Shared Mobility and the Transformation of Public Transit (2016)

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.

B-1 A P P E N D I X B Survey Methodology and Additional Data

B-2 Shared Mobility and the Transformation of Public Transit The user survey was distributed through both private shared-mobility operators and transit agencies in September and October 2015. The survey sample frame included adult residents of the study regions who have used one or more shared-use modes, including transit. The researchers requested distribuon by transit agencies and shared mobility operators in all of the seven study markets, and also in New York City. The recruitment method was through invitaons emailed and distributed via social media by cooperang agencies and operators, inving customers to complete a web-based survey instrument. A link was directly emailed by distribuon partners to more than 75,000 email recipients in addion to a large number of newsle‚er and social media followers, and received 4,551 at least paral responses. Provider-specific links, called collectors, allowed tracking of response sources and permi‚ed deacvaon of parcular channels at the end of a two-week open period. The overall count represents a net response rate of 6.0% for the sources the researchers were able to track. Distribuon partners in each market are listed in the table below. Table B-1. Survey distribuon partners, dates, and response counts. Market Agency or operator Field dates Total responses Ausn Car2go 9/17/15–10/1/15 539 Boston Movate/Hubway 10/8/15–10/22/15 69 Chicago Movate/Divvy 9/24/15–10/8/15 424 Los Angeles LA Metro 10/6/15–10/20/15 653 New York City Movate/CiBike 9/23/15–10/7/16 508 San Francisco Bay Area BART 9/18/15–10/8/15 (staggered samples) 179 San Francisco Bay Area Movate/Bay Area Bikeshare 9/16/15–9/30/15 5 Sea‚le Car2go 9/17/15–10/1/15 992 Sea‚le Movate/Pronto Cycle Share 9/15/15–9/29/15 30 Washington, DC WMATA 9/16/15–9/30/15 830 Washington, DC Movate/Capital Bikeshare 9/17/15–10/1/15 74 Washington, DC Car2Go 9/17/15–10/1/15 248 Total 4551 The survey contained an inial screening queson about overall experience with new shared-use modes, asked whether respondents had “ever used a shared form of transportaon like bike-sharing, car-sharing, or a ride-sharing service like Uber or LyŸ.” Respondents who answered “No” went immediately to a poron of the survey that only asked about transit technology, followed by collecon of demographic informaon, including home zip code (mapped in Figure B-1 for all respondents). The geographic distribuon of home zips generally matches the distribuon locaons.

Survey Methodology and Additional Data B-3 Figure B-1. Reported home zip codes of survey respondents. Sampling consideraons Because the researchers were limited to working with convenience samples in each market—those individuals able to be reached via the partners who agreed to distribute the survey, all of whom were people who had previously supplied their email addresses to the agencies or operators—we must be cauous about inferring to the wider populaon of shared mobility and transit users and certainly to the general populaon. The survey was administered via an online form, and links to this form were distributed by email. This implies a basic level of technological facility, and also a willingness to parcipate in research about transportaon. Also, the survey took place in several of the largest, densest, and most expensive cies in the country, which were chosen for this study specifically because of their known high levels of shared mobility usage. Thus the sample is likely over-representave of higher-income, more highly educated individuals compared to the general U.S. public. We should also make note of the small sample sizes in some markets relave to others: we received only 69 responses via the Boston channel, and fewer than 200 in San Francisco. In addion, we might expect some bias related to the mode of the distribuon channels for the various surveys. In Los Angeles and San Francisco, the survey was distributed almost exclusively via the transit agencies; in Boston, Chicago, and New York, the survey was distributed solely via bikeshare operators; and in Ausn and Sea‹le the primary channel was a carsharing operator. One subpopulaon this distribuon method might miss would be people who use ridesourcing exclusively among shared modes, including transit. Unfortunately the researchers don't have a way of esmang the size of this populaon because so li‹le systemac knowledge currently exists about levels of ridesourcing usage in urban areas and among the traveling populaon overall. Ongoing research—from other behavioral surveys, public and private operator data, personal travel inventories, and other data sources—is needed to connue building the understanding of the use and effects of ridesourcing and other shared modes. Overall, the familiarity with and level of informaon about shared-use modes in the general populaon is likely to differ somewhat from what we found from our respondents. However, since the subject of this report is the

