MBTA Buses & Infrastructure

As we have established, SL4/5 and the Washington St Corridor are important for two groups of riders: Dorchester/Mattapan transfers, and Washington St locals. The former would see a quicker journey downtown by using the Fairmount Line if it actually had <10 minute headways.
Under both the current map and the BNRD map, Nubian is the biggest bus hub in the entire network. Unless that hub were relocated to a Fairmount stop, the SL4/5 corridor will remain critical for Roxbury/Dorcester/Mattapan transfers. With that in mind, the I-93 "express to Nubian" line that @Riverside and @Teban54 propose makes a lot of sense.
 
I also don't think SL4/5 will ever be "de-branded". They'd literally have to tear the line off of the map -- the headlines write themselves. There's no way the T looks at the PR storm that would prompt and sees it as worth the eventual rehabilitation of the Silver Line brand for use elsewhere in the system; rehabbing the brand just isn't worth the headache.

I hope this isn’t the case as I think this attitude exacerbates the problem.

The T loses credibility every time it over-promises and under-delivers. Every day that the Silver Line is displayed as rapid transit on a map to a first time T rider, they are being lied to in the form of an over-promise that is not being delivered. What you are predicting is continued dishonesty because to the MBTA, admitting the ugly truth is worse than continuing the shameful lie.

It comes down to priorities. Which do you think is a higher priority:
  1. Accurate wayfinding that exposes the reality of the situation.
  2. Continuing a lie in the hopes that it causes less of a PR problem.
I can’t get behind #2, but I understand why you think the MBTA would.
 
Last edited:
I hope this isn’t the case as I think this attitude exacerbates the problem.

The T loses credibility every time it over-promises and under-delivers. Every day that the Silver Line is displayed as rapid transit on a map to a first time T rider, they are being lied to in the form of an over-promise that is not being delivered. What you are predicting is continued dishonesty because you think admitting the ugly truth is worse than continuing the shameful lie.

It comes down to priorities. Which do you think is a higher priority:
  1. Accurate wayfinding that exposes the reality of the situation.
  2. Continuing a lie in the hopes that it causes less of a PR problem.
I can’t get behind #2, but I sympathize with your inclination.
How an about option #3: Fix the Silver Line so it is closer to BRT standards, and hence not a lie.

I know, the T cannot "fix" anything.
 
Under both the current map and the BNRD map, Nubian is the biggest bus hub in the entire network. Unless that hub were relocated to a Fairmount stop, the SL4/5 corridor will remain critical for Roxbury/Dorcester/Mattapan transfers. With that in mind, the I-93 "express to Nubian" line that @Riverside and @Teban54 propose makes a lot of sense.
All key bus routes that connect Dorchester/Mattapan and Nubian already stop at a Fairmount Line station. The 15 and 41 stop at Upham's Corner, the 19 and 23 stop at Four Corners/Geneva, and the 28 stops at Blue HIll Ave. Admittedly this is quite far out, but since the 28 most of its time parallelling the Fairmount Line, and around half of all riders currently start or end their trip within a 10 minute walk of a Fairmount Line station, all the stations (Except Newmarket) would still see many riders switch from the 28 for downtown commuting, likely at least 3000 riders per day across the line.
 
As we have established, SL4/5 and the Washington St Corridor are important for two groups of riders: Dorchester/Mattapan transfers, and Washington St locals. The former would see a quicker journey downtown by using the Fairmount Line if it actually had <10 minute headways.
I think you are overlooking the need for those beyond Nubian. Even just on the 28 corridor -- and even before it gets close to the Fairmount Line -- the population is more than enough to justify better transit.

To be more specific, I'm talking about the Warren St corridor, from Nubian to around Grove Hall. TL;DR: The corridor's residential profile and density is comparable to many other cities that frequently receive calls for better transit, most notably Everett and Chelsea.

For this analysis, I use the Census OnTheMap's "total primary jobs with homes inside the area" as a proxy for number of residents. This is mostly for convenience of drawing any polygon on the map. Now look at the following polygon: Centered on the Warren St corridor, starting just south of Nubian, largely delimited by midpoints to Orange Line to the west, and Fairmount Line to the east.
1711054111461.png


This area has 10,247 job takers in 1.194 square miles, for a density of 8,582 jobs/mi^2.

In terms of the number of residents, the Warren St corridor almost matches exactly with Chelsea, which has 10,251 job takers, as seen below. Chelsea is a little bit denser: with 0.907 mi^2 in my selected area, it has a density of 11,302 jobs/mi^2.
1711054354432.png

(Note that I excluded the industrial areas of Chelsea.)

