Report Shows Muni Was Less Reliable and More Packed in Late 2009

3013734814_3d6f55712b.jpgIn the final three months of 2009, buses broke down more often than in previous quarters. Flickr photo: WarzauWynn

With grim news on most Muni-related fronts, including massive service cuts coming in two weeks, Muni managers also have to contend with a new quarterly performance report [PDF] that shows on-time performance suffering and more missed runs than usual in the last three months of 2009.

Service got worse in several key areas, including customer-observed schedule adherence, headway adherence, scheduled service hours delivered, vehicle availability, and the percentage of buses and trains exceeding maximum load during peak periods. Some categories, like headway adherence, were worse than the annual average for any of the previous five years.

As abysmal as those trends are, they’re not exactly up-to-date: they reflect the two months before Muni implemented service changes in December, as well as the first several weeks of those service changes. Unlike the 10 percent service cut due to arrive on May 8, the December service changes were virtually neutral in terms of the total amount of service hours Muni is scheduled to deliver.

The numbers suggest that one of the most serious problems is a lack of proper vehicle maintenance. While last year’s report was marked by a dramatic worsening in the mean distance between failure (i.e. breakdowns) for light rail vehicles, the end-of-2009 quarterly report suggests that trolley coaches (electric buses) and motor coaches (diesel buses) were suffering the worst maintenance issues. Overall, the mean distance between failure for buses dropped for a second straight quarter.

One of the few bright spots in the report is that operator absenteeism improved, dropping from 14.8 percent in July through September to 12.9 percent in October through December. But that positive trend may already have reversed, judging from daily service reports in early 2010.

Overall, the report paints a picture of the dangers of not properly funding vehicle maintenance. When Muni delivers fewer service hours than scheduled because of vehicle breakdowns, it’s ultimately the same to riders as if there was an official service cut — except it’s less predictable.

Of course, many of the additional budget-cutting moves that Muni made late last year to balance its mid-year budget deficit didn’t happen in time to show up in this report. When the report for the first three months of this year comes out, it could show even more of the consequences of cutting core areas like maintenance.

Earlier this year, Muni learned a similar lesson when it cut overtime spending: Not having operators to fill in for absences also leads to de facto service cuts. Just how bad the consequences of budget-snipping will be for the transit operator won’t be clear until the the quarterly report for early 2010 is released.

  • Nick

    It’s suddenly become clear to me why MUNI drivers must make so much money: they need to afford cars so that they don’t have to rely on MUNI for transportation.

  • Yesterday, going back from SOMA (on 3rd near SFMoMA) we got on the 8x. Sweet, this will take us right to Green and Columbus without all that Chinatown mess. Oh wait, we get to Kearny and Sutter and the bus stops. The driver announces she is ahead of schedule and needs to sit for 6 minutes. We get off and start walking. Low and behold a bus comes up Kearny once we reach California. We hop on, and it is a different driver. Ok cool, we don’t look like fools. Sadly, the bus gets to Columbus and stops. No announcement but the doors stay open. My wife goes into shop at the store while I stand on the street and watch as people get confused and get off. Finally, I’d say 3-4 minutes later, the original bus comes and passes the one now sitting. I think it ended up leaving a couple minutes after that.

    What the hell is the point of leap-frogging each other? Who sets these stupid schedules on a busy route like the 8x. So what, you are six minutes early, let me get to the end of the line and you can sort it out then.

    MTA management is the one who needs a ballot measure calling for their jobs. How can you justify paying a driver to sit with a half full bus because some make believe schedule isn’t being adhered to?

  • If you want to improve schedule adherence without addressing core issues, such as traffic congestion or drivers not leaving when they are supposed to, then you add layover times to the middle of the routes and you allow more time for a bus to complete each trip. This means that a bus is more likely to be able to adhere to the schedule even if the bus is sometimes stuck in traffic. Of course this also means that you are paying for buses and drivers to just sit around for minutes at a time. Plus you are taking the biggest problem with Muni (it takes too long!) and making it even worse for passengers.

    The publicly voted schedule adherence mandate is now most likely the reason for the system being even more problematic.

    And yes, one of the reasons that NextBus exists is so that passengers can know when their bus is going to arrive, even if the bus is not on schedule.

  • cr

    I agree with you about the 8x and other frequently served routes. But schedule adherence is important on routes with longer headways between buses. Try taking the 48 down 24th to meet Caltrain. Headways are every 24 minutes after 8:24 am, and that’s about to get worse. If the bus is early and you miss it, then you just missed your Caltrain connection, and you might be an hour late to your final destination.

    Next Bus may help you catch an early bus but it’s not always accurate, and even then it’s not a good solution for people who have fixed connections, inflexible work schedules, etc.

    As a rule of thumb, if headways are longer than 15 minutes, I think the buses need to be on schedule, and if that means waiting at stops, then that’s what they need to do.

    Buses that are ahead of schedule tend to pick up less passengers and get even further ahead of schedule, while the next bus picks more passengers and gets even further behind. It doesn’t help overall system efficiency.

  • cr, fair enough. But I definitely wasn’t advocating dropping schedule adherence on less frequent buses.

