Incentives shift BART riders out of the morning rush

“BART Perks,” a six-month test incentive program offered by BART and the San Francisco County Transportation Authority (SFCTA), concluded on February 28 and preliminary results reveal that incentives can successfully shift the travel behavior of BART riders.

The Perks program incentivized riders to shift travel times away from the peak morning hour of 7:30 a.m. to 8:30 a.m. to reduce crowding.  During the six-month trial period, an average of 250 Perks participants shifted their ride either before or after the peak morning rush hour each weekday. That amounts to the equivalent of two full BART cars being freed up each weekday during BART’s busiest hour. About 2,600 Perks participants traveled during the peak hour each day before the program, meaning about 10% of these riders targeted by the program did in fact shift.  

The first program of its kind in North America, Perks was modeled after successful international transit rewards programs, and showed comparable results to a similar program deployed in Singapore.  

"BART's cars may be the oldest in the country, but in many areas BART is at the cutting edge,” said BART Board member Nick Josefowitz.  “This program shows the benefits we can deliver to our riders and to the region by strategically deploying new technology.”

Exceeding expectations for 10,000 sign-ups, almost 18,000 BART riders participated in the Perks program trial that ran from the end of August 2016 through February 2017. Participants were eligible to win cash rewards by signing up for the test program and were incentivized to shift their commute outside of the morning rush hour by receiving additional rewards for trips starting before 7:30 a.m. or after 8:30 a.m. Participants frequently shared the results of their cash rewards on social media, helping to spread the word about the program.  

“Through BART Perks, we see that incentives can be used to shift behavior and create strong customer interest in rewards programs,” said Tilly Chang, Executive Director of the San Francisco County Transportation Authority, which serves as the congestion management agency for the county. “While a complete program analysis is still to come, we are encouraged by the program’s results to date, including the ability to manage peak demands while garnering overall rider satisfaction and employer support.”

Perks’ incentive model is based on the economic theory of nudging, wherein even a small reward can lead to adjustments in behavior. Each month Perks awarded an average of $35,000 to all program participants, and roughly $210,000 was awarded over the life of the program. On average, participants earned close to $3/month, with about 10 different participants per month being paid $100 or more based on their participation in the “Spin to Win” game. Each month, rewards were transferred to participants’ PayPal accounts.

The program complemented BART’s long-term efforts to reduce crowding and congestion on trains and in stations, which includes redesigning cars to increase capacity and running additional ten car trains with the arrival of the Fleet of the Future and a new train control system. These upgrades are particularly important for the approximately 28,000 riders who travel to San Francisco through the Transbay Tube on BART during the peak hour each weekday morning. During the peak hour, BART carries almost double the number of people carried over the Bay Bridge in private vehicles (there are 14,200 estimated private vehicle riders).

BART and the SFCTA measured the effects of the Perks program by comparing the percentage of rush hour trips participants made on an average weekday morning before the program to their behavior during the program. Based on BART’s overall ridership data, staff performed this same calculation for the commute pattern of non-participants in order to understand whether there were other factors that could have caused the shift in rider behavior. While the program was of too small a scale to result in noticeable crowding reduction on BART, initial results reveal the program did successfully reduce peak hour travel among participants.

Perks included an employer partnership program in order to increase employer support for greater adoption of flexible work schedules. Fifteen employers signed up as BART Perks Partners and were committed to promoting the program among their employees, including UCSF Benioff Children’s Hospital, Alliant International University, CRI, and Integral Group.  

How BART Perks Worked

Beginning in August 2016, participants could sign up for Perks at BARTperks.com using their email address and Clipper Card number. Perks offered riders points for all travel on BART, and up to six times as many points by starting their trip either 6:30 a.m. to 7:30 a.m. or 8:30 a.m. to 9:30 a.m. Points could be exchanged for small cash rewards or used to play the “Spin to Win” game for the chance to pick up additional points or random cash rewards from $1 to $100.

The program was funded primarily through a Federal Highway Administration grant program that supports innovative approaches to reducing congestion through pricing and incentives.  Funding was also provided by BART and San Francisco’s half-cent sales tax for transportation.  

Lessons Learned

To achieve even greater levels of rider shifting, future travel incentives programs for BART would need to be designed to better target individuals who are frequent riders during the busiest periods on the transit system.

In addition to testing whether crowding can be reduced through rider incentives, Perks also aimed to increase customer satisfaction among BART riders. In December 2016, BART and SFCTA assessed the satisfaction of BART riders who opted into Perks, and results showed that 67 percent of participants reported being satisfied with the program.

What’s Next

SFCTA and BART are in the process of a full evaluation of the program’s results which is expected to be completed by fall 2017. Based on the findings, the agencies will consider how to proceed with further strategies to reduce rush hour crowding.