A few weeks back, I decided to exercise my freedom in going into the office more than I needed to and commuted downtown on a Friday. The first thing I noticed was how much emptier my usual bus was. Of course, it shouldn’t come as a surprise that Friday ridership is down relative to the other days of the week, although none of the local transit agencies publicly report this kind of breakdown.
Many employers now operate a hybrid working mode, with workers in-office some days of the week, and at home the others. The problem is that “hybrid” is a spectrum: some companies don’t require in-person work but provide an office for those who want it, some companies mandate a few days in-person, chosen at the discretion of employees, and some companies mandate specific days in-person. This lack of predictability makes it difficult to plan for the right amount of service across the week.
Transit agencies have already shown willingness to cut trips and even entire routes for being low-performing. But it raises the question: for commuter routes, especially, does an altogether different schedule for Fridays make sense? This would create potentially four schedule categories for routes that have weekend service: Monday-Thursday, Friday, Saturday, and Sunday.
There are some immediate downsides that come to mind:
- Schedule complexity: riders might have increased difficulty sorting through the right schedule and not mistakenly consulting the wrong schedule.
- Regularity for M-F riders: those who still need to take transit all five weekdays might be unduly burdened by a different Friday schedule. This is particularly true of those in the trades or other professions that do not accommodate remote work.
On the flip side, trimming extremely unproductive trips on Fridays might provide a useful source of hours that can be reinvested in other lines or bolstering all-day service elsewhere, particularly with an uncertain future financial outlook. That said, employment patterns continue to evolve post-COVID so looking at day-of-week ridership will be an increasingly useful tool for planners to consult in the coming years.
Honestly, I’d take the argument a bit further: is there a point to having separate schedules for weekday, Saturday, and Sunday service, now that the peaks have been flattened? I’d rather have a route that is consistently (say) 10-minute headways 6AM – 10PM with decent overnight coverage, than a route that has a very frequent (and expensive!) service in tight bands and 20 or 30 minute headways outside of it.
Back in 2022, ST managed to give the 550 15-minute service all week just by removing a couple peak-direction trips during the week. Keeping a consistent schedule not only helps riders, but also keeps cost down for the agencies and lets them expand non-white-collar service even more.
I think that makes sense assuming there is no crowding. Looking at the RapidRide E, it runs every 6 minutes peak direction, then every 7.5 minutes in the middle of the day. Does it really need to run every 6 minutes during peak now? I have no idea. It wouldn’t shock me if it does. On the other hand, it wouldn’t surprise me if it runs that often to benefit those riders. Those riders may outnumber riders during the rest of the day (even if service was identical) but even then the extra trips might not be justified. While fewer people would benefit with a different bus running every 15 minutes instead of every 30, the benefit is much greater. Then there is the impact of Link (on that particular route). A significant portion of riders from the north will just take a different bus connecting to Link, which might suggest a reassessment then.
The cost to run extra peak trips gets complicated. In general it costs a lot more to run the buses then. The more you can reduce peak service the more savings you are likely to have. I can definitely see us erring towards better service throughout the day, rather than just during peak. This was the approach taken by the G. At one point they considered 5 minute service during peak and 10 minute during the day and went with 6 minute throughout the day. But even that gets messy (as service drops way too rapidly in the evening). We may be contractually obligated with RapidRide routes to provide that level of service, but I could see us erring towards more of a widespread baseline, rather than high frequency on a handful of routes. For Seattle for example the goal should be that every bus runs 15 minutes or better throughout the day. If not, the bus has to be considered a coverage route — which means it should be quite a ways from other routes. For a lot of routes this would require a restructure, but that could be a good thing. In my opinion we need to focus on improving frequency and a restructure would allow us to do that.
My formerly busy urban route which had frequent peak service and an express variant is now down to 15-minute clock-face headways at all times of day including the peak; other formerly mainline routes have fallen to less frequent peak service than that. But at least clock face headways are consistent and free you from having to look at the schedule. Reducing further would reduce frequency to an unattractive level and reduce consistency that removes a barrier from transit use.
Another set back: Route 5 Saturday headway was increased to 20 minutes in fall 2020. It had been 15 minutes for decades.
I would believe that regularity is more important than frequency reductions.
Also, Fridays vary throughout the year as well as the weather. Plus the various sporting events can also affect demand. Holiday weekends and the school breaks affect demand too — and play a factor in Friday demand. Constantly changing schedules by day of week to meet weekly fluctuations would be frustrating for regular core riders and drivers picking their scheduled runs.
Finally, Friday afternoons have notoriously slower traffic. So a bus round trip can take longer than on other afternoons, meaning that the route frequency ends up lower even with the same number of buses. Thanks to the many lane reductions around Seattle in the past 15 years, signals on several bus routes are taking 3-4 light cycles to get through where a few years ago buses would easily get through on one — and it’s often worse on Friday afternoons.
