Florida. They found Studies on Pedestrian Access Similar differences were found in little evidence that accessibility to retail the share of walking trips. Besides affected mode choice or vehicle-hours A number of studies, besides those proximity, this study matched traveled per person, and actually found examining ridership by walking communities on the basis of household a classic 1970s planned unit proximity to stations, have examined incomes and transit service intensity, development neighborhood factors influencing walking behavior. thus removing the influences of these (Wellington) had the shortest shopping Because all transit trips are linked trips factors. and recreational trips because of ample in the sense that walking is used to stores and recreational facilities within some degree to access stops or stations, With the exception of the last of the community. research on pedestrian behavior is the citations mentioned previously, highly relevant. By default, transit- none of the neighborhood studies Another study addressed the friendly environments must also be controlled or attempted to control for influence of mixed uses on work and pedestrian-friendly. other factors that are known to shopping trips in the Seattle area, using influence mode choice travel diary surveys for several hundred In Urban Space for Pedestrians, urban residents (Frank 1994). These Pushkarev and Zupan (1975) developed data Figure 18. Daily trip generation by all modes in the San Francisco Bay Area. (Source: Febr and Peers, 1992)

Figure 19. Actual counts vs. ITE trip generation projections for average daily traffic, Portsmouth, New Hampshire. (Source: White Mt. Survey Co., City of Portsmouth Traffic/Trip Generation Study 1991) Figure 20. Workers taking transit in 1980 Montgomery County, MD. (Source: MNCPPC[1992]) 38 relationships between land use and land-use environment in influencing Design Guidelines pedestrian travel demand for a pedestrian behavior (Rice Center 1987; downtown setting. The amount of Cervero 1993). Downtown Houston has Local jurisdictions and transit office, retail, residential and restaurant four times the employment density and agencies are increasingly developing space was considered as the 23 percent more sidewalk footage per design guidelines to promote transit- independent variable and the number of 1,000 workers than Uptown, a supportive development. Design pedestrians walking along block-faces suburban activity center 6 mi west of guidelines are perhaps the most visible was the dependent variable. downtown. Compared to West and prevalent means used by transit Relationships for midday and evening Houston's Energy Corridor, an axial agencies to inform and assist public and peak-period pedestrian traffic were strip along the Katy Freeway corridor private development decisions. developed. The relationships closely dotted with office parks, downtown matched expected trip generation rates Houston is nearly 10 times as dense In Oregon, a new rule in the for the area developed independently. and averages 76 percent more statewide land-use planning system Proximity to transit entrances (evening sidewalks. Downtown Houston also has requires local jurisdictions to change peak) and the amount of pedestrian skywalks and such pedestrian amenities their comprehensive plans and zoning walking space provided were also key as parks, civic plazas, benches, street codes to encourage walking, bicycling, independent variables. This report sculptures, and protection from the and transit use. Local jurisdictions are demonstrated the ability to develop elements from overhangs and trees. The typically changing standards for meaningful travel-demand land-use built environment is also more building access, parking lots, pedestrian relationships for pedestrian-scale travel. interesting downtown, consisting of an walkways, bicycle paths, and transit assortment of street-level shops, station areas. Other cities such as Untermann (1984) has examined eateries, and storefronts. Conversely, Sacramento and Toronto are also Americans' walking behavior closely. walking in Uptown and the Energy developing standards to orient His research shows that most people Corridor requires long waits at busy development towards transit (Moore are willing to walk 500 ft, 40 percent intersections, wading through and Thorsnes 1994). will walk 1,000 ft, and only 10 percent expansive surface parking lots, and will walk half a mile. These figures do passing indistinguishable urban spaces. About one-quarter of U.S. transit not specify purpose of the walk trip, As a consequence, walking/cycling agencies have established design however; for more crucial trips, such as account for around 30 percent of all guidelines to encourage development to work, Stringham (1982) has shown trips (made outside of buildings) in that supports transit, and more are acceptable walking distances to be downtown Houston, compared to 7 developing them (See Tables 6 and 7). farther. Both Untermann and Stringham percent in Uptown and only 1.9 percent Guidelines typically focus on one or have shown that acceptable walking in West Houston. The research more of three core topics: land use distances can be stretched considerably estimated that every 10 percent (type, mix, density, and location of (perhaps as much as doubled) by increase in pedestrian amenities (e.g., uses), site design (siting of buildings, creating pleasant, interesting urban lineal ft of sidewalk, number of parking, pedestrian access, street spaces and corridors. There is a need benches) is related to a 15 percent configuration), or the transit facility for more systematic research in this decline in motorized trips. (transit centers, bus stops and turnouts, area. bus shelters, bike facilities, pavement Research for 1000 Friends of and grading). Guidelines can provide Untermann contends that 10 min, Oregon (1993) examined the role of the information on transit supportive or 2,300 ft, is the maximum distance pedestrian environment at the development and support coordination Americans are willing to walk, while neighborhood level in affecting between transit agencies, developers, Canadians and Europeans are more apt vehicle-trip generation, mode choice, local jurisdictions, and the public. to walk farther. Untermann's research and VMT. The pedestrian friendliness also shows that transit passengers are of neighborhoods was evaluated using Design guidelines have, however, less sensitive to walking distances as four factors--topography, ease of street had little impact on actual development service frequency increases. crossing, sidewalk connectivity, and to date. Less than half of the transit Additionally, demographics also has street connectivity. A composite agencies with design guidelines could some bearing on willingness to walk-- Pedestrian Environmental Factor (PEF) identify projects that were influenced research shows females, those without was assigned to each zone in a regional by their guidelines. Moreover, when driver's licenses, and young people are model. Multiple regression analysis projects were identified they usually more amenable to walking. showed that increasing zonal PEF from incorporated only a modest degree of below average to average would reduce design features such as benches at bus Studies of activity centers in vehicle-trip generation per household stops or special drop-off lanes for greater Houston underscore the by 7 percent in that zone, after buses. Few examples of dense, mixed- importance of pedestrian amenities as controlling for household income, size, used well as the and auto ownership. 39 centers that support transit use were the past 20 years, most notably the One of the first efforts of this kind identified. One problem is that transit suburbanization of employment, this has been undertaken by the Puget agency guidelines are seldom binding work needs to be updated. In addition, Sound Council of Governments in on development (Cervero 1993). a number of other aspects of this Seattle, Washington (1990). The relationship needs to be systematically project, known as Vision 2020, explored and synthesized. involved analysis and comparison of Microlevel Analyses several long-range regional transportation and land-use plans in a Few evaluations of transit demand 1.5 INTERACTIVE IMPACTS OF manner similar to that of the MSM have been conducted at the individual TRANSIT AND URBAN FORM (1991) study previously cited (See also site/building level. The NCHRP Watterson 1990). The Puget Sound suburban activity centers data set has Thus far, the citations in this work compared and contrasted six yielded several studies that reveal the literature review have focused on land-use and transportation packages. sensitivity of transit demand to building unidirectional impacts on transit and The results are shown in Table 8. The densities, on-site services, and parking urban form. While acknowledging that titles of each of the alternatives are supplies (Hooper 1989; Cervero 1991). transit and urban form interact, the indicative of their characteristics. The researchers have found little empirical "existing plans" alternative involves the Several site-level studies have data of significance on this subject to implementation of existing land-use examined what happens to commuting feature in this literature review. plans and regional transportation behavior when downtown office projects (a mix of highways and transit workers are relocated to a suburban At the same time, it must be system improvements). The "major work location. Cervero and Landis acknowledged that the interactions centers" and "multiple centers" (1992) found that transit modal splits between transportation and land use are alternatives and the dispersed growth fell from 58 percent to 3 percent for a central issue among researchers alternative represent clear choices for office workers that were located from today. A number of individuals and the region in terms of density, downtown San Francisco (well-served organizations have developed and nucleation, and activity mixing. The by BART) to three suburban campus refined experimental procedures-- "preferred alternative" is a hybrid of the locations (that were poorly served by including interactive transportation and major centers and multiple centers bus). Similar work on office relocation land-use models--to simulate these alternatives. impacts in England (Wabe 1967; interactions. The state of the art in Daniels 1972, 1981) and Canada (Ley interactive transportation and land-use Using an interactive transportation 1985) found that commute distances modeling is summarized best in and land-use model system, the Puget typically fell slightly after jobs Webster, et al. (1988) and Wegener Sound researchers forecast variations in suburbanized, however there was a far (1994). Their work is summarized in transit mode split at peak hour from 5.6 more dramatic switch in commuting 1000 Friends of Oregon, Vol. 1 (1991). percent to 14.8 percent. Percentage modes, from public transit to the increases in transit ridership over private automobile. While it is beyond the scope of current levels varied from 20 percent to this literature review to summarize or 241 percent for the alternatives studied. describe the various model systems The results provide evidence in support Conclusions available to simulate the interactions of of the hypothesis that interactions transportation and land use, it is between transportation investments and Understanding how the densities, appropriate to note that a number of land development patterns at the settlement patterns, land-use metropolitan organizations, regional scale may result in divergent compositions, and urban designs of universities, and nonprofit outcomes for regional transit patronage. cities and neighborhoods influence organizations have undertaken and are transit usage is of vital importance to conducting research into these Recently Metro (1994a and transit planners and decision makers. interactions, using one or another of the 1994b), the Portland, Oregon regional Whether a future light rail extension models described in the citations government, evaluated five different will be a cost-effective investment or mentioned. Metropolitan planning land-use and transportation scenarios in whether headways should be increased organizations in U.S. cities such as Los a project known as Region 2040. on a conventional bus route hinges Angeles, San Francisco, Chicago, New Summary results are shown in Table 9. critically on whether the built York, Seattle, San Diego, Phoenix, The base case continues current environment will support these Washington DC, and elsewhere have trends of low-density, automobile- changes. The seminal work on such undertaken simulations as part of long- oriented suburban development for questions by Pushkarev and Zupan range regional and transportation most growth. The urban area is (1977) has provided some guidance. planning efforts in recent years. expanded by about 50 percent Because of significant changes in the including 63,900 acres of landscape of urban American over

