Of the many pains a city bus can inflict—from the stop-and-go ride to the eye-to-armpit interaction—the harshest is the wait. Unsympathetic as they often seem, transit agencies don’t enjoy this situation any more than riders do. A long enough wait might drive some people to drive, which not only creates more traffic for buses to wade through but reduces revenue that might go toward better service.
Unless an agency has enough money to invest in more frequent service, their options for reducing those frustrating waits are limited. Fortunately, one of the newest tools at their disposal is proving to be an extremely effective one. Real-time transit data—and the mobile applications that use it to tell riders where a bus is at any given moment—is changing the waiting game in ways that experts are only starting to measure in hard numbers.
“While this strategy does not actually improve the reliability of service, it helps riders to adjust their behavior in the event that a vehicle is running behind schedule,” said transit scholar Candace Brakewood of the City College of New York during a recent webcast for the Center for Urban Transportation Research. “It helps improve the perception of reliability.”
Working with Kari Watkins of Georgia Tech, who helped develop the OneBusAway app, Brakewood recently studied the effects real-time information had on rider behavior in Tampa (where the app was deployed in summer 2013) and Atlanta (where the full launch came in February 2014). In both places, riders using real-time services experienced shorter wait times and greater trip satisfaction; there were also signs the apps might increase ridership (fitting with previous evidence from Chicago).
Let’s take a closer look at the Tampa study. The researchers recruited about 200 regular HART bus riders and asked them questions about their transit habits. Then roughly half of the participants received access to an early version of OneBusAway, while the rest continued riding without real-time information. At the study’s conclusion, the riders once again answered questions about their transit habits so the researchers could see what had changed.
The “perception of reliability” definitely shifted. Compared with the start of the study, riders without the app felt less satisfied with both their wait times and with bus arrival times by the end of it. But real-time users felt significantly more satisfied on both counts. They felt less frustrated, too. The share of app users who reported being “frequently” or “always” frustrated with the wait fell from 25 percent to 18 percent during the study; for non-app users, frustration levels slightly increased.
More significant was that the actual wait times improved, too. HART riders in the OneBusAway group went from waiting an average of more than 11 minutes at the start of the study to an average of 9.5 minutes once they used the app. Wait times for the other group of riders remained fairly steady—no surprise, since they had no updates to inform their actions. All told, the OneBusAway riders waited on average about a minute and a half less than non-app riders by the study’s end.
Now for Atlanta. Using slightly different methods—a before-and-after analysis that also tracked individual riders via smart-card data—Brakewood found similar trends for riders on MARTA trains and buses. In this case, 71 percent of the riders using real-time apps perceived a decrease in wait times for MARTA trains. (The actual wait times weren’t available in this case.) About 60 percent of real-time riders felt more satisfied with the train service.
There was a hint of evidence that MARTA ridership rose as a result of real-time apps, too. People who used real-time apps took nearly 12 more trips a month after the release of the real-time services than they did before. Those who didn’t use real-time services took only 5 more trips, on average.
Now, the statistical significance of transit use disappeared when Brakewood added some controls to the data—removing any MARTA smart cards that might have been shared with other riders, among others. While that might mean real-time data had no impact on ridership, it’s also possible the controls made the sample too small to detect a change. The study also focused on existing riders; a better test of ridership change would also consider drivers who might be open to switching modes on account of real-time benefits.
None of the findings are terribly surprising, but that’s far from a bad thing. Cash-strapped city agencies can now point to more evidence that releasing real-time data will improve the transit experience even in the absence of massive service investments. As any bus rider knows, sometimes the best outcome is the most predictable one.
*This is Cross-posted with special permission from Citylab, a project of The Atlantic.
Head Photo: Orin Viriyincy, Flickr, Cropped by 100RC