Increasing RRT ridership: case study of Guangzhou

presented in Working paper of U de Montreal, 1999Increasing Ridership for Rail Rapid Transit : A Case Study of Guangzhou* Zhang yunquan
(Zhongshan University and Université de Montr閍l)AbstractThe paper will become Chapter 1 in the doctoral dissertation to be presented in 2000. In the context of urbanization and urban traffic congestion, in particular developing countries and China, the significance of constructing rail rapid transit (RRT) in metropolitan areas is addressed. Three aspects of literatures on RRT definition, characteristics and evolution ; transit ridership studies; and market segmentation are reviewed.The methodology consists of comparative study based on case cities --Montreal, Toronto, San Francisco, Hong Kong and Singapore; the case of Guangzhou; and sample survey. Transferable suggestions will be made for application to urban areas in developing countries and China.keywords: rail rapid transit; ridership; market segmentation; Guangzhou1. Context A philosopher once said ‘ transportaton is civilization, perhaps you think the saying is somewhat exaggerated, but it is substantially right. At least, modern transportation is a symbol of civilization and progress of human kind. It is well known that the most fundemental needs of human--food, clothing, housing, and travelling, are closely concerned with transportation, directly or indirectly. From the experiences and lessons in many countries, it is necessary to recognize and rethink the future and orientation of urban transit. In developed countries , traffic congestions not only arises in urban areas, but in freeways and suburbias. In developing countries, the situations in urban areas are being worse. Revitalization of urban transit and intensive utilization of land use is accepted as concoidant awareness by public and foresighted personnels. Rapid rail transit is an important component of urban transit, it will contribute to benefits for mankind, However, due to its cost, construction time , communication and automatic signal controls, many issues and impacts are worth studying and approaching.The findings will be very useful for decision-making and construction of rail transit in China and other developing countries. 1.1 Rapid Urbanization in Developing Countries, in Particular China In early 19th century, industrial revolution broke out and led to the great improvement of productivity. While early cities were typically the center of production and commerce due to the effectiveness of aggregation and economy of scales in Western countries, the cities were like a gigantic magnet attracting a great amount of farmer from rural areas into city for making their livings. Thus, urban sprawl took place with economic development * The author would like to be grateful to Dr C.Comtois and Dr B. Slack for helpful comments on my early draft. and urbanization is the result of industrialization. Then global urbanization level was about 10%. After World War II, world population are dramatically increasing ,in particular developing countries. In 1990, the world‘s urbanization reached 46.1% ( Xu,et al 1995). In 1995, there were 5.7 billion population in the world, it is estiamted that the sum will be up to 9.4 billion in 2030. About 90% of the increase occur in developing countries. The 90% of new incremental population will concentrate in urban areas (Laquian, 1996). The number of cities with 1 million people increased from 197 in 1981 to 309 in 1995 ,and there were 14 metropolises with morn than10 million population , over 50% of which were in developing countries(Liu,1998). China is the most populours in the world, with 1.224 billion in 1997, accounting for 21%, 55.5% in APEC. The urbanization level was 26.4%, less than the world’s average (46.1%), and urban population was 0.214 billion. But by 2000, the urbanization will be up to 47% (Industry Canada,1997).The total cities increased from 191 (1978) to 467(1990), and 666(1997)( Xu, 1995; Wu,1998).1.2. Large scale of investment in transport infrastructure Transportation , with energy and communication are the dominant factors restraining the high speed development in Chinese economy. With in-depth reform in ecomonic system and transition of socialist market economy, the emphases are placed on the investment of large scales in the transportation . The budgets will be reached US$100 billion in transport infrastructure and $10-15 billion in rail rapid transit(Wu,1998; Pendakur,1992). For example, by November,1998, 180 billion RMB ($21.7 billion ) was pooled in highway infrastructure, total road length was 1.26 million km, and freeways exceeded 6000 km(People抯 Daily,1998,12.10).1.3 The need for rapid rail transit(RRT) in urban areas It was reported that rapid urban growth had taken place and there were 666 cities total ,33 cities with 1 million population in China (1996). The rapid economic growth is the principal contributor to urban development in contemporary China. China has become the rapidest in annual rate of 11% during 1981--1994. Though the ASEAN financial crisis, the GDP rate was 7.8 % through expanding internal demand to stimulate economic growth in 1998, when the rate turned negative increase in Hongkong, Thailand, Phillipine. Rapid economic growth led to vast flows in passengers and freight transportation. Traffic congestion is becoming an extensive , serious problem confronting with municipal authorities. The bus operating speed runs 15-20 km/h, furthermore,10-15 km/h in central urban areas (Stares and Liu, 1996). The issue of transportation not only has impacts on economic development, but on environmental externalities, such as poorer air quality, longer trip time and traffic accidents. It is well known that rapid rail transit ,as one of modern public transport modes with higher capacity, speed,safety, reliablity and environment-friendly, is playing more and more role in most of metropolitan regions. However, the system’s non-flexibility, longer construction period and lion-share vast capital and operating costs usually aggravate pressure and burdens for most metropolis, even though rapid rail systems have completely operated for years. Rarely can you find that no subsidies are supplied by municipalitis except for Hong Kong (Allport, 1997). Obviously, transit riderships are closely related to revenues or profit of operators. Currently, there are increasing trends of diversified investors, public and private sectors, in China, and confronting with a dilemma both to increase passenger volume and to reduce traffic congestion under insufficient resources.2.Problem How to attract ridership for new RRT?3. Literature Review3.1 Definition of rapid rail transit (RRT) A RRT system generally serves on urban area, using high-speed,electricially power passenger rail cars operating in trains in exclusive right-of-way, without grade crossings and with high platforms. The tracks may be in underground tunnels, on elevated structures, in open cuts, at surface level, or any combination. Other terms are: metro; subway, underground, heavy rail transit, rapid rail, metropolitan railways, etc(TCPR ,1996)3.2 Characteristics and Historical studies of RRT3.2.1. Characteristics of rapid rail transit Rapid rail transit system is main part of urban transit systems, which include bus transit, light rail systems , rapid rail transit (metro,or U-Bahn in Germany) ,and suburban rails or commuter train . In boarder sense, rail transit systems involve metro, light rail transit (LRT), suburban rail and trolleyways or streetcars. The rapid rail transit (RRT) in the paper refers to so called metro, subways or underground railways. RRT invariably operates on completely exclusive rights-of-way, which may be underground , elevated or in cutting, and at high speeds up to 120 kph. Average operating speed between two stops is 30--35 kph, it can provide the highest capacity of 70,000 passengers per hour per direction. The train typically carried about 750 passengers, and station spacing is 1 km or less in the urban areas , not more than 2 km in the suburbs. Construction of metro is a difficult work, there are several types of building subways , such as ‘cut-and cover’, deep-level boring by shield, ‘ umbrella’, or laying pre-cast concrete tunnel underwater. Due to high passenger volumes, RRT systems usually require sophisticated signalling and control devices. The stations are built with wide,high-level platforms facilitating rapid loading and unloading , and usually equipped with escalators. Uderground RRT needs comprehensive ventilation, particularly in hard climate to keep comfortable in train cars. Meanwhile, the systems require advanced technology in vehicles, civil construction and operating management. The disadvantages of the systems are costly in construction. The investment of US$120 million /km or above ,operating cost US $0.10-0.15 per passenger-km may be needed. It is mainly dependent on constructing methods, geological and hydrological conditions. The systems lack more flexibility than bus modes, needing be supplemented by buses to provide feeder services for rail stations. If whole public transit systems are involved in diversified ownership, such as public and private sectors, there usually exist in difficulty of how to share operational revenues. In addditon, the construction of rapid rail system would take longer --at least five years, which obviously would affect urban surface transit to a larger extent. Farthermore, the costs of modifying original routes are likely to be enormous, but it is permitted to extend rail routes from terminals. The advantages of RRT can provide passengers with quickness, comfort, safety and reliability. Using clean energy--electricity, leads to no pollution to atmosphere. Its operating cost is relatively lower than that of other modes. RRT can increase modal choice and business efficiency for passengers, promoting urban and regional development, upgrading urban image and amenity of communties. RRT 憇 characteristics is shown in Table 1 in comparison with other transit modes. Table 1 Key characteristics of RRT compared with other transit modesCharacteristics LRT Bus Commuter Rail Automated Guideway RRTSystem cost: Initial M L/M L/H H H++ operating M H H H LAttributes: reliability excel. fair good super excel. grade separation vary less more 100% 100% automatic No No No Yes Yes entrained vehicle Yes No Yes Maybe YesPublic perception: ride quality good fair good good good route identity easy hard easy easy easy/hardsocial acceptability H L H H H/LRailroad involvement: operating labour No No Yes No Nofreight coordination maybe No maybe No No