In terms of residential density, Warren St is a good match for Everett. The area below approximates walksheds of stations at Everett Square and Glendal Square, with the west and east delimiters halfway to Orange Line and the 111 corridor. While Everett has more residents at 16,680 job takers, they're spread over a larger area of 1.772 mi^2, resulting in a density of 9,413 jobs/mi^2. This is pretty close to Warren St's 8,582, even if a little bit above (9% higher).
1711054518865.png


As a side note, while the Everett and Chelsea polygons consist of 60%+ white residents (though half of Chelsea is Hispanic/Latino), Warren St is 60% black.

What about comparing it to places that just got their transit extension in real life? The following polygon in Somerville, centered along the GLX Medford branch corridor, has 18,022 job takers (76% more than Warren St) and a density of 11,235 jobs/mi^2.(30% higher than Warren St). Even though both figures are higher than Warren St, keep in mind that Somerville is known for being one of the densest cities in the region. Oh, and it's also significantly wealthier and whiter than anything above.
1711055137451.png


Lastly, to compare to a heavy rail extension, I chose Malden. (I would have looked at a few others like Quincy, Lynn, Southwest Corridor and West Roxbury, but don't have time at the moment.) The area below is quite large at 2.220 mi^2, so while it does have 16.057 job takers (57% more than Warren St), its density of 7,232 jobs/mi^2 is actually 15% lower than Warren St! This is despite my attempts to give Malden benefit of doubt, by excluding the area NW with lower density. Also, Malden had seen TOD in the decades since they got rapid transit, while the opposite is true for Warren St.
1711055585009.png

This shows Warren St alone has virtually the same population as Chelsea, and almost as good of a residential density as Everett. Coincidentally, while Chelsea is primarily served by the T111, which has about twice as much service as an individual Frequent Bus Route, Warren St has two overlapping FBRs -- except they're nowhere near as good as T111 in rapid transit connectivity.

We have seen countless proposals bringing rail transit, and even specifically radial transit, to Everett and Chelsea. So why shouldn't the core part of Roxbury get the same? Or, put it another way: If it's fine for those at the southern tip of my Warren St polygon (Grove Hall) to do a somewhat indirect 8-min walk to Four Corners/Geneva station, while everyone to the north either walk 20+ mins or take the bus... Then it should also be fine for Everett and Chelsea to get the exact same treatment of Commuter Rail stations (with Sweetser Circle infill) and no additional rail transit.

And remember, this analysis is solely talking about Warren St corridor. It doesn't consider the continued density further south all the way to Mattapan (in stark contrast to Everett), even though at that point the Fairmount Line gets much closer. It doesn't consider the T15, another FBR. (I doubt you'd convince all riders from these places to switch to Fairmount.) It doesn't consider other connecting buses at Nubian, with the 42 being arguably the most notable. And it doesn't consider the sheer size of Nubian as a bus transfer hub, with the highest passenger volume in the system, easily surpassing Harvard and Sullivan (while not even counting riders to Ruggles).

Last but not least... I'm not aware if anyone from Roxbury and Dorchester have considered Fairmount Line improvements "equal or better" than the El. And this is a corridor where we really need to place extra care into what the residents say about their own transit issues, and not make assumptions on their behalf.
What I'm saying is that Roxbury has a firm idea on what Roxbury's transit issues are. But the last half-century of treating Roxbury's transit issues are has involved outside parties dictating to Roxbury choices that were not fashioned by neighborhood need but by other neighborhoods' projects being shotgunned on Roxbury's back and Roxbury being told a vat of half-truths about what is/isn't in their best interests. The end result is that transit shares in the neighborhood are LOWER than they were 3-1/2 decades ago prior to the breakage.

Roxbury is...generationally...D-O-N-E being talked down to about what is/isn't good for them. To the extent that's going to make beneficial transit improvements more difficult is all a matter of approach. Anything foisted on them from the outside is likely to get spat on with suspicion. We saw that with the original 28X in 2009 when utterly mangled public comment period artificially compressed and hostilly carried out sent the neighborhood into auto-shutdown mode. Reject. We are now, in 2020, seeing the unveiling of 28X v2.0...same basic project with little beyond cosmetic tweaks, but achieved through years of slow-cook workshopping between neighborhood and City. There wasn't even any state involvement until 11th hour when they were pretty much done and ready to unveil. The lighter touch and inside-out collection of actionable data was the sum total difference between the 28X that was violently rejected last decade and the 28X that was just proposed to broad consensus now.
 
I also don't think SL4/5 will ever be "de-branded". They'd literally have to tear the line off of the map -- the headlines write themselves. There's no way the T looks at the PR storm that would prompt and sees it as worth the eventual rehabilitation of the Silver Line brand for use elsewhere in the system; rehabbing the brand just isn't worth the headache.

I hope this isn’t the case as I think this attitude exacerbates the problem.

The T loses credibility every time it over-promises and under-delivers. Every day that the Silver Line is displayed as rapid transit on a map to a first time T rider, they are being lied to in the form of an over-promise that is not being delivered. What you are predicting is continued dishonesty because to the MBTA, admitting the ugly truth is worse than continuing the shameful lie.

