231 Transportation 6 (1977) 231--247 © Elsevier Scientific Publishing Company, Amsterdam - Printed in the Netherlands EVIDENCE OF LAND USE IMPACTS OF RAPID TRANSIT SYSTEMS ROBERT L. KNIGHT LISA L. TRYGG De Leuw, Cather & Company, San Francisco ABSTRACT This paper draws from the findings of published empirical studies and observations of the impacts of rapid transit systems on urban development. Analysis is based on comparisons of impact findings by different researchers and for different cities. An initial set of key issues is proposed, against which available information is arrayed and com- pared. It is concluded that rapid transit can have substantial growth-focusing impacts, but only if other supporting factors are present. Introduction In recent years studies have been conducted in a number of metro- politan areas in the United States to assess the feasibility of various public transit alternatives. More such studies can be expected in the future, partly because of continuing public concern over some of the externalized costs of automobile use such as energy waste and environmental degradation. Not unexpectedly, one result is that requests for capital grants to construct transit facilities already far exceed the available Federal funds. Particularly in such a competitive grant market, the heavy initial costs of proposed new transit systems must be offset by demonstration of massive public benefits. In their search for evidence of such benefits, many of the local agencies involved in these studies have attempted to include consideration of land use impacts. The usual premise is that a new rapid transit system will stimulate, revitalize, order, compact and/or create infrastructure economies in urban development in contrast to what would have occurred without such a system. This is plausible and even appealing in some ways. However, a review of these studies shows that there are many more intriguing questions than convincing answers regarding the impact of transit on land use (Frye, 1971; Marcou, O'Leary, 1971; Libicki, 1975). Some of these include the following: 232 - Will a transit system promote attraction of wealth or population from other regions? - Do different transit technologies (such as conventional rail, light rail, PRT, express bus) have different land use impacts? -What different physical and policy settings have significant bearing on these impacts? - Can a region's development be focused or its average density increased by transit? - Will declining CBD's be strengthened, or will suburbanization be speeded by a new transit system? - How quickly will any such impacts appear? Unfortunately, most of these questions cannot be answered satisfac- torily at present. In most cases the needed empirical research has not been done, often because methodologies or data for identifying such impacts are not available. Hopefully some current research, such as that of the Federal Department of Transportation's studies of the impacts of the new Bay Area Rapid Transit (BART) system in San Francisco and of empirical evidence of transit's land use impacts nationwide, will advance our understanding of these issues. However, in the interim there is a substantial need for a concise summary of available information to guide efforts in the transit feasibility studies now or soon to be in progress. This report seeks to provide such a summary. Approach The findings reported here are based on an extensive review of available literature and some firsthand observation. The paper's purpose is to assemble and present findings in order to identify Common threads, as an early step in the development of sound theory and understanding of the ways in which transit may influence land use. From the outset our investigation was guided by an implicit, emerging theory: The conventional model is one of land use succession driven primari- ly by changes in access, in turn wrought by transportation improvements and only marginally influenced by other factors (usually modeled as \"attrac- tiveness\" and \"developable land area\"). However, from experience in urban transit alternatives analyses we knew that modern :urban transit systems rarely, if ever, provide a major effective increase in accessibility, because the areas served tend to be already more accessible by auto. Therefore, the conventional model seemed much less appropriate today than it was in earlier generations when transit and later highway innovations generated almost revolutionary increases in accessibility, overwhelming the influence of most other factors. 233 We felt, then, that the achievement of major land use \"impacts\" around transit stations today must require the concerted action of other powerful forces in addition to transit-induced accessibility increases. This led to a more general view of the land development process as one driven by a complex of forces each adding to or detracting from a site's development potential and timing. Some of these might be essential, but others were expected to be substitutable. In different situations, different factors might be dominant, and several might sometimes be much more critical than even the access afforded by transit. If true, this implies a need for a substantial broadening of the range of concerns which must be encompassed in the planning of transit systems. The relevance of some factors was obvious enough: accessibility as well as overall level of demand, land assembly risks, zoning incentives and constraints, size and other physical characteristics of the land available were clearly among those involved. But the importance of each in different situations was unclear, as was the degree to which they could complement or substitute for one another. This study, then, sought to assess the importance of some of these factors in development near actual recent transit improve- ments. It also sought to provide some