Shaheen, PhD Honda Distinguished Scholar in Transportation, Institute of Transportation Studies-Davis, University of California, Davis and Post-Doctoral Researcher Partners for Advanced Transit and Highways (PATH), University of California, Berkeley 1357 S. 46th Street, Building 452; Richmond, CA 94804-4603 510-231-9460 (O); 510-231-5600 (F); E-mail: sashaheen@nt.path.berkeley.edu


ABSTRACT In the U.S., the automobile accounts for over 95 percent of all person miles traveled (1), while transit accounts for less than three percent of all trips (2). Between the private auto and traditional transit, niche markets exist for other transportation services, such as airport and transit feeder shuttles and carsharing. Commuter-based carsharing, in which individuals share a fleet of vehicles linked to transit, could potentially fill and expand one such niche; complement existing services, particularly transit and feeder shuttles; and develop into a viable transportation alternative. This paper includes a brief overview of transit feeder services, particularly in Northern California. Next, it describes the CarLink field test (a commuter-based carsharing model developed and tested in the San Francisco Bay Area in 1999) and highlights key behavioral findings. The author concludes with recommendations for strengthening the viability of this innovative service, based largely on experience from the CarLink field test. Empirical evidence suggests that CarLink could be sustainable if there is strong cooperation with local businesses and transit; parking limitations and congestion are prevalent; system feasibility is demonstrated; incentives are employed (e.g., preferred parking and transit pass subsidies); higher user fees are generated; and technology and management costs are lowered. Key Words: Carsharing, transit feeder services, and market niche potential. INTRODUCTION At the end of World War II, transit reached its peak in the U.S. with almost 24 billion passenger trips. Within ten years, ridership levels dropped by almost 50 percentage points. By the late 1950s and early 1960s, the U.S. mass transit industry—primarily owned and managed by private operators—was on the brink of collapse (3). Meanwhile, automobile trips were growing, and in 1956, the motor vehicle’s role was further solidified through passage of Interstate Highway System legislation (4). Coupled with an integrated road network, the auto offered unparalleled freedom and mobility, and its popularity grew (5). In contrast, the transit industry was in decline. In response, Congress enacted the Urban Mass Transportation Act of 1964. It provided capital grants for transit operations, as well as research support for demonstration projects to support and encourage local experimentation with innovative technologies and services (6). However, transit operations and productivity were not revitalized, as planned. Indeed, transit reported a $300 million annual operating deficit just six

Shared-use cars can offer instant and convenient access to destinations that are not conveniently accessible by transit alone (11). which was based out of the BART Dublin-Pleasanton station in 1999.g. Day Use). vehicles are shared by employees at an employment center for multiple uses throughout the day (i. today’s shift toward local transit policy (7).S. which transported suburban residents to employment centers. present. Due to rising costs. both carsharing and shuttles are gaining popularity (3.S. or both. such as transit feeder shuttles and commuter-based carsharing. could also provide connectivity to traditional transit. (See Shaheen et al. vehicles are also shared by participants (i. several . Carsharing vehicles provide a shared community resource at transit stations. and supported by subsidies. This paper also includes a description of CarLink (a commuter-based carsharing model) benefits. business parks in the East San Francisco Bay region deployed the first rail-based shuttle services (3). Shaheen 2 years later (3). and key findings from this study. Today. CarLink can provide individuals with a shared fleet of vehicles for accessing transit on the home. and linked to the Lawrence Livermore National Laboratory (LLNL) (12). Homebased Users). In the San Francisco Bay Area. most commuter subscription services have been taken over by local authorities. to fill in service gaps between the auto and traditional transit (3). the 1999 CarLink field test. While most shuttles rely heavily upon local support (e. On evenings and weekends. 