Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
CHAPTER III FEEDER SERVICE INTRODUCTION AND MAJOR FINDINGS Many persons with disabilities are able to board and ride fixed route service. This is particularly true if the fixed route system is accessible. A sigruficant percentage of these potential fixed route riders are not, however, able to get to or from the nearest fixed route stops. It has been estimated that, even if fully accessible fixed route service is available, about 30% of persons determined ADA eligible would still require paratransit service due to an inability to get to and from fixed route stops and stations. Feeder services can enable ~ndivicluals with disabilities to use fixed route service by providing paratransit to and/or from the nearest stop. Passengers are usually picked up at their homes (although the trip could originate from any location), and are transported for the first leg of their journey to a bus or train stop. This is usually the shorter segment of their journey to a bus or train stop. Feeder service may also occur in reverse, wad passengers being picked up at the bus or train stop and transported home. Feeder service is also an attractive alternative to transit pronciers seeking to offer appropriate but cost-effec~ve service to customers with disabilities. This is particularly true for lengthy para~ansit trips that parallel fixed route service. The greater portion of the trip that is provided on fixed route rawer Man parabansit service, the greater Me cost-effic~encies that can be reali~ect. Lewis, D., Hickling Corporation, Preliminary Regulatory ·mpactAr~alysis of Transportation Accessibi~i~ Requirements for file Americans with Disabilities Act, U.S. Department of Transportation, Washington, D.C., April 1991. TCRP B-l21 III- 1 Draft Final Report
The feeder service case studies presented In this chapter Include BC Transit In Vancouver, British Columbia, and Pierce Transit in Tacoma, Washington. Key findings from research efforts at both sites Include the following: . . Based on experience at Pierce Transit, about 9% of ADA paratransit eligible nclers are capable of using feecler service for aD of their trips. Another 10% can use it for some of their trips. Feeder trips were shorter than direct paratransit trips In Vancouver where frequent light raid service is available and the comparable direct trip would need to use inconvenient and very congested bridge crossings. At Pierce Transit, where feeder is used more extensively and transit service consists of a bus network, feeder trips are much longer than comparable clirect paratransit trips, mainly because of long wait times. Feeder trips at Pierce Transit take about 19% longer than comparable trips made using fixed-route trans*. Riders In Vancouver prefer feeder service to direct service with respect to travel time, scheclule convenience, service availability, and sense of independence, but not wad respect to personal effort or comfort level. Riders at Pierce Transit prefer direct trips with respect to aD aspects of service. Requiring riders to use feecler service at Pierce Trans* has reduced ADA paratransit usage by those riders by 54%. Of those riders required to use feeder, 46% have ceased riding ADA paratransit altogether. Of those who have stopped riding, 45% say they do not go out as often as before. The implementation and planrung cost for feeder service has been minima. TCRP B-IA III-2 Draft Final Report
. . . At BC Transit and Pierce Transit, feeder trips cost less than half as much per trip as direct paratransit service for comparable origins and destinations. The savings was $14.44 per trip at BC Transit and $16.38 per Hip at Pierce Transit. At Pierce Transit, planning one round hip by feeder costs about $4.51. Tog into account repeated tnps, trip planning cost an average of $~.26 per one-way tnp. Mandatory feeder service has enabled Pierce Transit to continue providing paratransit In fun compliance with ADA while reducing cost by about 7.5%. Most of the cost saving comes from reduced demand. CASE STUDY SITES Evidence on the effectiveness of feeder service is available from case studies concluctec3 for this research at BC Transit in Vancouver, British Columbia, and at Pierce Transit In Tacoma, Washington. These systems were chosen for case studies because they had more extensive feeder service directed primarily to serving riciers with clisabilities than any other medium-to-large transit system In North America. General information about the two case study systems is presented in Table ITI-~. BC Transit began purchasing lift-equipped buses in 1989. At the time of the study, 30% of the fleet was accessible. Both the SkyTra~n light rail line and Me SeaBus ferry are fully accessible. Pierce Transit has had a 100% accessible bus fleet since 1992. The bus system uses the time transfer concept. Routes are laid out in a grid pattern and meet at nine major transfer centers. TCRP B-IA III-3 Draft Final Report
Table IlI-~. Case Stucly Sites ~ Key Statistics - . BC Transit Pierce Transit Vancouver Service Area Population ~605,000 ~1.8 minion ~ TotalF=ed-Route Fleet | 172 buses | 699 buses | 244 trolley buses light rail passenger ferries Percent of Service Accessible | 100% ~30% of buse 100% of light rail and ferry service Annual FL'ced-Route Ridership 11.5 million 117.5 millions Paratransit Fleet l 80 | 190 | Annual Paratransit Ridership T 530,000 T 800,000 Annual Paratransit Budget T $11,400,000 T $10,220,00C (U.S. Currency) a Revenue passengers April 1,1995 - March 31,1996 Ad fleet data pertain to the time of We case studies. TCRP B-IA III-4 Draft Final Report
DESCRIPTION OF PARATRANSIT AND FEEDER SERVICE BC Transit BC Transit's handy DART paratransit service Is prowled by eight private operators under contract to BC Transit. Each operator provides its own caulking, scheduling, and dispatching within a primary service area. Transfers between operators are common The paratransit operators provide about 800 feeder Dips per months, primarily to SkyTra~n and SeaBus, out of about 67,000 total handyDART trips per month. Feeder service was not estabIished as a formal program Rather it has evolved, primarily as a way to provide long trips between Me suburbs and central Vancouver which would overwise be too expensive or tone consuming due to traffic congestion. Travel between the suburbs ant! We arty of Vancouver is limited because of water barriers crosser! by very few bridges, ant] because there are no freeways Cat provide connections into the cibr. SkyTra~n prowdes a direct link between the southeastern suburbs and downtown, while SeaBus provides a link between the northern suburbs and downtown, ~ transfers available directly to SkyTrain. Because of the way feeder service has evolveci, there are no formal cutena for determining which riders or trips are offered feeder service. According to the schedulers ~nter~ewed, about ~ree-quarters of Me feeder trips provider} are mandatory. That is, the ncler is politely offered a feeder trip or no trip at all. Paratransit eligibility criteria are similar to those under the U.S. Americans wad Disabilities Act. As part of the certification process, the i. . operators categorize riders in terms of functional ability, and these coded categories appear on the computer screen when a scheduler is planning a trip. TCRP B-lA TIl-5 Draft Final Report
