Map of south sections of Metro routes 3 & 4
Map of south sections of Metro routes 3 & 4

About ten months go, Zach wrote a post asking whether the southern tail of Metro’s Route 4 is redundant in the context of Metro’s current bus network:

I have a distinct impression that it no longer serves any unique transit market and in fact diminishes the performance of Routes 3, 8, and 48, all of which serve unique destinations.  From 3rd/James to 23rd/Jefferson, the shared 3/4 provide 7-15 minute headways until 1am.  Once the 4 turns south on 23rd, it duplicates the 48.  From its turn at Dearborn it runs in a couplet on 24th and 26th, needlessly threading the needle between 23rd (Route 48) and MLK (Route 8).

In this post I’ll drill down into the stop-level data to answer a complementary question: Are people using the unique part of Route 4 to Judkins park and beyond? I’ll also examine stop-level data for Route 3, and suggest an inexpensive capital modification to the trolleybus network that could dramatically improve the reliability and almost double the capacity of these workhorse routes.

Route 4

Route 4 Stop Level Data
Route 4 Stop Level Data

For those not familiar with this chart format, head on over to my post about Route 36. Here’s what I see in this chart:

  • Crush loads from downtown to Harborview in the AM peak. The average load tops out at 45, just off the chart. Keeping in mind that a Gillig trolley nominally seats 42, this means every coach is full as it heads up James. Presumably this overcrowding is already dissuading additional choice riders. This is another point in favor of the First Hill Streetcar: even though it’ll be slower, it’ll be a much more comfortable way for suburban commuters to get from the Downtown Transit Tunnel to First Hill, especially if Metro is unable to add capacity to this route in the near future. More on this below.
  • Very little activity after Jefferson & 23rd, except at the two stops on 23rd at Yesler and Jackson. These two stops, in addition to frequent north-south service from the 48, have faster one-seat rides to downtown with similar headways from the 14 and 27 respectively. South of these two stops, loads are light in all time periods; boarding activity and loads increase briskly as the bus heads down Jefferson. (I’m suspicious of the apparently large number of boardings at the terminal stops on Walker; even if those numbers are true, those people have many better bus options to almost anywhere).
  • Very little use of the Lighthouse for the Blind stop on Plum & 25th. While providing the ability for people who cannot drive to live and work independently is a vital function of transit, the stop here is so thinly patronized — about 15 boardings and four deboardings per day —  as to suggest that, in the context of the minimal use and almost complete redundancy of the long tail of this route, finding alternative ways to serve this facility are in order.

I think Metro’s riders have answered Zach’s question: the long tail is indeed redundant. There is little to be lost and much to be saved by simply terminating the 4 at 23rd & Jefferson, making those trips turnback routings of the 3.

Route 3

Route 3 Stop Level Data
Route 3 Stop Level Data

The data for the 3 confirm several trends observed in the data for the 4. Harborview ridership dwarfs everything else; Jefferson performs well. Ridership on Cherry is weak, making a decent showing only in the peak, but, unlike the 4, it does fill what would otherwise be a mobility gap in Metro’s network. Late night ridership seems to be almost nonexistent past 23rd Ave, and it may make sense to trim back service there, although it wouldn’t save much.

Switching from James to Yesler

Comparing Yesler to James
Comparing Yesler to James

This chart is similar to the one in my post about Route 16, and you should refer to that post for details on how to read it. The underlying information is time point data for routes 3, 4 and 27. It’s evident that from Broadway to 3rd Ave, the 27 is much faster and much more reliable, consistently clocking three minutes during the day and two minutes in the evening. By contrast, the 3 and 4 averages swing from 6 to 10 minutes, with much larger error bars associated with each run, indicating very unpredictable travel times even at the same time of day. [UPDATE: To be clear, I’m proposing moving only the segment that runs on James down to Yesler. The route on Jefferson would be unchanged. Metro’s timepoint data just happens to run from Broadway.]

Of course, this isn’t quite a fair comparison as the 3 and 4 have to make two additional turns, and board and deboard far more riders than the 27. Even so, it’s clear that moving the 3 and 4 to Yesler would make them faster and more reliable. Keep in mind that this doesn’t just improve service for the people who make this trip — the 3/4 continue to Queen Anne, and lateness on arriving at 3rd Ave is likely to translate in the lateness all the way to the north terminal.

There are additional points for this change:

  • If the trolleybus replacement program at Metro recommends the correct mix of 60′ and 40′ trolleys, and sufficient additional layover space were found, the peak capacity of this corridor could be doubled at minimal additional operating expense by switching to 60′ trolleys. Routes that use James and other streets with severe hill breaks (such as Seneca and Madison) are constrained to use 40′ coaches, a problem which does not exist with Yesler.
  • This move removes one of two lane-blocking left turns that buses have to make on 3rd Ave southbound (the other being the 2 onto Spring). The turn at Yesler has its own lane. Congestion on 3rd southbound was identified as the biggest problem with surface street reliability once the RFA is eliminated.

There are also some downsides:

  • The cost is relatively high for a trolleybus project — roughly $12 million according to people I spoke to at Metro. This is because the Dilling Bridge and the bridge over I-5 have some technical issues that make the trolley wire supports difficult to install.
  • Yesler has no ADA access to or from surrounding streets between 3rd & 8th 6th, only two narrow staircases. Riders currently boarding at 5th & James will have to walk, typically less than five minutes, most likely to the stop on 3rd.
  • Riders at 8th & James will have to walk up the steep hill to 9th & Jefferson for the 3/4 or several blocks to Madison for the nearest service on the 12. This creates a small mobility gap in a very steep part of the city that is difficult to address.

On balance, I think the benefits far outweigh the costs for this change. If the city’s $60 VLF were to pass in the fall, the $20 million allotted to trolleybus improvements would make this very affordable, and while projects such as this aren’t sexy, they do improve the transit system and the experience of the people who use it.