In a perfectly efficient world, the schedules would change hourly and daily. But doing that would turn waiting for a bus more into a guessing game.
So rather than reduce service, why not vary the pricing instead? If there is surplus bus capacity on a Friday and the weekends, entice more riding by making it potentially cheaper! Have rider drawings for free bus passes! Give away low-cost rewards on Fridays like shopping bags or pens! Offer “Free Fridays” for a few hours!
I would think that it’s much better to entice more riders during low demand Fridays to even demand to more like other days than to go down the destructive cycle of service reductions only on Fridays.
Finally, Friday afternoons have notoriously slower traffic.
Prior to the pandemic Thursday was worst. I doubt it has changed — if anything the trend has increased. My guess is people just remember the Fridays before a three-day weekend being really bad (and they are). But a typical Friday isn’t as bad as other days.
Thanks to the many lane reductions around Seattle in the past 15 years, signals on several bus routes are taking 3-4 light cycles to get through where a few years ago buses would easily get through on one
The problem isn’t a reduction in lanes, it is simply more driving. The freeways are basically unchanged or they have added lanes. Same goes for streets like Denny or 5th Avenue Northeast (close to Northgate). Mercer actually got an upgrade to increase throughput (so it wouldn’t be so “messy”). The problem is more people are using these streets. They are more crowded, and this spills out into other streets. The region has grown rapidly, and transit hasn’t kept up (e. g. a lot of HOV lanes still say “HOV-2” even though the average speed is well below the recommended 45 MPH). With slower average transit speeds you have more congestion (Downs–Thomson paradox).
My first intuition is that, with only one Friday per week, the returns from doing this are going to be very limited, especially if the intent is to reinvest the hours into service that runs 7 days/week. I don’t think it’s worth it.
Should we reduce Monday too? The people I see with hybrid schedules go in mostly Tuesday, Wednesday, and/or Thursday. So both Monday and Friday may be light.
On the other hand, it may be premature to shift Friday to weekend service while so many jobs are still Monday-Friday. One issue is frequency; another is expresses. If you eliminate the 218 on Fridays, then it takes an hour on the 554, and that can push the boundary of what travel time people can tolerate.
That’s a good point. If people respond to their express bus not running on Friday by working from home on Friday, the end result becomes downtown on Fridays becoming even more of a dead zone than it already is.
It also makes it more difficult for a company to have a mandatory return to office that applies on Fridays if some people’s bus commute will take twice as long as the other days, due to King County Metro penny-pinching.
And it’s not just people going to downtown office jobs. It’s also people going to other parts of Seattle where the 218 is just part of the trip, so the total travel time is even longer than to a downtown office. Some people take the expresses when they can for miscellaneous trips, even non-work trips, and some long trips just aren’t feasible when the expresses aren’t running.
For me that’s like getting to Kent. Sometimes I can arrange it so I can go southbound in the PM peak (on Sounder), but then I have to come back the conventional way which takes over an hour. As a Seattle resident I can sometimes leverage the PM expresses to get to all the suburbs or outer Seattle neighborhoods, so that at least one half of my round trip is better than both halves taking a long time. I also used the 218 to get to that transit hike in Issaquah we had one day.
I would love to see more Friday evening service, to reflect the reality that people do more traveling then than on other evenings. That goes particularly for light rail.
Not everyone has a smart phone, but most do, and looking up schedules specific to dates is really easy.
Providing date-specific information is not a real hurdle to offering different levels of service depending on the day, if the service levels are set ahead of time. Indeed, Metro and ST have become quite adept at having online schedules reflect cancelled runs. Live-day, not so much, but the route-by-route alerts partially make up the difference. Thank ST that we can look up alerts while a train is stalled in a tunnel!
The real hurdle to right-sizing service is scheduling drivers. But consider that NFL, MLB, and MLS schedules are set months ahead of time. Same with Taylor Swift concerts. Work pick packets that vary by week can be devised with these events in mind.
But varying by days of the week is really simple. I just would not frame it as reducing service on some days. That’s toxic. I’d frame it as adding service on the days and times that need the most help. Since ridership is in a rapid growth mode now, the focus will be on where to add, anyway.
All that said, I hope more Link connector routes can be brought up to 10-minute headway over time.
Right now, I count seven Metro routes with 10-minute-or-better weekday daytime headway: A, C, D, E, H, 7, and 36. The A is the only one among them that doesn’t go downtown.
CT will be adding the Blue and Orange Swift lines as 10-minute Link connectors on Friday, August 30. I hope they don’t get right-sized to less frequency within a couple years.
I think the 44 is 10-12 minute headways, not sure if that would count as well but could still be a bit better.
“Providing date-specific information is not a real hurdle to offering different levels of service depending on the day, if the service levels are set ahead of time.”