Figure IlI-~. Passengers Entering Accessible Sky Train Car ·.. :': - ? ~ _~.:', ..,~ TCR~ B-IA III-6 Draft Final Report
Sea_ ad Sinus Selves ^ f ~: i ~6 7 ~ID-7 ~
Typically, a feeder trip is offered if the eligibility information indicates that a ricLer can board SkyTra~n and/or SeaBus; and the trip destination is close to a SkyTra~n stop or can be easily coordinated with another operator whose primary service area is close to that stop; and: Lee requested destination would require a lengthy handyDART hip; or The requested trip occurs during handyDART peak hours; or The ricler asks for a feeder trip. Many riders have learned to ask for feeder trips because Hey know that direct paratrans* service to their destinations curing specific service hours are unavailable. When a passenger transfers from hanclyDART to SkyTrain or SeaBus, the paratrans~t vehicle drops the passenger off without waiting for He train or ferry to arrive. Since SkyTra~n runs every 2.5 to five minutes, wait Hines are generally short. SeaBus runs about once every 15 minutes so wait times are reasonably short Were as wed. For pick ups from SkyTrain or SeaBus, the hanclyDART vehicle is directed to meet the train or ferry at its scheduled arrival fume. Operators indicated t hat, more often then not, passengers are aIreadv waited when the van arrives. The vans will wait un to five minutes. Fare coordination is handed In one of two ways. BC Transit provides the handyDART operators with ~ree-zone Fare Saver tickets to be used as transfers, which will pay for the longest trip on BC Transit's fLxed-route transit services. Transit, to be provided to feeder riders when Hey pay their handyDART fare. However, most operators find it supper to forgo the paratransit fare and have riders pay the discount ACRE B-IA III-8 Draft Final Report
fare on the fixed-route system. For return trips, riders must produce a transfer that has been validated at one of the SkyTrain stations or the SeaBus terminal. Pierce Transit Pierce Transit has operated a paratransit system, known as SHUTTLE, since 1980. Until recently SHUTTLE provided direct, door-through-door, 2thour a day service throughout Pierce Transit's legal service area. After passage of the ADA, ridership began growing rapidly, reaching 530,000 in 1995. To control costs, which had reached over 25% of the total agency budget, Pierce Transit scaled back its paratransit program during the year prior to the case study by reducing service closer to minimum ADA service criteria. Changes implemented during 1995 included a rigorous ADA certification process, a reduced service area, hours comparable to fixed-route service hours, a travel training program, and mandatory feeder service for riders who, based on the ADA certification process, are capable of using the service. Fares were Increased from $.35 to $.45, matching an increase In the local bus discount fare. Pierce Transit employees are responsible for ah car-taking and scheduling for SHUTTLE service. Pierce Transit also dispatches the bulk of weekday trips, which are carried on vehicles operated by a combination of agency personnel and employees of a private contractor. Overflow trips, plus evening and weekend trips, are scheduled on vehicles operated and dispatched by two local cab companies. As of May I, 1996, out of 7,479 applicants screener! for ADA eligibility, 705 or 9% had been found eligible only because their disability prevents Hem from going to or from a bus stop. Pierce Transit refers to this group as "3B eligible," a reference to the Gird category of ADA eligibility under DOT regulations. This group is offered TCRPB-IA III-9 Draf~FinalReport
Dime "I-~. sniff Rider lransfe~ing to Plxed Dome Bus ~1II~ ~. IIIIII~ it. ~SS!sSsS~ ~:~i~i~i~i~i~i~i~i~ii~i~i~i~i~i~i~i~i~iii~i~i~i~i~ e 6-7 ~IH-10 ~- ma/ Of
mandatory feeder service. Indivicluals who could use feeder service only some of the time, because their clisabilitr would prevent Hem from boarding or riding a fixed-route bus some of the time, are permitted to decide for ~emseIves whether to make a specific trip by bus. As of April 1996, nine months after the start of feeder service, SHUTTLE was providing about 600 feeder trips per month out of a total 40,000 paratransit trips per monk. Pierce Transit's ADA certification a~ninistrator estimates that at least 900 people will have been designated 3B eligible once ADA eligibility screeriing is complete. If these people use feeder service at the same rate as current registrants, then 1 here win be about 750 feeder trips per month. When a rider designated as 3B eligible requests a SHUTTLE trip, the call taker checks to see if the distance from the ricler's trip origin to the nearest transit center is less than to the ricler's Resonation. If it is, and the trip is a repeat trip, then it is entered as a feeder trip In the daily trip schedule immediately. If the request is for a new trip, the caretaker arranges to can the rider back within about half an hour. The car-taker then plans how a feeder trip can be arranged and cans the rider back with a trip plan. If the rider accepts the proposed trip, it is entered into the schedule. Almost aD transfers between SHUTTLE and fixed-route service are done at one of Pierce Transit's nine transit centers. These are mostly at shopping malls with accessible telephones and bathrooms widen a relatively short distance of the transfer location. If the rider's trip does not meet the distance criterion, Hen He rider is offered a direct SHUTTLE trip. Feeder tries are schecluled using SHUTrT~E's standard criteria, which require . . riders to be ready to be picked up any time In the hour before the desired arrival tome at TCRPB-IA III-ll Draft FinalReport
the cleshnation. Drop offs may occur any time before Me desirer! arrival time. In the case of a feeder trip, this means that the rider may be dropped off at the transit center any time In the hour before the scheduled bus departure. Pierce Transit's fixed-route buses operate on 15-minute headways In the core area during peak hours. At other times and places, routes operate on headways ranging from 30 minutes to an hour. For trips Involving a transfer from fixed-route to paratransit, the trip plan wiB generally be designed to have the passenger at the transfer point an hour before the desired arrival dine at We final desdnation. Feeder riders pay a fare on the fixed-route portion of their trip, and pay noting on the SHUTTLE portion. METHODOLOGY The case studies included analysis of feeder trip records, surveys of feeder riders, and ~nterwews with operations staff. At BC Transit, me following data sources were used: . . . . TCRP B-1A Interviews with dispatchers and managers of the four operators which . . schedule the most feeder trips, and with staff of BC Transit responsible for overseeing paratransit. Detailed analysis by dispatchers and Me consulting team of frequent feeder trips made by 24 individuals who account for half of Me feeder trips provided. A mail-back survey sent to ah 77 feeder participants served by the four operators; 49 responses were received. At Pierce Transit, the data sources included: III- 12 Draft Final Report