9). this same region was the home of the CarLink field test..e. taxes and vehicle registration fees). Given this history of transit and highway subsidies. and future of carsharing internationally. A. Interestingly. feeder services almost exclusively serve commuter markets and connect to a single workplace or neighborhood (e. carsharing offers a feeder service alternative with commercial potential.or destination-end of a transit trip (10). However. these services are much less common. it is not surprising that only a few viable markets (e. In the 1980s.g.. however. and commuter-based carsharing primarily in Northern California.S. station cars. both of these programs were repealed in the early 1990s (3). however. and employment centers. neighborhoods.. airport shuttle services) have evolved in the U..e. Today. such as CarLink. Homebased Users and Workbased Commuters). In the last 15 years. They transported Walnut Creek Bay Area Rapid Transit (BART) District commuters to and from the Bishop Ranch and Hacienda Business Parks. In 1974. TRANSIT FEEDER SERVICES: AN OVERVIEW In the 1970s. They bridge the gap on the origin. a controversial step in national transit policy (6). employers. A commuter-based carsharing model. fostered by federal ISTEA and TEA-21 legislation.g. private subscription commuter vans and buses. particularly employersponsored shuttles. 1999 (8.e. Furthermore. Indeed. transit feeder servicessupported by employers and developershave emerged that help attract and retain employees and promote travel demand management. it explores several ways to strengthen CarLink viability. who drive them home from transit stations at the end of the day. suggests that an opportunity may exist for more sustainable market niche services to emerge (3). 8.. 9) for a discussion of the past. This paper provides an overview of transit feeder services. Finally.and work-end of a trip (i. Residents and developers versus employers (3) typically organize such services.. were popular throughout the U. new legislation was adopted to provide federal aid for both operations and capital improvements. It argues that carsharing can provide a complementary niche service (not a replacement) to traditional transit and feeder shuttles.) As a supplement to traditional transit. apartment buildings) destination (3).

“station car” concept. although other types of vehicles can be used. which are not mutually exclusive (12). Bernard (14) defines carsharing and station cars as two separate concepts. is a European development that usually begins as a local cooperative with one or two vehicles parked in several residential neighborhoods.. Furthermore. circulating from a transit station to one or more employment sites during peak commute periods. Majority of commute time on traditional transit versus shuttles (i. service limitations do exist. A few major employers. In the San Francisco Bay Area. CarLink is a station car program. many San Francisco Bay Area shuttles include one 22-passenger van. 13).S. Shaheen 3 innovative partnerships between employers and rail transit operators have formed recently to provide feeder services. there are over 100 shuttles in the Bay Area (3). definitions are evolving.000 per vehicle a year to operate. flex hours. 12. Station cars are typically small electric vehicles for environmental reasons. versus many small businesses. such as the BART Station Car Program and CarLink (3.e.. A. The author of this paper regards CarLink as a blend of both the station car and carsharing concepts. unpredictable schedules.000 to $80. Next. which are seldom used for commute trips. Serve similar origin-destination patterns for a substantial number of employees.e. Typically. These service gaps provide a complementary niche for commuter-based carsharing programs. The “station car” concept includes several to many cars parked at central locations. . or preference for personal vehicles). a shuttle trip is less than 15 minutes). Common characteristics of successful employer-based transit shuttles include: • • • • • • • Linkage to concentrated employment centers. there is a potential for unmet demand. “station car” rental program at the Fremont station. though somewhat overlapping. willing to support the service. in May 2000. the author reviews six reasons CarLink could provide a complementary service to traditional transit and feeder shuttles. either fully subsidized or shuttle costs included in price of a monthly transit pass) (10). Frequent service levels during peak commute periods with timed transfers. Under this definition. Indeed. It is interesting to note that CarLink has its roots in both European carsharing and the U. Since shuttle capacity is somewhat limited. according to Bernard. Although feeder shuttles are quite successful in the Bay Area. BART and Hertz jointly launched a commercial.. First. transit feeder shuttle services continue to gain popularity in Northern California. and Employ integrated fare policies (i. Carsharing. they include peak-period services (or headways that are timed with transit schedules) to reduce wait times at transit stations. As with any developing concept. CarLink could supplement such van services and perhaps attract customers who are unwilling to take a shuttle service for a variety of reasons (e.g. for subscribers to make local trips. transit feeder shuttles cost approximately $75. TRANSIT FEEDER SERVICES: IS THERE AN UNMET DEMAND? Today. outside walking distance of a transit station. such as transit stations and businesses. Congestion and parking are barriers to auto use.S.

