Click for next page ( 29


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 28
RELATED INFORMATION AND IMPACTS Extent of Vanpooling and Buspooling Numbers of Vanpools The first vanpool program is credited to the 3M Company, implemented in 1973 at their 3M Center outside St. Paul (Comsis and ITE, 1993). During the remainder of the 1970s and into the early 1980s, vanpooling grew dramatically. The number of vanpools in organized U.S. and Canadian programs doubled each year in the 1974 to 1980 period, reaching 8,100 in 1980. In early 1981, the U.S. Department of Energy estimated that there were about 12,183 such vanpools at 853 sites in the United States, under sponsorship of 697 employers, third parties, and other formal organizations (Pratt and Copple, 1981). In addition, as of 1979 it was thought that there might be 3,000 to 5,000 owner-operator vanpools (Pratsch and Starling, 1979). Taking the lower estimate for owner-operator vanpools, there may have been, circa 1980, some 15,000 vanpools in the United States. A major impetus for the vanpool growth leading up to the early 1980s was the oil crises of 1974 and 1979, with associated gasoline shortages and longer term gasoline price increases (Pratt and Copple, 1981). With lower energy costs in the 1980s, vanpooling decreased. The most precipitous decline may have been in Houston, for energy-industry-related reasons described in the previous section. Houston vanpooling slipped from a peak of 1,885 vanpools in October 1981 to 453 in August 1988, a 76 percent drop, while in the rest of Texas, the vanpool census remained close to 580 vans (Texas Energy, 197888). A 1984 estimate placed the U.S. total at 10,000 vanpools with 100,000 participants. This estimate was generally accepted for the next 10 years. In 1991, the Nationwide Personal Trans- portation Survey found about 0.3 percent of all work trips nationally being made in a shared ride vehicle with 5 or more occupants (Comsis and ITE, 1993; van der Knaap, 1996). Transportation Systems Management in general, and thus vanpooling specifically, received a boost following the passage of the Clean Air Act (CAA) of 1990 and its subsequent implementation (Comsis and ITE, 1993; van der Knaap, 1996). However, the mandatory aspects of the Employer Commute Options (ECO) element of the CAA, otherwise known as the Employer Trip Reduction program, were relaxed at the end of 1995. This relaxation was thought to have had an adverse impact on van- pooling nationwide, even though alternative voluntary programs (VEMPs) have a role in mobile source emissions reduction. The Vanpool Council of the Association for Commuter Transportation (ACT) estimated there were about 8,500 vanpools operating as of early 1999 (Boylan, 1999). A more recent but less comprehensive review has concluded that the numbers are now again increasing (circa 20012002), and puts the total in the United States at 10,000 vanpools once more. This review attributes resurgence to Commuter Choice tax benefits now available in the United States (Enoch, 2003). Tax free benefits for vanpool costs were established at $60 per month by the Comprehensive Energy Policy Act of 1992. These benefits, like the similar transit pass benefits, were raised from $65 to $100 tax free per month for 2002 (with provisions for escalation) under the Transportation Equity Act for the 21st Century of 1998 (Federal Transit Administration, 2004). Meanwhile, a 9-fold growth of the transit provider component of vanpooling between 1984 and 2001, measured in fleet size, is documented in the next section. This growth brings transit system vanpooling up to a total approaching 4,000 vans. As if to underscore the difficulty of getting a good handle on total numbers of vanpools, especially owner-operator units, a 1995 study done in support of tax revenue analysis for the Commuter Choice tax benefit legislation arrived at a much larger estimate of vanpooling than any of those 5-28

OCR for page 28
reported above. Starting with the 1990 Census, this study estimated there were 309,886 employees vanpooling in 1995, a 0.3 percent mode share. That figure was translated into a vanpool parking demand nationwide of over 44,000 parked vanpools (KPMG Peat Marwick, 1995). This result is equivalent to a vanpooling estimate 3 to 5 times the size of any other published estimate encoun- tered for any year between 1980 and 2002. Vanpool Operating Organizations Major shifts have taken place over time in the types and mix of vanpool operating organizations. In the mid 1970s, employer programs dominated and the only other type was owner-operator van- pools (Pratt and Copple, 1981). Third-party vanpooling independent of one-on-one employer involvement began to emerge in the 197678 period with the resolution of significant institutional barriers (Heaton et al., 1981; Pratsch and Starling, 1979). By the mid-1990s, the estimated mix of program types was 25 percent employer sponsored vans, 65 percent third party vanpools, and 10 percent owner-operator vans (van der Knaap, 1996). The 1999 ACT estimate is composed of about 2,000 vanpools (24 percent) operated by individual employers, 4,000 (47 percent) operated through "municipal" organizations (including transit pro- viders), and 2,500 owner-operator vanpools (29 percent). Note the definitional differences. Within all of these three categories, but particularly within the "municipal" category, many