41 farmland (EFU for Exclusive Farm Use relationships are revealed by what we in progress that will enhance further the in Table 9). Growth Concept A, or know about location behavior and how state of knowledge. This section "growing out," accommodates growth proximity to transit shapes ridership. reviews some of the more significant by expanding the metropolitan urban As noted, those living near rail stations research now under way. growth boundary by about 25 percent are more likely to commute by transit mostly on lands currently zoned for than those living farther away, all rural residential development. Some things being equal. The reality, Transit Impacts on Urban Form and growth is concentrated in transit however, may be that all things are not Land Use corridors. Road improvements, three usually equal. There is increasing new freeways, and a radial high- evidence that one reason station-area Research is soon to begin on the capacity transit system are included in residents are more inclined to use update of the study by Knight and this scenario. Growth Concept B, or transit is that they work downtown and Trygg (1977) described in a previous "growing up," accommodates growth in other locations well-served by section of this report. This work will without expanding the urban growth transit. Voith (1991) has found strong focus on intermediate and macroscale boundary. Instead both employment evidence of residential sorting based on impacts of transit on land use. and residential growth are concentrated employment location in the Researchers are reviewing the 1977 in centers that support greater use of Philadelphia region. He estimated that work to identify those criteria that walking, bicycling, and transit. No the percentage of the labor force in a effectively indicate and measure the freeway additions are built, some street census tract that work in the CBD is 12 land-use patterns and the economic improvements are made, but most percent higher for tracts with SEPTA development benefits that result from transportation improvements are in the and PATCO commuter rail services construction of fixed guideway transit transit network. Both radial and ring nearby. Thus, transit and land-use facilities. Research and data collection high-capacity transit are developed. In interactions are underscored by the for the update of the Knight and Trygg Growth Concept C, or "satellite cities," tendency of those working in areas with study were initiated during Fiscal Year a third of regional housing and good transit service to chose to live in 1994. employment growth occurs in existing areas with good transit access. This cities outside the metropolitan urban induces higher ridership, which in turn At the same time, work at the growth boundary. The remainder is should--over the long run--lead to macroscale is underway through accommodated by expanding the urban improved services (e.g., more frequent updates of studies conducted in the growth boundary modestly and headways). Finally this will induce 1970s on the impacts of heavy rail concentrating more growth in centers even more residents to search for home systems. In the San Francisco Bay area, than in Concept A. Concept C includes sites near transit stops. the Atlanta region, and the Washington, the full range of transportation DC metropolitan area, studies are improvements in Concept A plus high- The interactive nature of transit planned or are in progress to examine capacity transit service to the satellite and urban form, while complex, can the effects of the BART, MARTA, and cities. Finally, after evaluating the base potentially be conveyed through a WMATA systems, respectively. These case and growth concepts, a balance of modeling work and carefully efforts should shed light on the impacts recommended alternative was constructed empirical investigations of "new generation" rail investments on developed. that look at the joint influence of transit station areas, corridors, and urban and on residential location and ridership. At regional form a generation after initial The recommended alternative a minimum, studies of simultaneous construction. concentrates development in regional and lagged relationships require good and town centers plus transit corridors times-series data and often require the At the microscale of analysis, a and station areas. It includes addition of various statistical controls. research project is underway at the considerable redevelopment and small The data and modeling requirements Lincoln Institute of Land Policy, additions to the urban growth for probing transit and urban form Cambridge, Massachusetts, to examine boundary. Improvements are made in interactions can be considerable; the factors that contribute to both the transit and roadway system, however, the policy insights that can be residential-location decision making. with an emphasis on frequent-service gained from such efforts can be The key question in this study is the bus corridors and arterial street service. substantial. extent to which changes in employment By concentrating growth where it can location trigger residential relocations be served by transit and connecting at the household level. This question is regional centers with frequent bus 1.6 RESEARCH IN PROGRESS central to the validity of many land-use service, transit ridership is the highest models currently employed by of any alternative. While significant and valuable metropolitan areas in the United States. work has been completed on transit and Data from a sample of households in Besides simulation work, the urban form relationships, a number of the dynamics of transit and urban form important research projects are currently TABLE 8 Puget Sound transportation systems performance by alternative (Source: PSCOG [1990]) TABLE 9 Region 2040--comparison of alternative--summary

* The base case did not have parking factors and pedestrian factors modeled consistent with the other growth concepts.