the CarLink service includes a guaranteed ride service (e. either residential. could decrease traffic by reducing the number of cars needed by households and encouraging commuters to walk. the CarLink model.S. A. at least for part of their trips. such as Silicon Valley. are typically unable to use a shuttle service. shuttles normally operate only during peak periods. who work late or variable hours.e. the number of subsidized shuttle services deployed in a region each year is limited. COMMUTER-BASED CARSHARING: POTENTIAL BENEFITS A shared-use system. Shuttles are often co-funded by employers. the Transportation Fund for Clean Air (i. It is important to note. CarLink can provide a more demand-responsive alternative to individuals who may need to travel at times different than those covered by the shuttle service. For commuters and employers especially. In the case of employer-based shuttles. taxis) in case a vehicle is not available. Based on these arguments. For many shuttle programs. It is important to note here that CarLink vehicles could carry up to five passengers (carpooling is highly encouraged and facilitated by advanced technologies).” and “Day Users. the author supports that a potential niche market exists for commuter-based carsharing. Fourth.. Fifth. AB 434 funds are generated from California vehicle registration fees to support air quality management programs. Third. which is complementary to existing services. or take transit are typically restricted to the work site during the day.. that these funds are limited. feeder shuttles serve only one side of a transit commute (i. Hence. researchers found that the shared-use fleet actually increased the mobility options of participants who carpooled. such as feeder shuttles. The remainder of this paper describes potential carsharing benefits. and local transit providers. many smaller employers (the predominant model in Silicon Valley) are unable to support a shuttle service. shared-use vehicles could offer a . such as CarLink. AB 434 funds cover another 25 percent. vanpooled. bike.g. however. the 1999 Northern California field test. thus.e. CarLink offers a parking management solution to transit providers since shareduse vehicles can serve multiple transit customers per day with a single parking space. or took transit to work (12). Shaheen 4 Second. and key study findings.or employer-sponsored). and recommendations for strengthening the viability of a commuter-based carsharing service. Furthermore. the guaranteed ride service was never used (12). most vans are fully subsidized for employees. individuals who vanpool. CarLink could serve employers of almost any size (by scaling the number of vehicles contracted) without the level of local subsidy needed by a traditional feeder shuttle service. During the CarLink field test.” (See “CarLink Field Test Description” below for an expanded description of the CarLink model and study results.) Finally. CarLink could also provide a supplementary mobility option to individuals who carpool or take transit by offering an on-site vehicle fleet for business and personal trip making during the day. In addition to providing a more demand-responsive alternative. Furthermore. so funding is limited. carpool. biked. With CarLink. Assembly Bill (AB) 434 funds). Indeed. services are typically limited to a few employees during peak periods. Funding for shuttles is highly competitive. CarLink can give transit providers a means of attracting new customers through a more efficient use of their parking spaces.. timed shuttles can only provide connectivity to individuals whose schedules are within service hours. the same vehicle fleet can serve all “Homebased Users. hence. Individuals.” “Workbased Commuters. it is not uncommon that employers are unable to secure a shuttle service in a highly congested region. transit authorities pay 50 percent of shuttle costs. and use transit. and employers fund the remaining 25 percent. It is interesting to note that during the ten-month CarLink field test.

