A Brief History of the Fairmount Line in Boston
So the Fairmount Line in Boston was in the news again last week with word that Representative Michael Capuano was asking the DOJ & FTA to look into whether Keolis’ train cancellations on the Fairmount Line were in violation of Title VI. I wanted to provide some history, and data-driven context to a decades-long struggle for better service in the predominantly Black and minority neighbourhoods along the Fairmount Corridor.
US Rep. Michael Capuano asks DOJ and FTA to review whether Fairmount Line cancellations violated civil rights laws: https://t.co/jkiahye7vf— Nicole Dungca (@ndungca) October 24, 2016
As I wrote about in my thesis (but not yet in a blog post), Title VI of the Civil Rights act prevents transit agencies from disproportionately affecting visible minorities. If Keolis were disproportionately cancelling trains on the Fairmount Line (which passes through the predominantly minority neighbourhoods of Roxbury, Dorchester, and Mattapan) versus other Commuter Rail (which I will guess are predominantly White) then Keolis would be in violation of this regulation.
I was particularly struck when reading this headline because, as I detailed in my thesis, and I’ll get to at the end of this post, fast, frequent service with fair fares on the Fairmount Line is the #1 intervention to improve transit travel (and commuting) for Black and African Americans commuters riding the MBTA.
This is not simply angst over late or canceled trains. Historically, residents of the affected neighborhoods – populated largely by people of color – have been an afterthought in conversations revolving around economic development, specifically access to job centers in the city. For decades this disenfranchisement meant that people who could see the skyline soaring above the city’s financial heart did not have the means to get to jobs Downtown and in the Financial District in a timely manner, and found themselves effectively disqualified by virtue of lack of transportation.
As Becky Koepnick from the Boston Foundation writes in the quote above, this is more than just about cancelled trains, Keolis’s actions are yet another setback in a history spanning three decades and more for better transit service along the Fairmount Corridor.
Orange Line Relocation
The Orange Line used to run elevated along Washington Street through Dudley Square until Forest Hills. In the 1980s the elevated line was torn down and put into a trench next to mainline railroad tracks along the SouthWest Corridor which had been cleared for a highway project (I-95) that never happened.
A lot of people will still argue today that they were promised light rail as replacement service on Washington Street and instead got 15 years of bus service and eventually the Silver Line “Bus Rapid Transit”. You can read more on community resentment at the Silver Line in the Sierra Club report “Taxpayers get less for more” (pdf). I will say most bus lines still terminate at Dudley Square, as if it were still a Rapid Transit station, leading many riders to have to transfer there to complete their journey. More on this when I cover my thesis research
There once was a train line that didn’t stop in Black neighbourhoods
This section comes primarily from the article “Bypassing Equity? Transit Investment and Regional Transportation Planning” by Kate Lowe available here (paywall)
In the late 1990s, a grassroots community group (the Greater Four Corners Action Coalition) began asserting that residents in the black communities of Dorchester and Mattapan suffered from diesel fumes but could not access service on the Fairmount Commuter Rail Line.
Part of the environmental mitigation negotiated by the Conservation Law Foundation (CLF) for the “Big Dig” in 1990 was improved Commuter Rail service on the Fairmount Line. Following advocacy for the Fairmount/Indigo Line in 1999-2000, the MBTA pledged half of the money necessary to build stations along the Fairmount Line in 2002. The remaining necessary funds were put into the MassDOT budget in 2005-06 following a lawsuit by the CLF for inadequate completion of Big Dig mitigation measures (this also included an agreement to finally start building the Green Line Extension). Mostly from the Fairmount Corridor Coalition, the subsequent timeline of improvements is:
- 2007-Uphams Corner and Morton Street station renovations are complete
- 2009-Four Corners/Geneva Avenue Station starts construction
- 2010-Talbot Avenue and Newmarket Stations start construction; Blue Hill Avenue/Cummins Highway Station begins design planning process
- 2012-The Boston Redevelopment Authority launches the Fairmount Indigo Planning Initiative
- 2012-13 Three new stations open: Talbot Avenue, Four Corners/Geneva Avenue and Newmarket
- 2014-MBTA announces one affordable fare, new weekend service, and an RFP for DMUs to improve service to rapid transit-like speed and frequency
- 2015-New Governor Charlie Baker scraps plan to purchase DMUs
- 2016-Blue Hill Ave Station receives remaining funding for construction from State
Ari Ofsevit has argued that instead of DMUs, electrifying the line would likely be cheaper and more environmentally friendly.
Study Says: Black Commuters Spend Longer in Transit
A study conducted by now Secretary of Transportation, Stephanie Pollack (along with Liz Williams and Chase Billingham), used response data from the American Community Survey to estimate commute times by race and mode, an abridged summary is in the table below:
|Mode||Bus||Rail||All Trips||Bus||Rail||All Trips|
|Journey Time (min)||47.1||44.2||45.9||38.7||40.8||40.1|
Pollack had the following to say about the study in this news article
“If we care about equity in our transportation system, we have to pay attention to the bus system, which serves so many low-income and people of color,” Pollack said. “We have to do better. We want a system where nobody’s commute is longer because of the color of their skin,” she said. “What would it take to create that system?”
