UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Suburb-to-suburb commuting and transit planning : a case study of Surrey, B.C. Murray, Peter S. 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1993_spring_murray_peter.pdf [ 5.84MB ]
JSON: 831-1.0086141.json
JSON-LD: 831-1.0086141-ld.json
RDF/XML (Pretty): 831-1.0086141-rdf.xml
RDF/JSON: 831-1.0086141-rdf.json
Turtle: 831-1.0086141-turtle.txt
N-Triples: 831-1.0086141-rdf-ntriples.txt
Original Record: 831-1.0086141-source.json
Full Text

Full Text

SUBURB-TO-SUBURB COMMUTING AND TRANSIT PLANNING:A CASE STUDY OF SURREY, B.C.byPETER S. MURRAYB.A., The University of Victoria, 1991A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF ARTS (PLANNING)inTHE FACULTY OF GRADUATE STUDIESSchool of Community and Regional PlanningWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAApril 1993© Peter S. Murray, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature) Department of^rr,The University of British ColumbiaVancouver, CanadaDate /7,DE-6 (2/88)11ABSTRACTRapid growth in suburb-to-suburb commuting has created a problem for transit providers: thedispersed commuting patterns are very difficult to serve with transit, and are characterized by lowtransit use. This thesis aims to determine which markets have the best potential for transit, and whatfactors could increase this potential.Surrey, B.C. is typical of the rapidly growing areas where suburb-to-suburb commuting ismost prevalent. Commuting between Surrey and other suburban areas has increased sharply in recentyears. A detailed examination of commuting patterns within Surrey revealed the highly dispersednature of the work trip flows; the only flows which were concentrated to any degree were thosebetween nodes with relatively high population and employment densities. A correlation was foundbetween density, especially employment density, and transit use. Inter-nodal trips, which alreadyhave the greatest transit use among suburb-to-suburb trips, will be a key market for transit in thesuburbs. Inter-nodal express service would help to address complaints that suburb-to-suburb transitservice is too slow and indirect. Trips to and from the nodes will also be an important market. Intra-nodal trips, which presently have low transit use, form another key market which could possibly beserved by a paratransit shuttle service. In Surrey, efforts have begun to address the issue of suburb-to-suburb transit in a comprehensive manner, but there has been little substantive progress to date.The case study results were used to develop a conceptual framework for suburb-to-suburbtransit planning which could then be applied to other suburban areas facing similar problems. Theframework calls for a wide array of transit and paratransit services, each filling a different marketniche, which can be combined to create an integrated but flexible system. This system must bereinforced with land use strategies to promote greater densities, and more pedestrian and transitfriendly design. Transportation demand management must also be used to encourage transit use byincreasing the costs of driving an automobile. This three-pronged, comprehensive approach shouldallow transit to compete successfully in some suburban markets.TABLE OF CONTENTSABSTRACTTABLE OF CONTENTSLIST OF TABLESLIST OF FIGURES^ viGLOSSARY viiACKNOWLEDGEMENTS^ viiiCHAPTER 1. INTRODUCTION 11.1 Problem Statement^ 11.2 Purpose^ 41.3 Methodology 61.4 Assumptions and Limitations^ 61.5 Thesis Organization^ 7CHAPTER 2. LAND USE AND PUBLIC TRANSIT^ 92.1 The Relationship Between Transit and Land Use 92.2 Land Use Planning^ 132.3 Suburban Nodes 152.4 Summary^ 19CHAPTER 3. STRATEGIES FOR INCREASING SUBURBAN TRANSIT USE^203.1 Transit's Challenge in the Suburbs^ 203.2 Suburban Transit Strategies 223.3 Paratransit^ 253.4 Transportation Demand Management^ 273.5 Ending Automobile Dependency 323.6 Summary^ 33CHAPTER 4. CASE STUDY: BACKGROUND INFORMATION^ 344.1 Introduction^ 344.2 Surrey and Greater Vancouver^ 354.3 Urban Form 374.4 The Transportation System 414.4.1 Roads^ 414.4.2 Public Transit^ 434.5 Changing Commuting Patterns in Greater Vancouver^ 454.6 Summary^ 51iiiivCHAPTER 5. CASE STUDY: RESULTS^ 525.1 Introduction and Data Limitations 525.2 Trends in Surrey Commuting Patterns^ 545.3 Commuting Patterns: Traffic Zone Level 595.4 Nodal Patterns^ 625.4.1 Internal Surrey Trips^ 655.4.2 Trips Between Surrey and Other Suburbs^ 675.5 Effect of Density on Transit Use 705.6 Surrey in 1993: An Update^ 745.7 Summary^ 74CHAPTER 6. CASE STUDY: THE RESPONSE IN SURREY^ 756.1 Introduction^ 756.2 The Response by the Municipality^ 756.3 The Response by BC Transit 786.4 Public Attitudes Towards Transit Service 846.5 Summary^ 88CHAPTER 7. THE PLANNING FRAMEWORK^ 897.1 Introduction^ 897.2 Strategies for Suburb-to-Suburb Transit^ 897.2.1 Conventional Transit^ 907.2.2 Paratransit^ 917.2.3 Express Service 927.2.4 Land Use Strategies^ 937.2.5 Transportation Demand Management^ 957.3 A Conceptual Framework for Suburb-to-Suburb Transit Planning^967.3.1 Examples of Planning Frameworks 967.3.2 Outline of the Planning Framework^ 987.4 Summary^ 101CHAPTER 8. CONCLUSIONS^ 1028.1 Introduction 1028.2 Planning Implications^ 1028.3 Recommendations for Further Study^ 1058.4 Summary^ 106BIBLIOGRAPHY 107LIST OF TABLES1.1 Summary of U.S. Metropolitan Area Work Trips, 1960-80^ 32.1 Transit Modes Related to Residential Density^ 114.1 Population Growth in Surrey, 1951-91 354.2 Changes in Inter-Municipal Commuting Flows, 1971-81^ 484.3 Commuting Flows in Vancouver CMA, 1971-81 495.1 Commuting Flows for Surrey Residents, 1971-81^ 555.2 Commuting Flows by Mode for Surrey Residents, 1985 555.3 Commuting Flows for Surrey Workers, 1971-81^ 575.4 Commuting Flows by Mode for Surrey Workers, 1985 575.5 Commuting Balance in Surrey, 1981^ 595.6 Summary of Surrey Commuting Flows 605.7 Size Distribution of Commuting Flows Between Surrey TZ Pairs^ 625.8 TZs Ranked by Total Suburban Flows^ 635.9 Summary of Surrey Town Centre Characteristics^ 655.10 Nodal Commuting Flows Within Surrey 665.11 Commuting Flows Between the Town Centres^ 675.12 Commuting Flows Between the Nodes and Other Suburbs^ 675.13 Commuting Flows From Other Suburbs to Surrey 685.14 Commuting Flows From Surrey to Other Suburbs^ 687.1 Transit Options For Different Commute Types 90VLIST OF FIGURES1.1 Transit Mode Split by Type of Commute^ 32.1 Population Density and Transit Use 132.2 Examples of Transit Oriented Design^ 184.1 Vancouver CMA and its Component Sub-Regions^ 364.2 Surrey OCP Land Use Designations^ 384.3 Population Density in Surrey, 1985 404.4 Highway System in Greater Vancouver^ 424.5 Greater Vancouver's Transit System 444.6 Transit Routes in Surrey^ 464.7 Employment Distribution, 1971-1991^ 474.8 Transit Mode Split by Type of Commute, 1985^ 504.9 Commuting Flow Types by Mode^ 505.1 Traffic Zones in Surrey^ 535.2 Commuting Flows Between TZ Pairs^ 615.3 Town Centres in Surrey^ 645.4 Population Density and Mode Split^ 715.5 Employment Density and Mode Split 716.1 Plan for Surrey City Centre^ 766.2 Circumferential Suburban Transit Routes^ 796.3 Inter-nodal Transit Service in Surrey 816.4 Potential Locations for Paratransit Service in Greater Vancouver^ 826.5 Importance of Service Attributes^ 866.6 Attitudes About Transit^ 867.1 Conceptual Framework For Suburb-to-Suburb Transit Planning^ 99viviiGLOSSARYCensus metropolitan area (CMA) This is defined by Statistics Canada as a large urbanized core andadjacent urban and rural areas with a high degree of economic and social integration. It is used forstatistical purposes.Density Densities quoted in this thesis refer to gross densities, calculated by dividing the populationor employment of some geographic unit by its total area. Net residential densities, which excludevacant and other non-residential land, would be considerably higher.Greater Vancouver Regional District (GVRD) This is a form of regional government made up ofthe eighteen municipalities and two electoral areas in Greater Vancouver. It has specified powers andfunctions, including transportation modelling and forecastingHigh occupancy vehicle (HOV) Usually refers to a vehicle with three or more occupants. Thiscould be a car or vanpool, or a transit vehicle.Single occupant vehicle (SOV) A private automobile with just one occupant, the driver.Suburb As used here, this refers to any part of a metropolitan area outside the legally defined centralcity.Suburb-to-suburb commute This is a trip which begins in one suburban area, and ends in the sameor another suburban area.Suburban Node This is a concentration of office, retail, residential, and other uses located in asuburban area. These have been called suburban employment centres, edge cities, regional towncentres.Transportation demand management (TDM) This involves a wide range of strategies aimed atchanging people's travel behavior in order to reduce vehicle travel and congestion. This isaccomplished by reducing the total number of trips made, shifting trips from single occupant vehiclesto other modes, and lowering demand during peak periods.Traffic zone (TZ) This is a small geographic unit used for transportation modelling and forecasting.The traffic zones used in this thesis are the same as those used by the Greater Vancouver RegionalDistrict in the 1985 Metropolitan Vancouver Origin-Destination Survey.ACKNOWLEDGEMENTSI would like to thank my supervisor, Dr. Setty Pendakur and my second reader, Jim Chim forthe guidance and support they have given me during the writing of this thesis. I would also like tothank Jim Prokop, Martin Kobayakawa, Deming Smith, and Karen Hodgson at BC Transit, MikeDickenson at the Surrey Planning Department, Mike Lai at the Surrey Engineering Department, andRalph Perkins at the GVRD who helped me to find the information I needed, and took the time todiscuss this issue with me. Thanks also to my family and friends for their support and understandingover the past year.viii1CHAPTER 1. INTRODUCTION1.1 Problem StatementPopulation and employment growth in the suburbs has long outstripped that of the central city,and suburban areas now account for not only the majority of residents, but also the majority of jobs inmajor North American metropolitan areas. As a result, suburb-to-suburb commuting has becomeincreasingly dominant while the traditional commute to the central city continues to decline inimportance.This trend has some serious implications for the provision of transit service. Most suburbanareas were built after the Second World War and consist of low-density, automobile-orienteddevelopment. In contrast to the single, compact employment centre of the traditional CBD, suburbanemployment tends to be much less concentrated and is generally scattered over numerous locations.Added to this is the rapid growth of service sector employment, which tends to be highly dispersed.The resulting commuting pattern is much more difficult to serve by public transit than adowntown-oriented radial pattern (which most transit systems have invested heavily in serving).Although the total number of suburb-to-suburb commuters is large, the highly dispersed nature of thetrips means that the number of commuters on any one route is usually too low to be economicallyserved by conventional transit. Because of this, most suburb-to-suburb transit commuters are facedwith much poorer transit service than those bound for the central city: low service frequencies, lack ofdirect routes and inconvenient transfers are common. By contrast, the automobile is very convenientfor suburb-to-suburb trips, and it is here that it has its greatest speed advantage over transit.Incentives such as free parking at virtually all suburban workplaces only add to the attractiveness ofdriving. As a result, transit accounts for only a very small fraction of suburb-to-suburb commutes.INTRODUCTION /2Suburb-to-suburb commuting is not a new phenomenon; even in the late 1950s it wasbecoming important in some cities (Taaffe et al 1963). Taaffe compared travel patterns of commuterswho worked in Chicago's western suburbs (referred to as "peripheral commuters") with those whoworked in the CBD. The major finding of this study was the high degree of automobile dependencedisplayed by peripheral commuters. The mode split for CBD-bound commuters was 69 percent bytransit and 30 percent by automobile, while for peripheral commuters the split was only 9 percent bytransit and an overwhelming 84 percent by automobile (p. 122). Taaffe also found that suburbanemployees lived much closer to their work places than did CBD employees.Many of the trends identified by Taaffe have become much more pronounced in recentdecades. Pisarski (1987) undertook a comprehensive study of changes in U.S. commuting patternsbetween 1950 and 1980. He found that three major trends have affected commuting patterns duringthis period. The first major trend has been a 65 percent increase in the labor force due to populationgrowth and an increasing participation rate. Secondly, more than two-thirds of both population andemployment growth has been concentrated in the suburbs. Finally, there has been a boom inautomobile use as vehicle availability per household increased from 1.0 in 1960 to 1.6 in 1980 despiteshrinking household size. Collectively, these three trends have created an explosion in automobilecommuting, particularly in the suburbs. The automobile's share of work trips increased from 70percent in 1960 to 84 percent in 1980, while transit's share decreased from 12.6 percent to 6.2percent (p. 6). Information from the recent 1990 U.S. Census indicates that this trend has continuedwith the automobile's share increasing to 87 percent and transit's share declining to 5.3 percent(Barringer 1992).Some of Pisarski's findings are summarized in Table 1.1. He found that the suburb-to-suburbwork trip was the fastest growing type of commute, accounting for 58 percent of commute growthbetween 1960 and 1980, and increasing its share of total commuting flows from 29 to 40 percentduring the period. Commutes within the central city declined from 49 to 33 percent of the total duringINTRODUCTION /3the same period. In metropolitan areas with more than 1 million people, suburb-to-suburb commutesaccounted for 45 percent of trips in 1980 (p. 40). In fact, these numbers underestimate the growth ofsuburb-to-suburb commuting since many rapidly growing sunbelt cities include suburban areas withinthe boundaries of the central city. Between 1960 and 1980 the absolute number of suburb-to-suburbautomobile commuters more than doubled, from about 10 million to over 25 million.Table 1.1 Summary of U.S. Metropolitan Area Work Trip Flows, 1960-80.Type of Flow 1960Trips (millions)^Percent of Total1980Trips (millions)^Percent of TotalPercent Increase1960-80Within Central CityCentral City to SuburbSuburb to Central CitySuburb to SuburbTOTAL18.8 48.6% 20.9 33.1% 11.2%2.0 5.2% 4.2 6.7% 110.0%6.6 17.1% 12.7 20.1% 92.4%11.3 29.2% 25.3 40.1% 123.9%38.7 100.0% 63.1 100.0% 63.0%The problem for the transit industry is that this fastest growing segment of the commutermarket is also the least likely to use public transit. Figure 1.1, which is again based on Pisarski'sfindings, clearly illustrates this. People who commuted to central cities in 1980 were six times morelikely to use transit as those commuting to suburban jobs. Overall, transit's share of suburb-to-suburbtrips was only 1.6 percent (Pisarski 1987, 40).Figure 1.1 Transit Mode Split by Type of CommuteWithin Central CitySuburb to Central CityCentral City to SuburbSuburb to SuburbTOTAL0%^2%^4%^6%^8%^10%^12%^14%^16%Transit Mode SplitINTRODUCTION /4Another trend during this period, noted by Rosenbloom (1990), is the increasing number ofelderly and poor that are living in suburban areas. These groups are often captive transit riders andthere is a strong argument that government has an obligation to provide transit services for thesegroups so that they may better take advantage of the social services and employment opportunities thatare increasingly located in the suburbs.Suburban transportation problems emerged as a major issue during the 1980s, as growth insuburban traffic far exceeded any increases in the capacity of transportation infrastructure. In manyhigh growth areas, what were once country roads are now major commuter routes. Between 1982 and1987, the suburbs' share of metropolitan office space in the U.S. increased from 43 to 58 percent, andin some cities such as Phoenix, more than 90 percent of office space is in the suburbs (Cervero 1984,536). Much of this development over the past 15 years has located in so-called suburban employmentcentres, concentrations of office, retail, and other uses which became a part of the suburban landscapeof many cities. While these suburban employment centres have probably been the most severelyaffected, the rise in suburb-to-suburb commuting has created transportation problems throughout thesuburbs.1.2 PurposeWith suburban areas continuing to outpace central cities in population and employmentgrowth, suburb-to-suburb commuting will likely become even more dominant in the future. If wewish to reduce our dependence on the automobile and the problems that go with it, we clearly have toaddress this problem.According to a report by the Transportation Research Board (1987, 94), a research priority inthe area of public transit should be to determine "...the circumstances most conducive to servicereconfiguration and the service delivery patterns most cost-effective for targeting new [suburban]INTRODUCTION /5markets...". This thesis will look at the growth of suburb-to-suburb commuting and investigate howsuch trips can be better served by transit. The central hypothesis for the thesis is that it is impossiblefor transit to comprehensively serve the entire suburb-to-suburb commuting market but that it cansuccessfully compete with the automobile in certain niche markets. Transit's success in these marketscan be enhanced with a variety of planning strategies. Specific objectives for the thesis include thefollowing:1. To look at the implications for transit of increased suburb-to-suburb commuting.2. To determine which sub-markets of suburb-to-suburb commuting are the best candidates fortransit service, and to determine what types of transit/paratransit service would be mostsuitable for each market niche.3.^To determine what strategies are most effective for increasing suburb-to-suburb transit useincluding the following:a) changes in land use such as employment or residential densitiesb) transportation demand management strategies such as road pricing or parking controls.4.^To investigate how different land use and trip characteristics such as population density orcommute length affect transit use.5. To study specifically the situation in Greater Vancouver and in Surrey as to how wellsuburb-to-suburb commutes are being served by transit and what improvements could be made6. To derive from these investigations a conceptual framework for suburb-to-suburb transitplanning.INTRODUCTION /61.3 MethodologyThe first part of this thesis involves a review of relevant literature on public transit andsuburb-to-suburb commuting. The next part of the thesis will be a case study of Surrey. Backgroundinformation on commuting trends in Greater Vancouver and in Surrey will be collected, followed bymore detailed information on suburb-to-suburb commuting in Surrey. Rather than attempt to collectoriginal data, existing data on commuting and transit in Greater Vancouver will be used, including theMetropolitan Vancouver Origin -Destination Survey (GVRD 1987) and BC Transit's Usage andAttitude Survey (1992). Suburb-to-suburb commuting patterns in Surrey will be analyzed and potentialmarkets for transit service will be identified. The relationship between transit use and density willalso be further explored. In addition, a brief survey of planning agencies will determine whatstrategies are presently being employed in Surrey to deal with this issue. The results of the case studywill be used to determine which transit strategies would be most effective for Surrey. From thisprocess, a conceptual framework for suburb-to-suburb transit planning will be developed.1.4 Assumptions and LimitationsThe term "suburb-to-suburb" is used throughout this thesis to describe a trip that originates inone suburb and ends in the same or a different suburb. The emphasis here will be on those suburb-to-suburb trips which do not follow traditional radial commuting flows since these present the greatestchallenge for transit. Such trips are sometimes termed circumferential commutes. It should also benoted that a suburb is defined here as any part of a metropolitan area outside of the central city. Aswas mentioned earlier, this can result in problems of comparability since the legal boundaries of thecentral city do not always coincide with what is "urban" versus "suburban" in character.This study will be looking at work trips only. There are several reasons for this. Work tripsaccount for the largest single travel demand, about 31 percent of all vehicle travel in GreaterINTRODUCTION /7Vancouver in 1985 (GVRD 1987). Their importance is further magnified because they tend to beconcentrated during the two peak periods, and most transportation capacity problems are associatedwith these peaks. Due to the less variable nature of work trips, they are generally easier to serve withtransit than any other trip type. Thus, this would be the logical first step in providingsuburb-to-suburb transit service. A more practical reason for confining this study to work trips is thatthere is more accurate data available for these.Most of the available literature on this topic is American; there are few Canadian studies fromwhich to draw. Given the important cross-national differences between U.S. and Canadian cities inmany aspects of transportation and urban form (Goldberg and Mercer 1986), it will be of interest todetermine if the trends in suburb-to-suburb commuting documented in American cities also apply toCanada.1.5 Thesis OrganizationChapter 2 reviews the literature that relates public transit and land use characteristics, andhow land use planning can be used to encourage transit use.Chapter 3 includes a review of the literature on the problems transit faces in the suburbs, thetypes of transit and paratransit that can be used for suburb-to-suburb commuting, and transportationdemand management strategies which can be used to increase transit use.Chapter 4 provides a general overview of Surrey and Greater Vancouver, the transportationsystem, and recent commuting trends and patterns.INTRODUCTION /8Chapter 5 analyzes the suburb-to-suburb commuting flows in Surrey and determines whichwould be best suited for transit service. It also looks at the relationship between transit use anddensity.Chapter 6 takes a look at the response in Surrey to the problem of increased suburb-to-suburbcommuting. It also looks at public attitudes towards suburban transit.Chapter 7 presents a discussion of the results from the Surrey case study, leading to thedevelopment of a conceptual framework for suburb-to-suburb transit planning.Chapter 8 offers conclusions and the planning implications of the thesis, as well asrecommendations for further study.9CHAPTER 2. LAND USE AND PUBLIC TRANSIT2.1 The Relationship Between Transit and Land UseThere is clearly an important two-directional relationship between transportation and patternsof urban land use. New transportation modes have helped shape the historical development of cities,but at the same time, the success of these modes has depended on the land use environment in whichthey are operating. Transit is particularly sensitive to land use because it is less flexible than theautomobile. Thus, transit faces a serious challenge operating in the low density environment oftoday's suburbs.Deakin (1991) reviewed the land use/transportation literature. She first looks at the impact oftransportation on land use and concludes that new transportation facilities can have an impact on landuse patterns, but the magnitude of this impact is unclear. In terms of new