150 Replies to “Is Route 4 Redundant?”

  1. I tend to agree that the 4S may be redundant. But as a counter argument, I will offer these suggestions for improving the 4S. Start the improved 4S at Mt. Baker TC, thread the needle through Judkins Park (as it currently does), but turn onto Jackson St. and head downtown mimicing the 14’s route. This plan would offer additional service on Jackson between Rainier Ave and 23rd Ave where there are lots of housing units, but only 30 minute service from the 14. New wire would have to be installed from the Lighthouse to the MBTC and a switch installed at 23rd/Jackson. BTW, the ridership at Rainier/Walker is real, I pass by that stop regularly. The 3S would continue to serve Harborview to 23rd Ave at 7-8 minute headways with one bus every 30 minutes to Cherry Hill.

    1. Interesting idea, but I don’t see the point. Basically, wherever they’re going, the people on Rainier/Walker are better off on the 7, 8, 9 or 48. I think the current service levels are adequate on Jackson.

  2. I would argue it doesn’t even matter if it’s redundant, because whatever the solution is, it won’t result in any meaningful savings. You’d still have to have the same level of service on this popular workhorse route. So even if you deleted the 4, and doubled the level of service on the 3, which duplicates the 4 for 90% of the line, it would be a zero-sum game. And if you’re talking about it just not completing the last mile of its route to 25th and Walker, there are so many low income and disabled living facilities on that portion of the route, like Center Park, that even if Metro wanted to, they would have a tough time pulling the 4 out of that area.

    1. You’re wrong in every respect. There are considerable savings in run time and cycle time from axing the 4. The 4’s deviation from the 3’s turnback terminus more than doubles the length of its trip south of downtown.

      1. Ok Bruce, let’s pretend Metro and I (an award-winning internationally respected transportation expert) are wrong, and you are right. Why does Metro continue to follow my recommendations on the route 4, and not yours?

      2. Well, Metro was going to axe the 4S if the CRC hadn’t passed, which suggests they are very aware of how unnecessary it is. It will probably be one of the first to be cut if/when the new service guidelines are implemented.

      3. Internationally respected urban planners gave us slum clearance and disastrous “urban renewal.” Internationally respected financial minds gave us speculative trading in derivatives. Internationally respected statesmen gave us Augusto Pinochet.

        I’m not sure when you started claiming to be an internationally respected anything, but if you’re a one-seat defender then you are a dinosaur. Your presumptions are wrong, and you are the reason for anemic American transit everywhere. Full stop. Pens down. Dismissed.

    1. The idea has been kicking around for a while. My purpose here is mostly to examine the data behind it.

    2. Anyone who has ridden the 3/4 knows how terrible it is just getting through First Hill. Yesler would be a huge improvement.

    3. Yesler Terrace is going to be re-developed and it should be possible to lay out a straighter street grid through YT that would make it easier for the 3 to get from Yesler to Harborview. And once the 3 is re-routed onto Yesler, the 27 could be diverted to run north-south on 12th Avenue and provide the service that the neighborhood activists around Seattle U were hoping to get by hijacking the Broadway streetcar line.

      1. The 27 is not redundant. Not ideal, perhaps, but necessary in the context of our bus network.

      2. The sword of Damocles is hanging over the 27 even now. Cutting it to peak-only was in the “low impact” list of Metro cuts, and it’s probably going to happen within the next 5 years. Even though it’s got the best trip times from the CD to Downtown, it’s got the worst ridership. Once the 3/4 is inevitably rerouted through Yesler Terrace, Metro will have even less reason to keep it around.

        The only reason it’s still around at this point is to preserve the only service in Leschi. But hardly anyone in Leschi is riding it, anyway.

        I still advocate making the 27 a Leschi – Mt. Baker TC – Columbia City Station circulator. If it’s going to be kept around simply to fill in some small system gaps, it might as well fill in as many as possible, and leverage Link at the same time.

      3. Cutting the 27 to peak-only was not in any low-impact cut scenario that I’m aware of, and in the 600k cut it was a bottom-priority “cuts on appropriately served corridors” reduction in midday frequency. In fact, it’s identified as an under-served corridor in the 600k document.

        I do agree that moving the 3/4 to Yesler will eat most of the 27’s lunch. Other than the part near Yesler terrace where the hill is a barrier to mobility, it’s really only the unproductive part on Lakeview that matters for mobility. I was told that Metro considered an extension of the 27 south at one point, but there’s an intersection somewhere where they felt that a 40′ coach couldn’t safely make the required turn. I don’t know where that is.

      4. I just double checked, and I did misremember the 600k cuts.

        It’s listed under phase 2 as “convert to peak-only”. Right next to a bunch of common-sense cuts like eliminating the 4’s and 14’s tail.

      5. Also, I’m amused that Metro is concerned about a 40′ coach driving on the sidewalk to make a turn (I’m assuming it’s Lake Washington Blvd & Lake Park Dr, or maybe Spokane & 38th), when they’ve got the 8 running over the sidewalk & through the planting strip 4 times an hour with 60-footers at Madison & MLK.

      6. This proposed reroute has especially one big drawback. After it/they (3 and 4) turns on 3rd it would have to compete for the already very full stops and could not serve the County and City offices as it does now. This route also provides access for others from downtown to these offices. Believe or not there are citizens who still attend hearing and meetings and testify. These stops are important rides up a hill that is difficult for many. It is a longer route. I feel that it is only speculation that this would save time, and it definitely would not save the walk up the hill, which is impossible for many.

      7. All statistical evidence and all anecdotal evidence proves — not “suggests,” but proves; I know this is Seattle, but sometimes facts are important that the six-block James Street clustercluck adds 5-10 minutes to every trip.

        If you’re part of the 0.001% of the population that is incapable of walking uphill for 1 or 2 blocks and too stupid to access the system of elevators available within every affected block, then waste your own time riding to Pine, traversing the flat surface to 5th, and riding back down.

        You don’t have an inalienable right to waste the other 99.999%’s time too!

  3. The 3 and 4 are good because they go through neighborhoods, not on a fast car-oriented road. As someone who lived on Jefferson, putting the route on Yesler would suck. It is easily a 10 minute walk. I only very rarely chose to use the 27 and only on Sundays when I needed to get downtown and the 3/4 were off schedule. Even then I often just walked downtown – faster than walking to the 27, waiting and then riding downtown. Psychologically, you have to give yourself more time (even if you know how long the walk takes) for a bus stop that is further from home. Moreover, considering the next east-west bus is Union, that’s a substantial walk to get to a downtown bus for someone in the middle. It is about 0.8 miles between Union and Yesler meaning anyone in the middle has about a mile walk they didn’t have before. Having fewer distinct east-west bus routes (on different streets) makes it less convenient for neighborhood users – consolidation of routes on the same roads is good for commuters, but not nearly so much for people living in a neighborhood.

