» San Francisco fights to speed up buses and trains by encouraging customers not to buy their tickets up front.
Unlike underground metros or elevated trains, road-running streetcars and buses suffer from a significant slow-down: The time wasted waiting for people to board. The process is dreadfully sluggish in cities with well-used transit systems as large numbers of customers at popular stops are forced to line up at the front door and swipe their tickets or pay their fares in cash. In most cases, customers are forbidden from entering the bus at the rear door, even if they have unlimited ride cards.
In dense cities, the result of these boarding difficulties are buses and trains that practically crawl down the street, even on corridors without much competing automobile traffic. In San Francisco at least, a solution is being studied: Allowing passengers to board at all doors, starting with a pilot program on the Muni Metro J-Church light rail line, which runs from downtown south into the Noe Valley and Balboa Park neighborhoods.
There’s nothing particularly controversial or revolutionary about San Francisco’s proposal. Indeed, the concept of allowing people to get on a transit vehicle at any entryway is is not only standard on most rail networks and a basic component of most bus rapid transit investments, but it is also already in place for some customers on San Francisco’s Muni Metro lines, which operate in a tunnel under Market Street downtown but for much of the remainder of their routes operate in shared lanes like streetcars. What’s different here is the goal to extend the process to all customers on all services.
San Francisco has some of the slowest transit speeds in the U.S., with the average Muni train or bus moving from place to place at a measly eight mph. Those slow speeds are an impediment to easy mobility throughout the city and discourage people from taking advantage of transit.* The causes of the slow speeds are multifarious: The fact that most rail and bus corridors are shared with automobiles, the high density of stops, and, of course, the requirement to board up front. The result have been disappointing reliability statistics: Most services arrive at their destinations on time less than 80% of the time.
Municipal officials have for years been arguing that Muni’s services require an upgrade, and the recent Transit Effectiveness Project (TEP) recommended a number of measures to speed vehicles. Though corridors were analysed differently, the recommendations for the J-Church line (which carries about 17,000 riders a day) were broadly indicative of citywide principles: Increasing the distance between stops — there are currently five stops along the 3,500-foot section of track between 24th and 30th Street, for instance — and allowing all-door boarding. In order to do the latter, Muni will install ticket machines on the sidewalk at stops and require customers to have a proof-of-payment once they get on board at any door. People holding the Bay Area’s universal fare card Clipper can already do as much by tapping their cards at fare readers near the back doors on all rail, though not bus, lines.
Muni has installed card readers at the back of its buses in preparation for a greater roll-out.
Though the agency has not articulated just how much time it expects these trains to save along their route, it does expect to increase reliability as vehicle bunching caused by passengers lining up to get on board and pay will be reduced.
Advocates in San Francisco have been arguing for years that similar measures be extended to all routes, including buses, but the extent to which the city can rely on installing ticket vending machines at every bus stop is questionable. Not only are these machines expensive, but they require constant upkeep and supervision. The question is whether all-door boarding must rely on such on-street machines or whether there is an alternative.
Indeed, one of the great advantages of the Clipper Card — intended to be the fare media for the Bay Area and already useable on BART and AC Transit, among other agencies — is its convenience. It can be purchased in drug stores or by mail. It can be reloaded in train stations or online. Why is it necessary to have ticket purchasing machines at every corner? From a cost-benefit perspective, is installing hundreds of outdoor ticket machines any better than the existing system, where you must buy a ticket in line at the front of the vehicle?
There are other options: Some cities have allowed local merchants to sell tickets one at a time to customers without fare cards, sometimes with a slight markup. Others have coordinated sales with ATM providers, a particularly in-touch approach considering those machines’ ubiquity today. Agencies have encouraged the sale of ten-packs of tickets sold at slightly reduced prices so that customers don’t have to buy a new one every time they get on. In any case, in a city with as many stores and as much pedestrian activity as San Francisco, it is possible to envision a situation in which customers are not provided ticket machines on the street and are not allowed to buy them on board and go about their business just fine.
There are definitely some negatives related to the use of fully off-board ticketing. Fare evasion is already apparently a major issue in San Francisco, with many people either not paying to get on the bus or simply getting on through the back doors, no matter the law. These problems are likely to get worse as it becomes acceptable to board without paying. Either the city steps up vigorous on-board enforcement by police of Clipper Card validation and proof-of-payment from external ticket machines, or it will experience increasing criminality when it comes to fare payments. Considering the speed benefits the other riders are likely to experience, that may not be the worst thing in the world.
These are important issues for San Francisco because as the city works to improve its transit system, it will need to find ways to discourage ticket-buying on buses and trains. A roll-out to the entire network of buses is planned for later this year, if the trial on the J-Church goes as planned. If Muni develops a reasonable and cheap way to do so, it would be the model of every other city working to speed up its own transit network.