Source: Metro 1994b 44

Minneapolis-St. Paul metropolitan area At the neighborhood scale, at least the role of transit and transit supportive have been collected and are awaiting three ongoing studies are designed to land uses. analysis for this purpose. examine the role of density, mix, and amenities on travel behavior. In the The principal effort is ongoing in state of Washington, University of the Portland, Oregon metropolitan area. Land-Use Impacts on Transit Washington researchers have gathered The LUTRAQ study, for 1000 Friends Demand data about travel in several "traditional" of Oregon, will begin use of an neighborhoods (pre-1945) through a interactive land-use model to simulate A key research project is well travel diary of a sample of residents. A the effects of alternative transportation under way for the Federal Highway variety of travel behavior statistics for systems and land-use plans on urban Administration whose purpose is to this sample will be compared with form over a 40-year period. Three examine the effects of site design at the norms for the region as a whole different mixes of transportation workplace on journey-to-work mode (Washington State Transportation investments (one emphasizing light rail choice. A large sample of work sites in Commission 1993). transit and buses) will be modeled for a southern California, for which journey- 150-sq-mi suburban study area. to-work data are available, have been Aggregate measures of travel behavior analyzed for their urban design features In California, a similar project is will be developed as a result of the that contribute to pedestrian and transit being conducted by the University of simulations for the study area as a friendliness. These features are being California, Davis. Six neighborhoods in whole. These measures will include analyzed for their ability to explain the San Francisco Bay area are being indicators of mode choice for varying observed variations in mode choice in examined, each differing in density, trip purposes. journey to work, in tandem with land-use mix, and income. Aggregate demand-management programs in measures of household travel behavior This study will also include an place at the employer level. at the neighborhood level will be explicit comparison of the results of a compared and contrasted in these simulation using an interactive neighborhoods. The study is designed transportation and land-use model with Also at the microscale, work by to build on prior work by Holtzclaw a simulation using a traditional travel- Bruce Douglas of the TCRP research (1990). While data collection is demand forecasting model in which team is continuing in Montgomery complete, full data analysis awaits the land use is specified as an input. County, Maryland, where travel diaries involvement of other researchers. Forecast travel behavior in the year are being collected for employees at 2010 will be compared. suburban work sites of varying densities and land-use characteristics, Interactive Impacts of Transit and A number of similar exercises all located near rail stations. These data Urban Form may be conducted by other could complement the available data metropolitan planning organizations, from the southern California study by In addition to the interactive transit agencies, and other groups in the focusing on rail-oriented work sites. simulation previously described (Puget next few years. Many are based on (The southern California data focuses Sound Council of Governments 1990, model systems calibrated on actual on bus- and auto-oriented sites.) Metro 1994a, 1994b), there are several travel behavior in the region being Information on urban design features at other examples of experimental studied. Work in this important area these Montgomery County sites is research under way to simulate the will continue in the 1990s. being assembled. interactions of transportation and land use, emphasizing 45

BIBLIOGRAPHY

The Impact of Transit on Urban Form

Allen, W. R. et al. The Impact of Rapid Transit on Black, J. T. et al. Downtown Office Growth and the Role of Suburban Residential Property Values and Land Public Transit. Urban Land Institutes, Washington, DC Development: An Analysis of the Philadelphia High-Speed (1982). Line. Philadelphia, University of Pennsylvania, Wharton School (1973). Blayney Associates. The Study of Property Prices and Rents: BART Impact Study. Berkeley, Metropolitan Alonso, W. Location and Land Use: Toward a General Transportation Commission (1978). Theory of Land Rent. Cambridge, Harvard University Press (1964). Bonsall, J. A Bus for All Seasons. Ottawa, Ottawa-Carleton Regional Transit District (1985). Al-Mosaind, M. A. et al. "Light-Rail Transit Stations and Property Values: A Hedonic Price Approach." Boyce, D. The Impact of Rapid Transit on Suburban Transportation Research Record 1400. Transportation Residential Property Values and Land Development. Research Board, Washington, DC (1993) pp. 90-94. Washington, DC, U.S. Department of Commerce (1972).

Arlington County, Department of Community Planning, Boyce, D. et al. The Impacts of Rapid Transit on Suburban Housing and Development. Development in the Metro Property Values and Land Development: Analysis of the Corridors. Arlington, Virginia, DCPHD (1990). Philadelphia-Lindenwold High Speed Line. Philadelphia, University of Pennsylvania (1972). Arrington, G. B. Portland's Light Rail: A Shared Vision for Transportation and Land Use. Portland, Oregon, Tri-Met Boyce, D. et al. "Impact of Rapid Transit on Residential (1992). Property Sales Prices." Space, Location and Regional Development. London, Pion Press (1976) pp. 145-53. Armstrong, R. J., Jr. "Impacts of Commuter Rail Service as Reflected in Single-Family Residential Property Values." Burkhardt, R. Summary of Research: Joint Development Paper presented at the 73nd Annual Meeting of the Study. New York, Administration and Managerial Research Transportation Research Board (1994). Association (1976).

Baldassare, M., R. Knight, and S. Swan. "Urban Service Calgary, City of. LR T Impact Study Part I. Home Interview and Environmental Stressor: The Impacts of the Bay Area Survey. Calgary, Transportation Department (1983). Rapid Transit System on Residential Mobility." Environment and Behavior. Vol. 11, No. 4 (1979) pp. 125- Callow, J. "Impacts of Transit Facilities on Land Use." ITE 41. Journal. (January 1992) pp. 37-39.

Barney and Worth, Inc. et al. Evaluation of Banfield Light Cambridge Systematics. The Impacts of UMTA Programs- Rail Transit Station Area Planning Program: Summary Case Studies of Four Urban Areas. Washington, DC, Report for Tri-Met. Portland, Oregon (1993). Urban Mass Transportation Administration (August 1988).

Bernick, M. "The Bay Area's Emerging Transit-Based Cervero, R. "Light Rail Transit and Urban Development." Housing." Urban Land. (July 1993) pp. 38-41. Journal of the American Planning Association. Vol. 50, No. 2 (1984) pp. 133-47. Bernick, M. and M. Carroll. A Study of Housing Built Near Rail Transit Stations: Northern California Experiences. Cervero, R., M. Bernick, and J. Gilbert. Market Berkeley, Institute of Urban and Regional Development, Opportunities and Barriers to Transit-Based Development UC Berkeley, WP-546 (1991). in California. Berkeley, National Transit Access Center, University of California (1994). Biehler, A. "Exclusive Busways Versus Light Rail Transit: A Comparison of New Fixed-Guideway Systems." TRB Chicago Area Transportation Study. The O'Hare Transit Special Report 221. National Research Council, Extension Before and After Analysis. (1986). Washington, DC (1989). 46