    I also suspect the tail of the 4 (esp inbound) are real riders (“suspicious” is a somewhat strange word choice.) What bus do you think it is redundant with? The 8 and 48 don’t duplicate with it for more than 10 blocks. The 9x is useless off peak (duh) and only intersects at Harborview. The 42 is, uh, useless. The 7 does go downtown which is a significant duplication – but at easily a 10+ minute time penalty. After we moved to I-90 and Rainier area, we once took the 4 to Judkins Park and walked. It was almost faster for us despite the extra walk to Rainier — convenient enough that we’ve taken it whenever the 7 is more than 5 minutes out and we’re at a downtown stop for both routes. For someone living in Judkins Park, the time difference is probably more significant because the walk to/from the 7 stop is obligatory rather than a nice stroll home through a park. 

    1. Though I just realized you may have only been arguing for the 3/4 on Yesler up to Broadway but then you are skipping the highest ridership outside of downtown and it’s not one that can easily walk. I don’t really understand why this would bea good idea.

      1. Because it would probably cut in half the run time for the Harborview downtown segment and make it far more reliable.

      2. I understand the idea is to make the ride more reliable. I don’t understand where the route would go. Up 8th from Yesler? Is that road even big enough for a bus? Basically it seemed like you were proposing taking away the Harborview stop which seemed like madness.

      3. Indeed, taking away Harborview service would have been madness. Yes, the bus would go up 8th and then 9th.

      4. So now that that makes more sense, I have to question the necessity! :)

        Why is James so unreliable? The drive up the hill? Is it the turn from 3rd Ave? Why would the turn one block down be that much better? I guess it’s easy to say that an alternative is better by the numbers, but why is it? Arguably detouring a couple blocks south should not be faster. Would making James bus-only (except local traffic) help? Cherry is a much better street for thru cars anyway.

      5. From the Central District News article:

        a key cause of delay exists around Harborview where the buses have to negotiate a tight left turn from 9th to James, fighting all the way with a mess of other vehicles trying to get to and from I-5.

        One option that jumped out was Yesler Way, currently only served by the infrequent #27 route. It’s a much less congested roadway and doesn’t have any access to I-5, allowing it to avoid the rush of hospital employees that clog up James during rush hours.

      6. aw has it spot on. James is a major freeway access point. There’s no realistic way we could close that street.

        And again, the proof is in the last chart. Yesler runs like clockwork, James is a disaster.

      7. I meant make local only *on* James, not close it entirely. I didn’t realize that many drivers use James to get onto I-5 — the I-5 guide signs from First Hill direct you to Cherry I thought, not James, even if you can get to the onramp from James.

      8. In general, one of the biggest problems with downtown buses going east-west is their tendency to use the same streets that cars use to access I-5. Any change in routing that gets the buses free of that congestion will speed up service. Unfortunately the 8, which gets stuck behind traffic accessing I-5 from Denny, has no other way to get across the freeway.

      9. Unfortunately the 8, which gets stuck behind traffic accessing I-5 from Denny, has no other way to get across the freeway.

        Not true — look at the snow route. It’s not as pretty as running the route straight down Denny, but if it dramatically improves reliability, it could be worth it…

      10. “I didn’t realize that many drivers use James to get onto I-5 — the I-5 guide signs from First Hill direct you to Cherry I thought, not James, even if you can get to the onramp from James.”

        Cherry is not an arterial street on First Hill, and there is no way to get to I-5 southbound from there without moving to another street to cross the freeway. Are you thinking of Cherry St east of Seattle U, which turns into James as it moves through the SU campus to Broadway?

      11. 8th is suited for buses, one of the special UW Medical routes currently accesses harborview by Yesler-8th-9th.

        James is a huge I-5 access point, and is frequently backed up for blocks in both directions. Signs at Boren direct I-5 traffic to James, for both directions. It’s cruel to ask buses to wade through that.

    2. “As someone who lived on Jefferson, putting the route on Yesler would suck.”

      I’m suggesting putting it on Yesler only for the section up to 9th Ave where it would turn to Harborview and then to Jefferson, similar to the current 3/4 does from James to Harborview.

      ““suspicious” is a somewhat strange word choice”

      I should have been more clear. A lot of terminal stops show abnormally high boardings/deboardings, I suspect because of bus drivers get on/off for their breaks.

      Is the 7 really 10 minutes slower to downtown than the 4 from Rainier/Walker? It doesn’t look like that from the schedule.

      1. The 7 rarely seems to run in time. If I didn’t use onebusaway I would go mad. And not everyone has onebusaway and instead go out to bus stops per schedule, which is another reason peolple in Judkins Park might prefer the 4. Walking out to a stop 10+ minutes away is unrewarding if you don’t know it will come on time and you do know it will go thru the ID which is very slow.

      2. I agree that the unreliability of the 7 is a serious problem, a problem that we should address directly, rather than maintaining duplicative unproductive routes that go to the same place a different way.

        Also, for commuters, the 7X stops at Walker and I believe is reasonably reliable.

      3. I have three points:

        1. I wasn’t necessarily saying to keep the tail of 4 just giving a reason why ridership is real. :)

        2. The service is only duplicated if you’re going downtown. We in Seattle have a major problem only thinking in terms of downtown rides. For example, in this particular case, someone going from Judkins Park to Seattle University probably just added 50% or more to their commute (the fastest route is probably to take the 48 and walk on both ends). You can probably think of other useful destinations on the current 4 route that are incredibly inconvenient without the 4.

        3. Partially duplicated service is not a bad thing. It gives the system resiliency and reliability in the face of partial failure. If duplicated service were bad, all car drivers would take feeder roads to main thoroughfares (of which there would be very few) and there would be very few “thru” streets overall, possibly only one per neighborhood.

