Ben She and Yihong Hu | Dude, where’s my ride?

Picture this: it’s 8:30 p.m. You’ve just had three classes in a row, without time for dinner. You run to the PennBus stop at the Penn Bookstore, unlock your phone, and open up the PennTransit Mobile app. Opening the timetable on map, you find that the Penn Bus West just passed the Bookstore, and another one isn’t coming for 18 minutes. So a PennRide it is. You’re about to reserve it, but you notice the next vehicle won’t come for at least 45 minutes! Sighing in resignation, you begin the walk home in pitch-black darkness. What went wrong here?

The Penn Transit services have not changed substantially since the early 1980s. These include the Penn Bus, a circulator bus serving West Philadelphia and Center City; PennRides on Request (“PennRides”), which operates as a door-to-door ride-hailing (aka microtransit) service, somewhat like Uber Pool; and lastly the Walking Escort service.

The services were originally created as a response to concerns about public safety on and off campus, and had limited operating hours and frequencies. Today, though, students are living farther and farther off campus, with ever-higher demand for a frequent and reliable service that extends far beyond typical peak hours for public transit.

When the PennRides app was launched at the start of 2019, it immediately skyrocketed in popularity. Students used to have to remember and dial a phone number and wait from anywhere from one to 20 minutes, without any tracking and notification when one’s ride had arrived. Now, they could download an app and, in theory, know their shuttle’s position far more reliably. In fiscal year 2022, Penn Transit boasted record ridership numbers of 340,000 passengers, or close to 1,200 weekday riders.

However, with that success comes a familiar problem if you’ve ever studied economics: the tragedy of the commons. At peak hours, which are typically when late classes end from 7 to 9 p.m., the app is inundated with requests that extend waiting times to 30 minutes or more. Journey times lengthen, routings get noticeably more circuitous, and service quality and efficiency overall decrease dramatically. Penn’s strategy thus far has been to double down on on-demand ride-hailing, with an expanded service contract to 40 drivers along with new vehicles. 

But brute-forcing the problem here seems futile considering continued demand, with operating and capital costs scaling very poorly as well. In addition, this has come at a time when Penn Bus frequency has degraded from every 15 minutes per route last year to 20 minutes, signaling that Penn Transit has already had to save service costs elsewhere to compensate and feed the on-demand beast.

In a way, this phenomenon reflects the trend seen at many transit providers across the country — for universities and entire cities — as they find themselves at the crossroads between traditional fixed-route and insurgent on-demand service. With alarming bus ridership declines even before the COVID-19 pandemic, and the new normal of commute patterns post-pandemic, several transit agencies have switched out traditional routes in low-performing areas with on-demand microtransit. But critics have warned that University microtransit, like its private counterparts such as Uber or Lyft, run the risk of cutting into the busiest and most congested transit areas without sufficient restrictions and controls on its usage. What may seem complementary at first becomes competitive, and with far higher operating costs than a traditional 40 to 60 seat bus, ends up costing more in the long run.

Penn Transit, for its part, is very aware of this trend, but thus far seems stymied in its quest to get more students using Penn Bus. In every on-demand shuttle, there is a small 8.5-by-11 inch flier posted encouraging students to take the Penn Bus next time, promoting its “convenient” 20-minute service frequency. But if you’ve never taken the Penn Bus before, it’s pretty daunting where to even start.

Finding the nearest available on-campus stop is a struggle, especially at night. Signs are small, dark, and often torn off altogether, as is the case by the School of Dental Medicine on 40th Street. No information at the stop about which routes or what frequency they might be expected to come on the signs — it’s on you to download the app.

As you leave campus, you must now peer outside and discern street signs flashing by. When you think you’ve reached your stop, you must pull the cord or shout at the driver exactly before the bus passes it, as outside campus stops are only made by request. With such ambiguity, it’s no wonder most students are unaware PennRides can even be used for inbound trips to campus rather than just going home, leading to a bias for on-demand rides. 

With this many potential points of failure from the user’s POV, the message is clear enough: if Penn Transit wants more students taking Penn Bus, they need to promote their campus-wide bus service and improve user legibility and accessibility in all aspects. 

One welcome change: switching to a new PennTransit Mobile app, replacing software and management providers from DoubleMap to TripShot. Now, there are better trip planning features to get directly from origin to destination, along with stop pop-outs where app managers should include helpful user directions. Along with the app improvements, new printed brochures should indicate how to use Penn Buses step by step, with picture references for every campus stop. Additionally, off-campus stops should be clearly marked, both in-app and physically, around every two blocks to remove any ambiguity about where to get on or off. Penn Transit could even share the same bus stops and shelters with SEPTA to reduce confusion on stop locations. 

But perhaps the most necessary policy change revolves around service provision: the Penn Bus needs to get back to 15-minute, and ideally 10-minute frequency as more students get familiar with it. As transit planner Jarrett Walker said, "frequency is freedom": a service that comes at least every 15 minutes, without having to read a timetable, is the most attractive to use. But in order to sustain this service increase, we argue that the on-demand service should not be allowed for destinations within a five-minute walk of the Penn Bus — in other words, creating a geofence. Though admittedly controversial, this is a common provision transit agencies leverage to ensure that microtransit doesn’t cannibalize their existing bus service, and it makes sense here as well to reduce PennRide demand for short trips, improving wait time for students who live farther out.

As great as these changes would be, improving Penn Transit is only one piece of the mobility puzzle. Philadelphia already boasts many abundant mobility options, and University administration would do well to further partnerships with Indego bikeshare and SEPTA, bundling fares and passes with employee benefits and student tuition fees, as Penn Medicine has already done. In this way, all employees, faculty, and students, no matter where in the region they come from, are able to get to campus cheaply. But we think Penn Transit has the opportunity to start right in its own backyard with straightforward fixes to its own services. 

BEN SHE is a graduate student in the School of Design studying urban spatial analytics. His email is bensh3@design.upenn.edu.

YIHONG HU is a graduate student in the School of Design studying city and regional planning. Her email is yihonghu@design.upenn.edu.