. . . Interviews with Pierce Transit staff responsible for planning, management of the SHUTTLE program, caB-takers, schedulers, ant] dispatchers. Detailed analysis by SHUTTLE car-takers and dispatchers of 230 trips taken by the 16 most frequent feeder riders. A mail-back survey sent to ah 625 ~ndiv~cluals certified as 3B eligible as of April 1996; 208 responses were received. · A focus group with I! 3B registrants who had used the service, and a group interview win Pierce Transit's Transportation Advisory Committee. FINDINGS OF THE CASE STUDIES Data were assembled on the characteristics of feecler trips, the length of feecler mps compared to direct paratransit service and fixed-route service, operational impacts of feeder service, and demand Impacts. Implementation issues and rider concerns were also Investigated. Characteristics of Feeder Trips Differences In the way that feeder service has been unplemented at BC Transit and Pierce Transit have produced markedly different trip characteristics. For example, according to the operators In Vancouver, most feeder trips are for social/recreational purposes and for employment, and very few people use feeder service for medical trips. In contrast, the survey of Pierce Transit 3B riders shows that, of those who use feeder service SI% used SHUTTLE for medical trips, 42% for shod tries and 40% for ~ ~v ~ social/recreational trips. However 59% of 3B respondents who use SHUTTLE stated that they have never been required to use feeder service; only 25% indicated that most or aD of their SHUTTLE trips require transfers to a fixec3-route bus. TCRP B-IA III- 13 Draft Final Report
In Vancouver, 63% of feeder users who responded to me mail-back survey used a wheelchair, compared to 23% for handyDART ridership overall. Based on the review of trip records, almost ad of the frequent feeder riders are wheelchair users. At Pierce Transit, 27% of feeder users who responded to the mail-back survey used a wheelchair. These differences may reflect the more comprehensive nature of Pierce Transit's feeder program. In Vancouver, riclers who find feeder service difficult to use, either because of their disability or the nature of the trip, can often ride at other dines or to other places for which direct service is available. In contrast, Pierce Transit 3B riders probably face more limited choices. The great majority of feeder trips provided at BC Transit are standing order, repeated trips. No statistics are available on the precise percentage. At Pierce Transit it is estimated, based on estimates by Pr staff, that standing orders account for 60% of feeder Dips provided. Length and Duration of Feecler Trios . . The detailed analysis of frequent trips provided information about feeder trip lengths, ant} the components of the trip. This information is summarized In Table HI-2. The paratransit portion is somewhat longer at BC Transit than at Pierce Transit (4.9 miles and 21 minutes comparer} to 3.1 miles and IS minutes), and the transit portion is shorter at BC Transit (7.1 miles and 20 minutes compared to 8.0 miles and 31 minutes). The similarity of the mileages suggests that, despite differences in procedures, both systems end up assigning similar trips to feeder service. Most trips involve a paratransit leg only at the home end of the trip, i.e. the "to transit" portion in TCRP B-IA III- 14 Draft Final Report
Table IlI-2. Length ant! Duration of Feeder Tnps Tnp Component ~BC Transit Pierce Transit Length Duration Length Duration (Miles) (Minutes) (Miles) (Minutes) || ?aratransit portion l l l l 11 To transits 3.2 13 2.5 14 From transit 1.7 8 0.6 3 Total 1 4.9 ~21 1 3.1 1 17 j Transf er/Waitingb unknown 32 Transit portion 7.1 20 8.0 31 . . 11 'otal ~12.0 ~41 + wait time ~11.1 ~80 aTimes and distances were estimated based on trips originating at riders' homes. The "to transit' and "from transit" portions would be approximately reversed for return trips. bIncludes wait for the initial paratransit trip and waiting at transfer locations. Probably on tibe order of 15 minutes or less. See text for discussion. Source: Estimates by paratransit dispatchers from samples trips by 24 frequent riders at BC Transit and 16 frequent riders at Pierce Transit. TCRPB-IA III-IS Draft FinalReport
Table IlI-2. At Pierce Transit, only four of the 16 people had paratransit legs at the non- home end of their trips. The duration of the bans* portion is much shorter at BC Transit than at Pierce Transit. This difference occurs because most trips use the SkyTra~n light rail line and require no transfer between fixed-route vehicles, while aD Pierce Transit hips are made on buses, and three quarters of them require transfers. The paratransit portion accounts for 41% of the total trip mileage at BC Transit and 28% at Pierce Trans*. Note that waiting and transfer time accounts for a large portion of feeder trip duration at Pierce Transit. These long wait times stem from long headways on some routes, Me neec] to make transfers between fixed-route buses for most trips, and Pierce Transit's scheduling method which often delivers riders to the bans* center long before the scheduled bus departure time. The mailback survey of 3B eligibles provides some actditional information about wait times. Of 49 respondents who use feeder, over 73% nclicated they usually wait more than 10 minutes for He bus to arrive after berg dropped off by SHUTTLE, and 26% stated they usually we* more than 20 minutes. Respondents estimated similar wait times for the SHUTTLE portion of their feeder trips. Waiting and transfer dines are not known for BC Transit. However, they are probably much less than Hose at Pierce Transit for several reasons. First, the bulk of feeder trips connect to the SkyTra~n service which operates on 2.5-minute headways In peak hours and five-minute headways In off-peak hours, so wait times for SkyTra~n win be very short. However, there is probably some wait lime for transfers from SkyTrain to paratransit because, according to He operators, passengers more often Han not are already present at the transfer location when the paratransit vehicle arrives. Second, TCRP B-IA III-16 Draft Final Report
