Click for next page ( 122


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 121
Travel Planning in the Randstad: An Evaluation Based on ReMOVE This review of Dutch travel plans by Touwen in 1999 concluded that travel plans consisting of communication/marketing measures, basic measures such as carpooling and cycle leasing, and organizational measures such as flextime achieved an average reduction of 8 percent in vehicle kilometers traveled. However, if luxury measures or disincentive measures were added, the aver- age reduction was about 20 percent (Cairns et al., 2002). Smarter Choices--Changing the Way We Travel This study by the Cairns organization itself tracked and reported on its own selection of 26 workplace travel plan case studies from locations in Birmingham, Bristol, Buckinghamshire, Cambridgeshire, Merseyside, Nottingham, and York. Although the study reports trip reductions, the strategies employed were not specified. The report seems to have been more oriented to exploring the process by which the programs were engineered, their costs, and projections of their possible effects at dif- ferent levels of geographic resolution (Cairns et al., 2002). Site- Versus System-Level Impacts The vast majority of travel demand management effectiveness data presented up to this point in Chapter 19 has been measured and analyzed at the site level, providing findings about TDM impact on travel to and from particular offices, campuses, and other specific employment and institutional sites engaging in specific sets of TDM strategies. Most of this data addresses only the effect on travel produced by the journey to and from work or school. Some results have been presented at a pro- gram level, such as employee or student pass programs facilitated by public transportation agen- cies but implemented at the site level in conjunction with individual employers or institutions. This focus on site-level work-commute-trip impacts is consistent with the scope of this chapter, which (in accord with its title) addresses employer and institutional TDM strategies and not programs more or less independent of employers such as residence-based programs or area-wide agency pro- grams. Examples of broader agency-implemented programs are wide-area transit improvements and regional rideshare matching programs. Indeed, a number of important objectives associated with TDM apply at the site level, including employee and through-traffic congestion mitigation at site access points, site traffic pollutant emissions reduction, parking facility needs reduction, and employee benefits. Nevertheless, the question remains about the broader effects of site-based TDM on all employees within an urban sector, on overall peak- period area street traffic, on major transportation facilities, and on regional vehicle trip making and VMT. Information about such broader effects would be useful for a quantitative understanding of TDM efficacy in addressing objectives such as regional transportation facility congestion mitigation, energy conservation, improved regional air quality, and reduction in global warming gas emissions. Overview of System-Level Impact Relationships As one moves from the site level toward the broader area level or regional traffic facility level the following stages of impact dissipation apply (Pratt, 1990): Stage 1. A leveling out of effects among differing participating employer and institutional scenar- ios, reflecting various types and sizes of employers and TDM programs, and averaging out of high-effectiveness and lower-effectiveness programs. 19-121

OCR for page 121
Stage 2. Dissipation of work-trip effects as average site-level effects are mixed with the unaffected commuting associated with non-participating employers. Stage 3. Additional trip-reduction dissipation as work-commute trips are mixed in with non-work site-generated travel produced by both participating and non-participating employers and institutions along with other interspersed land uses. Stage 4. Further leveling-out of impacts as locally generated trips and traffic are intermixed with other trips and traffic. On transit systems and local area streets only intra-regional through movements will normally be of significance, but on regional highway facilities, other traf- fic will include intercity travel. These stages are referred back to, often with examples, in the discussion which follows. Impacts at the site level vary according to employer type, most notably whether an office, retail, or industrial employer is involved, and size of employer. Although most studies have found no significant relationship among TDM program effectiveness levels for large, medium, and medium- small employers, it has generally been presumed that fewer TDM actions are workable for firms of less than 100 employees (Comsis and ITE, 1993). According to 2004 U.S. Census data, 36 percent of all employees work for firms with less than 100 employees (U.S. Census, 2007).16 Degree of Employer Participation Site-level TDM impacts, even once averages are calculated (Stage 1 dissipation as identified above), apply