There’s two kinds of frequency: baseline policy level (which may be 15 minutes all day on a certain route), and extra to compensate for peak crowding (which may be 8-10 minutes on the same route).
Then there’s peak-only routes, which serve a wider variety of functions in different cases. In some cases they’re simply another way to relieve crowding. In others they compensate for peak-hour congestion delays. The 15 seems to do both. In others they make longer trips feasible (like the 218 example above). In others they give service to a neighborhood that doesn’t have all-day service (like Arbor Heights is or was after the 2014 cuts). So some peak service is simply adding frequency, while others are doing other things too.
Waiting for a bus is one of the biggest things that deters riders or decreases satisfaction. People would rather ride 10 minutes longer than wait 10 minutes longer or arrange their travel around that pulse. So the main issue is, what’s the baseline frequency? If deleting extra peak service means merely reverting to 15 minutes, that’s more tolerable than reverting to 20 or 30 minutes or twice as long travel time. Ideally we’d have 10-minute service all day and evening on all core routes, but first Metro needs to achieve smaller steps. Fill in the gaps in the 15-minute spans, and get at least a handful of routes up to 10 minutes like they were in 2017-2019. If that means some peak service needs to drop from 10 to 15, and that wouldn’t cause overcrowding now, then that’s something Metro should look at.
I don’t think we should modify the base frequency but changing the extra peak frequency might be fine. Of course this is a bit complicated to parse out what is ‘peak’ and not.
Peak was historically defined as 6-9am and 3-6pm Monday-Friday, but it has since spread into the shoulder periods as the population and activities increased. PM peak service now ends around 7pm, which makes it easier to do a stopover after work or get to an evening activity. The AM peak may start around 5 or 5:30: that’s when I’ve encountered increasing frequency. Colleges and tech companies tend to have a later peak extending to 10:30am, which is reflected in routes serving those. Then there’s I-5 between Northgate and downtown, which is now packed from 1pm to 7pm: the entire afternoon, including on weekends. So those are some of the general trends in peak surges. The agencies know all about them, even if they’re not quite clear to laymen.
I think it’s important to mention that jobs that are Friday “optional” are frequently white collar jobs, especially at a middle or senior management level.
In contrast, the typical transit rider is more likely to have a lower-paying service job that they must arrive onsite for.
So I’m totally fine with less crowded buses on Fridays. Those that must come in to work deserve a less crowded bus or train. Keep in mind that our higher frequency routes serve some of the most transit dependent populations — and that seems to be where any Friday service cuts would happen.
“the typical transit rider is more likely to have a lower-paying service job”
In some cities, yes, but Seattle and King County have a large percent of middle-class riders. It goes back to our geographic barriers, linear north-south orientation, Metro’s large network (it got an award for the biggest bus-only transit network in the US in the 1990s), the lingering concentration of jobs downtown after 1960 (a larger percent than most US cities), and Metro’s ubiquitous peak expresses to downtown (which attracted middle-class people to transit). Some middle-class people who start riding transit for work, then later start using it for other things too, or even get rid of their cars. That happens more in Seattle/King County than it does in most of the US outside the ten largest cities.
I don’t dispute your recent history, Mike.
My observation is simply that our commute needs have changed since the 1960’s (60 years ago now). While Metro has been effective at getting white collar commuters to use transit to reach Downtown Seattle jobs through the 2010’s, work-from-home is something that only evolved in the last 5-15 years with the advent of laptops and higher speed internet connectivity. It was incrementally growing in popularity until Covid, which has now changed the commute landscape in what seems to be at least generational if not permanent.
It’s actually the normal human condition to work at or near home. The idea for anyone to spend more than 30 minutes to get to work was rare before 1700 even in Europe. It’s no wonder that it’s appealing to those that can do it!
The topic of this post — Fridays — is driven by a reaction to partially working from home popularity . I’m simply pointing out that reducing Friday service has some implicit bias with it. First, I don’t see Metro reducing the long-distance commute routes on Fridays because there would be outcries by their vocal white-collar riders. Metro won’t reduce suburban routes on just one day a week if it’s already not running frequently. So the administrative decisions to cut Friday service hours would almost surely instead fall to the busiest routes that can accommodate the service reduction without major rider complaining — and those routes (like Route 7) aren’t as full of white collar commuters.
I think a more likely “end state” is for peak transit demand to be increasingly less pronounced. Not only has peak hour ridership dropped, but Saturday ridership is increasing, for example (on Link that number is running at 90% of weekday)! Add to that the growth of flexible hour options for white collar jobs, along with the spreading of non-white collar jobs to requiring a wider set of working hours (as stores and restaurants stay open longer hours).
I’ll briefly mention here too that the opening of the ST2 Link system in the next 24 months is a sea change in how we look at transit service in this region. Trying to hone Friday service hours merely would be making a ripple when the larger restructured network is set up to work very differently in less than two years.