vans are now supplied and maintained through for-profit vanpool service organizations. At the time, the largest such provider was VPSI with some 3,500 to 3,700 vans. Enterprise was next with only 100 vans, and there were several with still smaller fleets (Boylan, 1999). As noted, transit providers are included within the "municipal" category of the ACT vanpooling estimates. Examination of National Transit Database (NTD) totals for U.S. public transit agency vanpool operations indicates steady vanpool growth in this category on two counts. First of all, the size of individual operations has continued to grow overall, even though individual systems may have setbacks related to local economic conditions. Secondly, new vanpool system starts that more than counterbalance system closures have been expanding the number of operators. The combined effect is a transit provider vanpool growth from 447 vehicles in maximum service in 1984 to 3,932 vehicles in 2001, an expansion by almost 9 times in 18 years. Much of the growth has been fairly recent, as illustrated in Table 5-11 (Wambalaba, Concas, and Chavarria, 2004). The one caveat that must be emphasized is that, at least in some of the largest operations, average transit provider van- pool loadings have been decreasing (see Table 5-4, for example). Thus it is doubtful that there has been a 9-fold increase in transit system vanpool passenger trips. Table 5-11 Growth in U.S. Transit Provider Vanpools Operated in Maximum Service Year Vanpools Year Vanpools Year Vanpools 1984 447 1990 612 1996 1,919 1985 488 1991 930 1997 2,545 1986 524 1992 1,045 1998 3,329 1987 581 1993 1,227 1999 3,580 1988 661 1994 1,503 2000 3,692 1989 486 1995 1,533 2001 3,932 Source: Wambalaba, Concas, and Chavarria (2004). 5-29

OCR for page 28
More in-depth transit provider vanpool statistics have been extracted for 1994. In that year, 55 out of 5,973 U.S. public transit agencies (0.9 percent) operated vanpools. The 2,361 vanpools involved (apparently a fleet size statistic) represented 2.0 percent of the total transit vehicle fleet. The 6 mil- lion trips that these vanpools carried were 0.07 percent of all transit trips, but produced 204 mil- lion passenger miles, 0.5 percent of the total, at an operating cost of $17 million, 0.09 percent of all transit operating expenses. The average one-way vanpool passenger trip length was 32.4 miles, compared to the average unlinked transit trip length of 4.9 miles (Gross and Feldman, 1996). Relative Buspool Market Share Buspooling surfaced as a recognized urban transportation mode in the late 1960s, at least half a decade before the invention of formal vanpool programs. For many markets, buspools have been superseded by vanpools with their lower unit cost and ability to serve smaller trip concentrations. Various buspool or subscription bus applications remain viable, however, as described under "Response to Vanpool and Buspool Programs"--"Buspools (Subscription Bus)." In 1990, a nation- wide survey of commuter transportation organizations indicated that there was roughly one bus- pool commuter for every 10 vanpool commuters. More precisely, the average buspool versus vanpool split of subscription commuting was 8.3 per- cent buspools and 91.7 percent vanpools for programs of public agencies including transit providers, 12.7 percent buspools and 87.3 percent vanpools for employer programs, and 2.9 percent buspools and 97.1 percent vanpools for Transportation Management Associations (TMAs) and similar non- profit organizations (Spence, 1990). Although the survey in question was not primarily focused on determining mode shares, and thus not structured statistically toward that end, the information is some of the most detailed available for the distribution among ridesharing modes. It is reproduced in Table 5-12. Mode shares for the vast body of uninvolved companies and other uninvolved orga- nizations are by definition not included. Table 5-12 Average 1990 Ridesharing Mode Shares of Commuter Program Populations Served Type of Organization Buspool Share Vanpool Share Carpool Share Non-Profits (TMAs; other ridesharing or 0.02% 0.67% 13.77% commute management organizations) Private Companies (any entity offering 0.7 4.8 6.4 commute programs to their employees) Public Agencies (governments at all levels; 0.1 1.1 3.8 regional bodies; transit agencies) Note: Of survey respondents, 43 percent were in California, 13 percent in the remainder of the West, 22 percent in the Midwest, and 22 percent in the East. Many types of organizations being involved, the population served ranged from eighty (80) to seven million (7,000,000). Source: Spence (1990). Demographic Characteristics of Riders Golden Gate, Maryland, Chicago, and Seattle vanpooler characteristics and attitudes that can be directly compared are included in Table 5-13. These data suggest that a high percentage of the van- 5-30

OCR for page 28