. called CarLink. and ended on November 15. In 1999. low-hassle alternative to private vehicles. • • • Homebased Users drove a CarLink vehicle between home and the Dublin/Pleasanton BART station daily. transit operators could benefit by tapping a larger potential market. Workbased Commuters took BART to the Dublin/Pleasanton station and drove CarLink vehicles to and from work at LLNL. and weekend riders. LLNL). and society might benefit by diverting travelers from single-occupancy vehicles to transit for part of their trips. Sharing vehicles could even free up parking spaces. carsharing could reduce the need for additional household vehicles. The cars were based from premium parking spaces at the Dublin-Pleasanton BART station. was developed and tested in Northern California. 1999. . and East Bay communities.e. Oakland. Commuter-based carsharing could reduce government spending on arterial street systems and mass transit by increasing transit ridership through added reverse commuters and midday. each group paid a distinct fee according to the duration of car use. carsharing could help air quality by incorporating low-emission vehicles into shared-use fleets. keeping the car overnight and through the weekends for personal use.50 per hour and $0. There was a fee of $60/month per car. which was shared with a co-worker by carpooling. Participants did not pay for work trips because LLNL donated the compressed natural gas (CNG) fuel for this program. Fifty-four individuals enrolled in the program and shared 12 natural gas powered Honda Civics. The CarLink field test combined short-term rental vehicles with communication and reservation technologies (i.” During the field test. 1999. and LLNL. transit links for “Workbased Commuters.. Moreover. California Department of Transportation (Caltrans). CARLINK FIELD TEST DESCRIPTION The CarLink field test was launched on January 20.S. smart technologies) to facilitate shared-vehicle access. Davis. There was a fee of $200 per month for this package. Project partners included American Honda Motor Company. INVERS (a German-based smart carsharing technology company) and Teletrac provided the advanced carsharing and vehicle tracking technologies. a commuter-based carsharing model. the BART District. Day Users employed CarLink vehicles for business trips or personal errands during the day. The participants were from San Francisco.” and shared vehicle access at employment sites through “Day Use. Partners for Advanced Transit and Highways (PATH). The model incorporated traditional and reverse commute travel patterns and a day-use fleet application. It could aid employers by attracting and retaining employees (e. to the suburbs) and reducing the need for additional corporate vehicles to support travel needs. Individuals and employers could benefit by gaining the mobility of a car (linked to transit) without individually carrying the full ownership costs. tested at an employment center (i. Shaheen 5 low-cost. The fee was $1. evening.e. A brief description of each user group follows. A. The CarLink model includes three separate user structures: a “Homebased User” lease. The tenmonth demonstration was implemented and researched by two teams at the Institute of Transportation Studies-Davis (ITS-Davis) at the University of California. by serving multiple users each day.. Furthermore.10 per mile for personal trips. vehicles would spend less time parked at transit stations and employment centers.g.

Carsharing. and carpooling resulted in a net commute reduction of approximately 20 vehicle miles (or 32. CarLink resulted in at least 20 new BART trips each day. since most Homebased Users already used transit prior to CarLink... Shaheen 6 All user fees included fuel. minivans and pickup trucks).e. Using questionnaires. preferred parking at the BART station). BART. they would sell one of their personal cars. smart carsharing technologies). Principal CarLink program costs included vehicle depreciation costs.. Workbased Commuters said they were more hesitant about selling a private vehicle until transit services improved and CarLink supplied more lot locations and vehicle variety (e. insurance. Most Workbased Commuters interviewed said that they would return to solo driving after CarLink ended.e. CarLink implementation staff supported the program by cleaning and occasionally refueling the vehicles. cost and revenue data.g. behavioral findings. The combination of CarLink. This is not surprising since the field test objective focused on gauging participant response rather than on optimizing costs and revenues. 54 enrolled). Although the CarLink participant sample was not statistically significant (i. Key study findings are: • • Even though many CarLink users’ commutes took longer (on average. CarLink findings include operational understanding. This paper explores the potential of shuttle vans and commuter-based carsharing in filling key transportation service gaps. and maintenance costs. and monthly operational costs. as well as maintaining e-mail and phone contact with users. maintenance and administration. but would try to carpool more frequently than they had previously (12). researchers explored CarLink attitudes and use over time. Carsharing’s commercial potential is appealing in . or startup. and focus groups.2 kilometers) per day (on average) across the fleet due to a transit mode shift. This reduction was primarily due to Workbased Commuters. It is important to note that during the CarLink field test. researchers did not strive to reduce costs (12).g. fuel. CarLink drivers used personal vehicles less than before they joined the study. and directions for future research (12).. and the COCOS and Teletrac technologies (i. Approximately 20 percent of program costs were covered by the field test. perhaps because they became more familiar with the transit system and had easier access to it (e. For a preliminary CarLink economic analysis. household interviews. participant profiles. in particular. which would greatly reduce their transportation costs. has the potential to become an economically viable. they found them less stressful. costs were separated into fixed. Several Homebased Users said that if CarLink became a permanent service. insurance. carsharing and transit shuttles are gaining popularity as modal alternatives that provide connectivity and increase transit use. A.S. approximately 10 minutes longer). Those in the Workbased Commuter group also increased their use of BART for recreational travel. Furthermore. demand-responsive service that complements existing transit and shuttles. • • • • • CONCLUSION Today. In addition to vehicle support services. valuable lessons may still be drawn from these survey and interview results. Roadside assistance and an emergency taxi service were also provided.