@Mike Orr,
“ Metro’s large network (it got an award for the biggest bus-only transit network in the US in the 1990s),…..,”
That is an absolutely true statement, but it isn’t necessarily a good thing. It’s actually a symptom of this regions traditional underinvestment in rail transit, which is not a good thing.
The transit part of Metro actually came into being for this very reason – the transit parts of Forward Thrust (which would have built rail) had recently been defeated, and the private transit operator on the Eastside was facing bankruptcy. Enter Metro as a sort of “Plan B”.
The idea of Metro running buses was simply to rescue the operations on the Eastside, while consolidating county wide service into some sort of cohesive system. And it was sort of thought to be a temporary situation until some future date where rail would be built.
No one expected that wait to be 20 years, but oh well. At least we are making progress now.
Also note that the express bus system was pioneered by Seattle Transit with routes like the #41 Blue Streak. Thus some of these commuter type routes actually pre-date Metro.
@Al S,
“ the opening of the ST2 Link system in the next 24 months is a sea change in how we look at transit service in this region. ”
This is an absolutely true statement. In 2 years transit in this region will be completely transformed. It’s actually sort of hard to comprehend how much things will change, and how fast they will change.
But I look forward to that change, and I think it will serve the region well.
Six days and 15 hours until the ELSL opens.
“work-from-home is something that only evolved in the last 5-15 years”
Only a quarter of jobs are even compatible with work-from-home. It appears to be a larger portion because tech workers feel entitled and are easily visible in a few places, and the Eastside is particularly tech-heavy.
However, those non-WFH jobs are more scattered and many don’t have as good transit options, so a smaller proportion are transit commuters.
“The idea for anyone to spend more than 30 minutes to get to work was rare before 1700 even in Europe”
It’s still unusual. The average commute time is 20-30 minutes. People start to feel it gets too long when it surpasses the 30, 45, or 60 minute mark.
“the administrative decisions to cut Friday service hours would almost surely instead fall to the busiest routes that can accommodate the service reduction without major rider complaining — and those routes (like Route 7) aren’t as full of white collar commuters.”
That won’t happen. The 7 and 36 are ultra-frequent because they’d be overcrowded if they weren’t. Metro has increased them several times for that reason. And they’re in the equity-emphasis area that Metro and King County have been directing resources to since 2020. The people who would complain are progressive activists, and they currently have the government’s ear.
I wouldn’t say it’s completely transformational as its effect on the City of Seattle is modest — mainly in the Judkins Park area and far north Seattle.
I do think there will be some surprises that will result. Just a few:
1. Sounder will lose riders. The peak nature of Sounder service both north and south will result in those that can to drive to a nearby Link station instead. The peace of mind and the return trip flexibility of using Link will overpower the faster Sounder travel time for many current Sounder riders.
2. Niche Metro Express routes will lose riders. Similar to Sounder, those riders will move towards Link as it’s all-day frequency is so attractive.
3. Some local suburban transit routes will lose riders that will instead drive to a Link station. With thousands of new spaces combined with frequent trains, some riders that lose one-seat long-distance routes will drive to Link rather than drive to a parking lot and ride a local bus to Link. Lots like Eastgate, South Everett and Redondo Heights will particularly lose demand.
4. There will be more drop off and pick up at Link stations. People will perceive that Link stations are better lit, quieter and more patrolled than a nearby bus stop — along with more frequent trains. So for many, just getting to Link will be more preferred. A parent may not want their 13 year-old waiting at a dimly lit bus stop, but will accept them waiting at a bright Link platform, for example.
5. More SeaTac airport riders. SeaTac airport drop off and pick up is very congested. It takes quite a bit of time to reach from anything north or east of Downtown Seattle; rather than drive to their airport, people will ask to be dropped off or picked up at nearby Link. Even if Link is a few minutes longer than driving, the unpredictably of traffic congestion (maybe missing a flight) will result in more Link users. (This assumes that ST is astute at preventing service disruptions — either sudden or planned.) At some point, the last train leaving SeaTac will be later than the current ridiculous 11:45 pm.
As to your points Al:
1) I don’t think Link expansion will have much impact on Sounder. If you are in Everett and ride Sounder you just like Sounder. Right now the alternative is an express bus (which will continue to be offered) or a two-seat ride to Link. I doubt that the fact that this transfer point will change from Northgate to Lynnwood will make much difference. Same sort of thing is true with Mukilteo and Edmonds. That doesn’t mean that ST won’t cancel North Sounder, but they could do that now (based on the enormous subsidy per rider). For South Sounder I don’t see any expansion making much difference. For the most part they serve different areas. The one overlap would be the Tacoma Dome and Sounder would be faster and more comfortable.