Ciocca, D. G. "Financing Transit Based Development in Edner, S. M. and G. B. Arrington. Urban Decision Making California." Paper presented to the Urban Land Institute, for Transportation Investments - Portland's Light Rail San Diego, California (April 1994). Transit Land. Washington DC, Urban Mass Transportation Administration (March 1985). Clark, W. and J. Burt. "The Impact of Workplace on Residential Relocations," Annals of the Association of Emerson, D. J., Office of Planning, UMTA. "A Framework American Geographers. Vol. 70 (1980) pp. 59-67. for Analyzing the Land Use and Urban Development Impacts of Fixed Guideway Transit Projects." (Presented at Czamanski, S. "Effects of Public Investments On Urban TRB Williamsburg Conf. on Transportation and Economic Land Values." Journal of American Institute of Planners. Development 1989). (July 1976) pp. 204-217. Fejarang, R. A. "Impact of Property Values: A Study of the Damm, D. et al. "Response of Urban Real Estate Values in Los Angeles Metro Rail." (Paper presented at the 73nd Anticipation of the Washington Metro." Journal of Annual Meeting of the Transportation Research Board, Transport Economics and Policy. (September 1980) pp. Washington, DC, 1994). 315-35. Figueroa, O. "Santiago Metro: Integration and Public Davis, E. et al. Transit-Linked Development: A Case Study Transport." Rail Mass Transit for Developing Countries. of Atlanta's MARTA System. Washington, DC, U.S. London, Thomas Telford (1990). Department of Transportation (1985). Folgelson, R. The Fragmented Metropolis: Los Angeles Deakin, E., R. Cervero, and J. Landis. BART at 20: from 1850 to 1930. Cambridge, Harvard University Press Research Design. Berkeley, Institute of Urban and (1967). Regional Development (1994). Gannon, C. and M. Dear. "Rapid Transit and Office Dear, M. "Rapid Transit and Suburban Residential Land Development." Traffic Quarterly. Vol. 29 (1972) pp. 223- Use." Traffic Quarterly. Vol. 29, No. 2 (1975) pp. 223-42. 42.

Dewees, D. "The Effect of a Subway on Residential Gatzlaff, D. H. and M. T. "The Impact of Miami Metrorail Property Values in Toronto." Journal of Urban Economics. on the Value of Residences Near Station Locations." Land Vol. 3 (1976) pp. 357-69. Economics, Vol. 69, No. 1 (1993).

Dingemans, D. "Rapid Transit and Suburban Residential Land Use." Traffic Quarterly. Vol. 32, No. 2 (1978) pp. Gilpin, G. Results of the 1990 Transportation Systems 289-306. Management Program. Pleasanton, California, City of Pleasanton, Planning Department (1992). Donnelly, P. Rail Transit Impact Studies: Atlanta, Washington, San Diego. Washington, DC, Urban Mass Giuliano, G. "Is Jobs Housing Balance a Transportation Transportation Administration (1982). Issue?" Transportation Research Record 1305. Transportation Research Board, Washington, DC (1991) Dornbush, D. "BART-induced Changes in Property Values pp. 305-312. and Rents." Land Use and Urban Development Projects, Phase I, BART Impact Study. Washington, DC: U.S. Giuliano, G. "Land-use Impacts of Transportation Department of Transportation and U.S. Department of Investments: Highway and Transit." The Geography of Housing and Urban Development (1975). Urban Transportation. New York, Guilford Press, (1986) pp. 247-279. Dueker, K. et al. The Portland Mall Impact Study. Washington, DC, U.S. Department of Transportation, Urban Mass Transportation Administration (1982). Gladstone Associates. Economic Benefits Study: 16th Street Transitway Mall: Denver, Colorado. Washington, DC, U.S. Department of Transportation, Urban Mass Transportation Dyett, M. et al. The Impact of BAR Ton Land Use and Administration, Office of Planning Assistance (1982). Urban Development: Interpretative Summary of the Final Report. Gomez-Ibanez, J. A. "A Dark Side to Light Rail? The San Francisco, BART Impact Program, Prepared for U.S. Experience of Three New Transit Systems." Journal of the Department of Transportation (1979). American Planning Association. Vol. 51, No. 3 (1985) pp. 337-51. 47

Gomez-Ibanez, J. A. "Transportation Policy as a Tool for Larwin, T. LRT Impact on Real Estate Values. San Diego, Shaping Metropolitan Development." Research in MTDB (1992). Transportation Economics. Vol. 2 (1985) pp. 55-81. Lerman, A. et al. The Effect of the Washington Metro on Green, R. D. and 0. M. James. Rail Transit Station Area Urban Property Values. Cambridge, MIT (1978). Development: Small Area Modeling in Washington, DC Armonk, New York, M.E. Sharpe (1993). Lerner, J. Curitiba: The Ecological Revolution. Curitiba, Jaime Lerner Institute (1993).

Harrison, D. The Impact of Transit Systems on Land Use Lutin, J. and G. Benz. "Key Issues in Light Rail Transit Patterns in the Pre-Automobile Era. Discussion Paper D78- Station Planning and Design." Transportation Research 21, Cambridge, Harvard University, Department of City Record 1361, Transportation Research Board, Washington, and Regional Planning (1978). DC (1992) pp. 117-124.

Harrison, D. Transportation and the Dynamics of Urban Meakin, R. "Hong Kong's Mass Transit Railway: Vital and Land Use. Discussion Paper D78-22, Cambridge, Harvard Viable." Rail Mass Transit for Developing Countries. University, Department of City and Regional Planning London, Thomas Telford (1990). (1978). McGean, T. J. et al. Assessment of the San Diego Light Rail Harrison, D. and J. Kain. "Cumulative Urban Growth and System. Urban MassTransportation Administration, Urban Density Functions." Journal of Urban Economics. Washington, DC (November 1983). Vol. 1 (1978) pp. 61-98. Metropolitan Washington Council of Government Heenan, W. "The Economic Effect of Rapid Transit on (MWCOG). Trends Before Metrorail. Washington, DC, Real Estate Development." The Appraisal Journal. Vol. 36 MWCOG (1982). (1968) pp. 212-24. Metropolitan Washington Council of Government Hilton, G. "Rail Transit and the Pattern of Cities: The (MWCOG). "Metro: Before and After Program, Assorted California Case." Traffic Quarterly. Vol. 67 (1968) pp. Materials and Reports." Washington, DC, MWCOG (1980- 379-93. 84).

Hoover, E. The Location of Economic Activity. New York, Meyer, J. and J. A. Gomez-Ibanez. Autos, Transit, and McGraw-Hill (1948). Cities. Cambridge, Harvard University Press (1981).

Institution of Civil Engineers. Rail Mass Transit for Meyer, M. "Economic Development Impacts of Rail Developing Countries. London, Thomas Telford (1990). Investments." Georgia Tech, (1989).

ITE Technical Committee 6Y-38. Impacts of Transit Middleton, W. The Time of the Trolley. Milwaukee, Facilities on Land Use. Washington, DC, Institute of Kalmbach Publishing (1967). Transportation Engineers (1992).