Source: Adapted from Schumann,J. W. (1989)3.2.2. Historical studies of rail transit In 1863, the first subway in the world was opened in London, then it was the biggest city around th e world. The event remarked the beginning of raid rail transit. In fact , RRT is a kind of application of convertional railways or railroads in urban areas. Hence, telling of the history of rail trasit , it is reasonable to date back history of railway advances. Unlike shipping, road and water teansport with a history of several centuries , raiways have a relatively recent history of about 150 years (Simon P. Ville, 1990). The context, iron rail and the steam locomotive were separately developed. The earliest railways or wagonways, were simple wooden tracks with lower frictional resistance. In 18th century, wooden tracks were useed in transporting coal to riverside staiths. Then the tracks were replaced by iron because of its greater durability and lower friction. In 1821, John Birkinshow patented of rolling wrought iron for rails. In the second half of 19th century, steel rails replaced iron. In 1804, Richard Trevithick developed a high pressure steam engine, later Stephenson innovated the engine into useful locomotive called ‘ Rocket’, which 1829 was applied to route from Liverpool to Manchester, the 1st steam-driven passenger railway in the world. Britain led the world in the construction of railway system, then transfered to other countries in Europe. By 1850, nearly 10,000 km (9797 km) of railways had been built,German and France ranked the 2nd and 3rd , 5856 km and 2915 km respectively. By 1910, total railways in Europe reached 307,535 km. Most European railway systems had been completed in 1914 (Simon, 1990). In 1869, the transcontinent railways between New York and San Francisco were opened. In 1881, another great railway , Montreal to Vancouver in Canada was completed. In 1964, high-speed trains were produced in France, then in Japan, at operating speed of more than 200 kph (Rodrigue, 1998). Nowadays, supertrain TGV is able to run up to 350 kph, becoming effective rivalry to expressway and airlines for intercity or interregional services. In April, 1998, the SNCF company in France launched a new package of easier fares for rail passengers, in last 3 years there have occurred in strong increase of international rail passengers, SNCF expected a good return of profitability for the year of 2000 (Pepy, 1998). On the other hand, rapid rail transit almost kept pace with development of mainline railways.In fact, underground railways may be of same loading gauge as surface railways. It is the power supply that may be different from surface railways. The current supply in underground is ususally made with the 3rd rail pickup. After 1860 in London , some main line railways developed commuter traffic, and acute traffic congestion led to the construction of world抯 first underground railway, and in 1863 the 1st phase was opened.Then the 1st deep-level tube was completed in 1890. From then on , construction of RRT was quickly widespread to Europe, North America, later Latin America and Asia. Approximate openning years of the 1st phase of subway in the world are shown as followings.Table 2 Historical records of subways in the world Regions and city Year opened Europe London 1863 Glasgow 1896 Budapest 1896 Paris 1898 Berlin 1902 Hamburg 1912 Madrid 1919 Barcelona 1924 Athens 1925 Milan 1964 Rome 1955 Moscow 1962 Stockholm 1972 Munchen 1972 Rotterdam 1968North America New York 1892 Chicago 1892 Boston 1901 Philadelphia 1907 Cleveland 1955 Toronto 1954 Montreal 1966 San Francisco 1972 Washington, D.C. 1976 Atlanta 1979 Edmonton 1980 Baltimore 1983 Miami 1984 San Diego 1970sLatin America Mexico 1970s Sao Paulo 1970s Buenos Aires 1970sAsia Tokyo 1927 Seoul 1970s Hong Kong 1979 Osaka 1970s Nagoya 1970s Singapore 1987 Tunis 1970s Beijing 1969 Tianjin 1984 Shanghai 1995 Guangzhou 1999
Source:Nock,1973;Vuchic,1975;Pushkarev,1982; Hall, 1985; Stare, 1997; www.subways.net Light rail transit (LRT) can be considered as an advanced development of the conventional streetcar, which can be dated back 1890-1910. Urban rails are relatively recent. In Germany, LRT is called Stadtbahn or S-bahn. With modern , medium-capacity vehicles operating on streets in mixed traffic , in reserved street lanes and highway medians, in activity center mall and rights-of -way shared with other rail transit modes and in subways , it represents a mode of transportation between bus, streetcar and rapid transit. It operates usually in tunnel in congested city core areas. Rapidly increasing cost in heavy rail development and uncertainty regarding new transit technology served as an incentive to search for a different rail technology. The research led to Europe and especially West Germany, where the use of LRT was widely used. In 1960s-1970s, many of cities, such as Cologne, Bonn, Bielefeld and Frankfurt in Germany provided a few example of how LRT can be integrated into existing urban environments and multimodal transportation systems. Meanwhile, the research also led to Boston, San Francisco Edmonton, Calgary, Vancouver, Toronto, Buffalo, Denver, Dayton, Detroit, San Diego, Cincinnati, Portland, Mexico and Philadelphia for making use of existing electric lines and urban facilities, revitalising in light rail transit. Table 3 North America LRT systems
city year opened km passengers/d No. of cars downtown operation Boston 1897 52 70,000 230 old subwayBuffalo 1985 10 21,000 27 mallCalgary 1981 22 36,000 78 mallCleveland 1920 22 17,500 48 undergroundEdmonton 1978 10 25,000 37 underground Forth Worth 1962 3 6,500 8 undergroundLos Angels 1989 64 100,00 54 undergroundMexico City 1900 34 47,000 50 not centralNewark 1935 7 12,000 26 undergroundNew Orleans 1893 11 21,000 35 streetPhiladelphia 1892 160 127,000 236 ols subwayPittsburgh 1891 36 20,000 90 undergroundPortland 1986 24 21,000 26 mallSacramento 1987 29 21,000 26 mall-streetSan Diego 1981 33 23,000 30 mall-streetSan Francisco 1897 39 133,000 141 undergroundSan Jose 1987 32 40,000 50 mall-streetToronto 1892 73 334,000 284 all streetVancouver 1985 22 80,000 114 old tunnel
source: Adapted from Parkinson (1989) LRT currently demonstrates its effectiveness as a part of the total urban transport system in more than 300 cities of the world. There have been 2800 new light rail vehicles (LRVs) put in service.But it is not being advocated as a replacement for all bus, all private cars or heavy rapid transit system.The characteristics of LRT approximately is: technical simplicity; medium capacity (maximum 30,000 passengers per hour); medium cost and speed.(TRB Report 161, 1975; TRB Report 182, 1978; Alan ,1986) By 1980, miles of line of grade-separated rail transit (including LRT) in the world were 2175 miles (3500 km ), and 498 km miles (800 km ) were under construction, as shown Table 4. Table4. The distribution of rail trasit systems in the world(1980)
Rregion rail network length(miles) % of total length