poolers in major metropolitan areas hold professional, technical, or management jobs. Vanpoolers holding either professional/management or support/sales jobs constitute 86 to 96 percent of these four samples. The Golden Gate and Maryland vanpooler income data, and comparisons with overall service area demography, indicate a predominantly middle to upper-middle income market for vanpooling. The market is characterized by employees with stable employment and fairly regular hours traveling long distances (Dorosin, Fitzgerald, and Richard, 1979; Stevens et al., 1980). In suburban Chicago, Pace vanpool participants tend to have somewhat higher household incomes than other Pace tran- sit riders, and to be largely indistinguishable from the general population (Michael Baker et al., 1997; Pace Suburban Bus Service, 1993). Vanpooler characteristics surveyed in Minneapolis suggest income levels similar to Golden Gate and Maryland, but with fewer workers in the managerial and professional categories (47 percent as compared to 71 to 73 percent). Minneapolis demographic characteristics were found to be indis- tinguishable from those of auto commuters at the employment sites served. Of Golden Gate Corridor vanpoolers, 93 percent rarely worked overtime and 95 percent rarely needed their car for work; the corresponding percentages for Minneapolis were 86 percent on both counts. In Norfolk, 80 percent of vanpoolers reported regular working hours. Knoxville vanpoolers had lower incomes, only 20 per- cent were in managerial and professional categories, and 7 percent reported no automobile available (Heaton et al., 1981). Table 5-13 Demographic Characteristics of Vanpool Riders in Major Metropolitan Areas Characteristic Golden Gate Maryland Seattle Chicago Average Age 40 years 41 years 42 years 44 years Sex 63% male 57% male 55% male 46% male Marital Status 78% married 72% married n/a n/a Average Income ($ 1985) $37,000 $45,000 $37,000 (family) Household Income ($ 1993) $50 - $75,000 Occupation Professional, Technical 55% 58% 53% 32% Manager, Administrator 16% 15% 14% 42% Clerical, Sales 18% 19% 19% 22%a Crafts, Operators, Laborers 8% 2% 10% 1% Service 1% 0% 3% 1% a Other 2% 6% 2% 1% Overall Satisfaction (good/adequate or better) 99% 91% n/a 94% Note: a Sales included under "Service." Sources: Dorosin, Fitzgerald, and Richard (1979); Stevens et al. (1980); Conway Associates (1986); Pace Suburban Bus Service (1993). 5-31

OCR for page 28
As the limited Knoxville data barely hints at, the vanpooler characteristics data of Table 5-13 can- not possibly be fully representative on all counts. The prevalence of vanpooling and buspooling in connection with large shipyards suggests that there must be another largely undocumented potential vanpooling and buspooling market spectrum. In 1980, on the order of 200 owner-operator vanpools were carrying 15 percent of all 12,750 AM peak hour person trips entering the core area of Newport News, Virginia, dominated by the Newport News Shipyard and Drydock Company. Another 10 percent were carried in some 30 pri- vately operated buspools (Pratt and Copple, 1981). Nearly half of the 25,000 military and civilian employees of the shipyard in Portsmouth, Virginia, were reported to be commuting via buses, vans, or carpools around 1990 (Keesling, 1991). In 1998, some 5 to 8 percent of civilian employees at the Bremerton, Washington, Naval Shipyard were commuting via buspools using a program in place since World War II (see "Response to Vanpool and Buspool Programs"--"Buspools (Sub- scription Bus)"). Other Bremerton shipyard workers are served by Kitsap Transit vanpools. Al- though a number of the van and buspool riders at the Newport News, Portsmouth, and Bremerton shipyards may be administrative personnel, a substantial percentage must be blue collar riders, given both the gross numbers involved and anecdotal evidence as well. Finally, there is the Talihina, Oklahoma to Ft. Smith, Arkansas shuttle, also reported under "Buspools (Subscription Bus)," oper- ated for welfare-to-work poultry processing workers. Data from the late 1970s third-party vanpooling demonstrations indicate that drivers tended to be slightly older, better educated, and from higher income households than passengers, with nearly all of them being married males (Heaton et al., 1981). In 1980, Massachusetts found that 74 percent of their regular and backup drivers were male (Morris, 1981). It is not known whether this aspect of the driver profile has changed over time or not. Sources of New Ridership and Vanpooler Turnover In examining sources of new vanpool (or buspool) ridership, it is potentially useful to differenti- ate between new or relatively new vanpool programs and ongoing programs. All but the most recent available surveys of vanpooler prior travel modes have focused on new or relatively new programs. For ongoing vanpools, new vanpoolers are needed only to replace driver and rider turnover. The only available information on turnover itself pertains to rates of turnover, presented toward the end of this subsection. Prior Mode of Travel When vanpools serve central area employment in corridors with heavy transit service, a substan- tial proportion of the vanpoolers may be drawn away from transit use. For example, both Mont- gomery Ward Chicago vans and Golden Gate Vanpool Demonstration Project vans serving the downtown San Francisco commute attracted over half their riders from conventional bus or rail services. This is shown in Table 5-14. The Golden Gate van-versus-bus competition was a deliber- ate attempt to head off further expansion of the deficit financed bus service without sacrificing highway lane productivity (Dorosin, Fitzgerald, and Richard, 1979; Johnson and Sen, 1977). Table 5-14 also illustrates what is presumably the effect of external events. The shift in the latter stages of the Golden Gate Demonstration toward more vanpoolers who previously drove alone probably reflected not only a greater focus on suburban employment destinations, but also the impact of the 1979 gasoline shortage and price-at-the-pump increases (Dorosin, 1982). 5-32