indication of the resulting scale and timing of development. Thus the findings presented here are oriented toward the key questions outlined earlier. Although they reflect a careful study of the references found, errors of omission as well as interpretation may well occur. Hopefully others will undertake to extend the framework provided here by adding information not available to the present authors and also by sharpening these initial interpretations. Findings Will a transit system promote interregional attraction of investment or population ? Here the oft-stated thesis is that a modern rapid transit system may increase the attractiveness of one region over others to such an extent that a healthy rate of local private investment and a steady supply of qualified labor can be more easily maintained. In certain economically declining cities in the northeastern United States, for example, this may be a major hope. The only formal study found to address this issue was the BART Impact Program's analysis of that system's impacts on the regional economy (MacDonald and Grefe, 1976).* No such interregional-transfer impacts were *BART is the Bay Area Rapid Transit system, in operation in the San Francisco region since 1972 and the subject of a comprehensive impact evaluation sponsored by the Federal Department of Transportation, Department of Housing and Urban Development and the Bay Area's Metropolitan Transportation Commission. 234 found by that study, despite its detailed reviews of the region's economic activity trends by sector and even including tile effects of BART construc- tion activity and the sale of BART bonds. Some published material has asserted major increases in a city's assessed property valuation due to introduction of a transit system. In Toronto, for example, one observer has argued that of a $15 billion total increase over the ten year period following construction of the rail transit system's first segment, $10 billion was directly attributable to tile transit system and hence in effect paid for the system many times over (Heenan, 1968). However, this conclusion was based only on a simple comparison of the rates of tax assessment increases downtown near the transit system with those in the rest of the metropolitan area. No consideration was given to the effects of any other factors which encouraged downtown office development .in Toronto, nor were any comparisons made with other Canadian cities. Many policymakers have been misled by this widely-publicized overstatement. Kovach (1974) did compare Toronto's growth with that of other Canadian cities. Her study revealed that while Toronto had a very large population growth during the 1960s, other large Canadian cities with no plans for transit systems exhibited similar or even larger growth rates. Do different transit technologies have different land use impacts? Since a common objective in transit feasibility studies is the identifica- tion of some optimal system, it would be useful to be able to differentiate among competing technologies such as rapid rail, light rail and express bus by their land use impacts. However, in this review no studies of land use impacts of urban transit technologies other than conventional rapid rail have yet been found. This is largely because there has been almost no use of technologies other than conventional rail and express bus in modern North American systems. Even express bus systems are typically limited in scale, and no case studies of the impacts of such bus transit systems have come to our attention. Clearly more knowledge is needed, especially in view of the federal DOT trend toward emphasis on less capital-intensive transit solutions such as express buses. However, given the lack of effective research on this topic, the remainder of this paper must be limited to impacts of conventional rail rapid transit systems. Users of this information must be careful in attempting to apply it to any comparison of transit technologies, since there is no documented evidence to guide such applications. Surely other modes, pa r - ticularly modern light rail, have similar if lesser potential effects; however, in the absence of empirical evidence, assessments of such effects must remain intuitive. 235 What different physical and policy settings have significant bearing on these impacts ? Given a conventional rail technology, what seem to be its land use impacts under various circumstances? Stated another way, where will a rapid transit system be most likely to generate such impacts? A scattered and uneven but substantial body of literature on this topic was found. Our observations and review of this literature suggested that several character- istics of the area served are significant in determining land use impact. A convenient grouping of the available evidence into factors influencing impact is the following: Local government land use policies Regional development trends and forces Availability of developable land - Physical characteristics of the area LOCAL GOVERNMENT POLICIES The issue here is whether local governmental bodies, through their land use policies, can or do exert a significant influence on the generation of land use impacts by transit. In Toronto, most observers agree that the rapid transit system appears to have significantly shaped and intensified develop- ment, particularly along the Yonge Street corridor (Heenan, 1968; Marcou, O'Leary, 1971; Kovach, 1974). However, the transit system alone did not cause these impacts. Introduction of the rapid transit system was soon followed by several pro-development public policies. These include: (1) aggressive marketing of air rights and available excess land parcels by the transit commission; (2) allowance of liberal floor area ratios (in some cases increased from 3:1 to 12: 1) and density bonuses by the city in certain locations, especially around