USSR 223 10.2Asia 273 12.6Latin America 92 4.2North America 658 30.2Europe 965 44.4total 2175 100.0 source: adapted from Pushkarev, B.S., 19823.3 Transit Ridership Studies The Toronto subway has small but significant positive impacts on work trip generation rates and better access accounts for a 1.5% increase. By quantitative analysis, it is argued that simply providing better service will not necessarily attract more riders to the public transit modes because issues other than accessibility have an important impacts on mode choice(Wells,1994). Considering trip time, transfers have significant impacts on transit ridership (Vaga and Shortreed,1981). About 33% all Canadian transit trips involve transfers. The reduction in demand is thought to vary between 10% and 40% depending on the quality of the transfer itself. Data from Edmonton and Kitchener-Waterloo suggested that for a timed transfer of less than 5 minutes the demand reduction is about 15-20%, which atransfer with 5-10 minutes waiting time has about 25-30% reduction in demand. US DOT (1980) conducted the research on rider segments--handicapped persons. The ridership estimates are based on many assumption and incomplete data, the projections assume that transportation handicapped will exhibit characteristics similar to those of the non-transportation handicapped, and that discrimination will no longer occur throught the economic and physical environment. Queueing delay will be necessarily influenced on the ridership in rapid transit station Hall,1987). The model--Passenger Delay in a Rapid Transit Station was presented , concentrating on the delays encountered by exiting passengers. Because arrival pattern is lumpy.The method requires estimates of the average number of people exiting each train , average headway and average service rate as imputs. The average delay is easily calculated with a hand calculator. The transit ridership usually fluctuated on day-of-week or part-of-month basis (McCord and Cheng,1986). Results for a study of bus ridership during a 2-year survey for the Central Ohio Transit Authority indicated that ridershipp tended to decrease from Monday through Friday , and was higher at the beginning than at the end of each month. Possible reasons lie in disposable income difference or diversion of trips from transit to automobile. For new transit facilities , how to attract and sustain the ridership to increase revenue? Promotion marketing method might be effective by disseminating service information by free ride coupons (Capo and Messmer,1987). It showed that most coupons users were existing high-frequency, transit-dependent riders. Survey findings also suggest that with direct-mail promotions , it is not sufficient to target areas bordering bus routes, as many current and potential riders may not live in these adjacent areas. Taking the viewpoint of potential riders, users are not interested in technology but in the level of service the transit system provides(Jessiman,et al,1975). From several observations, it is found that level of service is a complex combination of many system attributes such as travel time, cost, comfort and convenience. Different user groups (market segments) make different trade-offs among these attributes. To attract maximum ridership, the system should be tailored to the particular needs and constraints of the market segments it is serving. Using disaggregate, behavioural travel demand models, they further addressed the ridership response of different market segments to different level-of-service packages. The findings indicated that no single service will be sufficiently attractive to all potential riders, but rail transit would be the most attractive mode of public transportation. Michelle (1987) concentrated on commuters as the target market segment, analyzing their response to transit marketing at the workplace. Two marketing theories--Peer Pressure Theory and Utilitarian Theory are tested. Santa Clara County Transit was used as the case study. Several multivariate statistical techniques-- correlation analysis, principal component analysis and multiple regression analysis were used to analyze the data. The result showed that new riders could be predicted with standard error of 1.6, yielding a multiple correlation coefficient of 0.8 between the number of predicted and observed new riders. In fare structure, redistributive effects are concerned with socioeconomic groups pay for a particular program and which socioeconomic groups benefit from it. It is concluded that off-peak reduced fare is more efficient than reduced fare at all hours. The redistribution , due to the reduced fare funding mechanism used in Illiois, appears to favor the poor (Rock,1979). Kemp(1974) investigated the impacts of the short-range improvement program on system ridership levels. For the major fare out of March 1972, the elasticity of ridership with respect to base fare level was roughly -0.15 to -0.20. The mass transit industry has long focused on regular riders, large numbers of trips and daily commuting. However, recent research challenges these views from on-board surveys, fare structure changes and transit voucher (commuter check) fare subsidy plan. The findings suggested that infrequent riders are a critical transit market and perhaps the key to building transit ridership and revenues (Oram et al,1996) It is not easy to accurately forecast ridership given a specific pattern of land use and other data as imputs, conventional ridership forecasting methodology typically calculates time and out-of-pocket cost to project ridership on each modal configuration for a selected design year. The deficiencies result from neglecting mode choice component and land-use impacts on potential ridership. For example, LRT and bus modes vary in their attributes. Specific modal attributes such as stations, passenger space and seating, ride quality, air pollution, noise, schedule reliability , safety , system identity , public orientation and familiarity can be rated for different modes. According to survey of real estate developers in 8 US cities , rail transit is rated the highest in this comparison, because LRT and other rail transit modes have stronger potential to induce adjacent real estate development (Herny,1989). Similarly, Harmatuck(1986) studied the seasonality of urban transit ridership. Short-term forecasts can be developed from time series and econometric methods. Time series methods are particularly useful when the factors remain unchanged. However, monthly transit ridership levels follow a seasonal pattern, a combination of time series and econometric methods can solve these problems. Khattak et al (1997) reported the results of a comprehensive behavioural survey conducted in Brussels, using ordered probit models of mode and departure time. The models show that commuters change their travel patterns in systematic ways when adverse weather occurs. In ridership forecasting, measures of ridership should be firstly identified , including average weekly passengers carried by the rail transit line or system; total ridership by all transit modes for corridor or entire urban area; and the chang in total transit ridership accompanying in introduction of a new transit facility. In order to estimate the impact of route changes on system ridership, Canadian transit operators used a forecasting model, its major steps are: Define catchment area boundaries; Characterize catchment area; Estimate total daily home-based work trips; Split total daily home-base work trips into downtown and elsewhere ; Apply mode split rates; Estimate peak-hour work and school transit trips from daily total; Estimate total transit trips from work and school trips; and estimate new and diverted riders (CUTA,1985). Transportation demand management (TDM) as transit policy played active role in disencouraging automobile use and increasing transit share. Dagang (1993) suggested 15 employer-based TDM measures in the speadsheet-based model. They are: Commute information program; In-house ride matching services; Transit pass subsidies; Employee transportation coordinator; Home-based telecommuting; Compressed workweeks; Reduction of employer subsidized parking; Preferential parking for carpools and vanpools; Bicycle lockers and showers; Guaranteed ride home; Shuttle to transit stations; Vanpool program; Reduction of parking supply; Direct monetary incentives for use of alternative transportation; and Transportation allowance. A framework to study the effectiveness of employer trip-reduction program was developed by L.A. Waldo (1993). The basic components include public policy; employer factors; travel mode characteristics; employee attributes; employee mode choice and program impacts. Combridge Systematics Inc. (1994) based on interviews with transit managers, concluded the activities and initiatives that have the greatest positive effects on ridership.The 5 major categories of ridership-oriented initiatives include: Service adjustment; fare and pricing adaptions; marketing and information initiatives; planning orientation and service coordination, consolidation and market segmentation. Charles River Associates (1997) made the research on transit’s ridership and its share of the market in US and Canada. They have examined a number of different policies potential for increasing or maintaining transit market share under unfavorable market conditions. Each case study provides valuable insight into the workings of particular types of policies in a designated local setting. These experiences are worhty of consideration for application elsewhere.The public policies include multi-agency cooperation, innovative use of existing infrastructure, method of serving low density service territory, coordinated planning effect, rapid transit expansion, land use, HOV (high occupied vehicles) lane, fare integration, feeder service, etc. It is well known that transportation is dependent on land use patterns to greater extent. From developer survey and pedestrian intercept surveys conducted in California, one possible strategy for increasing transit ridership would be to concentrate more housing and workplaces around urban