OCR for page 28
Table 5-14 Former Commute Mode of 1970s Chicago and San Francisco Vanpoolers Chicago San Francisco Golden Gate Demonstration First 9 Months First 9 Months Last 10 Months Former Mode Wards (Downtown) (Suburban) (All Markets) Drove car alone 15% 10% 25% 33% Carpool 29 23 74 33 Drop off/other 2 -- -- -- Regular transit 53 62 1 34 Buspool -- 5 -- -- Source: Johnson and Sen (1977); Dorosin, Fitzgerald, and Richard (1979); Dorosin (1982). Most vanpool and buspool operations tap a predominantly new travel market as compared to more traditional mass transit. It is a market shared to a degree by carpooling, however. The late 1970s demonstration projects illustrate typical results, summarized in Table 5-15, with respect to ridership sources. Table 5-15 Former Commute Mode of Late 1970s Demonstration Project Vanpoolers in Four Areas Former Mode Knoxville Norfolk Golden Gatea Minneapolis Drive Alone 36% 52% 15-33% 27% Carpool 54 33 35-33 65 Transit 10 3 50-34 8 Private Hauler 0 12 0 0 Notes: a First 9 months - Last 10 months. Source: Heaton et al. (1981), Dorosin (1982). More recent findings are similar to those displayed in Table 5-15. The "Caravan" third-party van- pool program in Massachusetts launched 34 vanpools in 1980. Of its participants, 46 percent pre- viously drove alone, 44 percent carpooled, and 10 percent took the bus (Morris, 1981). A 1987 survey in the Hampton Roads, Virginia, area found the prior mode of vanpool participants to be roughly one-third solo driving, one-third carpool, 13 percent transit riders, and somewhat less than a quar- ter "another vanpool." This may be one survey that reflects mainly turnover-replacement vanpooler characteristics (Keesling, 1991). Table 5-16 presents additional prior mode data that unquestion- ably reflects predominantly turnover-replacement vanpooler prior modes. This information was obtained in a circa-2000 cross-sectional survey of the vanpools of King County Metro, serving the central county of Greater Seattle, and presently and for many years the nation's largest vanpool operator. 5-33

OCR for page 28
Table 5-16 King County Metro Vanpooler Prior Commute Modes, Systemwide Sample Prior Modes Samples Percent Prior Modes Samples Percent No Commute 24 1.9% Bike 2 0.2% Drive Alone 642 51.2 Walk 3 0.2 Carpool 263 21.0 Bus 192 15.3 Vanpool 127 10.1 Total 1,253 100% Note: The existence of bike and walk prior modes suggests that some travelers surveyed may have been involved in a different (shorter) commute prior to vanpooling. Source: Cambridge Systematics and Urban Analytics (2003). The survey samples summarized in Table 5-16 were used to develop a behavioral model to estimate the former mode of vanpoolers, intended for application in conjunction with an experience-based vanpool usage calculation. The model shows high drive-alone operating cost to be an indicator of above-average likelihood to have switched from another high occupancy mode, while good em- ployment accessibility via transit indicates above-average likelihood to have switched from public transit specifically. Higher household auto ownership is, as expected, associated with higher inci- dence of the drive-alone prior mode (Cambridge Systematics and Urban Analytics, 2003). Based on the findings from new vanpool programs and the two available examples from ongoing programs, the one major compositional difference in the prior modes of new versus ongoing pro- grams is that ongoing programs do have vanpooling as a significant prior mode. This presence of vanpooling as a former mode appears to be roughly counterbalanced by fewer prior carpoolers. In any case, prior solo drivers constitute roughly a quarter to slightly over a half of the vanpoolers and buspoolers in examples of commuting to non-CBD workplaces. Total prior auto drivers, count- ing in carpool drivers (but discounting alternate drivers), are in the 45 to over 65 percent range (Heaton et al., 1981; Morris, 1981; Keesling, 1991; Pratt and Copple, 1981; Cambridge Systematics and Urban Analytics, 2003). Vanpooler Turnover The Hampton Roads survey indicates a moderate degree of stability among ridesharing arrange- ments. Of those surveyed, two-thirds had been in current arrangements for 1 year or more while one quarter were enrolled for 6 months or less (Keesling, 1991). In the late 1970s third-party van- pooling demonstrations, passenger drop-out rates averaged well under one rider per month per van in Norfolk and Minneapolis, and less than 5 percent of all registered vanpoolers in the Golden Gate Corridor demonstration. Nine months into the Golden Gate demonstration, 32 drivers had been used to operate 30 vans. The average driver turnover rate in Knoxville during the last 6 months of the demonstration was 2.6 drivers per month, representing 7 percent of the operating vanpools. Principal reasons for leav- ing a vanpool, as reported in the Minneapolis and Golden Gate surveys, appeared to be higher than anticipated vanpool fares, inability of low income passengers to pay a monthly fare, insufficient flexibility and convenience, and changes in commuting needs (Heaton et al., 1981). The Spring of 1993 survey of Pace vanpoolers in suburban Chicago revealed that 4 percent had been Pace vanpool members for less than 3 months, 24 percent for 3 to 6 months, 66 percent for 5-34

