First Hill Streetcar at Capitol Hill Station / photo by SounderBruce

Correction: In Metro’s Phase 3 proposal, route 64 goes to South Lake Union. (New route 361 from Bothell will also go to South Lake Union via Northgate Station.) Also, route 309 is renumbered as 322 to reflect the detour to Roosevelt Station.

One of the most notable features of King County Metro’s North King County bus route restructure proposed for September 2021, when Northgate Link is scheduled to open for service, is the continued use of north-end and Shoreline express bus service for First Hill. The rest of the express bus service from the north end and Shoreline to the Central Business District will go away.

Metro plans to have four First Hill express routes in operation after Northgate Link opens, three of them competing with Link Light Rail:

  • Route 193 serves Federal Way Park & Ride (S 320th St), Federal Way Transit Center, Star Lake Freeway Station, Kent – Des Moines Freeway Station, and Tukwila Park & Ride before expressing to First Hill.
  • New route 302 would replace some 301 and 304 service, but going to First Hill, with a stop at Northgate Station.
  • Route 303 serves Shoreline Park & Ride, Aurora Village, Northgate Transit Center, and then expresses to First Hill. Routes 302 and 303 are planned to provide alternating service between Northgate Station and First Hill.
  • New Route 322 would essentially be a renumbering of route 309 (Bothell to First Hill), but with a detour to Roosevelt Station before jumping on I-5 to get to First Hill.

The First Hill expresses only operate during peak hours, and only in the peak direction. Given the 24/7 nature of all the medical buildings, this specialty service is mostly irrelevant to a large chunk of First Hill employees, unless they are the lucky ones working the latte shift.

Broadway corridor

The First Hill Streetcar is a straight shot to the Broadway sides of First Hill and Seattle U (albeit perhaps not as fast as biking, running, or scootering) from Capitol Hill Station. Unfortunately, it lacks enough fleet to be as frequent as Link during peak hours, and have a streetcar there waiting for each departing southbound trainload wanting to transfer to a streetcar to get the last mile to these major employers. It can do it off-peak, if the timing can be made reliable. Since the September service change, the streetcar now runs every 15 minutes mid-day on weekdays and all day until 11 pm on Saturdays. During peak, it runs every 12 minutes, which maxes out the fleet of five streetcars. For the duration of the pandemic, and the accompanying 15-minute off-peak headway on Link, there can be a streetcar waiting for each off-peak southbound trainload, including through the evening and on Sundays, if SDOT budgets the service hours to do so, or if First Hill employers provide gap funding to top up the service hours.

For as long as Link’s peak headway is 8 minutes, the streetcar could serve every other southbound trainload. Filling in the gap could be a “streetcar” bus alternating with the streetcar. Since the bus’s routing is more flexible, it could turn at E Jefferson St to serve the Cherry Hill Swedish campus, and provide two-way service to get graveyard shift workers back to Capitol Hill Station. RossB has graciously provided a map of how that might look.

It bears mention that diesel buses stuck in I-5 traffic have a higher carbon footprint than transferring to Link at Northgate or Roosevelt Station, taking the mostly wind-powered train to Capitol Hill, and then jumping on the electrified streetcar. The express buses then double that carbon footprint by deadheading back.

Harborview corridor

Route 60 provides the north-south path through the west side of First Hill, serving Harborview Hospital, Swedish Medical Center’s First Hill campus, Virginia Mason Hospital, and various other medical facilities. The route was the luck recipient of a weekday frequency bump in September, so it now runs every 15 minutes or better from 6 am to 7 pm on weekdays. That takes care of the weekday off-peak rider connections to and from Capitol Hill Station. Evenings and weekends, when the bus runs every half hour or worse, not so much. Bumping evening and weekend headway to 15 minutes, and restoring span of service, might provide for the needs of more First Hill employees than the express routes do. Given the length of route 60, having a short-run version of the route just between CHS and Harborview might be implementable at a fraction of the cost.

Station congestion

Picking up passengers at Northgate Station on routes 302 and 303 to head to First Hill, and dropping them off again in the afternoon, will take up precious bus bay space and time that might have otherwise allowed more Snohomish County commuters to transfer at Northgate. The extra Community Transit buses that have to head downtown have a carbon footprint, too.

67 Replies to “Last-mile alternatives to legacy First Hill express service”

  1. I agree. What is true of these routes to First Hill is also true of the express routes to South Lake Union. There are connecting bus routes from Link stations, and often the distance is short enough to walk. Running these buses is just not a good value. It is the type of thing you do with a robust system, when service is saturated (i. e. buses running frequently on every major corridor). That clearly isn’t the case. This is a bad idea, and Metro should put the money into service:

  2. I recall suggesting the idea of using buses to supplement the First Hill Streetcar during peak hours as a way to boost frequency without needing to actually buy more streetcars (and downtown real estate to store them).

    It’s certainly more efficient than Northgate->First Hill express buses, but it requires people at Metro to think outside the box and acknowledge that it’s the route and frequency of the First Hill Streetcar that really matters, not the steel wheels and overhead wire.

    In an ideal world, the N/S transit route through First Hill would go straight down Broadway/12th and not take the 14th Ave. detour. In the real world, if the streetcars are taking the detour, the streetcar buses probably should too, just to maintain consistency and avoid rider confusion, while also maintain consistent frequency on all the segments.

    I would also like to use this comment to point out that the $150 million spent on constructing the First Hill Streetcar tracks would have been extremely useful right now to keep all-day Link frequency at a higher level. The exact same service through First Hill could have been provided at far less cost (and saving many cyclists a trip to the hospital) by simply making it a bus route.

    1. We don’t need the buses to follow the exact same route as the streetcar. Only part of the route needs to be doubled up, for the reason you mentioned. The button hook is terrible, and hurts ridership. Those that depend on it can still take the streetcar. Everyone else can take the streetcar or bus — whatever comes first. The only area that needs doubling up is Broadway.