2) Those niche Metro express routes were largely put in place because of Link. The idea is that places that are basically downtown from a demand standpoint (South Lake Union and First Hill) are not well served by Link, and since people have to catch a bus to Link, you can replace a 3-seat ride with a one-seat express. They have yet to perform well and Metro has cancelled most of them. To be fair, I think the latest set (based on Lynnwood Link restructure) looks better than previous routes, although they may still fail.
3) Some local suburban transit routes will lose riders that will instead drive to a Link station.
I think local routes will be about the same. But parts of express routes will lose ridership. For example riders used to take the 41 from Pinehurst to downtown. Some of those riders switched to the two-seat ride involving the 75. Others just drive to Northgate Station. The same thing will happen in other areas. I agree that the biggest loss will likely be in the parking lots, and many of the express routes are based on them. A lot depends on the geography though. The Issaquah Highlands are a long ways from South Bellevue. South Everett is not that close to Lynnwood. It is worth noting that some of the Community Transit buses were largely geared around the park and ride lots, but many ran alongside relatively dense areas (even if there was a park and ride lot nearby). Swamp Creek was like that. Thus it will more like the 75-to-Link replacing the 41, instead of everyone just driving to a different park and ride lot.
4. There will be more drop off and pick up at Link stations. Yes, but that is true with plenty of other Link stations (especially Northgate). Northgate has a very popular drop-off location, as well as plenty of taxicabs waiting for rides. Some of that also occurs with the express buses. There really isn’t much difference between the Lynnwood Station and the (existing) Lynnwood bus stop for the 512.
5. Yes, I expect an increase in SeaTac ridership as the train gets to Federal Way. A fair number of the workers at SeaTac live south of the airport, and the train will gather a larger set of the riders. Bus service just isn’t good right now.
To the north I don’t expect much of an increase at all. SeaTac ridership peaked in 2016. It was the only station that didn’t see increased ridership when Angle Lake was added. It is still nowhere near what it was back then, and I doubt expanding further north will have much an impact. The loss of ridership at SeaTac could probably be attributed to two things. First there is item 4 you mentioned. People from the south used to drop off riders at SeaTac where they would then catch the train to Seattle. They switched to using Angle Lake. Then again maybe it was people from the south (riding the RapidRide A) who decided to switch to the train earlier. It is basically a “terminus effect” — the end station of a route in a large, sprawling area (like what we have here) is bound to be especially popular with people who park there, drop people off, or transfer from buses farther away. Thus I will make a bold prediction:
6) Lynnwood Link will cause ridership at Northgate to go down. Northgate is still a destination. It is still a major feeder for buses in the area. But it will no longer be the main feeder for buses coming from the north. For those driving to the station, it will no longer be the northernmost station. Riders will switch to using the other stations. Angle Lake will likely see a bigger drop off than Northgate (when it is no longer the terminus).
@Al S,
“ I wouldn’t say it’s completely transformational as its effect on the City of Seattle is modest”
I would disagree.
Sound Transit’s mission is regional, and that mission doesn’t end at the Seattle city line, and nor is it just a function of the number of stations inside the Seattle city limits. It’s all about improved regional mobility.
Just the opening of Lynnwood Link will increase the number of station pairs available on Link as a one seat ride by approx 50%. Once the full ELE gets tied in it gets even better. And both capacity and frequency will double in the urban core.
The next 2 years will be truely transformational for regional transit. The greater Seattle region has never seen this much improvement becoming available in such a short period of time. Ever.
It is going to be a very exciting 2 years.
I think what Al S. was alluding to with Sounder is the possibility that some people who drive to Sounder at Everett might choose to drive to Link at Lynnwood instead. Either way, it’s car->train, no additional transfer.
The counter argument is that the traffic and parking situation for the car part of the trip is likely to be much easier by sticking with Sounder. Driving to Link would mean 15 miles each way down a congested I-5, in the middle of rush hour, with no access to the HOV lane. Plus, the parking is probably much more likely to fill up at the Link station than the Sounder station.
I can see people doing it if they need to stay downtown late, rendering Sounder back not an option, but for a commute schedule that is compatible with north Sounder both ways, it is simpler and easier to just keep riding it.
I find it rather amusing that some posts strongly disagree with me expecting merely a bit of a shift in transit behavior. I merely said that some people would make different choices when Link soon doubles its number of stations — not everyone! Those choices have all sorts of contributing reasons — home location, work location, other drivers available in a household, need for work schedule flexibility, local traffic congestion, anticipated parking availability, after-work activities, even bad weather!
At the end of the day, Link is going to open unless something catastrophic and unpredicted happens. Some behaviors (but not all behaviors) will change because of it. My musings don’t seem to be that radical and no matter what I or others say now will change the outcome anyway.
ST does not seem to make details about surveying how people get to and from stations or make design choices from such surveys. Other systems like BART study these things more publicly and frequently. So we are left to guess what will happen. It’s kind of sad how we commit to building things like specific sizes of ST parking garages even though the background surveys haven’t been conducted to define the need.