stations; (3) encouragement of coordinated station design efforts with developers desiring direct access from office, retail or apartment buildings; and (4) city zoning classification changes in certain districts, notably near metropolitan area transit stations, to permit much higher intensity development (City of Toronto, 1963, 1971). Toronto plan- ners and developers interviewed in our check of these results agreed that these incentives were essential in inducing the extent and type of develop- ment which occurred. Two other more general policy factors were also important in Toronto. One is the Canadian government's policy on income tax benefits of home ownership. In contrast to the United States, the Canadian homebuyer receives no special income deduction for mortgage interest, so this important encouragement to suburban living is absent. As a result, the postwar American exodus from urban apartments to suburban houses was never paralleled in intensity in the major Canadian cities, including Toronto. The 236 second factor is in the Toronto governmental structure, which in 19 was changed to a two-tiered regional form with a metropolitan authority respon- sible for regional functions and five municipal subunits (the City of Toronto plus four boroughs) to deal with primarily local concerns. Although the member municipalities retain zoning powers, Metro still exerts substantial influence on land development through its authority over regional transpor- tation and the coordination of land uses among the boroughs. Such an empowered regional perspective is approached only in a few American cities, notably Nashville and Miami. In addition, other observers (National League of Cities et al., 1973) note that the Toronto city government was actively involved in developing basic overall plans for the Eglinton Station, which later turned out to be the focus of one of the most heavily developed areas along the line. As an added boost, the city government constructed the first building in the complex near the Eglinton Station. Essentially, then, the local government, while encouraging intensive development, at the same time successfully attempted to direct and control it where it was most favorable. The San Francisco BART experience is less dramatic but still somewhat similar to that of Toronto. For instance, density bonuses have been per- mitted for buildings providing direct access to downtown San Francisco BART stations. Smaller bonuses have been allowed for nearby development, based on proximity to stations up to a distance of 750 feet (Libicki, 1975). Development has been very heavy in this downtown area, with BART consistently acknowledged as one of the key forces involved (Lee and Wiech, 1972). However, as in Toronto, many other forces also acted to encourage and concentrate Bay Area office development in this location. Moreover, there is little evidence of any substantial land use impacts outside the CBD station areas (Gruen, 1977; Lee, 1973; Wells, 1973). The observations of Boyce and his associates (Boyce et al., 1972) on the character of development along the Lindenwold rapid transit line in Philadelphia (opened in 1969) also illustrate the potential impact of specific local government policies. Two towns along the suburban line, Lindenwold and Voorhees, appear to have experienced different development patterns between 1960 and 1970 due to such policies. Boyce noted that while Lindenwold promoted multi-family development, Voorhees discouraged such development by not approving apartment projects and instead sup- porting single-family construction. This was further evidenced by the fact that developer willingness and available land were present in Voorhees but did not result in significant amounts of multi-family construction. Specifical- ly, during the 1960's, Lindenwold doubled its housing stock, with apart- ments accounting for over 80 percent of the new construction. Voorhees, in contrast, saw a smaller increase (60 percent) in new housing with over 80 percent of the new construction in the form of single-family homes. Several 237 other suburban Philadelphia communities not in the Lindenwold corridor matched or exceeded this level of growth. This trend has only recently been changed when Voorhees Township approved construction of a regional shopping center which includes provi- sions for 3,700 apartments and town houses. Boyce reported that the project was approved and located there mainly because it was the only suitable place with a large amount of available land. This, only the second large-scale development project in Voorhees since 1960, was then scheduled for com- pletion in 1975, six years after the opening of the Lindenwold Line. In Washington, D.C. a study by the National League of Cities (1973) reported that station area development along the METRO transit system in that city has been influenced by government involvement. One example involves the Gallery station at 12th and \"G\" Streets in downtown Washing- ton, D.C., which presently is not a highly desirable area in terms of character and development. Public capital was to be used to construct the city's nearby Convention Center, and consequently greatly enhance this presently declining area. Although this project has yet to materialize, due to Congressional delays in the approval process, the Washington approach bears further study. The city's Redevelopment Land Agency (RLA) concentrated their land purchases around a few major METRO stations, including the Gallery station, for later resale to developers. This policy sets the stage for strong guidance of development, although it is still early to judge the outcome. In contrast, a study by Marcou, O'Leary and Associates (1971) con- cludes that the transit lines and extensions in Cleveland and Chicago provide very little evidence of transit-related development, partly due to the lack of demand