      I like Brent’s ideas because they are small — they work within the system, rather than do a big overhaul. I’m in favor of major changes in the area, but it is unrealistic to expect that. On the other hand, a small route like so: would be just the ticket. That uses existing turnaround/layover spots (it could also turn-around at Republican). This is a short route with plenty of potential riders. At worse it would run only during rush hour (which is all these express routes provide). That means it could easily be timed to run opposite the streetcar southbound in the morning, and opposite the streetcar northbound in the evening. At the same time, it would run bidirectionally, since there are plenty of people headed every direction, and even if it is untimed, it complements the streetcar.

      That is a short route. The round trip (with full service) is likely cheaper than any one of these express buses. It costs a lot to deadhead. Traffic in the reverse direction is often really bad, and there are no HOV lanes. As a result, Metro has to allot a huge amount of extra time in case the bus is late (along with the break time for the driver). Thus a short round trip in-service is similar to a long round trip involving a deadhead. If it ran every 12 minutes from 7:00 to 9:00 AM, that’s 10 round trips. Add another 10 for 4:00 PM to 6:00 PM. Throw in 4 more to round out the edges. The 64 is expected to make 24 trips. Thus for the about the cost of just one of the express buses — the cheapest, least frequent one — you could have very good bidirectional service on First Hill.

      1. Thanks for providing the map. It points out one more major employer if the new bus route continues just a few more blocks in a straight line on Jefferson: Garfield High School. Heck, it might even serve students living in the increasingly residentially dense parts of First Hill, and for those reasons work quite well as a peak-only service.

    2. asdf2, we might indeed want to investigate how much that streetcar track and trolleywire could bring on the scrap metal market. But RossB, no chance we can just use these buses, or not, on a “whatever works” basis?

      For a “culture”, which I learned at my seventh grade science fair is what you call virus-food in a petri dish, COVID-19 far as I’m concerned can eat the politico-transit past in good health.

      Honest, guys, you both do so well when you’re operating in the grammatical tense called “Future.” Work in the direction that makes you strong. You’ve got so many really solid tools at hand.

      Mark Dublin

  3. Given the length of route 60, having a short-run version of the route just between CHS and Harborview might be implementable at a fraction of the cost.

    I would probably run it between Beacon Hill Station (where there is an existing layover space) and the current northern terminus. That way you still have some of the connections with the 60 (like the 7, 14, 27, 36 and 107). That would also increase frequency along the north part of Beacon Hill (14th). That section is in good shape, but could be better.

    1. I thought long and hard about suggesting what you suggest here, not so much for all the bus connections, but more to serve passengers transferring from Tacoma Dome / Federal Way / Angle Lake Link to get to First Hill faster via shorter headway.

      Maybe splitting route 60 at Beacon Hill Station, or a variation on that theme, is the real solution. As a frequent rider of the southern portion of route 60, I would totally approve of such a split.

      1. I thought long and hard about suggesting what you suggest here, not so much for all the bus connections, but more to serve passengers transferring from Tacoma Dome / Federal Way / Angle Lake Link to get to First Hill faster via shorter headway.

        I thought about that as well. I don’t think there is a strong argument for that though. If you are headed towards the north end of First Hill, then it is probably better to just round the horn on Link. If you are headed to Harborview, it is about the same to get off at Pioneer Square and take the 3/4. I suppose this combination could be more frequent than the 3/4, but that sounds like a flaw in our system (the 3/4 should be very frequent).

        In general I like the other idea better. It is shorter, straighter, and might survive a restructure. It makes up for the weakness of the streetcar (i. e. you have a limited number of special vehicle, and only part of the route is good). As a peak service addition, it looks like a good value.

        The 60, on the other hand, is a mess. No bus should make all those turns. We should have an all-day bus on Boren. That can go to Beacon Hill or Mount Baker. The 49 should go straight as well, to one of those places. That means the squiggly part of the 60 gets replaced with one of those sections (with service to South Lake Union or the UW).

      2. I want that Boren straight shot, too.

        But I don’t think removing the front-door service at Harborview will happen, or should happen.

        Nor does the Boren route do much good for northern Link to Virginia Mason, or add value to the Swedish connection (which is actually done best by the streetcar, but route 60 shaves a couple blocks for those on the west side of the sprawling campus).

        The Boren route does many things, but connecting northern Link riders to First Hill isn’t really one of them. And as you point out, it probably doesn’t compete well with routes 3/4 for Link riders coming from the south or east. If it did, route 9 probably wouldn’t be in a coma.

        If we had money to throw at the problem, leaving routes 36 and 60 as are, and extending route 49 down Broadway, and on to Beacon Hill Station, would be the politically sale-able approach to getting the north-south straight shot done. Of course, we don’t have that money now, and are struggling to figure out how to fund just keeping what we have.

        I agree with you that small ball is the game right now.

      3. I don’t think removing the front-door service at Harborview will happen, or should happen.

        The 3/4 serves the front door. It also connects to Link.

        The 60 makes a terrible route diversion ( Those that just want to keep going straight have to wait for the bus to twist and turn its way back. It gets away with it, because it is the only bus in the area.

        Running a bus on Boren *and* a bus on Broadway changes that. You create a grid, with buses running frequently and straight on the major corridors. For First Hill that means:

        Pike/Pine, Union (Revised 2)
        Madison (RapidRide G)
        James/9th/Jefferson (3/4)
        Broadway from Roy to Yesler (49)
        Boren from South Lake Union to Yesler (New bus)

        Again, I don’t have a strong feeling as to whether the Boren bus goes to Beacon Hill, or whether the Broadway bus goes to Beacon Hill. But it is essential that we have both (and that they both run frequently).

        This creates a system that works with the overlapping street grid. Every bus connects to Link (and the 49 connects three times). Just about every connection is fast and frequent. For example:

        Link to Virginia Mason — Get off downtown and take the G.
        Link to Swedish — Get off at CHS and take the 49 (or streetcar). Or get off downtown and take the G.

        There is no reason to have buses zig-zagging, making half a dozen turns in a half mile. There should be a frequent bus on Broadway, and a frequent bus on Boren.

        In many ways routes like the 60 are simply outdated. Without a doubt the hospitals around there are the main attraction. But there was a time when they were the only attraction. There was nothing but churches, parking lots and hospitals. Now there are apartments, offices, and just a bunch of building. It isn’t clear where one hospital ends, and another begins. The best way to serve an area like that is with a network that is as grid-like as possible.