rail stations (Cevero,1993). The measures might result in other secondary benefits: improved air quality, higher revenues , inner-city redevelopment and increases in affordable housing and infill development and more efficient urban form. 3.4 Review of literature on market segmentation It is more than 130 years since the world抯 first subway opened in London in 1863. Nowadays, there are over 300 cities with RRT or light rail transit(LRT) system around the world. At the turn of 21st century, people have well recognized that RRT or LRT is an alternative contributing to urban sustainable development and work efficiency. In the era of growing mature in rail transport technology, the risk is lower in applying proven technology in tunnel boring, cut-and-cover, viaduct, underwater concrete tube positioning, signal automatic control, optimal manufacturing for trains and cars,etc. But how to convert cost-intensive into cost-effective? The issue is critically important for planners, operators and researchers. The hypothsis is to increase ridership through market segmentation in existing rail transit system. It is well known that the ridership is closely related to operating revenue.Higher fareboxe revenue can help to cover the operational expenses, and reducing governmental subsidies for public transportation. Market segmentation is an effective technique of attracting or increasing ridership. Generally speaking, market segmentaton is classified the category of economics. In the past, the theory was ignored or little highlighted by researchers. Many people today believe that market segmentation is the key strategic concept in marketing. However,the basic idea behind segmentation has been around since the beginning of trade. It was only in the 1950s that this idea was formalized by researchers and pursued earnestly by large number of business firms and other types of organizations(Rebecca, 1998a). For the market segmentation was originated from business and trade, it is not surprising that most of documentations have been engaged in production (Winter,1997; Kinnucan,1997), labor market (Orr,1997; Grimshaw,1998; Funkhouse,1997; Leontaridi,1998), occupation(Anker,1997), export (Erm and Mengue,1997), tourism ( Andersen,1997; Cha,1995;Sheauhsing,1992 ), environmental assessment (Howenstine,1993) , residential property( Hoesli, 1997), as well as transport (Jukubiak, 1990; Roberts,1986; Trombly,1986). The declines of transit ridership have become a generized trends around the world. The results lead to rapid growth in governmental subsidies and great deficits on the account of transit operators. In one recent survey of 147 transit agencies, three out of ten agencies reported that in the past three years, ridership on their system had decreased. In another survey of more than 200 transit agencies in the United States, 44% of all managers said ridership--particularly the problem of declining redership--is one of the most important current issues for their agency. Even funding is related as a lack of rider results in lower farebox recovery rates (Rebecca, 1998b). In developing countries, due to insufficiency in transit facilities or a lack of effective management for new transit infrastructure or fluctuaion in system transition, the phenomenia occur in many cities. In order to attract ridership in new rail transit system using market segmentation, it is essential to identify the implication, connecting segmentation with transport attributes, making a design in research methodology as well as translating maket segmentation into transit strategies and tactics in case study.4.3.1 Characteristics, functions and limitationsCharacteristics· the process by which market segment are identified;· the foundation for an overall marketing strategy;· a process of analyzing markets, finding a niche,and developing and capitalizing on a superior position with that niche; and· central to a market and customer needs and wants.Functions· design responsive products to meet the needs ogf the marketplace;· develop effctive and cost efficient promotional strategies;· provide insight on present marketing strategies;· provide data on which to base resource allocation decisions; and· have available important data on which long-range planning for market growth or product development can be basedLimitations· Findings that may only provide a composite profile of the market;· changing lifestyles has made segmentation more difficult; · segmentation is not a remedy for other marketing organization deficiencies;· segmentation抯 effectiveness is limited by management抯 ability to use the results; · many segmentation studies emphasize methodological and statistical procedures over substance; and· the marketing research required for market segmentaiton may be expensive.4.3.2. Application in transport industry Generally speaking, market segmentaton is classified the category of economics. In the past, the theory was ignored or little highlighted by researchers, transportation planners, transit operators and managers, considering transportation as a service industry is rather different from manufacturing, processing and production sectors centering on their final products, while transportation only provides service to meet the demands of customers. They generally thought service had substancial differentiality with products. The idea lasted a relatively long time, especially during product-oriented period. All operationg activities were rotated around product, emphasizing on production yield and quality, no matter what needs in the marketplace or customers. After a large scale of production, Fordist as remarking cornerstone, oversupply of products led to fierce competition in the market with numerous producers. Under the historical background, contemporary marketing theory was established. The primary principle of marketing theory is at the core of customers or market, thus firms or enterprises produce what the customers want rather than make a decision on expectations or wants of producers themselves. The implicatiom of products were extended, and a perfect product should contain internal quality, external quality--package ,and service. When users buy the product, they are entitled to obtain relative service--presale or postsale service. Expanding the concept of product leads to application of market theory in transporation. In comparison with production or trade fields, the fewer documentations are available in transportation industry. These applications are distributed on several regimes, such as transportation policy, planning and management, energy and environment, travel pattern and behavior, transit marketing , travel demand analysis, and mode choice model,etc. In transport policy, a survey of all current Austrian Car-sharing Organization(ACO) members is used to quantify urban local market segment potentials and applied to two residential areas. Results indicate a substantial reduction of aggregate private vehicle mileage. Combining behavior impact with market segment size results in the quantification of emission reduction and car ownership reduction due to car-sharign, which is a decentralized demand-side transport policy(Steininger,K.et al,1996). Toh,R.S. et al(1990) identified the behavioral , attitudinal and demographic correlates of frequent flier members and non-members as well as light, medium and heavy users of air travels in the US using multiple discriminant analysis and other statistical techniques. These membership categories are viable bases for segmentation and the appropriate marketing mixes are then designed to comform to the discriminating characteristics of the target markets of different airlines. Dobson et al(1978) illustrated a relatively uncomplicated and effective perceptual market segmentation procedure for transportation policy analysis through a flowchart describing the technique, an empirical application and the tests of the reliability of the derived market segmentation structures across split halves of a data set. The procedure was calibrated on a sample of Los Angeles CBD workers.The segmentation structure was correlated highly with appropriated mode choice patterns. It was also observed that perceptual segmentation membership was a stronger determinant of mode choice than zone network time and cost. The segmentaion can be used in developing short-range forecasting models and its potential for developing information aids to target groups of travelers. Echols, J.C. (1983) identified several issues on Workshop on Systems Operations, including 1). Deregulation impacts; 2). Financial consquences; 3). Market segmentaton ; 4). Social goals;5) Users subsidies; and 6) Equity. Margolin, J.B. et al (1978) examined consumer motivation concerning ride sharing, particularly carpooling ,according to a market segmetation approach. The major market segmentation involved dividing the sample by commuting mode and pattern and by occupation type. The factors discussed include, 1) attitudes toward costs or interpersonal aspects of carpooling, time variables, carpool routes, parking management and convenience issues; 2) demographic characteristics of the two types of commuters. A special analysis focuses on the attitudes of those solo drivers who stated that they were interested in