transit facilities, she agreeswith other authors (Knight and Trygg 1977; Taebel and Cornehls 1977) that supportive factors mustbe present for these impacts to occur. Perhaps more applicable to this study, Deakin also looks at theeffect that land use can have on transportation demand, in particular she looks at the effect ofgovernment land use policies. She points to this as a prime area for future research. Most studies todate have focussed on specific projects and how such factors as density, development size, land usemix, and design features can explain differences in travel patterns. To date, however, she finds therehas not been sufficient work done in this area, and so there is little empirical evidence from which todraw any firm conclusions.One of the most comprehensive studies on the basic relationships between transit demand andpatterns of urban land use was undertaken by Pushkarev and Zupan (1977). This study helped todemonstrate some of the difficulties of providing transit in areas of low density sprawl. Pushkarevand Zupan looked at several U.S. cities, but focussed on the New York metropolitan region. In someLAND USE AND PUBLIC TRANSIT /10ways this is a unique case, although many of the basic relationships they discovered are applicableelsewhere. They found a strong relationship between transit use and two aspects of urban form:residential density and the presence of large, high density central business district (CBD). Differencesin population density could explain 57 percent of the variation in transit use, and when combined withdifferences in CBD size, could explain 79 percent of the variation (p. 26).Their study showed that increasing residential density both increased transit use and reducedautomobile use. As residential densities increase, the number of potential transit riders in a given areaincreases, allowing service levels to be raised. The better service, in turn, attracts even more riders,further increasing transit use and creating a positive feedback loop. The result is an exponentialincrease in the number of riders in a given area as densities increase. Pushkarev and Zupan found thatin the U.S., no form of transit was economically feasible at densities below 3 dwelling units/acre(roughly equal to a gross density of 800 persons/km2) except perhaps park and ride facilities. Atthese low densities, virtually all transit trips are to large downtowns. Transit use remained low up todensities of about 7 du/acre (2000 persons/km 2), but above this level there was a sharp increase intransit use, especially "local transit". A density of 15 du/acre (4500 persons/km 2), for example,could support frequent local bus service. The point at which transit use suddenly increases couldrepresent some critical density above which the positive feedback loop goes into operation. Table 2.1summarizes the relationship between residential density and the level of transit service supported.Higher residential densities were also accompanied by a decrease in automobile use for worktrips. For example, a ten-fold increase in density resulted in an 80 percent decrease in automobile use(p. 32). There was also a closely related decrease in automobile ownership rates. This is in part dueto the increased attractiveness of transit because of the generally higher service levels in high densityareas. Walking may also become an attractive alternative to driving at higher densities, since traveldistances are generally much shorter. Automobile use may be further discouraged at high densitiesLAND USE AND PUBLIC TRANSIT /11Table 2.1 Transit Modes Related to Residential DensityMinimum NecessaryResidential DensityMode^ Service^( dwelling units per acre)^RemarksDial-a-bus Many origins to many^ 6^ Only if labor costs aredestinations^ not more than twicethose of taxisDial-a-bus Fixed destination or 3.5 to 5 Lower figure if laborsubscription service costs twice those oftaxis; higher if thricethose of taxisLocal bus^"Minimum," 1/2 mile^ 4route spacing, 20buses per dayAverage, varies as aLocal bus^"Intermediate," 1/2 mile^ 7^ function of downtownroute spacing, 40^ size and distance frombuses per day residential area toLocal bus^"Frequent.; 1/2 mile^ 15^ downtownroute spacing, 120 busesper dayExpress bus^Five buses during two^ 15^ From 10 to 15 miles— reached on foot^hour peak period Average density over^away to largest down-two square mile^towns onlytributary areaExpress bus—reached by autoFive to ten buses^ 3^ From 10 to 20 milesduring two hour Average density over^away to downtownspeak period 20 square mile^larger than 20 milliontributary area square feet of non-residential floorspaceLight rail^Five minute headways^ 9^ To downtowns of 20or better during^Average density for a^to 50 million squarepeak hour. corridor of 25 to 100 feet of nonresidentialsquare miles^ floorspaceRapid transit^Five minute headways^12 To downtowns largeror better during^Average density for a^than 50 million squarepeak hour.^ corridor of 100 to 150^feet of nonresidentialsquare miles^ floorspaceCommuter rail^Twenty trains a day^ 1 to 2 Only to largest down-towns, if rail line existsSource: Pub/k. Trans/vita:Yon and Land Use F'o&y,Pushkarev and Zupan, 1977.LAND USE AND PUBLIC TRANSIT /12because there is less space for storing and driving them and there tends to be more traffic congestion,further increasing the attractiveness of alternate modes.Pushkarev and Zupan also found that transit use for work or shopping trips to anon-residential node increased with the size and density of that node. They estimated that if a newdowntown with 1 million m2 of non-residential space were added to a community, transit ridershipwithin a 5-8 km radius would be 50-70 percent higher than if the same amount of space was added asa traditional highway strip (p. 174). Changes in non-residential land use appeared to be even moreimportant than increasing residential density. Doubling the size of a non-residential node from 1 to 2million m2 would increase transit use 3-4 times as much as doubling residential density from 15 to 30du/acre (p.175). These findings agree with Hendrickson (1986) who found a strong correlationbetween transit use and the proportion of a region's labor force employed in the CBD, usually thelargest and densest non-residential node in the city.Newman and Kenworthy (1989) looked at various transportation and urban form parametersfor 32 major cities in North America, Europe, Asia, and Australia. They found a strong positivecorrelation between population density and the proportion of total travel on transit. There was aneven stronger correlation between transit use and job density. They also found a weaker but stillsignificant correlation between transit use and percent of jobs in the CBD. These results supportPushkarev and Zupan's findings, and indicate that this strong relationship between transit use andurban form is not restricted to New York or even the United States. Figure 2.1 uses data Newmanand Kenworthy collected from the 32 world cities to illustrate the relationship between populationdensity and transit use.What is the explanation for this relationship between transit use and non-residential land usecharacteristics? Large, dense nodes can best support frequent and convenient transit service, while atthe same time, congestion and lack of parking often makes them relatively inconvenient for the•■• ■■ ■t^ ■LAND USE AND PUBLIC TRANSIT /13automobile driver. The effect of these nodes in encouraging transit use appears to decline rapidlywith distance, so residential areas located close to these nodes will tend to have especially high levelsof transit use. Thus putting more housing close to a downtown or a suburban node would be moreeffective than merely increasing densities throughout a region. For example, Pushkarev and Zupanfound that doubling residential densities in an area from 5 to 10 du/acre would be 17 times moreeffective if the area is 1.5 versus 15 km from a 2 million m 2 non-residential node (p.175).Figure 2.1 Population Density and Transit Use70%60%350%6,5 40%cnc, 30%20%210%0%0^2000^4000^6000^8000^10000^12000Population Density (persons/km2)It is clear from this discussion that transit will have a very difficult time competing in the lowdensity environment of the suburbs. Increasing suburban residential densities should improve transituse, especially for local trips; however, higher employment densities appear to be even moreimportant for improving transit's appeal. A combination of the two would likely be most effective.2.2 Land Use PlanningLand use planning could be used to guide suburban development in a way that encouragesgreater transit use. This would involve promoting many of the urban form characteristics discussedLAND USE AND PUBLIC TRANSIT /14above, including higher population and employment densities, as well as greater centralization ofemployment (although not necessarily into a single centre). Several studies have looked at varioustypes of urban form in order to determine which allows for the most efficient transportation system.A study by Rice (1977) is particularly helpful because it looks at both highway and transitsystems. Rice compared six urban forms (central core, homogeneous, multi-centred, radial corridor,linear, and satellite) using land use statistics from many cities to determine the land use configuration.He found that, overall, the multi-centred form resulted in the most efficient transportation system withthe least number of person-hours for work trip travel and the shortest average work trip length.While transit use was not as high for the multi-centred model as for the central core, the radialcorridor, or the linear forms, it was higher than for the homogeneous form, which probably mostclosely resembles the present situation in the suburbs.Schneider (1981) looked at the role transit can play in encouraging the development of apolycentric urban form by providing enhanced accessibility to certain well defined nodes. In turn,Schneider believes, a city with several relatively high density nodes is better able to support anarea-wide transit system since the high employment density in these nodes is easier to serve by transit,and makes transit use more attractive. If the nodes also include areas of high residential density,transit use among these residents, as Pushkarev and Zupan explained, can be very high. Thus thedevelopment of the polycentric city, which appears to be a natural market trend, should not be fearedby the transit industry. Rather, it should be guided in such a way as to better support economical andefficient transit. In addition, the polycentric city has other advantages including reduced city-centrecongestion, reduced total travel, and a greater degree of urbanity and self-sufficiency in the suburbs.Schneider surveyed 46 metropolitan planning organizations throughout North America. Hefound that 36 of these had land use plans, with the majority envisioning a polycentric form for theircities (p. 50). The greatest amount of work and the most progress in implementing these plans was inLAND USE AND PUBLIC TRANSIT /15the two Canadian cities he surveyed, Vancouver and Toronto. In Vancouver, planning for apolycentric urban form was through the GVRD's Livable Region Program (1975). A similar plan wasdeveloped in Toronto by the Metro Toronto Planning Department (1976). A major goal of both planswas to create a balance of jobs and population in each part of the region. This would be accomplishedby encouraging the development of suburban nodes. In Vancouver, these would consist of 4-6 regionaltown centres, while in Toronto, there would be a series of both major and minor centres. Thus,suburban residents could work and shop in their nearest centre, reducing the need to commute andrelieving pressure on the transportation system.Both plans would encourage suburban transit use through the creation of a more transitfriendly urban form, with compact, mixed-use development in the nodes. In both cities, some of themajor nodes would be linked with rapid transit, providing these nodes with good accessibility andhelping to concentrate development in them. While the primary goal of the plans was to reducecommuting, they would make it easier for suburb-to-suburb commuters to travel by transit. Rapidtransit links between suburban centres would carry some suburb-to-suburb commuters, and expressbus service between those centres not connected by rapid transit was also suggested. Clearly, suchregional land use plans could play an important role in increasing suburb-to-suburb transit use.2.3 Suburban NodesIn the last decade, suburban non-residential development has shown an increasing tendency toform relatively high density clusters. These have been called suburban employment centres, edgecities, suburban downtowns, and numerous other variations, but here they will be referred to simplyas suburban nodes. They are usually dominated by office development; however, retail, hotel, andother commercial uses are also common, and some suburban nodes have a residential component aswell. Most major North American metropolitan areas now contain one or more suburban nodes.Because of this trend, a large amount of the literature on suburban transportation now focusses on theLAND USE AND PUBLIC TRANSIT /16problems associated with these suburban nodes. It is worthwhile to look at this information becausesuburban nodes will likely play a key role in suburb-to-suburb transit planning. The nodes will beimportant transit destinations in their own right, but with their relatively high densities, they may alsobe one of the few viable markets for transit in the suburbs.Cervero has looked extensively at transportation problems related to suburban nodes. In onestudy (1989), he found that land use characteristics had a major influence on the travel behavior atsuburban nodes. The three most important variables were density, size, and land use mixture.Cervero found that in larger and denser nodes there was a greater reliance on transit, ridesharing,walking, and other alternatives to the automobile. It is important to note, however, that the automobilewas the overwhelming travel mode in all of the 57 suburban nodes surveyed: the average transit splitin the nodes was less than 2 percent, and only 4 nodes had a share greater than 7 percent, the U.S.metropolitan-wide average (p. 64). Suburban nodes in Canadian cities tend to have somewhat highertransit mode splits (often 10-20 percent), but they are still very much dependent on automobile traffic(Cervero 1986; Schneider et. al. 1979).The degree of land use mixture in a suburban node was found to have the greatest influence ontransportation behavior, and Cervero concentrates on this. He found that mixed use development canreduce transportation problems in a number of ways. Motorized travel can be reduced since mixeduse suburban nodes allow for more internal trips, which can often be made by foot. For example,office workers do not need to drive to restaurants or shopping during their lunch hours. Commuterswho do not need a car for midday errands are more likely to consider transit or carpooling. Becausedifferent uses are busiest at different times, traffic tends to be more evenly spread throughout the dayat mixed use developments. This phenomenon of staggered peaks also allows for a greater degree ofshared parking which reduces the overall need, and in turn can allow for higher densities and a morepedestrian oriented environment. Cervero (1988, 431) found that while a 10,000 m 2 office complexcould be expected to generate 1230 vehicle trips per day, 10,000 m 2 of development divided equallyLAND USE AND PUBLIC TRANSIT /17into office, retail, and apartments would generate only 1000 vehicle trips per day, not even taking intoaccount further reductions from more internal walking trips. Cervero performed a regression analysisand found that as the amount of retail space in a complex increased and the dominance of office spacedeclined, ridesharing, walking, and biking all increased.One of Cervero's (1989) major arguments is that the way suburban nodes are designedcontributes to their automobile dependence.. While they are dense by most suburban standards,distances are often too far to make walking convenient. Sidewalks are frequently lacking and it isoften difficult to cross busy streets. The large amounts of surface parking is a feature of suburbannodes that distinguish them from traditional CBDs, and this makes walking distances even further.Transit stops are often in inconvenient locations, especially compared with parking. These designfeatures discourage transit use among commuters to suburban nodes. They also leave driving as theonly convenient alternative for midday trips, which further necessitates commuting by automobile.The Snohomish County Transportation Authority (1989) prepared a set of guidelines fortransit friendly design which can be applied to suburban nodes. These guidelines begin by shifting thepriority from automobile drivers to transit riders and pedestrians (since transit riders end their trip onfoot). Walking distances need to be reduced and walking needs to be made safer and more pleasant.This can be achieved through a number of measures. Better transit access into developments (i.e.door front service) reduces walking distances. Moving parking to the back of developments so thebuildings can be brought up to the sidewalk also greatly reduces walking distances, and results in amore interesting and pleasant pedestrian experience, especially if buildings are outward-looking andprovide an engaging streetscape. Walking can also be made safer and more pleasant by providingadequate sidewalks, weather protection, lighting, landscaping, and crosswalks. Barriers such as walls,berms, and parking lots also need to be eliminated. Bus stops themselves should include bus bays,seating, and shelter to improve safety and comfort for transit users. Some of these design features areshown in Figure 2.2.Transit related developmentAutomobile related developmentGibRoadwayI Bus)!ExclusiveBike Parking• ParkingJ 0BikewayHOVParkingILAND USE AND PUBLIC TRANSIT /18Figure 2.2 Examples of Transit Oriented DesignSource: A Guide to Land Use andPub/ic Trahsportatim /1"/- Snohomish Claw* WashingicaSnohomish County Transportation Authority, 1989.LAND USE AND PUBLIC TRANSIT /19Hooper (1989, 2) looked at 6 typical suburban nodes: Bellevue (Seattle), South Coast Metro(Los Angeles), Southdale (Minneapolis), Parkway Center (Dallas), Perimeter Center (Atlanta), andTysons Corner (Washington). The first 3 were classified as small suburban nodes while the latter 3were classified as large. The only suburban node with a significant transit mode split was Bellevue (7percent), while the share was below 1 percent for all the others. Lunch hour trips were twice as likelyto be internal at large suburban nodes. However, most of these lunch hour trips were made byautomobile: only 6 percent of these trips were made on foot (25 percent in Bellevue). Hooper foundthat automobile dependence for these midday and after-work errands was extremely important indiscouraging transit and ridesharing. He makes a series of recommendations. Transit service shouldbe improved, with the suburban node serving as a focus for radial bus service. He suggests apractical limit may be a 6 percent transit mode split. Complexes require a greater degree ofclustering, with a reduction in the inward focus of the individual buildings. More attention must alsobe paid to the relationship between transportation and land use.2.4 SummaryThe literature indicates that there is a strong relationship between public transit and land use.High population and employment densities appear to be particularly critical in order for transit to besuccessful. One way to achieve these densities in the suburbs is through the development of suburbannodes. These nodes offer one of the best potential markets for suburb-to-suburb transit service sincethey help to reduce the dispersed pattern of suburban trips. In many North American cities, suburbannodes have evolved through market forces, but often these are not very transit friendly. They need tobecome even more compact, with a greater diversity of uses, and it must be easier to travel withinthem by means other than the automobile. Land use policies and guidelines can be used to helpachieve these goals. Land use strategies, and suburban nodes in particular, will play a critical role insuburb-to-suburb transit planning.CHAPTER 3. STRATEGIES FOR INCREASING SUBURBAN TRANSIT USE3.1 Transit's Challenge in the SuburbsWhile suburban transportation problems have become a major issue in the last decade, there islittle optimism that conventional transit can play a significant role in solving these problems. Orski(1985), for example, believes that transit will be able to serve only a very small segment of the totaltravel demand in the suburbs, and as this suburban market becomes increasingly dominant, transit'sshare of the travel market will inevitably shrink. Cervero (1986) warns that conventional transit'sfuture in the suburbs does not look promising.A statement made at the annual meeting of the American Transit Association in 1940 is veryprophetic and shows that the changes in urban form that have made transit uneconomic were notcompletely unexpected:Can we not pause long enough on this headlong decentralization process to see where we aregoing. The mass transportation industry is caught in a strong tide which is sweeping this andmany other businesses toward disaster... [The] situation calls for strong expression andvigorous leadership (Cervero 1986, 105).Fifty years later the transit industry is still struggling to adapt to these changes. Weigle(1988) argues that the transit industry twice failed to become a part of the post-war suburban boom:first in the 1950s and 1960s as people moved out to the suburbs, and again during the 1970s and1980s as employment became suburbanized as well. Instead, transit has continued to concentrate itsresources on the CBD-oriented market. While Weigle agrees that this is the market in which transitwill continue to perform the best, it occupies a decreasing share of the regional commuter market; theopportunity for growth lies in suburbia.20STRATEGIES FOR SUBURBAN TRANSIT /21In order for transit to grow in the suburbs, Weigle believes that it must market itselfaggressively and play a more active role in the design stage of suburban development. It mustestablish itself in suburban areas as they are being developed, so it can better shape developmentpatterns and travel habits of the residents. He believes that transit must be able to respond faster inthe rapidly changing environment of the suburbs, possibly through joint ventures with the privatesector. A key point he makes is that rather than attempting to provide comprehensive service, transitneeds to identify and fill local market niches. This would help to establish a foundation for transit inthe suburbs from which it could grow.Schofer (1983) also discusses ways in which the transit industry will have to change torespond to suburban development. Like Weigle, he agrees that transit will have to tailor itself tospecific markets. It will have to move away from fixed route, fixed schedule, line haul servicetowards paratransit and ridesharing. This would help make transit more efficient. In light of this,Schofer finds the trend toward building rail systems odd since most of the literature says that busesare cheaper in low density markets. He recommends deregulation and returning to the situation ofmany smaller providers, but with regional coordination. The private sector should be encouraged tobecome involved in transit through contracting out and private operators. Transit operators should notbe afraid that private operators will "skim the cream"; often they help to reduce the deficit by takingover unprofitable routes. Overall, Schofer, like Weigle, believes the transit industry needs to be muchmore flexible and open to change.Despite much pessimism, there appears to be a general consensus that transit does stand achance in the suburbs if it becomes more willing