in general but also because of the absence of pro-development public policies and incentives. The potential role of such policies and incentives was recognized later in Cleveland when air rights were leased above the station parking lot at the eastern terminal point of the transit system. However, the developer was never able to. secure financing and nothing has been built. Similarly, air rights granted at sites along the seven year old airport extension have also spurred developer interest but no development has yet occurred, due largely to zoning delays. The above evidence suggests that local government policies are impor- tant factors affecting development, with transit being an important but not sufficient condition for such development. For instance, policies such as allowance of liberal floor area ratios, density bonuses at designated locations, changes in zoning plans, marketing of air rights, sale of excess land parcels, and urban renewal - all implemented at strategic locations near a transit station or along the corridor - may have a very significant impact on development. Not only can intensive development be encouraged (or dis- couraged as has been the case with Chicago and Cleveland) through the use 238 or lack of effective public policies, but there is strong evidence from these studies that it call be controlled to a significant extent both in terms of where it will occur and what type will occur. Thus tile transit system can become tile focus or catalyst around which tile suggested policies are implemented, which in turn affect development ill the area. DEVELOPMENT TRENDS AND FORCES Among North American examples, Toronto seems to have experienced the most spectacular rate of growth and development in parallel with construction of a major transit system during tile post-World War II era. As noted in an earlier section, it seems unlikely that this regional growth is due to the transit system. However, the presence of suchgrowth appears to have been a major force behind the development which subsequently occurred around the transit system. Part of the high rate of economic growth which took place in Toronto during the transit system's early years may be attributable to a coincidental easing of credit and new availability of development capital. This resulted in freeing a large amount of pent-up demand for offices and apartments in the late 1950s. With this demand, capital, and the zoning centives already noted, station areas on the Yonge Street Line became competitively attractive for investment with more expensive downtown sites. This explanation seems to be supported by the lack of an equal amount of construction associated with opening tile Bloor-Danforth Line, the second phase of the subway system; once the supply and demand for offices and apartments achieved equilibrium and many more station areas were put on the market, the monopolistic position experienced by earlier station areas was destroyed and a much smaller scale of focused development took place. Downtown San Francisco was experiencing an increasing amount of construction activity even before bonds were approved for financial backing of the BART transit system. Also, like Toronto, San Francisco has been considered as a regional financial and business capital. It has long been thought of as an attractive place to live and thus has experienced significant growth right up until as well as after BART. Thus, attempts to isolate BART's impacts from those due to the construction boom period have been difficult, Given the city's favorable attributes, much of the rapid rate of investment and construction in the downtown area, especially during the system's earlier planning and construction period, may be due to conditions completely independent from the presence of BART. For example, Lee and Wiech (1972) have estimated that much if not most of this new downtown development may be due to conditions completely independent from the presence of BART. Downtown Philadelphia, in association with the Lindenwold Line has 239 not experienced the same development as downtown Toronto and San Francisco. It did not see a large amount of new office and business develop- ment after construction of the Lindenwold transit system. Looking at economic conditions of the city at this time, 1969 to 1971, Gannon and Dear (1972) noted that the general trend was not favorable (unlike in Toronto and San Francisco) during the construction period of the transit system. Their data show that the City of Philadelphia experienced a decline in population between 1960 and 1970, and only a relatively small expansion in total employment. It appears, then, that the single Lindenwold Line was too small to have a substantial effect on the Philadelphia CBD. Camden, just across the river from downtown Philadelphia and the first suburban stop on the Lindenwold Line, is an example of a declining older subcenter apparently not aided by the transit system. Gannon and Dear report that, if anything, the transit system initially triggered a spatial displacement of activity out of Camden rather than the generation of new activity there. Although evidence for such a conclusion was thin (a small survey of some 20 firms which relocated to Haddonfield), it is nonetheless clear from observation that Camden is not experiencing any significant redevelopment due to the transit line. Thus the Line may have been a factor in speeding the decline of Camden's aging CBD, and certainly contributed little or nothing to its revival. In Toronto the trend toward middle-class suburbanization and resultant decline of central areas has never been so pronounced as in major cities of the United States. A strong middle-class element has tended to remain in the city's older, inner areas. This may have been reinforced by the provision of a modern downtown-oriented short-line transit system before expressways