        As for the reliability of the 7 or 7x, I don’t know. I didn’t even realize there was an express for many months because the trip from the I-90 stop always seemed to take 15-20 minutes. It took me noticing the driver passing up a stop with people with at it (I usually am reading) for me to realize I had taken an express. :)

      4. “1. I wasn’t necessarily saying to keep the tail of 4 just giving a reason why ridership is real.”

        OK, gotcha. And as you can probably see, I’ve revised the piece to make more clear what I mean about James and Yesler.

        The service is only duplicated if you’re going downtown.

        … or to the commercial district and library on 23rd/Jackson which is a major ridership center. And from 23rd & Jefferson during the daytime there would buses headed downtown at least every seven to ten minutes — probably more like five minutes.

        We in Seattle have a major problem only thinking in terms of downtown rides.

        Yes and no. That is an issue in some parts of the city like the north and south ends where you have almost no crosstown service, but I disagree that’s a problem for your neighborhood. The 7, 8, 9 and 48 provide service to a raft of non-downtown ridership centers. I agree that the 9’s hours of service would ideally be extended.

        Partially duplicated service is not a bad thing.

        I disagree. Service duplication wastes money that could otherwise be focused on improving a smaller number of routes.

        Take the 7. I agree that the 7 is slow and unreliable, and I see why people would take the 4 instead from your stop. Running the 4 1.6 miles out of its way (almost empty, most of the day) robs money that could be used for capital improvements on the 7 or service improvements on the 9. The reliability of the 4 at that stop helps you and a handful of others, but doesn’t help the thousands of other people who have to put up with the crap service of the 7.

        all car drivers would take feeder roads to main thoroughfares (of which there would be very few)

        That is, in fact, precisely what happens. The city designates some roads as arterials (of various levels) that are designed to prioritize capacity and speed, while neighborhood routes are designed to maximize safety and access. You can see a map of Seattle’s arterial system here:

        I forget the numbers, but arterial streets, even though they comprise a small fraction of street miles, carry a majority of vehicle miles traveled.

      5. It would be great if the 9 could become a full-time route. That would make for a nice parallel crosstown to the 48.

      6. Yes, we should fix the 7 (and the slowness of any bus on Jackson) but the existing state of affairs is a good reason why anyone living near the 4 might prefer not taking a ten minute walk to an unreliable bus. Given the 7 already runs incredibly frequently, what good would the 4’s service hours do for the 7? If you’re going to take away a neighborhood’s bus, it seems reasonable to require they have some promise of improved service sooner rather than later. You haven’t given either the amount saved or how it could be used to improve the 7. What would that be? The 7 already runs 7-10 minute headways most of the day. Adding more buses will just mean more bunched up buses. You mentioned capital improvements — what capital improvements are there that would increase reliability? Considering we’re talking about a Judkins Park user who would now be facing an extra 15 minutes for their trip (ten minute walk assuming single adult not children, give five minutes to arrive early enough to ensure not missed by a slightly early bus), the savings would have to be quite a bit better to be a fair trade to current users.

        There is also a social justice argument that has been barely touched upon — yes, not all routes are “efficient” and they may never be. Unless we’re prepared to run Access at a much higher standard than we do now, we need the “non-performing” routes just to be fair to people who can’t drive. In other words, the tail of the 4 doesn’t “rob” users of the 7 or 9 of better bus service — it makes it possible for people who live in that area to go out and participate in society even though they can’t manage the 4/10 of a mile to Rainier. That distance is non-trivial to many users (women with young children, wheelchair users, the blind — sidewalks in this neighborhood are sometimes a cruel joke because they often exist but become impassable at random intervals).

        As an aside, does anyone know where the per route on time data is? I couldn’t find it on metro’s site and the aggregate data is depressing (only 75% of runs can manage an on time standard of no more than 5 minutes late) so it’d be nice to see that broken out by route so I’m not quite as depressed…

      7. Given the 7 already runs incredibly frequently, what good would the 4′s service hours do for the 7? […] You mentioned capital improvements — what capital improvements are there that would increase reliability?

        If you read my suggestion carefully, it was that I would spend the money either on increasing the span of service on the 9 (which is indeed inadequate) or capital improvements on the 7. I agree that extra service on the 7 would be pointless.

        Capital improvements on the 7 would include bus bulbs on Rainier (so the bus is not delayed pulling out into traffic), and, in the long term, could include bus lanes, bus signals, and maybe helping SDOT pay for the proposed reconfiguration of the MLK/Rainier intersection (by Mount Baker Station) as a couplet with a contraflow transit lane and queue jump.

        the savings would have to be quite a bit better to be a fair trade to current users.

        Turning back at Jefferson saves boatloads of money. It cuts the cycle time of that south section of the 4 nearly in half. I’m hoping to write a post in the next week or so detailing exactly how much better service from First Hill to SPU could be once the 4’s tail is abolished. Basically (along with some reconfigurations at the Queen Anne terminals) it can get you five minute peak headways and 7.5 minute midday headways for zero extra cost.

        There is also a social justice argument that has been barely touched upon — yes, not all routes are “efficient” and they may never be. Unless we’re prepared to run Access at a much higher standard than we do now, we need the “non-performing” routes just to be fair to people who can’t

        That distance [0.4 miles] is non-trivial to many users[…]

        Social justice and mobility issues are why we operate service like the 27 (past MLK) or the 1 (past 1st & Mercer) or night owl (2AM – 4AM) service. Few people use those routes, but they serve people who would otherwise be stranded. That said, we can’t operate fixed-route service to everyone’s doorstep, and, in fact, attempting to do so is arguably what got Metro into its current financial pickle.

        Everyone who might the tail of the 4 is either within walking distance of a one-seat ride to downtown on the 3, 7 or 14, or a two-seat ride on the 8 or 48 (the transfer requiring very little walking and being less than 7 minutes). People in other areas of the county that are comparably dense to Judkins Park would probably be shocked that you consider this somehow inadequate.

        As an aside, does anyone know where the per route on time data is?

        Metro doesn’t publish route-level data like that at present, and it would probably be a fairly heavy investment of staff time to compile such statistics regularly. The charts I assemble for my posts on STB are non-trivial to make.

