Behind the scenes with MBTA data.

An analysis of bus crowding by street as opposed to bus route.

The Office of Performance Management and Innovation and MIT researchers have developed a model to measure crowding on MBTA buses in time and space. This is the second post in a series on the tools we are building to analyze where and when on the street network crowding is taking place, different temporal patterns of crowding by route, and the causes of crowding. 

After determining how often MBTA bus passengers are crowded, one of the first questions we asked was “where on the street network does crowding occur?” While transit agencies traditionally analyze bus capacity by route, for this analysis we wanted to add up all crowding per street, regardless of how many bus routes use that street. The result of this systemwide overview shows those places where the most concentrated crowding occurs on buses – places where changes to bus service or road configuration could have a major impact on passenger comfort. 

The map below displays where crowding occurred across almost all of the MBTA bus service area, regardless of bus routes. The colors represent the number of crowded, or uncomfortable, inbound passengers. (Due to a limitation in our dataset, the lines connect one bus stop to the next.) 

Average weekday of crowding for inbound passengers

Inbound bus passengers experiencing crowding, average weekday, Fall 2015.

The street-level view

We chose to draw this map using a simple count of passengers who experienced crowding. This simplification makes drawing the map easier, and matches existing roadway counts of vehicle traffic, pedestrians, and bicyclists. However, it can’t be directly compared to the crowding metrics in our previous post which also consider the amount of time each passenger was crowded. 

The colors on the map start at a faint pink to designate very few crowded passengers traveling inbound per weekday. They climb through darker reds to black, representing the observed maximum of 1,291 crowded passengers. For reference, 200 crowded passengers represents about 3 to 4 crowded rush-hour buses. 

How we drew the map

This visualization simplifies the actual Boston-area street network by drawing straight lines between adjacent bus stops on a route. Most of the time these lines closely match the actual streets that buses travel on, because most bus stops are close together along major roads that curve gradually. However, when you see longer straight lines, an actual bus might have followed a winding path. For example, the below image shows the actual path traveled by route 505 from its last stop in West Newton to its first stop in downtown Boston in blue. The straight line between these stops hides the actual path taken. Most other long straight lines on the map correspond to places where buses use highways.

Example showing a long segment

The path of route 505 express from Washington St at Prospect St in West Newton (“A”), to Lincoln St at Beach St in Downtown Boston (“B”).

Why does crowding grow then “disappear”?

You might have noticed that few passengers are crowded near the outer fringes of bus routes, and that crowding grows as more inbound passengers board, and then jumps as different bus routes merge to follow the same street. However, the lines representing large numbers of crowded passengers can suddenly disappear far before they get downtown or to other major employment centers. The map below zooms in to show two of these places, which turn out to be Davis and Porter Squares on the Red Line. Instead of disappearing, passengers are transferring from crowded buses to rapid transit. In cases where a bus route continues past a rapid transit stop, those passengers that remain on the bus are generally no longer crowded.

Example showing transfers to rapid transit network

Uncomfortable inbound bus passengers, zoomed in and with transfer points to rapid transit added

How does crowding vary through the day?

All the above maps depicted the total number of crowded inbound passengers across an average weekday. However, bus crowding varies throughout the day, depending on bus ridership, how much bus service is scheduled, roadway traffic speeds, and other factors. To fully appreciate how the number of crowded bus passengers varies both in space and in time, we created a video stepping through every half hour of bus service that the MBTA operates.

Just as with the maps above, the below video shows where crowding occurred across the street network, regardless of bus route. It displays the number of bus passengers experiencing crowding during every half hour over an average weekday. Times from 4:00 a.m. to noon show inbound passengers; times from noon to midnight show outbound passengers.

The crowding video shows some interesting patterns. Crowding starts in the half hour of 4:30–4:59 a.m., and grows quickly to show a first peak between about 5 and 6 a.m. on a small number of streets in Everett, Chelsea, Lynn, and Hyde Park. These crowded passengers are most likely traveling to jobs that start at 6 or 7 a.m. Crowding then grows across the bus system, peaking about 7:45 a.m. Most of these passengers likely begin their workdays around 8:309 a.m., but travel to work via a transfer to rapid transit. Finally, routes from South Boston which deliver their passengers directly downtown hit their peak crowding at 8:30–8:59 am. A similar pattern plays out in reverse for outbound crowding in the afternoons, clouded slightly by school schedules and passengers conducting errands after work or school. Strikingly, though, some crowding remains late into the night, especially approaching Everett and Chelsea.


This crowded passenger visualization helps identify problem areas where streets carry a lot of crowded bus passengers. It is also useful for understanding systemwide crowding patterns, especially patterns that vary by time of day. However, because it combines all routes together, it doesn’t pinpoint the route contributing most to crowding, or why in particular this route is crowded at certain times and places. Stay tuned for a post that provides more insights at the level of individual bus routes.