were built; the very heavy postwar immigration of Europeans accustomed to urban apartment life certainly aided as well. Finally, there is no significant racial problem in Toronto; there has been no discriminatory treatment of any racial group. This may contribute to the fact that crime is generally low and both urban and suburban areas are considered safe places to live. All of this evidence shows that there are economic and social factors which can affect growth and development and are at least in part indepen- dent of the presence or absence of a transit system. Such factors include population growth (and a subsequent demand for new development) and the general economic and social condition of the specific areas served by the transit system's stations. This implies that transit may help to direct growth to one area rather than another (e.g., to attractive areas near transit stations rather than somewhere else) even though it alone probably does not generate or create new regional growth. When the general character of the area is favorable toward develop- ment, as in downtown San Francisco or Toronto, transit may further enhance such development. That is, transit-related growth seems more likely 240 to occur in areas already ripe for development near a newly constructed transit station where some activity would probably have taken place even if the transit system had not been built. This suggests that the transit system can enhance and speed the already existing development process. Conversely, this conclusion also suggests that an area possessing deteriorating attributes, such as the City of Camden, may have that trend further accelerated rather than reversed by the introduction of transit. As a result, redevelopment may occur sooner in such areas than would otherwise be the case, but the transition from decline to renewal may be socially disruptive. AVAILABILITY OF DEVELOPABLE LAND Market forces - primarily the availability of land for development - may significantly affect the location and degree of development above and beyond other influences such as those already discussed. Land may become available through several means, some of which do not necessarily depend on the introduction of a transit system. These include new access to existing vacant or unused space, redevelopment of occupied space to more intensive uses, and assemblage of privately-owned parcels to form a space large enough to accommodate significant development. In Toronto a significant amount of development has been attributed to introduction of the transit system. However, it has been pointed out that such development, particularly the apartment boom, was delayed in starting because of the scarcity of available land to build on. Over a period of time enough demand pressures were generated (perhaps by the transit system) to cause an increase in land value to a point where a change in land use intensity (acquisition and clearance of existing properties) proved to be \"economically justified. The point is that significant development was restrained until usable land was somehow made available (Wacher, 1970; Anderson et al., 1971). Some studies have been made of impacts near specific BART stations. Lee reported, for example, that in downtown Hayward the anticipation and arrival of BART in the area had not (in 1972-3) appeared to have a significant impact on commercial property values within a half-mile of the BART station. Other market forces have acted to direct new commercial facilities to the newer outlying areas where inexpensive vacant land is much more available. Wells (1973) argued that in the City of Fremont, which is BART's south-eastern terminus some 25 miles from the San Francisco CBD, avail- ability of land seemed to have a potentially significant role in development associated with BART. It is one of the few areas along the system with large amounts of vacant land available for immediate use. However, development near this BART station has been occurring at a slow pace relative to that of Fremont as a whole. This suggests that mere accessibility and land avail- ability are not enough. 241 Boyce (1972) points out that the Township of Voorhees, which is oil tile Lindenwold Line, had no central business and residential area of con- sequence until the Echelon Mall was built one year after the opening of the line. The Mall, a large regional shopping center combined with a major apartment and townhouse developlnent, has become a highly successful community center. The developer, in selecting a site for the Echelon Mall, needed to find a large parcel of available land (in this case an area which was previously a general aviation airport) and a location not near any existing regional center. The Voorhees site was virtually unique in meeting these basic requirements, neither of which had anything to do with transit. Thus, while the Lindenwold Line could be viewed as a positive factor in bringing customers to the area, it could not be considered as one of the most important determinants in the creation of the Mall. In Toronto, at the intersection of Bloor and Yonge Streets where the two transit lines cross, no new development has yet occurred on two of the four abutting blocks, some 23 years after the first line segment opened. Private realty spokesmen attributed this to the difficulty and risk of land assembly, since many small parcel owners were involved. The two corners which have been developed into high-rise office buildings and department stores each involved assembly of five or fewer parcels; the other two (undeveloped) corners involved nearly 25 parcels each. Under such con- ditions development is likely to be delayed indefinitely. It can be concluded that availability of suitable, assemblable land is a necessary prerequisite for development on any significant scale with or without the presence of a transit