    What about two or three buses leap-frogging each other or all bunched up? I think schedule adherence and a lot to do with that and only hurts system efficiency.

  • cr

    If buses are adhering to schedules, by definition they can’t get bunched up. That can only happen when a bus gets behind schedule or a bus gets ahead of schedule. (And I would guess that a common cause of a bus getting behind schedule is the previous bus getting ahead of schedule.) Good schedule adherence should mean good headway adherence, which should mean no leapfrogging buses.

    But yes, we agree, on frequently served routes, Muni should just forget about the schedules, and run as many buses as they can as fast as they can.

  • Manish

    One of the issues, is that the on-time requirement in the Charter states that a bus is considered late if its any more than 1 minute early or 4 minutes late relative to the schedule. That means going faster than the schedule is frowned upon and actually penalizes MUNI. For buses that are relatively frequent (say 15 minute headway or better) headway adherence is probably a better yardstick than on time. I don’t care if the 14 Mission shows up when it is published to show up. I don’t know or care about the published schedule. The main thing I care about is that if there is a 5 minute headway that I don’t have to wait more than 5 minutes for a bus to arrive.

    btw, do you have a link to the study?

  • cr, I get it now. You are correct.

  • Just got the SFMTA press release regarding the May 8th service cuts … I see my former neighbors up on Potrero Hill who rely on the 10 bus got the shaft pretty good with service wrapping up at 7:20pm … sadly, many folks who don’t work at government jobs work well past that time and still need to get home somehow. This reminds me, I’m going to a neighborhood meeting this evening … need to take my Fix Muni Now petitions to that meeting!

  • Is the Fix MUNI Now petition the one about TWU pay? Don’t hold your breath that is going to solve anything more then continuing to push off parking meter extensions.

  • DT

    The City & County of San Francisco is NOT known for performing routine maintenance of ANYTHING. One look at our streets and parks would ascertain this. They will, of course, maintain their palace at Van Ness and McAllister.

    They would rather spend on social services for out-of-towners and float Bond Measures.

  • david vartanoff

    Yeah drivers playing leapfrog to stay “on time” is so obsolete. Muni’s Inspectors w/clipboards need to go. And route still running 4 or more BPH should be headway managed by the Inspector w/laptop and a radio watching the buses in real time.

    As to the maintenance debacle, check out the Daily Reports for the LRV deficits.

  • Nick

    Blast from the past: people standing at MUNI stops in the 80’s with a paper copy of the “Times Tables” and then a confused look on their face.

  • @Jamie, doesn’t the 10 end around that time anyway? I can’t recall seeing one after 7:30 anyway. Of course, I haven’t taken it since the cambios significantes left it even more packed than usual and with longer headways. If the bus doesn’t come for 25 minutes and it takes a half hour to walk home, I walk.

  • marcos

    Most all of the fleet was replaced between 1995 and 2000 or so. The trolley coaches and diesel vehicles are at the midpoint of their 20 year useful lifespan. This means that they are more inclined towards failure and that any skimping on maintenance over the first half will show up markedly from now to the end. Anyone who remembers the Flyer trolley coach fleet’s demise in the 1990s has a clue as to what is in store.

    The MTBF numbers do not bode well for the Metro fleet in a few years time, as they are just before the halfway point of their 30 year useful lives.

    One intention of Prop E in 1999 was to ensure that Muni was maintained through a dedicated funding stream.

    If it turns out that The MTA has been draining capital from the fleet due to deferred maintenance, then the imperative for structural reforms cannot be clearer.

    I support eliminating on-time performance as the chief metric for evaluating effectiveness in favor of appropriate metrics by line category.

    Changing standards is not going to maintain the vehicles to reduce MTBF, ending the practice of using the MTA as an ATM through work orders will, raising revenue will.

  • JohnB

    I think the focus is too much on schedules which are near impossible to adhere to in a town like SF anyway, even if the drivers weren’t so idiosyncratic.

    If the schedule is “every 10 minutes” and a bus is running 10 minutes late or early, then it really shouldn’t make any difference to the passenger.

    On a frequent service like the 38, the schedule is much less important. If it runs every 5 minutes, then you don’t set out to meet a specific service but rather just catch the first one that comes.

    While on less frequent services, NextBus (when it’s accurate; it often isn’t) should at least in theory do a better job than hope and prayer.

    But nothing is more frustrating than a service simply stopping with a full load. Does anyone know if the apocryphal story about the N train is true i.e. passengers threatening and forcing a driver to complete his route when he tried to cut it short? Either way, there is something about that story that almost every muni rider can sympathize with.

  • The Muni Metro Breda trains are still a few years from their mid-life points and were supposed to undergo manufacturer-required overhauls to keep them running another 15 years, but years ago the funding for that disappeared.

    Funding to cover required maintenance (overhauls were estimated at $1.5 million per-car) started disappearing in a big way in 2005 to cover T-Third Street cost overruns and delays. Many of the wrecked vehicles (we’re now up to 16 completely non-servicable vehicles now I think) will be rebuilt, but otherwise the rebuilds have been scaled back to fixing doors, trucks and other critical components in-house.