only to employers actually participating in the employer-dependent TDM strategies. These impacts are diluted, on an employment-area level, by non-participating employers (Stage 2 above). The degree of employer participation or non-participation is heavily impacted by the regulatory environment. Where no legal requirement compels participation, employer involvement is volun- tary, and participation generally reflects the employer's self-interest. Where a legal requirement for TDM program participation derives from the administration of land use and zoning regulations, the requirement may directly apply only to new sites and not to pre-existing employment. Full mandatory participation, such as applied in Southern California's air quality Regulation XV pro- gram, does not really apply to all employers. "Full mandatory" participation typically applies only to employers of 100 (sometimes 50) or more employees. Further background on effects of employer type and regulatory influences is provided in the "Underlying Traveler Response Factors" section under "Voluntary Versus Regulatory Employer Motivation" and "Characteristics of Employer." Degree of employer participation has been approximated in some analyses by means of "employer participation rate" averages. Evaluating degree of site-level impact dilution by non-participating employers requires associating employer subgroups with their average employer participation rates (Comsis and ITE, 1993). Table 19-34 gives default experience-based employer participation rates for voluntary employer participation from the "FHWA TDM Evaluation Model" and indicates their empirical and/or judgmental basis. As can be seen, these rates were based on very limited expe- 16 National employment statistics such as this provide only an approximate indication of employment site size distributions and almost certainly understate the prevalence of smaller-size establishments. Self-employed persons without payroll are excluded, and large employers may have secondary places of business which, from an employee transportation perspective, are similar to small businesses. 19-122

OCR for page 121
riential data and considerable extrapolation therefrom. (See the "Additional Resources" section for more on the FHWA TDM Evaluation Model.) Empirical voluntary employer participation rates have been reported in the aggregate for the United Kingdom. Termed "level of travel plan take-up," the observed rates apply to all degrees of travel plan scope and intensity. (The United States equivalent of a travel plan is a Transportation Management Plan, or TMP.) The concept of the travel plan arrived in the United Kingdom from the United States in the mid-1990s, after first being taken up by the Netherlands. The U.K. Department of Environment, Transport, and the Regions (DETR) is pursuing "widespread voluntary take-up of travel plans" according to a 1998 white paper. The highest U.K. employer participation rates have been observed among local government enti- ties. In 1997/1998, 3 percent had TMPs on a permanent basis and another 4 percent had TMPs on a pilot or trial basis, for a total of 7 percent participation. The total had risen to 24 percent by 2001 and 62 percent by 2006. Participation rates are observed to be much more modest in the U.K. pri- vate sector. Of firms with 100 or more employees, 4 percent had permanent TMPs in 1997/1998, with no data available on pilot or trial TMPs. The total of permanent and experimental TMPs equated to a 7 percent participation rate in 2001 and rose to 11 percent participation in 2006. Small businesses of less than 100 employees have been found to be relatively neglected. Circa 2000, only 19 percent of small businesses were even aware of the travel plan concept (Enoch and Zhang, 2008). 19-123

OCR for page 121
Table 19-34 Experience-Based Voluntary Employer TDM Participation Rates Program Voluntary Employer Basis of Value Sources of Observed Type Participation Rate a (Rates Observed) Participation Rate Values Carpool 37% for firms with 100 or Average for firms in Transportation Research Programs more employees Atlanta, Cincinnati, Center, Urban Mass Trans- Houston, and Seattle portation Administration, 4% for firms with less than encompassed by "National Ridesharing 100 employees Ridesharing Demon- Demonstration Program: stration Program b Comparative Evaluation Report." Cambridge, MA (1985). Variable Same as carpool programs Weighted average of 6% Pratt and Copple (1981) Work (deemed to fit logically for six North American Hours with the observed values, regions and 28% for six which weren't available CBDs (16% weighted by firm size) average overall) Other Non- Same as carpool programs None available None available Monetary TDM Transit 7% for firms with 100 or Taken at 20% of carpool Pratt and Copple (1981) Fare more employees program rates based on Incentives comparisons available for 1% for firms with less than Boston c, d 100 employees Other Same as transit fare None available None available Monetary incentives TDM Notes: a As utilized in the FHWA TDM Evaluation Model. b The National