to adapt to the circumstances. Future transit systemsmust be far more than standard buses on radial, CBD-oriented routes; they must includemulti-destinational grid systems, paratransit, private involvement, and other innovations. Perhaps mostimportantly, the transit industry must also give up the idea of serving the entire suburban market, butSTRATEGIES FOR SUBURBAN TRANSIT /22instead should concentrate on those sub-markets in which it can best compete. Only by making thesechanges will transit begin to make up the ground it has lost over the past five decades.3.2 Suburban Transit StrategiesWhat types of transit strategies are most appropriate for suburb-to-suburb travel? Thefollowing is a summary of some of the strategies which have been discussed in the literature.In very large cities, conventional transit, even rapid transit, might be viable for somecircumferential suburban routes. Young (1984) cites an example in Chicago where suburb-to-suburbrapid transit is being considered. The Regional Transit System (RTS) wants to establish itself in thesuburb-to-suburb market by building a rail beltway 24 km west of the CBD, connecting with theregion's commuter rail network. The line would run through a 10x30 km corridor which contained587,000 jobs in 1980, almost as many as the CBD, but spread out over a much larger area. This isthe highest volume traffic corridor in Chicago's suburbs. The RTS projects 30,000 riders per day.Toronto is also planning a non-CBD rapid transit line which would stretch along the region's northernsuburbs and connect several major suburban nodes.Many of the proposals for suburban transit service are organized around suburban nodes.Huth (1983) advocates a multi-nodal urban form with well developed suburban transit servicefocussing on suburban nodes as well as the CBD. Such a multi-destinational transit system wouldconsist of three types of transit service: (1) local service between nodes and their surroundingresidential areas (2) express service between different nodes and from the nodes to the CBD (3)internal circulation within the nodes. Huth cites Minneapolis and Denver as examples of U.S. citieswhere this form of development is being encouraged. Vancouver has at least some of these features inits transit system, as will be discussed in later chapters.STRATEGIES FOR SUBURBAN TRANSIT /23Schneider and Smith (1980) also advocate a multi-nodal transit system. They suggest usingregional shopping centres as nodes since these are important trip generators in their own right andthey usually serve about 100-200 thousand people, which is also about right for transit nodes. Amajor advantage of a multi-nodal system is that by concentrating passengers into heavily-travelledcorridors between the nodes, it provides the opportunity for improved high-frequency express service.Schneider and Smith also believe that such a system would be more user-friendly: people would learnthat once a node is reached, other destinations throughout the urban area can be reached easily.The transit networks described by both Huth and Schneider would likely use some form oftimed focal point transfer system. Priest and Walsh-Russo (1983) stress this system as a way ofcreating a multi-destinational transit network that would greatly improve suburban transit service.Timed transfer points are nodes in the transit network where several routes converge, usually within afew minutes of each other, to allow for convenient transfers and thus better connectivity. These timedtransfer points are usually located at shopping centres or other suburban nodes, tapping a market thattransit has often ignored. The timed transfer points serve as hubs for local transit service and providethe connection with express services which operate to the CBD and sometimes to other transfer points.Priest and Walsh-Russo cite Edmonton and Toronto as successful examples of this type ofmulti-destinational transit service. Vancouver's system also uses timed transfers in the suburbanareas. Transit ridership in Edmonton increased by 45 percent in the five years following theintroduction of a timed transfer system which allows 90 percent of the service area to be reachedwithin 50 minutes (Cervero 1986b, 307). In Portland, two timed transfer points were introduced in1979 and ridership in the area increased 40 percent (Priest and Walsh-Russo 1983, 42). There havebeen several examples where entire radial systems have been successfully converted into timedtransfer systems, including Edmonton and Austin, Texas. Timed transfer systems seem to beparticularly well suited for relatively low density areas where existing headways are typically 30STRATEGIES FOR SUBURBAN TRANSIT /24minutes (Bakker et al. 1991). A major drawback of using timed transfers is the difficulty and cost ofscheduling such a system, and the need for all routes to have compatible service frequencies.An alternative to the timed transfer system, but one that also provides multi-destinationalservice, is the grid transit system. Usually these systems are found in areas with high transit use sincethey generally work best with frequent headways. This is because, given the large number of transferpoints, it is impossible to schedule the service so that transfers are timed in any way; thus waitingtimes tend to average one half the headway. In Vancouver, a grid system is used in the City itself,where service is most frequent. There may be potential for grid transit systems in areas of lowerdensity and lower transit service, however. Nelson and O'Neil (1983) describe the case ofAlbuquerque, New Mexico, a city which had a population of about 400,000 in 1980 and a multi-nodalform with no dominant activity centre. Albuquerque introduced a grid transit system, with busesrunning on all major arterials at 30 minute headways. Transit passengers can now go nearlyanywhere in the city with a maximum of two transfers. Most locations are served by at least tworoutes and users thus have a choice of routes when planning a trip, often catching whichever buscomes along first. Nelson and O'Neil report that the system is quite successful, indicating thepossibility that such a system could be adapted for use in multi-nodal areas with relatively low transituse, including the suburbs.Cervero (1986) advocates a combination of grid and timed transfer systems to replace existingradial transit systems in the suburbs. The result would be a cobweb-like network with high routeinterconnectivity and greatly enhanced service to important suburban nodes. The major goal would beto better serve the diffuse travel pattern of the suburbs and, in particular, to reduce the inconvenienceof making transfers, which are a necessary part of transit when origins and destinations are sodispersed. Cervero points out that people perceive the waiting time during transfers to be three timesas long as it actually is (p.12). The hybrid system that Cervero describes is the kind of flexiblesolution that transit will need if it is to be successful in the suburbs.STRATEGIES FOR SUBURBAN TRANSIT /25Cervero (1986, 103) also discusses some of the potential problems such a transit system wouldface. Often transit services do not penetrate suburban complexes because developers are reluctant forsafety and image reasons, while transit authorities may fear headways will lengthen or schedules willsuffer if buses are detoured into these complexes. Another problem results because timed transfersseem to need residential densities of 1200 to 2400 persons/km 2 in order to operate efficiently. Whilethis is a relatively low density, most U.S. suburban areas have residential densities which are evenlower, and thus have difficulty supporting timed transfer systems. For example, while 12 bus routesconverge at South Coast Plaza in Los Angeles, only 5 percent of trips are by transit. A majorproblem is that only 21 percent of South Coast employees live within a 5 minute walk of a bus route,the maximum time most people will walk. Fairfax County, Virginia ran a cross-county bus service toTysons Corner, a huge suburban node, in the early 1980s, but it attracted an average of only 6 ridersper trip. Clearly it will take more than simply adding bus service to these nodes to improve transituse.Another problem for transit, of course, is that of funding. Orski (1985) is a strong supporterof cooperative financing, in which the private sector helps to pay for transportation improvements.An example of this would be benefit assessment districts such as those established around transitstations in Los Angeles and Portland. These are a way of taxing landowners who benefit from nearbytransportation improvements. In other cases, the private sector might pay for part or all of certaintransit improvements which increase the accessibility of a development. To date, this has mostlyinvolved rapid transit projects such as joint station developments, but this could conceivably beextended to other forms of transit or even paratransit.3.3 ParatransitParatransit includes demand responsive services like shared ride taxis and dial-a-bus, as wellas ridesharing such as car or vanpools. It is often suggested as a solution to providing transit serviceSTRATEGIES FOR SUBURBAN TRANSIT /26in areas where densities and levels of demand cannot support conventional transit. Paratransit wouldalso help reduce the need for transfers by providing more customized service; often this is"door-to-door". Most authors agree that paratransit must play a larger role if transit is to besuccessful in the suburbs (Reichart 1979; Schofer 1983; Orski 1985; Humphrey 1990). In order forparatransit to play this larger role, many believe that the monopoly of standard bus public transitneeds to be broken by deregulating the industry and allowing more private sector involvement(Schofer 1983; Cervero 1986). Of course, there have been many examples of publicly operatedparatransit services, but these have tended to have very high operating costs per passenger,particularly the demand responsive services such as dial-a-bus (Bhatt 1979). Rosenbloom (1990)surveyed 22 medium-sized U.S. cities and found that despite the many paratransit options available,only two of the cities were involved with non-traditional paratransit services.Ridesharing may be a more promising and less expensive option for suburb-to-suburbcommuting. Carpooling programs in particular can be provided at a low cost since they merely makeuse of a resource which at present is vastly underutilized: all the empty seats in private automobiles.Ridesharing is less expensive than driving to work alone. At the same time, it is generally faster,more comfortable, and more convenient than regular transit since it offers door-to-door, expressservice with a guaranteed seat. Ridesharing seems to be especially well suited to long commutes,where the cost of driving can be quite high and the distance required to pick up fellow passengers issmall compared to the actual "line haul" trip. Teal (1987, 207) found that carpoolers in the U.S. hadcommutes that were 43 percent longer, on average, than those who drove alone. Most car andvanpools have a single destination, so large employers and employment concentrations makeridesharing more attractive (Stevens 1990). A potential problem with ridesharing programs is thatthey often compete directly with transit services, something which may or may not be desirable.Likely targets for ridesharing programs in suburban areas might be suburban nodes or a large isolatedemployer which presently has poor transit service.STRATEGIES FOR SUBURBAN TRANSIT /27Shuttle buses could represent an important means of circulation within suburban nodes, onethat could help end the dependence on the automobile for midday trips. Higgins (1992) looked atseveral shuttle bus services in suburban nodes, office parks, and other locations in the United States.Many of these services were privately run, and most operated either as midday circulators or feederservices to rail and bus lines. Some had a fixed schedule while others were on demand; some serviceswere free while others charged a small fare. He found that most of these services were costly andgenerally had low ridership, in most cases amounting to less than 2 percent of all employees.However, Higgins did find some successful examples. These tended to have certain characteristics:service frequency was at least every 10 minutes, parking in the area was usually restricted, and theshuttle buses did not compete directly with transit but they were connected to well used transit lines.Thus, shuttle buses may have some potential under certain circumstances.Given that many potential suburban market niches consist of relatively small flows, paratransitmight be an appropriate way to fill these niches. In many cases, paratransit could be combined withmore conventional transit to fill a particular market need. For example, paratransit could be used as afeeder service to a timed transfer point in an area where population densities do not support full busservice (Cervero 1986). Roos (1979) describes this as the "family of services concept", which headvocates as an alternative to the "one size fits all" approach to transit provision presently used inmost North American cities. Since no single service or provider can satisfy all of a community'stransportation needs, a combination of different services and providers must be interfaced to producean overall system, one that includes an array of both transit and paratransit services.3.4 Transportation Demand ManagementTransportation demand management (TDM) strategies include a broad range of measuresaimed at changing travel behavior in order to reduce vehicle traffic and congestion. This can beaccomplished in three ways: reducing the total number of trips made, shifting demand away from peakSTRATEGIES FOR SUBURBAN TRANSIT /28periods, and shifting trips from single occupant vehicles (SOVs) to alternate modes. Here, the focuswill be on those strategies which could shift suburb-to-suburb trips from SOVs to transit andparatransit. TDM can consist of mandatory, punitive measures which discourage SOV trips byincreasing the cost or time involved (sticks), or they may be voluntary incentives which encouragealternative modes (carrots). Examples of the former include road pricing and parking management,while an example of the latter might be provision of priority lanes for transit and high occupancyvehicles (HOVs). Stick TDM measures are generally much more difficult to implement than carrotTDM. Because it actually changes travel behavior, TDM could be a very powerful tool forencouraging transit and paratransit use for suburb-to-suburb commuting.Road pricing, by increasing the cost of using an SOV, is one of the most effective means ofchanging this behavior. Road pricing simply means applying tolls on the use of bridges andhighways; these may be collected at standard toll booths, or by using more newly developed electronicsensor systems. The rate structure of the tolls can be used to shape travel demand. For example, carand vanpools could be encouraged by allowing these to travel toll free or at a reduced toll. Therevenue generated from road pricing can be used to provide improved transit services as an alternativeto driving. A major criticism of road pricing is its regressive nature: the impact of tolls would bemuch greater on low-income motorists than on those who could afford to pay (GVRD 1993). Longdistance suburb-to-suburb commutes would be a prime target for road pricing.Parking reforms can also be used to encourage alternatives to the automobile. Free employeeparking is almost universal at suburban jobsites. This is a benefit that employers grant to automobiledrivers, at a cost of hundreds of even thousands of dollars per space annually, while transit usersgenerally receive no such subsidy. Pickrell and Shoup (1979A) estimate that 20 percent of those whodrive alone would switch to transit or carpools if they were required to pay for their parking. Theyalso recommend that if employers subsidize automobile drivers, they should extend this to transitusers in the form of subsidized passes. There are also many actions which the public sector couldSTRATEGIES FOR SUBURBAN TRANSIT /29take to increase the cost of parking, including priced parking permits in designated areas, a direct taxon parking spaces, and increased rates at parking meters and municipally owned parking lots (GVRD1993A).Pickrell and Shoup (1979B) argue that minimum parking requirements result in more parkingthan the market would supply naturally, and have depressed the market value of a parking space, thusencouraging more people to drive to work. Reducing these requirements, or even putting a cap onmaximum allowable parking (as has been done in Bellevue, Washington), would help to change this.Cervero (1991, 483) compared commuting habits at 83 different suburban office buildings and foundthat those with fewer parking spaces per employee did have a significantly lower automobile modesplit. Instead of fixed limits, parking requirements could be negotiated with the developer in order toget more transit friendly development. For example, a reduction in required parking for adevelopment could be granted in exchange for the provision of a bus loop. Another strategy is toprovide preferential parking for employees who carpool to work. Parking reforms will be particularlydifficult to carry out in suburban areas, where free parking is taken as a birthright and there are fewalternatives to the automobile. These reforms will be easier to implement if they are accompanied byimprovements in transit service.Another increasingly common form of TDM involves setting trip generation ceilings or modesplit targets. These may be voluntary, employer based programs, or they may be mandatory targets adeveloper commits to meet before approval for a project is granted (Orski 1985). The developerachieves these targets through measures such as ridesharing programs, staggering work hours, andparking management. Encouragement of transit use is usually an important element of these schemes.Orski (1987) cites some examples where such traffic mitigation programs have been quitesuccessful. In Pleasanton, California an ordinance required developers and employers to reduce peakSOV trips by 45 percent over four years, and in the first two years of the program they had alreadySTRATEGIES FOR SUBURBAN TRANSIT /30been reduced by 36 percent (p. 470). In Silver Spring, Maryland a TDM district was established witha goal of 25 percent mode split for transit. The county provides incentives such as discount transitpasses and HOV parking These types of programs are often quite flexible; jurisdictions negotiatewith developers, using carrots such as density bonuses and parking reductions, or sticks such aswithholding development approvals. Cervero (1986) agrees with Orski on involving the private sectorin TDM strategies. An innovative scheme being used by some municipalities in the San FranciscoBay area involves impact fee ordinances, in which developers pay a fee based on the expected trafficimpacts a development will produce. This provides developers with an incentive to design andmanage their projects in such a way as to encourage alternatives to the automobile.Developers and employers in many suburban centres have banded together to formTransportation Management Associations (TMAs) to cope with transportation problems in suburbanareas, as well as in response to the imposition of the regulations discussed above. TMAs areparticularly prevalent in centres which are poorly served by transit. They assist members in trafficmitigation obligations, lobby for transport improvements, plan for long range transport needs, andoperate transport services such as internal circulators, shuttle systems, or contract buses. There were24 TMAs operating in the U.S. in 1985, a number which had grown to more than 100 by 1990(Humphrey 1990). TMAs could play a crucial role in providing innovative transit service forsuburb-to-suburb commuters.To illustrate just how effective TDM regulations can be in reducing automobile dependence,Cervero (1991, 489) cites the example of Pacific Northwest Bell in Bellevue. Bell provides its 1150employees with just 402 parking spaces, over half of which are reserved for car and vanpools. Thecompany charges $60 per month for single occupant vehicles to park, but two-person carpools arecharged only $45 per month and carpools of three or more park for free. At this particular officebuilding, only 50 percent of employees solo commute while 37 percent rideshare and 12 percent useSTRATEGIES FOR SUBURBAN TRANSIT /31transit. A block away at an office tower with 730 spaces for 650 workers, 85 percent of employeessolo commute and only 8 percent rideshare.A major study which compared eleven different sites (seven of them suburban) found that lowoccupancy vehicle trip demand can be significantly reduced by TDM strategies (U.S. FederalHighway Administration 1990). Trip rate reductions varied from 6 to 48 percent, with the averagebeing about 20 percent. The degree of success did not appear to be related to the size or density of thesites. The specific strategies used in the traffic reduction programs, in particular the presence of legalor economic pressure to induce employers and employees to cooperate, had a much greater influence.Most importantly, a viable alternative to the single occupant vehicle must be present, and in order forthis alternative to compete, subsidies for private automobiles, such as free parking, must be ended.While a 10 percent trip reduction has often been viewed as acceptable, this would be an insignificantcontribution to the rapidly growing gap between transportation demand and road capacity, especiallyin suburban areas. However, the study concludes that much more significant trip reductions of 20 to40 percent could become the norm rather than the upper limit if subsidies for drivers are eliminatedand TDM programs are universally implemented to create a level playing field for all employers andall commuters. Guiliano (1992) agrees with this conclusion. While she found that most TDMprograms had a minimal effect on traffic congestion, the effect was significant when programs weremandatory and strong incentives were present.TDM strategies will play an important role in encouraging greater transit use forsuburb-to-suburb trips. Providing transit or paratransit services for these trips is not enough; thismust be combined with both land use measures and TDM strategies that both encourage transit useand discourage - even penalize - automobile use. These are difficult to carry out because they arepolitically unpopular and they require the cooperation of employers and developers in the privatesector. However, if transit is to succeed in the suburbs, this kind of comprehensive approach isnecessary since there is no one solution for the problems it faces.STRATEGIES FOR SUBURBAN TRANSIT /323.5 Ending Automobile DependencyThe innovative suburban transit services, land use changes, and transportation demandmanagement strategies discussed above represent only one step towards the larger goal of ending ourdependence on the automobile. While these strategies play an important role, they should also be seenin context of the broader issues. Ending automobile dependency will require some fundamentalchanges in society's values and priorities. We must decide whether we will continue to design ourcities almost entirely from the point of view of the private automobile, or if we will begin to give thesame amount of consideration to alkriative modes. It is essential that we change the way we assessall the costs and benefits of automobiles and public transit.Newman and Kenworthy (1989) discuss many of these broader issues. They argue thattransportation impacts on all aspects of how we live, and automobile dependence has created manyfar-reaching problems for our cities and our society. Some of these problems are directly related tothe physical structure that automobiles have helped to create in our cities. For example, urban sprawlhas created a need to travel long distances