    Budget cutting has taken its toll on the maintenance crews though and as of 2007 the staff charged with maintaining the LVR fleet had been cut by 38%.

    To top it off, federal matching funds for rail vehicles only covers replacement once every 30 years, so if we don’t keep them running we’re not getting help replacing them until 2022.

  • Sounds like we need a central subway!

  • I think you’re right Josh … the 10 stops around 7pm today, so no change on May 8th. I guess I just didn’t realize it stopped running so early already .. I used to ride it home from work in the FiDi until at least 10pm or so when I lived over at Texas and Mariposa Streets.

  • Alex

    The problem with NextBus is that it’s so dependent upon schedule adherence that the predictions are nearly useless. If a bus goes MIA (or was never there in the first place), a NextBus sign will still bleat predictions (presumably based upon the schedule). If you’re unlucky enough to be near the end of a route (+/- 20 blocks for the rail lines at least) you might not even get predictions.

    @mikesonn and @Manish Schedule and headway adherence are hugely important. However, adhering to the schedule shouldn’t mean leapfrogging. If all of your vehicles bunch up and run early, you’re left with huge gaps in service. For a brief moment in time you’ve got great service. And for a longer period of time you’ve got NO service.

    The problem is, of course, that the MTA fails at sticking to both headways and schedules.

    @david I was watching one of the street supervisors turn some N’s around at La Playa the other day. Actually I was hoping to catch one. It probably took about five minutes to turn each train around (while about 4-5 two car trains backed up down Judah). You don’t need fancy laptops or radios. You need employees who will a.) cooperate with their coworkers and b.) communicate with their coworkers.

  • Andy Chow

    I am wondering whether wheelchair loading is a cause of the delays, and how significant that is.

    Loading a wheelchair passenger can be a hassle on crowded lines. The driver has to clear the space for that passenger, and the driver have to lower him/herself to put the seat belts on. On lines that have close headway (or a corridor with multiple trolley bus routes) than the impact can be significant.

    Not as if Muni can do anything to deny wheelchair riders from boarding, but paratransit like wheelchair rider only service could be deployed to provide disincentive to wheelchair riders from using certain regular routes.

  • JohnB


    Yes! I have often argued that there is a huge hidden cost for wheelchair carry, especially on busy lines like the N or 38. Holding up 100 or 200 people for up to 10 minutes, comes with a considerable cost i.e the value of up to several dozen man hours.

    And causes disruption to the buses and cars behind, as well as general traffic.

    I am yet to be convinced it wouldn’t be cheaper to put on minibuses, or even taxi’s, for the non-ambulant.

  • James Figone

    I find myself in agreement with John B. on this one. If we could provide bus stop spacing every 4 blocks instead of every block, the efficiency gain might be enough to provide excellent dedicated service to those who have mobility impairments. A separate (not equal but actually better) system may be better for everybody.

    I am in favor of providing great service to everyone, including the disabled. Using the taxis more effectively, perhaps as an adjunct part of the public transit system for those that need it is an idea that has merit and should be looked at. I understand that taxis are part of the paratransit program. I have been told by paratransit users that it is unreliable. Perhaps someone here could comment further on this.

  • @alex, fancy laptops and radios are indeed nessacery to effectively balance out service. Being cooperative and communicative isn’t enough for a street supervisor to know the exact position of every vehicle on the line which are ahead or behind schedule and where bunching is occurring.

    SFMTA already has the technology it needs but so far the line management center in their main building goes unused. It was built to be mission control with four monitors at each workstation which show supervisors at a glance where every vehicle on a line is with color coded makers showing how early and late it is. In the situation you described the supervisor would have been able to see there was a line of trains coming down the track which would be effected by turning one around short.

    Pressure needs to be put on them to use the technology already at hand.


Mayor Newsom’s State of the City Address Breezes Over Muni

Mayor Newsom delivers the State of the City address. Image: SFGTV For anyone still looking for bold action from the Mayor to improve the city’s transit system and rescue it from financial calamity, last night’s State of the City speech didn’t offer any hope. The Mayor did apparently defy his staff, though, including Muni briefly […]

Will Muni’s Largest Service Increase in Decades Have Staying Power?

Muni is making major service improvements and shoring up the basics of running buses on schedule, and this time, officials say, the improvements will stick. “This is long term, focused and systematic,” Muni Operations Director John Haley told reporters last week, calling upcoming “Muni Forward” upgrades the largest increase in service since the Market Street subway […]

Report: SFPD’s New Approach to Crime on Muni “Promising”

The Ingleside station district saw the greatest spike in Muni crime of any police station district in the city during the last several years. Image: City Controller’s office. Until a year ago, there was no written agreement between Muni and the city’s police department specifying what services Muni was paying for and how much, despite […]
Photo: SFMTA

Op-ed: We Need Police Visibility on Muni

Over the weekend, 40 to 60 teenagers swarmed a BART station and train, robbing seven people. The story grabbed the headlines because it was so out of the ordinary. But several disturbing crimes have also hit Muni over the last few months. At 9:40 p.m. on April 12, a mob of about 10 teenagers beat […]