      4. My personal opinion is that after First Hill, the 60 should be going to South Lake Union, not Capitol Hill.

        My first choice would be straight-up Boren, but if the front door of Harborview is important, even that could be accommodated without much delay since, when heading to South Lake Union, turning west is actually on the way:

        Sending the 60 to Capitol Hill made sense 20 years ago when Link did not exist and South Lake Union was just a bunch of warehouses. Times have changed, and the time has come for bus routes to change with it.

      5. My personal opinion is that after First Hill, the 60 should be going to South Lake Union, not Capitol Hill.

        For what it’s worth, in Metro’s long range plan, they merge the 49 with the 36. The 60 ends at Beacon Hill. Thus the only bus going north from Beacon Hill on 14th eventually becomes the 49. It would be RapidRide (very frequent).

        A different bus goes from Lower Queen Anne to South Lake Union and then on Boren to MLK (and beyond).

        I wrote something similar when I came up with a plan for the greater Central Area:

        In general it is six of one, half dozen of the other. From a network standpoint, I don’t see much difference in swapping those destinations (although that would mean an additional turn for both). I think one advantage of the proposed pair is that 36 and 49 both run on wire. It wouldn’t take much to link them up (they could probably run off wire between there).

        Both Metro and I have a bus that runs on 9th (I call mine the new 60). It is clearly a coverage bus. I’m not a big fan of the route. It makes sense to have a coverage bus for Leschi (the east part of the 27). But a coverage route a couple blocks away from a frequent (and faster) route is a prescription for low ridership and eventual removal.

        Which really gets at the heart of things. Imagine Metro does both, and allocates the same amount on both routes. The route on 9th makes several extra turns, and thus runs less frequently. It is closer to the hospitals, but farther away from a lot of other destinations. The bus on Boren would just get a lot more riders. It is faster, more frequent, and you can get to just as many places (just slightly different ones).

        The bus on 9th becomes like a lot of things in our system — a “nice to have”. After we have good frequency on the main corridors, it would be nice to have that bus. The same can be said for the express buses that are the subject of this post. After we have really good frequency on all the main corridors — both north-south AND east-west — then express buses like these are a good idea. We are nowhere near that, though. In fact, we are moving in the wrong direction.

  4. Refresh my memory. How much signal priority does the South Lake Union Streetcar get? Any chance that could be increased? And could motor traffic more effectively kept out of the streetcars’ way?

    I also seem to remember that, since the dual-wire does not follow the eastward diversion of the streetcar line approaching Jackson, existing trolley-bus overhead provides faster ride downtown. Signal pre-empts could help there too.

    And the sixty-foot MAN 2000’s on the Route 9 directly connected the Broadway District with the Rose Street turnback via Rainier Avenue. So I think there’s a good chance we’ve “got this covered.” Am I right?

    Mark Dublin

    1. I don’t have the answer, Mark.

      But consider that many of those intersections have a crucial trade-off: Signal priority for east-west bus routes, including the future RapidRide G Line.

      1. If it’s got wire overhead, with a negative-included or positive-only, let’s treat it like a railroad. If all the signals are run from a single control-board, coordinating them shouldn’t be a problem.

        Especially if every single driver, controller or steering wheel + power-pedal, gets trained to stay on-page. Also, from what I’m reading and noticing, an increasing number of streets aren’t used by automobiles anymore.

        For bus trolleywire north-south, I’d like to see the old Route 9 wire- what is it called now?-duplicated for Broadway to Beacon Avenue to the rest of the Route 36 down to Othello Street Station.

        All we’d have to buy is the extra “wiring pans” either side of Jackson at 12th. The rest is already hanging there for the taking.

        Mark Dublin

    2. Mark, to and from the north, the Broadway overhead leads to several turnaround loops in the U District, the John Street wire (10 and 43), and the East Aloha Street turnback. To and from the south, the former Route 9, before fall 2005, accessed Rainier via South Jackson Street. there are not switches between Broadway and Jefferson for south to east and west to north. The 12th and Jackson switches do not include a grand junction; straight north-south movement is not provided for today. 9th Avenue has overhead with connections to Seneca and Jefferson. The FHSC could have used ETB between Pioneer Square and Aloha but the ST Board wanted a monument. The FHSC overhead complicated the ETB overhead; exceptional special was required; some ETB turns were lost. The cycle track shifted the ETB and FHSC to the same lane.

      1. eddiew: Metro’s Trolleybuses have batteries now. They don’t need to worry about what sections of wire allow turns when they can just drop poles through the section.

      2. “Just drop[ping] poles” means “just raising poles” too. While that’s fine for temporary detours or dewirings for construction it is not acceptable for permanent service.

  5. When I lived in the area, I always felt that a frequent “short turn” 60 would be useful. I would have the southern end of it turn off 8th Ave. on to Yesler and then down to ID station, which is more direct than the streetcar routing and the current 60.

    It’s a shame that SDOT has never seriously looked to acquire a sufficient streetcar fleet to provide frequent service, and even the streetcar expansion is not envisioned as being frequent along the entire route (by frequent meaning Link peak frequency or better).

    1. Unfortunately, the CCC project was supposed to double FHSC service frequency and add vehicles. To get more vehicles (which I’d support) appears to be tied up in the relatively duplicative and expensive set of tracks between ID and Westlake.

      1. But, as the author illustrated, boosting peak frequency on the streetcar route need not require actually buying more streetcars. All that’s needed are buses with the words “First Hill Streetcar” printed on their headsign.

        The CCC has always been a big waste of money and, post COVID, makes even less sense than before, since building it implies starving the basic bus system in order to pay for it.

        The right solution is rubber tired vehicles supplementing the streetcar, although really, if people were just a bit less lazy, they could just walk, and we wouldn’t feel compelled to do anything.

      2. The current plan is to just overlap the two routes on that stretch, and then turn them back, so only the new section gets the frequency boost.

        They want a forest. (Except the mayor and my rep seem to want a clear-cut.) They don’t seem to understand that requires planting more trees.