carpooling versus those who stated that they were not. The purpose is to highlight the motivation of a prime target group toward three carpool strategies: carpool match method, parking management, and dedicated carpool routes. In fare elasticity, results of the Chicago Transport Authority(CTA) surveys revealed an apparent lack of sensitivity to fare increases related distance traveled. Cummings,C.P.et al(1989) used these results and market segmentation, together with historical analysis of ridership response to three recent CTA fare increases, several elasticities are derived for use in systematically analyzing potential ridership response to future CTA fare structure options. In market research, Rebecca (1998a ) focuses on the statues of market research as practiced in transit agencies and identifies major market issues confronting them, evaluating market research strategies appropriate for transit and provides guidance to integrate and institutionalize market research into decision-making processes of transit agencies, analyzing some institutional barriers to limit the use market research . Jakubiak et al (1990) used market research to improve management of transportation system. It is designed to aid officials at many levels with a transport agencies.They presented techniques of modern market research, including data collection( focus group, panel surveys, and telephone, mail and intercept surveys), data analysis (cross-tabulations, importance-performance analysis ) and multivariate statistics( regression analysis, factor analysis , discriminant analysis, multidimensional scaling, psychographic market segmentation, perceptual mapping and conjoint analysis). Walb and Booth (1985) reviewed current marketing practice at representative transit agencies in North America, revealing a number of promising new practices including market segmentation, target marketing, direct contact marketing, electronic user information aids and consumer orientation training. The role of marketing in the transit industry has undergone numerous changes in the past decade, expanding as the task of attracting new riders or retaining existing riders has become more complicated. The complexity of the marketing function encompasses market reserach, service development, pricing, promotion (radio and televison advertising, use of car cards and billboards), consumer aids( schedules and timetables, telephone information system and system maps) and evaluation. In travel pattern and behavior, Roberts, R.A. (1986) predicted the future of the Portland transit market using demographic trends. Demographic changes affect the the market for urban transportation in many ways. Identifying theses changes and monitoring demographic trends can give the planners better insight into the future nature of the transit market. An inductive approach is used to examine some of the demographic variables most important to the transit market. Then the market segmentation that exists in the Portland standard metropolitian statistical area is identified. Tischer, M.L. (1983) summaried alternative analytical approaches to the traditional planning and modeling process. Survey strategies and market segmentation are reasonably well utilized in practice , and attitudinal analyses have been performed in a few transit , carpool and pedestrian planning studies. Travel behavior analysis techniques are most suited in strategic planning and project planning. Blankenship, D.P. (1976) conducted segmentation analysis of transit users and nonusers. A stratified random sample of households was used to ensure accurate spatial representation. Market segmentation was used to ascertain the attitudinal framework. The survery identified various demographic and psychographic profiles that may impact current, future, private and public transportation modes. The factors were concerned in the survey including social status; funding public transit; psychological automobile captivity to transit dependency; vehicle availabilityand trip market profiles; conceptual awereness and anticipated use; anticipated demand for public rapid transit; marketing and advertising analysis. In transport planning, Trombly, J.W(1986) reviewed the use of market segmentation analysis in transportation energy planning. Two statewide telephone surveys completed to identify groups sharing a similar response to energy price increases and supply short falls.These responses were then used to construct matrices of interitem correlations. The action factors contained shop travel,nonautomobile mode choice, vehicle replacement, automobile maintenance, moving nonautomobile work travel, changing leisure activity and cancelling a vacation. The results yielded a consistent set of factors across survey years that are use in a cluster analysis algorithm to identify segment sharing a similar response to conservation. The results of the study can be used to design programs and policies to meet future planning needs. Nicholaidis,G.C.et al(1977) described a comparison of market segmentation using six different basis--two based on demographic variables, two on travel choices constraints, and two on attitudinal variables. The six segmentations were compared with respect of five criteria judged to be important considerations in transportation planning: measurabilty, statistical robustness, substainability, relation to travel behavior , and relation to planning of service options.The results showed that no single segmentation base was superior. In mode choices , Dehghani and Talvitie (1980) using market segmentation to identify following segments: one-versus two car houshoulds; commuters bound for the central business district(CBD) versus others; and low-and high-income households that also value service attributes unequally. A simplistic method is used to aggregate rail and bus modes in five-and four-alternative models, respectively. Finally, coefficients estimated by means of obervation of level-of-service attributes or by means of multinomial logit models that estimate these attributes are compared. Stopher,P.R.(1977) discussed the development of individual-choice models of the destination choice of nongrocery shopping locations. Two key features of the approach are used of perceptual data for characterizing alternative destinations and attempt to segment the population before the model building on the basis of homogeneity of perceptions of destinations. Data were obtained about the perceptions of shoppers of several shopping locations and on their preferences for various attributes of shopping locations. The attributes were selected as those that make up the image of a shopping location independent of the transportation system. Several techniques are discussed for segmenting the population by perception, all of which are based on analysis of the psychological distance between shopping locations. Given the special properties of psychological distance, two forms of were undertaken. First, correlations were computed for the set of interpoint distances for each socioeconomic group identified in the data. Second, the group interpoint distances were used as inputs to an individual scaling process . Market segmentation was sought through a hierarchic, fusion clustering process on the axis weights for each socioeconomic group. The results of these analyses converge well. Lengh of residence and age were found to be important segmentation variables. Sex and income were not found to be very powerful segmentation variables, but occupation may be worth study as a basis for segmentation. In travel-demand analysis, Salomon ,I et al (1982) suggested the employment of the life-style concept as an improved basis for segmentation . Lifestyle is defined as the behavioral pattern that results from three major life decision: the decision to form a household, the decision to participate in the labor force, and the orientation toward leisure. By using available socioeconomic variables, an attempt is made to identify lifestyle groups and to use them as market segments in a joint mode and destiantion choices model. Two tests were presented. One is the use of lifestyle specific variables in the model specification ; and the other is the estimation of separate models for each market segment. Both approaches have shown an improvement in the model performance compared with either a pooled model or an income-base and a life-cycle/ occupation-based segmentation. Dunbar,F.C. (1977) reported that disaggregate travel-demand models can be applied to market segments. The method leads to improved accurancy of travel forecasts because appreciation bias is reduced. The work-trip modal split is simulated on National Personal Transporation Survey data. Travel demand is forecast under a variety of transportation policies that involve automobile controls and transit level-of-service improvement. In ridership, Rebecca, E.Y (1998b) provided an overview of market segmentation--what it is and why is relevant to public transit agencies. It serves as an introduction for managers to the basic concepts and approaches of market segmentation and provides steps and procedures for marketers or market researchers who have the responsibilty for implementing a market segmentation program to increase transit ridership. He