system. Thus, this market force can significantly affect the location and degree of development above and beyond the influence of the transit system and its increase in accessibility. Placement of a transit line through areas in which developable land is not readily available, then, is probably unlikely to lead to substantial land use intensification without at least a long delay. PHYSICAL CHARACTERISTICS OF THE AREA The newer extensions of the Chicago rapid transit system are located in rights-of-way of the Dan Ryan, Eisenhower, Congress and Kennedy express- ways. Thus, access to the surrounding developed area is difficult. In addition, these areas are themselves completely built up, and many are not attractive locations for intensive new development. Such area characteristics have apparently made it difficult to encourage any substantial new growth (Department of Public Works, 1967; Marcou, O'Leary, 1971 ). The first portion of the Cleveland transit system, which was opened in 1955, was built on an existing rail right-of-way. In one study it was pointed out that the line runs primarily through established or declining industrial 242 areas (Marcou, O'Leary, 1971). Our own observations support this view. In addition, the line is removed from the direction of new downtown develop- ment. Such major constraints imposed by the surrounding area may have essentially neutralized any impetus for significant development generated from the transit system. Significant development interest did not come until the opening of the airport extension in 1968, which provided for an increased scope of service throughout the metropolitan area. This extension also introduced several new stations with transit system-owned parking lots potentially compatible with air rights development. However, only limited interest has been shown in the air rights and virtually no nearby development has occurred. The experience of the Lindenwold Line in Philadelphia seems to support one of the conclusions reached by Spengler (1930) in his early study of New York's transit system: \"It appears that a subway reflects the conditions of the section through which it passes, in any influences which it might exert upon land values.\" Along the Lindenwold Line, office develop- ment at the Haddonfield station has been associated with introduction of the transit system. But while the transit line may have served as an impetus for office development here, the Haddonfield Borough itself has been charac- terized by Gannon and Dear (1972) as a highly desirable location, irrespec- tive of the presence of the transit system: The most important qualities of Haddonfield are the intrinsic attraction of the ...area, the extent, quality and proximate juxtaposition of other commercial activities, especially retailing, and the desire on the part of many office users to escape the undesirable aging environment of the City of Camden, yet still maintain good accessibility to the administrative and business centers of Philadelphia and Camden. It may be concluded from these observations that physical character- istics of the area to be served are important factors in determining transit's potential land use impacts, The evidence found suggests that where transit lines are built alongside an existing rail line or in rights-of-ways of express- ways or freeways, where access to the surrounding area is difficult, develop- ment seems to be discouraged. Similarly, development associated with a transit system may be discouraged or deterred when a line runs through predominantly industrial areas or in a direction away from new or increased development. On the other hand, areas characterized by desirable attributes such as a good location in terms of a large population served, the presence of compatible land uses or intense development, and/or close proximity or direction toward new development might well find further development noticeably encouraged by rapid transit construction. 243 Catla region's develot)meltt be .focused or its deltsity htcreased by tralzsit? Can transit strengthelz the CBD, or will it speed suburbalzization? It is often alleged or hoped that construction of a rapid transit system will promote orderly growth focused on its stations, in contrast to a previous pattern of urban sprawl or low density suburbanization. Such an effect is at least plausible both at a corridor level and on a smaller scale very near stations. The current interest in \"activity centers\" gives added impetus to this hope. References cited earlier in this study support the view that some focusing of a region's development around rapid transit stations can be achieved, although favorable conditions of public policy, regional develop- ment trends, land availability and physical constraints seem to be essential. Most of such localized development impacts to date appear to have occurred around downtown stations, especially in Toronto and San Francisco. In Cleveland, conversely, CBD impacts appear to have been minimal due at least in part to lack of the favorable conditions just noted. We found little documentation of the geographical extent of transit- related impacts. Gannon and Dear (1975) suggest from their Lindenwold study that most new transit-related office development occurs very close to a station. They noted that nearly all the new development in Haddonfield had taken place within a 5-minute walk (about 1,000 feet) of the station. The Marcou, O'Leary study (1971), which included some site visits andinter- views as well as literature review, yielded an estimate of 1,200 to 1,800 feet from a station as the radius of the likely impact area. They cited planners' opinions and supporting incentive ordinances ranging from 500 feet (Washington) to 2,500 feet (London) in support of this contention. Some of this reasoning here may be somewhat circular, since such incentives would tend to create