Conclusion

Description of image

First, not all lines and stations are affected equally. The Red and Orange Lines consistently contain more traffic, serving major employment centers and residential neighborhoods. Downtown hubs like South Station, Downtown Crossing, and Park Street dominate ridership, but their usage patterns differ dramatically, some showing intense commuter peaks while college-area stations maintain strong midday activity reflecting class schedules.


With that, the pandemic had long lasting effects on Boston’s transit. We saw ridership plummet from over 6 million monthly riders to under 1 million almost overnight in March 2020. Even after restrictions were lifted, ridership stabilized around 4-4.5 million, meaning we've permanently lost 25-30% of pre-pandemic riders. MBTA ridership reflects lasting changes in work patterns, with remote and hybrid work becoming the new normal.


Also, external factors dramatically shape ridership patterns. Major sporting events like Red Sox games can greatly boost station entries at nearby stops like Kenmore, while extreme weather, like cold and snow, consistently reduce ridership across all transit modes. These represent substantial shifts in how people choose to travel.


Lastly, timing matters. Our heatmap analysis revealed that Boston's transit system has distinct daily rhythms, with clear AM and PM peak patterns at most stations, but more distributed usage at certain locations throughout the day.


Understanding these patterns can inform service planning, resource allocation during events and severe weather, and decisions around capacity improvements. As Boston continues to evolve post-pandemic, tracking these trends will be essential for ensuring the MBTA effectively serves the city's changing needs.


Some next steps for this project would be to expand our analysis to buses and not just trains. Another would be to look at how service and repairs for stations affect ridership. Finally, to get a deeper sense of how class effects the transit system, creating a visualization that looks which neighborhoods in Boston have the best access to public transit and see how ridership correlates with income level.



Home Data Visualizations