* It is interesting to note that while transit speeds are quite low on Muni, regional rail provider BART features some of the fastest speeds in the country thanks to the long average distances between its stops and an entirely dedicated right-of-way.
Image above: Route of Muni Metro J-Church Line, from Flickr user jdeeringdavis (cc)
30 replies on “In San Francisco, All-Door Boarding Catches On”
Muni vehicles have long had all-door boarding everywhere so long as you have some form of prepaid fare — a pass, a transfer ticket, a ticket from an underground or aboveground station, etc. Presumably this is why people find it easy to sneak on board via rear doors, as getting onboard without paying on that vehicle is nothing unusual. Maybe the money for ticket machines would be better spent on more fare inspectors?
1) SFMuni was a transit pass pioneer with their “Fast Pass”.
2) Clipper cards are accepted on all seven (7) of the major transit agencies in the SF Bay area (link below).
3) All-door boarding is typical for the LRV’s in SF. There’s an ordinance on the books that prohibits rear-door boarding on the diesel and trolley buses. That ordinance tends to be ignored at major stops (e.g. Geneva + Mission, Balaboa Park BART Stn.).
San Fran would probably be better off with an outright sales tax to cover fares or go strictly to a card or non-cash system where you can replenish via vending machines – think parking meters.
My family took the streetcar from Ferry’s Landing to Fisherman’s wharf this past Sunday. The cars were packed like sardines, typically summer weekend, and all the driver did was get us to squeeze in more. We end up paying on a return trip but had to chuckle to myself when the driver noted that we paid a quarter too much.
I’m glad they’re doing this, however, I don’t really see the J-Church as a good test. That’s because, as mentioned above, riders who have already paid their fares are allowed to enter through any door on all Muni Metro lines (including the J). So yes, this will speed things up a bit, because now passengers that have yet to pay won’t all be boarding the front door, but it’s not as substantial as it might be on a major bus line.
When I lived in San Francisco and attended SF State, we often boarded the M-Ocean View at whichever door we wanted, and then proceeded to the front of the train to buy passes. There were even ticket machines installed at that station, but most people forewent those machines on the assumption that they would only give you 90 minutes, whereas a transfer torn by an otherwise occupied driver would often give you two hours or more. All this to say which door we boarded was never much of a problem on lrvs in San Francisco.
I don’t want to complain about a small improvement like this, I just worry that this might bring negligible results, leaving critics to claim that the costs outweigh the benefits. But moving toward an all-door boarding system for Muni’s bus lines would be a major improvement far surpassing anything that could be gained on the J, which Muni does not even designate as running on a major corridor.
Yeah, all-door-boarding on the J is nothing new.
TVMs, trolleybus bypass electric lines, and most importantly fewer stops are the major proposed improvements to bus lines — along with all door boarding.
They’ve tried to do some enhanced fare evasion enforcement, leading to a shootout and death of a wanted individual by officers. There are renewed calls from citizens to eliminate police enforcement of fares in parts of the city, regardless of the shootout details.
F-Train is always crazy to the Wharf on the weekends. They tried to raise fares for just the F train (ie: make it like cable cars, $5/use, no transfers), but locals blocked the proposal.
Rear door boarding rules are ignored at any stop with more than say 3 people at it. I’d figure the percentage that aren’t paying is similar to the rest of the system. Current rules treat busses as POP areas, so fare enforcement can be done with periodic sweeps of busses.
I wish Geary would get some serious transit capitol improvements. Its a tragedy that the Richmond is cut off from the city the way it is, the place would change dramatically if it didn’t take 40 minutes to get there from a fairly central part of the transit system. (4.5 miles, takes 17 minutes by taxi, 45 minutes by bus) Hilarity is that its nearly as fast to take a metro train inbound to the major east-west bus on Geary (38 38L etc, 56k daily boardings!) and take it back out instead of taking the worlds slowest north/south bus, the 22 Fillmore (19k daily boardings measured in TEP).
Scott, I do believe that Geary is slated to receive a BRT route. I don’t know right now just where things stand right now on it but I do know that it’s been in the works for several years now.
Local merchants and SMS tickets are the right answert to the ticket distribution question.
This sounds like a good idea as for the five transit stops in 3,500 feet they could at least cut three of them in that as far as walking goes a 1000 feet of walking to a bus stop is not that bad.
Question and comment.
Question: What is California’s traffic law re. the right-of-way for buses to re-enter traffic? Here in Wisconsin, the major factor slowing buses down is that a bus has to WAIT until all the cars pass before it can re-enter traffic. If cars had to wait instead, buses could have more stops AND be considerably faster.