because of destinations to which trips are scheduleci, few feeder trips Involve transfers between f~xed-route vehicles. Third, the boarding and alighting time on SkyTra~n is short because the system is a high-platform, level-entry operation, without wheelchair securement. Flunky, most feeder Hips are for standing orders, so operators and passengers have an opportunity to learn to time transfers between SkyTra~n anct paratransit to reduce wait fume. The trip ciata were also used to compare feeder trip lengths to: the estimated length of the same trip if it had been made using direct paratransit service; the estimated length of the same trip if it had been macle entirely using fixed-route transit (for Pierce Transit only); and the average trip length for aD paratransit trips. These results are summoned in Table IlI-3. Feeder trips are actuary shorter than the comparable direct trip at BC Transit, ant! probably somewhat faster, depending on wait times. The reason appears to be that the SkyTrain service provides more direct and faster connections to many locations In central Vancouver than We automobile routes that would be available to a paratransit vehicle. At Pierce Transit, the average feeder trip is 2.3 miles (26%) longer than a direct paratransit trip between Me same origin and destination. It appears Mat feeder adds a reasonable amount to the total distance traveled. The duration of an average feeder trip, however, is 42 minutes (~%) longer than a direct paratransit trip between the same origin and destination. Most of the time difference results from Me average 32 minutes spent waiting on a feeder trip. Pierce Transit staff also estimated the time that it would take a non-disabled rider to make a trin using f~ced-route transit alone between the same origins and ACRE B-1A - -A- III-1 7 Draft Final Report
Table IlI-3. Length and Duration of Feeder Trips Compared to Alternatives l Trip type ~ BC Transit ~Pierce Tran: it Length Duration Length Duration If | (Miles) | (Minutes) | (Miles) | (Mi cutest | Feeder Tnp 12.0 41 + wait 11.1 tilne Same Tnpon Direct Paratransit ~15.3 ~52 ~8.8 Difference (Feeder - Direct ~-3.3 ~-11 ~2.3 _ _ Same Trip on Fixed Route NA NA | NA 67 . _ Difference (Feeder - Fixed-Route) NA NA | NA 13 . Average ParatransitTnp | 5.0 | NA | 8.8 I NA l Sources: Estimates by paratransit dispatchers from samples of trips by 24 frequent riders at BC Transit and 16 frequent riders at Pierce Transit. Direct paratransit length and duration account for detours due to shared riding at Pierce Transit but not at BC Transit. Average paratransit trip for Pierce Transit from FIA Section 15 data for FY 1994 (prior to reduction in service area). TCRP B-IA III-18 Draft Final Report l
destinations served by the feeder Hips. The estimated time includes walking time, waiting time, and time riding the bus. As shown In Table IlI-3, travel by feeder takes an average of 13 minutes (19%) longer than travel using fixed-route transit. This comparison suggests that, if transfers between paratransit and f~ced-route vehicles could be scheduled more closely, feeder travel times could probably match or be faster than fixed-route travel times. Incidentally, the comparison also shows Mat direct paratransit service is much faster Wan travel by fixed-route transit. This comparison shows that feeder service is in compliance with ADA guidelines to be comparable to fixed-route service, and that direct paratransit far exceeds the ADA guidelines. Delays to Vehicles and Passengers Depending on how feeder service is implemented, it could result in delays for vehicles and passengers, both on paratransit and on fixed-route service. Such delays would occur if paratransit or fixed-route vehicles dropping off feeder passengers for a transfer waitec! for the other-mode vehicle to arrive, if fixed-route drivers were alerted to expect a transfer from paratransit and waited for the paratransit vehicle to arrive, or if paratransit vehicles picking up feeder passengers at a transfer point consistently waited more than the normal waiting time at pick up points for Me fixed-route vehicle to arrive. In practice, none of these policies are used at either case study site, so they do not cause any delays to vehicles or to other passengers. It is possible that Implementation of feeder service has increased the total level of fLxed-route wheelchair boardings, which would result in delays due to the time required for a wheelchair passenger to board and alight. At Pierce Transit, the 600 feeder trips per month in April 1996, would result in roughly 190 total minutes of delay per average TCRP B-IA III-l9 Draff Final Report
weekday, assuming that most feeder trips occur on weekdays, that the average feeder trip requires .75 fixed-route transfers (i.e. a total of 1.75 boardings), and that each boarding or alighting takes two minutes. This delay represents about 0.2% of Pierce Transit daily revenue hours.3 Looked at another way, feeder service creates one wheelchair boarding for every 37 hours of bus operation, more or less.4 BC Transit's SkyTra~n is an automated system using high platforms, and no on- board wheelchair securement. Therefore, feeder trips result In no measurable delays. Service Quality Feeder service results In a different quality of service for those riders who choose to use it or are required to use it. Service quality can be compared based on Me survey of riders, a focus group of riders, and the trip duration data presented earlier. Results of the rider survey comparing feeder and direct service are given In Table ITI~. The most positive and important aspect of feeder service for handyDART riders appears to be its availability. (Recall that feeder is primarily used for trips the operators find too long or time consuming to provide.) A slight majority of riders also believe that feeder service is faster than direct paratransit. This result is consistent with We trip duration data presented In Table TIl-2 showing ~at, excluding wait dines (which are probably short) feeder trips are I! Granites faster than a direct paratransit Hip between the same origin and destination at the same dine of day. Furler, 83% of feeder riders surveyed felt that their transfer from a handyDART vehicle to a bus, train, or ferry was "no problem" 3 [~600 feeder trips/mon~) x (1.75 boardings/trip) x (2 boardings~alightings) x (2 minutes) / (22 weekdays/mon~] / [~527,461 vehicle-hours/year) / (300 weekday-equivalents/year) x (60 minutes~our)] = 0.002. 4 [~527,461 vehicle-hours/year3 / (300 weekday-equivalents/year)] / [~600 trips) x (1.75 boardings/~ip) / (22 weekdays/mon~] = (1758 vehicle-hours/weekday) / (48 boardings/weekday) = 36.8. TCRP B-IA III-20 Draft Final Report
Table III4. Customer Comparison of Feeder and Direct Service at BC Transit Feeder Is ~ About the Direct Is ~ Service Aspect Better Same Better . l | TravelTime | 37% | 8% | 31°~ Personal Effort 18% 20% 43% | || Schedule Convenience | 42% | 34% | 247 . . || Service Availability | 45% | 12% | 16, || Comfort Level | 16% | 18% | 475 | Sense of Independence | 39% | 20% | 20°7 Source: Survey of 49 BC Transit feeder users. TCRP B-IA III-21 Dray Final Report
