In San Francisco, not all bus stops are created equal. That’s based on data soon to be published in the Journal of Public Transportation and garnered by in-person visits to nearly 3,000 San Francisco Municipal Transportation Agency (SFMTA) bus stops from May to June 2020.
Finding gaps in public data and wanting to understand why bus stops in his own San Francisco neighborhood at the time exhibited such wide-ranging differences, Marcel Moran, an avid Muni rider and PhD candidate at UC Berkeley, conducted an inventory of benches, bus shelters and signage at every street-level bus stop in SFMTA’s system. (He excluded those under construction or part of other transit agencies such as SamTrans and AC Transit).
The results of his study or “census” found that two-thirds of SFMTA bus stops in San Francisco lack seating or shelter, about one-third of stops were obstructed by on-street parking and more than 1 in 10 stops lacked legible route signage of any kind.
More strikingly, the study also found that disparities in bus stop features were heightened between stops in the northern and southern halves of the city, and that routes with the longest headways (or intervals between bus arrivals and therefore the longest wait times) provided on average the least seating, shelter or clear, unobstructed curbs for bus boarding.
“If you're looking at riders from the south side of the city, [versus] riders from the north side of the city, they're effectively experiencing a different bus system,” Moran told Here/Say.
While there’s much to unpack behind the disparities, what is clear to Moran is that SFMTA is not meeting the city’s transit-first policy, which aims to make public transit an “attractive alternative to travel by private automobile” and encourage “use of public rights of way by pedestrians...and public transit,” in certain categories.
For Moran, there are too many stops with too many inconsistencies and inadequacies to meet San Francisco’s transit-first policy.
A way to make this point hit home, says Moran, is to imagine if the city “used higher or lower quality buses for different routes in the system. Imagine the number 10 bus is like a brand new electric bus, and the number 22 is like a gasoline bus from the 1980s,” he says, noting that the public would likely be outraged by such a blatant show of disparity.
“But the stops are kind of in different eras,” he says.
According to Moran’s data, 45% of the city’s northern bus stops provide seating and 42% feature shelters whereas just 22% of bus stops in the city’s southern half provide seating or shelter.
Nearly all of the city’s clusters of stops lacking seating, or “coldspots,” lie in the city’s southern half, with very few stops featuring seating in Bayview/Hunters Point (San Francisco’s historically Black community) and a cluster of blocked curbs covering a broad portion of the city’s southeastern quadrant. Nearly all obstructed curbs, also known as “coldspots,” also lie in the city’s southern half.
Conversely, clusters of stops providing seating and clear curbs, or “hotspots,” are concentrated in the city’s northern half, including downtown San Francisco and residential neighborhoods running to the west.
However, Moran’s study did not find the same degree of disparities between bus stops in the eastern and western halves of the city. For instance, seating was provided at the same percentage of stops in the east and west (34%) and there was only a two percentage point difference between stops providing shelter in the east and west sides of the city (32% and 30% respectively).
Moran ran a “mediation analysis” to see how much of the north-south stop disparity could be explained by route frequency and determined that geographic differences in route frequency only accounted for 40% of the effect of distribution of bus stop seating, meaning that “60% of the geographic effect documented is unexplained by route frequency,” Moran wrote in his study.
Moran’s analysis did find that bus stops in census tracts with a higher than average share of white residents were more likely to feature seating, shelter and clear curbs in comparison to tracts with higher than average populations of people of color.
For every 1% increase in a tract’s white residents, the odds that a given bus stop also featured seating, shelter and clear curbs also increased by about 1% for each category. While Moran’s study did not find a relationship between income or neighborhood density and bus stop amenities, Moran did observe some higher-income neighborhoods that featured sidewalk designs and strips of vegetation that left “bus riders little room to wait.”
But Moran stresses that further research is needed to explore any possible connections between transit amenities and population demographics.
“Me generating this data set now gives us the chance to push it forward to some of these social demographic questions,” he says.
SFMTA spokesperson Erica Kato told Here/Say that the agency keeps a database of stop amenities and monitors amenities based on federally mandated Title IV guidelines. The agency also analyzes amenities at stops in minority and non-minority census blocks as well as in low-income census blocks and non-low-income census blocks.
According to a graphic shared by SFMTA, minority and non-minority census blocks share the same percentage of stops with shelters (34%) and the percentage of low-income census blocks with shelters is higher (38%) than non-low-income census blocks (31%).
“As you can see, stops in a minority census block are just as likely to have a shelter as stops in a non-minority neighborhood, and more stops in low-income census blocks have shelters than non-low-income areas,” wrote Kato.
Kato also told Here/Say that “if a stop reaches a minimum of average daily boardings” SFTMA attempts to place a shelter at the location, but noted that denials of permits by other city agencies can prevent a shelter from being erected. She also noted that San Francisco’s older streets with narrow rights of way make it difficult to comply with the pedestrian safety mandates of the Americans with Disabilities Act and also accommodate shelters.
“We do have several types of shelters that try to account for locations with narrow sidewalks,” she added. “If the sidewalk is too narrow we go to a smaller footprint shelter that does not have seating. … if the sidewalk is even too narrow for that we might try to place a single info panel that at least has a system map. However, that leads to… [issues with] power access, which is not possible or prohibitively expensive in many of the outlying neighborhoods.”
Moran also found that routes with the most frequent service, or shorter “headways” (10 minutes or less between arrivals), had the highest share of seating and shelters (51% respectively) and unobstructed curbs (88%). In contrast, routes with the least frequent service (20 to 30 minutes between bus arrivals or the longest headways) had the lowest percentage of stops with seating (17%), shelters (15%) and clear, unobstructed curbs (44%).
