One City’s Ambitious Plan To Ease Overcrowded Trains? Pay Riders

With help from an urban tech startup, San Francisco is using psychology and gamification to aid its overburdened transit system.

One City’s Ambitious Plan To Ease Overcrowded Trains? Pay Riders

BART, the regional transportation system in the San Francisco Bay Area, has a new strategy to help cope with commute-hour congestion that’s packing train platforms and cars to the gills: BART Perks, a rewards program that uses cold hard cash to make off-peak travel more enticing. BART’s leadership is betting that this unconventional short-term solution will be a big enough patch until longer term plans come to fruition.


As cities grapple with ways to better manage their transit systems, BART is using a psychology-fueled strategy to solve the age-old problem of getting people to work efficiently and on time. If the grand experiment proves successful, it sets a benchmark for other municipalities struggling to alleviate commute-hour congestion on trains. Will it work?

[Photo: Flickr user Alex R]

A Booming City With An Aging Transit System

When the system first opened in 1972, it saw only 170,000 passengers weekly. Today, about 430,000 riders take BART every day of the week. Thanks to San Francisco’s booming economy, the area has experienced surging growth that’s placed extra pressure on public transportation. Since 2010, BART ridership has increased 30% system wide and 43% on the lines that connect the East Bay to San Francisco, specifically. The system sees more than 2 million trips every week.

Solving the congestion problem is more complex than putting more trains into service; it requires redesigning cars to accommodate more passengers and overhauling the train-control systems. (Right now, it’s impossible to send more trains through the Transbay Tube–the underwater train tunnel that connects Oakland to San Francisco–because the computerized controls require a certain distance between trains.) BART is working on those strategies right now, but the full effect won’t start to be felt for years. Overcrowding is a real problem now.

Just as building more roads won’t improve traffic–it actually makes congestion worse–the situation is similar with other infrastructure. “Demand management is essential,” says Balaji Prabhakar, a Stanford professor and founder of Urban Engines, the consultancy BART hired to work on its incentive program (which was recently acquired by Google). “The ability to build capacity for physical networks–transportation, energy, communications, water–is limited. When you attempt expansion projects, the cost is incredibly high. There’s an arsenal [of strategies to help commuting congestion]. The addition of supply needs to be done, no question. But better ways of rethinking mobility is critical.”

Some population growth experts estimate that by 2050, 70% of people will live in urban areas. As cities become more crowded, the pressure on public transit will increase. The slow wheels of public policy mean that adequate funding to improve infrastructure to accommodate growth doesn’t come as easy or is often many years away.


San Francisco is not alone in the struggle. As cities try to do more with less capital, unconventional strategies like this one can reduce some of the burden.

How It Works

Ryan Greene-Roesel, a senior transportation planner at the San Francisco County Transportation Authority, worked closely on BART Perks. She and her team estimated that if 1,200 people shifted their commute, it would be the equivalent of putting another 10-car train into service, which would make more room for riders who can’t shift their schedule. In a survey of riders who commute into San Francisco, Greene-Roesel found that most people worked in the tech sector and 20% of respondents had no barriers to changing their schedule.

To use BART Perks, riders register their transit card, called Clipper, which tracks what time they enter a train station. They receive points for every trip they take, but if they enter a station during “Bonus Hours”–which are between 6:30 and 7:30 a.m or 8:30 and 9:30 a.m.–they earn more points. The points are like frequent flyer miles, and as a rider earns more points, their status–bronze, silver, gold platinum–rises. (The exact rules on how riders level up are on BART’s website.) As the points accumulate, riders can cash out (1,000 points equals one dollar) or play a Spin to Win game for the chance to get $100. The program, which runs until February 2017, is budgeted to dole out up to $50,000 per month and is capped at 25,000 participants. About 10,000 people signed up for the program within two days of its August 30th launch.

In essence, BART is paying people not to ride at the most impacted time, between 7:30 a.m. and 8:30 a.m. “We wanted to make an incentive and make it fun and addictive, like how you track your steps [on fitness apps],” Alicia Trost, a BART spokeswoman, says. “We’re trying to hone in on that behavior.”

So just as Fitbit documents steps to help its users stay active, as airlines lure repeat customers through frequent flier miles, and your local cafe offers a punch card for free coffee, BART is applying gamification to public transportation.


The Psychology Of A Rewards System

The only other transit system in the world using an incentive program to pay commuters is Singapore, where Urban Engines implemented a similar successful program that launched in 2015. Instead, most transit systems use congestion pricing, a fairly common practice of charging commuters more to travel at a certain time. For example, drivers pay more to cross the San Francisco Bay Bridge during peak traffic hours.

So why did BART create a rewards system that costs it money, when something like congestion pricing could actually generate revenue?

For one thing, BART learned through its research that it would have to raise its fares several dollars to achieve the same reduction in ridership during peak hours, which would not be in the best interest of its customers and would likely affect low-income riders disproportionately.

Then there’s the subtle behavioral design that informed the program. A 2011 article in Monitor, the American Psychology Association’s journal, cited a number of scenarios, such as criminal justice, substance abuse, and HIV care, where positively reinforcing good behavior–also known as contingency management–was more effective than punishing bad behavior. With the incentive system, ticket prices stay the same for everyone regardless of when they travel, but people who can change their behavior get a little cash prize.

The format of the reward–a lottery–is based on prospect theory, a behavioral economic theory that explains how people make decisions. For small everyday choices–like buying a cup of coffee or not as opposed to buying a car–decision-makers look at potential gain over potential loss. “Lotteries administered as a game of chance engage people more,” Prabhakar says. Urban Engine’s original pilot study took place in Bangalore from October 2008 to April 2009, and in the course of the incentive program, off-peak travel doubled and the average commute time dropped from 71 to 54 minutes.


“When you’re rewarding good behavior, you’re taking advantage of some people having more flexibility,” Greene-Roesel says. “Some might not have flexibility and penalizing then seems punitive.” Trost puts it another way: “We wanted to give the carrot instead of the stick.”

But Will It Work?

BART Perks’ success will depend on the number of people who actually use the program.

Singapore saw a 7.49% shift from peak to off-peak travel after the implementation of its program, but it remains to be seen if those who have already registered for the game in San Francisco will follow through and change their commute. As with most games, novelty wears off, and since the program is in its nascent stages, it’s too soon to tell if it will be a smash success. In the Singapore program, it took about four weeks of participation to change behavior. If a meaningful shift in ridership has occurred by the program’s conclusion in early 2017, BART may consider making the pilot program permanent (though a new funding source and new contract would be needed for that to happen).

The project could become a precedent for using gamification in public transportation and planning. But if it doesn’t work as expected? Bay Area denizens will continue to endure cramped commutes–until BART puts its new trains into service, or conjures up a new plan.

Visit for more.

About the author

Diana Budds is a New York–based writer covering design and the built environment.