A few weeks ago, Lizz reported that the union representing Metro bus drivers, mechanics, and service supervisors (among others) approved a new collective bargaining agreement with significant changes to work rules. The most notable of these changes is that part-time Metro drivers can work on weekends. In exchange, union negotiators secured two big concessions from Metro. First, Metro accepted a lower ceiling on the number of part-time drivers, with the limit changing from 45 percent to 33 percent of all drivers. Second, Metro agreed that no drivers, either full- or part-time, will have to work split shifts on weekends. The changes will be effective in September 2018.
Far from being the arcane, inside-baseball news you might think, this is a major win-win for Metro and the union, and a big deal for Metro riders. If implemented well by Metro, it could result in more service on the road for the same amount of money. Depending on how drivers ultimately choose to pick their work under the new rules, it could also make driver recruitment easier—important in an era where Metro has been struggling to keep enough drivers to operate current service, let alone add significant service hours as funded by the City of Seattle and contemplated by Metro’s own Long-Range Plan.
To explain why this is such good news, we’ll have to dive into the murk of bus-driver work rules a bit, below the jump.
There are essentially two types of bus assignments. The first is all-day work, where a bus stays on the road from early morning until sometime at night using multiple drivers. Most bus riders have had the experience of waiting through a driver change while riding a bus that’s on an all-day assignment. The other kind is “tripper” work, where a bus stays on the road only for a few hours and then returns to base. Almost all tripper work exists to add capacity during rush hours, and thus happens on weekdays.
Buses on all-day work may be on the road for widely varying lengths of time, anywhere from 11 to 25 hours. This means most all-day bus assignments don’t divide evenly into the 8-hour shifts usually worked by full-time drivers. To make most all-day bus assignments work, there has to be at least one shorter shift, anywhere from 2 1/2 to 7 hours long. Such a shorter shift may be worked by a part-time driver, or it may be worked by a full-time driver doing a split shift.
Under the old work rules, with no part-time drivers allowed to work weekends, Metro had unsatisfactory solutions for this problem. On Saturdays, it assigned an extraordinary number of full-time drivers to work split shifts. (During my 3+ years as a full-time driver at Metro, I never became senior enough to avoid a split shift on Saturday.) This solution was efficient, but resulted in terrible working conditions for many drivers. On Sunday, no split shifts were permitted. All-day bus assignments that did not fit neatly into a multiple of 8 hours were scheduled using shorter or longer shifts. Drivers working shorter shifts were still paid for 8 hours, despite working as little as 5 1/2. Drivers working longer shifts, sometimes as long as 11 1/2 hours, got paid overtime for every minute over 8 hours. This solution was pleasing to drivers, but spectacularly cost-inefficient. It, in large part, explains why Sunday service is so much less frequent and shorter-spanned than Saturday service.
Now, with the ability to schedule weekend part-time shifts, Saturdays will get much more pleasant for drivers, while Sundays will get much more efficient. In theory, this should allow for more service hours on weekends. Metro’s Scott Gutierrez confirmed by email that Metro expects weekend scheduling to become more efficient as a result of the agreement, even with the improvement in working conditions for full-time drivers on Saturday. Asked whether improved efficiency would translate into Sunday service additions, he responded: “The CBA doesn’t really change Metro’s service plans, although it complements and supports the direction Metro is headed with an increasingly all-day service profile as envisioned in Metro Connects.” In other words, if additional service hours can be added, they will be added consistent with Metro’s overall service planning priorities, with no special attention to Sunday.
To match the “increasingly all-day service profile” Mr. Gutierrez described, Metro has been increasing its full-time workforce over the last few years, in conjunction with the economic recovery. Today, part-time drivers make up just 34.8% of Metro drivers, down from percentages in the low 40s several years ago. All-day service will make up even more of Metro’s total service hours under Metro Connects, as hours now devoted to long-distance peak express routes that will be replaced by Link or Sounder are redirected into high-frequency, all-day feeder services. Accordingly, Metro’s concession on the number of part-time drivers ends up aligning well with the agency’s expected future needs.
While all of this is encouraging, it will be some time, and require some careful parsing of data, before we can tell just how beneficial the new contract is. If Metro’s total cost per service hour does not rise as fast as its total expenses in 2019, it is very likely that the new contract deserves at least some credit. On the other hand, if cost per service hour rises unabated, then the new contract may not be as helpful as we hope.
Is there a limit to how many part time drivers can work on the weekends? Normally the most efficient runcut would result in as many as 35% split shifts on weekends. No split shifts on weekends is a huge win for the union.