for everyday activities; this has led to increasing congestionand a massive waste of energy. It also means that people are spending as much or more timetravelling as in the days before the private automobile. The automobile-oriented city creates problemsnot only for transit, but the long distances have also made walking and cycling impractical in mostcases. In addition, sprawl creates massive requirements for roads which are extremely costly tobuild, use up huge amounts of land, and result in severe environmental impacts.Automobile dependence also has many social impacts. There is severe hardship for children,the elderly, the poor, the handicapped, and others without access to an automobile. The separation ofactivities in our cities has also led to a loss of neighborliness and vitality. The excessive privacy andisolation which the automobile fosters has resulted in a lack of community, and this has contributed toan increase in crime and other social problems. From their discussion, Newman and KenworthySTRATEGIES FOR SUBURBAN TRANSIT /33illustrate how the issues of transit and automobile use can have impacts far beyond the realm of whatmight be considered "transportation problems".At present, our society is so highly geared towards the automobile that we often do not realizeall the advantages we give to automobile use. Perhaps this is most obvious in the way we viewfunding and subsidies for our transportation system. A recent study estimated that the subsidy fromgovernments and individuals to private automobile use costs each Lower Mainland resident roughly$1300 per year, which means automobile drivers are being subsidized for about 23 percent of theirtotal costs (GVRD 1993B). The total subsidy for private automobiles is 7.5 times that for publictransit in the region, and yet most people are more aware of the transit subsidy. If transit received thesame subsidy as the private automobile, service levels could be vastly improved. These fundamentalchanges in our values and priorities go far beyond limited changes discussed in this thesis, but it isimportant to keep these larger issues in mind3.6 SummaryT-Ansit will have to become more innovative if it is to survive in the suburbs. The traditionalsingle centred system needs to be replaced with a multi-destinational one that better reflects the realityof the suburbs. Different types of transit services should be employed for different segments of themarket; this includes the use of paratransit and ridesharing. These more innovative services must alsobe combined with transportation demand management strategies that will encourage people to switchfrom driving to transit. Transit can compete effectively in some suburban markets, but it will take acomprehensive approach to meet this challenge.34CHAPTER 4. CASE STUDY: BACKGROUND INFORMATION4.1 IntroductionA case study approach was chosen to illustrate some of the implications for public transit ofincreased suburb-to-suburb commuting. The case study consists of three parts. In the first part (thisChapter), background information to the case study will be presented. The situation in Surrey will beput in context of Greater Vancouver and the changing commuting patterns within the region. In thesecond part of the case study (Chapter 5), commuting patterns to, from, and within Surrey will beinvestigated in greater detail, and the effect of land use patterns will be analyzed. The third part ofthe case study (Chapter 6) looks at the response to this issue in Surrey: what strategies are being usedto provide transit service for suburb-to-suburb commutes, how is transit use being encouraged, andwhat are the public attitudes about transit service in Surrey? Overall, the case study gives both aregional and historical context to the issue of suburb-to-suburb commuting in Surrey, then provides adetailed picture of the situation and the response to it.The case study approach was chosen for a number of reasons. First, it contributes to theempirical evidence on the impact of land use patterns on transit use, which a review of the literaturehas shown to be quite sparse. A detailed examination of the specific commuting patterns in Surreyshould provide clues about the different variables which might affect suburban transit use, as well aswhich types of commutes are most amenable to improved transit service. By analyzing thisinformation, potential markets for transit in Surrey can then be identified. Again, there have been fewstudies which have specifically looked at transit use for different types of suburban trips. A keyreason for choosing a case study approach is that it can then serve as a model for other cities, leadingto the development of a conceptual framework for suburb-to-suburb transit planning.CASE STUDY: BACKGROUND INFORMATION /354.2 Surrey and Greater VancouverThe District of Surrey is part of the Vancouver Census Metropolitan Area (CMA). For thepurposes of the case study, the CMA has been divided into nine sub-regions, each consisting of one ormore municipalities which are commonly grouped together and generally form a functional unit.Figure 4.1 shows these sub-regions and their constituent municipalities. Of these sub-regions, theNorth Shore, Burnaby/New Westminster, and Richmond will be referred to as the inner suburbs,while the remaining areas will be termed outer suburbs. Surrey, along with the City of White Rock,forms one of the outer suburban sub-regions. Most of the data presented in the case study refer to thiscombined area of Surrey and White Rock. It will be stated where this is not the case.Surrey, which is located roughly 25 km southeast of downtown Vancouver, was chosen forthe case study because it is fairly typical of the communities which are facing the most seriousproblems from the rapid rise in suburb-to-suburb travel. In terms of both population and area, Surreyis the largest suburban municipality in the Greater Vancouver region. As shown in Table 4.1, Surreyhas grown rapidly as a major recipient of Vancouver's post-war suburbanization. During the last twodecades in particular, it has accounted for a very large proportion (nearly 30 percent) of the region'spopulation growth. It has also been the location for a smaller but growing proportion of employmentgrowth in the region, as will be discussed in more detail later (GVRD 1991).Table 4.1 Population Growth in Surrey, 1951-91.*Surrey Vancouver CMA % of CMAin SurreyPopulation Increase DuringPrevious DecadePopulation Increase DuringPrevious Decade33670 586054 5.7%77291 43621 827335 241281 9.3%108950 31659 1082185 254850 10.1%160688 51738 1268197 186012 12.7%261487 100799 1602502 334305 16.3%*Includes White RockSource: Statistics Canada.19511961197119811991PM MeadowsNEW WESTMINSTER-^ - -NORTH SHOREWest Vancouver North VancouverNORTHEAST SECTORMAPLE RIDGEVANCOUVER BURNABY/Port CoquillarnCoquIllarnSURREYLANGLEYRICHMONDDELTACASE STUDY: BACKGROUND INFORMATION /36Figure 4.1 Vancouver CMA and its Component Sub-RegionsCASE STUDY: BACKGROUND INFORMATION /37Surrey displays a high degree of suburb-to-suburb commuting due to its distance fromdowntown Vancouver and its central location among Vancouver's southern and eastern suburbs.Increasingly, Surrey is being viewed as the second city of the Greater Vancouver Region, and themajor centre south of the Fraser River. Downtown Vancouver's location in the northwest corner ofthe region has helped to create a real need for such a second centre to serve the rapidly growingsoutheastern suburbs. The psychological and sometimes real barrier to commuting provided by theriver has also helped to promote suburb-to-suburb commuting among South of Fraser municipalities.These conditions have made Surrey a very important player in suburb-to-suburb commuting in GreaterVancouver.4.3 Urban FormLike many suburban areas, Surrey is a sprawling community with a multi-nodal form.Surrey's Official Community Plan (1986) encourages this multi-nodal form through the creation of aseries of town centres which act as foci for commercial and higher density residential development.The largest of these town centres is Whalley, which has also been designated a regional town centre inthe GVRD's Livable Region Program. At present, Whalley consists mostly of a regional shoppingcentre, along with strip retail and commercial development, but it is projected to evolve into a muchdenser and more mixed-use "downtown" over the next 20 years. In fact, Whalley is now beingpromoted as Surrey City Centre, one of the future twin downtowns of the region (along withdowntown Vancouver). The second order town centres include Guildford (home to Surrey's largestregional shopping centre), Newton, Cloverdale, and South Surrey/White Rock. Figure 4.2 showsSurrey's basic urban structure as outlined in its Official Community Plan.Residential development in Surrey is predominantly single family (about 70 percent in 1986),but multiple family housing is becoming increasingly important.^According to the OfficialCASE STUDY: BACKGROUND INFORMATION /38Figure 4.2 Surrey Official Community Plan Land Use DesignationsLEGENDME ormTCR Town CentreCOM co.mwooenil AM Multiple Re ...Wel. 1111. MB I.e..SOO SuburbanINA IndustrialIBM AGA prIcullwalSpectat slueISource: Surrey Official Community Plan, Surrey Planning Department, 1986.CASE STUDY: BACKGROUND INFORMATION /39Community Plan, multiple family housing is encouraged in the town centre areas, with the highestdensities in Whalley and Guildford. About 70 percent of multiple family housing is in the towncentres, nearly 40 percent of it in the Whalley-Guildford area. Multiple family housing makes upabout half the total housing in this area. Farther out from the town centres are larger areas of low tomedium density housing designated as "urban". The areas designated "suburban" are for low densityresidential development and provide a transition between Surrey's urbanized area and the belt of ruralagricultural land that runs through the centre of the municipality.Population densities in Surrey are generally quite low, averaging 635 persons/km 2 in 1986.(All density figures presented here represent gross densities; net residential densities would beconsiderably higher.) As shown in Figure 4.3, there is considerable variation in population densitywithin the municipality, however. There are three areas of higher density. The largest of these isNorth Surrey with about 127,000 people (65 percent of the total) at an average density of 1300persons/km2 . This area is contiguous with North Delta to the west. South Surrey/White Rock hasabout 39,000 people also at an average density of 1300 and Cloverdale has a population of 8500 at adensity of 1700. The remaining central and eastern part of the municipality makes up 55 percent ofthe land area, but contains only about 21,000 people with an average density of only 123persons/km2 . This area is rural in nature with extensive agricultural land (much of it in theagricultural land reserve) and only scattered, very low density development. This large rural expanseseparates South Surrey/White Rock from the remaining built-up area.Overall, about a third of the area's population lives at densities above 2000 persons/km 2 ,about a third live at densities between 1000 and 2000, and about a third live at densities below 1000.The "average" Surrey resident lives in a neighborhood of predominantly single family homes with adensity of about 1300 persons/km 2 , or a net residential density of about 12 dwelling units to thehectare.CASE STUDY: BACKGROUND INFORMATION /40Figure 4.3 Population Density in Surrey, 1985Persons per km2under 500500-10001000-2000over 2000 CASE STUDY: BACKGROUND INFORMATION /414.4 The Transportation SystemA brief overview of the transportation system in Greater Vancouver and in Surrey, withemphasis on the transit system, is helpful in understanding the discussion of commuting patternswhich follows.4.4.1 RoadsAs shown in Figure 4.4, Greater Vancouver has a fairly limited freeway system, particularlywithin the City of Vancouver itself. A four km stretch of the Trans-Canada Highway skirtingVancouver's eastern border is the only freeway segment within City boundaries, and there are nofreeway links to the CBD. Circumferential freeways or beltways, which have helped to encouragesuburb-to-suburb travel in many cities, are also largely absent in Greater Vancouver. There are afew freeway segments, such as the Richmond Freeway (Highway 91) connecting Surrey and NorthDelta with Richmond, which serve a beltway function, but these constitute only fragments of a truebeltway. Given its relative lack of freeways, Vancouver is more dependent on surface streets andarterials compared with most North American cities.Several highways connect Surrey with the surrounding suburban municipalities. Three roadbridges provide access across the Fraser River from Surrey to Burnaby, New Westminster, andCoquitlam; however, the bottlenecks which result tend to discourage travel across the river to someextent. Not surprisingly, Surrey enjoys the best access with the adjacent suburban areas of NorthDelta to the west and Langley to the east. North Delta is a continuation of the built-up area of NorthSurrey, and access is by numerous city streets. Built-up areas of Langley are separated from Surreyby extensive rural areas. The Trans-Canada Highway provides high-speed access between NorthSurrey and Langley, while Highway 10 and the Fraser Highway also link the two municipalities.Highway 99 provides good access from South Surrey to South Delta and Richmond. Access to111.A.,,.....,_,„ _4 a - •lilm..... ism Nel^Aqalkallaj lall■O'.:. :4411116,70111.■ A,14,1^:;.471 IPPAIlli^„ Air fl ,-.'-.;..i, ,...46. ..4,11h,-m, .........:,„.airaii mu u.........^5,1,2::^dill b- 4441p.. .....00.7 illirillriv.......:-.....v.„- .4........19.1.1 1 ,2,,,m.,m^,,,,,,„ 1„lip"......-i....16-..1..„,-.4,-,110 -Iirtm.mimmiiiivo„.impeum....1.„,,,u0.----Qaar a.-,.........:_,..uwimh. .4.4,b7:_:, _,.- "ArlinEirm "1" - "•s"ail■1gg'■■■110All.1111E511111111.1■1CASE STUDY: BACKGROUND INFORMATION /42Figure 4.4 Highway System in Greater VancouverNETWORK CLASSIFICATIONFreewayExpresswayOther HighwaysArterial RoadsCASE STUDY: BACKGROUND INFORMATION /43Richmond from other parts of Surrey has been greatly improved with the construction of Highway 91and the Alex Fraser Bridge. The existing freeway network is not really designed for travel within thesprawling municipality, and rapid growth has led to problems of traffic congestion.4.4.2 Public TransitThe Metropolitan Vancouver Origin Destination Survey found that transit accounted for nearly9 percent of 24-hour travel in Greater Vancouver and about 13 percent of work trips (GVRD 1987).This is fairly typical for a Canadian city, but is considerably higher than the average U.S. city.Greater Vancouver's transit system is operated by BC Transit and consists of both diesel and trolleybuses, two commuter ferries, and a rapid transit line. The system carried about 128 millionpassengers in 1991/92.Buses carry the majority of transit passengers in Vancouver. Within the city itself, there is agrid system consisting mostly of trolley buses. The north-south routes generally loop throughdowntown while the most east-west crosstown routes do not. In the suburban areas, where service isless frequent, the system is arranged around a series of major transit exchanges, often located atshopping centres. Local service operates between the surrounding suburban areas and the exchanges,most of which have express service to downtown Vancouver. There is also local or express servicesbetween some of the transit exchanges. Figure 4.5 shows Greater Vancouver's transit system.SkyTrain, Vancouver's rapid transit line, forms the backbone of the system. The mostlyelevated line runs 25 km from downtown Vancouver, through Burnaby and New Westminster, toSurrey. SkyTrain is closely integrated with the bus system; the majority of the 175 bus routes in thesystem connect with one or more stations. Many of the suburban transit exchanges are located atSkyTrain stations, with SkyTrain providing the express service.CASE STUDY: BACKGROUND INFORMATION /44Figure 4.5 Greater Vancouver's Transit SystemSkyTrain RouteSeaBus RouteBus RoutesCASE STUDY: BACKGROUND INFORMATION /45The transit system is strongly oriented to downtown Vancouver. This results in fairly goodservice for those suburb-to-suburb trips which follow traditional radial routes: the best example of thisis the very high level of service for suburb-to-suburb trips along the SkyTrain line. However, for themajority of suburb-to-suburb trips, those which follow a more circumferential pattern, the service isgenerally poor. These types of trips are much more difficult to serve since they have such widelydispersed origins and destinations, and they cannot easily "piggyback" on the downtown-orientedservices.Transit service in Surrey is focussed on the Scott Road SkyTrain Station. From here,SkyTrain provides excellent express service to parts of New Westminster, Burnaby, and Vancouver.An extension of SkyTrain to Whalley will go into operation in 1994. Local bus routes in Surreyconnect with at least one of the six transit exchanges located at the town centres. More frequentservice connects the exchanges with each other and with Scott Road Station. Bus routes from NorthDelta and Langley also operate to some exchanges and to Scott Road Station, providing transitconnections between Surrey and these adjacent suburban areas. To make these connections, onegenerally must first travel to one of the transit exchanges. As mentioned, the best suburb-to-suburbtransit service from Surrey is to New Westminster and South Burnaby, along the SkyTrain route.Connections to other suburban areas generally involve circuitous routes or inconvenient transfers.Figure 4.6 details the transit system in Surrey.4.5 Changing Commuting Patterns in Greater VancouverBefore looking at the specific case of Surrey, it is helpful to first put the situation in ahistorical and region-wide context by examining the changing travel patterns and increasing suburb-to-suburb commuting in Greater Vancouver as a whole over the past two decades.Delta14:1154 355 3565/ 153SI 390.45 312 115 11/ .9 1/0 321'124 125 139 390 341 IL391 395_.I ./.6^IvlON EAC33 GII 332. 323.1249. 334 33I 340390 39330770.4.0 311 312 314.316.11 115,122 340 391640 ON PIMA^64 AK Cloverdale340^ 140130^31-7,1MUD BAYA UM.351 15244.1.3t DaWHITE ROCKSE MIAHMOO BAY 3514 354CASE STUDY: BACKGROUND INFORMATION /46Figure 4.6 Transit Routes in SurreyCASE STUDY: BACKGROUND INFORMATION /47Like most North American metropolitan areas, Vancouver has experienced rapid growth inboth suburban population and employment, leading to large increases in suburb-to-suburb commuting.Between 1971 and 1981, the share of regional employment in Vancouver's suburbs increased from 41to 51 percent, and most of this increase occurred in the outer suburban areas (GVRD 1985). Theemployment share of the inner suburban areas increased slightly, from 29 to 34 percent, but all of thisincrease occurred in Richmond, while Burnaby/New Westminster and the North Shore maintained astable proportion of regional employment. By contrast, employment in the outer suburban areasincreased from 14 to 21 percent of the regional total. GVRD estimates indicate that these trendscontinued during the 1980s, with suburban areas accounting for 58 percent of regional employment in1991. Of this, the inner suburbs accounted for 34 percent while the outer suburbs held 25 percent ofthe region's jobs. Figure 4.7 illustrates these changes.Figure 4.7 Employment Distribution, 1971-1991Source: GVRD, 1992.CASE STUDY: BACKGROUND INFORMATION /48Table 4.2 highlights another important trend during this period: an increasing tendency forpeople to live in one municipality and work in another. This type of commute grew at nearly threetimes the rate of commutes within a single municipality, and accounted for nearly 70 percent of worktrip growth. The region is thus becoming more economically integrated. Obviously, this rise in longdistance commuting has serious implications for transportation planning due to the increased pressurethis puts on highways and other transportation infrastructure.Table 4.2 Inter-Municipal Commuting Flows in Vancouver CMA, 1971-81.1971^1981^Increase^% Increase1971-81 1971-81Live and work in same municipalityCommute to different municipalityTOTAL204795 266730 61935 30.2%171195 315365 144170 84.2%375990 582095 206105 54.8%Percent Distribution^% of Total Flows^% of Increase1971^1981^1971-81Live and work in same municipalityCommute to different municipalityTOTALSource: Statistics Canada54.5% 45.8% 30.1%45.5% 54.2% 69.9%100.0% 100.0% 100.0%Types of work trips are further classified according to origin and destination in Table 4.3.Trips with suburban destinations, dominated by suburb-to-suburb trips, were clearly the fastestgrowing types, increasing at nearly twice the rate for all work trips. Suburb-to-suburb trips becamethe dominant type of commute during this period, increasing their share of total commuting flowsfrom 35 to 44 percent while internal Vancouver trips dropped from 36 to 26 percent. Meanwhile,"traditional" commutes from the suburbs to Vancouver grew at roughly the same rate as for all worktrips and maintained a constant share of about 22 percent. Suburb-to-suburb trips accounted for 61percent of commuter growth in the region between 1971 and 1981.CASE STUDY: BACKGROUND INFORMATION /49Table 4.3 Commuting Flows in Vancouver CMA, 1971 -81.Commuting Flows^1971^1981^Increase^% Increase1971-81 1971-81Internal VancouverSuburbs-VancouverVancouver-SuburbsSuburb-SuburbTOTALPercent Distribution136255 150735 14480 10.6%83455 125470 42015 50.3%24450 47560 23110 94.5%131830 258330 126500 96.0%375990 582095 206105 54.8%% of Total Flows^% of Total Increase1971^1981 1971-81Internal VancouverSuburbs-VancouverVancouver-SuburbsSuburb-SuburbTOTAL36.2% 25.9% 7.0%22.2% 21.6% 20.4%6.5% 8.2% 11.2%35.1% 44.4% 61.4%100.0% 100.0% 100.0%Source: Statistics CanadaVancouver appears to be fairly typical of large North American metropolitan areas in terms ofsuburb-to-suburb commuting. Such trips accounted for 45 percent of all commutes in U.S.metropolitan areas with more than 1 million people in 1980 (Pisarski 1987, 40). Among Canada'sthree largest metropolitan areas, suburb-to-suburb trips made up 48 percent of the total in 1981(Statistics Canada 1983).Figure 4.8 shows the transit mode split for the different types of commuting flows, based onthe Metropolitan Vancouver Origin -Destination Survey (GVRD 1987). This survey will be used for amore in-depth analysis of Surrey's suburb-to-suburb commuting patterns in the next chapter. The keypoint to note from the figure is the very low transit use for suburb-to-suburb work trips: only 4percent of such commutes were on transit compared with 13 percent for all work trips and 26 percentfor work trips within the City of Vancouver. Thus, as Figure 4.9 illustrates, suburb-to-suburbcommutes accounted for 46 percent of all work trips, but they made up less than 15 percent of transitwork trips. The City of Vancouver is, by far, transit's most important market in the region; it is thedestination for nearly 80 percent of transit work trips, and roughly half of all transit work trips haveCASE STUDY: BACKGROUND INFORMATION /50Figure 4.8 Transit Mode Split by Type of CommuteInternal VancouverSuburbs-VancouverVancouver-SuburbsSuburb-SuburbTotalT0%^5%^10%^15%^20%^25%^30%Transit Mode SplitFigure 4.9 Commuting Flow Types by Modeffl Internal Vancouver^TM Suburb to Vancouver .6 Vancouver to Suburb ^ Suburb to SuburbAll modes Transit100%90%80%70%60%50%40%30%20%10%0%CASE STUDY: BACKGROUND INFORMATION /51both origins and destinations within the City. As was shown in Table 4.3, however, the internalVancouver market is the slowest growing type of commute, and even the suburbs-to-Vancouvermarket is growing only half as fast as the suburb-to-suburb market. Thus, transit's growth potential inthese markets is limited.4.6 SummaryGreater Vancouver, like other North American cities, has seen a dramatic increase in suburb-to-suburb commuting