Comment: Consider a public transit system in which there are NO fares. Or consider a transit system in which all fares are paid BEFORE boarding. The latter is what exists on the University of Wisconsin-Madison campus. The university pays the transit agency an agreed fee so no one on campus has to take time paying a fare, nor explain why a fare card was forgot in the rush to make it to class on time. People board the bus without having to do anything right then — they’d already paid ahead of time. General taxpayers could do the same thing.
It’s simple, the city already has a electronic payment card, just make paper tickets twice the price. When London did it, 90% of ticket purchases were electronic within a month. You soon remember to check you have got your Oyster card, before going out, when there is such a price difference. Even many people who don’t commute into the city often have one for the occasional trip.
Boarding a bus happens really quickly when everyone is just slapping their card against the reader. No need for rear boarding on standard buses when everyone uses a card.
Too true. I’d got so used to this that I’d forgotten you can pay cash outside of central London. Was astounded and frustrated yesterday when one person paid with a bill, driver struggled for change and what should have been a three-second transaction took the better party of a minute.
I can hardly imagine front-door-only boarding city bus, being used to short and frequent stops (central europe), even without it, especially older people tend to pack at the front door (because stop sign is there) and it slows down things as hell. If all passengers had to pass through the first door, it would really add at least minutes to an average journey and block other vehicles at tram/bus shared stops, etc.
Sure, with a simple paper ticket proof-of-payment system, fare evasion is quite high and getting a ticket isn’t always trivial (you can get one from the driver, but for a much higher price). There is, of course, a very high ratio of passengers with month transit passes. But most likely, it’d be so expensive to slow down the system so much with front-door only boarding (you’d have to add more buses to keep the same level of service) that the lower fare evasion would not compensate it, ever.
Front loading bus is pretty common in the UK, most still pay by cash to the driver. The speed of loading in London is very quick because of its card system. With one door for loading and another for leaving it works fine. Old people who can’t walk for do sit near the front but they won’t be getting on or off the bus quickly no matter how they pay. Besides in the UK all retired people have a free national bus pass, so no old person pays at all.
People underestimate the level of non payment on honor systems, unless their is rigorous enforcement. It’s expensive to have large teams of inspectors. They are fine for trains but they can’t police a bus system very well.
For systems that recieve very high subsidies maybe fare evasion does not matter to them, but most British systems are profit making, local council may subsidise certain rural routes or some evening and weekend services but the bus companies here want maximum revenue.
Also widespread fare evasion encourages other criminal activity from people who have the spare time to loiter and cause trouble.
I’d like to know how you know that lower fare evasion rate won’t would not compensate.
Trams used to have conductors on who would collect fares and ensure order. In modern tram systems the trams are much bigger and carry many more passengers, this combined with much higher wages out paid to conductors. The solution was pre paid paper tickets, it’s not ideal but it mostly works. As there are large of passengers on a tram it makes is economic for inspection teams to swoop down on them. The temptation has always been to cut back on these to save money but it is a false economy. Amsterdam used to have a big problem with evasion and petty crime but it stomped down.
Both the UK’s DLR and Tyne & Wear Metro are ungated and travel through some dicey neighbourhoods, but they have very heavy on board ticket inspection. When I was in Newcastle I was checked about 40% of the time, on the DLR, every off peak trip was checked.
Nearly all UK buses now only have a single entrance as they require payment or showing a multi-use pass to the driver on entry. Outside major metropolitan areas buses are now usually single deck with a capacity of less than 60 passengers and generally only operate Mon-Sat 7am-7pm.
The few UK tram (streetcar) systems (e.g Blackpool/Sheffield) have larger vehicles with multiple entrances and some employ on-board conductors (not the driver) to collect fares; ticket machines at stops are prone to vandalism/theft.
Not true. In Montreal, the speed of boarding actually went down after the introduction of the Opus cards.
Also, we sometimes have lines of up to 100 people (I even counted that once at midnight!) going single file into an articulated bus with three doors, each having to wait the 1-2 seconds for the Opus card to register.
You do the math.
Doesn’t Montreal have any validators at the bus stops?
Nope, I wish they’d do something SBS-like – given that they have a frequent service network and even some rush hour lanes, they should really look into something like that.
Validators exist in at the front doors of buses, turnstiles exist for the metro, and commuter rail has validators in the stations.
Sigh. Maybe they’re just waiting for Bombardier to develop validators so that they can give it a no-bid contract.
You’d think they’d put on more buses if that many people are waiting at one stop. Don’t they run very frequently then?
At that particular stop the next bus will basically pull in once another has left. To be fair, this is at a Metro stop and a starting point for many (but not all) buses. I think frequency is not an issue as much as slow boarding.