      8. Chiming in as a former student who lived in Judkins Park while attending Seattle U. At the time, the 9 only ran about once an hour, so it was fastest for me to take the 30 minute walk and forgo the bus. Even now the 9 has so few stops in that area that it’s not always useful. The transit situation for SU students living south of the school is pretty miserable, and it shouldn’t have to be.

      9. Making the 9-express more frequent may be worthwhile until University Link opens, but ultimately the 9-local should be restored.

      10. Making the 9-express more frequent may be worthwhile until University Link opens, but ultimately the 9-local should be restored.

        As I expressed elsewhere, I still think the more logical change is for the main RV route to head up Boren and Fairview, especially once the FHSC opens. Broadway won’t need any more service, but Boren desperately will.

  4. Wee suggestion: These analyses are interesting and valuable. I think they would be far more legible/comprehensible if they included a route map (ideally showing other relevant routes like the 3, 8 and 48) and relevant land-uses like the Lighthouse for the Blind.


      1. Thank you, I wouldn’t have made half my original comment if I had realized the exact route change you were suggesting.

  5. I lived at Terry & Jefferson for two years, which is one block from the 3/4 but three hilly blocks from the 27. The 27 is less frequent (hourly in the evenings). But still I used to cheer when the 27 came first because it was a faster smoother ride. The 3/4 is not just slow, it stops for a minute or more every block up James, and at the turns at 3rd and 9th, and it has that freeway traffic.

    I always felt sad for those who lived on the tails of the 3 and 4, having to go through the Harborview bottleneck every time, simply because their route happens to go through there. I’d move.

    I also wouldn’t live on the 3 east of 23rd because of the half-hour frequency. If it were 15 minutes, I might move there. The Central District is a surprising island of little transit even though it’s so close to downtown. Transit choices drop off dramatically east of 23rd, and get as bad as Magnolia.

    1. When the 27 breaks free of 3rd Ave, it just blazes up that hill. 2 or 3 minutes to 12th, every time.

      Throws into terrible relief the absurd routings/operations/signal-timing that cause the 3/4/12/2/10/11/49 to take 5-10 times longer to cover the same.

    2. One nice thing about being so close to downtown is that even if transit sucks, you’re not stuck with it.

      According to Google, a bicycle ride from the Central District to downtown is just 15-20 minutes, door-to-door, much better than any bus.


      1. Your map perfectly illustrates why most transit riders choose not to ride their bikes to make these trips.

      2. Yeah. 15 or 20 minutes. Maybe quicker. But it would take me 45 minutes to push my bicycle back up that hill!

        That’s more of a reflection of Google than anything else, though.

      3. I would avoid the hell out of any bike ride from the CD to downtown. The hills are as bad as any in the city, and unlike many other neighborhoods, there’s no good flat reroutes to take.

        I’ve done it many times, and there’s several moderately popular E/W biking routes out of the CD. But it’s utterly exhausting if you plan on going any further than Capitol Hill or the ID.

      4. Actually, Google estimates only a 3 minute difference going up the hill compared to going down. Still beats a 10 minute wait followed by a 20-30 minute bus ride.

      5. East-west is bad for biking, but MLK makes a pretty flat trip from Madison to Rainier Beach.

    3. I also wouldn’t live on the 3 east of 23rd because of the half-hour frequency

      So we use the service hours to make the 3 a frequent route, rather than adding more confusing turnbacks. Problem solved. Ridership past 23rd would probably grow dramatically, and the somewhat-secluded business district on 34th gets more accessible.

      1. This wouldn’t be adding another turnback. This would be reducing the number of termini.

        The ridership out there on Cherry is minuscule, and the Madrona commercial district has frequent service from the 2. Unless that extension is revenue neutral due to cycle-time effects, I don’t think it’s worth the effort.

      2. I agree with Bruce – and I would benefit from increased service on the 3 east of 23rd personally. But I still don’t think it makes good sense to do that through-route all the way to the current terminus. The current turnaround seems logical based on ridership.

      3. Most regular transit riders be it here, NY, or DC have learned to usefully use their time on the bus, often reading or texting or whatever.

    4. live between Cherry and Union east of 23rd and you can split the difference between the 2 and the 3 :)

  6. Fond Memories Dept:
    I was driving the 4 some years back when the snow really started down in the afternoon approaching the turnback wire on Judkins. My leader bus got off wire, and by the time nightfall arrived there were about 6 buses stuck along Judkins.
    One neighbor make hot coco for all the drivers sitting in one bus, another brought out homemade cookies, and we all told war stories until a push truck finally arrived about 9 that night.
    Much better than a picnic on a sunny day.

    1. Well, there’s virtually no downside to it. The gap in operational costs between a 40′ and 60′ trolleybus is minuscule. So if it’s standing-room-only this consistently, why not take the plunge?

      1. Aside from the fact that a new turnback in Madrona would be needed due to the turns and short layover space.

  7. It seems like most of the tail end of the 4 is within a 1/2 mile’ish walk from the I-90/Ranier station, from which the 550/554 can get you downtown in as little as 2 minutes. For cross-town trips, there’s still the 8, 48, or possibly 9 if you happen to be traveling at the right time of day and day of week. So, I would be inclined to say, yes, the 4 does appear redundant through that stretch.

    That being said, I don’t ride through that area very much and, when I have, my primary use has been jogging along Lake Washington Boulevard on a weekend.

    1. A half mile is a long way if you’re blind, using a wheelchair, have small children or are sick or out of shape.

      1. (N.B. my comment on it being a long way for someone who is blind is because I know the quality of the sidewalks in this area are poor.)

  8. I’m fundamentally opposed to turnback routings, but still think this is a good idea. The 4’s tail is a fossil, a vestige of the old streetcar network. Newer, high frequency routes on MLK and 23rd with numerous downtown & crosstown connections (including Central Link!) render it completely redundant.

    The numbers show that it’s practical. Even trusting the terminus numbers, you get about 5 riders down at Rainier, people who are simply avoiding the 7’s trudge down Jackson (so fix the 7 on Jackson). You pick up 3 or 4 riders in the residential stretch between there and Jackson – They’ll need to walk to the 7/14 or add a transfer from an existing high-frequency route.
    Then we hit the busy transfer point at 23rd & Jackson, and pick up maybe 5 more. Transfers from the 8/48 can move down to Jefferson/Cherry, so that’s not a problem. A large part of the walk-on (non-transfer) traffic is probably folks just taking the first “DOWNTOWN” bus they encounter, so those riders can take the 14 (again, fix Jackson if you have a complaint about that). There’s a couple more riders at Yesler, with the exact same alternatives (except that the 27 is a much faster route downtown).