over the past two decades. Similarly, Surrey, which plays a central role in theregion's commuting patterns, is typical of other suburban areas experiencing rapid population andemployment growth, along with the resulting transportation problems. The stage has now been set fora more detailed examination of suburb-to-suburb commuting in this key municipality.CHAPTER 5. CASE STUDY: RESULTS5.1 Introduction and Data LimitationsThis second part of the case study involves a more detailed examination of suburb-to-suburbcommuting to, from, and within Surrey. First some recent trends in the travel patterns of Surreycommuters will be reviewed. Work trip flow patterns and transit mode splits for different commutetypes will then be analyzed using data from the GVRD's Metropolitan Vancouver Origin DestinationSurvey. The effect of residential and employment densities on transit use in Surrey will also beinvestigated.Greater Vancouver has been divided into Traffic Zones (TZs), which are based on CensusTracts, and are used by the GVRD and its municipalities for transportation modelling and forecasting.In this part of the case study, data at the TZ level will be used for Surrey and White Rock, while theother suburban municipalities have been grouped into the same sub-regions as in Chapter 4. Most ofthe TZs in built up areas are about 2.5 km2 , larger in lower density areas. The 55 Traffic Zones inSurrey/White Rock had an average population of 3400 people in 1985. Figure 5.1 shows a map ofSurrey Traffic Zones.The information from the 1985 survey is beginning to get dated, especially in light of rapidpopulation and employment growth in the region, as well as the introduction of SkyTrain and thereorganization of the area's transit system. The GVRD conducted a new origin-destination survey inthe fall of 1992 and data on employment location from the 1991 Census will be released in 1993.Unfortunately, these new data were not available in time for this study. A brief section updating someof the factors affecting commuting in Surrey has been included at the end of the chapter. While therehave been many changes, most of the basic relationships which are being investigated should remain52662-----Th \ j-659CASE STUDY: RESULTS /53Figure 5.1 Traffic Zones in Surrey65666366643655^ 652657645644 660646661.647658654646653649650 620 "-----..,.639621638 633 632637 640 634 631627642 622630641 629626636 635 625626624623708707703704706709700705701702CASE STUDY: RESULTS /54the same. This study can serve as a base to describe the situation in 1985, and offers the opportunityfor further study of the situation as conditions change.There are certain limitations in the data. The Origin-Destination Survey was based on asample of roughly 5 percent of the households in the Vancouver CMA. This sample size results in arelatively high margin of error for small disaggregate flows such as those between traffic zones. Thisis particularly a problem for transit flows, which tend to be much smaller than automobile flows.Wherever possible, data have been aggregated to reduce the margin of error. For this reason, theanalysis looks at 24-hour work trip flows; while AM of PM peak flows would provide moreinformation on the actual demand during these critical periods, the 24-hour data provide a largersample size with less sampling error, yet still results in a good picture of commuting patterns inSurrey. Finally, the data used were only for vehicle trips, so cycling and walking trips will not bedirectly considered in the analysis.5.2 Trends in Surrey Commuting PatternsFurther background to the 1985 situation in Surrey was provided by investigating recenttrends in commuting patterns between Surrey as a whole and other sub-regions in Greater Vancouver.Table 5.1 shows the employment location for Surrey residents in 1971 and 1981. Rapid employmentgrowth south of the Fraser River has resulted in a declining proportion of work trips that cross theriver to the traditional employment locations in Vancouver, Burnaby, and New Westminster. Theproportion of Surrey residents working in these municipalities declined from 45 to 35 percent duringthe period. By contrast, the share of Surrey residents working in their own municipality increasedfrom 40 to 43 percent, and the proportion working in the neighboring South of Fraser suburbs ofRichmond, Delta, and Langley increased from 9 to 16 percent. While Surrey's labor force doubledduring this period, work trips to Richmond, Delta, and Langley nearly quadrupled. Work trips toSouth of Fraser locations (including Surrey itself) accounted for nearly 70 percent of commuterCASE STUDY: RESULTS /55Table 5.1 Commuting Flows for Surrey Residents, 1971-81.Place of work^1971^1981^Increase^% Increase1971-81 1971-81SurreyVancouverBurnaby/New West.North ShoreNortheast SectorRichmondDeltaLangleyMaple RidgeTOTALPercent DistributionPlace of work:SurreyVancouverBurnaby/New West.North ShoreNortheast SectorRichmondDeltaLangleyMaple RidgeTOTAL12375 27660 15285 123.5%7240 13175 5935 82.0%6770 9490 2720 40.2%540 1230 690 127.8%1030 2150 1120 108.7%1115 4035 2920 261.9%1035 3325 2290 221.3%770 3275 2505 325.3%70 300 230 328.6%31060 64695 33635 108.3%% of Total Flows^% of Total Increase1971^1981 1971-8139.8%23.3%21.8%1.7%3.3%3.6%3.3%2.5%0.2%42.8%20.4%14.7%1.9%3.3%6.2%5.1%5.1%0.5%45.4%17.6%8.1%2.1%3.3%8.7%6.8%7.4%0.7%100.0% 100.0% 100.0%Source: Statistics CanadaTable 5.2 Commuting Flows by Mode for Surrey Residents, 1985.Place of Work^All Modes^% of Total^Public Transit^% of Transit^TransitFlows Flows^Mode SplitSurreyVancouverBurnaby/New West.North ShoreNortheast SectorRichmondDeltaLangleyMaple RidgeTOTALSource: GVRD, 1987.19304 36.7% 864 23.5% 4.5%11214 21.3% 1975 53.8% 17.6%8311 15.8% 452 12.3% 5.4%1036 2.0% 107 2.9% 10.3%2137 4.1% 29 0.8% 1.4%4409 8.4% 118 3.2% 2.7%3244 6.2% 60 1.6% 1.8%2689 5.1% 67 1.8% 2.5%294 0.6% 0 0.0% 0.0%52638 100.0% 3672 100.0% 7.0%CASE STUDY: RESULTS /56growth. The trend in Surrey seems to be an increasing dominance of circumferential suburb-to-suburbcommutes, while more traditional radial commutes to Vancouver and the inner suburbs are becomingrelatively less important.Table 5.2 shows employment locations for Surrey residents based on the Origin DestinationSurvey, including a breakdown by mode. Because of differences in data collection, the absolutenumbers are not directly comparable with those in Table 5.1, but the percentage distribution of totalflows is comparable. The most important thing to note from Table 5.2 is the generally low rate oftransit use by Surrey residents: the transit mode split for work trips was 7 percent, compared with 13percent for Greater Vancouver as a whole in 1985. Not surprisingly, transit use from Surrey wasmuch greater for trips to Vancouver (17.6 percent) than for trips to suburban destinations (4.0percent). While Vancouver was the destination for 21 percent of all work trips from Surrey, itaccounted for 54 percent of transit work trips made by Surrey residents. For work trips withinSurrey, the transit mode split was 4.5 percent and this was the second largest transit market (24percent of all transit work trips). Caution should be used when looking at the transit use betweenSurrey and some of the other suburban areas since inaccuracies may arise due to the small samplesize. Overall, 3.6 percent of work trips between Surrey and other suburban areas were by transit.Transit use was particularly low for fast-growing circumferential commutes: Delta, Langley, andRichmond were the destinations for 20 percent of all work trips from Surrey, but only 6.6 percent oftransit commutes from Surrey were to these areas.The other side of Surrey's commuting picture involves those commuters who work in Surreybut live elsewhere in the region. Table 5.3 shows the place of residence for people who worked inSurrey in 1971 and 1981. During this period, employment in Surrey increased by 2.5 times. Theouter suburban areas formed the fastest growing segment of Surrey's commuter market. Particularlyimpressive is the growth of commuters from Delta and Langley, which more than quadrupled, from2230 to 9325 during the decade. The proportion of Surrey workers commuting from these two areasCASE STUDY: RESULTS /57Table 5.3 Commuting Flows for Surrey Workers, 1971-81Place of Residence^1971^1981^Increase^% Increase1971-81 1971-81SurreyVancouverBurnaby/New West.North ShoreNortheast SectorRichmondDeltaLangleyMaple RidgeTOTAL12375 27660 15285 123.5%1205 2795 1590 132.0%1640 2960 1320 80.5%205 665 460 224.4%730 1920 1190 163.0%300 830 530 176.7%1335 5220 3885 291.0%895 4105 3210 358.7%110 480 370 336.4%18845 46680 27835 147.7%Percent DistributionPlace of Residence:^% of Total Flows^% of Total Increase1971^1981 1971-81SurreyVancouverBurnaby/New West.North ShoreNortheast SectorRichmondDeltaLangleyMaple RidgeTOTAL65.7% 59.3% 54.9%6.4% 6.0% 5.7%8.7% 6.3% 4.7%1.1% 1.4% 1.7%3.9% 4.1% 4.3%1.6% 1.8% 1.9%7.1% 11.2% 14.0%4.7% 8.8% 11.5%0.6% 1.0% 1.3%100.0% 100.0% 100.0%Source: Statistics Canada.Table 5.4 Commuting Flows by Mode for Surrey Workers, 1985.Place of Residence:^All Modes^% of Total^Public Transit^% of Transit^TransitFlows Flows^Mode SplitSurreyVancouverBurnaby/New West.North ShoreNortheast SectorRichmondDeltaLangleyMaple RidgeTOTALSource: GVRD, 1987.19304 55.7% 864 64.8% 4.5%2272 6.6% 106 7.9% 4.7%2531 7.3% 165 12.4% 6.5%419 1.2% 0 0.0% 0.0%1634 4.7% 42 3.1% 2.6%555 1.6% 0 0.0% 0.0%3730 10.8% 138 10.3% 3.7%3785 10.9% 0 0.0% 0.0%420 1.2% 19 1.4% 4.5%34650 100.0% 1334 100.0% 3.8%CASE STUDY: RESULTS /58increased from 12 to 20 percent while the proportion commuting from Vancouver and Burnaby/NewWestminster declined from 15 to 12 percent. Surrey residents held a majority of the jobs in themunicipality throughout the period, although the proportion did decline slightly.Table 5.4 is based on the Origin-Destination Survey and shows the place of residence forSurrey workers, broken down by mode. Transit use for work trips to Surrey is very low,considerably lower than for work trips from Surrey (3.8 versus 7.0 percent). Much of this differencecan be accounted for by the sharp contrast in transit use between trips to and from Vancouver. Whilethe transit mode split for work trips from Surrey to Vancouver was 17.6 percent, it was only 4.7percent for Vancouver-to-Surrey work trips. This might indicate that the very high employmentdensities in downtown Vancouver are more influential on transit use than are the higher residentialdensities in Vancouver. This issue will be explored in more detail later. Overall, most transit worktrips which ended in Surrey tended to be relatively short. Thus, 65 percent of these trips were madeby Surrey residents while the next largest sources were residents of nearby Burnaby/New Westminster(12 percent) and Delta (10 percent). Only 8 percent of transit work trips to Surrey were made byVancouver residents. Transit use to Surrey from other areas was negligible. This suburbanworkplace market offers tremendous growth potential, but also many challenges, for public transit.Table 5.5 looks at Surrey's commuting balance, the number of people commuting into Surreycompared with the number of people commuting out of Surrey for each sub-region. As of 1981,Surrey was a net exporter of jobs to Vancouver and the inner suburban areas of Burnaby/NewWestminster, the North Shore, and Richmond. By comparison, it had a fairly even commutingbalance with the Northeast Sector and was a net importer of workers from the other outer suburbanareas, particularly from Delta. These patterns clearly have important implications for transit planning.Transit service between Surrey and Richmond, for example, would need to focus on bringing Surreyresidents to their jobs in Richmond while service to Delta or Langley would need a greater emphasison Surrey as the place of employment.CASE STUDY: RESULTS /59Table 5.5 Commuting Balance in Surrey, 1981.Origin/Destination:^Commute^Commute^CommutingInto Surrey^Out of Surrey^BalanceVancouverBurnaby/New West.North ShoreNortheast SectorRichmondDeltaLangleyMaple RidgeTOTAL2795 13175 -103802960 9490 -6530665 1230 -5651920 2150 -230830 4035 -32055220 3325 18954105 3275 830480 300 18046680 64695 -18015Source: Statistics Canada.5.3 Commuting Patterns: Traffic Zone LevelThe Traffic Zone (TZ) data were used to provide a more detailed picture of commutingpatterns both within Surrey, as well as between other suburban areas and specific parts of Surrey.Four origin-destination matrices (trip tables) were constructed. For internal Surrey work trips, originand destination was recorded by TZ. For external suburb-to-suburb trips, the trip tables consisted ofall work trips between each Surrey TZ and each of the other seven suburban sub-regions (as definedin Chapter 4). Trips between Surrey and Vancouver were not considered. In each case there was onetable for all work trips and one for transit work trips.Table 5.6 summarizes these commuting patterns, including transit mode split. A total of53,536 suburb-to-suburb work trips were made to and from the 55 Surrey TZs. About 40 percent ofthe trips were internal, both beginning and ending in Surrey. Nearly one-tenth of these trips bothbegan and ended in the same TZ (intra-TZ). The remaining internal Surrey trips began in one TZ andended in another (inter-TZ). External suburb-to-suburb trips made up 60 percent of the total. Nearlytwo-thirds of these trips were from Surrey to other suburbs while just over one-third were from othersuburbs to Surrey.CASE STUDY: RESULTS /60Table 5.6 Summary of Surrey Commuting FlowsAll Trips^% of Total^Transit Trips^TransitFlows Mode SplitInternal Surrey TripsIntra-TZInter-TZExternal Suburban TripsSurrey-Other SuburbOther Suburb-SurreyTOTAL21014 39.3% 704 3.4%1984 3.7% 0 0.0%19030 35.5% 704 3.7%32522 60.7% 1185 3.6%20495 38.3% 837 4.1%12027 22.5% 348 2.9%53536 100.0% 1889 3.5%Source: GVRD 1985 Origin-Destination SurveyThe transit mode split was consistently low for all commute types. There was little differencein transit use between internal and external trips. Transit mode split was highest for trips from Surreyto other suburbs, while trips from other suburbs to Surrey had a significantly lower mode split. Alsoof note is the 0 percent transit mode split for intra-TZ trips. Transit use is very low for these shortsuburban trips (which would generally be less than a kilometre given the size of most TZs). Lowservice frequencies are likely an important factor: the waiting time would be considerably longer thanthe travel time for such short trips, making transit a very unattractive option. Many of these tripswould likely be made by walking or cycling and are thus not recorded in the data.The TZ-to-TZ trips made within Surrey display a highly dispersed pattern. While the 21,014internal Surrey work trips represent a large total flow of commuters, the number of trips between anygiven pair of TZs is generally small. Table 5.7 shows a distribution of such trips by size of flowbetween TZ pairs. Of course, the size of the TZs is quite arbitrary and this will affect the size of theflows; the point of Table 5.7 is simply to illustrate the dispersed nature of the commuting flows. Asshown, TZ-TZ flows of less than 50 commuters accounted for nearly 75 percent of total work tripswithin Surrey. Only 24 out of 3,025 possible TZ pairs had flows greater than 100. Figure 5.2 showstwo-way TZ pair flows of greater than 100 commuters, graphically illustrating the dispersed nature ofcommuting patterns in SurreyCASE STUDY: RESULTS /61Figure 5.2 Commuting Flows Between TZ Pairstwo or more trips)CASE STUDY: RESULTS /62Table 5.7 Size Distribution of Commuting FlowsBetween Surrey TZ PairsSize of Flow^Number^Number^Percentof Flows of Trips of Total Trips150 +100-14950-990-4910148129202105167453631565110.0%8.0%25.5%74.5%All Flows 3025 21014 100.0%Source: GVRD 1985 Origin-Destination Survey5.4 Nodal PatternsWhile commuting patterns in Surrey are highly dispersed, there are some clusters of TZswhich attract a greater concentration of work trip flows. Most of these TZs are located in and aroundthe five Surrey town centres discussed in Chapter 4. Table 5.8 shows the top 25 TZs when rankedaccording to the total number of suburb-to-suburb work trips made to and from each TZ. Of the 25TZs shown, 15 are located in the town centre areas; in fact, all of the first 12 TZs on the list are intown centres. Travel to, from, and between these nodes likely represents the most promising potentialmarket for transit since these are usually the largest and most concentrated flows. If more travel canbe routed through these nodes, they will become even more attractive for transit service. It istherefore helpful to examine the commuting patterns of these nodes in more detail. (In the discussionwhich follows, it should be noted that the terms node and town centre are used interchangeably.)For the purpose of this study, all TZs that include part of a town centre were groupedtogether. The five town centres are Whalley, Guildford, Newton, Cloverdale, and White Rock (whichincludes the South Surrey town centre as defined in Surrey's Official Community Plan). The towncentres are shown with their constituent TZs in Figure 5.3. This definition for the town centresresults in overbounding; areas of lower density residential and even vacant land are sometimesincluded depending on the TZ boundaries.CASE STUDY: RESULTS /63Table 5.8 TZs Ranked by Total Suburban FlowsTZ Internal Surrey Other Suburbs TotalFrom TZ To TZ From TZ To TZ Suburban Flows660 856 1774 747 1085 4462654 541 2086 371 1453 4451700 688 1594 418 320 3020701 1223 902 550 215 2890657 512 1173 212 578 2475641 568 497 763 570 2398661 579 714 612 466 2371655 854 389 860 116 2219640 275 1001 293 539 2108634 787 345 822 72 2026625 635 699 375 293 2002705 1002 470 426 89 1987646 810 266 738 135 1949659 683 132 1074 0 1889648 818 192 710 158 1878645 410 301 804 239 1754709 721 389 526 21 1657638 176 613 292 525 1606643 403 285 743 148 1579658 348 477 535 166 1526704 718 38 584 58 1398662 246 336 416 343 1341706 349 424 347 196 1316637 245 259 517 195 1216649 443 183 465 83 1174Source: GVRD 1985 Origin-Destination SurveyCASE STUDY: RESULTS /64Figure 5.3 Town Centres in SurreyDefined by Traffic ZonesCASE STUDY: RESULTS /65 'Table 5.9 summarizes some of the features of the five town centres. In 1985, these areascontained about 44 percent of Surrey's population and 58 percent of its jobs. Both population andemployment densities are considerably higher in the town centres than in the rest of Surrey. As willbe discussed in more detail later, this contributes to higher transit use for trips to and from the towncentres. The transit mode split for all work trips originating in the town centres was nearly 50percent higher than for the rest of Surrey. For trips going to the town centres the difference was evengreater: transit use was about twice as high as for work trips to other Surrey locations.Table 5.9 Summary of Surrey Town Centre CharacteristicsPopulation^Employment^Transit Mode SplitTotal 1985^Per km2^Total 1985^Per km2^Ori in^DestinationWhalleyGuildfordNewtonCloverdaleWhite RockTotal Town CentresOther SurreyTotal Surrey19874 2322 6772 791 11.4% 5.2%12971 1782 4457 612 8.6% 5.5%17251 1887 3516 385 10.6% 3.9%8696 1689 1415 275 0.0% 0.0%24201 1566 4200 272 11.2% 2.4%82993 1821 20360 447 9.6% 4.1%104292 395 14521 55 6.7% 2.0%187285 605 34881 113 8.0% 3.2%Source: GVRD 1985 Origin-Destination Survey5.4.1 Internal Surrey TripsThe town centres form nodes which are very important in the overall commuting patternwithin Surrey, as shown in Table 5.10. Nearly one-third of commutes within Surrey both began andended in these nodes. Of these "nodal" trips, over half were made within a single node (intra-nodal)while the remainder were made between nodes (inter-nodal). Nearly half of all internal commuteswere between the nodes and other parts of Surrey. Of these, about two-thirds were directed to thenodes while a third were from the nodes. The remaining commutes, which represented only aboutone-fifth of Surrey work trips, had neither an origin nor a destination in the nodes.CASE STUDY: RESULTS /66Table 5.10 Nodal Commuting Flows within SurreyFlows:NodalIntra-nodalInter-nodalBetween Nodes and Non-nodesNon-nodes to NodesNodes to Non-nodesNon-nodalTotal Internal FlowsAll Trips^% of Total Transit Trips^TransitFlows^Mode Split6893 32.8% 361 5.2%3861 18.4% 38 1.0%3032 14.4% 323 10.7%9765 46.5% 255 2.6%6350 30.2% 232 3.7%3415 16.3% 23 0.7%4356 20.7% 88 2.0%21014 100.0% 704 3.4%Source: 1985 GVRD Origin-Destination SurveyTable 5.10 also includes information on transit mode split. The transit share for nodalcommutes was about twice as high as for other commute types. The 5.2 percent mode split for transitis still very low, however, given the advantages transit has in this market (large, relatively non-dispersed flows) and indicates further potential for transit. Another important point to note fromTable 5.10 is the much higher transit share for inter-nodal trips compared with intra-nodal trips (10.7versus 1.0 percent). A similar argument can be made here as for the intra-TZ trips: transit use is verylow for these short trips because of the ratio of access and waiting time to travel time. These shorttrips could represent a critical new market for transit and will be discussed in more detail later. It isnot surprising that non-node to node trips have a much higher transit mode split than node to non-node trips (3.7 versus 0.7 percent).Table 5.11 shows the total flows between the five town centres. Whalley is most important asboth an origin and destination. The largest inter-nodal flows (200+ commuters) were to Whalleyfrom Guildford, Newton, and White Rock, and from Whalley to Guildford and Newton. These largeflows represent an excellent niche for transit. White Rock has a much greater proportion of intra-nodalflows than the other town centres, indicating perhaps it is less integrated with the rest of Surrey thanare the other nodes.CASE STUDY: RESULTS /67Table 5.11 Commuting Flows Between Town CentresTo:^Whalley^Guildford^Newton^Cloverdale White Rock^All TownFrom:WhalleyGuildfordNewtonCloverdaleWhite RockAll Town CentresCentres734 314 288 22 45 1403439 573 148 42 56 1258446 134 483 95 83 124119 95 38 291 38 481329 187 133 81 1780 25101967 1303 1090 531 20621 6893Source: GVRD 1985 Origin-Destination Survey5.4.2 Trips Between Surrey and Other SuburbsThe flows between the nodes and other suburban areas also represent important markets fortransit. Table 5.12 shows that about 41 percent of commutes from Surrey to other suburbs began inthe nodes, while about 54 percent of other-suburb-to-Surrey trips ended in the nodes. Among trips toother suburbs, the transit share was much higher for those originating in the nodes (7.3 versus 1.9percent). Among trips from other suburbs, however, the transit share was slightly lower for trips tothe nodes (2.7 versus 3.1 percent). This is very surprising but could be due to sampling error.Table 5.12 Commuting Flows Between the Nodes and Other SuburbsAll Trips^Node/Non-node^Transit Trips^TransitDistribution Mode SplitNodes to Other SuburbsNon-node to Other SuburbsTotal Surrey to Other SuburbsOther Suburbs to NodesOther Suburbs to Non-nodeOther Suburbs to Surrey8348 40.7% 609 7.3%12147 59.3% 228 1.9%20495 100.0% 837 4.1%6448 53.6% 176 2.7%5579 46.4% 172 3.1%12027 100.0% 348 2.9%Source: GVRD 1985 Origin-Destination SurveyTable 5.13 shows the flows from the seven suburban sub-regions to the five town centres inSurrey. Delta and Langley are by far the most important sources of commuters to Surrey, and theTable 5.13 Commuting Flows From Other Suburbs to SurreyTo:^Whalley^Guildford^Newton^Cloverdale^White Rock^All Nodes^Non-nodes Total SurreyFrom:North ShoreBurnaby/New West.Northeast SectorRichmondDeltaMaple RidgeLangleyTOTAL56 73 65 0 18 212 203 415387 292 166 0 46 891 1185 2076281 423 65 16 79 864 700 1564109 46 42 0 55 252 176 428780 325 535 81 415 2136 1483 361956 89 59 16 0 220 163 383626 453 347 240 207 1873 1669 35422295 1701 1279 353 820 6448 5579 12027Source: GVRD 1985 Origin-Destination SurveyTable 5.14 Commuting Flows From Surrey to Other SuburbsFrom:WhalleyGuildfordNewtonCloverdaleWhite RockAll NodesNon-nodesTotal SurreyTo: North Shore^Burnaby/^Northeast^Richmond^Delta^Maple Ridge^Langley^TOTAL^New West.