Then Secretary of Transportation, Rich Davey, had this to say in response to the study:
The state is also spending tens of millions of dollars to add stations to a commuter rail route known as the Fairmount Line that has run for years through Dorchester and Mattapan with few stops. But it lacks the money to run trains on weekends or with the frequency of subways and trolleys, undercutting the line’s usefulness. That could change if lawmakers approve sustainable funding, Davey said. Longer-term and at greater expense, planners are intrigued by self-propelled cars known as “diesel multiple units” (DMUs) that provide transit-style service on shorter commuter rail routes in Europe and Asia, without the lengthy locomotive-and-coach sets found here.
You can find the study here (paywalled).
Fast, Frequent Service with Fair Fares on Fairmount
My thesis research at MIT’s Transit Lab involved determining the origins and destinations of transit trips based on fare transaction data (AFC), in a software we called ODX. When it came time to find an application of this inferred data I sought to answer unanswered questions following the study by Pollack and her associates: Where were Black commuters going? Why did it take them so long to get there? What could be done about it?
I’ll spare you the technical details but briefly: I inferred for every transit user who appeared to be regularly commuting an area where they probably lived and then compared that to the demographics of public transit commuters in that area. I could not determine any individual commuter’s race, but will instead talk about commuters from areas with predominantly Black transit commuters (66%) and predominantly White commuters (83%). The proportions aren’t equal because more White people live in areas with higher concentrations of White people than Black people live in areas with higher concentrations of Black people (choosing thresholds was complicated). For these two areas, the average observed transit (excluding walking to and from transit) travel times are below. The all trips column has a higher value since it includes trips with both modes which, by definition, have 2 stages and are therefore typically longer.
|Mode||Bus only||Rail only||All Trips||Bus only||Rail only||All Trips|
|Transit Journey Time (min)||26.8||27.3||32.6||25.5||25.4||29.5|
Why the reason for this difference in travel time? On average, commuters from predominantly Black areas travel shorter distances, on slower modes (slower bus vs bus, bus vs rail), while having to transfer more times than commuters from predominantly White areas. You can read more details, with tons of charts, here.
Armed with this knowledge and my large trip dataset I tried to devise different (simple and quick) policies to improve travel times. I got pretty frustrated at how few of these seemed to do much on average (but could save individual commuters 5 minutes or more!). Until my advisor suggested I look into running faster, more frequent trains on the Fairmount line. The analysis was somewhat crude, and made were as follows:
- Users access origins (Fairmount stations, or bus or subway stations) from the centroid of their home location (the most likely point of residence)
- Users within 800m (½ mile) walk to the Fairmount station
- User walking speed is 3km/hr and walking distances are Euclidean
- Fairmount headways are equivalent to peak Orange Line scheduled headways of 5 minutes (average waiting time of 2.5 minutes)
- Travel time to Blue Hill Avenue station is half the travel time between Fairmount Station and Morton Street Station. This extra stop adds 1 minute to the running time. Currently the Fairmount schedule includes no explicit estimate of dwell time: departure time is equal to arrival time.
- Fairmount running times decrease by 16% due to decreased dwell time and faster acceleration versus current equipment despite the current scheduled running times on the branch are already the 25 minutes predicted from the introduction of a previous study of the introduction DMUs.
The results of this analysis shows by having 8.6% of trips (~17,000 trips in the morning) saving around 7 - 23 minutes per trips (depending on number of stages, origin, destination), the overall travel time difference could improve by as much as 35%. Note that this analysis only considered trips by certain categories instead of using a routing engine nor did it assume improving the bus network to better serve the Fairmount Indigo Line.
|Trip Type||Average Time Saved (min)||Monthly Commutes that Might Switch to Fairmount (% of all Black Trips)||Impact on Travel Time Differential|
|Origin and Destination within walking distance of Fairmount Line||13.4||4900 (2.5%)||10%|
|Home location within walking distance of Fairmount Line Destination requiring a transfer at South Station||12.4||7900 (4.0%)||16%|
|Origin on a bus route intersecting with the Fairmount Line Destination requiring a transfer at South Station||12.2||2600 (1.3%)||5.2%|
|Origin on a bus route intersecting with the Fairmount Line Destination within walking distance of Fairmount Line||12.1||1600 (0.8%)||3.2%|
Nevertheless, it is evident that fast, frequent service on the Fairmount line could benefit mainly (predominantly Black) commuters by quite a bit. So it is depressing to see that not only is the purchase of the necessary equipment cancelled, but the reliability of the line is ruined by regular cancellations by its private operator.
Still unanswered questions
I will confess that during my masters there was minimal emphasis on soliciting input from the Roxbury, Dorchester, Mattapan, Hyde Park neighbourhoods about their experiences and what solutions they would like to see. There’s a particularly problematic question to what degree housing affordability can be preserved when improving transit. I don’t know.
To end on a positive note though, it’s not all complaints about cancelled trains. Last year the Fairmount Corridor hosted a party train celebrating local artists:
Let me know what you think of this article on twitter @dumasraphael!