OCR for page 28
6 months to one year, and 6 percent for over a year (Pace Suburban Bus Service, 1993). These results reflect in large measure the major influx of new vanpools, roughly a doubling of the fleet, when Sears moved to the suburbs about 6 months previous. Even in a more stable situation, with vanpooling in operation for one to two decades and more, rider and vanpool turnover may be high. Puget Sound area operators identified an annual turnover in both ridership and vanpools of 40 to 50 percent. Most of this turnover is ascribed to job-related reasons (WSDOT, 2000), and probably results not only from job changes and residential reloca- tions, but also reassignments within multi-location firms. An important aspect of attracting vanpool participants is arranging matches. An alternative to computer matching is provided by the Commuter's Register. The Register is published every month for distribution to 70,000 to 100,000 commuters in Connecticut, New York, New Jersey, and Massachusetts. It has over 1,500 listings for ridesharing, as well as transit route and schedule infor- mation. Monthly telemarketer monitoring indicates a rideshare success rate of 25 to 35 percent. In a June 1990 survey, publishers found that of people finding a high occupancy solution for their commute, 36 percent began ridesharing, 8 percent increased the size of their pool, and 50 percent began taking the bus (Urban Transportation Monitor, July 20, 1990). Indicators of Market Potential Vanpools and buspools are almost exclusively oriented to serving work trips. Vanpools are nor- mally most successful where one-way trip lengths exceed 20 miles, work schedules are fixed and regular, employer size is sufficient to allow matching of 5 to 12 people from the same residential area, public transit service is inadequate, and other conditions exist such as congestion or a short- age of parking. Nevertheless, strong employer commitment in cases of either employer-sponsored programs or partnerships of employers and third-party operators can help overcome conditions that are otherwise not ideal. Vanpooler Trip Lengths The average person trip lengths for vanpools tend to be much longer than for carpools or transit. Vanpool pickup and dropoff time becomes less onerous in the context of a longer overall trip, and cost savings increase, adding to the attraction of vanpools for long trips. Practically all vanpool program one-way trip length averages fall within a range of 24 to 54 miles, with the lower end of the range being more common. Drawing upon data presented elsewhere in this chapter, it can be shown that this range covers the averages for the El Segundo, California, Aerospace Corporation employer vanpool program (35 miles); the third-party demonstrations in Knoxville, Minneapolis, Norfolk, and the Golden Gate Corridor (27 to 30 miles, see Table 5-3); and third party programs in Connecticut (36 miles) and Massachusetts (33 miles). (The Aerospace Corporation and Connecticut values are van rather than person mileage, and thus somewhat overstated.) The 7 largest transit provider and other public agency vanpool programs as of 2002 also fall within the 24 to 54 mile range (see Table 5-5 and the discussion immediately following). These large programs are, in fact, what has been used here to define the range (National Transit Database, 2002). These vanpool person trip length figures also bracket the results of a 1990 nationwide survey of commuter transportation organizations, in which average vanpool one-way trip lengths were reported as 32 to 35 miles (Spence, 1990). The 3M employer vanpool program is the most notable 5-35

OCR for page 28
exception: in the 1970s their person trip lengths averaged approximately 17 miles one-way (Owens and Sever, 1974 and 1977). The 24 to 54 mile vanpool person trip length range stands in contrast to the national solo-driver average one-way commute trip length of 10.5 miles reported in the 1990 National Personal Transportation Survey, and the transit rider unlinked trip average of about 5 miles. Service Attractiveness Guidelines The ratio of maximum passenger pickup and delivery time to line-haul travel time was proposed in the early days of vanpooling as a useful rule of thumb measure with which to judge the attrac- tiveness of individual vanpools and buspools. This "Utility Ratio" or "service ratio" describes the travel time quality of the vanpool trip in terms of the ratio of residential pickup time to line-haul time. Although users accept long vanpool travel times, there is a limit to the time spent picking up and dropping off passengers, perhaps relative to driving time with a full load or perhaps in the absolute, that will be tolerated. The Utility Ratio measure assumes the limit is relative. This concept is examined further within the case study, "The 3M Company Employer Based Vanpool Program." Pace vanpoolers complained about the time required to pick up passengers in a 1993 survey (Pace Suburban Bus Service, 1993). In the original 3M pilot vanpooling program, vanpools with a ratio of residential pickup time to line-haul time of up to 1.0 proved successful, while problems were encountered with forming vanpools in areas where the ratio would be greater than 1.0 (Owens and Sever, 1974 and 1977). Other