It comes down to priorities. Which do you think is a higher priority:
  1. Accurate wayfinding that exposes the reality of the situation.
  2. Continuing a lie in the hopes that it causes less of a PR problem.
I can’t get behind #2, but I understand why you think the MBTA would.
Oh to be very clear, I'm not in favor of the T continuing its charade of pretending SL4/5 are rapid transit. I'm just describing why I think the T will overwhelmingly see it as in their best interest to do so. If you are asking whether I think the T will continue to be dishonest, both in general and in this specific case, my answer is yes, to both.
All key bus routes that connect Dorchester/Mattapan and Nubian already stop at a Fairmount Line station. The 15 and 41 stop at Upham's Corner, the 19 and 23 stop at Four Corners/Geneva, and the 28 stops at Blue HIll Ave. Admittedly this is quite far out, but since the 28 most of its time parallelling the Fairmount Line, and around half of all riders currently start or end their trip within a 10 minute walk of a Fairmount Line station, all the stations (Except Newmarket) would still see many riders switch from the 28 for downtown commuting, likely at least 3000 riders per day across the line.
Beyond @Teban54's analysis (which I agree with), one significant shortcoming of the Fairmount Line is its lack of connectivity to the rest of the system. No connection to Green, Orange, or Blue. And South Station is good for access to downtown, but it doesn't cover everything. I do think that (near-)rapid transit frequencies on the Fairmount Line will be transformative in their own right, but I don't think they fill the gap being discussed.
It's the segment around Watts/Compton where in (Mainly) runs in the median. My understanding is that since it has crossing gates it basically has ultimately signal priority at all intersections, which is how it accomplishes the very high running speeds necessary.
Right -- that's why I think that example is not super relevant here. Yes, if we ran tracks down the middle of Washington St and put down quad gates at every intersection, and had only 1 or 2 stations, yes, we'd get high speeds and travel times equal to the El. But I'd argue that that speed comes directly from the aspects of that which would be unviable on Washington St.

Put another way: if there's a scenario where a train may have to yield to pedestrians, there's just only so fast it can go. And I can't imagine a scenario where residents would accept building a train line down Washington that would never yield to pedestrians.
As we have established, SL4/5 and the Washington St Corridor are important for two groups of riders: Dorchester/Mattapan transfers, and Washington St locals. The former would see a quicker journey downtown by using the Fairmount Line if it actually had <10 minute headways. But this solution does nothing for the former group. SL4/5 buses are already very crowded even with insanely tight headways that are under 3 minutes at some peak times, so more capacity is badly needed here. Even low-floor trams would represent a major capacity improvement, even single type 10s would roughly double the number of pax per train. Depending on how Boylston and Park St are used, you could maybe even convert some platforms to run high-floor trains, which as stated previously would be even better for capacity. (Although for an almost entirely street running route with high stop density I'm not sure it's worth the trade off of more expensive, less accessible stops and platforms.) Yes, for the people along Washington St, it might not be quite as fast as the El was. (~10 mins vs ~8 mins, but it would also have more local stops than the El did, likely resulting in faster travel times for some people compared to the El. I would argue pretty strongly that this service could and should be considered "Equal or Better." In fact, if we're only concerned about travel times and not capacity, you could even get most of the benefit with median bus lanes and the existing buses and not change anything else. (Although the loop at DTX/South Station gets a little messy on the surface, light rail can terminate at Park St which is also better for transfers.)
To be clear: as shown on the map I included, I advocate for two branches to Nubian, for exactly the two groups you describe here: an I-93 el for Dorchester/Mattapan transfers, and a modern streetcar line for Washington St locals. Yes, SL4/SL5 are bursting at the seems and yes, they would benefit from the greater capacity of light rail (low floor or high floor -- thank you for finding those stats, by the way!) And yes, I agree that center-running bus lanes would be a big improvement even with the current rolling stock.

What I'm saying is that all of that is necessary but insufficient for Equal or Better.
Yes, for the people along Washington St, it might not be quite as fast as the El was. (~10 mins vs ~8 mins)
The "10 minutes from Nubian to Boylston" claim is where I remain skeptical. Yes, the D Line and LA's A Line can clear 2 miles in 10 minutes, by making 1 or 2 stops, running in a dedicated ROW, where they never have to yield. That seems vastly different from any conditions possible on Washington at ground level.

To bring this back to buses:

1) Yes, build center-running bus lanes if the community can be brought onboard
2) Yes, aim to replace SL4/SL5 with a branch of the Green Line
3) Relieve the Washington surface service (whether bus or rail, Silver or Green) by building a grade-separated branch of the Green Line to replace the old El, so that buses/surface trains can focus on local ridership
 

Back
Top