      3. No. The Murray-Kubly SDOT CCC Streetcar plan did not improve the FHSC frequency (much). Both the SLU and FH lines would have 10-minute headway with five-minute headway provided between 8th Avenue South and Republican Street in SLU. Of course, that plan was severely flawed by the poor reliability of both tails, leading to bunching in the common CCC section. Another flaw was that the FHSC does not provide its promised 10-minute headway due to the longer running times than forecast by SDOT; this was probably due to the deviation to 14th Avenue South and the increased congestion on Broadway due to the cycle track taking capacity and increasing congestion. Another serious flaw of the Murray-Kubly plan was the lack of new service subsidy; it was to be funded by fairy dust or the farebox revenue from the STOPS ridership forecast. The Durkan study found that flaw but has not yet suggested a service subsidy. Durkan has further delayed the CCC due to the fiscal crisis. Waiting for Godot.

    1. Huh? You’d hate doubling the frequency of a bus?

      Or would you hate losing your imaginary First Hill Express route from Bellevue?

  6. Swedish has their own bus system that shuttles staff and patients between their two first hill campuses, as well as to King St Station and Colman Dock. (At least, they have in the recent past. I couldn’t find anything info online, which leads me to believe that it’s either suspended or internal.) This system could fairly easily be improved to connect to Link (and potentially connect to other hospitals if they wanted to combine forces). Add to this the FHS, and routes 9 and 60 from CHS, and routes 3 and 4 from Pioneer Square Station, and I think it’s pretty well covered.
    But a First Hill Station along the new downtown Link tunnel would make the most sense…

    1. Good point. With a little bit of coordination, this could be done without much in the way of new service.

    2. UW also has one between the UWMC, Roosevelt clinic, and Harborview. Might come in handy for those coming in from the north via Northgate/Lynnwood Link and possibly even from the east side on 520 busses. They had to cut it back from every 15 minutes to every 30, but presumably it will eventually be restored to 15.

  7. If I were a route 309 Kenmore to First Hill rider, I wouldn’t want my one seat ride broken up. No rider will jump for joy when told their less frequent one ride will be broken up into more frequent three rides. A transit planner will see it as an improvement, but a rider won’t.

    1. Sure. But you could say the same for people who currently ride the 41. They won’t like this change (they liked the express to downtown). The problem is that it is expensive to keep that route, and a poor value, once Link gets here. We only have so much money. Running very expensive buses like the 309, which don’t carry that many people, means other people don’t get service at all. There are apartments in Kenmore with no bus service within a ten minute walk (and that bus service is infrequent, even during rush hour). Lots of people hate it when their bus is very infrequent. It is a zero-sum game. The more money we put into routes like this, the less we have for routes that save people a huge amount of time, and carry more riders per service hour.

      Why should be build up a route that adds no additional coverage and is a poor ridership value?

    2. Route 309 is a slow service to and from First Hill. It provides better speed to and from SLU. The current proposal deletes Route 309; see routes 322 and 361.

    3. Of course. But mass transit decisions have to be made based on getting the most value for the most people, not privileging a few riders over everyone else.

      1. No, they don’t. No transit agency exclusively seeks to maximize ridership. Ridership is one goal among many.

      2. Generally speaking, agencies balance ridership with coverage. These express routes do neither. They are favoring riders who have other options, while telling some riders that they should buy a car. No transit for Victory Heights, but Kenmore Park and Ride users can get a one-seat ride to South Lake Union (must be nice).

  8. Lots of these First Hill routes serve Swedish Cherry Hill. That puts a bit of a wrench into routes only focusing along Broadway.

    I think that a great connecting bus should run between Capitol Hill Station and Judkins Park Station using Jefferson, then either continue on 23rd and using Walker or McClellan to climb up to Beacon Hill Station or cancel Route 4 east of Harborview in favor of continuing this new route to Mt Baker Station — and put those service hours to this new Capitol Hill/ SE Seattle route.

    1. The couple of routes I mention have already been there for decades, wrench-free. And now connectible with no extra overhead.

      But you’re a absolutely right about the Capitol Hill-Judkins Park line. Light rail fact of life has always been the chance that the line might suddenly need to be run by buses at very short notice.

      Mark Dublin

  9. I have ridden the street car from Pioneer Square to my dentist’s office across from Swedish Hosp., considering my dentist’s building charges $8/hour to park underneath (including the 20% tax). The walk from the Smith Tower to Occidental park where the run starts is pretty quick.

    The street car is very, very slow, and meandering, especially the “hook” as it is referred to above, and does not run very often. Whereas I could drive to my dentist’s office in less than 10 minutes (straight up Washington, then to Yesler to Broadway) the street car takes at least 30 minutes each way, including the wait. The fare I believe was $2.75 each way. So not much savings.

    I could also take Uber, which probably would be quickest and easiest since I would not have to get my car out of the garage and park at the dentist’s office. I imagine that would cost $6 to $8 each way including tip since it is not a peak time.

    For me taking an hour just to take the street car to Capitol Hill and back is not time I can afford at work, so I never took it again.

    I also lecture at Seattle University Law School a couple of times per year. If there were some kind of fairly efficient transit I would consider it since parking can be tough near the university, although now the university provides a faculty parking space.

    Pioneer Square and Capitol Hill are two areas of the city I think should be tied more closely together, and there should be much more efficient transit between the two. Personally I don’t get the street cars (and neither do bicyclists). They don’t have the charm like San Francisco’s trolleys, (in fact they look like a bus), which are mostly tourist attractions, and the routing and frequency make them uncompetitive against driving and Uber/Lyft. If I remember correctly there was a Seattle Times piece in which the reporter was able to jog the trolley lines faster than the trolleys themselves.

    I think it is a mistake for Metro to not understand that when it comes to travel within Seattle’s urban core, especially Pioneer Square to Capitol Hill, its prime competition is Uber/Lyft, especially since if there is more than one rider and the fare is shared Uber/Lyft cost the same or less.

    1. DPT, you neglect the bus options from Smith Tower to Broadway/Jefferson because the routes you want are the 3 or 4 (split after Cherry Hill), take <10 minutes, and are a shorter walk for you on both ends.

      Streetcar is $2.25. Bus is $2.75, or free for scofflaws.

      Uber/Lyft competition is [ot].