Isard, W. Location and Space Economy. Cambridge, MIT Mullins, J. III et al. "Land Use Impacts of the Houston Press (1956). Transitway System." Transportation Research Record 1237, Transportation Research Board, Washington, DC Knight, R. and L. Trygg. "Evidence of Land Use Impacts (1989) pp. 29-38. of Rapid Transit Systems." Transportation. 6 (1977) pp. 231-47. Nelson, A. C. and S. J. McCleskey. "Improving the Effects of Elevated Transit Stations on Neighborhoods." Landis, J., S. Guhathakurta, and M. Zhang. "Capitalization Transportation Research Record 1266. Transportation of Transportation Investments into Single Family Home Research Board, Washington, DC (1990) pp. 173-180. Prices: A Comparative Analysis of California Transit Systems and Highways." Berkeley, Institute of Urban and Neuwirth, R. "Economic Impacts of Transit on Cities." Regional Development, University of California, Working Transportation Research Record 1274. Transportation Paper, 619 (1994). Research Board, Washington, DC (1990) pp. 142-149. 48

Paaswell, R. and J. Berechman. An Analysis of Rapid U.S. Department of Transportation. "Rail Transit Impact Transit Investments: The Buffalo Experience. Washington, Studies: Atlanta, Washington, San Diego." Urban Mass DC, Urban Mass Transportation Administration (not Transportation Administration, U.S. Department of dated). Transportation (March 1982).

Page, J. A. et al. Catalog of Transit Station Impact Case U.S. Department of Transportation. San Diego Trolley: The Studies. RSPA, U. S. Department of Transportation, First Three Years. Urban Mass Transportation Washington, DC (1983). Administration, Washington, DC (November 1984).

Pisarski, A. New Perspectives on Commuting. Washington, Vance, J. E. Geography and Urban Evolution in the San DC, U.S. Department of Transportation, Federal Highway Francisco Bay Area. Berkeley, Institute of Governmental Administration (1992). Studies (1964).

Rabinovitch, J. Environmental Innovation and Voith, R. "Changing Capitalization of CBD-Oriented Management in Curitiba, Brazil. Washington, DC, The Transportation Systems: Evidence from Philadelphia, 1970- International Bank For Reconstruction and Development, 1988." Journal of Urban Economics, Vol. 33 (1993) pp. Urban Management Programme, Work Paper 1 (1993). 361-376.

Rice Center for Urban Mobility Research. Assessment of VNI Rainbow Appraisal Service. Analysis of the Impact of Changes in Property Values in Transit Areas. Houston, Light Rail Transit on Real Estate Values. Prepared for the Rice Center (1987). Metropolitan Transit Development Board (1993).

Rybeck, W. Metrorail Impacts on Washington Area Land Warner, S. B. Streetcar Suburbs. Cambridge, Harvard Values. Washington, DC, Subcommittee on the City, University Press (1962). Committee on Banking, Finance, and Urban Affairs, U.S. House of Representatives (1981). Watterson, W. T. "Estimating Economic and Development Impacts of Transit Investments." Transportation Research SANDAG. San Diego Trolley: The First Three Years. San Record 1046. Transportation Research Board, Washington, Diego (1984). DC (1985) pp. 1-9.

Schaeffer, K. and E. Sclar. Access for All: Transportation Webber, A. Theory of the Location of Industries. Chicago, and Urban Growth. New York, Columbia University Press University of Chicago Press (1929). (not dated). Webber, M. "The BART Experience: What Have We Shaw, J. "Transit-Based Housing and Residential Learned?" Public Interest. Vol. 12, No. 3 (1976) pp. 79- Satisfaction: Review of the Literature and Methodological 108. Approach." Transportation Research Record 1400. Transportation Research Board, Washington, DC (1993), Wilson, R. and J. Anderson. "Planning for Transit-Oriented pp. 82-89. Development in San Diego and Vancouver, British Columbia." (Paper delivered at ACSP Annual Meeting, Smerk, G. "The Streetcar: Shaper of American Cities." October 1993) Traffic Quarterly. (October 1967) pp. 134-46. Wolpert, J. "Regional Economic Growth and Highway Spengler, E. Land Values in New York in Relation to Congestion." Woodrow Wilson School, Princeton Transit Facilities. New York, NYTA (1930). University (1991).

Stockholm, City of. The Development of Stockholm, Stockholm (1989). Zimmerman, S. "UMTA and Major Investments: Evaluations Process and Results." Reported in Transit Transportation Research Board. "Light Rail Transit: New Administration and Planning Research. Transportation System Successes at Affordable Prices." TRB Special Research Record 1209. Transportation Research Board, Report 221. National Research Council, Washington, DC Washington, DC (1989) pp. 32-36. (1989). Zupan, J. Rail Transit as an Economic Development Tool. U.S. Department of Transportation. Land Use Impacts of New Jersey Transit (1993). Rapid Transit: Implications of Rapid Transit. (December 1977). 49

The Impact of Urban Form on Transit Demand and Cervero, R. "Transport Changes Generated by Travel Behavior Decentralized Offices: A Second Survey." Regional Studies. Vol. 15, No. 6 (1981) pp. 507-20.

Baerwald, T. "Land Use Change in Suburban Clusters and Clark, J. The Effect of Urban Form on Travel Demand and Corridors." Transportation Research Record 861, Transportation Energy Consumption: Theoretical Transportation Research Board, Washington, DC (1982) Considerations. Ph.D. Dissertation, University of pp. 7-12. Washington (1976).

Beimborn, E. and H. Rabinowitz. Guidelines for Transit Daniels, P. "Transport Changes Generated by Sensitive Suburban Land Use Design. Washington, DC, Decentralized Offices: A Second Survey." Regional U.S. Department of Transportation (1991). Studies. Vol. 15, No. 6 (1981) pp. 507-520.

Bookout, L. "Neotraditional Town Planning." Urban Land, Douglas, B. Comparison of Commuting Trends Between Vol. 51, No. 2 (1992) pp. 10-15. Downtown, Suburban Centers, and Suburban Campuses in the Washington Metropolitan Area. Washington, DC, Burby, R. et al. New Communities USA. Lexington, Parsons Brinckerhoff Quade & Douglas, Inc. (1992). Massachusetts, Heath (1974). Dunphy, R. and D. Fisher. "Transportation, Congestion, California Air Resources Board. The Linkage Between and Density: New Insights." (Paper presented at the 73nd Land Use, Transportation and Air Quality: Draft Report. Annual Meeting of the Transportation Research Board, Sacramento (June 1993). Washington, DC, 1994)

Cervero, R. "Managing the Traffic Impacts of Suburban Dyett, M. "Site Design and Its Relation to Urban Form. Office Growth." Transportation Quarterly. Vol. 51, No. 3 Transportation, Urban Form, and the Environment." TRB (1984) pp. 533-550. Special Report 231. National Research Council, Washington, DC (1991) pp. 117-126.