rather Han "difficult." However, based on information from the focus group and the survey, most feeder riders do find that feeder trips take more personal effort and are less comfortable than direct paratransit trips. The situation at Pierce Transit is quite different. As shown In Table TIl-3, feeder trips take much more time than equivalent direct paratransit trips. In addition, 73% of feeder riders surveyed described their transfer from SHUTTLE to a bus as "difficult." Approximately 80% of riders rated direct service as better than feeder service win respect to ad aspects of service, Including time, convenience, comfort, independence, and safety. The reasons for the different outcomes are clear. BC Transit uses feeder only for trips that are very difficult to provide on direct paratransit, while Pierce Transit requires feeder for essentially aD trips for which it is feasible (a practice consistent with ADA). BC Transit provides most of its transfers to a high-frequency, light rail line, with level- entry boarding and alighting, while Pierce Transit provides transfers to a bus network, portions of which run with long headways and which requires transfers between buses for many trips. BC Transit has an informal system of determining which riders can use feeder, while Pierce Transit has a comprehensive process that requires many riders to use feeder. Demand for Parat~ansit Depending on whether riders experience feeder service as providing a higher or lower level of service than direct paratransit, it may be expecter] that He demand for para~ansit win be increased or reduced if riders are required to use feeder for some trips. Data from Pierce Transit, shown in Table IlI-5, suggest that He service level TCRP B-IA III-22 Draft Final Report
Table IlI-5. Demand Impacts of Feeder and Other Service Changes . ~ One-way paratransit tI ips in first two weeks Category of Rider of April Change . _ | Before Feeder(Apri} | After Feeder 1995) (April 1996) . Feeder Riclers 1,235 473 -62% (414 3B-eligibles) _ _ Non-feeder Riders ~ 23,319 ~ 19,338 ~ -17% ~ Sources: Feeder riders-from Pierce Transit computerized dispatch records of 414 riders registered prior to April 199S, later determined to be Ineligible, adjusted to count feeder and distribution links as one trip. Non-feeder riders-from Pierce Transit records of total paratransit ridership, adjusted for feeder usage. TCRP B-1A lII-23 Draft Final Report
differences described earlier have resulted In a dramatic reduction in demand. Paratransit ridership by a group of 414 paratransit riders, later determined to be 3B- elig~ble, declined service by 62% based on a comparison of two weeks of data before and after implementation of feeder service. (Note that the "after" data represent conditions approximately nine months into the feeder program.) The other service and policy changes implemented during the same period (see earlier description) resulted In a 17% reduction in overall paratransit demand. Since the 3B riders faced all the same changes as other riders, Me demand reduction due to feeder alone can be calculated as having been approximately s4%.s These results provide a basis for escorting the change in annual demand for SHUTTLE service resulting from the feeder program as follows: Prior to service changes: Reduction due to non-feeder service changes: After non-feeder service changes: After feeder and other changes: Change due to feeder: Pierce Transit's ADA eligibility screening administrator has estimated that there win be approximately 900 registrants certified as 3B eligible by the end of the ADA certification process. Using the trip rates just given, total SHUTTLE ridership by this group will be about 26,640 trips per year, a reduction of 31,320 trips per year due to the feeder program. This reduction represents about 6% of total SHUTTLE ridership. Trips per 3B-Eligible per Year 77.6 17% 64.4 29.6 34.8 Despite the mandatory nature of Pierce Transit's feeder program and the associated demand reduction, it appears that 3B eligibles continue to make a majority of their SHUTTLE trips without transferring. The data shown in Table IlI-5 indicate that s Calculated as (1 - 0.38/0.83) = 1 - .46 = .54. TCRP B-IA III-24 Draft Final Report
the 414 people who had been SHUTTLE riders since at {east April 1995, were making about 2.5 trips per person per month in April 1996. If this same trip rate applied to all 721 people who had been certified as 3B eligible by May I, 1996, then they would be making about 1,800 trips per month. Since actual feeder usage at this time was about 600 per month, it appears that feeder trips account for roughly a third of total SHUTTLE travel by 3B eligibles. This result is reasonably consistent with the findings of the survey of 3B eligibles in which 41% of those using SHUTTLE said they had been required to use feeder service at some time, and 25% said most or all of their trips require transfers to f~xed-route service. Impact on Mobility The survey of 3B eligibles explored the impact of the demand reduction on individuals' lives. If most of the trips no longer being made on SHUTTLE were now being made on fixed-route transit, that would be a positive accomplishment from the point of view of meeting the objectives of ADA. The survey was sent to all those certified as 3B eligible, regardless of how often they used SHUTTLE. Of 208 respondents, 72% had used SHUTTING prior to becoming 3B eligible. Of these riders, 46% had stopped using SHUTTLE completely since their 3B designation. Their reasons, given in Table ITI-6, show that many stopped riding for personal reasons, unrelated to the feeder program. However, many stopped riding because of the expected difficulties of using the bus instead of the SHUTTLE. Respondents who had stopped riding SHUTTLE or were riding it less were asked how they travel now. As shown in Table TTT-7, those who usually ride on the TCRP B-1A III-25 Draft Final Report
fable IIL6. Reasons for No Loner Using Sbu"Ie . . Reason Percentage of Respondents (~=~} ~ Donlt need it more 6~ . Pied He rues Ho long Ruben S4~ . b~s~rr~g ~ ~ bus . . Cam Dab outside for long 53 ~ . Had He bus Ho d~cuR ~ use 49~ Others 46~ . Ides add ~ Cog ~1~ He ~ file Spoke. boost hewed -~ ''ogre' masons were reload ~ demoed physical condition of me mspo~ent. ~ 7 ~IIL26 ~/~'
Table IlI-7. Current Movies of Fanner Shuttle Riders . . Percentage of . Response Respondents : (n=152) . Usually ride with a family member or friend 25% . Usually ride on Pierce Transit buses 9% Do not go out as often as before 45% Ot hers 21% aIncludes "call a cab" and "condition has worsened." TCRP B-IA III-27 Draft Final Report
Pierce Transit buses make up only 9% of the respondents, while those who do not go out as often as before make up 45%. These results indicate that the mandatory feeder program has resulted in a significant reduction in personal mobility, even though the feeder option complies fully with ADA requirements. AD respondents were asked whether the availability of free fixed-route fares would increase their usage of the buses. Only 9% said they would use the buses more often, and 40% stated that they cannot use the bus. Additional User and Community Perspectives Participants in the focus group of Pierce Transit 3B registrants expressed views consistent with those found in the survey. Participants described very long wait times, sometimes in inclement weather (possibly similar to those experienced by users of fixed rate bus service). Some indicated that they had reduced the number of their leisure trips since being designated 3B eligible, or had begun conducting more errands by mail or telephone. Two participants did describe positive impacts of feeder service which, based on the survey results, are atypical. One was forced to try out the bus and now uses it exclusively. A second had to move out of his parents' home into the Pierce Transit service area and now uses fixed-route service on a regular basis. He enjoys his new independence but resents the fact that, in his view, Pierce Transit forced him to make these changes. Members of Pierce Transit's advisory committee were generally supportive of the concept of feeder service. The Independent Living Center (TALC) representative strongly endorsed the concept as fully consistent with the ILC's Philosophy of ~-- -r ~ mainstreaming people with disabilities. However, she believes that the certification TCRP B-1A III-28 Draft Final Report