Finally, I agree with Ross that end stations will have fewer boardings as the line gets extended and the feeder buses go to other stations. Again, that’s such an obvious outcome that I didn’t even mention it!
The next 2 years will be truly transformational for regional transit.
I think the wording here is very important. These are big changes in regional transit, but not that big of a change as far as transit in the region. Regional trips (e. g. Lynnwood to Seattle, or Mukilteo to Seattle) are about to be faster and more frequent. But not that many people take those sorts of trips. In contrast the transit trips that a lot of people take are in the city, and they are the ones that benefit the most from tunnels or elevated structures.
But people still make regional trips. People travel from the suburbs into the city, and within the various suburban areas. This effects other cities, that aren’t really suburban in nature (like Tacoma). One thing I’ve mentioned numerous times is that this will dramatically improve the interface between the main mode for our cities (buses) and our subway system. For buses coming from the north the Northgate transfer is awkward. A bus has to leave the HOV lanes and then slog through the exit and local streets to get to the station. The connection to the southern end of Link is even worse. A few buses connect to the airport, but they tend to be infrequent (running every half hour in the middle of the day). These improvements will make a big difference.
But it needs to be put into perspective. There are not that many people trying to get from Tacoma to the airport. If ST decided to truncate all of the Tacoma buses in Federal Way, then a lot of people will have a worse trip between Seattle and Tacoma. Lynnwood Link is a clear improvement, but it probably won’t effect that many people, and the improvement for many of them won’t be huge. Simply doubling the frequency of Metro buses would probably help more people, and the improvements for many of those riders would be bigger.
Partly this is because regional bus service is actually quite good. It is really the best thing about our transit system. Our buses are often stuck in traffic; the network has some big flaws, and frequency is not great. Our subway system — even with the big expansion — is not that good either. Various urban areas have no service even though there are some urban corridors (like the 8 and 44) where grade-separated rail could make a huge difference. In contrast the regional bus system is exceptional. You can take a bus from Seattle to Everett every 15 minutes, all day long! The trips are quite fast considering it serves places along the way. Tacoma was slated to have the same sort of service. I’m not saying the train won’t be an improvement — it will definitely be a big one in the north end — but regional service isn’t terrible right now.
In contrast, trying to get from Roosevelt to Capitol Hill used to be terrible. There was no direct bus. You had to transfer in the U-District to a surface bus that was slow and had to deal with the draw bridge. The time savings for Northgate, Roosevelt or the UW to Capitol Hill was huge. I only see that with a handful of locations as Link expands north and south.
I do think the situation for the East Side is different, and more like an expansion inside Seattle. Bellevue can be considered an “inner suburb”. Downtown Bellevue is huge, and any improvement there is big. Several emerging neighborhoods east of there will benefit as well. But again, there are weaknesses. A lot of the improvement is based on the way things are now — a recent low point, essentially. When they kicked the buses out of the tunnel it dramatically increased the time it gets to get from Downtown Bellevue to Downtown Seattle. Ridership in the 550 peaked back then (not 2019, like most buses). East Link will be an improvement over what existed back in the day, but not a gigantic improvement for Mercer Island or Downtown Bellevue. The areas to the east of Downtown Bellevue also improve. That may be the biggest improvement. They have a faster trip and avoid a transfer. But if you go far enough east, the riders are actually better off with an express to downtown. Meanwhile, riders from Issaquah have to transfer to get downtown right at the point where their trip was about to be really fast. The same thing happened with the 41, but riders there got a fast trip to Roosevelt, UW and Capitol Hill as part of the deal. In this case it is simply a transfer compared to what they had before the buses got kicked out of the tunnel.
Again, I’m not saying it isn’t good — it definitely is — but it needs to be put into perspective. This isn’t like U/Northgate Link. Nor is it like projects that could be built in the city (that would benefit a lot more people, and save those people a lot more time).
I think what Al S. was alluding to with Sounder is the possibility that some people who drive to Sounder at Everett might choose to drive to Link at Lynnwood instead. Either way, it’s car->train, no additional transfer.
Yeah, I’m just saying that you can count those people on one hand. Sounder is a different beast. It serves a different area.
First consider Federal Way Link. Those that are closer to Link than Sounder probably just take an express bus right now. Those that are willing to drive a little extra to Sounder will continue to do so. Link doesn’t really change things for peak riders, since that is when the express buses run. Sounder will remain more comfortable (and fast in comparison).
Now consider North Sounder. If you are in Everett or Mukilteo it is too far of a drive. If you are in Edmonds it could work, but I don’t think that many people are driving to Sounder in Edmonds. There are only 64 spots in the (paid) parking lot, and I’m sure most of those folks are taking the ferry.