    The only people I see really missing out are the riders on the Jefferson corridor who lose a one-seat ride to the commercial center at 23rd & Jackson. However, with the Kress IGA downtown now, that’s not anyone’s grocery lifeline.

  9. Yesler has no ADA access to or from surrounding streets between 3rd & 8th, only two narrow staircases. Riders currently boarding at 5th & James will have to walk, typically less than five minutes, most likely to the stop on 3rd.

    This is no longer true. Accessible sidewalks have been completed on 6th, and there are no longer any sidewalk gaps. You can wheel down 6th from James to Yesler, or from Jackson to Yesler. Steep slopes are involved, but this is true of James as well.

  10. I can’t say how I feel about the #4 vs just more #3s. Remember these two routes directly serve important areas of SU, Harborview, and Swedish with high ridership and very crowded. I am not sure why people agitate to change them from that route especially since they serve extremely important employment centers and have a high ridership. Perhaps the #4 is not important around Judkins, but the number of #3s would have to be increased to serve the already served areas along E. Jefferson. This is a terrible idea. Doesn’t ridership usually rather go down at either end of most routes? The current route also allows more easy access to many CD residents to the city and county buildings. Moving it to E. Yesler would make for a greater walk for more residents to any of the major institutions. Remember many of the users are also patients and their families who may not in great shape. Yesler would not be a suitable substitute.

    I have occasionally used the #4 to get to the Judkins area. But, find the #48 suitable for my purposes. However, the #4 does serve a small area that otherwise does not have easy access to a bus to downtown. I don’t know why you question the boarding numbers at Walker.

    1. The point is that routes are different from segments. People ride in a segment, and downtown to Harborview is a high-ridership segment. That doesn’t say anything about which route should serve it, and where it should go at the end of the segment. When people say “Keep the 4” or “Keep the 3”, they’re not considering what other routing may end up serving more people, or would connect several segments more efficiently. For instance, one method is to reconfigure two routes to join their high-ridership segments into a frequent route with an articulated bus, and join their lower-ridership segments into a less frequent route with a small bus.

      You’re also discounting the real problem that for people on the tails of the 3 and 4, it takes significantly longer to go the same distance to downtown, than their luckier neighbors who live near the 2, 12, and 27. We should not disenfranchise Harborview riders, but at the same time we can ask whether Harborview and the eastern Central District have to be served by the same route. Or whether we can move the route three blocks south to Yesler and improve most people’s trips. Note that the Yesler routing WILL serve Harborview; it just wouldn’t serve the part of James from 5th to 9th, which means the jail and the apartments east of it.

      1. I think that access to major institutions should be a consideration. Oh you mean duplicate further and serve just Harborview with yet another route?

    2. The proposed reroute to Yesler is only on the part to Harborview. It does not miss any major employment centers.

      Increasing the number of #3s is precisely what would happen. It’s about half the time (thus about half the money) to run a 3 to 23rd & Jefferson than all the way down to Walker.

  11. Harborview is an important and big one. Also remember that the First Hill Street Car will also be traveling on that portion of Broadway if you are thinking of having it travel from Jefferson to Yesler on Broadway. There are already a lot of transit already stopping in front of the County and City Buildings on 3rd and the current stops on James don’t compete there. Where would the stops be for the #3 or #4 on 3rd for access to the City and County buildings.

    1. I’m proposing that instead of the bus going up James then zigzagging to Harborview and go down Jefferson, that it instead go up Yesler and then go up 8th to Harborview then down Jefferson.

      Where would the stops be for the #3 or #4 on 3rd for access to the City and County buildings.

      On 3rd Ave, just south of James, where the current stop for the 27 is.

  12. I don’t thing there are any major advantages and that the disadvantages outweigh them. As for the Judkins corridor the trolley wires are there and unless there is new and great idea for them I don’t see taking the service away. By the way what is the ridership at the other end on Queen Anne? Is it high? Where is the major ridership at that end?

    1. Did you actually read my post? I have laid out the advantages in great detail. There is a considerable cost to running the 4 down to Judkins, yet very few people use it, and all who do have other options. There are much better ways to spend that money. Just because trolleybus wire exists does not mean that those routes are then free to operate, or that we are obliged to operate them forever.

      Queen Anne has pretty good ridership until the 4 starts its long stupid circle in East Queen Anne. I will cover the restructures for Queen Anne in a subsequent post.

  13. There are always some outlying areas that need service. Even if it the connection was along 9th with a jog on Spruce to Yesler, I am not convinced that would give any advantages since then turning onto third would put in competition with a lot of other transit and traffic and still not serve the City and County buildings. Even if it stopped on 3rd in the area it would be competing with other buses on third for stops and the walk to those buildings be a slog up the hill. This route may be one of the few that does a good job of providing ADA access to the City and County buildings.

    The route is longer and would have to compete with more buses for those blocks on third.

    1. The bus stops on 3rd south of James do not have any capacity problems (mostly because of all the West Seattle routes that get on the viaduct at Columbia).

      As you can see from the chart, the 27 is much faster and more reliable getting between 3rd & James than the 3/4 are getting between 4th & James. Moving the 3/4 to Yesler will make it faster because James is so congested and slow, even though the bus will go farther.

      I agree that James serves the county buildings better (this is one of the disadvantages I mention to the Yesler reroute in my post) I still think these buildings are within reasonable walking distance of either 3rd & James or 6th & Yesler.

      1. The basic observation behind the move from James to Yesler is that non-freeway buses do not belong on freeway onramps — and that’s precisely what James is! (Incidentally, the same is true for Denny eastbound, west of I-5.)

  14. Riders at 8th & James will have to walk up the steep hill to 9th & Jefferson for the 3/4 or several blocks to Madison for the nearest service on the 12. This creates a small mobility gap in a very steep part of the city that is difficult to address.