process is resulting in people being offered feeder service who should be traveling on direct SHUTTLE service. This ILC has appealed the 3B certifications given to their clients. Over a period of two years, 35% of the 3B certification appeals have resulted in the appellant's eligibility category being upgraded. Participants In the focus group of BC Transit feeder riders appreciated the independence offered by fixed-route service. They find transferring to SkyTrain more comfortable and faster than handyDART for very long trips. The participants did have trouble coordinating with two handyDART operators. More than half of the participants used three-legged trips on a regular basis, with a different operator on each end of the fixed-route portion. Participants noted problems with open, wind-swept SkyTrain stations, especially during winter, and fear of crime at night. Economic Evaluation One motivation for implementing feeder service is to reduce We cost of providing para~ansit service, particularly complementary paratraTIsit service required by the ADA. Feeder service can also be used to increase mobility and provide more hips within a fixed budget. A full calculation of the economic impact of feeder service requires consideration of the foDow~ng factors: . Implementation and planning cost Continuing added operational cost for scheduling, coordinator, and trip planriing. Condn`iing operational cost savings resulting from reduced paratrans* mileage per trip for feeder service compared to direct para~ansit service. TCRP B-IA III-29 Draft Final Report
. Continuing operational cost savings resulting from reduced demand for paratransit. Each of these factors is considered in turn. Implementation and Planning Cost Since feeder was not implemented as a distinct program at BC Transit, no implementation and planning costs could be identified. At Pierce Transit, there would have been some costs associated with developing criteria and scheduling procedures, eligibility certification, and outreach to Me riders. Staff did not keep track of the costs of these activities and regard them as mom al, since they were integrated into other work O involved In SHUTTLE operations and policy changes. For example, 3B certification was completed as part of a larger program of recertifying aD registrants according to ADA · - criteria. Scheduling, Coordination, and Trip Planning At BC Transit, all the schedulers interviewed agreed that, for two-legged trips, feeder trips required no more telephone scheduling time than direct paratransit trips. As for three-legged trips, involving both a feeder and distribution leg on paratransit, these typically involve two different paratransit operators, and most of the operators require the rider to make arrangements win both operators. One operator which does coordinate with the other operator esUmatec! mat coordination with the other operator acids five to ten minutes to the scheduler's time for a three-legged feeder trip. At Pierce Transit, an analysis of time spent by call takers planning feeder trips indicated that the average time for trip planning was about 16 minutes. At the average TCRP B-IA III-30 Draft Final Report
wage for call-takers of $16.90 per hour (including administrative overhead and fringes), the cost of trip planning is about $4.51 per round trip. Based on discussion with Pierce Transit staff, we estimate that the roughly 600 monthly feeder trips provided as of April 1996 required the following trip plans: 210 non-repeat trips (about 35%) require a trip plan each time. Assuming each trip plan covers a round trip, 105 trip plans per month are needed. About 30 individuals who make repeated trips require about one trip plan per year, or 2.5 per month. About 60 trip plans are prepared and then rejected by riders. Overall about 168 trip plans would be needed per month costing about $757 per month, or about $~.26 per trip provided. At the estimated feeder demand of 750 trips per month, once ADA certification is complete, the cost of pre-tr~p planning would be about $~l,340 per year. Operational Savings from Reduced Paratransit Vehicle Time Based on Me data shown earlier, the paratransit vehicle time needed to provide the feeder and/or distribution legs of a feeder trip is much less than Mat needed to provide a direct paratransit trip between the same origin and destination. Table ITI-S shows a calculation of the operational cost savings from this reduction In paratransit vehicle time at BC Transit and Pierce Transit, which comes to about $14.44 per trip at BC Transit and $16.41 per trip at Pierce Transit. Since BC Transit provides about 800 feeder trips per monk, the reduced cost per trips results In an annual savings of about $139,000, roughly I.3% of the annual TCRP B-1A III-3 1 Draft Final Report
Table TIl-~. Calculation of Paratransit Vehicle Time Cost Savings Element of Calculation BC Transit Pierce Transit Average Duration of Parabansit Portion of Feeder Trips 21 minutes Average Duration of Direct Para~ansit Trip between Same 52 minutes Origin and Destination Reduction In Passenger-Minutes 31 minutes Average Vehicle Occupancy I.Oa Reduction In Vehicle-Minutes 31 minutes . Cost per Vehicle Hour (U.S. $) $28.10 Cost per Trip Savings | $14.44 17 minutes 38 minutes - 21 minutes 1.09b 19.3 minutes $51.00 $16.38 a Estimation of paratransit trip durations made no allowance for detours to serve other passengers. b Estimated from 1994 Section 15 data: (4,197,672 passenger miles) / (3,840,040 vehicle-miles) = 1.09 passengers per vehicle. Note: Results may not appear exact due to rounding. TCRP B-IA III-32 Draft Final Report