emphasized on the problems likely to arise in implementing segmentation studies, ways to encourage the use of market segmentation at transit agencies, and methods to translate segmentation findings into strategy. Private and public sectors examples of market segmentation analyses are extensively used. Consequently, a great amount of literature on market segmentation are related to labor market,production,trade,tourism and environment assessment,etc. And few are involved in transport sector, including transport,fare elasticity,transit market research, transport planning, mode choice, demand analysis and ridership.4 Theoretical Approach4.1 Definition of market segmentation Different groups of people depending on their socioeconomic and locational characteristics will make different trade-off among level-of-service characteristics. Each market segment can be expected to behave differently from other market segments. Therefore, characteristics of unique market segments must be identified , such as income, automobile ownership, trip purpose, life cycle (age, family status), occupation, transportation affecting handicaps, and accessibility, etc. So called market segmentation refers to the identification of groups of customers-- or market segments-- that have similarities in characteristics or similarities in needs who are likely to exhibit similar purchase behavior and/or responses to changes in the marketing mix(Rebecca,1998) . Or market segmentation is the division of a total population of travelers into groups(segments) that are relatively homogeneous with respect to certain personal characteristics ( the segmentation base)( Nicolaidis, 1977) Market segments consist of groups of people or organizations that are similar in terms of how they respond to a particular marketing mix or in other ways that are meaningful for marketing planning purpose. Marketing mix is a set of product, price, postion and promotion, and can be briefly considered ‘4Ps’.4.2 Component of market segmentation Market segmentation can be classified into 4 categories--Physical attribute segmentation,product usage segmentation, psychographic segmentation and benefits or needs segmentation.· The physical segmentation includes demographic, socioeconomic and geographic segmentation, which provides important information about individuals with specific markets. The segmentation is easy to collect data and less expensive than other forms of segmentation research, but failed to explain consumption behaviors.· The product usage segmentation is based on usage rates.Usage rates are one the simplest basis variables used for segmentation efforts. The segmentation is particularly applicable to transit. But it is often determined on a priori basis, and leading to inherent organizational biases.The solution is to use a post hoc approach--product related attitudinal segmentation, however, it is time-consuming and costly.· The psychographic segmenattion seeks to describe human characteristics of consumers they may respond to products. Psychographic variables include product attributes, lifestyle and psychology, so it can be viewed as the practical application of behavioral and social science to marketing research. It can identify target market and provide an insight into consumer mind to feed back for market changes.But it is difficult to accurately quantify indicators, and costly and complicated.· Benefits or needs segmentation is based on benefits, which are the sum of product advantages or satisfactions meeting an individual抯 needs or wants. The benefits segmentation relies upon causal factors rather than descriptive factors. But it is difficult in data collection and analysis. Its cost is higher.4.3.Market segmentation and rapid rail transit· physical segmentation Physical segmentation consists of demographic, geographic and socioeconomic segmentaion. These segmentations are closely related to increasing ridership. The emphases will be placed on geographic segmentaion,connecting with rail transit-based land development and urban land changes.· Product usage segmentation Product usage segmentation can translated into ridership frequency,dividing non, light, medium and heavy riders. The segmentation can be conducted through on-board survey .· phychographic and benefit segmentation The segmentation is applied to transport field using the theory of behavior science.It is linked to riders’ behavior, such as lifestyle or lifecycle, revealing the response to transit marketing strategies for riders. The kind of segmentation is difficult in data collection and more expensive. Due to limits to time and cost, it will become future in-depth research topic. 5. Research Hypothesis and Objectives· Increasing ridership for RRT system through market segmentation;· Application to uban areas of developing countries6.Methodology6.1 Comparative Study Selecting Montreal, Toronto, San Francisco, Singapore and H.K. as case cities, their characteristics, location, organization, market segmentation and ridership are addressed. Emphases will be placed on geographic segmentation and level-of- service.6.1.1 Montreal ,Canada· demographics: riders(adult,student,senior,disabled)· geographics: zoning control;5 metro lines,65 stations,64 km, full subway underground city, integrating of buildings with metro exit· socioeconomics: integrated fare system in STCUM, transfer ticket, magnetic card reader machine· product usage: quality service, amenity of art with architecture in station, rubber tire,bike accessible· benefits: routing and scheduling,safe, comfortable and reliable· ridership: 1,230,000/weekday, annual rate: +12%, metro share 58%,revenue/cost: 46.5% Table 5. Fare and rates in Montrea
Local Fare Regular Fare Reduced Fare
CAM Pass $45 $19Weekly CAM $12Tickets 6/$8 6/$4Cash fare $1.85 $1.0Tourist Pass $5/1day -- $12/3days --Regional fare $73 $40(STCUM,STL,STRSM)6.1.2 Toronto, Canada· demographics: rider(adult,student,senior,disabled)· geographics: zoning control;2 lines, 66 stations,56.9km, underground pedestrian, transit-based property joint development· socioeconomics: integrated fare system in TTC,transfer ticket, magnetic card reader machine· product usage: quality service,network: RRT(640 vehicles) + LRT(28) + Bus(1701)+streetcar(248),park-and-ride· benefits: safety measures (transit community watch, public telephone, security mirror, designated waiting areas and passenger assistance alarms), cost and timesaving· ridership: 994,000/weekday(1986), per captia trips :186/a, public transit share:38%,revenue/cost=68% Table 6 Fare and rates in Toronto Item Fares and Rates Memo
Adult: cash fare $2.00 tickets/tokens 10/$16.00 50/$80 metropass(monthly) $83.00 metropass discount $76.00Students/Senior Citizen: cash fare $1.35 tickets 10/$10.70 metropass(monthly) $73.00 metropass discount $67.00Child: less than 24 month: free cash fare $0.50 tickets 10/$4.00 2-12 yearsDay/Family Pass $6.50(less than 6 persons, adult 2or less) GTA Weekly Pass $35.00 6.1.3 San Francisco, USA· demographics: rider(adult,young,senior,disabled)· geographics: regional control,4 lines,115 km(40.6 km surface,37 km aerial and 37.4 km underground),37 stations (10 surface, 13 aerial and 14 underground), slight impact on land use· socioeconomics: independent fare system between BART and MUNI · product usage: quality service, modern train and cars(680 vehicles), transbay tube,Network=RRT+LRT+Bus+Trolley+ treetcar,park-and-ride(free charge for parking, bike and motocycle accessible), bus feeder service· benefits: safe, convenient, time and cost saving· ridership:250,000/weekday. 32 million(1975),58.9 million(1986) and 75 million (1998,estimated)Table 7 Fare and rates In San Francisco
Types Rates
New High-value Ticket $48 ($3 discounted)BART Blue Ticket (B) $32 ($2 discounted)BART Red Ticket(R) $4 ($12 discounted for the disabled and 5-12-year child)BART Green Ticket(G) $4($12 discounted for the seniors more than 65 years)BART Plus $28-61 (8 different values on zones, the tickets are integrated between BART and Muni. Expiray is 2 weeks from 1st-15th, or 16th - the end day in the month)BART Express Bus Ticket (EB) ( It is good for 20 riders, $19 for zone1 and $30 for zone2)Notes: BART tickets are available from a variety of Tickets-to-Go Vendors and the Lake-Merritt, Montogemery Street and Embarcadero Stations or by mail.6.1.4 Hong Kong, China· demographics: rider(adult,young,senior,disabled)geographics: physical planning, 4 lines(KwanTong,TsuenWan, Island and TungChung),43.2km,38 stations. Transit-based property joint development:H.K.station(415,000sq m), Kowloon station (1.09 million sq m),Olympic station(628,000sq m), Tsing Yi station (291,870 sq m) and Tung chung 843,120 sq.m)· socioeconomics: independent fare system,metro fare:HK$4-26, Octopus ticketing system(electronic cash and ticket system)· product usage: quality service, network = RT (43.2km) +LRT(32km) +Tramways (16.3 km)+Bus (3 main companies-KMB,CMB and Citybus) + Minibus (4350 vehicles, private)+ Peak Cable Tram+ Ferry.· benefits: safe, reliable,frequent, comfortable and convenient· ridership: 2.38 million/weekday, 817 million passengers(1997),10 million /weeday by public transit. RRTRevenue/cost: +100%! Fare revenue HK$5,078 million,other recurring revenue HK$1,093 million, net profit HK$1,535 million (1996) Table 8 Fare and rates in Hong Kong (HK$)