Yes, the categories of part-time work that include weekend work are limited to 17% of the total weekly assignments systemwide, so roughly half of part-time operators, and some of those are required to have one weekend RDO. One of the big questions I have is whether it will be possible to schedule the weekends optimally with just that number of part-timers and the Group D full-timers. It doesn’t seem like quite enough to me.
I suspect many full-time drivers will be delighted to shift weekend work to part-timers, to some extent.
It is a good thing Link will help the transition away from the peak-direction express model, with most of the remaining peak expresses coming just from the south end after 2024.
However, peak service will always require more drivers than off-peak service, in order to maintain at least the same headway, because peak runs take longer. Metro, unfortunately, does not take advantage of the faster off-peak runs, by scheduling off-peak runs to take as long as peak runs. So drivers stop for a minute or two along the line here and there, to keep from leaving timepoints early, and thereby fulfill the schedulers’ prophecy.
For the love of passengers’ time and operators’ health, please fix the schedules and move recovery time into the layover period.
And they need to also add time to all the peak schedules. In the middle of some routes during peak, the only on-time bus is a bunched bus
Your third paragraph is false. Looking up a route at random, the 27 is scheduled for 29 minutes to downtown at morning peak, and 25 minutes during midday. Some routes are more dramatic: the 5 is scheduled for 41 minutes for its last midnight run, and 60 minutes during peak. Even a peak express route like the 312 has a schedule that varies from run to run; it’s scheduled for 47 minutes on its first run of the morning, but 67 minutes at the peak of peak.
I don’t doubt that there are runs here or there that could be scheduled for less (or more) time, but Metro does do exactly what you ask.
Those late-night trips on the 5 will only run on time if the bus grows wings. A mere lead foot isn’t nearly enough.
There are actually a number of places in the system where night and weekend schedules are too tight, even though there’s enough time at peak under normal conditions. The 120’s Sunday schedule is another example, as is the 255’s night schedule.
In general, it seems like the process of updating schedules to reflect real world changes in run times is very manual and reactive.
Ideally, with the wealth of data out there, the process of adjusting run times would be more automated.
That’s a good point, especially since we could then have that automated process change other schedules in response. For instance, if the 41 keeps getting stuck in traffic and getting to Northgate late, we could adjust the 345/346/347 in response so they leave after the 41 arrives instead of just before.
Many routes vary by 1-2 minutes every few hours to reflect changing travel time. The 150 is one of the most dramatic:
150 weedays southnound: 6am 58 minutes, 7am 59 minutes, 9am 61 minutes, 12pm 63 minutes, 2pm 66 minutes, 5pm 67 minutes, 7pm 60 minutes, 10pm 54 minutes, 1am 43 minutes.
Saturday at 7am is 51 minutes, 10am 54 minutes, 12pm 55 minutes, 3pm 56 minutes, 5pm 53 minutes, 6pm 52 minutes, 8pm 49 minutes, 11pm 47 minutes.
Sunday at 7am is 46 minutes, 10am 47 minutes, 12pm 49 minutes, 3pm 50 minutes, 5pm 58 minutes, 8pm 46 minutes, 11pm 46 minutes.
As to how good the estimates are, I’m not an expert. But trying to impose a seasonal schedule on random traffic bottlenecks and collisions is a fool’s errand. That’s why we’re building Link, and why the suburbs are so keen on it, and why the buses bunch. The longest catchup stops I see are on minor holidays when the schedule is still weekday.
In general, as long as OBA is reliable, I would rather the schedules error on the side of being optimistic, rather than pessimistic. That way, on light-traffic days, the bus will actually move, rather than stop and wait every few minutes.
As long as a reliable way exists to know where the bus actually is, it makes no difference whether it’s on-time or 15 minutes late. For instance, route 540 is routinely delayed 15-30 minutes (due to no layover before reverse-commute runs), but even with these delays, my actual wait time is usually only a minute or two, since I use OneBusAway to decide when to leave the office. (Of course, for those trips where OBA shows only a scheduled arrival for the 540, I just take the 255 – I have learned to never, ever trust a “scheduled arrival” for the 540 without real-time arrival info to back it up).
Thanks, David. And to STB, for the fine and respectful display of an emblem that will always mean a lot to me. But your last paragraph is your most important one, for both operating personnel and the system they operate.
“Service Hours” are the wrong calculation. Real “Gold Standard” for transit is accumulating cost of every machine in MINUTES, from depreciation to fuel to maintenance to wages and beyond, moving slower than design speed. Especially stopped. Somebody who knows: per division and vehicle…What Is It?