and reservations. there is a potential to reduce labor and management costs. Some incentives are currently available and should be promoted in conjunction with carsharing (e. thus. and billing of a shared-use fleet and help keep personnel costs down. The CarLink field test enabled researchers to specify the carsharing technology requirements needed that could be provided at a more affordable cost. others include federal and state credits and sales tax exemptions (15).g. the author provided six reasons that commuter-based carsharing might supplement existing services. reservations. and 3) employing alternative fuel vehicle purchase incentives to reduce vehicle costs.. Although the preliminary economic analysis of the CarLink field test revealed revenue shortfalls.. A potential way to reduce CarLink capital costs would be to buy cleaner vehicles and utilize alternative fuel vehicle purchase incentives. Several incentives have been established on federal. • Carsharing Technologies Streamlined. shuttle services are typically provided only during peak commute periods. carsharing would contribute further to the environmental benefits associated with a shared-use vehicle system. approximately 75 percent of total costs). 17).e. primarily the employment side). shuttle capacity is somewhat limited (e. Some are in the form of direct air quality management district incentives. or vanpool to work is often restricted to an employment center during the day. . and regional levels to encourage the purchase of low-emission vehicles. would be more challenging to implement. Third. By employing alternative fuel vehicles. billing. First.e. increased revenues.S. shuttles only serve one side of a transit commute (i. in marketing materials). so service availability is also limited.. A chief advantage of a smart carsharing system is the ability to expand program size with only modest increases in operations personnel. policy incentives. Disincentives. CarLink can provide a parking management solution to transit providers and employers by serving multiple users with a single parking space throughout the day. Alternative Fuel Vehicle Incentives. Fourth. Second. Program Expansion with Limited Personnel Increase. By streamlining vehicle tracking. 2) expanding program size with limited personnel increases. CarLink can provide a supplementary transportation option for business and personal trip making throughout the workday. most are 22passenger vans). and local program support. including subsidized transit passes and premium parking at transit stations. including cost-reduction strategies. Each of these areas is described below. • • Policy Incentives A range of policy incentives should be explored to encourage carsharing and transit use. state. carpool. Nevertheless. individuals who work variable or late hours are often unable to use them. In contrast. Cost-Reduction Strategies Several cost reduction strategies include: 1) streamlining carsharing technologies and reducing program management costs. CarLink can serve transit connections on both the home. such as restricted parking. Finally. Smart technologies can greatly facilitate the management. Fifth. shuttle funding is competitive. In this paper. when parking is limited individuals are more likely to use another mode (16. A. Shaheen 7 Northern California since shuttle vans rely heavily on Bay Area subsidies (i. CarLink’s long-term potential and viability could be strengthened through a combination of approaches.and work-end.g. the mobility of individuals who take transit..