Ridesharing Demonstration Program report indicated a higher overall voluntary employer participation rate for minimal-effort, low-impact carpool programs but without any breakout by firm size. c Of Boston area firms selling transit passes to their employees, 7.6 percent subsidized them. This was 20 percent of the carpool program employer participation rate for firms with over 100 employees. The number of firms subsidizing transit passes (56) was also 20 percent of the number of firms with variable work hours programs (285). d Programs such as Denver's Eco Pass and Seattle's FlexPass (see "Response by Type of Strategy" -- "Changes in Fare Categories" -- "Unlimited Travel Pass Partnerships" in Chapter 12, "Transit Pricing and Fares") offer the potential for updated participation rate calculations but none such have been encountered. Source: Pratt (1992). 19-124

OCR for page 121
Intermediate Effects of Dissipation Continuing briefly with U.K. experience, it is of interest that two estimates have been prepared of national-level commute trip vehicle travel reductions attributable to travel plan implementation. DETR research published in 2004 produced estimates that commute trip vehicle kilometers of travel (VKT) reductions in the range of 0.4 to 3.3 percent had been achieved. This finding is complemented by a slightly earlier independent estimate that the travel-plan-attributable reduction in commute VKT had possibly been 1.143 billion kilometers per year, or 0.74 percent of the auto commute total (Enoch and Zhang, 2008). These U.K. estimates address the dissipation in Stages 1 and 2--and to a certain extent Stage 4--as described above, but not the Stage 3 intermixing of commute travel with travel for all other trip purposes. In the United States, as of the early 1990s, results for a number of programs that had taken an area- wide approach to TDM were not suggestive that TDM was an effective broad-area strategy. Many of these efforts were associated with TMAs that had largely depended on voluntary cooperation by employers or developers and lacked the technical guidance and legal clout to move beyond a marketing-based approach. Lacking decisive actions, these programs had not proven to be instru- ments of significant change (Comsis and ITE, 1993). During a somewhat longer contemporary time frame, California saw widespread adoption of mandatory employer-based trip reduction, implemented primarily through city and county ordi- nances and air quality district rules. The objectives were congestion mitigation or pollutant emis- sion reduction. These programs were terminated or drastically altered by California legislation signed in January 1996 that prohibited mandatory programs. The reported results of these manda- tory trip-reduction programs do not encompass the full sequence of employer-based TDM impact dissipation, but do provide a look at the work-commute-trip reduction averages achieved by all non-exempted firms across whole political jurisdictions (Stage 1 dissipation as identified earlier). Exempt firms were typically those with fewer than 100 employees (Dill, 1998). Table 19-35 lists 13 such programs, including two from outside California, and indicates the drive- alone mode share and reduction averages achieved. Drive-alone rate reductions somewhat overstate vehicle trip reductions, since the carpooling alternative diminishes but does not fully erase vehicle trips. Nonetheless, they are a useful indicator. As can be seen by examination of the table, the aver- age drive-alone work-commute reduction was 3-1/2 percentage points. The range, excluding the highest and lowest outliers, was from no effect to a 7 percentage-point drive-alone trip reduction among covered employees. Results from SCAQMD for the 5-year period 19881993 provide additional information for the greater Los Angeles region. Employers with over 100 employees were required to submit plans and status annually, although somewhat longer times between plan submissions was not uncommon. Implementation of the mandatory program was gradual, and started with the region's largest employers. Data for 817 sites with 4 or more plans give an indication of average vehicle ridership (AVR) progress over time. The aggregate (weighted average) AVR for these sites was 1.196 at the time of the first plan, 1.220 for the second plan, 1.271 for the third, and 1.288 for the fourth. This equates to a 7.7 percent improvement in AVR, over a span of 3-plus years, equivalent to slightly over a 7 percent reduction in journey-to-work vehicle trips per employee. 19-125

OCR for page 121