Anybody who knows (which delivered responsibly is labor’s real power) quick report on LINK accident in Rainier Valley a couple of nights back. Only transit’s own performance pertinent. Quick report on general handling, multiplied by the minute. The World needs to know. Especially those of you who’ll be on-scene next repeat.
On principle, I have to take responsibility for a lot that hasn’t worked in the Tunnel since 9/15/1990. But worst regret is that none of us on the Employee Advisory Committee on the Downtown Seattle Project had a successor to fix what we screwed up. Because the committee didn’t either. I should have fought harder to keep it.
30 years is long enough for any missing operating part to stay unreplaced. Coldest accounting: By-the- minute -money saved should pay for a lot of deferred maintenance on personnel yourselves. And every entity and person you serve Through the Union, start organizing the effort to get the authority your knowledge and experience deserve. Call your share Consulting Fees.
You’re doing good, David.
Mark Dublin
According to Metro’s most recent figures** it costs Metro an average of $140.86 per hour to operate a bus.
I believe that’s the “all-in” cost… depreciation, fuel, maintenance, and wages.
So that’s $2.35 per minute of delay — or 4¢ per second of delay.
** http://metro.kingcounty.gov/am/reports/annual-measures/financial.html
As regular commentators know, until last year I was a Republican. In part because I was anti-gov’t union and turned off hugely by the tactics of a non-transit gov’t union. But after watching how technocrats almost trashed Island Transit, are trashing Trimet, and how it’s the public sector unions holding the line for us transit dependent folks… my views are changing.
So congrats to the ATU Local on this one!
Please reconsider, Joe. The Republicans who started Metro Transit had no problem with unions. A shame that the Southern Democrats who failed to take Washington DC in 1864 eventually got it without firing a shot.
So please keep faith with all the Republicans now pathetically pretending to be Democrats. Republicans like these actually started the Progressive movement. Which title the Democrats should feel bad about stealing when they got scared to call themselves Liberals.
Look at real Republicans like the balance wheel that keeps a dynamic engine running smoothly. Also as suppliers of a necessary break from the stress of governing every few years when the Democrats need to regroup and refresh.
And even more necessary the farther left anybody moves. More than one Socialist revolution got slaughtered after it won because nobody knew how to set up an agency that could handle water, sewage, and both buses and trains. Rail/BRT street fighting really alienates formerly friendly populations.
You don’t have any problems with actual Republicans, do you David?
Mark
“You don’t have any problems with actual Republicans, do you David?”
I’m married to one, albeit one who is finding herself rather estranged from much of her party at the moment.
You don’t have any problems with actual Republicans, do you David?
This is silly. Party’s change over time; it’s a normal thing political parties do. The radical, Trumpist Republican party of 2017 is every bit as much “Republican” as the party of 1970, 1920, and 1870. If you imagine yourself a 70’s-style Rockefeller Republican, or a Teddy Roosevelt-style progressive Republican, or whatever, but you remain in the party in the era of Trump, you’re in denial about the fundamental nature of parties. Choosing a particular point in the past and saying the mainstream of the party at that point is more “real” than other times is arbitrary and sentimentalist. Political parties are what they are, not what we wish they were.
Thank you djw. I seem to be happy in the “Independent” zone at the moment.
A question: I suspect that Saturday and Sunday traffic would be more even (level?) than weekdays. And two shifts would cover from 6am till 10pm or thereabouts??
Yes, the peak to base ratio is very close to 1:1 on the weekends. However, there is often more running time and sometimes a higher frequency between 12 PM and 6 PM that requires more buses in the afternoon. Also, not sure how relief points work in Seattle but normally there will be only one relief point where the operators can begin or end their shift, which is usually the closest point the bus comes to the garage. So a long route like Route 120 could only a bus driver start or end their shift every 2 hours or so when it comes close to a base. Not sure if it operates from South Base, but if it does it could only be relieved at Burien and you would also have to factor in the deadhead to and from the garage for the operator. All of this extra time on a straight shift would either be overtime or penalty time.
Typically Metro frequency on weekends is constant from about 8 am – 6 pm and then from 6 pm – 10 pm.
Some routes have multiple relief points, but most just have one, and scheduling the road reliefs is an issue. The 120 is operated out of Atlantic and reliefs happen downtown, where travel time to the relief is paid but travel from the relief is not. On suburban routes, operators will be paid for travel time in a car to and from the relief point.