    2. Yeah, the streetcar sucks. It is a bad mode, and a terrible route.

      But it does have some value, that should not be ignored. The main thing it provides is some level of service on Broadway, from Yesler to John. But there could be more service along that corridor, which is Brent’s point.

    3. The 3 and 4 are so slow you wouldn’t wish them on your worst enemy. It’s better to build the entire solution around the streetcar and supplemental buses, and leave the 3 and 4 to their current riders.

      (Of course, the 3/4 should be speeded up by rerouting it via Yesler, but Metro rejected that alternative due to status quo advocates arguing for the jail stop, so that option is not available. Other alternatives would be to make the 27and 12 more frequent until the G can take some of the load. But people are coming to the hospitals not from 3rd Avenue but from everywhere, so some could just as easily shift to Capitol Hill station or Intl Dist station if access from there were improved.)

      1. Is the 3/4 really that slow? Everything I’ve read suggests that it is really slow during rush hour (it has to deal with freeway traffic). That is bad — and should be fixed — but the vast majority of riders (and potential riders) don’t ride during rush hour.

      2. They’re always excrutiatingly slow between 3rd Ave and 9th Ave. They’re significantly slower than the 7, 14, 10, 11, or 49. The 2 and 12 are second worst, but their problems are mostly limited to rush hour. I can’t believe there hasn’t been more outrage about it from people living east of 17th where the travel time gets especially long. The 27 is a safety valve and in the eastern part you can walk to the Jackson Street buses. I lived right near Harborview in 2005 and worked at Harborview before that, and it hasn’t gotten any better since then. I used to rejoice when the 27 eastbound was coming soon. Jackson wasn’t much of an option because of the steep hill. The bottlenecks are the turn at 3rd, 7th, and the turn at 9th, and the fact it makes three turns between Jefferson and 3rd instead of one. Turns in other locations aren’t as bad, but there they are.

      3. The bottlenecks are the turn at 3rd, 7th, and the turn at 9th, and the fact it makes three turns between Jefferson and 3rd instead of one.

        This is why I had trouble getting very excited about the proposed change: On the one hand, it seems like a better overall route. It avoids freeway related traffic, while also covering Yesler. On the other hand, it actually has *more* turns. I still think it is a good idea, but it is hard to see it as the best overall solution.

        I think I would just straighten out all the buses. Have the 3/4 just keep going on James all the way to Cherry, and move the bus stops over there. That would mean one turn (3rd and James) and that’s it until 23rd (and half the buses would just keep going straight). You would need to add some BAT lanes or bus lanes close to the freeway, but that doesn’t seem too hard. There is nothing special about James in terms of the freeway — drivers can access the same ramps via Cherry. You basically make it harder to drive on James, and people will move over to Cherry or other streets, while the bus goes in its own lane much of the way.

        Then all you do is run the 27 a lot more often. The tail of the 27 is weak, so you could have a shorter version, or a split (the 3/4 has both). East of 12th, the spacing of Union, Cherry and Yesler is better the current spacing of Union, Jefferson and Yesler (the streets are more evenly spaced).

        Having buses twist and turn just doesn’t make sense to me in the long run. I can see why folks wanted to run the 3/4 on Yesler — it would make for some good connections while avoiding rush hour traffic. But it wouldn’t be especially fast, any time of day.

      4. At one point Metro had a goal of eliminating turns on 3rd Avenue between Stewart Street and Yesler Way. The 7 (49), 14 (47), and 11/125 were split pursuant to that. A 2 split appeared in multiple restructure proposals (RapidRide C/D or E, the 2014 cuts) but was defeated by status quo activists. (The loudest activists were always for the 2, the second-loudest for the 12.) It will probably be done with RapidRide G and the 2’s move to Pine-12th-Union. The 3/4 would move to Yesler.

        The new turns shouldn’t be problematic because the 27 and 60 are turning there now. The 60’s turns aren’t great but they don’t wait as long as the current 3/4 turns. None of the other routes do, except when the 12 turns at 6th in the PM peak. The 2 bogs down on Seneca Street and maybe Spring Street — I don’t remember specifically about the Hubbell Place turns.

        Before the Yesler alternative I’d long thought the 3/4 should remain on James/Cherry rather than turning to Jefferson. Jefferson is such a small street there’s no apparent reason why it should be chosen for a bus. But it’s kitty-corner from Harborview’s main entrance and ER, so that’s probably a rigid requirement. There’s no way to get from Jefferson to 3rd without turning three times.

        Here, hear for making the western part of the 27 frequent. If the 3/4 must remain as-is, then the 27 should be turned into the main route to the southern CD rather than forcing everybody onto the snail-pace 3/4.

  10. Brent, you did not keep up. The current network before the public for comment includes routes 302, 303, and 322 oriented to First Hill; the first pair would serve Northgate Link; Route 322 would serve Roosevelt Link. Routes 63, 64, and 309 would no longer serve First Hill; routes 63 and 309 would be deleted; Route 64 would be revised to serve SLU only. New Route 361 would serve SLU. See:

    All your strategic points are valid: speed, duplication of Link, climate, and network efficiency. The First Hill market is nearly 24 hours; essential workers are better off with two-way all-day service.

    The headway of the streetcar could be improved if it was truncated at 5th Avenue South. Southbound riders oriented o First Hill could straddle the streetcar and Route 60 on Broadway near Denny.

    One could characterize the service design of routes 193, 302, 303, and 322 as transfer adverse; but their loopy pathways on First Hill take minutes for riders oriented to the eastern institutions; riders would be better off using Link and transferring to and from a connecting service. The survey tests this rider preference. Should the buses and hours go toward direct trips or shorter waits?

    The STB has already commented on the weak and cheap ST SIP for 2021. Short Link headways are needed for integration. Before Covid they planned four car trains and six-minute headway for Northgate. ST needs to get past Covid; that crisis will probably end; no one knows when. The 2021 network has to be resilient and last a long time.

    1. Right, there are significant changes. But they are all variations on a theme. A theme that has failed in the past, and is more likely to fail in the future.

      Consider the 64. The old route SLU, then First Hill. It performs poorly. Will splitting it make it perform better? Of course not. It will be faster, but less efficient. If the bus isn’t full with South Lake Union and First Hill riders, why will it be full with just South Lake Union riders?

