Click for next page ( 88


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 87
Pre- and Post-TOD Travel Modes Travel modes of TOD residents or workers before and after relocating to a TOD have been captured in a very few studies, and even fewer have proceeded to translate such data into overall net shifts in travel modes (Hendricks, 2005). The following extraction from available information starts with disaggregate before-and-after residency-change data from California. It then examines overall net mode shifts primarily using documented net shift findings but also on the basis of extrapolation and drawing of inferences. These documented and inferred travel mode shifts upon relocation into TODs range from 2 percentage-point or smaller gains in the transit mode share for commuting (average of surveyed California sites, two statewide studies) to a 15 or 16 percentage-point transit commute mode share gain (two Portland, Oregon, studies, one an 8-site average). The smaller Portland study (Center Commons) also examined non-work trip mode shifts, found to be similar in order-of-magnitude transit mode share impact. The one instance encountered of fully supported overall mode shift data applying to place of work relocation to a TOD is provided last. Disaggregate Mode Shifts with Residency Change to TOD The 1992 California transit-focused development study included an analysis of the usual commute mode at the prior residence in comparison with the usual commute mode at the current, station- area residence. Results are presented in Tables 17-39 and 17-40, disaggregated by current mode and prior mode, respectively. Light rail transit (LRT) station area residents in Santa Clara County were excluded because of the newness of their LRT system. Also, the data analyzed included only station-area residents "whose workplace location did not change between their former and present residence" (Cervero, 1993). This data definition, along with findings reported, suggests inclusion of respondents who changed workplace addresses after moving in as well as those who kept the same employment site. The study specifically noted that the majority of the 15.7 percent of car drivers who switched to rail transit also changed their workplace address. This was inferred to be one more indication of the importance of proximity of the workplace location to rail transit in determining mode choice (Cervero, 1993). 17-87

OCR for page 87
Table 17-39 Pre- and Post-Station-Area Living: Distribution of Prior Mode to Work for the Current Mode to Work Usual Mode to Work Before Living in Station Area Current Usual Mode to Work Drove Rode Car Rail Bus Walked Other Drive 82.0% 2.0% 9.3% 2.6% 3.2% 0.9% Ride Car 65.5 10.3 6.9 10.3 6.9 0.0 Rail 28.8 3.9 42.5 13.7 4.6 6.5 Bus 23.5 5.9 23.5 41.2 5.9 0.0 Walk 40.0 0.0 13.3 20.0 20.0 6.7 Other 20.0 0.0 0.0 30.0 15.4 34.6 Note: Shows distribution of prior mode of users of each current mode (each row totals 100%). For example, 28.8% of current rail commuters drove to work before moving to the station area. Source: Cervero (1993). Table 17-40 Pre- and Post-Station-Area Living: Distribution of Current Mode to Work for the Prior Mode to Work Usual Mode to Work Before Living in Station Area Current Usual Mode to Work Drove Rode Car Rail Bus Walked Other Drive 75.5% 38.5% 18.3% 11.4% 41.2% 25.6% Ride Car 4.8 7.7 1.2 2.9 5.9 5.0 Rail 15.7 46.2 76.8 54.3 41.2 70.0 Bus 1.2 7.6 2.4 17.1 0.0 0.0 Walk 2.0 0.0 1.2 5.7 11.8 0.0 Other 0.8 0.0 0.0 8.6 0.0 0.0 Note: Shows distribution of current mode of users of each prior mode (each column totals 100%). For example, 15.7% of residents who drove to work before moving to the station area are now rail commuters. Source: Cervero (1993). The 1992 California transit-focused development study did not explicitly report either the overall before and after mode shares for survey respondents combined or what percentage of current tran- sit riders were previously transit riders. The 2003 California TOD travel characteristics study derived information similar to that above, but only for the subset of survey respondents who did change workplace location when moving into their TOD (Lund, Cervero, and Willson, 2004a). The latter study also provided certain summary information for this subset, presented below along with comprehensive overall mode shift data for LRT-based TODs in Portland, Oregon. 17-88