evidence, provided by Maryland vanpooling and buspooling experience, suggests the Utility Ratio is often lower than 1.0. The total time spent picking up and dropping off passengers was 14.0 minutes for the average Montgomery County vanpooler com- pared to 40.1 minutes enroute time, an average Utility Ratio of 0.35. The corresponding figures for other Maryland vanpoolers were 22.6 minutes pickup or dropoff and 37.4 enroute, for an aver- age Utility Ratio of 0.60 (Stevens et al., 1980). Although the Utility Ratios for individual vanpools assuredly vary significantly around these mean values, it is easy to imagine that most lie well below 1.0. There is some evidence to support the alternative proposition that the tolerance limit for time spent picking up and dropping off passengers is an absolute, rather than relative, limit. In the Min- neapolis third party vanpooling demonstration project analysis, it was found that the absolute cir- cuity time increment was roughly constant regardless of commute distance. It has been noted that this finding was consistent with empirical evidence from Australia on carpool spatial structure. In Minneapolis, the average vanpool time increment over the drive-alone time was found to be about 12 minutes for vanpool passengers and 22 minutes for van drivers (Heaton et al., 1981). A much simpler measure has also been offered: the suggestion that the economics and time analyses only begin to look favorable for vanpooling when one-way trip lengths approach 20 miles (Comsis and ITE, 1993). Theoretical Market Potential The 1993 Federal Highway Administration report Implementing Effective Travel Demand Management Measures estimated the potential market for vanpooling by looking at the distribution of U.S. worker population by size of employer and one-way trip distance. (The distribution is reproduced in Table 5-17). The analysis relaxed the 20-mile threshold, and assumed that the potential vanpool market would include trips of 11 or more miles for the largest employers, 16 or more miles for medium- large employers, and 21 or more miles for medium-small employers. The market potential thus cal- culated was 11 percent of all U.S. workers. 5-36

OCR for page 28
Table 5-17 Cumulative Distributions of U.S. Work Trips by Employer Size, Trip Length, and Both Parameters Combined Trip Distance (in Miles) 30+ 21+ 16+ 11+ 6+ All Employer Cumulative 3.4% 8.4% 14.9% 25.0% 46.3% 100.0% Size Distribution 500+ 25.0% 0.8% 2.1% 3.7% 6.3% 11.6% 25.0% 100+ 50.0% 1.7% 4.2% 7.5% 12.5% 23.2% 50.0% 50+ 61.6% 2.1% 5.2% 9.2% 15.4% 28.5% 61.6% All 100.0% 3.4% 8.4% 14.9% 25.0% 46.3% 100.0% Source: Comsis and ITE (1993). Next, a success rate of 50 percent of the resulting market was assumed. With this, a vanpooling goal of 5 percent of the U.S. worker population was obtained (Comsis and ITE, 1993). Restricting the analysis to include only those workers with trips of more than 20 miles, but with the same assumed success rate, yields an alternative overall vanpooling goal of 2 to 3 percent of U.S. (or region-wide) work trips. It is instructive to compare actual vanpooling experience in the Greater Seattle area of Puget Sound with these theoretical market potential estimates of 5 percent, or alternatively 2 to 3 percent, of work trips. The observed Greater Seattle vanpool mode share had already reached 2 percent of the overall commuter market as of 1999 (WSDOT, 2000; Enoch, 2003). This market share has been achieved in the context of geographic and institutional factors capable of being replicated in large measure but not completely in other areas. The geographic feature that cannot be replicated is a large body of water immediately west of Seattle and Everett that is directly crossed only by ferries, which in turn offer priority vanpool access with substantial time savings including certainty of get- ting on board in peak loading hours. The Washington State Ferries also waive vanpool vehicle and driver fares for registered vanpools. As previously noted, ferries in 1999 were carrying 11 percent of public system vanpools and roughly 60 percent of the approximately 200 private vanpools (WSDOT, 2000). The institutional factors of the Puget Sound area, largely amenable to replication, include an exten- sive HOV lane system and--most importantly--a series of legislative acts that cause many large employers to proactively support alternative transportation for their employees, require adequate public facilities for new developments (leading to further travel demand reduction efforts), and provide trip reduction support and assistance (WSDOT, 2000; Enoch, 2003; Samdahl, 1999; Kavage and Samdahl, 2004). The importance of this state legislation in encouraging vanpooling is under- scored by the previously mentioned statistic identifying employers involved with Commute Trip Reduction as being served by 93 percent of area vanpools (WSDOT, 2000), implying relatively lit- tle vanpooling to non-involved employers. The Washington State Department of Transportation conducted a study of the market potential for vanpooling in the Greater Seattle area that concluded a 7 percent market share could reasonably be attained. This estimate encompasses all those projected (on the basis of a survey) to rely on the automobile for the trip to work, commute at least 10 miles each way, and have an interest in van- 5-37

OCR for page 28