“The routes in which riders need to wait the longest for the bus, the least frequent buses, have the least amenities with which to wait,” he says. “Another way to think about it is depending on what route you're taking, you're basically using a different bus system.”
Moran says that his counterintuitive finding on headways could be explained by a well-observed trend in transportation literature: transit authorities often determine amenities based on ridership rates—the logic being that more frequently used bus stops or routes should be serviced with more features to serve more riders. However, Moran believes that logic creates a “self-fulfilling prophecy.”
“The problem that I'm trying to identify with that logic... is that the quality of the stop itself influences the ridership,” says Moran. “[People] may not ride there, board there, enter there because of the quality of the stop.”
Moran’s findings also shed light on how SFMTA is meeting San Francisco’s transit-first policy, which not only aims to encourage walking and use of alternative transportation but even has recommendations for parking policy, something that Moran believes SFMTA is struggling with. He found that 32% of bus stops were obstructed by on-street parking, meaning that there was not enough space for buses to pull up or riders to board safely from the curb. (Moran counted bus stops as obstructed by parking if there were no signage or curb markings prohibiting parking; he also noted if parking meters or signage allowing parking were present.)
In his assessment, not having clear and accessible pathways for riders to board or exit buses is a failure in meeting the city’s criteria and brings up bigger questions about accessibility in SFMTA’s system.
“No rider should have to navigate via on-street parking to get out onto the bus,” says Moran, adding that bus stops obstructed by parked cars could be especially difficult for some riders to handle and force riders to exit onto the street into oncoming traffic.
“Imagine your grandparent trying to use this bus stop. Imagine someone with a stroller trying to use this bus stop. Imagine someone with a personal shopping cart trying to use this or with a suitcase. … I would argue that any bus stop where we're allowing on-street parking in front of it, that plainly fails that transit-first policy,” he says.
San Francisco’s transit-first policy states: “Decisions regarding the use of limited public street and sidewalk space shall encourage the use of public rights of way by pedestrians… and public transit” and “parking policies for areas well served by public transit shall be designed to encourage travel by public transit and alternative transportation.”
Kato told Here/Say that one-third of SFMTA’s stops are “flag stops,” where curbed parking is typically not removed.
“For the most part these are low ridership stops and located in residential areas where on street parking removal is [a] trade-off residents are typically not willing to make,” wrote Kato in an email. “Additionally, pulling into and out of traffic to access a curb stop creates reliability issues (i.e. getting ‘trapped’ at the curb by traffic congestion in the travel lane) and increases conflict points between the transit vehicle and other automobiles.”
In order to mitigate these issues, Kato says SFMTA is implementing a program called Muni Forward “to give transit priority over other automobiles.”
Kato told Here/Say that SFMTA has installed approximately 150 “transit bulbs” in recent years, which extend the curb to the travel lane so that transit vehicles can stop in the travel lane and allow riders to board and alight without the vehicle having to pull out of or back into traffic. The agency anticipates installing “many more” transit bulbs over the next decade.
Another worrisome trend Moran observed: wide inconsistencies in signage throughout SFMTA’s system. While he spotted legible signage at 89% of stops, the type of signage varied widely: 41% were metal street poles, 23% had shelter markings, 19% were paint on pavement, 18% had metal signs and 7% were paint on telephone poles. Working ETA screens were only present at 21% of stops and route maps were only present at 30%.
Moran observed bus ID stickers “worn out from the sun and illegible,” pavement paint so worn or obstructed by cars it was difficult to read or locate and bus stickers so high up on poles they were hard to read. More than 1 in 10 stops lacked “legible route signage of any kind,” he wrote in his study.
Moran says signage variance could make it harder for first-time riders, riders with low-vision, tourists and riders whose first language isn’t English to navigate the system.
“I think that all these types of things just make it less likely for people to ride,” he said.
SFMTA responded that the agency is aware of the signage variance and is “initiating a program to install a physical bus stop sign at all 3,000 bus stops in the city to make transit access more visible, user-friendly and uniform.”
Moran says his study is in no way “an attack on SFMTA.”
“I'm a happy Muni rider,” he says, noting that SFMTA faces tremendous challenges as a result of the pandemic and budget cuts. “But what I'm trying to do with this paper is point out ways in which the orthodoxy of American transportation planning around bus stops themselves creates counterintuitive problems”—like routes with the longest wait times having the fewest amenities for waiting.
“I'm not saying every stop and every neighborhood needs to be the exact same because a bus stop on Market Street is serving an order of magnitude more people than a bus stop in Glen Park. These are not equivalent,” Moran says. But two-thirds of stops not having seating or shelter and one-third of stops being blocked by on-street parking is, in Moran’s view, “an inconsistency and an inadequacy in stop amenities that does not meet a transit-first policy.”
Moran hopes his paper can offer possible pathways forward for improving SFMTA, like removing on-street parking from bus boarding areas or adding smaller, design-friendly benches and shelters to the stops that need them. Another solution that could have impact: adding bus stop seating and shelter information to trip planning and navigation apps like Google or Apple Maps to help riders make informed decisions about their rides.
Ultimately, Moran wants to bring more attention to bus stops, which he says don’t get a lot of love in transportation planning but convey so much about a city’s values to riders and non-riders alike.
“It’s a symbol of the city,” he says. “I'm just trying to raise the profile of the bus stop, and San Francisco is an ideal case to do that.”
Christina Campodonico can be reached at email@example.com