OCR for page 87
Overall Mode Shifts from Before to After TOD Residency California. Given the paucity of information on overall mode shifts for persons moving into TODs, the Handbook authors have undertaken to estimate before and after overall commute mode shares from the 1992 California data reproduced in Tables 17-39 and 17-40. This exercise introduces various statistical approximations and uncertainties including recognition that the production of aggregate estimates may have been shunned by the original research team for sound statistical reasons.21 Taken at face value--but making allowance for exclusion from "before" data of Santa Clara County LRT station-area residents with their lower transit and walk shares--the approximated results seem to indicate virtually no change in prevalence of driving a car to work and less than a percentage point mode share increase in taking transit to work or in making the commute by transit, walk, or auto passenger in total. Probably the only overall work mode share changes with statistical significance were a roughly 5 percentage points increase in rail transit commuting with rail transit station-area residency, and a smaller increase in carpooling, counter- balanced by decreased use of bus transit and non-motorized commute modes (Handbook author computations). Work trip mode shares in 1992 after moves to California rail station areas were, for all survey respondents combined, 71 percent drive car, 4 percent ride in car, 19 percent rail transit, and 2 percent each for bus transit, walk, and other (Cervero, 1993). The 2003 California TOD travel characteristics study, as noted above, did publish selected overall prior versus current mode share summaries for survey respondents who "changed both their residential location and place of work." The assumption was that this subset of respondents had a high degree of choice in both their residence and workplace locations. The documentation notes that "before" commute mode data was based on "typical mode used" whereas the "after" commute mode data was based on "actual mode used to commute to work on the day of reported travel" (Lund, Cervero, and Willson, 2004a). This difference introduces uncertainties in the comparison that may possibly be reflected in either understatement or overstatement of individual mode shifts.22 At all surveyed locations across California combined, 11.5 percent of respondents reported shifting from driving alone or carpooling to using rail transit (no one reported shifting to bus) at their new rail-served TOD residence, while 9.7 percent of respondents reported shifting 21 To Work/Return Home mode shares of station-area residents from Table 4.17 of the source document (Cervero, 1993) were used as weights for application against the prior usual commuting mode shares dis- aggregated by current usual mode, enabling computation of weighted-average overall prior commute mode shares. One known approximation introduced in this process is the application of weights based on all responding TOD residents to data from which selected groups had been excluded (Santa Clara County LRT station-area residents and persons changing workplace at the time of their move). A second approximation is the comparison of current mode shares for the full universe of surveyed trips against the prior mode shares derived with the selected groups removed from the universe. 22 "Typical mode used" versus "actual mode used... on the day of reported travel" introduces a definitional difference similar to that between U.S. Census journey-to-work data ("normal" mode) and conventional regional travel data ("survey day" mode). In the case of Census data, the definitional difference has led to application of adjustment factors to the normal mode to obtain survey-day-equivalent data. Questions have arisen about results of applying these factors (Siaurusaitis and Saben, 1998), leaving uncertainty as to the nature of discrepancies introduced by such definitional differences. 17-89

OCR for page 87
from rail or bus transit to driving alone or carpooling. This produced a net estimated shift of 1.8 percent of respondents to transit. When walk, bike, and carpooling commute modes were included along with transit, however, the shifting into these modes from automobile commuting (14.6 percent) was estimated to be exceeded by the shifting out of these modes (17.6 percent). The proportion of transit users selecting rail over bus clearly increased very substantially with rail sta- tion TOD residency, but some elements of this shift were not reported. The 2003 California sample size was deemed large enough in the case of the San Diego Trolley LRT and the BART HRT residential surveys to separately examine the mode shifts of persons moving into TODs along those lines. At surveyed locations along the San Diego Trolley, 10.3 percent reported shifting from automobile commuting to use of LRT, while 6.9 percent shifted from rail or bus to automobile, a net estimated shift of 3.4 percent of respondents to transit. When walk, bike, and carpooling commute modes were included along with transit the shifting into these modes (14.1 percent) was nearly in balance with the shifting out of these modes. At TODs along BART, 17.9 percent of respondents reported shifting from driving alone or car- pooling to HRT, while 13.7 percent of respondents reported shifting from rail or bus transit to driv- ing alone or carpooling. This produced a net estimated shift of 4.2 percent of respondents to transit. When walk, bike, and carpooling were included with transit, the shifting into these modes (19.4 percent) was more than the total shifting out of these modes (17.8 percent). Possible factors listed for the superior performance of BART-centered TODs in inducing desirable commute mode shifts included the greater maturity of San Francisco Bay Area transit systems (including BART), the associated larger share of workplace destinations readily accessible to transit, and the relatively higher prevalence of both parking charges and employer alternative mode subsidies (Lund, Cervero, and Willson, 2004a). The traffic-free rapid transit characteristics of BART HRT in combi- nation with commuter corridor highway congestion surely played a role as well. Portland, Oregon. Resident travel mode changes associated with moving into a TOD-like envi- ronment were derived from 2005 surveys of Portland-area TODs and transit-adjacent development as part of research conducted for the TransNow Center. Residents at 8 different developments at 4 stations on Portland's MAX LRT Blue Line were surveyed. Survey response rates at the individ- ual developments ranged from 24 to 43 percent. Details of the surveys and findings are provided in the "Travel Findings for Individual Portland, Oregon, Area TODs" case study. Given in Table 17-41 are the transit and non-transit mode shifts, for the commute trip, from before residing at the current home to after (Dill, 2006b). A full breakdown of old and current commute modes for all 8 sites combined is displayed in Table 17-54 of the case study. The use of all forms of public transit for commuting in all neighborhoods combined increased by 156 percent (nearly 16 percentage points). Biking and walking increased as well, by 38 percent (a little over 2 percentage points). Data on carpooling was not obtained. Auto use including the "multiple modes" category dropped by 21 percent (18 percentage points) (Dill, 2006a). The good survey response rate, the coverage of multiple developments, and the consistency of mode use definitions between before and after conditions, together make this a robust data source. It pertains, however, only to LRT-based TODs under Portland's highly transit- supportive conditions and medium-sized urban area environment with moderately low "before" condition mode shares. Additional cautions with regard to interpretation and use are provided in the case study. 17-90