    This is a stupid question (because I don’t live in the CD and so I don’t really know the ridership patterns), but what about rerouting the 3/4 from Jefferson to James/Cherry? That is:

    – Head from downtown (3rd Ave) to Yesler
    – Turn onto 8th and head up to Harborview
    – At 9th, turn onto James/Cherry
    – Continue to the 3’s terminus

    This seems like it serves First Hill better than turning at Jefferson. It also provides extra service to Seattle U (which needs it); it better follows Newton’s First Law of Bus Routes (avoiding the turn from Jefferson to Cherry at 23rd); and it seems to provide a more even spacing between the 2 and the 27.

    That said, I’m not sure whether this would cause the same I-5 problems that we have now (is James an I-5 onramp in First Hill?), and I also don’t know whether Cherry in the CD has a sufficient walkshed compared to Jefferson.


      1. Argh, apparently if you copy-paste from the address bar, you’ll get the *old* link, not the new one. Finally:

        via Cherry

        Mods, feel free to delete the intermediate posts and update the original if you want…

    1. I like it. Crosses the freeway at Yesler, drives right through Harborview, and is overall fairly straight.

      I’m no expert on buses in that area, but it certainly looks logical.

  15. Your suggestions makes sense for those trying to access Harborview and other First Hill health care destinations from downtown, but not for those who want to access these from the south end. I catch the 4 on MLK and Massachusetts and get off at HMC or the next earlier stop every day, and I don’t see another alternative that would work for me if the #4 goes away. There are others from the 4 who also want a one-bus option for getting to HMC, Swedish, Virginia Mason, etc.

    1. To defend Bruce a bit since I’ve been giving him such a hard time [1], the idea he’s getting at here is perfectly reasonable and is actually a common idea on this blog: not everyone gets one seat rides to their destinations. So for someone like you, you choices would be:

      * take the 48 and transfer to the 3
      * take a bus downtown (such as the 7, 7x or any of the flyover stop buses at I-90 and Rainier) and then transfer to something going up the hill
      * possibly the 9x directly

      I find these all perfectly feasible if the transfers work and the frequencies work out. But, unfortunately, very few of them seem to when I try them — the 48 and the 7 are two of the highest frequency and I generally won’t do a transfer to/from them if I am under a tight schedule (or I leave very early.) So, I completely understand the resistance people have to losing their one-seat rides when it’s not at all clear any savings would actually go to making it possible for them to make the same trips in about the same time.

      [1] I’ve been giving him a hard time I think because I have a pretty strong reactionary position sometimes (transit should be better! taking service away is bad!), but more specifically because the original post as written was very clear that there was a lot of savings in the changes but not at all clear to me that the savings could be effectively used to make transit better for those whose existing service would change or be lost. But I’ve definitely been too hard on him and not reading as carefully as I could. Sorry Bruce. :(

      1. Someone (I think it was d.p.) once pointed out that, despite being a major arterial and a direct connection between two major employment centers, Boren has virtually no transit. So what about rerouting the 7 to First Hill/SLU via Boren/Fairview, also replacing the 9 (and maybe even the 70)? Think Martin’s Rainier Valley Mobility plan, except that the new 7/9 now goes via Boren instead of Broadway.

        This gets you the direct connection to First Hill and SLU from much of the 4’s area (as well as most of South Seattle). It’s not quite door-to-door service to Harborview, but it’s pretty close, and it’s arguably *better* service to First Hill if you want to go anywhere else (e.g. Swedish, VM).

        This seems like a nice compromise. You replace a bunch of complex routes with a smaller number of very straight ones, which improves reliability and legibility.

      2. @Rachel I value your input, and I realize I haven’t laid out exactly what the benefits are from removing the tail of the 4. Look for a post late next week on that subject.

        People such as yourself, who live in the neighborhoods where transit planners and bus nerds suggest these major restructures, are precisely the people we need to convince and explain ourselves to.

      3. I think we need to make the other changes first before messing with any of Metro’s most successful routes (7, 36, etc). There’s a real chance of making things worse for people even if it looks good theoretically. There may be latent demand for a Rainier-Boren-SLU-Seattle Center route, and some people may find Boren close enough to their destinations, but there’s also a chance it would screw a significant percentage or Rainier Valley transit riders. So let’s reorganize the smaller routes first and see how well it goes, and then come back to the 7 and 36 and reevaluate them.

      4. The problem with that, Mike, is if you create new, straight, potentially useful, “corridor-ized” routes, but then run them infrequently or with limited hours, they don’t actually become that useful. Their potential remains untapped.

        The classic example is the 46. It connects Ballard to Fremont (sometimes).
        It provides an east-west through route without the bottlenecks of the 44 (sometimes). It has the potential to connect Fremont, the north-lake areas, and the U-District (without being subjected to ship-canal-crossing delays like the current options), and it could therefore be a key east-west addition to a better-gridded system.

        But currently it runs 8 times a day. On two unrelated paths. So it’s useless, and its numbers are terrible.

        Newly proposed service corridors should only be implemented as all-day frequent service and nothing less. The 7 runs constantly and yet it sucks anyway. Why shouldn’t we be messing with it?

      5. I said “most succesful routes”, meaning approximately the top ten. The 7, 36, 48, etc. Not the 46, 14, 39, and other moderately-used or little-used routes. The 7 is a perfect example of potential BRT: people make different trips all along its length, both within the Valley, Valley-Jackson, Valley-downtown, downtown-Jackson, etc. In contrast, the 4 and 14 have high ridership on the part of the route that happens to Jackson and Harborview, which argues for high-frequency routes there but not necessarily the 4 and 14. (Of course the 14 is full peak hours; that’s a separate issue.)

        The 46 is a mostly-ineffective route whose main effect is to highight the hole in all-day service between Ballard-Fremont and Ballard-Golden Gardens. (I don’t know what it’s peak-hour ridership is.) I won’t say Ballard-Fremont-UW because the 30 and 31 are doing a good job for Fremont-UW, and it’s not clear that Ballard needs two one-seat rides to UW. The 44 needs to be sped up of course. Maybe ultimately we’ll have to make Ballard-Fremont-UW the main route and give up on improving the 44.