      The only chance it has is if it poaches riders who would otherwise ride Link to downtown. Before the pandemic, Westlake was the most popular Link station, with twice as many riders as any other downtown station. It stands to reason that a lot of those riders will just catch the 64 all the way downtown, rather than bother with a transfer. Using that logic, we should just keep the 41. It gets way more riders.

  11. Daniel, I hope I haven’t started giving the impression that everything wrong with transit from streetcars to bullet trains does not make me ten times angrier than anybody else on our home Planet.

    True, the only streetcar I ever drove was the 90′ Breda that Vastraffik’s chief instructor ordered me to take through Downtown Gothenburg at eleven in the morning, my trackway filled with pedestrians, pets, bikes, baby carriages and everything else in Sweden. His opinion of the car’s supplier, same as mine.

    Might be better Topicality to put my thoughts about my former employers’ still-unvarying attitude toward things like signalling, vehicle coordination, and operator training into a posting. [OB]scenity might be lesser offense than that tiresome [OT].

    And all my guilty [HOMINEMS] are either retired or begging on their knees they live to be. But no, Lake Washington Community college pays no employee head-tax for me. Just guilty about the mess I’m leaving its freshmen. Both genders.

    And I certainly would not mind a declaration from Mercer Island’s business community how fed up they are about the foot-dragging that’s preventing Mark Dublin from cementing the Streetcar Sister City relationship that’ll put every FHS driver aboard Icelandair to learn make a headway mean what it says.

    Age thing, but can’t help you with the Ballard Runaway Problem. However, while your folks are filing the police report to get you back (like SPD even answers the phone!) Nordic National Museum owes me a favor like getting you EU Asylum.

    Your sheer effort’s exemplary. Your Island Heritage can be proud of you.

    Mark Dublin

  12. People are not going to bus->Link->bus for a commute; the transfer penalty is far too high.

    This sounds like a great way to drive the few riders on these expresses into cars. Which is fine if you don’t want to serve them, but let’s call a spade a spade.

    1. Thank you for calling a few “the few”.

      Those who don’t have a latte shift will have to do that bus-to-Link-to-bus trip, anyway, unless they take Uber/Lyft, bike, or park at the Northgate garage.

      The exorbitant investment in these peak-direction-only expresses means there is less money available for the all-day grid. As Mr. Lawson described in his analysis, the all-day grid took a big hit in the Phase 3 proposal.

    2. It depends on the trip, the person, and the quality of the two bus routes. Before Covid I had several ways to get to work, including Link+bus+bus. The total travel time was the same as taking two buses via one route or two buses via another. My decision was based on the frequency and reliability of the bus routes, and how much walking at each end. Link was more frequent then so that was never the issue. Going to work the Link+bus+bus transfers were especially bad, I’m more time-sensitive, and the walk at the end was more objectionable. Coming back from work, I felt more willing to do a 5-minute walk between bus+Link and a walk at the end (which is downhill and has two nice sidewalk forests).

      If I were going from 522-land to First Hill, my criteria would be the same. Plus there’d be the added factor of avoiding freeway bus congestion. Transferring to Link at Roosevelt avoids any traffic jams around the Ship Canal bridge or South Lake Union. The worst part is transferring to the 3/4 downtown. But if I can transfer at Capitol Hill Station to the streetcar or a Broadway bus, that’s much better and I’d be more willing to. I could even walk if I’m going to Swedish or Virginia Mason or don’t mind a walk to Harborview. Staff are more able to walk than disabled patients, and rush-hour trips are mainly staff.

    3. Three seat rides are common in our system. I would bet that most of our commuter rail riders drive or take a bus to the train in the morning. Once they get off the train, a lot of them get on Link or a bus (to get to the other side of downtown). This is true of our commuter rail, and a lot of commuter rail systems around the world (unless, of course, the commuter rail system has more downtown stops).

      If people simply refuse to take a three-seat ride, then we have no hope for riders the rest of the day. This runs counter to previous truncations. Lots of people in northeast Seattle have been living with a three seat ride to First Hill, Queen Anne and South Lake Union for a while now. Ridership increased during that period.

      So far as I know, there is very little evidence about the transfer penalty (i. e. how it effects ridership). Jarrett Walker asked about it, but in a little more specific manner. He was interested in the quality of the transfer ( In the comments you can see how Toronto and New York (the two biggest transit agencies in Anglo North America) handle it. In both cases they just consider it a slower vehicle (based on the assumption that riders won’t accept a really long commute). Here are some numbers, based on their approach:

      1) Walking time to Link platform: 5 minutes
      2) Waiting time at Link platform: 3 minutes
      3) Waiting time at bus stop: 3 minutes (this assumes an improvement as Brent suggested. If not, then assume 6 minutes.)

      Using MTA’s approach, that means we multiply that time by 1.75, which works out to about 20 minutes. That assumes that the bus would get you there 11 minutes sooner. The train is faster (not counting the waiting) so let’s subtract a bit from that. So, using MTA methodology, the three-seat ride has a roughly 15 minute penalty, even though it isn’t that much slower (it might even get the rider there faster).

      The Toronto approach is simpler. They add a ten minute penalty for the transfer. There is also an additional penalty for waiting, but that gets complicated (it might take longer to wait for the express bus than the shuttle one that gets you to Link). Anecdotally, I know more than one person who refuses to wait for the express bus, even though the express bus will get them to work faster. They take the more frequent bus (followed by the more frequent train).

      But I think considering this a 20 minute penalty is quite reasonable. Would you have riders using this if it took 20 minutes longer than it does without traffic? I think the answer is absolutely. There might be some who find another way there, but this is a very reasonable way to get from one place to another. This express only runs during rush hour — a time when a 20 minute penalty sounds like nothing. Driving may encounter a similar penalty on a regular basis.

      To be clear, there may be some who give up. So what? A lot of people have given up on a system that has trips like this: This is not an obscure or unusual trip. Two major, transit rich areas, with lots of people and businesses. This is about a 10 minute drive, yet it takes over 45 minutes to get from one place to another, not counting the initial wait time (oh, and it is a three-seat ride as well). The fastest travel time involves a 20 minute walk. Wonderful.