Cervero, R. "A Tale of Two Cities." Journal of Eager, W. "Accommodating Land Use and Transportation." Transportation Engineering (1985). Land Use In Transition. Washington, DC, The Urban Land Institute (1993) pp. 32-37. Cervero, R. Suburban Gridlock. New Brunswick, New Jersey, Rutgers University Press (1986). Edwards, J. and J. Schofer. "Relationships Between Cervero, R. America's Suburban Centers: The Land Use- Transportation Energy Consumption and Urban Structure: Transportation Link. Boston, Unwin-Hyman (1989). Results of Simulation Studies." Transportation Research Record 599. Transportation Research Board, Washington, Cervero, R. "Land Uses and Travel at Suburban Activity DC (1976) pp. 52-59. Centers." Transportation Quarterly, Vol. 45, No. 4 (1991) pp. 479-91. Ewing, R., P. Haliyr, and G. Page. "Getting Around a Traditional City, and Suburban PUD, and Everything In- Cervero, R. and J. Landis. "Suburbanization of Jobs and the Between." (Paper presented at the 73rd Annual Meeting of Journey to Work: A Submarket Analysis of Commuting in the Transportation Research Board, Washington, DC the San Francisco Bay Area." Journal of Advanced 1994). Transportation. Vol. 26, No. 3 (1992) pp. 275-97. Federal Highway Administration. Edge City and ISTEA- Examining the Transportation Implications of Suburban Cervero, R. et al. Ridership Impacts of Transit-Focused Development Patterns. Searching for Solutions, Policy Development in California. Berkeley, Institute of Urban Discussion Series No. 7. Washington, DC (1992). and Regional Development (1993).

Cervero, R. et al. Transit-Supportive Development in the Fehr and Peers Associates. Metropolitan Transportation United States: Experiences and Prospects. Berkeley, Commission Bay Area Trip Rate Survey Analysis. Oakland, California, Institute of Urban and Regional Development Metropolitan Transportation Commission (1992). (1993). 50

Fehr and Peers Associates. Effect of Stockton's Proposed Holtzclaw, J. "Manhattanization versus Sprawl: How Suburban Village Center Development on Travel Mode Density Impacts Auto Use Comparing Five Bay Area Choice and Auto Use. Stockton, City of Stockton (1992). Communities." Proceedings of the Eleventh International Pedestrian Conference. Boulder, City of Boulder (1990) Frank, L. "The Impacts of Mixed Use and Density on the pp. 99-106. Utilization of Three Modes of Travel: The Single Occupant Vehicle, Transit, and Walking." (Paper presented at the Hooper, K. Travel Characteristics at Large-Scale 73rd Annual Meeting of the Transportation Research Suburban Activity Centers. Alexandria, Virginia, JHK & Board, Washington, DC, 1994) Associates, NCHRP Project 3-38(2), Transportation Research Board, National Research Council (1989). Friedman, B. et al. The Effect of Neo Traditional Neighborhood Design on Travel Characteristics. Hoyt, H. The Structure of Growth of Residential (Presented at the 1992 ITE District Annual Meeting, Neighborhoods in American Cities. Washington, DC, U.S. Anchorage, Alaska, July 1992). Government Printing Office (1939).

Fulton, P. Changing Journey-to-Work Patterns: The JHK & Associates. Analysis of Indirect Source Trip Increasing Prevalence of Commuting Within Suburbs in Activity at Regional Shopping Centers. Prepared for Metropolitan Areas. (Paper presented at the Annual California Air Resources Board, Report A132-094 (1993). Meeting of the Transportation Research Board, Washington, DC, 1986). JHK & Associates. Development-Related Ridership Survey I. Washington, DC, Washington Metropolitan Area Transit Giuliano, G. "Land Use Impacts of Transportation Authority (1986). Investments: Highway and Transit." The Geography of Urban Transportation. New York, Guilford Press (1985). JHK & Associates. Development-Related Ridership Survey II. Washington, DC, Washington Metropolitan Area Transit Giuliano, G. and K. Small. Subcenters in the Los Angeles Authority (1989). Region. Berkeley, University of California, Center for Transportation, Working Paper (1990). Kain, J. and R. Fauth. "The Effect of Urban Structure on Auto Ownership and Journey to Work Modes." Gomez-Ibanez, A. "A Global View of Automobile Transportation Research Record 648, Transportation Dependence." Journal of the American Planning Research Board, Washington, DC (1977) pp. 9-17. Association. Vol. 57, No. 3 (1991) pp. 376-79. Kulash, W. et al. Traditional Neighborhood Development: Gordon, P. and H. Richardson. "Gasoline Consumption and Will the Traffic Work? Washington, DC, Real Estate Cities: A Reply." Journal of the American Planning Research Consultants (1990). Association. Vol. 55, No. 2 (1989) pp. 342-45. Leinberger, C. and C. Lockwood. "How Business is Hall, P. "Land-use Change and Transport Policy." Habitat Reshaping America". The Atlantic Monthly. (October 1986) International. Vol. 7 (3/4) (1983) pp. 67-77. pp. 43-51.

Handy, S. Regional Versus Local Accessibly: Implications Lerner, H. "Curitiba's Transit Success." Berkeley, Institute for Non-Work Travel. Doctoral Dissertation, University of of Urban and Regional Development (1993). California, Davis, Institute of Transportation Studies (1992). Lerner-Lam, E. et al. "Neo-Traditional Neighborhood Design and Its Implications for Traffic Engineering." ITE Journal. January 1992) pp. 17-24. Handy, S. How Land Use Patterns Affect Travel Patterns: A Bibliography. CPL Bibliography 279, Chicago, Council Levinson, D. M. and A. Kumar. "Is Residential Density a of Planning Librarians (1992). Transportation Issue?" (Submitted to Urban Studies, 1993, Unpublished). Hanson, S. and M. Schwab. "Accessibility and Intraurban Travel". Environment and Planning. H 19 (1987) pp. 735- Levinson, H. and F. Wynn. "Effects of Density on Urban 748. Transportation Requirements." Highway Research Record. Vol. 2 (1963) pp. 38-64. Harvey, G. Relation of Residential Density to VMT Per Resident: Oakland. Metropolitan Transportation Commission (1990). 51

Ley, D. "Work-Residence Relations for Head Office 1000 Friends of Oregon. The LUTRAQ Alternative/Analysis Employees in an Inflating Housing Market." Urban of Alternatives: An Interim Report. Portland, Oregon Studies. Vol. 22, No. 1 (1985) pp. 21-38. (October 1992).

Long Island Regional Planning Board. The Journey to 1000 Friends of Oregon. Making the Land Use Work to Major Employment Centers. Hauppuage, New Transportation Air Quality Connection, Modeling York, LIRPB (1984). Practices, Vol.1. Portland, Oregon (October 1991).

Maring, G. E. "Pedestrian Travel Characteristics." Highway Orski, K. "Suburban Mobility: The Coming Transportation Research Record 406. Washington, DC, Highway Research Crisis?" Transportation Quarterly. Vol. 39, No. 2 (1985) Board (1972). pp. 283-96.