pooling. To achieve this market share would require major penetration of worksites not presently involved with Commute Trip Reduction, to the point where a majority of vanpools would be focus- ing on such employers (WSDOT, 2000). Employer Participation The one major consideration not addressed by these theoretical market potential analyses is the propensity for employers to get (or not get) involved in vanpooling programs even when urged, either as employer-sponsors or in partnership with third party operators. Under present condi- tions, a relatively small proportion of U.S. employers overall are under any type of mandatory trip reduction requirement. Although voluntary rates of employer participation have never been researched for vanpool pro- grams per se, the proportion of larger and smaller firms offering ridesharing assistance in the early 1980s was examined in Atlanta, Cincinnati, Houston, Portland, and Seattle as part of the National Ridesharing Demonstration Program. The average rate of employer participation in ridesharing was found to be 36.8 percent for firms with 100 or more employees and 4.0 percent for smaller firms (Booth and Waksman, 1985). Applying these percentages to the national goal calcu- lations presented at the start of the previous subsection results in horizon estimates for nation- or region-wide vanpooling somewhat less than 2 percent of work trips for the 5 percent goal, and 1 percent for the alternative 2 to 3 percent goal.5 A return to nationwide mandatory trip reduction would move these horizon estimates closer to the national goal calculations. It bears repeating that the present national utilization of vanpooling is estimated at some 0.3 percent of all work pur- pose travel. Impacts on VMT, Energy, and Environment Vanpooling is the least energy intensive of four-or-more-wheeled urban transportation modes, which is to say that vanpooling is estimated to consume the least propulsion energy per passen- ger mile. The reduction in number of vehicle trips and VMT that results from commuters switch- ing to vanpooling, taking into account prior travel modes and all possible energy requirements, leads to substantially reduced fuel consumption. There has not been comparable evaluation of bus- pooling, but buspools probably have an energy intensiveness similar to or somewhat better than conventional bus service, depending on the extent to which the bus vehicles are or are not parked at the trip origins and destinations (Pratt and Copple, 1981). A 198081 analysis of the then-new third-party vanpool program in Massachusetts found the daily round-trip VMT per participant had dropped on average from 43.1 to 10.5 miles, a reduc- tion of 76 percent. These estimates were the result of taking into account the mix of previous modes of travel and the access mode to the vanpool; the average vanpooler round trip was actu- ally 66 miles. The 76 percent decrease in VMT was somewhat more than the estimated percent- age decreases in fuel consumption and emissions, because the van and short-distance auto-access trips had higher per-mile fuel use rates than long automobile line-haul trips. Each Massachusetts vanpool saved an average of 26.2 gallons of gasoline daily or about 6,548 gallons per year. For 5 Employer participation calculations such as these are applied in the "Projected Effectiveness of Individual TDM Strategies" section of the "Implementing Effective Travel Demand Management Measures" report (Comsis and ITE, 1993), and are examined further in Chapter 19, "Employer and Institutional TDM Strategies," of this TCRP Report 95, "Traveler Response to Transportation System Changes" Handbook. 5-38

OCR for page 28
each vanpool group, the fuel reduction was 66 percent with a per commuter reduction of 1.9 gal- lons per day. The same study also calculated hydrocarbon emissions reductions from the Massachusetts van- pools using information on VMT reduction and vehicle cold starts. Each vanpool was estimated to reduce the non-methane hydrocarbon (NMHC) emissions by 2.62 pounds each day of operation. On an annual basis, this equated to an emissions reduction of 0.33 tons (55 percent). For the aver- age vanpool group, the 4.79 pounds per day released by the vanpoolers in their previous modes dropped to 2.17 pounds per day for the vanpool group (Morris, 1981). Both the energy and emis- sions savings would be different and presumably less today, with nearly two decades of automotive fuel economy and pollution control improvements. Nevertheless, these early 1980's computations serve as a model for taking into account prior mode and access mode influences that are all too often inappropriately ignored. The Pace VIP vanpool program serves as a component of the Chicago region's air quality improve- ment program. The 1996 daily impacts estimated for the 252 vanpools then in operation are listed in Table 5-18. The calculations are adjusted for mode of access. They do not rely on vanpooler reports of prior mode, because of the large number of person trips involved that have relocated to the suburbs from Chicago's central area. Instead, the estimation relies on rider survey reporting of current alternative modes. The estimated volatile organic compounds (VOC) reduction of 0.0666 tons per day constitutes 2.5 percent of the 2 to 3 tons budgeted for Transportation Control Measures (TCMs) in the 15 Percent Rate of Progress SIP for 1996. The TCM-generated emissions reductions