^Sector114 1161 287 193 195 24 132 210686 700 366 147 99 0 226 162444 766 322 491 406 19 155 22030 289 76 157 40 0 112 67441 501 37 627 205 0 330 1741285 3417 1088 1615 945 43 955 8348636 5028 1082 2289 1342 109 1661 12147921 8445 2170 3904 2287 152 2616 20495Source: GVRD 1985 Origin-Destination SurveyCASE STUDY: RESULTS /69town centres are the destination for 50-60 percent of these commuters. Some of these flows are quitelarge and could represent excellent potential transit markets. For example, the flows from Delta andLangley to Whalley each consist of 600-800 commuters. It must be noted, however, that the triporigins within the sub-regions are likely to be quite dispersed, as is the case in Surrey. Notsurprisingly, the largest flows to each town centre are from those sub-regions with the besttransportation access to that centre. For example, Delta is the most important source of commuters toWhalley, Newton, and White Rock, while Langley is the most important source of commuters toGuildford and Cloverdale. Burnaby/New Westminster and the Northeast Sector are the next mostimportant sources of commuters. Burnaby/New Westminster commuters appear to be less oriented tothe nodes than those from other sub-regions. The primary nodal destination for Burnaby/NewWestminster commuters was Whalley, for which there are already quite good transit connections. Theprimary destination for Northeast Sector commuters was Guildford (it is closest to the Port MannBridge which connects Coquitlam to Surrey). This was one of the largest flows and represents apotential market which currently has poor transit service. The North Shore, Richmond, and MapleRidge are generally unimportant sources of commuters.Table 5.14 shows the commuting flows from the town centres to the suburban sub-regions.By far the most important single destination was the more traditional radial commute to Burnaby/NewWestminster, with its fairly good transit service from Surrey. However, as we have seen, this is oneof the more slowly growing commuter markets. Richmond, which was the next most importantdestination, is a much faster growing commuter market and it has much poorer transit connections.There were particularly strong flows to Richmond from White Rock and Newton. The NortheastSector was also an important and rapidly growing destination for Surrey commuters which again haspoor transit connections. Most trips to the Northeast Sector were from North Surrey, with Guildfordbeing the most important origin. Routes from South Surrey to Richmond and from North Surrey tothe Northeast Sector might be good candidates for improved transit service. Of course the problemCASE STUDY: RESULTS /70again arises that destinations in Richmond, the Northeast Sector, as well as the other sub-regions arealso likely to be quite dispersed.5.5 Effect of Density on Transit UseThe literature contains numerous studies examining the positive effect of urban densities ontransit use. However, in looking at this relationship, most of these studies have compared suburbanareas with the central city or the CBD. An examination of the effect on transit use of differentresidential and employment densities within a suburban area, in this case Surrey, is therefore valuablebecause it may help to identify what types of suburban development are most amenable to transit.The relationship between the residential density and transit use was investigated first, usinggross population density for each Surrey TZ and the transit mode split for work trips originating ineach zone. A linear regression was performed on the data, and transit use was found to be positivelycorrelated with residential density, although the relationship was fairly weak. The R-Squared(coefficient of determination) was found to be 0.342, meaning that only 34 percent of the variance intransit mode split can be explained by differences in residential density. By comparison, Pushkarevand Zupan (1977, 26) found that residential density could explain 57 percent of the variance in transituse. Figure 5.4 shows this weak positive relationship between these two variables, and theconsiderable variation of the points around the regression line.This weaker than expected correlation likely occurred for a number of reasons. The modesplit data are subject to considerable sampling error due to the small sample size for transit riders inindividual TZs. Another factor is the level of transit service, which obviously affects transit use.Some low-density TZs are bordered by well served transit routes, producing mode splits that arehigher than expected based on density alone. A major source of weakness in the data results from theuse of gross densities rather than net residential densities, which were not available. Most studiesCASE STUDY: RESULTS /71••^•AFigure 5.4 Population Density and Mode Split20%16%^•4%0%0^500^1000^1500^2000^2500^3000Population Density of Origin TZ (persons/km2)••• A•• At•••• •Figure 5.5 Employment Density and Mode Split•0%0^200^400^600^800^1000^1200^1400Employment Density of Destination TZ (jobs/km2)10%8%6%4%2%CASE STUDY: RESULTS /72have related transit use to net density, but this can vary for any given gross density depending on theamount of non-residential land in the TZ. For example, some TZs have a low gross density due tolarge tracts of vacant land, when in fact the majority of residents live in a small part of the zone at arelatively high net density. The three data points in Figure 5.4 which are furthest from the regressionline represent such TZs (638, 640, and 709).The relationship between transit use and population density is likely not a purely linear one.As was noted in Chapter 2, Pushkarev and Zupan found that there was a greater change in transit useat the upper and lower extremes of density. The Surrey data appear to support this. For tripsoriginating in TZs with residential densities below 500 persons/km 2 , the transit mode split was wellbelow the mean for Surrey (3.2 versus 7.0 percent). In fact, the mode split was 0 percent for 10 ofthe 17 TZs in this density range. In the middle range of densities (500 to 2000), the mode splitshowed much more variability, and there was little or no discernible linear relationship. Whendensities exceeded 2000, the mode split averaged 9.5 percent, and the data points were once againmore clustered around the regression line. Thus, transit use appears to be most sensitive to populationdensity at the upper and lower extremes.Next, the relationship between gross employment density, as measured by jobs/km2 for eachTZ and the transit mode split for work trips to each zone, was investigated. Linear regressionanalysis of the data indicates that there is a positive correlation between transit use and employmentdensity. Figure 5.5 shows the relationship. The R-Squared was found to be 0.535, meaning thatemployment density can explain about 54 percent of the variance in the transit mode split. Thus thecorrelation with employment density appears to be stronger than with population density. This agreeswith the finding earlier in this chapter that transit use was much higher for work trips from Surrey toVancouver than vice versa. The effect on transit use of Vancouver's relatively high employmentdensities was greater than the effect of its relatively high population density.CASE STUDY: RESULTS /73As with population density, the relationship between transit use and employment densityappears to be stronger at the extremes of density. Transit mode split was extremely low foremployment densities below 200 jobs/km2 (0.5 versus an average of 3.8 percent for all work trips toSurrey), and 27 of the 30 TZs in this range had a 0 percent transit mode split. The increase in transituse at higher densities was not quite so dramatic, with a mode split of 5.1 percent for TZs withemployment densities above 500. The same sources of error discussed for population density apply tothe employment density data as well. Employment density is even more susceptible to the errorresulting from the gross densities, since employment is rarely spread evenly throughout a TZ.These observed relationships generally agree with those of Pushkarev and Zupan who werecomparing downtowns with suburban areas. The case study findings support the idea of thresholddensities, above or below which there can be significant changes in transit use. In the case of Surrey,the lower threshold is most apparent, resulting in the most significant changes in transit use, with themode split dropping sharply at population densities below 500 persons/km2 and at employmentdensities below 200 jobs/km2 . Increases in transit use above the upper threshold were not as great asthose observed by Pushkarev and Zupan, likely because there are few areas in Surrey with highpopulation or employment densities. The case study also agreed with Pushkarev and Zupan in thefinding that employment density had a greater effect than population density on transit use.It is important to note that at least part of the difference in observed transit use can beexplained by the greater level of transit service already in place in areas with higher population andemployment densities. However, these high service levels are largely a result of the higher densitieswhich made the transit service feasible in the first place; this relates to the positive feedback loopdescribed by Pushkarev and Zupan in Chapter 2.CASE STUDY: RESULTS /745.6 Surrey in 1993: An UpdateSurrey has continued to experience rapid population and employment growth. The populationof the Surrey/White Rock area increased from 187,285 in 1985 to 261,487 in the 1991 Census, anincrease of 40 percent in just six years. Employment increased from 57,000 in 1986 to an estimated71,000 in 1991 (GVRD 1992). There has been rapid development at the town centres. Whalley inparticular is poised for spectacular growth with the arrival of SkyTrain in 1994. Growth has alsobeen very rapid in Newton, which serves the area where much of the residential growth has beenconcentrated. There have also been some important changes in the transportation system between1985 and 1993. The major change in the road system has been the addition of the Alex Fraser Bridgeand Highway 91, providing much better connections from the Newton area to Richmond and to NewWestminster. The major change in the transit system has been the addition of SkyTrain. This hasbeen accompanied by improved and reorganized local service in Surrey to connect with SkyTrain.5.7 SummaryThis detailed investigation of suburb-to-suburb commuting in Surrey has resulted in someimportant findings. It has shown just how highly dispersed these work trips are, yet it has alsorevealed that there are some flows within and between the nodes which are more concentrated, andoffer the most promising market for transit. Transit use was relatively high for the inter-nodal flows,but there is clearly room to expand transit's role. For the intra-nodal flows, transit use was extremelylow; this represents a key market with great potential for improved transit service. There is alsopotential for better transit to serve fast-growing circumferential commutes between Surrey and othersuburban areas like Richmond and the Northeast Sector. Finally, this investigation has shown that,even within a suburban area, there are important relationships between transit use and differences indensity, particularly employment density.CHAPTER 6. CASE STUDY: THE RESPONSE IN SURREY6.1 IntroductionThis final component of the case study will examine how Surrey, the GVRD, and BC Transithave responded to the challenge of promoting transit use for suburb-to-suburb commuting. Publicattitudes toward transit in Greater Vancouver, which could provide some important insights on themost effective strategies to adopt, will also be reviewed. The information presented here shouldcomplement the quantitative analysis of the previous chapter, and help to create a clearer picture of thestate of suburb-to-suburb transit service in Surrey.6.2 The Response by the MunicipalitySurrey's efforts to promote transit use have mostly been through land use strategies (SurreyPlanning Department 1986 and 1992). In particular, Surrey plans to actively promote its town centresin order to prevent further sprawl and encourage a denser development pattern that is more pedestrianoriented and more easily served by transit. The Official Community Plan calls for net residentialdensities in the town centres to range from 30 to 330 units per hectare, compared with the very lowdensities of around 10 units per hectare in a typical single family area in Surrey.Surrey is also beginning to promote transit and pedestrian friendly design features. Thesefeatures are particularly evident in the design for Surrey's major town centre, Whalley, which is beingdeveloped along the SkyTrain route. Figure 6.1 shows the plan for Surrey City Centre (Whalley'snew official name). It consists of high density commercial and residential development clusteredaround the SkyTrain stations. An important feature of the plan will be the creation a pedestrianoriented shopping spine along 135th Street which aims to move the focus of the community away fromthe automobile strip development along the King George Highway.75CASE STUDY: THE RESPONSE IN SURREY /76Figure 6.1 Plan for Surrey City CentreSource: Whalley Town Centre Plan, Surrey Planning Department, 1992.CASE STUDY: THE RESPONSE IN SURREY /77The municipality is also looking at ways to improve transit service between the town centres(Dickenson 1992). An HOV lane along 104th Avenue is currently being proposed to provide fastertransit service between Whalley and Guildford, as well as to promote greater use of carpools. Thereare also a proposals for HOV lanes on the King George Highway connecting Whalley with Newtonand White Rock, and on the Fraser Highway connecting Whalley with Langley. The ultimate goalwould be a network of HOV lanes on the major arterials connecting Surrey's town centres.Currently, this initiative is at the conceptual stage. Surrey is also planning to reserve a right of wayfor any future rapid transit extension between Whalley and Guildford.Surrey is just beginning to consider transportation demand management and currently thereare no TDM strategies in place. One strategy being discussed would be to manage parking at the towncentres. Surrey presently negotiates with developers to reduce parking requirements if it is believedthat there will be substantial transit use to the development. It is also looking at restricting on-streetparking in some areas to residents only (Lai 1993). The parking restrictions that Surrey isconsidering are meant to target commuters. However, there is concern, particularly among businessowners, that any restrictions might discourage people from using the town centres. There are stillconflicting views on this among municipal staff and the idea is at a very early stage of consideration(Dickenson 1992).Only recently has Surrey even begun to face the challenge of encouraging greater transit use.The municipality has focussed largely on land use strategies, but even here the work is still mostly atthe planning stage, with little substantive progress to date. Other strategies such as TDM are only justbeing considered. As Surrey's transportation problems worsen, a more comprehensive approach willlikely be required.CASE STUDY: THE RESPONSE IN SURREY /786.3 The Response by BC TransitBC Transit is very aware of the increasing importance of suburb-to-suburb commuting and hasbeen attempting to deal with the issue. However, funding limitations mean that it has had to movemore slowly than it would like to in implementing any solutions (Kobayakawa 1992).One imminent improvement to Surrey's transit system is the extension of SkyTrain from ScottRoad to Whalley in 1994. This will greatly improve transit service to Surrey's major town centre,giving it more direct bus connections with other parts of Surrey, Delta, and Langley, as well as arapid transit link with other regional town centres like Burnaby's Metrotown or downtown NewWestminster. BC Transit also plans improved east-west services in North Surrey and North Delta, aswell as better connections between town centres, as part of this reorganization of Surrey's transitsystem (BC Transit 1992B).As was noted in Chapter 4, transit service in Greater Vancouver is predominantly orientedtowards downtown. This orientation also allows for reasonably good service for those suburb-to-suburb commutes which follow the traditional radial routes in the direction of downtown Vancouver,such as from North Surrey to Burnaby, but service is much poorer for circumferential routes. Serviceon these routes is generally slow, indirect, and inconvenient. A typical Newton area resident, forexample, would take over an hour and a half and require two transfers to go by transit to eitherCoquitlam Town Centre or Vancouver International Airport.There are some circumferential routes connecting Surrey with other suburban areas, as shownin Figure 6.2. One of these connects Surrey with Simon Fraser University via Burnaby's LougheedMall. There are three routes from Surrey, through North Delta, to 22nd Street SkyTrain Station inNew Westminster. Another two routes connect Surrey with South Delta. Service levels on all theseroutes are low, often just a few peak hour trips per day. Ridership is also generally low. BC TransitCASE STUDY: THE RESPONSE IN SURREY /79Figure 6.2 Circumferential Transit Routes in SurreyService to: Service Level(Peak) (0ff Peak)Regular Service:340 Cloverdale/Newton/22nd St. Delta, New Westminster 30 min. 60 min.640 Scottsdale MaIVLadner/Tsawassen Delta 60 min. 80 min.Limited Service:314 Scottsdale Mall/22nd St Delta. New Westminster 8 trips/day none315 Scott Road/Ladner Delta 4 trips/day none145 Scott Road/SFU Burnaby, Coquitlam 9 trips/day noneFuture Service:313 Scottsdale Mall/Richmond Centre Richmond329 Whalley/North Delta/22nd St. Delta, New Westminster640 Scott Road/Ladner/Tsawassen DeltaSource: BC Transit 1992-1993 Annual Service Plan.CASE STUDY: THE RESPONSE IN SURREY /80is planning to improve service on some routes that serve circumferential commutes, as well as addnew routes. However, other circumferential routes may be cut due to poor ridership. One proposednew route would be from Scottsdale Mall to Richmond Centre. This would be the first route todirectly connect Surrey and Richmond, and tap this large commuter market.BC Transit does provide inter-nodal service within Surrey. As discussed in Chapter 4, thenetwork is designed around a series of transit exchanges located at the town centres, using a timedfocal point transfer system. All bus routes in Surrey connect with at least one of these exchangesThere is a high level of service (15 minute headways) from Whalley to both Guildford and Newton,and this forms the core of Surrey's transit system. This core system is extended from Guildford toCloverdale, and from Newton to White Rock with 30 minute service. This is shown in Figure 6.3.Other pairs of nodes, such as Guildford-Newton or Guildford-White Rock, have much poorer servicedespite relatively high demand levels. None of these inter-nodal routes is an express route, they allprovide local service for points between the nodes. Inter-nodal routes which are not part of the coresystem tend to be indirect.BC Transit (1992A) is also considering expanding its use of paratransit. It presently operatesa paratransit "community service" in Maple Ridge using minibuses. These smaller vehicles bettermatch the lower demand levels in low density suburban areas (although cost savings are not alwaysthat great). They are also less intrusive than standard buses, which are not welcomed in manysuburban neighborhoods. These services emphasize comfort and convenience over speed. They stopdirectly in front of major destinations. BC Transit has identified several areas throughout GreaterVancouver where community services could be implemented, as shown in Figure 6.4. These servicesfall into two general types. One type could be used in low density suburban areas where demand doesnot warrant full service. These might operate as feeder services in these areas, connecting them tomajor transit routes. White Rock/South Surrey is a potential candidate for this type of service.Community services could also be used as circulators within town centres. Whalley is one of the areasCASE STUDY: THE RESPONSE IN SURREY /81Figure 6.3 Inter-nodal Transit Service in SurreyCASE STUDY: THE RESPONSE IN SURREY /82Figure 6.4 Potential Locations for Paratransit Servicein Greater VancouverSource: BC Transit 1992-1993 Annual Service Plan.CASE STUDY: THE RESPONSE IN SURREY /83being considered. This would help address the problem of getting around the town centres without anautomobile.In 1990, the Go Green program was initiated by BC Transit, with involvement from theFederal and Provincial Environment Ministries, as well as the Ministry of Transportation andHighways and the GVRD. The initial goal was to increase public awareness of the environmentalproblems associated with the automobile. This modest goal has been achieved mostly through anadvertising campaign (Hodgson 1992).An additional goal for Go Green - and one that will be more difficult to achieve - is to useTDM strategies to actually change travel behavior and reduce the use of single occupant vehicles. BCTransit and the GVRD are both considering various TDM strategies. One strategy involves a seriesof Employee Transportation Administration Seminars to teach representatives from regional employersabout the basics of TDM, so they can then design a trip reduction plan for their companies.Eventually, major employers may be required to submit trip reduction plans. Ridesharing is one wayto reduce trips, and this is being promoted through the Go Green Program. The initial focus has beento work with major employers in the region to help them set up car and vanpool programs. A carpoolprogram to Scott Road Station is also being considered. In the longer term, a regional carpoolmatching service may be established. While BC Transit, like Surrey, is beginning to consider TDMstrategies, actual implementation of these strategies has been minimal thus far (Smith 1992).BC Transit's role is evolving beyond that of merely a transportation provider and there is anattempt being made to develop a more integrated transportation and land use planning process in theregion. BC Transit has been working with the GVRD, the municipalities, and the provincialgovernment on Transport 2021 to develop a long range transportation plan. This plan will look notonly at new facilities, but will also assess the impact of various land use and TDM strategies. Forexample, one analysis using the GVRD's transportation planning model predicted that a package ofCASE STUDY: THE RESPONSE IN SURREY /84seven TDM measures could produce a nearly 50 percent rise in the morning peak hour transit modesplit when compared with a no-TDM base case for the year 2021 (GVRD 1993).6.4 Public Attitudes Towards Transit ServiceInvestigating public attitudes to transit may help explain why transit use is low for suburb-to-suburb trips, even when service is provided. It may also indicate the type of changes that must bemade to convince suburban commuters to use transit. BC Transit's Usage and Attitude Survey (1992C)was a telephone survey of 1,979 people designed to gather information on public attitudes abouttransit. The results are discussed below.Of those surveyed, 48 percent felt BC Transit service was quite good or very good. Anadditional 35 percent felt service was satisfactory. The main reasons people had for using transit werelack of alternative transportation (32 percent) and not wanting to worry about traffic (23 percent). Inthe Southeast suburbs (Surrey, Delta, and Langley), not wanting to worry about traffic, most likelyinvolving a commute to downtown Vancouver, was the most important reason for using transit (34percent). This