I could see some ferry-Sounder riders switching to ferry-bus-Link, mainly because it is more flexible. But again, most of those riders are taking the express bus right now (ferry-bus if you will).
I think it is quite likely we will see a bigger change in the number of people who drive on the ferry, especially in the middle of the day. That isn’t cheap, and this will be a big improvement. The connecting bus service (from both Edmonds and Mukilteo) isn’t great right now, especially if you are trying to get to Seattle in the middle of the day. Right now the vast majority of ferry users on both lines drive. Only 7% of the Mukilteo ferry riders arrive on foot, and only 13% of the Edmonds ferry riders do (in contrast about half the riders of the Bainbridge ferry walk on). There are about 16,000 people a day who drive (or are a passenger) for those two lines (combined) compared to only about 350 (total) riders for North Sounder. Even a small change in the number of people driving would be a big increase compared to people switching from Sounder.
I think it’s important to mention that jobs that are Friday “optional” are frequently white collar jobs, especially at a middle or senior management level.
In contrast, the typical transit rider is more likely to have a lower-paying service job that they must arrive onsite for.
I think that is a fair assessment, but there are other considerations. The lower paid worker is less likely to work 9-5, Monday through Friday. Thus if they are running buses during peak on Fridays (and they don’t have many people on them) it is quite possible they are shortchanging other workers who have different shifts. I think you would have to look at the numbers, but I think that is what this proposal is suggesting. For a long time there was a very strong Monday through Friday peak transit pattern (involving commuters). That hasn’t completely disappeared, but it is smaller than it used to be. If it Friday is like Saturday, then then some of the service during Friday might be better off spread throughout the week or to the weekends (where service tends to lag).
This debate is also similar to the question of whether to reduce service on a weekday that corresponds to a minor holiday (e.g. Veterns Day) where most people don’t get the day off. Or, for routes near the UW, days when the UW is not in session.
From a pure bean counter’s perspective, it may seem efficient. But, it causes rider inconvenience and confusion, and there just aren’t enough of these days for such service reductions to have a noticeable impact on the level of all-day baseline service.
Likewise more Thursday night service given Thursday is the after work happy hour day. I have a friend who takes the KC Water Taxi and has to race to catch the last boat on Thursdays at 6:40pm and can’t stay downtown for events or drinks after work. Plus West Seattle north of the Junction has poor bus service as an alternate to the water taxi.
Going back to the original comment, the logistics of scheduling a significant decrease in service on Fridays is significant. Like other full time jobs, bus operators work 5 days per week. Monday – Thursday is only four days. Yes, there could be some ten hour runs. Yes, there could be some people that work Sunday – Thursday.
That could probably be overcome. But the idea of having tailored schedules that would require a new bid/pick every two weeks would likely cost more money than it would save. I don’t know how KC Metro drivers bid their runs, but in places I’ve worked the process requires a lot of extra manpower, with additional drivers needed to replace drivers bidding, union representatives on site at all times, and supervisors entering run picks instead of supervising.
As others have noted, customer confusion strongly indicates a regular schedule for each weekday. I don’t think we want people thinking things like: “This is Friday, so the trip I normally take does not operate unless there is a Mariners home game – and, I forget if the Mariners are home this weekend or if they are away, and besides, is it really Friday? I want it to be Friday but it might be Thursday so I might have already missed the bus I usually take on Thursdays”.
The buses in south King County are already unreliable the way it is. I’ve been taking buses to work for a few months, including on Friday. My work-from-home day is Monday, not Friday. It’s been absolutely awful. I’ve been stuck chasing buses down because of poorly-timed schedules, waiting 30 minutes in the cold, and taking an Uber for $50 to get to work on time. Completely ridiculous. This is WITH me driving to a transit center and only counting on the bus to get me to another transit center with the assumption that I’ll have to take Lyft or Uber for my last mile. NO NO NO NO NO NO.
Which routes? Some routes are notoriously late every day, like the 28/131/132 or the 62. The 10 has gotten noticeably worse the past couple years, which is odd because it’s so short and has no obvious freeway-entrance/bridge bottlenecks. I don’t have an exhaustive list of unreliable routes, but my impression is some routes are much worse than others.
3 options:
ST578NB-ST574SB
KCM181WB-ST574SB
ST578SB-PT400WB
It may get harder for transit agencies to justify providing the same level of service as they were pre-pandemic for fewer riders, with more people now working from home at least part of the time and, as Josh at PSRC said, “doing Costco runs and errands” (on workdays). The problem is that not everybody is working at home on the same days. Still, this suggests some reduction in service for the days with the least ridership, as riders are smart enough to adapt, contrary to what transit planners believe. Transits also need to be more flexible with their BRT schedules. For most, they’re wedded to the concept of a certain frequency for most of the day, even when the ridership is primary during the peak hours. You’ll never see transits report other than total ridership, which impresses the masses, and they do it because that obscures these kinds of truths and the enormous cost of providing one hour of bus service, $200-300, for what may be a handful of riders at certain times of day.