      Of course we will lose a few riders if we eliminate these express buses. But we will lose *more* if we don’t. It is a zero sum game, and if we put extra money into these routes, then we take it out of other parts of the system. That reduces ridership too — a lot more. We know these routes perform poorly (and that was before there was the alternative of using Link). We should put the money into routes that will increase ridership, and save people more time, like the proposed 61.

      1. The transportation seminars I have attended suggest commuters won’t take three forms of transit to get to work, and often forgotten is the first form of “transit” is from door step to “second” form of transit. First/last mile access is the most critical part of the journey.

        This probably is more applicable to commuters, and those who have a car or other form of transportation to get where they going. These commuters rarely take transit except to get to work, because of congestion or the price of parking at work. Although there are many comments on this blog about ridership during non-peak times in the Seattle area, few eastsiders are going to ride transit during non-peak times. They are not on transit because they want to be.

        The reason this is important on the eastside is most of our first/last mile access is by car to a park and ride, which eastsiders like. It allows them to use their car after work to run errands, and depending on parking space drop kids off at school on the way to the park and ride, or pick them up from daycare on the way home, they don’t have to walk and deal with the elements, and it is perceived as safe. The biggest problem is lack of park and ride space, so the stalls fill very early, which is why eastside cities are demanding larger park and rides.

        This means this eastside commuter is already in their car — an alternative form of transportation other than transit — in the first form of transit, and so transit better be more convenient to get them out of that car, or they will continue in their car to work.

        Bus intercepts, or truncation, work much better if that bus is the first form of transit, from door step to a rail line, or even work (although as noted milk runs are frustrating). This works better in a dense environment like Seattle where there are more feeder buses, and the buses can be walked to. Since most eastsiders have to drive to a park and ride that option is not available. The drive to the park and ride is the first form of transit for them, and they generally won’t take a third.

        ST would like to go to a bus intercept or bus truncation system on the eastside, probably to goose ridership numbers on East Link, and according to Metro to save $1.5 million/year although that is nothing compared to the eastside ST subarea reserves, which will still require huge park and rides. Right now many of those who will not be served by East Link drive to a park and ride, and then catch an express bus to either downtown Seattle or Bellevue, but mostly Seattle. They don’t take a bus to Mercer Island.

        The irony is East Link should have followed the huge park and rides on the eastside, because now ST is having to build new park and rides, and until there is rail existing huge park and rides in Issaquah, Eastgate and Renton that will only serve “feeder buses” won’t be used.

        It is unlikely they will drive to a park and ride to catch a bus to go to a rail station to catch the train to Seattle. I just don’t see it. And that is assuming adequate capacity on East Link without a second transit tunnel.

        So, since they are already in their car they will drive past the park and ride for the bus and head directly to a park and ride next to a rail station. Since S. Bellevue’s park and ride holds 1500 cars that would be first choice (despite 405 during rush hour which WSDOT claims it will alleviate with its new $800 million design) which is not what Bellevue wants but cannot prevent, or they will drive directly to work which may be more convenient if working from home reduces traffic congestion during peak hours. Many of these commuters could afford the parking costs (and write it off).

        Since most eastsiders have cars, transit has to compete more than for those who don’t have cars. For eastsiders, ST 2 and 3 meant huge park and rides and a TRAIN. Just conceptually, driving to a large park and ride in your own car, parking, and getting on a TRAIN that goes to Redmond (and avoids the nightmare of 520) or downtown Bellevue (kind of, 112th) or Seattle, is more convenient during rush hour than driving, assuming congestion.

        It will be interesting to see how things turn out. I think congestion will permanently decline after Covid-19 due to working from home (and assuming bus and train riders are comfortable with the perception of risk after Covid-19 so buses and trains can go back to full capacity), especially on I-90 which has good capacity under R8A, and I think commuters in cities like Issaquah, Snoqualmie, Renton and Eastgate will object to taking a bus to a rail station to get to Seattle, although the problem there is the buses won’t access the rail station, and that badly hurt ridership on the 550 pre-Covid.

      2. The eastside used to have three First Hill Express buses: route 202 from Mercer Island, route 211 from Eastgate (IIRC), and route 265 from Redmond. They did not get good ridership, so they were axed long before the pandemic.

        Mercer Island got route 630, which added the Park & Ride as a stop, as a replacement for nearly-useless route 202. Route 630 did not survive the September massacre.

      3. > But I think considering this a 20 minute penalty is quite reasonable. Would you have riders using this if it took 20 minutes longer than it does without traffic?

        I disagree. I really do not think this is reasonable; traffic is bad, but not so bad that people are just going to take a 20 minute hit to their commute each way on the chin.

        > Of course we will lose a few riders if we eliminate these express buses. But we will lose *more* if we don’t. It is a zero sum game, and if we put extra money into these routes, then we take it out of other parts of the system.

        Then call it what it is, and don’t try to parade around a service cut like it’s supposed to be adequate for these riders.

      4. Also, I highly doubt that this would even be remotely as fast as you’re suggesting. Maybe if the stars align and there’s an eclipse, but generally speaking the buses here are so unreliable, and the general transfers so bad (usually requiring crossing one or more poorly timed crosswalk signals) that you’d more often than not exceed times on at least one of these trips, and one bad transfer is enough to screw up the entire trip.

        This is why more than one transfer is generally not palatable; because the amount of chances to screw up a “good” commute rise exponentially.

      5. I will give myself as an example and say that I avoid three-bus (or bus-Link-bus) transfers like the plague. Even in the Before Times, I never did them, and ended up moving rather than have to face one every day once Northgate Link opened up and the NE restructure took place.

        The 20 minute delay sounds about accurate to me. However, the issue is not that you get a 20 minute delay over driving; it’s that you get a 20 minute delay over whatever delay you get from riding transit over driving, and that initial delay is already potentially very long. So it can go from doubling your commute to almost tripling it (the commute I left behind when moving as mentioned could have been about 25 minutes by car; my two-bus thing was around 55 minutes; adding another transfer would have pushed it to about 70 because of the less direct route + extra transfer wait). At that point it became not worth it, so I voted with my feet. I’m not special and don’t expect transit to be geared to my needs, but I also think that it is not uncommon for people who are in a similar situation as I was to do something similar to what I did.