Peat Marwick and Mitchell. An Econometric Analysis of Maryland National Capital Park and Planning Commission Public Transportation Planning Norms. Washington, DC, (MNCPPC) and the Montgomery County Planning U.S. Department of Transportation (1975). Department. Comprehensive Growth Policy Study: A Policy Vision: Centers and Trails. Silver Spring, Maryland, Perfater, M. A. and M. J. Demetsky. Pedestrian Attitudes MNCPPC (1989). and Behavior in Suburban Environments. Charlottesville, Virginia Highway and Transportation Research Council Maryland National Capital Park and Planning Commission. (December 1974). Transit and Pedestrian Oriented Neighborhoods. Silver Spring, MNCPPC (1982). Pill, J. "Emerging Suburban Activity Centers in Metropolitan Toronto." Journal of Advanced McNally, M. and S. Ryan. A Comparative Assessment of Transportation. Vol. 17, No. 3 (1983) pp. 301-15. Travel Characteristics of Neo-Traditional Developments. Irvine, University of California, Institute of Transportation Potter, S. "The Transport Versus Land Use Dilemma." Studies (1992). Transportation Research Record 964. Transportation Research Board, Washington, DC (1984) pp. 10-17. Meyer, J. and J. Gomez-Ibanez. Autos, Transit and Cities. Cambridge, Harvard University Press (1981). Pucher, J. "Urban Travel Behavior as the Outcome of Moore, T. and P. Thorsnes. The Transportation Land Use Public Policy." Journal of the American Planning Connection. Planning Advisory Service Report No. Association. Vol. 54, No. 3 (1988) pp. 509-20. 448/449. Chicago, American Planning Association (1994). Pushkarev, B. and J. Zupan. Public Transportation and MSM Regional Council, The Impact of Various Land Use Land Use Policy. Bloomington, Indiana University Press Strategies on Suburban Mobility: Final Report. Princeton, (1977). New Jersey (1991). Pushkarev, B. and J. Zupan. Urban Rail in America. Bloomington, Indiana University Press (1980). Neels, K. et al. An Empirical Investigation of the Effects of Land Use on Urban Travel. Washington, DC, The Urban Pushkarev, B. and J. Zupan. Urban Space for Pedestrians. Institute (1977). Cambridge, MIT Press (1975).

Newman, P. and J. Kenworthy. "Gasoline Consumption Recker, W. Empirical Analysis of Household Activity and Cities: A Comparison of U.S. Cities with a Global Patterns: Final Report. Irvine, University of California, Survey." Journal of the American Planning Association. Institute of Transportation Studies (1980). Vol. 55, No. 1 (1989) pp. 24-37. Rice Center for Urban Mobility Research. Houston's Major Activity Centers and Worker Travel Behavior. Houston, Nowlan, D. and G. Stewart. "Downtown Population Houston-Galveston Area Council (1987). Growth and Commuting Trips: Recent Experience in Toronto". Journal of the American Planning Association, Richardson, H. and P. Gordon. "Counting Nonwork Trips- Vol. 65 (Spring 1991) pp. 165-182. The Missing Link in Transportation, Land Use and Urban Policy." Urban Land. (September 1989) pp. 6-12. 52

Rutherford, S. et al. Land Use-Transportation Linkage: White Mountain Survey Co., Inc. City of Portsmouth Working Paper. Washington State Transportation Traffic/Trip Generation Study. Portsmouth, and Ossippee, Commission (1991). New Hampshire (1991) pp. 479-91.

Schneider, J. B. and R. J. Beck. Reducing the Travel Requirements of the American City: An Investigation of Interactive/Modeling Alternative Urban Spatial Structure. Research Report No. 73-1. Washington, DC, U.S. Department of Transportation Anas, A. The Chicago Area Transportation-Land Use (1973). System Final Report. U.S. Department of Transportation. Springfield, Virginia (1983). Smith, S. A. et al. "Planning and Implementing Pedestrian Facilities in Suburban and Developing Rural Areas." Anas, A. Residential Location Markets and Urban NCHRP Report 294 A and 294 B. Washington, DC, Transportation. New York, Academic Press (1982). Transportation Research Board, National Research Council (1987). Association of Bay Area Governments. Increasing Transit Ridership and the Efficiency of Land Use While Maximizing Economic Potential. Working Paper 90-2 Smith, W. "Mass Transit for High-Rise, High-Density (1990). Living." Journal of Transportation Engineering. 110, 6 (1984) pp. 521-35. Association of Bay Area Governments. Assessing the Future: A Sensitivity Analysis of Highway and Road Stone, J. and C. Johnson. "Neo-Traditional Neighborhoods: Improvements on Growth in the San Francisco Bay Area. A Solution to Traffic Congestion?" Site Impact Traffic Working Paper 91-4 (1991). Assessment Conference Proceedings. Paaswell, ed., New York, American Society of Civil Engineers (1992). Alonso, W. Location and Land Use: Toward a General Theory of Land Rent. Cambridge, Harvard University Press Stringham, M. "Travel Behavior Associated with Land (1964). Uses Adjacent to Rapid Transit Stations." ITE Journal. 52, 4 (1982) pp. 18-22. Berechman, J. "Interfacing the Urban Land-use Activity System and the Transportation System." Journal of Thomson, J. Great Cities and Their Traffic. London, Pion Regional Science. Vol. 16 (1976) pp. 183-194. Press (1977). Berechman, J. and K. A. Small. "Modeling Land Use and Untermann, R. et al. Linking Land Use and Transportation: Transportation: An Interpretive Review for Growth Areas." Design Strategies to Serve HOV's and Pedestrians. Seattle, Environment and Planning. 20 (1988) pp. 1285-1309. Washington State Transportation Commission (June 1991). Cervero, R. et al. Transit Joint Development in the United Untermann, R. "Adapting Neighborhoods for Walking and States: A Review and Evaluation of Recent Experiences and Bicycling." Accommodating the Pedestrian. Van Nostrand an Assessment of Future Potential. Washington, DC, U.S. Reinhold (1984). Department of Transportation (1992).

Crane, R. Access and Travel: On Trip Frequency and Mode Urban Engineers, Inc. Cross County Metro Feasibility Split in the New Suburbs. Irvine, University of California Assessment. Philadelphia, Southeastern Pennsylvania (June 1993). Transportation Authority (May 1994). Cropper, M. and P. Gordon. "Wasteful Commuting: A Re- Wabe, J. "Dispersal of Employment and the Journey to examination." Journal of Urban Economics. Vol. 29 (1991) Work." Journal of Transport Economics and Policy. Vol. pp. 2-13. 1, No. 3 (1967) pp. 345-61. De Jong, G. C. and M. A. Bradley. Accessibility and the Washington State Transportation Commission, Innovations Choice of Residential Location and Workplace: Traffic and Unit. Travel Patterns in Mixed Use Neighborhoods. Transport Policy Between Rural And Urban Area, to an Working Paper 93.6, Olympia, Washington (1993). Integral Approach on a Regional Level. The Hague, Transportation Planning Colloquium (1989). 53

Frank, L. A Guide to Literary Sources Pertaining to the Muth, R. Cities and Housing. Chicago, Chicago University Interactive Nature of Land Use, Urban Design, and Press (1969). Transportation. Seattle, University of Washington, College of Architecture and Urban Planning (December 1991). Potter, S. "The Transport Versus Land Use Dilemma." Transportation Research Record 964. Transportation Gordon, P., H. Richardson, and H. Wong. "The Research Board, Washington, DC (1985) pp. 10-17. Distribution of Population and Employment in a Polycentric City: The Case of Los Angeles." Environment Prastacos, P. "Urban Development Models for the San and Planning A. Vol. 18 (1986) pp. 161-173. Francisco Region: From PLUM to POLIS." Transportation Research Record 1046. Transportation Research Board, Gordon, P., A. Kumar, and H. Richardson. "Congestion and Washington, DC (1985) pp. 37-44. City Size." (Presented at the 10th Pacific Regional Science Conference, Pusan, Korea, 1987). Puget Sound Council of Governments. Vision 2020 Growth Strategy and Transportation Plan for the Central Puget Gordon, P., A. Kumar, and H. Richardson. "The Spatial Sound Region: Final Environmental Impact Statement. Jobs-Mismatch Hypothesis: Some New Evidence." Los Seattle, Washington (September 1990). Angeles, California, University of Southern California, School of Urban and Regional Planning (1987). Putnam, S. H. The Interrelationship of Transportation Development and Land Development, Vol. 1. Washington, DC, Federal Highway Administration (1976). Gruen and Gruen Associates. "Changing Influences of Residences on Employer Locations." San Francisco, Putnam, S. H. Integrated Urban Models: Policy Analysis of Mimeo (1991). Transportation and Land Use. London, Pion Press (1983).