indicates that worsening traffic conditions for suburb-to-suburb trips could significantlyimprove the attractiveness of transit. Longer suburb-to-suburb commutes would be more susceptibleto this.People were asked what factors would most influence them to use transit. Rapid transitservice was the greatest motivator (66 percent rated this 4 or 5 out of 5). Other service improvementssuch as shorter travel times, express bus service, fewer transfers, and more frequent service also ratedhighly (58-60 percent). Next most important was helping the environment (57 percent), followed byincreased traffic congestion (56 percent). Higher parking, gas, or insurance costs rated lower (44-50percent). These results should be viewed with caution, however. While they seem to indicate thatgreater efficiency and speed of the transit system are better motivators than increased costs of driving,CASE STUDY: THE RESPONSE IN SURREY /85many studies have found that survey responses can often be very different from actual behavior. Forexample, actual use of new transit facilities has been found to be only one-third to one-fifth ofintended use (Couture and Dooley 1981). Because people tend to overstate their intentions to useimproved transit services, measures which penalize automobile use would probably have a greaterimpact relative to service improvements than these results would indicate.The survey asked about the importance of different transit service attributes and found thatservice frequency was most critical (77 percent rated it as important). On average, respondents feltthat 12-14 minutes was a "reasonable" time to wait for a bus, 9-11 minutes when transferring. Thenext most important service attributes were travel time/speed (69 percent) and number of transfersneeded (65 percent). Distance to bus stop (47 percent) and seat availability (42 percent) wererelatively less important. Figure 6.5 compares the response in the Southeast suburbs versus the Cityof Vancouver. Frequency of service was relatively less important in the Southeast suburbs while seatavailability, distance to bus stop, and the number of transfers was more important. This latter factorreflects the problem of the lack of direct service for many suburb-to-suburb trips.Suburban residents had quite different attitudes than city residents about the existing transitservice in their neighborhoods, as shown in Figure 6.6. Southeast residents were much less likelythan Vancouver residents to feel that service frequencies were adequate, and only half as manySoutheast residents could reach their destination without transferring. Not surprisingly, moreSoutheast residents felt that transit was less convenient than driving. However, Southeast residentswere more likely than Vancouver residents to say that transit was faster than driving during rush hour(although this likely applied only to suburb-to-downtown trips). Southeast residents were also slightlymore likely to agree that transit should be given priority over automobiles.In terms of planning options, the greatest support was for more rapid transit lines and moresuburb-to-suburb routes (73 percent for each). Support was lower for improved local area orCASE STUDY: THE RESPONSE 1N SURREY /86 Figure 6.5 Importance of Service AttributesService FrequencyTravel Time/Speed^ Southeast SuburbsNumber ofTransfersVancouverDistance to BusStopSeat Availability0^20^40^60^80^100Percent who rated attribute as importantFigure 6.6 Attitudes About TransitAdequate servicefrequencyCan travel with 1transfer or less^ Southeast SuburbsFaster than rushhour drivingVancouverLess convenientthan autoGive transit priority0^20^40^60^80^100Percent who agree with statementCASE STUDY: THE RESPONSE IN SURREY /87downtown service. There was also strong support (89 percent) for using other types of vehicles.Dial-a-ride was supported by 58 percent. These results indicate a strong level of support forimproved suburb-to-suburb transit service. A recent survey for the United Way (Pendakur andBowler 1987), looking at travel patterns for people living South of the Fraser communities, confirmsthis. People wanted better connections between the South of Fraser municipalities, especially toRichmond from Delta and Surrey. They wanted shorter waits for transfers or more direct busesroutes for these trips. Improving suburb-to-suburb service appears to be a much greater priority forsuburban residents than any improvements in downtown services, with which they are fairly satisfied.While there is clearly demand for improved service, it is not so clear whether people areprepared to pay for it. In the United Way survey, people wanted greatly improved service at lowercost. Similarly, in the BC Transit survey, support for transit improvements declined markedly ascosts increased. For example, while 86 percent supported the creation of bus lanes, this dropped to60 percent if it meant replacing a lane of automobile traffic. In the Southeast, support dropped evenmore, from 86 to 52 percent. This represents a major dilemma for transit planners.The results of the attitude survey provide some useful information for transit planners on thetypes of service improvements which would attract more suburban riders. A major theme was thedesire for faster, more frequent, and more direct service between suburban areas. This points to theneed for some type of suburb-to-suburb express service. One possibility, which will be discussed ingreater detail in the next chapter, would be to consolidate some suburban transit routes into fewer butbetter served express corridors. The attitude survey also highlighted the problem that, in order to getthis improved transit service, people do not seem willing to make sacrifices such as higher costs orlonger distances to bus stops. This demonstrates how difficult it will be to solve the problems facingtransit in the suburbs.CASE STUDY: THE RESPONSE IN SURREY /886.5 SummaryThis review of the response in Surrey helps to provide a more complete picture of the suburb-to-suburb transit "problem" than that formed in Chapter 5. Transportation planners in Surrey and inGreater Vancouver have begun to take a more comprehensive approach to solving the problem ofproviding suburb-to-suburb transit service. Land use strategies have been formulated, innovativetransit and paratransit services are being planned, and TDM strategies are being considered. To date,however, very little has actually been implemented. A survey of public attitudes shows that peopleare reasonably satisfied with the downtown-oriented transit service, but there is great demand forimprovements in suburb-to-suburb service. The survey reveals why people use transit and what typesof service attributes they consider to be important. It can thus serve as a helpful guide for transitplanners. Unfortunately, it also shows that the desire for better service improvements does notnecessarily translate into a willingness to pay for these improvements.89CHAPTER 7 THE PLANNING FRAMEWORK7.1 IntroductionIn this chapter, the results of the case study will be used to help decide which strategies forimproving suburb-to-suburb transit service would be most effective for Surrey and other comparablesuburban areas. As earlier discussion has indicated, there is no single solution to this problem; acombination of improved transit services, land use planning measures, and TDM strategies will berequired. In the final part of the chapter, a conceptual framework for suburb-to-suburb transitplanning will be developed based on the Surrey case study. This is meant to provide a way ofapplying the case study results more broadly, to solve similar problems in other suburban areas.7.2 Strategies for Suburb-to-Suburb TransitThe most obvious way to promote suburb-to-suburb transit use is to improve transit service.The problems facing suburban transit indicate that merely increasing the levels of existing servicewould likely not be sufficient to make transit an attractive alternative to the automobile. What isneeded is a greater range of innovative transit service options. According to the "family of services"concept which was discussed in Chapter 3, different services occupy different market niches,providing an overall system that better meets the full range of suburban travel needs. Table 7.1,which is based on the Surrey case study, summarizes the potential suburban markets for transit, alongwith the type of transit service which would most effectively fill each market niche.THE PLANNING FRAMEWORK /90Table 7.1 Transit Options for Different Commute TypesCommute Type^Transit OptionsIntra -nodal (within the nodes) Paratransit - shuttle serviceInter-nodal (between thenodes)Conventional transit except paratransit wheredemand is low.Also express service - bus or paratransitdepending on the level of demand.Between the nodes andsurrounding residential areasConventional transit if density and demandwarrantsParatransit if density and demand are low.Between non-nodal areas Paratransit if density and demand warrantsBetween nodes and othersuburban areasExpress service between nodes and othersuburban nodes(conventional transit or paratransit depending ondemand)Long trips between differentsuburban areasParatransit such as ridesharingNo direct transit service7.2.1 Conventional TransitThe most promising market for conventional transit in the suburbs is the inter-nodal market,such as the travel flows between the five town centre areas in Surrey. These are the only suburbanflows which are concentrated at both trip ends and, as noted in Chapter 5, they displayed the highesttransit mode split of any trip type in Surrey (10.7 percent). The relatively high population andemployment densities in and around the nodes enhances their attractiveness as transit sub-markets.While this is a relatively small market, accounting for about one-seventh of all work trips withinSurrey in 1985, there is potential for growth as the nodes increase in size and density. In Surrey, thetransit system is already structured around the nodes, using timed focal point transfers, and parts ofthis market are served quite well (Whalley to Guildford and Newton). There is room for improvedservice levels, however, and this may be an area for BC Transit to expand its conventional service.THE PLANNING FRAMEWORK /91Providing service to the nodes from surrounding residential areas is another market whereconventional transit might be viable. These flows are concentrated at only one end (the node) so thepotential of this market is not so high as that of the inter-nodal market. At present, BC Transitattempts to provide conventional service to the nodes from nearly all areas of Surrey. Since the casestudy results confirmed that there is a significant increase in transit use when residential densitiesexceed 500 persons/km2 , any improvements in conventional transit service to the nodes should befocussed on areas with densities above this lower threshold. Areas with densities below 500 likelycannot support conventional transit service to the nodes, and alternative service options such asparatransit should be considered. In Surrey, about 30 percent of commutes were to the nodes fromsurrounding non-nodal areas, and roughly two thirds of these were from TZs with densities above 500persons/km2 .7.2.2 ParatransitParatransit is a more flexible option which could replace conventional transit service in certainsub-markets. One potential suburban market for paratransit, just mentioned, is providing service tothe nodes from areas which cannot support conventional service. Another potential market isproviding service within the nodes.Paratransit might be a more viable means than conventional transit for providing service fromsome low density areas to the nodes (where it could connect with conventional transit services). Mini-buses with flexible routing and stopping procedures might be one option. BC Transit is alreadyexperimenting with these types of services in Maple Ridge. As was noted in Chapter 3, however,paratransit often results in minimal savings when compared with providing conventional transitservice. One reason: labor costs, the largest component of transit expenses, are the same for a mini-bus as for a full size bus. Organized car and vanpools, which are well suited to serving commutesTHE PLANNING FRAMEWORK /92with relatively dispersed origins converging on a single destination, may be a better option. Theyhave a major advantage over other forms of paratransit in that their operating costs are very low, andthey make more efficient use of existing private vehicles. They also provide fast, comfortable "door-to-door" service for the passenger.A key potential market for paratransit could be serving trips within nodes in the form of afrequent internal circulator. The case study results highlighted the extremely low transit mode split forsuch trips; most people do not find it convenient to use transit for such short trips, given the ratio ofaccess and waiting time to the actual travel time. Increased service frequency is a critical requirementfor this market. Intra-nodal trips account for nearly one-fifth of all work trips made within Surrey, aproportion that could increase as the town centres grow. Another major role for a circulator would beto serve shopping, recreation, and personal business trips. This has the added benefit, discussed inChapter 2, of encouraging more commuters to use transit, since they would no longer need their carsto make short midday trips. To date, little has been done in Surrey to serve this market, so there isgreat potential.7.2.3 Express ServiceExpress transit service, which is used for suburb-to-downtown trips in many cities, includingGreater Vancouver, also has potential in the suburb-to-suburb market. This would help to address amajor complaint which suburban commuters voiced in BC Transit's Usage and Attitude Survey: thatsuburb-to-suburb transit is too slow, infrequent, and indirect. Almost by definition, express servicewould have to be inter-nodal since it requires a high level of demand, such as exists between thelarger town centres (e.g. Whalley-Guildford or Whalley-Newton). Presently, most inter-nodal routesin Surrey double as local routes. More express-only routes could improve the appeal of transit.These express routes could be integrated into the timed transfer system already in operation in orderto connect with local service (just as express services to downtown Vancouver presently do). AsTHE PLANNING FRAMEWORK /93Schneider and Smith pointed out in Chapter 3, such a system would concentrate even more of the totaltrips into heavily travelled corridors between the nodes, which could then support shorter headwaysbetween buses, longer hours of peak operation, and other improvements in service. Where demandbetween nodes is still quite low ( e.g. Guildford-Cloverdale), paratransit, including ridesharing, mightbe a viable means of providing express service.Since express service becomes more appealing with increasing trip length, longer commutesbetween different suburban areas (e.g. Surrey to Richmond) form another important market. Suchtrips accounted for more than 60 percent of Surrey based suburb-to-suburb work trips in 1985, andthis proportion has been increasing. While express service would mostly serve trips between thenodes in different suburban areas, links with local service at timed transfer points could be used to tapeven more of this large market. In some parts of Greater Vancouver, suburban inter-nodal expressservice is already provided in the form of SkyTrain, but express service could be provided betweenother nodes at far less cost using buses or paratransit. Potential markets for such express servicesmight include Whalley-Guildford-Coquitlam Centre or Whalley-Newton-Richmond Centre. The casestudy results indicated that demand on these routes was relatively high and growing rapidly. Wheredemand levels are lower, ridesharing could provide the express service. As noted in Chapter 3,ridesharing becomes especially attractive for people making long trips, since travel costs increase withdistance and passenger pick-up time decreases as a proportion of total travel time.7.2.4 Land Use StrategiesThe discussion in Chapter 2 indicated that existing suburban land use patterns are not at allconducive to high levels of transit use, often resulting in low ridership regardless of the servicesoffered. Strategies to change this land use pattern and make it more amenable to transit use, thus,play an important role. These land use strategies generally involve increasing residential andemployment densities, as well as improving suburban design.THE PLANNING FRAMEWORK /94There is general agreement in the literature that increasing population and employment densityresults in increased transit use, and the case study findings supported this. Given the very lowdensities prevalent in Surrey and other suburban areas, strategies to increase these densities will be animportant means of encouraging greater transit use.The case study results confirmed Pushkarev and Zupan's findings, noted in Chapter 2, thattransit use does not increase steadily with increasing density. Instead, there appears to be upper andlower thresholds above or below which transit use sharply increases or decreases. In the middle rangeof density, there was considerable variation in transit use (indicative of the many other factors whichinfluence transit use). These findings applied to both residential and employment density.Given this non-linear relationship between transit use and density, targeting the nodes forincreased density would appear to be more effective than a general or area-wide increase in density, asa means of encouraging greater transit use. That is, threshold densities (above which there is a sharpincrease in transit use) could be more easily achieved in the nodes than throughout a suburban area.In fact, threshold densities could easily be exceeded in the nodes, providing a very favorableenvironment for transit. Both the literature review and the case study indicate that clusteringemployment into high-density nodes will have an even greater impact on transit use than increasingresidential density. According to Pushkarev and Zupan, proximity of a residential area to a highdensity employment node is more critical than the actual residential density. Thus, a potential landuse pattern might consist of nodes with high population and employment densities, surrounded bypredominantly residential areas of medium density (but above the threshold level that would supporttransit service to and from the nodes). This pattern would encourage transit use, provide for a rangeof residential density choice, and allow open space and rural areas to be preserved.THE PLANNING FRAMEWORK /95The Surrey case study clearly illustrated that transit is virtually nonexistent at low densities.This suggests that it may be futile for transit to try to serve those trips which both begin and end inlow density areas ( i.e. residential density below 500 and employment density below 200). Evenparatransit would not be viable in this market. Thus, determining which market niches have potentialfor transit service also means deciding which niches transit should not even attempt to serve. It isimpossible for transit to serve all areas effectively.Urban design can also be an important determinant of transit use that is particularly critical inthe nodes. As was noted in Chapter 2, most suburban nodes, although they are denser than typicalsuburban development, are not very transit friendly, since they have been designed almost entirelyfrom the point of view of the automobile driver. Urban design characteristics, such as widely spacedbuildings and lack of sidewalks, are likely another reason for the extremely low transit mode split forintra-nodal trips: travelling around the nodes is inconvenient and possibly even dangerous for transitriders and other pedestrians, so most people drive. In order to achieve transit friendly design,guidelines must be developed and followed. As was seen in Chapter 6, Surrey's plans for its towncentres, especially Whalley, promote the idea of transit friendly design, but there have been fewconcrete results to date.7.2.5 Transportation Demand ManagementTransportation Demand Management (TDM) is an important strategy for increasing suburbantransit use, particularly in the suburban nodes. TDM is not a panacea, and as discussed in Chapter 3,its impact can vary. However, it can potentially have a large impact since it acts to change travelbehavior. The impact is greatest when the program is mandatory and when there is a strong financialincentive involved. In addition, TDM is most effective when it is combined with improved transitservices and land use strategies.THE PLANNING FRAMEWORK /96TDM strategies which make it more expensive to commute by single occupant vehicle (SOV)are generally the most successful. The results of BC Transit's Usage and Attitude Survey presented inChapter 6 indicated that higher SOV commuting costs would persuade about half of GreaterVancouver commuters to consider taking transit. Reducing the availability of free employee parkingin the town centres, which would eliminate an important incentive for driving, is one strategy beingvery tentatively considered in Surrey. The difficulty of implementing such a policy fairly is majorbarrier. Road pricing is another way of increasing the cost of driving, and thus encouraging alternatemodes. This might be particularly effective for discouraging suburb-to-suburb commuting by SOV inGreater Vancouver given the many bridges which connect different suburban areas. Surrey andGreater Vancouver appear to be moving slowly towards adopting some of these strategies.7.3 A Conceptual Framework for Suburb -to-Suburb Transit PlanningThe case study findings clearly have broad implications, beyond the example of Surrey itself.A conceptual framework for suburb-to-suburb transit planning could be used to apply these findings,and develop transit strategies for other suburban areas facing similar problems.7.3.1 Examples of Planning FrameworksThere have been previous attempts to produce a method for developing a suburb-to-suburbtransit service strategy, and two of these will be discussed briefly below to provide some backgroundto the development of the planning framework. The first is concerned mostly with identifying marketniches for transit in the suburb-to-suburb market, while the other deals more with cost-effectivenessand implementation.Khasnabis et al. (1990) used a Detroit case study to develop a procedure for matching unmetsuburban travel needs with interested private sector providers. First they identified potential suburb-THE PLANNING FRAMEWORK /97to-suburb transit markets in the Detroit area using 1980 census work trip data. Suburban Detroit wasdivided into 41 zones and those zone pairs with the highest inter-zonal travel demand were chosen aspotential transit markets (53 zone pairs in all). Travel times and densities were also factored in sinceit was postulated that these would affect transit use. The zone pairs were rank ordered by traveldemand for peak period work trips, travel time (based on congestion levels and distance between thezones), population density at origin zone, and employment density at destination zone. Each pair wasthen scored, producing an overall ranking of the zone pairs as potential transit markets.The second part of the procedure involved asking local transportation providers if they wereinterested in providing suburban transit services. Those that expressed an interest were then shownpotential routes developed from the highest ranking zone pairs. Costs of providing service wasestimated for five sectors and it was concluded that privately run transit service was feasible for thesesectors.Unlike the previous example, Rosenbloom (1990) considered both work and non-work trips inher framework for developing a suburban transit strategy. First, major concentrations of potentialtransit users were located, differentiating between work and non-work trips. The destination of worktrips was determined by identifying suburban