For years, Baltimore students and parents have complained that trips to school on public transit are long and unpredictable. They describe long waits on dark street corners and full buses driving past them. Every day, they say, is a stressful and frustrating odyssey.

The Banner set out to quantify how bad the problem really is. To do that, we mapped every city middle and high school student’s commute and tracked the location of every Maryland Transit Administration bus for months. We determined how long it would take a kid to get to school under the best circumstances, how long it was actually taking and how often things would go awry.

This first-of-its-kind analysis shows it’s nearly impossible for city students to get to school on time every day on public transit.

Here’s how we did it.

Mapping 4,000 trips to school

City middle and high school students don’t automatically go to the school closest to their home, thanks to a school choice program that started 20 years ago. That means kids who live next door to one another could have vastly different commutes.

Through a public records request, we obtained an anonymized database that shows the U.S. Census tract where students lived and the school they attended. There were over 4,000 such combinations. That means more than 4,000 unique trips to school.

To map each journey, we used a routing model developed by researchers at Massachusetts Institute of Technology and Conveyal, a transit consulting firm whose clients include some of the nation’s largest transit agencies. It uses MTA’s local bus, subway and light rail schedules to estimate the quickest route from the tract where each student lives to their school.

The quickest route doesn’t take into account its distance or the number of transfers a student would have to make from one bus or train line to another. In reality, some students may opt for a slower route to simplify their trip. The model we used also could not take into account schedule variations imposed by things like canceled trips or late trains or buses. In other words, the model represents a best-case scenario.

Because we only know where students live by their Census tracts, which are geographical boundaries about the size of a city neighborhood, some trips are difficult to route. Some routes start and stop in the same tract; some tracts are bigger than others. We chose a starting point for each based on where most people live in the Census tract according to EU Global Human Settlement Layer data.

Our analysis found that when the transit system runs on time, the typical city student commutes about 40 minutes to school, more than twice the length of the average school bus commute in Baltimore County and longer than it takes the average Baltimore adult to get to work. Some trips are longer than 90 minutes each way.

We also determined that nearly 60% of students have to transfer at least once on the way to school, which increases the chance they’ll miss a connection and have an even longer commute.

We found that a missed bus adds, on average, 20 minutes to a trip. On some lines, that wait can extend past half an hour.

Tracking unreliable buses

We knew that our model for the best-case scenario only told part of the story. Dozens of students complained of buses that showed up at unpredictable times.

So we tracked the location of every MTA bus in the city, every five seconds, from 4 a.m. to midnight from the first day of school until mid-January. We used the same real-time data that powers transit apps that tell you when the next bus will arrive, ultimately compiling tens of millions of data points.

Comparing scheduled bus arrival times to reality gave us a systemwide look at how often buses don’t arrive on time. We used the MTA’s definition of “on time,” a nine-minute window from two minutes before a bus’s scheduled arrival to seven minutes after. By that measure, about 1 in 4 buses won’t arrive on time during students’ morning commuting hours (5 a.m. to 9 a.m.). One in 3 are not on time when they head home in the afternoon (2:20 p.m. to 4 p.m.).

There’s no industrywide standard for what counts as an on-time bus, and transit researchers say MTA’s window is long on purpose.

“[MTA] have a goal in mind for a percentage that they want to hit, and they calibrate the metrics to make that goal attainable,” said James Pizzuro, who runs Aries for Transit, a website that tracks public transit’s performance in Maryland, Virginia and D.C.

MTA officials have said they use the same standard as Washington, D.C.’s transit system, but it’s more common for urban transit agencies to define “on time” as up to one minute early and five minutes late. By that measure, only 62% of Baltimore buses are on time during students’ morning commutes; it’s 52% in the afternoon.

The difference between a bus arriving exactly on time and seven minutes late may mean more to a student than a typical commuter. Class starts with or without them. City middle and high school students have lower grades in first-period classes than the rest of the school day, we learned through data obtained via public records request.

To learn more about the academic harm students suffer, read our investigation.

For all of the details of our data analysis, visit our GitHub.

This story has been updated to correct the name of the Maryland Transit Administration.

About the Education Hub

This reporting is part of The Banner’s Education Hub, community-funded journalism that provides parents with resources they need to make decisions about how their children learn. Read more.