B-4 Shared Mobility and the Transformation of Public Transit interacon of shared-use modes and new mobility technologies with transit, the researchers believe it made sense to focus this inial descripve effort as we did. Response rate The survey was distributed through both private shared-use operators and transit agencies starng in mid-September 2015. The survey link was directly emailed to more than 75,000 email recipients plus a large number of newsleer and social media followers, and received 4551 at least paral responses, of which 3548 reached the end of the survey. The number of responses to individual quesons varied, with some respondents skipping individual quesons while answering others later in the survey. The overall count represents a net response rate of 6.0% for the sources we were able to track. Response rates from the collectors of individual agencies and operators ranged from less than 2% to more than 15%. Response rate was calculated as the proporon of responses to successful (non-bounce) email deliveries. Since we don’t have access to the web social media metrics of the distribung organizaons, we did not aempt to calculate response rate for those mediums. The circulated version of the survey instrument is aached in Appendix C. Addional data Table B-2, Figure B-2, and Table B-3 in this secon present addional data that informed the main findings. Table B-2. Frequency of use of all modes, by top mode. Frequency of use of all modes, by top shared-use mode Frequency by mode Top Mode Bikesharing Carsharing Ridesourcing Public bus Public train Public bus Daily/almost daily 3% 3% 3% 63% 14% 1-3 times a week 18% 14% 10% 23% 21% 1-3 times a month 39% 25% 23% 11% 28% A few times a year 31% 34% 35% 3% 23% <1/yr 9% 24% 30% 0% 14% Public train Daily/almost daily 11% 2% 4% 11% 77% 1-3 times a week 33% 3% 6% 13% 18% 1-3 times a month 39% 13% 18% 21% 7% A few times a year 16% 41% 33% 34% 1% <1/yr 2% 38% 38% 16% 0% Bikesharing Daily/almost daily 70% 1% 2% 2% 12% 1-3 times a week 24% 2% 6% 6% 20% 1-3 times a month 8% 5% 8% 6% 15% A few times a year 1% 10% 13% 9% 8% <1/yr 0% 78% 69% 68% 43% Cell shading reflects the relave magnitude of percentages, from lowest (red) to highest (green). Data reflect cross-tabulated responses to survey quesons 7 and 4 (see Appendix C). (Connued on next page)

Survey Methodology and Additional Data B-5 Table B-2. (Connued). Frequency of use of all modes, by top shared-use mode Frequency by mode Top Mode Bikesharing Carsharing Ridesourcing Public bus Public train Carsharing Daily/almost daily 1% 8% 0% 1% 0% 1-3 times a week 5% 31% 12% 16% 6% 1-3 times a month 14% 40% 33% 39% 15% A few times a year 23% 20% 29% 20% 21% <1/yr 53% 2% 23% 18% 55% Ridesourcing Daily/almost daily 1% 1% 7% 0% 1% 1-3 times a week 10% 14% 40% 12% 13% 1-3 times a month 27% 29% 42% 31% 29% A few times a year 21% 22% 8% 20% 23% <1/yr 37% 30% 2% 30% 32% Driving alone Daily/almost daily 9% 41% 49% 11% 13% 1-3 times a week 16% 24% 23% 23% 24% 1-3 times a month 21% 11% 9% 17% 17% A few times a year 19% 7% 6% 12% 20% <1/yr 33% 16% 13% 30% 26% Driving with family/friend Daily/almost daily 5% 19% 25% 7% 6% 1-3 times a week 20% 41% 40% 37% 28% 1-3 times a month 32% 19% 23% 28% 28% A few times a year 30% 11% 9% 15% 25% <1/yr 11% 9% 3% 9% 12% Cell shading reflects the relave magnitude of percentages, from lowest (red) to highest (green). Data reflect cross-tabulated responses to survey quesons 7 and 4 (see Appendix C). In addion to asking for home zip code, the survey asked respondents in which metro area they generally used their top shared-use mode—this disncon was intended to elicit informaon about where the shared-mode use actually took place, even if these services were not available near respondents’ homes (for example, people who use the public train and bikesharing when they’re in Washington, DC, even though their hometown only has bus service). As would be expected, all but a few respondents told us that they most commonly use shared-use modes in one of the eight metro areas where the survey was fielded (the seven study cies plus New York City). We received fewer than 100 responses to this queson for either Boston or San Francisco, reinforcing the cauon we must take with what we infer from our results about those areas due to their small sample size.

B-6 Shared Mobility and the Transformation of Public Transit Figure B-2. Metro area where respondents use top shared mode. Trains are prevalent as the top mode in every city with established heavy rail (Boston, Chicago, DC, NYC, and San Francisco), and Los Angeles has a significant train share even at a relavely early point in its rail system’s build-out. Ausn is notable as the only metro where the top mode is not a bus or train. It has the largest shares by far of respondents who choose carsharing or ridesourcing as top modes, reflecng a transportaon infrastructure based on solo driving that is only starng to be retrofi„ed with more fixed-guideway transit and emerging shared modes. Table B-3. Top shared mode by respondent metro area. Cell shading reflects the relave magnitude of percentages, from lowest (orange) to highest (blue). 418 61 424 684 231 413 98 834 27 0 100 200 300 400 500 600 700 800 900 Ausn Boston Chicago DC LA NYC San Francisco Sea„le Other Metro where respondents use top shared mode Austin Boston Chicago DC LA NYC SF Seattle Bikesharing 2% 38% 31% 5% 0% 39% 0% 2% Carsharing 36% 2% 0% 4% 3% 1% 0% 23% Ridesourcing 30% 3% 7% 5% 11% 1% 9% 10% Public bus 30% 7% 13% 14% 59% 3% 17% 61% Public train 2% 51% 49% 72% 28% 57% 73% 5% Mode Metro In what metro area do you most o™en make trips by your top shared mode? (mode as percentage of respondents choosing each metro)