Table 19-35 Drive-Alone Work-Commute Travel Mode Shares and Reductions with Mandatory TDM Programs Year Plan Time Span of Change in Drive- Jurisdiction Adopted Reported Data Alone Rate a Menlo Park, CA 1988 1990-1993 -7% Pleasanton, CAb 1984 1985-1995 +3% Central Contra Costa Co., CA 1992 1994-1995 -3% West Contra Costa County 1992 1994-1995 -3% East Contra Costa County 1992 1994-1995 -1% Contra Costa Centre (BART) 1986 1986-1990 -1% Walnut Creek, CA 1988 1988-1990 -4% San Ramon, CA 1989 1990-1995 -3% Concord, CA 1985 1987-1988 0% Northern San Mateo Co., CA 1983 1992-1994 -6% Southern CA (SCAQMD) 1987 1989-1994 -11% Phoenix/Maricopa Co., AZ 1988 1990-1995 -6% Washington State 1991 1993-1995 -4% Note: The source states: "Determining the actual impact of these programs is beyond the scope of this paper... the issue is to determine what is significant or worthwhile." a Percentage-point change in drive-alone share. b The focus of the Pleasanton program was on peak-congestion mitigation, more than trip reduction per se, and the program relied heavily on variable work hours measures. Source: Dill (1998). Looking at all 4,999 participating employment sites with valid information and two-or-more approved plans, the aggregate starting AVR was 1.205. The last reported AVRs, as of November 1993, aggregated to 1.257, up 4.3 percent. The corresponding trip reduction was 4.1 percent over the 1- to 3-plus-year time spans involved. The average commute-trip drive-alone share for the 4,999 sites decreased from 73.5 percent to 67.2 percent, a 6.3 percentage-point reduction. That shift translates into a 9.6 percent drive-alone share decline from the initial to the most current plan. It was largely attributable to increased carpooling, with the carpooling share increasing from 15.5 to 21.4 percent. Concurrently, vanpooling increased from 1.2 to 1.9 percent. The tran- sit share increase was slight, from 4.0 to 4.3 percent. Walk and bike, together 3.0 percent, did not change. While the CWW day-off share increased from 1.3 to 1.9 percent, the telecommuting share dropped from 1.4 percent to 0.3 percent (Young and Luo, 1995). Only 6 percent of regulated sites charged for parking. In their earlier plans, 69 percent of employers offered direct financial subsidies for alternative modes. This investment in vehicle trip reduction declined, however, to 53 percent in the last plans of the 5-year evaluation period. No more than a weak statistical relationship was found between AVR progress and site characteristics and between AVR progress and program duration. Location in the Los Angeles Central Business District (CBD) affected the mix of mode shifts, however. At the 188 CBD sites, carpool and transit shares started at 19-126

OCR for page 121
around 21 to 22 percent. Carpooling increased 4 percentage points and transit increased 3 percent- age points from the first-reported to the last-reported shares. A significant negative relationship was found between AVR progress, expressed as aggregate per- cent gain in AVR, and the initial AVR. Of particular note was the 14 percent average drop in AVR for the 582 sites starting with an initial AVR of 1.5 and above (Young and Luo, 1995). The research paper does not examine possible causes of the program failure affecting the group of sites with a high AVR to start with. Clearly overall average program results, though they remained positive, were dragged down by this outcome. The case study, "Overall TDM Program Effects over Time--Bellevue, Washington," (see this chapter's "Case Studies" section), looks at impacts of the first two stages of dissipation on employee travel mode shares. While estimated employee vehicle trip reductions at key employ- ers were on the order of 30 percent or more in response to TDM program implementation, the overall downtown employee drive-alone commute share reductions were on the order of 10 per- cent over a 20-year time span. (A second data source places the reduction at 2 percent.) In response to both employer TDM and extensive public improvements to bus transit services, the corresponding transit commute share doubled (see case study for sources). End Results of Dissipation Only two fully or partially empirical studies addressing all four stages of dissipation of TDM pro- gram effects, as identified above under "Overview of System-Level Impact Relationships," have been encountered. In the fully empirical study, dissipation as one moves outward from the involved employment area--primarily through intermixing with other travel (Stages 3 and 4 dis- sipation)--was measured in the early 1970s for the introduction of variable work hours for Canadian government workers in Ottawa, Canada. The relative impacts are highly illustrative even though the specific traffic-peaking reductions obtained are uniquely large because of the single- employer dominance of the national government in downtown Ottawa. The primary effectiveness measure in the Ottawa study, given that trip-reduction measures were not part of the variable work hours program, was percentage reduction in the peak-hour to peak- period traffic ratio. The construct of this measure was such that it was equal to the percentage reduction in peak-hour traffic in those instances where there was no change in peak-period traffic volumes. It was applied to both street traffic volumes and transit ridership. The street traffic effect was measured in the PM peak period. The reduction in the peak-hour to peak-period auto traffic ratio was 21.6 percent as measured at parking lots. Non-work travel and other traffic intermixing effects were such that the