employment concentrations, while the origin of non-work trips was determined by identifying residential concentrations of likely non-traditional transitusers (many of them transit dependent). The next step was to calculate the number of trips generated,and to estimate the potential non-traditional transit mode split for these trips. Potential routes werethen determined. Although work and non-work trips were considered separately at the trip generationstage, the method also stresses the need to overlap work and non-work transit service at the routeplanning stage in order to meet the midday travel needs of transit commuters.Rosenbloom also made an inventory of non-traditional services used by local transit authoritiesin 22 medium sized U.S. cities; these included vanpools, taxis, and paratransit feeder services.THE PLANNING FRAMEWORK /98Potential transit and paratransit options were then matched with the travel needs identified in thesuburban area and the cost-effectiveness of the various transit service options was assessed. The finalstep would be to develop an implementation strategy.7.3.2 Outline of the Planning FrameworkThe following planning framework is an attempt to develop a generalized planning approachfor a comprehensive suburban transit strategy. It was developed primarily from the Surrey casestudy. Figure 7.1 outlines the planning framework.The first step in the process involves gathering information about the problem. Populationand employment distributions and the associated travel patterns constitute critical information to becollected. Other useful information might include household and population characteristics, modesplit, and public attitudes towards transit. The type of data collected will depend on the particularneeds and problems which must be addressed (for example, whether to look at work trips only or peakperiod only). Most of this information can be obtained from an origin-destination survey if one hasrecently been undertaken in the community. Public input would also play an important role at thisstage, and in later stages of the planning process. Likewise, private developers and employers need toplay a greater role in finding the solutions to these problems.The next step is to analyze the data. A major goal of this analysis is to identify which travelmarket niches are viable for transit service. Density is a critical factor, and concentrations ofpopulation and employment should be noted. These and other activity centres form the nodes, whichplay a key role for suburban transit. Analysis of the travel patterns should identify major travel flowsbetween zones or groups of zones, with particular attention being given to those zones containing thenodes. An examination of current mode splits (along with current transit service levels) for differentflows gives some indication of how amenable the flows are to transit service and the potential forTHE PLANNING FRAMEWORK /99Figure 7.1 Conceptual Framework for Suburb-to-Suburb Transit PlanningTHE PLANNING FRAMEWORK /100expansion. Another important consideration might be the direction of the flows; whether they arebalanced or predominantly in one direction will clearly effect the proposed service. The result of thisstage will be the identification of a number of potential transit sub-markets, while other sub-marketsmay be abandoned.In the system development stage, the potential sub-markets identified by data analysis arematched with the most suitable transit service. Different sub-markets will require different types oftransit service, depending on factors such as level of demand, trip length, or trip purpose. Transit canbe classified according to mode (including rail, bus, mini-bus, vanpool, and carpool), and service type(express, local fixed-route, or flexible paratransit). Each travel sub-market is matched with the mostappropriate combination of mode, service type, and level of service. This family of services, eachcatering to a different sub-market, must then be integrated into a unified system. This involves suchstrategies as a timed transfer system to integrate local and express services. A system can also beintegrated to serve both work trips during the peak periods and non-work trips during off-peak times.The final result of this stage is a transit service plan.Since merely providing the service will not guarantee increased transit use, TDM and land usestrategies must be developed concurrently to complement and strengthen the transit service plan.Results from the data analysis stage can be used to determine what TDM and land use strategies wouldbe most effective. For example, how would commuters respond to increased commuting costs? Dothe nodes need to be strengthened or redesigned? If TDM or land use strategies could improve thetransit viability in a sub-market or encourage the use of a particular transit service, this informationneeds to be fed back into the process. Thus, there must be strong links of communication between thesystem development process and the TDM and land use strategy development processes at all stages inorder to ensure an integrated product. The final stage of the process is to combine all three to form acomprehensive suburban transit strategy.THE PLANNING FRAMEWORK /101There are some constraints on this process which should be mentioned. One type ofconstraint results when the transit operator has certain public objectives which it must fulfill. Forexample, a mandate to serve the transit dependent population in the community may result in theoperator providing service in sub-markets which would normally be abandoned as non-viable. Cost isundoubtedly the most important constraint on transit operators, and all transit services and strategiesmust be assessed for their cost effectiveness.7.4 SummaryThe results of the case study suggest that nodes or town centres will play an important role insuburb-to-suburb transit planning. The most promising suburban markets for transit appear to beserving inter- and intra-nodal trips, as well as trips to the nodes from surrounding areas. Increaseduse of express and paratransit services will also help transit expand its market. Combining theseimproved services with land use and TDM strategies results in a comprehensive transit servicestrategy. The generalized process for creating such a strategy was developed, resulting in a conceptualframework for suburb-to-suburb transit planning.CHAPTER 8 CONCLUSIONS8.1 IntroductionThe case study findings confirm that Greater Vancouver has experienced a fundamentalrestructuring of its commuting pattern which has seen suburb-to-suburb travel become the dominantcommuting flow in the region. Thus, Vancouver appears to be following a trend similar to that inmany other North American cities. As in these other cities, circumferential trips between outersuburbs with burgeoning employment make up the most rapidly expanding commuter market.Unfortunately, these suburb-to-suburb trips, particularly when they do not follow traditional radialpatterns, are extremely difficult to serve with transit, and generally have very low transit mode splits,as the case study demonstrated. The challenge for transit providers will be to create a comprehensivestrategy that will allow them to tap this growing market.8.2 Planning ImplicationsOne assumption made at the beginning of this thesis was that transit should not attempt toserve all markets in the suburbs. Some markets are much more amenable to transit service than others;this was illustrated in the case study by the variations in travel demand, associated densities, andexisting transit use for different suburb-to-suburb commute types. Transit can and should providesuburb-to-suburb service, but it must concentrate its efforts on those markets where it can bettercompete, and not waste its scarce resources trying to serve markets which are simply not viable.Travel within, between, and to suburban nodes is a rapidly growing market, and one whichlikely holds the greatest potential for transit in the suburbs. The case study clearly indicated theimportance of nodes for suburb-to-suburb transit service. As major activity centres, suburban nodesare important trip generators. This, combined with their higher densities, means that nodes account102CONCLUSIONS /103for most of the concentrated flows in the suburbs. If the transit system is oriented to the nodes, travelcan be concentrated even more into high volume corridors which support greater levels of service. Amulti-nodal transit system would have three major components: inter-nodal, intra-nodal, and service tothe nodes. Vancouver's system partly follows this structure, but improvements are needed.Inter-nodal trips constitute a large, rapidly growing market which is more amenable to transitservice than any other suburban commute type (as indicated by the high transit mode split for suchtrips in Surrey). This market could be better served if express service was provided between majornodes. Public attitudes clearly point to a desire for faster, more direct suburban transit routes, andthis is one market where such a high quality service might be viable. The explosive growth ofcircumferential commuting has made longer distance inter-nodal service another potential transitmarket. At present, circumferential transit service between nodes in two different suburban areas isvirtually non-existent. However, the case study indicated there is growing potential for such services,with the most promising routes from Surrey being Whalley-Newton-Richmond Centre and Whalley-Guildford-Coquitlam Centre. When demand does not warrant conventional transit service, this marketcould be served with paratransit or ridesharing.Short intra-nodal trips, which presently display very low transit use, represent a key growtharea for transit. In order to tap this critical market, some form of frequent shuttle service is requiredto reduce the ratio of access and waiting to travel time. A major reason many commuters resist anyalternatives to driving is because they require their cars for midday or after-work trips. Thus, anefficient shuttle system would not only increase transit use for intra-nodal trips, but might alsoencourage more people to travel to the nodes by transit, since they would no longer need their carsonce they had arrived.There is no one type of transit service that offers the perfect solution for suburb-to-suburbtransit. Instead, the "family of services" concept needs to be applied, in which different types ofCONCLUSIONS /104transit and paratransit services occupy different market niches. These services can then be integratedto form a comprehensive system.Land use strategies play a very important role in encouraging suburb-to-suburb transit use.The case study confirmed that transit use in suburban areas does increase with rising residential andemployment densities. The correlation appeared to be strongest for employment density. Therelationship between transit use and density was not linear. Transit use was extremely low below acertain threshold density and quite high at densities above an upper threshold, but the correlation withtransit use appeared weaker in the middle range of densities. Nodes generate transit use because theyhave densities well above the upper threshold. The benefits of density can be enhanced by employingtransit and pedestrian oriented design. Encouraging the development of mixed-use, transit andpedestrian oriented suburban nodes will likely be a major objective for any land use planning strategy.Transportation demand management strategies are also important, particularly given theirability to actually change travel behavior. The key to success for TDM seems to be the use ofmandatory economic measures which penalize automobile use rather than just encourage transit use.Parking reforms and road pricing are two types of TDM which could greatly impact the attractivenessof driving to work Implementation of TDM strategies can be complex and politically unpopular, butthe results can be significant.Clearly, there is no single solution to the problem of encouraging transit use for suburb-to-suburb trips; there must be a comprehensive approach. This would combine innovative new transitservices with TDM strategies to discourage automobile use, and land use strategies to create a moretransit- and pedestrian-friendly landscape. In order to be successful, strong measures which bothencourage transit use and penalize driving, must be complemented with the provision of a high qualityalternative to the private automobile. The example of serving intra-nodal trips provides a goodillustration of this comprehensive approach: an innovative paratransit service is combined with mixed-CONCLUSIONS /105use, pedestrian oriented development and TDM strategies, all of which combine to make it mucheasier and more attractive for people to get around without their cars. A similar approach needs to beapplied to other suburban travel markets. Comprehensive also means that transit providers mustbecome more involved with all aspects of the suburban development process. Greater cooperationwith private developers and employers is needed to solve some of these problems. The planningframework presented in the last chapter outlines this comprehensive approach to the problem ofsuburb-to-suburb commuting, and provides a means for applying the lessons learned in this case studyto other suburban communities facing similar problems.8.3 Recommendations for Further StudyVery little work has been done in the field of suburb-to-suburb transit planning. Theliterature review indicated a need for more research in this area, and in particular, there are fewCanadian examples. A parallel study in another suburban area of Greater Vancouver or anotherCanadian city would be helpful.An update of the situation in Surrey using the 1992 GVRD Travel Survey and the 1991Census would show the impact of SkyTrain, the Alex Fraser Bridge, and the rapid growth in GreaterVancouver, especially Surrey, over the past 7 years. A further investigation of other factors affectingsuburban transit use (such as trip length) is also needed. While this study has looked at work tripsonly, it would be helpful to investigate other trip purposes and the potential for transit to serve these,since suburban nodes are important generators of non-work trips as well.While the multi-nodal urban form advocated in this thesis produces greater transit use thansuburban sprawl, it can be argued that it is less amenable to transit use than a highly centralized urbanform. Further study is needed into the tradeoffs between a multi-nodal and a single-centred urbanform in terms of transit use, minimization of total travel, and other factors.CONCLUSIONS /1068.4 SummaryThis study has shown that the growth of suburb-to-suburb commuting is a serious challengefor the transit industry. While it is not reasonable to expect transit to serve all suburban travelmarkets, it will be able to compete with the automobile in certain sub-markets, particularly thoseinvolving suburban nodes, which will play a key role in suburb-to-suburb transit service. TDM andland use planning strategies will also be critical in making transit more competitive with the privateautomobile. If transit is to be successful in these suburban markets, it will require an innovative andcomprehensive approach.107BIBLIOGRAPHYBakker, J.J., J. Calkin and S. Sylvester. 1991. "Multi-Centered Time Transfer System for CapitalMetro, Austin, Texas". Transportation Research Record 1202: 22-29.Barringer, Felicity. "Census Data Show More U.S. Children Living in Poverty". The New YorkTimes, May 29 1992, p. Al2.Bhatt, Kiran. 1979. "Costs and Productivities of Innovative Urban Transportation Services" inUrban Transport Service Innovations Special Report 184. Washington: TransportationResearch Board.BC Transit. 1992A. 1992-1993 Annual Service Plan for the Vancouver Regional Transit System.Vancouver: BC Transit Planning Department.. 1992B. Whalley Bus-SkyTrain Integration Plan. Vancouver: BC Transit PlanningDepartment.^ . 1992C. Usage and Attitude Survey. Prepared for BC Transit by Angus Reid Group.Cervero, Robert. 1984. "Managing the Traffic Impacts of Suburban Office Growth". TransportationQuarterly 38 (4): 533-550.^ . 1986. Suburban Gridlock. New Brunswick, N.J.: Center for Urban Policy Research,Rutgers University.^ . 1986. "Urban Transit in Canada: Integration and Innovation at its Best".Transportation Quarterly 40 (3): 293-316.^ . 1988. "Land-Use Mixing and Suburban Mobility". Transportation Quarterly 42 (3):429-446.^ . 1989. America's Suburban Centers. Boston: Unwin-Hyman.^ . 1991. "Land Uses and Travel at Suburban Activity Centers". Transportation Quarterly45 (4): 479-491.Cervero, Robert and Bruce Griesenback. 1988. "Factors Influencing Commuter Choices in SuburbanLabor Markets: A Case Analysis of Pleasanton, California". Transportation Research A 22A(3): 151-161.Couture, M.R. and T. Dooley. 1981. "Analyzing Traveler Attitudes to Resolve Intended and ActualUse of a New Transit Service". Transportation Research Record 794: 27-33.Deakin, Elizabeth A. 1991. "Jobs, Housing, and Transportation: Theory and Evidence onInteractions Between Land Use and Transportation". Transportation, Urban Form, and theEnvironment, Special Report 231. Washington: Transportation Research Board.Dickenson, Mike. 1992. Surrey Planning Department. (Personal communication).108Flannely, Kevin J. and Malcolm S. McLeod, Jr. 1990. "Predicting Consumer Demand forAlternative Transportation Services Among Suburban Commuters". Transportation ResearchRecord 1280: 73-81.Goldberg, Micheal A. and John Mercer. 1986. The Myth of the North American City. Vancouver:University of British Columbia Press.Greater Vancouver Regional District. 1975. The Livable Region 1976/1986, Proposals to Managethe Growth of Greater Vancouver.. 1987. 1985 Metropolitan Vancouver Origin-Destination Survey. GVRD DevelopmentServices.^ . 1992. Greater Vancouver Key Facts. GVRD Development Services.^ . 1993A. Transportation Demand Management Measures and Their Potential forApplication in Greater Vancouver. Transport 2021 Working Paper.. 1993B. The Cost of Transporting People in the British Columbia Lower Mainland.Transport 2021 Working Paper.Giuliano, Genevieve. 1992. "Transportation Demand Management". American Planning AssociationJournal (Summer 1992): 327-332.Hendrickson, Chris. 1986. "A Note on Trends in Transit Commuting in the U.S. Related toEmployment in the CBD". Transport Research A 22A (1): 33-37.Higgins, Thomas J. 1992. "Shuttle Buses: Experience and Planning Implications". TransportationQuarterly 46 (3): 397-409.Hodgson, Karen. 1992. BC Transit. (Personal communication).Hooper, Kevin G. 1989. Travel Characteristics at Large Scale Suburban Activity Centers, SpecialReport 323. Washington: Transportation Research Board.Humphrey, Thomas F. 1990. "Suburban Congestion: Recommendations for Transportation and LandUse Responses". Transportation 16: 221-240.Huth, Mary Jo. 1983. "Toward a Multi-Nodal Urban Structure". Transportation Quarterly 37 (2):245-262.Khasnabis, Snehamay, Bharat B. Chaudhry, Mark E. Neithercut, and Neva A. Nahan. 1991."Developing Markets for Transit Privatization for Suburban Travel in Large MetropolitanAreas". Transportation Research Record 1297: 93-105.Knight, Robert L. and Lisa L. Trygg. 1977. "Evidence of Land Use Impacts of Rapid TransitSystems". Transportation 6: 231-247.Kobayakawa, Martin. 1992. BC Transit. (Personal communication).109Lai, Mike. 1993. Surrey Engineering Department. (Personal communication).Municipality of Metropolitan Toronto Planning Department. 1976. Metroplan: The Central Area andthe Sub-Centres.Nelson, David 0. and Katherine K. O'Neil. 1983. "Analyzing Demand for Grid Transit System".Transportation Quarterly 37 (1): 41-56.Newman, Peter and Jeffrey Kenworthy. 1989. Cities and Automobile Dependence, An InternationalSourcebook. Hants, U.K.: Gower Publishing Company Limited.Orski, C. Kenneth. 1985. "Suburban Mobility: The Coming Transportation Crisis?".Transportation Quarterly 39 (2):283-296.. 1987. "Managing Suburban Traffic Congestion: A Strategy for Suburban Mobility".Transportation Quarterly 41 (4): 457-476.Pendakur, V. Setty and Stephen Bowler. 1987. Urban Transport in the South Fraser Valley ofMetropolitan Vancouver, B. C. Prepared for the United Way of the Lower Mainland.Pickrell, Don H. and Donald C. Shoup. 1979a. Employer Subsidized Parking and Travel ModeChoice. Cambridge, Mass.: Harvard University Department of City and Regional Planning. 1979b. Land Use Zoning as Transport Regulation. Cambridge, Mass.: HarvardUniversity Department of City and Regional Planning.Pisarski, Alan E. 1987. Commuting in America. Westport, Conneticut: Eno Foundation forTransportation.Priest, Donald E. and Joseph L. Walsh-Russo. 1983. "Land Use Trends and Transit Operations".Special Report 199 Future Directions of Urban Public Transportation, pp. 37-46.Washington: Transportation Research Board.Pushkarev, Boris S. and Jeffrey M. Zupan. 1977. Public Transportation and Land Use Policy.Bloomington: University of Indiana Press.Reichart, James P. 1979. "Wanted: National Policy on Suburban Transit". Transit Journal (Summer1979): 37-42.Rice, R.G. 1977. The Effect of Alternative Urban Forms on Two-Mode Transportation SystemRequirements. Toronto: University of Toronto/York University Joint Program inTransportation.Roos, Daniel. 1979. "Relationship of Innovative Services to a Family of Urban TransportationServices" in Urban Transport Service Innovations Special Report 184. Washington:Transportation Research Board.Rosenbloom, Sandra. 1990. Developing a Comprehensive Service Strategy to Meet a Range ofSuburban Travel Needs. Washington: Urban Mass Transportation Administration, U.S.Department of Transportation.110Schneider, Jerry B. 1981. Transit and the Polycentric City. Washington: U.S. Department ofTransportation.Schneider, Jerry B. and Stephen P. Smith. 1980. "Synchrocentered Transit Systems: The Challengeof the 1980s". Transit Journal 6 (2): 39-48.Schneider, Jerry B., Stephen P. Smith, J.L. Heid, D.S. Saito, and K.E. Cotterall. 1979. IncreasingTransit's Share of the Regional Shopping Center Travel Market: An Initial Investigation.Washington: Urban Mass Transportation Administration, U.S. Department of Transportation.Schofer, Joseph L. 1983. "Does This Bus Go to the Future? Some Thoughts on the Future of UrbanPublic Transit". Special Report 199 Future Directions of Urban Public Transportation, pp.59-66. Washington: Transportation Research Board.Smith, Deming. 1992. BC Transit. (Personal communication).Snohomish County Transportation Authority. 1989. A Guide to Land Use and Public Transportationfor Snohomish County, Washington. Washington: U.S. Department of Transportation.Statistics Canada. 1983. 1981 Census of Canada, Cat. 92-908, Tables 8,19,21. Ottawa. Ministry ofSupply and Services.Stevens, William F 1990. "Improving the Effectiveness of Ridesharing Programs" TransportationQuarterly 44 (4): 563-578.Surrey Planning Department. 1986. Surrey Official Community Plan, By-Law No. 7600.^. 1992. Whalley Town Centre Executive Summary.Taaffe, Edward J. and Barry J. Garner. 1963. The Peripheral Journey to Work. Chicago:Northwestern University Press.Taebel, Delbert A. and James V. Cornehls. 1977. The Political Economy of Urban Transportation.Port Washington, N.Y.: Kennikat Press.Teal, Roger. 1987. "Carpooling: Who, How, and Why". Transportation Research 21A (3):203-214.United States Federal Highway Administration. 1990. Evaluation of Travel Demand ManagementMeasures to Relieve Congestion. Washington: U.S. Department of Transportation.Transportation Research Board. 1987. Research for Public Transit, New Directions, Report 213.Washington: Transportation Research Board.Weigle, Theodore G. 1988. "Causes of Change: Internal and External Pressure" in NewOrganizational Responses to the Changing Transit Environment, Special Report 217.Washington: Transportation Research Board.Young, David. 1984. "Chicago - Trying to get Around the Transit Dilemma from Urban Sprawl".Mass Transit XI (11): 30-35.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items