paratransit budget. From BC Transit's perspective, Me feeder service allows handyDART to accommodate another 10,400 passenger trips annually win no increase In budget. At Pierce Transit's estimated feeder trip rate of 750 trips per monk once ADA certification is complete, the reduced cost per trip would result In annual savings of about $147,000. Operational Cost Savines Due to Reduced Demand At BC Transit, feeder service provides travel opportunities that might not exist otherwise, and did not previously exist. Therefore, mere is no basis for calculating any savings from reduced demand. As calculated earlier, Pierce Transit's mandatory feeder program wid result in a reduction in paratransit clemand of about 31,320 trips per year once ADA certification is complete. At SHUTTT~E's average cost per trip of $22.63 (based on the first quarter of 1996), the demand reduction implies a cost savings of about $709,000 per year. Summary of Cost Savings For BC Transit, the only documented cost impact is a savings resulting from ~ ~ O shorter paratransit trips of about $139,000, roughly 1.3% of We annual parabansit budget, which translates to an ability to provide 1.3% more passenger trips. For Pierce Transit annual cost savings can be summarized as follows: Cost savings clue to reduced paratransit vehicle time: Cost savings due to reduced demand: Added cost of trip planning: Net cost savings: $147,000 709,000 -11 000 $845,000 TCRP B-IA III-33 Drag Final Report
This sum represents about 7.5% of the annual budget for SHUTTLE. Note that most of the savings results from reduced demand. If there were no demand impact, the cost savings would amount to about 1.2% of the annual SHUTTLE budget. Does Cost Savings Depend on Trip Length? In planning the evaluation, it was hypothesized that cost savings per trip would depend on trip length, since the cost of pre-trip planning would be constant while cost savings from reduced paratransit vehicle time would occur only for long trips. In the case of Pierce Transit, the cost of pre-trip planning is $~.24 per trip provided, while the savings on paratransit vehicle dine is $16.41 per one-way trip. The question is: does this savings depend on limiting feeder to trips of some minimum length? At SHUTTLE's operating cost of $2.51 per vehicle-mile, the pre-trip planning cost is equivalent to the cost of about 0.5 vehicle-miles of operation. Therefore, feeder win save money as long at the para~ansit portion of Me average feeder hip is 0.5 miles shorter than the direct paratransit mileage for the same trip. In practice, Pierce Transit achieves an average mileage difference of 5.7 nodes (3.1 miles paratransit portion vs. S.S miles for the same trips on direct paratransit). At BC Transit, the cost of pre-trip planning is considered negligible. Therefore feeder would save money as long as the paratransit portion is shorter by any amount than the direct paratransit mileage. In practice, the handyDART operators achieve an average mileage difference of 10.4 miles (4.9 miles paratransit portion vs. 15.3 miles for the same trips on direct paratransit). TCRP B-IA III-34 Draft Final Report
Standing Order vs. One-time Trips A question In implementing feeder is whether it is cost-effective for one-time trips which require a new trip plan every time, or only for standing order trips which can use We same trip plan many times. Since the savings per trip is much greater than Me pre-trip plar~iing cost in both case study sites, it appears that feeder would save money even if each trip hack to be planned, that is even if none of the feeder trips were standing orders. However, this calculation does not consider the inconvenience to riclers and operators on feecler trips, especially one-Ume and short feeder trips. IMPLEMENTATION ISSUES--LESSONS LEARNED The case studies provide a number of lessons that win be useful to other systems ~plement~ng feeder service. Voluntary or Mandatory Feeder Service The evaluation clid not prov~cle direct evidence of whether a voluntary feeder program would be effective in an ADA environment. Pierce Transit's program is mandatory. BC Transit's is theoretically voluntary, but riders do not have a direct paratransit alternative for most feeder trips. The two agencies' experiences suggest that a voluntary program win not attract significant usage unless travel Ones and convenience are viewed as better than direct paratransit. Achieving this may be relatively easy with frequent rail transit, but win require sophisticated schecluling capabilities with less frequent bus service. Eligibility Screening In a mandatory feeder program, an effective eligibility screening program Is essential. Pierce Transit unplemented a comprehensive program of deterIruning feecler TCRP B-IA III-35 Draft Final Report
eligibility in conjunction with Me recertification of its entire registrant base for ADA eligibility. Pierce Transit has been criticized by some members of the community for designating some people as 3B eligible who should really be getting direct paratransit service. The agency is beginning to use ~n-person assessments with a medical or rehabilitation professional when necessary to refine the determination of the distance that a person is able to walk or wheel unassisted. The long wait times that occur with feeder suggest a need to make a careful determination of the time that a ricler can wait outdoors, especially in cold or rainy weather. Feeder Trip Scheduling Pierce Transit has found Mat can takers do not always schedule the most efficient or quickest possible feeder trip. Therefore, where car-taking and schecluling are separate functions, there would be value in having schedulers or dispatchers involved in trip planning, or having them review trip plans made by call takers. Both BC Transit and Pierce Transit prepared trip plans manually. Automated scheduling systems, taking into account fixed-route service and accessibility information, may help prepare more efficient trip plans. The Critical Role of Wait Times The experience of both systems demonstrates the critical role of wait time at transfer points in creating reasonable feeder trip times. Where fixed route service floes not run frequently, modifications to standard paratransit scheduling practices may be cles~rable to create acceptable wait times at transfers. Coordination with fixed-route personnel, for example informing fixec3-route c3 rivers when a transfer is planned, may help recluce transfer problems. TCRP B-IA III-36 Draft Final Report
Community and Rider Relations Bow systems benefited from an excellent, cooperative relationship wad ctisablect riders and the communing. At Pierce Transit, a history of excellent and weD-funded paratransit, comb~nec! with an intensive educational campaign about Me need to reign in expenses, helped create support for feeder service despite We fact Mat most riders perceive * as a clear recluction in service level. At BC Transit, feeder is perceived as providing travel opportunities that would not overwise exist. Selecting Trips for Feeder In adclition to a system for determining which riders should use feeder, a system for determining which trips are appropriate for feeder is needed. Pierce Transit created an elaborate system for determining which paratransit trips should be provided using feeder. This system Is necessary for ensuring Mat ah feeder trips are at least feasible. The results of the economic evaluation suggest ~at, beyond determining feasibility, a strict trip selection process is not necessary to ensure that feeder service is cost-effective. A stricter selection process than used by Pierce Transit may be necessary if feeder is to appear as a clesirable alternative for riders. BC Trans* anolies feeder onIv for tries ~ ~ ~ r which are very difficult to provide using direct paratransit, with the result that riders clo perceive it as an attractive alternative. Standing Orders and Single Trips It may be thought that feeder service works best for repeated, standing order trips. Since most of Me feeder trips provided at bow case study systems are for repeated trips, it appears that riders and staff do find feeder more usable for these trips TCRP B-IA III-37 Draft Final Report