“Transit riders are smart enough to adapt.”
It isn’t that transit riders are or are not smart enough to adapt. It’s that life simply does not allow it!
Neither my wife nor I have that luxury. We have toddlers on a sleep schedule, shift work at a hospital, and client-driven meetings that dictate our schedules. Rarely, if ever, does a bus schedule work with that mix.
Speaking to folks with fewer resources, working in retail, customer service, manufacturing, construction, or food service has near-zero flexibility. Retail, food service, and customer service establishments have set hours and need to have employees to cover all shifts. At the end of the day, the boss is setting your schedule, not you. In most facets of construction, you are frequently limited to daytime working hours, either because of limited sunlight, or because of City noise ordinances. People who work in manufacturing need to keep an assembly line, whether small or large, operating. This requires all of the skilled labor to be on-site at specific times to keep things going. Service techs, even those using company vehicles, need to be able to service their operating hours, so if there’s a possibility of overtime or a need to arrive at the shop early to make an 8:00 am appointment, there isn’t flexibility in that job either.
PSRC and the transit agencies are accommodating real life for real working people. Most people can’t just be dedicated transit riders. Most of us are using as a tool to live our lives. Too many transit “advocates” like you forget that!
With that, I’ll resume my car-shopping online…
Just to echo what engineer wrote: frequency is very important. Jarrett Walker writes about it in his book and mentions it on his blog quite often: https://humantransit.org/basics/the-transit-ridership-recipe#frequency. One thing he mentions in the book (but I can’t seem to find online) is what engineer mentions Imagine you have are a shift worker and you get off work at 5:00 PM. You can’t leave early, and there is no reason to stay late. The bus runs every half hour and goes by your work at 5:20. If the bus runs twice as often you would be much better off (even if the timing isn’t ideal). It isn’t just workers, either. Medical appointments, meetings, shows — they have a particular schedule and if transit isn’t frequent, you are often out of luck.
It is probably one of the most misunderstood aspects of transit. Everyone understands the importance of speed but underestimates the importance of frequency. The studies are also very clear: Improve frequency and you get more riders. Of course you get diminishing returns, but that argues for less of a focus on peak, and not more. For example imagine a route is running buses every 6 minutes peak, and every 10 minutes in the middle of the day. Running the buses every 5 minutes during peak is a meaningful improvement, but it isn’t nearly as good as running buses every 7.5 minutes in the middle of the day — even though it would cost the agency the same amount.
The longer your trip the less important frequency is. Again, studies show this to be true (comparing commuter rail versus local metro service). It is also intuitive. If I’m taking a trip to Portland speed becomes a bigger issue and frequency is less important. I don’t care if the train runs every ten minutes — hourly would be fine. If I’m trying to get three miles down the road then frequency is quite likely the most important aspect to transit.
I’d like to add some corollaries to this.
1. A transit trip that requires a transfer is only as good as the worst frequency of the trip. Even if a trunk line runs every 6 minutes, a connecting bus at 30 minutes means that the effective frequency is 30 minutes and possibly longer if there is a fear about delays on the initial frequent leg.
2. The less frequent a bus or train is, the more important to get to the stop early so the trip isn’t missed. A parallel express service loses some of its travel time effectiveness if it isn’t frequent enough as riders will be adding extra time waiting at the stop to make sure they don’t miss it.
3. Limited hours that express transit service operates is another frequency issue, particularly if the service is only a few runs. I had a friend that recently accidentally hopped on a Sounder North train rather than a Sounder South train at King Street. That mistake cost him over 3 hours as well as a costly Uber ride for one leg. Had that mistake been made on Link, the hassle would have been negligible.
I had a friend that recently accidentally hopped on a Sounder North train rather than a Sounder South train at King Street. That mistake cost him over 3 hours as well as a costly Uber ride for one leg.
Oh man, that sucks. After a soccer game I once accidentally paid for a Sounder trip (instead of a Link trip). I realized what I had done fairly soon after but was with someone and didn’t want to go back and “un-pay”. I paid the max fair (Lakewood).
I’ve also taken a bus the wrong direction, as well as the wrong bus. You are right. The key is frequency, as you can often just turn around and go back.
> It may get harder for transit agencies to justify providing the same level of service as they were pre-pandemic for fewer riders, with more people now working from home at least part of the time and, as Josh at PSRC said, “doing Costco runs and errands” (on workdays). The problem is that not everybody is working at home on the same days. Still, this suggests some reduction in service for the days with the least ridership, as riders are smart enough to adapt, contrary to what transit planners believe.
This has already been happening. king county metro removed a lot of the extra express bus routes and merged them together with the regular bus routes. And on the flip side increased weekend bus service. It’s generally been a “flattening” of the demand and bus service to be less peaky