Category single-ride ticket(discount) Octopus card (discount)
trip on HK ,Kowloon 4.0 3.0 3.7 2.3 5.0 3.0 4.5 2.3 6.0 3.0 5.5 2.8 7.5 4.0 6.7 3.3 9.0 4.0 8.1 3.6cross harbour 9.0 4.5 7.8 3.9 11.0 5.5 9.9 5.0 13.0 6.0 11.7 5.6
Note:adult fare: HK$4.00-26.00; concessionary fare:HK$3-13 for elderly, chilren and students.Ticket Issuing machines sell single journey ticket and Octopus cards.6.1.5 Singapore· demongraphics; rider(adult,young,senior,disabled)· geographics: integrating phyical land planning with transport;2 lines, 48 stations (15 underground, 32 elevated and 1 surface),83 km( 19 km underground, 60.2 km elevated and 3.8 km surface)· socioeconomics: flexible fare system(monthly concession ticket,TransitLink Giro Farecard, Commemorative Ticket,Tourist Souvenir Ticket, TransitLink Park&ride ticket, Thematic Edition Tichet, Single trip Ticket and TransitLink Farecard), automatic fare collection (Automatic Fare collection Gates and Ticket vending machines)· product usage: trains(85 vehicles), 510 cars,average speed 45 pkh.· benefits: safe, reliable ,comfortable, affordable and user-friendly· ridership: 560,000 riders/weekday, and 1,673,000 riders/month. RRT share:62%. In 96/97抯, Fare revenue improved by 5.8%, ridership grew by 5.7% over the preceding year.6.2 The Case of Guangzhou6.2.1 Geography of Guangzhou Guangzhou is typical metropolis in China, located in gateway adjacent to Hong Kong and Macau, powerful and vital in economy. The economic power was ranked the 3rd following Shanghai and Beijing. In 1998, urban econonic growth was the first in top cities in China, the rate being 13%, while provincial and national average was 8.5 and 7.8%,respectively. The metropolitan area has 6.8 million population.Its goal is to be built into an international metropolis.The metroline 1 will open in early 1999, other rail transit lines are under construction or projected, and will have been completed by early 21st century. The experience and lessons in RRT from Guangzhou will be constructive and suggestive to other cities, such as Shenzhen(14.5 km), Wuhan, Qingdao, Congqing, Shenyang, as well as cities in developing countries.6.2.2 Increasing Riders Movement or Initiatives The city has planned or is under construction new transit facilities, such as RRT, LRT, urban expressway, ring freeways, transriver bridges; new international airport transit oriented land development and redevelopment; The major goal is to relieve traffic congestion and increase the velocity of passengers/freight flows,reducing trip time. In operation, public transit priority is being implemented and great shares of transport investment are pooling in for the 9th National Game to be hold. At the same time, high-speed economic growth strive to be maintained and 13% annually will be new GDP goal in 1999, according to latest news from Guangzhou Daily(1999.1.29). 6.3 Sample Survey The survey list or questionnaire will be tailored to adapt for Guangzhou case .Sample survey and field work can help to make in-depth research.The data wil be collected and obtained with the assistance of Guangzhou Metro Co., Guangzhou Real Estate Information Centre, Guangzhou No 2 Bus Co. and intercepting questionnaires by undergradutes.The collected data will be processed and analyzed with statistical tools and modelling to reveal the interrelationships between major factors.7. Conclusion and Suggestion Some strategies of increasing ridership for RRT system can be concluded as following:· Increasing utilization rates· Increasing customer retention· Attracting new and infrequent rider or potential users· Reducing cost by equipment domestization · Multiple use of fare card and rational fare structures· Increasing productivity· Upgrading level of service in reliability, safety, comfortable, countesy and affordability.· Providing special service for elderly and hanicapped· Creating the amenity of transit facilities with artistic style and oriental culture· Optimizing RRT network and bus feeder system· Constructing bike/motorcycle-and-ride, facilities· Marketing mix stategies· Educating and trainingReferencesAlan, A. W. 1986,Urban transit system: guidines for examing options. Washington:The World Bank Allport, R. 1997, Investment in mass rapid transit.In Stares, S and Liu Zhi (ed) .China urban transport development srategy. Washington: The Word Bank Allport, R.J.et al. 1990,Study of mass rapid transit in developing countries. CR188. U.K. Transport and Road Research Laboratory, 212ppAndersen,V.et al ,1997, Imagery of Demark among visitors to Danish fine arts exhibitions in Scotland. Tourism Management, 18(7),453-464.Anker,R. 1997,Theories of occupational segementation by sex: an overview. International Labour Review, 136(3),315-339BART website, 1998,WWW.trnasitinfo.org/BART/ --Bhattacharjee,D.et al . Commuters’ attitudes toward travel demand management in Bangkok. Transpotation ResearchBlankenship,D.P.1976, Segmentation analysis of transit users and nonusers.TRR 590, p.1-4, US Transportation Research Board(TRB)Blunden, W.R. 1971,The land use / transportation system. Oxford: Pergamon Press. Boorse, J. W.1968, Rapid transit in Canada. Philadelphia:Almo Press Boyce, D.E. et al .1972, Impact of rapid transit on suburban residential property values and land development. US Department of Commerce . Boyle,D.K. 1985,Are transit riders becoming less sensitive to fare increases? NewYork DOT. Planning Division,Report 183. Brown, H.J. et al .1972, Empirical models of urban land use: suggestion on research objectives and organization. New York: Columbia University Press Cambridge Systematics Inc.1994, Transit ridership initiatives. TCRP J-6.TRB.Capo, C.K. and Messmer, D. 1987, Effects of disseminating service information and free ride coupons on bus ridership. TRR 1144, TRBCervero, R.1996, Mixed land use and commuting :evidence from the American housing survey. transportation Research.A. 30(5), 361-377 Cervero,R.et al.1997, Twenty years of the bay area rapid transit system: land use and development. Transportation Research A. 31(4), 309-333 Cervero,R. 1993,Ridership impacts of transit-focused development in California. Working paper 176,UCTC. The University of California, Transportation CenterCha,S. et al . 1995,Travel moyivation of Japanese overseas travelers: a factor cluster segmentation approach. J. of Travel Research, 34(1), 33-39.Charles River Assocites,1997, Building transit ridership--An exploration of transit market share and the public policies that influence it.TCRP27, TRBCumming,C.P. et al .1989,Markety segmentation of trasnit fare elasticities. Transportation Quarterly, 43(3), 407-420Curtis, C.et al .1997,Targeting travel awareness campaigns. transport policy. 4(1), 57- 65CUTA. 1985,Canadian transit handbook. Canadian urban transit Association and the Road and Transportation Association of canada Dagang, D.A.1993,Transportation demand mangement cost-effectiveness model for suburban employers. TRR 1404. TRB. 64-72Davies, G.W.1974, The effect of a subway on the spatial distribution of population . Dept Economics, The University of Western Ontario Dehghani,Y.et al.1980,Model specification , modal aggregation and market segmentation in mode chois models: some empirical evidence. TRR 755, p28-34, TRB.Dobson,R. et al. 1978,Perceptual market segmentation techniques for transportation analysis. TRR 673, 145-152, TRBDodsworth, J.et al ,1997,Hong Kong, China, growth structure change and economic stability during the transition. Washington: International Monetary Fund Dunbar,F.C. 1997,Policy-contigent travel forecasting with market segmentation. TRR 637, 27-32, TRBEdwards, M.et al.1996, Developing new urban public transit systems. transport Policy. 3(4), 225-239 Else,P. 1996,Subsidy requiremens in a restructured rail network. Transpot Policy, 3(12).13-15Erem,T. et al.1997, Export market segmentation practices of Turkish firms. J. of Euromarketing, 6(3), 103-135.Fouracre, P.R.et al. 1990,The performance and impact of rail mass transit in developing countries. Research report of U.K.Transport and Road Research Laboratory, RR278,28ppFunchouse, E. 1997,Mobility and labor market segmentation: the urban labor market in EL Salvador. Economic Development and Cultural Change. 46(1). 123-153Giampietro, G.D. 1989,The relationship between transportation investments and the land use changes. CRT615, Universite de Montreal Grimshaw,D.et al. 1998,Integrating the internal and external labour market. Cambridge J. of Economics, 22(2).199-220.Hall, P.et al. 1985,Can rail save the city?the impacts of rail rapid transit and pedestrianisation on British and German cities. Gower Publishing Company . Hall, R. W. 1987, Quick estimation of queueing delay for passengers exiting a rapid transit statiion,TRR 1152, TRB, p.11-13.Harmatuck. D.J. 1986,The seasonability of urban transit ridership.Proceedings of 27th Annual Meetings, Canadian Transportation Reserach Forum, Sept.22-24, Seattle. 