When I was in San Francisco in 2008 I had a most confusing experience on the J train. I got onto it by mistake and by the time I realized it wasn’t taking me where I wanted to go I was quite lost. I eventually just got off and go on a J train headed back into downtown and backtracked to the tunnel and got on the right train from that point.
What was interesting was watching how at a certain point, everyone in the second car had to move to the first car once it began operating like a streetcar. The second car was closed and dead weight for, I assume, the rest of the trip. As well, at that point anyone who got on had to get on through the front doors and show their ticket (or buy one) from the driver. It all seemed very convoluted and strange, I had never seen a light rail system operate like that.
I’m quite glad from the tone of these comments and the article that this has been changed.
The J-Church on the surface has a number of stops that are only long enough for a single LRV. So when an outbound double “J” reaches Church + Market the second car goes out of service.
I think Melbourne’s trams could be a model for all door boarding in San Francisco. The trams have some seats removed, and in their place they installed a full-size ticket vending machine on board. It’s a proof of payment, so all door boarding happens, and I believe there is a smart card reader at each door as well as a place to “dip” your old-style magnetic fare. I don’t think you can pay the driver, but I can’t remember. Sydney doesn’t have vending machines on board buses but you can pay at all doors, which have a “dip” machine for fares. The front door even has 2 so you can skip the luddites who insist on paying cash to the driver.
To combat fare evasion, in addition to more inspectors the penalty should be higher. I can’t think of a POP transit line I’ve been on where you could expect to ever have to show a ticket to an inspector – its never happened to me on Sydney trains, Melbourne trams, Vancouver’s Skytrain, or the San Diego Trolley. Once in a blue moon it happens on LA Metro, although one time when they were checking fares at North Hollywood Station I just walked around them and nobody said anything. They checked when I rode a Melbourne train and you can be 100% sure that on a LA Metrolink Commuter Train they will walk through the train once checking tickets – but if you get on later you can ride for free.
San Francisco is interesting in the sense that there are so many things that they are great at, such as payment infrastructure, network design, and given their traffic constraints, on time performance is pretty damn good.
But then they go and blow all those advantages with their piss poor capital investment decisions and their unprecedented operating cost structure.
For once I partially agree with you. SF has made questionable cap investment decisions for Muni Rapid Transit projects.
First, after the 3rd Street Corridor completed, how did SF not make a 90,000 bus patron corridor (highest in the city) its highest priority? It should have chosen Geary Blvd as its top priority LRT extension out to the ocean.
Second, since it blew the Geary Blvd opportunity, the SF Muni board didn’t have the political guts to scrap the Van Ness BRT and Geary BRT projects to focus Rapid Transit funding dollars extending the Central Subway LRT line northwest to Van Ness and Bay Street. That would make a more successful LRT extension with higher patronage per mile for lower operating subsidy. It would have connected these attractors with a 1-seat ride: Fort Mason, Fisherman’s Wharf, North Beach, Chinatown, Union Square, Market Street, South of Market, South Beach, AT&T Park/Caltrain Station, Mission Bay, Central Waterfront, and City College of San Francisco – Evans Campus. The residents of Bayview-Hunters Point work also have access to more jobs.
All-door boarding has been allowed on Muni Metro trains since the introduction of proof of payment early in the last decade. Nothing new here.
Palal is correct. What this is is another spin exercise on Muni’s part trying to show they are “doing something” when the basic problems remain ignored.
Most days 20 to 25 % of drivers and maintenance workers are absent and total fit for service buses/trolleys are less than the minimum to put the scheduled service on the street. Exacerbating this, many Bredas break down during the day so that by evening rush the system is in crisis.
check out the San Francisco Transit Riders’ Union at sftru.org and at facebook.com/sftru
I’m a little puzzled by this statement: “The result have been disappointing reliability statistics: Most services arrive at their destinations on time less than 80% of the time.”
What does low travel speed and delays in boarding time have to do with reliability? On-time performance is simply a measure of actual travel time minus scheduled travel time. Unless there is extreme variation for the travel times from one day to the next (which could be the result of inconsistent operators and unexpected delays like bad weather or accidents), poor on-time performance is the result of schedules that do not reflect actual running times.
It’s certainly possible that extreme variation is at play here and there is little that adjusting the schedules could do to improve the situation, but equating poor on-time performance with low travel speeds fails the logic test.
The longer the bus/trolley/train is stationary boarding passengers(dwell in transit lingo) the slower the net ride time for all. This is why subways/elevateds figured out level/all door boarding and fare control a century ago. The time wasted by idiots fumbling for change AFTER stepping on is crazy. The fastest fare check by the driver is a “flash” pass or transfer. rfids are second and dip/swipe next. At heavy use stops, fare checking can cost two traffic signal cycles or more. Smart transit agencies use “rear door loader” fare checking in rush hour to speed service at high usage stops.