Evaluation of San Francisco Bay Area Rapid Transit’s Carpooling App Revealed that 88 Percent of Scoop Trips Either Reduced or Caused no Change in Vehicle Miles Traveled.

Independent Evaluation Results from San Francisco Bay Area Rapid Transit Integrated Carpool to Transit Access Program Pilot.

Date Posted

09/29/2023

San Francisco, California, United States

San Francisco, California,

United States

Identifier

2023-B01791

Mobility on Demand (MOD) Sandbox Demonstration: Bay Area Rapid Transit Integrated Carpool to Transit Access Program Evaluation Report

Summary Information

The Federal Transit Administration's (FTA) Mobility on Demand (MOD) Sandbox Demonstration Program offers a real-world platform for integrated MOD strategies with local partnerships. This study included the independent evaluation results for the Bay Area Rapid Transit (BART) Integrated Carpool to Transit Access Program pilot project in San Francisco, California. The project aimed to address carpool parking at stations by using a carpool matching platform to match drivers and passengers with similar station destinations into carpools, through a smartphone app. The pilot program initially launched at the Dublin/Pleasanton BART station in January 2017 and expanded to 17 stations. The evaluation of the pilot demonstration ended in April 2019, and the project ended shortly thereafter. Expert interviews with project personnel were also conducted to reveal lessons learned on implementation and policy.

METHODOLOGY

This study considered 13 hypotheses (e.g., carpooling to stations increases) for the purpose of evaluating the BART Carpool to Transit Access Program pilot project facilitated through a smartphone app, in comparison to legacy (traditional) carpooling requiring printed paper permits. The hypotheses were tested to explore the project impacts on carpooling, costs, enforcement, ridership, parking, and vehicle miles of travel (VMT), using activity data and user surveys conducted at three separate times; the first was in August 2017, the second was in February/March 2018, and the final one was in June 2019.

FINDINGS

The analysis found, with survey-derived mode shifts applied to the activity data, a total of 88 percent of carpool trips, that 44 percent of trips reduced VMT, and 44 percent imposed no change in VMT.
The evaluation indicated that the carpool service might have increased persons-per-parked-vehicle by up to five percent at Dublin/Pleasanton station and around one percent at other stations like Antioch, Warm Springs, and Orinda.
BART postulated that the additional marginal app-based carpool-specific enforcement activity added about one hour per week of labor, which amounted to a small increase of about 0.3 percent of additional labor costs resulting from carpool-specific enforcement.
Across all the app-based carpool trips to BART stations, system activity was estimated to have increased the number of persons per parked vehicle by 0.81 percent at the end of the evaluation period.
The evaluation revealed that without the app-based carpool services, many users would have opted for single-occupancy travel, like driving alone or using transportation network company (TNC) services. The survey showed 41 percent would've driven solo to BART, 14 percent would've used TNC services, and nine percent would've been dropped off. Altogether, 64 percent of respondents chose the app-based carpool service over solo automobile travel, indicating the pilot's success in reducing VMT.
Forty percent of app-based carpool users cited cost savings as a reason for using the service. Drivers, making up 30 percent, saw cost recovery as a motivation. Other reasons included station parking access (58 percent), flexible arrival times (46 percent), an alternative to driving (43 percent), faster travel (40 percent), and enhanced safety (18 percent).
BART observed fraudulent use in traditional carpool lots reaching up to 60 percent. In contrast, app-based carpool spaces saw a maximum of 30 percent fraud, possibly as low as 15 percent. This represents a 50 to 75 percentage improvement when comparing traditional to the app-based carpooling.