the impact area rather than respond to it. Evidence on the occurrence of broader corridor-level development impacts includes several statistical studies. The continuing Lindenwold residential property value investigation, begun by Boyce and continued by a succession of his students, is the main source of information on this topic. Boyce's study yielded the conclusion that the transit line had tended to enhance property values, though only slightly, in its market area in contrast with values elsewhere in suburban Philadelphia. This finding was later sup- ported in related but separate studies by Platt (1972), Mudge (1974), Slater (1974), Yang (1975) and Tang (1976). These later studies have tended to indicate a larger but still not massive impact on values. Minimal evidence was found either to support or reject the hypothesis of reduced suburbanization (i.e., increased overall density) or its converse (more sprawl in transit corridors) due to transit. Gannon and Dear (1975) suggested that in the case of office development, transit may simultaneously strengthen the CBD and promote location of offices along the lines, ap- parently at the expense of other suburban locations. However, with systems which extend far into previously low-accessibility suburbs such as Philadel- 244 phia's Lindenwold and San Francisco's Concord, the contention of increased residential as well as commercial suburbanization due to transit service merits further study. The BART Impact Program may provide some initial insights on this issue, although its early findings indicate that no such shaping of suburban development has yet occurred attributable to BART. How Quickly Will Such hnpacts Occur? The timing of transit-related development to be expected is a key factor in assessing likely impact and establishing mechanisms to limit hardship and inequities. However, here as with many other important issues empirical data are limited. Gannon and Dear (1972) concluded that definitive evidence of land value changes does not appear for several years following construction. Boyce et al. found little evidence that anticipation of the opening of the Lindenwold Line had an impact on residential sales prices before the line started to operate, but some of his students found at least some indication of small pre-operations impacts (Tang, 1975). Anderson (1971) concluded that the Toronto construction boom began about five years following the transit system's start-up; this coincided with a loosening of credit, which released a pent-up demand for office and apartment space. With BART, on the other hand, the intensive downtown development activity in San Francisco started even before the transit system construction was begun, although it is likely that much if not most of this early development would have occurred even without BART. However, apparent development impacts elsewhere along BART are generally minor even now, some five years after start-up (Gruen, 1977). One high-rise office building was built at the downtown Berkeley station, and another at Walnut Creek. However, both apparently had serious difficulties finding tenants, despite a continuing demand for office space in downtown San Francisco. In Oakland a major downtown redevelopment project is in progress, but its first build- ings have also been slow in reaching full occupancy. These findings suggest that development may have been premature in these areas outside downtown San Francisco. Conclusions This review of available literature yields conclusions both with respect to future research directions and for application in current transit feasibility studies. - Interregional effects.\" There is no evidence that a new transit system will increase the overall level of development of a metropolitan area. - Transit technology: Only conventional rail rapid transit appears in the literature. Study of differential land use impacts of competing tech- 245 nologies, especially light rail and advanced express bus systems, are dif- ficult but need to be done in order to allow balanced evaluation of transit alternatives. - hzfluence of surroundiugs. Most of the available studies of transit impact dealt with this issue. Four general factors were identified as important determinants of the degree of land use impact of a new transit system: 1) Local government policies encouraging development 2) Regional development trends 3) Availability of developable land 4) Physical constraints of the site For substantial land use impact to occur, it appears that nearly all of these factors must be favorable. For evaluation of impact potential for specific locations, then, these factors might be useful as the basis for composite rating of each site or corridor. -Focusing of del, elol)ment. Some concentration of development around transit stations can be achieved, given the conditions stated just above. This is most apparent in CBD locations, with little effect generally detected to date farther from downtown areas. Further study is needed, particularly on the growth-shifting effects from the central city and suburban areas not served by transit to suburban areas near the transit system. - Speed of impact: Most researchers have concluded that substantial land use impacts do not occur until several years after inauguration of transit service, although there is evidence of much earlier effect in the case of the Lindenwold Line. There is evidence that some commercial developers moved too quickly to build near suburban transit stations, with expected demand not materializing. However, timing of transit-related development may well be much more determined by the region's overall economic health and activity rather than by completion of the transit system itself. These conclusions are only suggestive; much remains to be understood. It seems from the evidence available that rapid transit improvements can provide an