are a small but still vital portion of the region's overall emissions reduction budget (Michael Baker et al., 1997). Table 5-18 Estimated Air Quality Benefits of 1996 Pace VIP Vanpool Program Measure of Effectiveness Effectiveness (Daily Impacts) Number of Vanpools 252 vanpools Number of Vanpool Commuters 2,423 commuters Daily Vanpool Person Trips 4,846 person trips Vehicle Trip Reduction 2,529 vehicle trips Vehicle Miles of Travel (VMT) Reduction 119,956 vehicle miles VOC Emissions Reduction 0.0666 tons NOx Emissions Reduction 0.156 tons CO Emissions Reduction 0.639 tons Note: VOC reduction adjusted for cold starts for 38 percent of participants and model improvements. Source: Michael Baker et al. (1997). The cost of obtaining the emissions reductions credited to Pace VIP vanpooling is essentially lim- ited to the purchase price of the vanpool vehicles, given that operating costs are almost entirely supported through fare revenue. With 252 vehicles having a standard useful life of 4 years and a replacement cost of $27,000 each, the cost of reducing 0.0666 tons of VOC emissions is estimated to be $7,000 per day, or $51 per pound of VOC emissions (Michael Baker et al., 1997). If the cost were to be distributed over other benefits, such as congestion mitigation, parking needs reduction and mobility, the emissions reduction component would obviously be much reduced. 5-39

OCR for page 28
Revenue/Cost Considerations Vanpooling has established itself as a comparatively cost-effective commuter service option. Al- though wide variation is possible in vanpool expenses, including program administration costs in particular, the use of a volunteer driver helps to hold costs down. Most employer-sponsored van- pool programs have been priced so as to recover vehicle and operating costs, but typically provide a private subsidy covering costs of program administration and support. Some third-party pro- grams seek to cover all costs, but most have elected to use public subsidies for certain program administration, overhead and promotional costs, or alternatively, for capital costs. For transit providers operating vanpool systems, the vanpools typically enjoy a high fare recovery ratio, which contributes to the overall transit agency performance (Suhrbier and Wagner, 1979; Michael Baker et al., 1997). Owner-operator vans are normally supported by user charges alone, although the owner may choose to absorb certain costs to keep the vanpool viable if the vehicle has other, personal value. The FTA's Capital Cost of Contracting program helped to fund the vanpool program of the San Diego Association of Governments (SANDAG). Each of the 130 vanpools operating in mid-1997 received a $300 per month subsidy from SANDAG. Participants in the FTA's subsidy program are required to report monthly ridership, travel time, and mileage data of the subsidized vanpools for the National Transit Database (MetroPool, 1997). Federal funds constituted 80 percent of the 1997 Pace (suburban Chicago) $28.6 million VIP vanpool capital budget. Congestion Mitigation and Air Quality (CMAQ) funds were the major component, along with Section 3 discretionary and Section 9 apportionment funds, and Surface Transportation Program and other flexible funds. The balance of the capital program was made up with Regional Transit Authority discretionary funds and Illinois DOT funds. Operating costs of the core Pace VIP vanpool program are virtually all covered by fares: the cost recovery ratio was 92.42 percent in 1995 and 105.27 percent (estimated) in 1996. The overall vanpool cost recovery ratio is lowered some- what by the inclusion of ADvAntage vanpools, which serve the physically and mentally disabled having regular employment or workshops to attend. The ADvAntage vanpools posted a 69 percent cost recovery ratio for 199596, still a considerable savings, given Americans with Disabilities Act (ADA) requirements, over the cost of serving these trips with regular ADA paratransit. The overall Pace cost recovery ratio for 199596 was 36 percent (Michael Baker et al., 1997). King County Metro annual vanpool program operating costs were $2.6 million in 2000. These costs were covered by vanpool fares, grants, and income from the self insurance reserve and sale of vans over 5 years old. These same sources also covered 45 percent of the $1.5 million annual adminis- tration costs. The other 55 percent was covered by public subsidy (Enoch, 2003). Most vanpool programs either charge a flat per person fee or a distance or zone based fare. Some programs may have additional fees for added services such as guaranteed ride home programs. Typically monthly fares as of 199899 were in the $70$120 range. Specific examples are provided in the "Response to Vanpool and Buspool Programs" section and in the case studies. Third-party providers keep fares low through economies of scale with large fleets and the benefit of federal cap- ital subsidies. Employer sponsored programs keep fares low by absorbing administrative, insurance, and sometimes maintenance costs. Vanpools have administrative time costs associated with their formation and the replacement of lost riders. Employee transportation coordinators can play an important role in minimizing these costs. Third-party providers often help with the marketing and administration of programs, including the recruitment of drivers and riders (Comsis and ITE, 1993). 5-40