      6. “don’t try to parade around a service cut like it’s supposed to be adequate for these riders.”

        It’s not a “cut” when you shift hours from one route to another. Transit agencies need to look at the good of all the residents in the service area, not just those in a particular corridor. Sometimes that means reducing a route to achieve a better balance. It may feel like a cut to those riders, but they’re not the only ones paying for the network or needing to get around.

        Whether transit is more time-consuming than driving depends on the trip. Using Link avoids traffic jams that make both bus trips and car trips take longer. Sometimes the traffic is every day, like Northgate to downtown between 2pm and 7pm. Other times it’s unexpected spikes, which can happen once a week or once a month or more.

      7. The transportation seminars I have attended suggest commuters won’t take three forms of transit to get to work, and often forgotten is the first form of “transit” is from door step to “second” form of transit. First/last mile access is the most critical part of the journey.

        Yet I provided a counter example in the very first paragraph of the comment you just replied to. I’m happy you went to a seminar, but that doesn’t make you an expert on transit. You wrote a dozen paragraphs and you failed to provide one shred of evidence to support your claim. In contrast, I provided several, and your only rebuttal to the approach taken by large North American transit agencies is that you went to a seminar once and someone said something different. We don’t know exactly what they said, or what their qualifications were, or what studies they cited (if any) but that’s what you remember.

        You seem to be using this discussion of additional service to First Hill as just as an excuse to profess your love of park and ride lots, and how the East Side rider (of which their is only one, apparently) will not ride transit unless they can find a place to park. Again, no evidence, pure conjecture.

        In contrast, we know that there is a very strong correlation between frequency and ridership. The less frequent the transit, the fewer people ride. (There are several studies listed here: and a simple internet search will yield a lot more).

        To be clear, I realize transfers can cut into ridership (all other things being equal). But it is overly simplistic to assume that no one will take a three seat ride to work when obviously they do. Focusing on one type of rider (someone who hates transfers) will likely result in fewer riders than if we spent the money on frequency (or other improvements).

      8. it’s that you get a 20 minute delay over whatever delay you get from riding transit over driving, and that initial delay is already potentially very long.

        It may also be nonexistent. Again, we are talking about rush-hour service here. From Kenmore to the freeway there are lots of bus stops, but there are also lots of HOV lanes. On the freeway the cars crawl along getting to First Hill. Quite often, you would spend less time on the bus/train than you would in your car. The reason transit takes a while is because of the waiting (and walking to the platform, etc.). But that was factored into the 20 minutes. In other words, door to door taking transit could be 8 minutes slower, but it feels like 20.

        I really do not think this is reasonable; traffic is bad, but not so bad that people are just going to take a 20 minute hit to their commute each way on the chin.

        You are saying that if a commuter finds that his transit trip takes 20 minutes longer than driving, they will always drive? Seriously? Have you ever been on the 40? It is backed to the gills with people from Ballard, making their way slowly to South Lake Union and downtown. Of course driving would save you 20 minutes. But people don’t want to mess with driving. It is a hassle. Parking costs money. It adds stress.

        In this case we aren’t talking an actual 20 minutes. It is quite possible that someone would get to work in the same time it would take them to drive. It is just a different kind of hassle. Instead of sitting on one bus, you transfer to a train. Then, instead of walking a short distance to your work, you walk farther (or you take another bus). That is a big hassle — it isn’t ideal. But it is extremely common in our system now, and will be common in the future.

        Since we are sharing personal transit stories, I’ll share mine. I used to commute from my home in north Seattle to Factoria. It was two buses, and it sucked. Frequency was terrible — it took over an hour. Eventually I drove, but that sucked too. It took forever, but was still better than transit. I would have loved to have a three seat ride, as long as each piece was reasonably frequent. I would take the bus to Northgate, then the train to South Bellevue, then the bus to my work. Again, that would have been wonderful, as long as the bus (and the train) was reasonably frequent.

        Consider the other alternative. We run buses from Northgate to Factoria. We run buses from Juanita to Lower Queen Anne. Phinney Ridge to Renton. These are all places with a fair number of people, and a fair number of jobs. The problem is, if we run all these buses, only a handful of people actually benefit. Those trying to get from Phinney Ridge to Factoria or out of luck. Not only do they still have a three seat ride, but that three seat ride still sucks. We don’t have enough money to make the buses frequent.

        That is what is so terrible about this. This does not solve the three-seat problem. It actually makes it worse. Oh, it solves it for a handful of riders going to a handful of places, but everyone else is worse off.

        This is a system that favors a certain type of existing riders over everyone else. This only runs during rush hour. So if you work the night shift, you are out of luck. Not only do you not get this express to work, but your commute is worse. This is a bad use of resources.

    4. Henry, it all gets to minutes. You can be sure that some riders transfer twice today. If the middle trip is Link, it will very reliable and fast and hopefully have a short wait. if connecting services are very frequent, the waits are short. intending riders will have to compare the cost of waiting for an infrequent one-seat ride that has an indirect pathway on First Hill v. using Link and transferring. Routes 302, 303, and 322 are or would be snakes.

  13. There is precedent for supplementing the streetcar with a bus. The South Lake Union streetcar couldn’t handle the loads to SLU offices, so Metro rerouted the C there and boosted the 40, 62 and 70 to supplement it. So boosting the 60 or at least its northern half would be similar. There’s also the 9 (although currently suspended).

    1. Was the South Lake Union Streetcar ever crushloaded other than right before or after a firewords show?

    2. Brent, Route 205 served First Hill and the U District; Route 202 served downtown Seattle; both were deleted in fall 2014.

    3. Mike, It was the STBD that decided to split lines C and D and paid for it. It was SDOT that was clever and provided the C Line and Route 40 priority across Mercer Street and along Westlake Avenue North. Really, the streetcar is the supplement to routes 70, 40, 62, and the C Line; the bus routes penetrate downtown Seattle; the SLU line kisses the edge a long block north Westlake Station.; the Nickels SLU line is 1.3 miles long. The STBD improved the midday frequency of Route 60.

Comments are closed.