Hamilton, B. "Wasteful Commuting." Journal of Political Putnam, S. H. "Urban Land Use and Transportation Economy. Vol. 90 (1982) pp. 1035-1063. Models: A State of the Art Summary." Transportation Research. Vol. 9 (1975) pp. 187-202 . Hanson, S. and G. Pratt. "Reconceptualizing the Links Between Home and Work in Urban Geography: Review, Ricardo, D. On the Principles of Political Economy and Critique, Agenda." (Unpublished Paper) Clark University Taxation. London, J.M. Dent and Sons (not dated). (1988). Schneider, J. Transit and the Polycentric City. U.S. Heilbrun, J. Urban Economics. Cambridge, Harvard Department of Transportation, Urban Mass Transportation University Press (1981). Administration, Seattle, University of Washington (1981)

Isard, W. Location and Space-Economy. Cambridge, MIT Schneider, J. et al. Defining Urban Form/Transportation Press (1956). Energy Relationships for Use in Planning Transportation Energy Conservation Strategies in Urban Area. U.S. Levinson, H. S. and H. E. Strate. "Land Use and Energy Department of Transportation, Seattle, University of Intensity." Transportation Research Record 812. Washington (1983). Transportation Research Board, Washington, DC, (1981). Scott, A. Metropolis: From Division of Labor to Urban Meyer, J., J. Kain, and M. Whol. The Urban Form. Berkeley, University of California Press (1989). Transportation Problem. Cambridge, Harvard University Press (1964). Shortreed, J. H., P. May, and E. Dust. "The Effect of Future Trends on Trip Patterns, Urban Commercial Structure, and Metro. Region 2040: Concepts for Growth. Portland, Land Use." Transportation Research Record 1046. Oregon (1994a). Transportation Research Board, Washington, DC (1985) pp. 50-58. Metro. Region 2040: Recommended Alternative Decision Kit. Portland, Oregon (1994b). Simpson, W. "Workplace Location, Residential Location, and Urban Commuting." Urban Studies. Vol. 24 (1987) pp. Mills, E. Studies in the Structure of the Urban Economy. 119-128. Baltimore, Johns Hopkins Press (1972). 54

Small, K. and S. Song. "Wasteful Commuting: A Chinitz, B. Metropolitan American in Transition: Resolution." Journal of Political Economy. (not dated). Implications for Land Use and Transportation Planning: Summary of Proceedings. Arlington, Virginia: Lincoln Solow, R. "On Equilibrium Models of Urban Location." Land Institute, Federal Highway Administration, and Essays in Modern Economics. Essex, Longman, Harlsow Federal Transit Administration (1993). (1973). Garreau, J. "From Suburbs, Cities are Springing Up in our Urban Land Institute. Market Profiles 1990. Washington, Backyards." Washington Post. (March 8, 1987) pp. Al, DC, Urban Land Institute (1990). A26-30.

U.S. Department of Transportation. Urban Transportation Garreau, J. Edge City: Life on the Frontier. New York, and Land Use. Washington, DC, U.S. Department of Viking (1991). Transportation (1976). Giuliano, G. and K. Small. "Subcenters in the Los Angeles Region." Regional Science and Urban Economics. Vol. 5 Voith, R., "Transportation, Sorting and House Values." (1991) pp. 305-12. AREUEA Journal. Vol. 19, No. 2 (1991) pp. 117-137. Gordon, P., H. Richardson, and H. Wong. "The Von Thunen, J. Der Isolierte Staat in Beziebund auf Distribution of Population and Employment in a Landwirtscahft und Nationalokonomie. Hamburg, Polycentric City: The Case of Los Angeles." Environment Germany, F. Pethes (not dated). and Planning A. Vol. 18 (1986) pp. 161-73.

Harris, C. and E. Ullmann. "The Nature of Cities." The Watterson, W. T. "Adapting and Applying Existing Urban Annals of the American Academy of Political and Social Models" DRAM and EMPAL in the Seattle Region." Science. Vol. 242 (1945) pp. 7-17. Journal of the Urban and Regional Information Systems Association. (Fall 1990). Hartshorn, T. and P. Muller. Suburban Business Centers: Employment Implications. Washington, DC, Economic Webster, F.V. et al. "Urban Land Use and Transportation Development Administration (1986). Interaction." Avebury, U.K., Interaction (1988).

Webster, F. V. and P. H. Bly. "Changing Pattern of Urban Hartshorn, T. Interpreting the City: An Urban Geography. Travel and Implications for Land Use and Transport New York, John Wiley and Sons (1992). Strategy." Transportation Research Record 125. Transportation Research Board, Washington, DC (1987) Heikkila, E., P. Gordon, J. Kim, R. Peiser, H. Richardson, pp. 21-28. and D. Dale-Johnson. "What Happened to the CBD- Distance Gradient? Land Values in a Polycentric City." White, M. "Urban Journeys Are Not 'Wasteful."' Journal of Environment and Planning A. Vol. 21 (1989) pp. 221-32. Political Economy. Vol. 96 (1988) pp. 1097-1110. Hoyt, H. The Structure and Growth of Residential Neighborhoods in American Cities. Washington, DC, USGP (1939). Other References Muller, P. "Transportation and Urban Form: Stages in the Adams, J. "Residential Structure of Midwestern Cities." Spatial Evolution of the American Metropolis." The Annals of the Association of American Geographers. Vol. Geography of Urban Transportation. New York, Guilford 60 (1970) pp. 37-62. Press (1986) pp. 24-48.

Atlanta Regional Commission. Transportation Problems Park, R., W. Burgess, and R. McKensie. The City. Chicago, and Strategies for Major Activity Centers in the Atlanta University of Chicago Press (1925). Region. Atlanta Regional Commission (1985). Pivo, G. "The Net of Beads: Suburban Office Development in Six Metropolitan Areas." Journal of the American Chapin, F., E. Kaiser, and T. Godschalk. Urban Land Use Planning Association. 56, 4 (1990) pp. 457-69. Planning. Urbana, University of Illinois Press (1994). 55

Schneider, J. Transit and the Polycentric City. Washington, Thomson, J. Great Cities and Their Traffic. DC, U.S. Department of Transportation, Federal Transit Harmondsworth, England, Penguin Books (1978). Administration (1981). Vance, J. The Scene of Man: The Role and Structure of the Schaeffer, K. and E. Sclar. Access for All: Transportation City in the Geography of Western Civilization. New York, and Urban Growth. Baltimore, Penguin Books (1975). Harper's College Press (1977) .