Canada revealed that when employers pay for parking. Premium parking was found to be a significant program incentive and should be emphasized whenever possible. If carsharing organizations employ alternative fuel vehicles.e. by changing zoning ordinances and eliminating free parking. they could obtain an identification sticker from the Department of Motor Vehicles that allows those vehicles to be used on HOV lanes even when there is just one driver. 3) encouraging public-private partnerships to initiate and support such programs. over 1. In the Bay Area. while potentially helping to meet air quality control requirements. particularly where parking is limited (12). which provides financial incentives to employees to use transit or vanpools. More Efficient Parking Use. Through carpooling. local governments could stimulate use of other modes.5% saving on payroll taxes. gasoline or alternative fuel) access on HOV or HOT lanes (3). Increase Revenues and Local Support Commuter-based carsharing viability can also be strengthened by increasing local support for such programs and through increased revenues. Use of HOV lanes typically requires two occupants. as either a supplemental employee benefit (payroll tax savings) or a substitute for salary (a pre-tax salary deduction) (18). researchers estimated that CarLink could reduce parking demand at a BART station by four spaces or $400 per month (i. including zero-emission vehicles. If employers purchase transit passes for their employees. guaranteed parking spaces were provided at the BART Dublin-Pleasanton station. which grants single-occupant vehicles use of HOV lanes for electric and alternative fuel powered vehicles. During the CarLink field test.e. Zoning Ordinances as A Mode-Shift Incentive.. Once demonstrated as reliable and convenient. 2000. particularly in congested corridors. a carsharing program could help reduce a transit provider’s or an employer’s parking needs (and increase the value of the CarLink program). whereas if individuals pay their own parking only 39 percent (on average) drive alone (19). A. ultra-low emission vehicles. and 4) conducting research on the non-monetary benefits of carsharing. carsharing users would be likely to pay more for this service. approximately $400 per year on income taxes). Shaheen • 8 • • • • User-Side Subsidies. During the field test. Indeed. Parking case studies from four Los Angeles regions and Downtown Ottawa.. Several strategies include: 1) increasing user fees. Local governments could also encourage carsharing use by allowing shared-use vehicles (e.S. all CarLink user groups appeared to be willing-to-pay more . According to Shoup (16). Preferred Parking at Transit Stations and Employment Centers. • Increased User Fees. 66 percent of employees (on average) will drive alone. transit and vanpooling benefits are now tax free to employees under TEA-21 (i. such as transit and carsharing. For instance. and super-ultra-low emission vehicles. each space is valued at $100/month) (12). On July 1. Such benefits should be emphasized to employers and transit providers when carsharing services are proposed. Increased user fees would certainly help support economic viability.500 employers participate in the “Commuter Check” program. they are also eligible for a 7.S.. 2) obtaining demonstration grants to support carsharing as a “developing” industry (3). High-Occupancy Vehicle (HOV) or High-Occupancy Toll (HOT) Lane Privileges.g. Commuter Checks (or transit vouchers) are sold to employers throughout the U. California enacted AB 171. Reduced transit fares and customer discounts should be explored as a means to encourage transit and carsharing use.

The CarLink field test provided a starting point for a full benefit-cost analysis of commuter-based carsharing. employers.) to individuals and society throughout the U.g. this new mobility option should be explored further. (2) Vincent. congestion relief. “Seed” funding could be used to support carsharing ventures. and local initiatives to foster carsharing’s long-term viability.J. Paratransit in America. John Wright and Nihar Gupta of PATH. Further Study of Non-Monetary Program Benefits. the University of California Transportation Center (UCTC). since the CarLink program was not large enough from which to extrapolate (12). policy incentives. etc.S.. neighborhood. PATH. Incorporated. Connecticut. commuter-based carsharing shows market niche and economic potential. REFERENCES (1) Davis. Demonstration Grants. using logic similar to that of the Urban Mass Transportation Act of 1964.. ACKNOWLEDGEMENTS The author would like to acknowledge Robert Uyeki of Honda R&D Americas. . Thus. Office of Highway Information Management. Demonstration grants could also support carsharing startups. (3) Cervero. LLNL. Partnerships among local transit providers. Oak Ridge National Laboratory. However. Edition 18. many societal benefits and costs were not estimated and economies of scale could not be calculated. Urban Travel Patterns. More research is needed to examine cost-reduction strategies. CarLink was based on a public-private partnership that was crucial to the field test’s deployment and success (12).S. including pollution reduction. However. Public-Private Partnerships. researchers should continue to explore the non-monetary benefits of carsharing. further study is needed to investigate the economic potential and non-monetary benefits of a range of carsharing models (e. the National Science Foundation. resorts. further investigations are needed to more accurately estimate market rates (12). In addition. M. Praeger: West Port. BART District. Such partnerships can foster and encourage the creation of unique transportation services that reflect the needs of the local community. American Honda. Department of Transportation. and the Dwight David Eisenhower Fellowship Program for their generous contributions to the CarLink research program. willingness-to-pay for services. and reduced land use impacts. In the future.S. Shaheen 9 • • • than they paid to participate in the CarLink field test. Thanks also go to the California Department of Transportation. and industry can be invaluable to the development and a viability of a commuter-based carsharing service. universities. M. 1994. and David Dick and Linda Novick of ITSDavis for their invaluable assistance gathering articles for the literature review and CarLink data collection efforts. commuter. U. FHWA. 1998. Keyes.A. Transportation and Energy Data Book. R. A. and M. 1997. This paper serves as a starting point for a market assessment of commuter-based carsharing from which other efforts might build. In conclusion. S. which provided grants for transit demonstration projects and encouraged local experimentation (3). Reed.