OCR for page 87
Table 17-41 Changes in Commute Mode Between Transit and Non-Transit for Residents Moving into Portland Station-Areas LRT Transit Continued Non- Continued Transit Blue Line to Non- to use Non- Transit to to use Percentage- Sample Station Neighborhood Transit Transit Transit Transit Point Gain Size Convention The Merrick 0.0% 74.1% 25.9% 0.0% 25.9% 54 Center Beaverton Beaverton 8.3 58.3 25.0 8.3 16.7 12 Central Round Elmonica/ Condos and SW Arbor Station 0.0 71.0 25.8 3.2 25.8 31 170th Orenco Arbor Homes 6.5 69.6 17.4 6.5 10.9 46 Orenco Original and 6.8 69.5 11.9 11.9 5.1 59 Club 1201 Orenco Sunset Downs 0.0 72.7 18.2 9.1 18.2 11 All 8 Sites 3.8% 70.4% 19.7% 6.1% 15.9% 213 Notes: LRT stations are listed east to west, from the new downtown area east of the Willamette River to the western suburbs. See Table 17-53 for additional details and background. "Differences between neighborhoods not significant." Source: Dill (2006b). Earlier mode change findings from two individual Portland TODs, LRT-served Center Commons and Orenco Station, are also reported in the "Travel Findings for Individual Portland, Oregon, Area TODs" case study. See especially Table 17-52, which for Center Commons presents Pre- and Post-TOD travel modes for both work and non-work trips. The two studies involved address only one individual TOD each, but have the advantage of excellent 39 to 44 percent survey response/completion rates. The one study that derived quantitative percentage-point mode shifts obtained commute mode results consistent with the larger Portland study summarized above. Both of the TODs involved, as noted in the case study, suffered from less than optimal location of the developed area with respect to its LRT station--at least at the time of the resident surveys reported on (Switzer, 2002; Podobnik, 2002). The Center Commons TOD features below-market- rate subsidized housing. Use of the transit mode increased by 48 to 60 percent (15 and 12 percentage points, respectively), depending on whether the trip purpose was commuting to work or non-work activities. Biking and walking decreased for work-purpose trips but increased slightly for non-work trips, while carpooling to work increased. Driving alone dropped by 21 to 24 percent (12 and 14 percentage points) (Switzer, 2002). Orenco Station's original component was and is, in contrast, an upscale market-rate development. More use of transit (bus or rail) than in their previous neighborhood was reported by 69 percent of surveyed households versus 6 percent reporting less. Households in a more typical neighborhood in the same suburban sector reported transit use changes upon moving at the rates of 18 percent more use versus 26 percent less (Podobnik, 2002). 17-91

OCR for page 87
Results Differences. Clearly there is a large gap between the mode shift results for Portland LRT-based TODs and the surveyed California rail-based TODs. What is particularly striking is the fact that the estimated percentage-point mode shifts to transit for commuting at HRT-based TODs along BART in the San Francisco region were barely over one-quarter the estimated shifts at LRT-based TODs in Portland. The various possible explanations include: A highly transit-supportive environment in Portland, including constrained parking in the healthy downtown, combined with a modest transit-use starting point compared to larger central-place cities such as San Francisco. Pre-existing commute mode shares for transit in the BART service area, basically the San Francisco and Oakland/Berkeley commutershed, that are sufficiently high to leave reduced leeway for additional shifts to transit use. Insufficient transit competitiveness in other parts of California vis--vis auto commuting to support either major attraction of potential transit users into TODs or commanding shifts to transit use. Survey and analysis differences in response, methodology, and definitions between the Portland and California studies. In areas where transit use is already high, the more noticeable shifts will not be so much from auto to transit for the primary commute as sub-mode shifts from conventional bus to fixed-guideway transit and mode-of-access shifts from auto to walking. Data bearing on bus versus rail sub-mode shifts are included in Tables 17-39, 17-40, and 17-54. No data on mode-of-access shifts per se were encountered, but large shifts to the walk mode of access from motorized modes may be readily inferred from the high walk mode of access shares reported for TODs compared to the lesser shares observed systemwide and particularly at non-TOD distances from stations. Mode Shifts with Workplace Change to TOD Mode changes upon moving a work location to a TOD were analyzed by the 2003 California TOD travel characteristics study on the basis of 102 surveyed office workers who were found to have changed their work location within the past 3 years. The shifts (or non-shifts) were reported as follows: Shifts favorable to vehicle trip reduction included 7.9 percent from auto (drive alone or multiple occupancy) to rail transit, 2.9 percent from auto to bus transit, 3.0 percent from auto to walking or biking, and 5.9 percent from driving alone to carpooling. Shifts not favorable to vehicle trip reduction included 3.9 percent from rail transit to auto, 4.9 percent from bus transit to auto, 5.9 percent from walking or biking to auto, and 3.9 percent from carpooling to driving alone. Other shifts and non-shifts included 7.8 percent continuing to take bus or rail transit, 47.1 per- cent continuing to drive alone, and 6.8 percent making unspecified shifts. 17-92