impetus toward generation of new nearby development. However, transit alone seems no longer enough to insure such development, in this day of very high accessibility often only marginally improved by the transit system. If development is to be created, other factors such as those cited here must be effectively brought into play at the start of the transit planning process. This calls for a more coordinated land use/transit planning process than has often been evident in the past. Unreasonable claims of transit's power to induce major land use change must be avoided. On the basis of the evidence reported here, claims of massive net increases in property value tax revenue because of a transit improvement seem hard to justify. The main effect, at best, seems to be a focusing of development rather than a net regional increase. It also seems 246 apparent that the major potential of transit in encouraging development may be that of a catalyst in the local, governmental process, providing the rationale needed to gain support for land development controls and incen- tives to focus growth. But these are still substantial benefits and an impor- tant role. Rapid transit will not automatically revitalize and reshape our cities, but it can do much - if we can learn to understand that role and the others which must accompany it. References Anderson, Warren et al (1971), \"The Impact of Rapid Transit on the Metro-Center,\" Baltimore: Department of Planning (November). Boyce, David et al (1972), \"Impact of Rapid Transit on Suburban Residential Property Values and Land Development,\" for the U.S. Department of Transportation, Philadel- phia: University of Pennsylvania. City of Toronto Planning Board (1963), \"Air Rights and Land Use Adjacent to the Bloor-Danforth Subway,\" (June). City of Toronto Planning Board (1971), \"Report on a Proposal to Rezone Lands Along the East/West Subway Route to Permit High Rise Apartments,\" (October 12). Department of Public Works (1967), \"Public/Private Development of Transit Stations,\" City of Chicago (December 14). Frye, Frederick F. (1973), Alternative Multi-Modal Passenger Transportation Systems, NCHRP Report 146, Washington, D.C. : Highway Research Board. Gannon, Colin A. and M. Dear (1972), \"The Impact of Rail Rapid Transit Systems on Commercial Office Development: The Case of the Philadelphia-Lindenwold Speed- line,\" for the U.S. Department of Transportation, Philadelphia: University of Penn- sylvania. Gannon, Colin A. and M. Dear (1975), \"Rapid Transit and Office Development,\" Traffic Quarterly 29:2 (April), pp 223-242. Gruen Associates, Inc. (1977), \"Indirect Environmental Impacts,\" TM 24-4-77, BART Impact Program, Berkeley: Metropolitan Transportation Commission. Heenan, G. Warren (1968), \"The Economic Effect of Rapid Transit on Real Estate Development in Toronto,\" The Appraisal Journal (April), pp. 213-224. Kovach, Carol (1974), \"On Conducting an 'Impact' Study of a Rapid Transit Facility - The Case of Toronto,\" presented to Joint Transportation Engineering Meeting, Montreal, July 15-19 ; Meeting Preprint MTL-23. Lee, Douglass B., Jr. (1973), \"Impacts of BART on Prices of Single-Family Residences and Commercial Property,\" BART Impact Studies, Berkeley: Institute for Urban and Regional Development, University of California. Lee, Douglass B. and D.F. Wiech (1972), \"Market Street Study,\" BART Impact Studies, Berkeley: Institute of Urban and Regional Development, University of California. Libicki, Martin C. (1975), \"Land Use Impacts of Major Transit Improvements,\" Urban Analysis Program, Office of the Secretary of Transportation, Washington, D.C. (March). MacDonald, Angus N. and Grefe, Richard D, (1977), \"Identification of Economic Sectors Influenced by Transportation Service,\" paper presented to Transportation Research Board Annual Meeting, Washington (January). 247 Marcou, O'Leary & Associates (197I), \"Transit Impact on Land Development,\" Washing- ton, D.C. (minleographed). Mudge, Richard R. (1974), The Impact oj Transportation Savings on Suburban Reside12- tial ProperO, Values, P-5259, New York City-Rand Institute discussion preprint. National League of Cities, et al. (1973), \"Transit Station Joint Development,\" prepared for USDOT and HUD, Washington, D.C.:U.S. Government Printing Office. Platt, Jeffrey (1972), \"Residential Property Value Gradients and Urban Transportation hnpacts,\" Ph.D. dissertation, University of Pennsylvania. Slater, Paul B. (1974), \"Disaggregated Spatial - Temporal Analyses of Residential Sales Prices,\" Journal of the A merican Statistical Association 69:346 (June), pp. 358 363. Spengler, E. H. (1930), \"Land Values in New York in Relation to Transit Facilities,\" New York: Columbia University Press. Tang, Foh-tsrang (1975), \"Detection and Estimation of Transportation Impact with Models of Suburban Residential Property Sales Prices,\" Ph,D. dissertation, University of Pennsylvania, Trygg, Lisa L. and Knight, Robert L. (1975), \"Transit System Impacts on Urban Land Use,\" Denver: Regional Transportation District. Urban Systems Research & Engineering, Inc. (1976), The Growth Shapers: The Land Use hnpaets of Infkastrueture Investments, prepared for the Council on Environmental Quality, Washington, D.C.: U.S. Government Printing Office. Voorhees, Alan M. & Associates, Inc. (1974), Baltimore Regional Environmental Impact Study, for the Interstate Division for Baltimore City. Wacher, T.R. (I 970),\"The Effects of Rapid Transit Systems on Urban Property Develop- ment,\" Chartered Surveyor (March). Wells, William R. (1973), \"Rapid Transit Impact on Suburban Planning and Development, Perspective and Case Study,\" Department of Industrial Engineering, Stanford, Califor- nia: Stanford University. Yang, Chin Ming (1976), \"Impact of a Rapid Transit Line on Suburban Vacant Land Values,\" Ph.D. dissertation, University of Pennsylvania.