In Transportation Research Record 1666. (11) Shaheen. National Research Council. Washington. (18) Commuter Check home page. and M. TRB. and R. Overview of Travel Demand Management Measures. D. 1994. UCD-ITS-RR-99-16. 1995.J. 14-28. T. and N. 1994. 3. S. Sturges. M.com. Wachs. (5) Salon. D. K. (12) Shaheen. Novick. Shaheen. Sperling. (14) Bernard. D. 1991. Volume 61. and L. Transit Cooperative Research Program Report 55. Employer-Paid Parking: The Problem and Proposed Solutions. 162-192. Sperling. Wagner. 35-52. http://www. 1997. and the Environment. 1998. New Mobility: Using Technology and Partnerships to Create More Sustainable Transportation. 10 (6) Ortner. D. 2-41.C.. 1999. D. American Planning Association Journal. Shaheen (4) Gordon. 1666. Number 1. and D. . (19) Shoup. Bernard. pp. (8) Shaheen. Oakland. and Federal Assistance in a Historical Context. (10) Urbitran Associates. University of California. TRB. UCD-ITS-RR-99-1. 118-124. (13) Nerenberg. No. D. Volume 1.C. and C. V. (17) Comsis Corporation. Carsharing and Partnership Management: An International Perspective. Davis. (7) Petersen. Volume 42. pp. Journal of the American Planning Association. Multisystems. and R. Sperling. and C. Davis. Evaluation Results of the San Francisco Bay Area Station-Car Demonstration. Davis. UCD-ITS-RR-00-4. CarLinkA Smart Carsharing System Field Test Report. California. Transit Subsidies in New York and Chicago: Local. Energy. (16) Shoup. Washington. 4. and Future. pp. No. 2000.C. Journal of Public Transportation. TRB.C. National Research Council. Dick. Steering A New Course: Transportation.S.E. 110-117.S. J. University of California. Wilson. U. Collins. Wagner (1998). Institute of Transportation Studies. Dynamics in Behavioral Adaptation to a Transportation Innovation: A Case Study of CarLink—A Smart Carsharing System. Incentive Policies for Neighborhood Electric Vehicles. 1999. D. SG Associates. pp. Guidelines for Enhancing Suburban Mobility Using Public Transportation. pp. (9) Shaheen. J.. Department of Transportation. 1992. D. University of California. Volume 52. Accessed Friday. 1998. Present. M. Transportation Quarterly. pp. FHWA and FTA. (15) Lipman. S. 2000. D. In Transportation Research Record.J. D. Institute of Transportation Studies. Bay Area Rapid Transit (BART) District. Davis. Washington. Number 2. Wright. 10-21. S. Sperling. State.commutercheck... pp. July 28. Institute of Transportation Studies. 1999. University of California. D. D..E.. 1999. DOT-T-94-11. Island Press. San Francisco Bay Area Station Car Demonstration Evaluation. Kurani.. 1999. January 1979. Collins. D. and N.. Washington. A. Institute of Transportation Studies. Cervero. S. An Opportunity to Reduce Minimum Parking Requirements. UCD-ITS-RR-94-20. S. National Research Council. Transportation Quarterly. The Cost-Revenue Squeeze in American Public Transit. E. Carsharing in Europe and North America: Past.