corresponding reductions as measured at the central Ottawa Cordon and at the Ottawa River screenline dropped to 10.2 and 5.7 percent, respectively. Corresponding PM transit ridership peaking reductions were 23.7 percent for gov- ernment employee transit riders, as measured at their workplace, down to 19.2 percent for all peak-direction transit riders at the central Ottawa Cordon. Peaking reductions for AM peak- period transit riders were 16.9 percent at the workplace and 8.4 percent at the central Ottawa Cordon (Safavian and McLean, 1975). There is no other known fully empirical research on system-level effects of TDM. It is doubtful that broad-area traffic volume reduction impacts of typical site-level employee-trip-reduction pro- grams are amenable to direct field measurement of traffic volume changes given the potential for confounding effects ranging from land development to shifts in economic conditions. Successful 19-127

OCR for page 121
empirical measurement requires rigorous surveys of person- and vehicle-trip making, planned suf- ficiently in advance of initial program implementation, and paired with good luck that brings min- imal or reliably quantifiable exogenous impacts. The partially empirical study encountered is particularly useful in that the methodology not only covers--in effect--all four stages of dissipation but goes further, estimating congestion and emis- sions reduction effects. By so doing, it addresses the importance of even modest traffic reductions to congestion mitigation. The subject of the partially empirical study was an 8.6-mile north-south stretch of I-5, centered on downtown Seattle, along with the employers along the way and their CTR programs. Employers within an approximately 3.6-mile band were considered, including all located in the Seattle CBD. Coverage included 189 employers with CTR programs. The empirical component of the study included the before and after trip reduction strategies of each employer with a CTR program, par- ticipation rates for the nearly 63,000 employees involved, and commute trip origin-destination data. The modeled component started with the estimation of the corresponding VTRs, done using EPA's COMMUTER Model (see "Additional Resources"--"Analytic Tools"). Non-participating employers were accounted for through the non-inclusion of any trip-reduction effects for employers without included CTR programs. The vehicle trips eliminated were then assigned to an already calibrated and evaluated current-conditions CORSIM microsimulation traffic representation of the corridor, to assess what conditions would be without CTR program TDM. For simplicity and con- sidering their function, the analysis treated the I-5 reversible express lanes as ramps and otherwise excluded them from the analysis (Georggi et al., 2007). Not all TDM strategies in place were credited in the analysis, again for simplicity. The included strategies were alternative work schedules, employer TDM support strategies, travel cost changes, and flexible work hours. The estimated VTR average for the employers with CTR programs was between 11.3 and 14.2 percent. The higher number assumed observed parking charges were all attributable to the CTR programs, while the lower number took into account that the basic charges might have existed anyway (but without carpool/vanpool discounts). The higher estimate was adopted for the extended analysis. Note that these average VTR estimates correspond to Stage 1 dissipation of the four stages outlined earlier. The completed traffic analysis (corresponding to Stage 4 dissipation) found a difference in peak period traffic volume on study corridor I-5 on-and-off-ramps averaging about 4 percent. The largest peak-direction changes on any individual ramp were about 50 to 55 percent depending on the peak period (Georggi et al., 2007). Corresponding performance measure effects, more substantial than the volume changes might at first glance imply, are covered in the "Cost-Effectiveness" and "Energy and Environmental Relationships" subsections to follow. In addition to these two empirically-based studies, a few forecasts or hypothetical estimates have been made of system-level effects inclusive of all four stages of impact dissipation. In the mid- 1980s, for example, a precursor of the FHWA TDM Evaluation Model and traffic assignment (sim- ulation) techniques was used to estimate the effect of potential Minneapolis airport-area TDM on adjacent I-494 traffic volumes. The "Low Scenario," employing modest TDM measures in combi- nation with assumption of mixed-use land development, was forecast to offer a 7 to 11 percent vehicle-trip reduction, compared to conventional development with no TDM, as measured at par- ticipating employer/land-development sites. After accounting for all types of impact dissipation, this was projected to equate to a 2 percent average workplace vehicle-trip reduction, including both new and old development, and a 1 percent reduction in peak traffic on the I-494 study-area segment. The "High Scenario," employing mandatory TDM with strong parking management 19-128