Man for single, non-repeated hips. However, the cost analysis shows that feeder can be cost effective for either type of trip. Travel Training Pierce Transit implemented an aggressive program of travel training for ah riders certifiect as 3B eligible. The evaluation did not determine to what extent this training contributed to riders' ability to use feeder. BC Transit die! not have an active program of travel training at the time of the case study. Should Parairansit Vehicles Wait at Transfer Points? In plar~iing feeder service, some systems may consider having paratransit vehicles wait at transfer points to assist In the transfer to a fixed-route vehicle. Neither of the case study agencies has such a policy, so the case studies did not determine what Impact it wouic3 have on vehicle scheduling, passenger ride times, or the attractiveness of leerier service. Pierce Transit's experience with wait Hines at transfers does suggest that there would be long delays for vehicles and riders unless scheduling policies can be tailored to provic ing very efficient transfers. These delays would have cost implications as well. Transfer Locations BC Transit uses SkyTrain and SeaBus stations, while Pierce Trans* uses the transit centers of its Anew transfer bus network. AN of these locations have accessible pathways, shelters, benches, telephones, and accessible rest room facilities. Considering the long wait times, and long total travel times, Mat occur for feeder trips at Pierce Transit and are likely to be repeated at many transit systems, these characteristics are probably essential ones for feeder transfer locations. ACRE B-IA III-38 Draft Final Report
Comparability to F=ecI-Route Travel Times The results from Pierce Transit suggest that much of the attractiveness (or even the usability) of paratransit may depend on travel times that are substantially better than those available on fvced-route transit. The analysis found that average travel times on hypothetical direct paratransit trips were about 43% less than those using fixed-route transit for the same trip. 3B riders were required to switch one-third of their trips to feeder service, which offers travel times about 19% longer than those using fixed-route transit and involves transfers for each trip. As a result, these riders reduced their paratransit travel by 54%. This result suggests that overall paratransit demand would probably be much less if paratransit travel times were more comparable to f~xed-route travel times, or if many trips required transfers between paratransit vehicles. Fare Coordination Because of the difficulty in keeping all handyDART drivers supplied with Fare Saver tickets to be used as transfers onto fixed route service, BC Transit's paratransit operators chose on their own to forego charging a fare for paratransit trips. Pierce Transit also elected to provide the paratransit portion of feeder trips at no charge. This arrangement may not be appropriate or practical for all systems. APPLICABILITY AND TRANSFERABILITY BC Transit's experience is most applicable to those systems that have high- frequency accessible rail service. The applicability of lessons regarding feeder service at BC Transit is limiteci for systems In Me United States by the fact that BC Transit is not subject to the requirements of the ADA. As a result, the service provided by handyDART is somewhat more limited than would be the case in many U.S. systems. TCRP B-IA III-39 Drays Final Report
However, systems with good raid or high-capaci~ bus service and a very congested road network may be able to offer feeder service win a similar travel time advantage over direct paratransit. BC Transit's experience also shows that it is possible to provide feecler service In a paratransit system with decentralized operation with multiple private providers. A number of features of service at Pierce Transit need to be considered In determining how the case study results would apply to other transit systems: . . . . The Pierce Transit SHUTTLE offered a very high level of paratransit service before the period In which feeder service was Implemented. Systems offering a lower level of paratransit service might experience less demand reduction due to feeder service than Pierce Transit. Pierce Transits timed transfer bus network provides a set of natural transfer points for feeder service at which bus departure Hines are closely controlled. Systems without such a network of transfer points wiD need to put more effort into determining appropriate locations for feecler transfers, and might have more difficulty scheduling transfers. Pierce Transit included strict screening for ability to use feeder as part of its comprehensive rescreening of registrants for ADA eligibility. Systems which have already completed rescreening without considering ability to use feeder will need to make special arrangements for feeder eligibility screening. Weather conditions in Pierce Transit's service area and in Vancouver are not as severe as those at many Midwestern and eastern transit systems which experience extreme coic! and snow. Systems In such areas win need to make TCRP B-IA III-40 Draft Final Rep or!
additional allowance for riders' ability to wait outside or limit transfers to more sheltered or heated locations. . . . Pierce Transit's paratransit scheduling process is structured entirely around ensuring an on-time arrival at destination points. This arrangement ensures that fixed-route transfers are not m~sseci, but may contribute to long wait times. Systems where schedules are created based on a w~nc30w around promised pick up time may need to implement special procedures to ensure smooth transfers. However, appropriate procedures may succeed in reducing wait times compared to those observed at Pierce Transit. Since Pierce Transit's paratransit fare Is equal to its discount f~xed-route fare, riders had no fare incentive to use feeder service compared to direct paratransit. As a result the evaluation gives no evidence on the extent to which a fare incentive helps to make feeder service attractive to riders. Pierce Transit provides all call-taking and scheduling, and most dispatching for paratransit trips using its own operations staff. This arrangement gives the agency great control over aD aspects of operations, and at least provides the opportunity for coordination with fixed-route operating staff. Systems with contracted paratransit service, especially those with decentralized service, may have more clifficulty establishing effective procedures. Note, however, that BC Transit's feeder service does work with contracted, decentralized para~ansit operations. TCRP B-lA III-41 Draft Final Report