323-31.Hamer, A.M. 1971,The selling of rail rapid transit. Lexing: D.C. Heath and Company. Harral, C.G.et al. 1992,Transport development in southern China. Washington: The World Bank Harris, N.G.et al. 1997,Railway profitability and station closures. Transport Policy.4(1), 41-47 Harris,R and Lewis,R. 1998,How the past matters: North American cities in the twentith century. J. Urban Affairs, 20(2), 159-174 Herny, L. 1989, Ridership forecasting considerartion in comparisons of light rail and motor bus modes. Special report 221,TRB p.163-189.Hill, P.et al .1994,Transport, energy and the environment in Asia. Hong Kong:The Unversity of Hong KongHoesli,M.et al. 1997,A hedonic investigation of the rental value of apartments in central Bordeaux. J. of Property Research, 14(1).15-26.Howenstine, E. 1993,Market segmentation for recycling. Environmental Behavior. 25(1).86-102.Hurst, M.E.E. 1974,Transportation geography. McGraw-Hill Industry Canada. 1997,Impact of China抯 trade and foreign investment reforms on the world economyJakubiak,S.et al.1990, Using market research to improve management of transportation systems. NCHRP 329, 81p. TRB.Jessiman, W.A.et al.1975, Attracting LRT ridership. In TRR 161.Light rail transit. TRB. 126-46.Kanninen, B. J.et al . 1996,Intelligent transportation systems: an economic and environmental policy assessment. Transportation Research, 30(1), 1-10Kemp, M. A. 1974,Transit improvements in Atlanta: the effects of fare and service changes. Report 1212-2-1. The Urban Institute. Kerr, J.W. 1983,Illustrated treasury to rail rapid transit systems and cars of North America. DPA-LTA Publication Khattak,A.J.et al.1997, The impact of adverse weather conditions on the propensity to change travel decisions: a survey of Brussels commuters. Transportation Research.A ,31(3). 181-203 Kinnucan,H.W. 1997,Marketing research paradigms for aquaculture. Aquaculture Economics and Management. 14(1).15-26.Kockelman, K.K. 1997,Travel behavior as function of accessibility, land use mixing and land use balance: evidence from San Francisco Bay Area. TRR 1607. TRB. 116-24.Kraft,G.et al. 1971,The role of transportation in regional economic development. Lexington.: D.C. Heath and CompanyLang, A.S.et al. 1964,Urban rail transit: its economics and technology. Cambridge: The MIT Press Laquian,A.A. 1996,Pearl river delta development in the world perspective. In A.G. Yeh(ed.). Planning H.K. for the 21st century. University of H.K. p11-36Leontaridi, M.R. 1998,Segmented labor markets: theory and evidence. J. of Economic Surveys. 12(1). 63-101.Liu, Kang. 1998,Global cities above million population surpassing 300! Geographic Knowledge, No.9 p71-3(in Chinese)Lowe, J. C.et al ,1975,The geography of movement. Boston: Houghon Mifflin CompanyMcCord, M.R. and Cheng, L.H. 1986, Day of week and part of month variation in bus ridership:empirical result.TRBMichelle,F.R. 1987, Increasing transit ridership through a targeted transit marketing approach. Ph D Thesis, Rutgers the State University of New Jersey.Milan, J. 1996,The trans European railway network: three levels of service for the passengers . Transport Policy. 3(3) 99-104Montreal website. 1998,WWW.STCUM.qc.caNicolaidis,G.C.1997, Evaluation of alternative market segmentations for transporattion planning. TRR 649.p23-31 , TRBNock, O.S. 1973,Underground railways of the world : NewYork: St Martin's PressOrr, D.V. 1977,An index of segmentation in local markets. International Review of Applied Economics. 11(2):229-247Oram. R.L.et al. 1996,Infrequent riders: one key to new trasnit ridership and revenue. TRR 1521.TRB.37-42O’sullivan, P.1980,Transport Policy. Barnes &Noble Books Parkinson,T.E. 1970,Passenger transport in canadian urban areas. Canadian Transport Commission Parkinson,T.E.1989,Advocacy for conventional light rail.Special report 221,TRB.Pendyala, R.M.et al.1997, An activity -based microsimulation analysis of transportation control meatures. Transport Policy. 4(3). 183-192 Pendakur,V.S.1992, Urban transport in China:Trends and issues.TRR1372, TRB pp3-10 Pesson, J.et al . 1995,The transportation and land use interaction in North York: a policy perspective. J.Andrey (ed.) Transport and policy issues. Canadian Association of Geographers Pepy, P.G. 1998,The train:simpler and cheaper. Transport. May-Jun, People Daily (oversea version), 1998.12.10 (in Chinese)Pickrell ,D.H. 1990,Forecast versus actual ridership and cost. US DOT.Report Purdy, H.L. 1972,Transport competition and public policy in Canada. University of British Columbia Press Pushkarev, B.S.et al. 1982,Urban rail in America. Indiana University pressRabinowich, B.et al.1980,Transit ridership and cost models for N.Y. State: How accurate were they? NewYork :DOT. Planning Division,Report 183. Rebecca.E.Y. 1998,A handbook: using market segmentation to increase transit ridership. TCRP 36. TRB Rebecca, E.Y. 1998, A handbook:Integrating market research into transit management. TCRP. 37. TRBResenbloom,S. 1998,Transit markets of the future: the challenge of change.TRB Roberts, R.A. 1986, Analysis of demographic trends and travel patterns: implications for the future of the Portland transit market. TRR 1067, 1-8. TRBRock, S.M. 1979, The redistributive effects of reduced transit fares for the elderly. Transportation Research.A. 13: 361-365.Rodrigue,J.P.etal .Transport Geography. www.ubishops.ca/geotrans,1998San francisco: 1998,www.ci.sf.ca.us/muni Salomon, I. et al.1982, Life-style segmentation in travel demand analysis. TRR 879. 37-45, TRBSavage, I. 1997, Scale economies in the United States rail transit system. TransportationResearch. A31(6). 459-473 Schumann,J.W. 1989, What抯 new in North American LRT projects? Special report 221,TRBSheauhsing, H. et al. 1992,Segmenting the international travel market by activity. Tourism Management , 13(2). 209-223.Simon, P.V. 1990,Transport and the development of the European economy 1750- 1918. The Macmillan Press Smith, P.N.et al. 1994, A method for the rationization of a suburban railway network. Transportation Research. A. 28(2). 93-107 Stares, S. and Liu zhi .1996, China's urban transport development strategy. Washington: The World Bank Steininger,k. et al .1996,Car-sharing organizations: the size of the market segment and revealed change in mobility behavior. Transport Policy, 3/4, 177-185.Stopher, P.R. 1997,Development of market segmentation of destination choices. TRR 649. 14-22. TRB.Stover, V.G.et al. 1988,Transportation and land development .Washington: Institute of Transportation Engineers SYSTAN, Inc. 1983,Timed transfer: an evaluation of its structure, performance and cost. U.S. Department of Commerce TCRP Report 13.1996, Rail transit capacity. TRB Tischer, M.L. 1983, Workshop on travel behavior characteristics ansd analysis: workshop summary. TRR 202. 72-73 . TRBToh, R.S. et al. 1990,A multiple discriminant approach to identifying frequent fliers in airline travel: some implications for market seegmentation, target marketing and product differentiation. Logistics and Transportation Review, 26 (2). 179- 197.Toronto Website. 1998,WWW.city.toronto.on.ca/ ttc Transportation Reserach Circular.1994,No.432. Workshop on safety Research Related to High-speed Rail and Maglev Passenger System. TRB TRB et al.1992, National Conference on Advanced Technologies in Public Transportation.Aug.16-19. San Francisco.TRB ,1975,Report 161. Light rail transit. Washington: Transportation Research BoardTRB ,1977,Report 182. Light rail transit: planning and technology. Washington: Transportation Research Board TRB,1997,Report 1607. Transportation forecasting and travel behavior. TRB TRB,1996,Report 1554. Advanced traffic management systems and high- occupancy-vehicle systems. TRB Trombly, J.W. 1986, An investigation into the use of market segmentation analysis for transportation energy planning. TRR 1092. 39-47 , TRBUS DOT Report. 1992,An evaluation of making rail transit system accessible to handicapped persons: a national summary of cost esimatesVega, K,et al . 1981,Impact of transfers on transit ridership. Annual Conference Preprints.Vol.1 Planning. Roads and Transportation Association of canada. D3-22Vuchic,V.R. et al.1975,Rail transit: characteristics,innovations and trends.TRR 552,TRBWalb, C. et al. 1985, Transit marketing: the state of the art. TRR 1039.TRB, P.9-16Walmsley, D.A.et al. 1992,The effects of rapid transit on public transport and urban development.London: HMSO Press Waldo,L.A. 1993, Conceptual framework to study the effectiveness of employer trip- reduction programs.TRR 1404. TRB,55-63Wells,S.S. 1994, Analysis of the differential impacts of transport modes on travel behaviour in greater Toronto area. Ph D thesis, University of Waterloo.Wingo, L. 1961,Transportation and urban land. Washington: Resource for the Future, IncWinter, N.A. 1997,Rual taxation in Ethiopia, 1981-1989: a policy analysis matrix assessment for net producers and net consumers. Food Policy. 22(25), 419-431.Wu, Jiahao.1998,Urban transportation planning methods in China. Proceedings of ICTTS’98. Beijing. P275-89Xu,X.Q. et al, 1995, The changing urban system of China: new development since 1978. Urban Geography, 16(6):493-504.Zhang,yunquan,1999,Translating market segmentation into increasing rail transit ridership: methodology and application. Working paper. Zhang,Yunquan. 1998, Market segmentation and its application in transit industry. working paper Zhang ,Yunquan. 1998,Some measures to increase rail transit ridership: case studies. Working paper Zhang, Yunquan,1998, A brief review of literatures on rapid rail transit ans urban land use, Working paper