      6. Every plan that has ever been concocted for cross-town improvements includes the 46 as a corridor:

        1. There’s no reason the eastern half of a north-of-canal corridor should rely on the labyrinthine 30 or the Nickerson-overpass- and Fremont-bridge-requiring 31.

        2. Ballard-Fremont as part of a continuous corridor rather than isolated.

        3. Lower Wallingford is actually a mile from 45th, in addition to the hill. An east-west through corridor is actually better than the scattershot 16/26 service it has now (which should probably become a single north-south service anyway).

        4. Ballard-UW needs a better through corridor than the 44 is clearly capable of providing. To the extent that they serve very different intermediate purposes, and given our general dearth of east-west north of the canal, both are valuable.

        For the last of those reasons, the commute-hour 46s are packed! But any overall reading of the 46’s metrics on paper would look horrible.

        I was theorizing, with a great deal of evidence, that a part-time or infrequent Boren corridor service would have horrible numbers on paper and would undersell the potential for the corridor.

        I wasn’t arguing for or against rerouting 7 service. Though seriously, imagine the number of connections that you would avoid slogging through downtown to make if Boren cross-transit were suddenly available! (But would you do the same if Boren were to infrequent to rely upon?)

      7. Oh, you’re right about the 46. I didn’t realize it ran on 40th. I thought it went southeast on Market, 45th, and northeast on Greenlake Way to 50th. I must be remembering an old route. Yes, the 46 could potentially replace the 30 and 31 on 40th, but it would have to be local then.

  16. Again, I live in the neighborhood and I still say if you want to practice tweaking this route start on Queen Anne where there aren’t so many major institutions and the ridership might not be as great. Who knows? But, wait I might want access to some of the Queen Anne neighborhoods even from the CD.

    Every bus route will grow in ridership as it moves and then thin out at major destinations and then pick up again. Nonetheless, the public interest lies in as much as possible assuring transit access to all. Everyone comes from different areas to major destinations. I’m not sure if this is some type official blog or exactly what your aims are. At this point you were positioned well enough to be referenced in another blog and should listen a bit better to the real concerns, The routes were designed with some need in mind. You might start by asking what that was and if then without asserting an answer ask a few questions.

  17. All neighborhoods should be well-served in Seattle. In the meantime, have you looked at something like the #12 and its total ridership. Could its numbers be improved? I don’t get why the #4 became your issue, when a vast area of this route is one of the most heavily used. There are many other little neighborhoods where the service could be more easily questioned. You should start with a less complex route to question.

    1. Re the 12: Funny you should bring that up. :)

      I can’t speak for Bruce, but for me, part of the attraction with fixing the 4 is precisely that it’s so heavily used. Fixing this corridor — even if it means folding the 4 into the 3 — has the potential to benefit a *ton* of people.

      One of the biggest problems with operational change for any system (transit or otherwise) is that current users are much louder than future users. Yes, removing the long tail of the 4 will inconvenience some people. But how many people currently don’t ride the 3 because of the painful slog on James, or because of the poor frequency past 23rd? None of those people are speaking up because they don’t think of themselves as riders of the 3.

  18. I want to add that all public schools should have one bus that allows families almost direct access to the school. The #4 is the only bus that would fit into that category for Thurgood Marshall.

    1. Eh? The 8 stops at the same stops as the 4, and the 48 comes almost as close on the other end…

  19. Yes, the the north/south buses #48 and #8 do stop nearby on 23rd and MLKjr Way respectively. The #4 is an east/west bus and has one stop that is a bit closer as it connects from 23rd to MLKjr Way. This may not be the best argument for the section. It is still a neighborhood in an odd place where without the #4 there would not be easy access to downtown. The light rail is a bit too far to walk for many and having to transfer seems unnecessary. I’m not sure why you were so certain that the boarding records at Walker were incorrect, that is too high. There must be dozens of route ends where the very end of the routes could have the same issues. I guess that is the analysis I would want to see. Who uses this one and how does it compare to all other route ends before being certain that this one deserves to be cut. Then as I said before the rest of the route is a heavily used one.

  20. I think one of the lessons here is that headlines for re-route posts need a more positive title. I’m not sure if readers who aren’t familiar with the blog would have had quite as negative an initial reaction if the title had been “Could the 4 get to Harborview more quickly on Yesler?”

    1. Good point. I also haven’t done a good job explaining how removing these goofy long tails from our bus network can allow us to have a very fast, frequent network. I will write a post about that next week.

  21. Getting to Harborview more quickly is not the only point here and from which direction? It doesn’t go to Harborview on E. Yesler unless 9th is the connector between Jefferson and Yesler. What about the access to the City offices? E. Yesler is not a direct route to Harborview.

    OK. As for the posts regarding negative reactions and long tails. Some long tails on otherwise high demand routes serve areas that would otherwise not be served. Many from all different areas want to access to, for instance, most educational institutions and major medical centers, and therefore, not every tail will be a high demand area as most buses move through major destinations to various neighborhoods. If you begin willy nilly choosing which tails are the ones you use least without doing an analysis of all the tails and noses then you are not doing the public any favors. Perhaps Judkins is a nose, not a tail. On top of that the whole area there is quite high density and growing and you yourself admitted the boarding rate at Walker was better than you anticipated. Why then go after that portion of the #4 without analyzing how it compares to all other tails and noses? I say the tails and noses of routes are often important. For instance, I have sometimes needed access to an area that is not otherwise that east to access. For those who provide services to not just one school or facility or for the disabled being able to move without transferring and having to worry about many different connections is sometimes more important than having all the routes be really fast and not quite getting to all the places necessary. Also please acknowledge that it is reasonable that tails and noses will generally have fewer passengers and gather them as they move along the route and then lose them as they go to the other end. I am not sure how you want the corridor fixed. During the day one bus or the other comes about every 7 minutes. Since I use it mainly for the corridor covered by both, I don’t really notice if one is late or not since they run so often. When I want to go to Judkins or Madrona on that route then I notice and haven’t found it to be outrageous. I admit that I do not use them to go in that direction a lot.

    1. What do you mean by noses? The fact is that half of 3/4 riders get on downtown and get off at Harborview. As opposed to say the 14 and 358 where people get off gradually along the entire route. (The 550 also has this issue: a large chunk of riders get off at Mercer Island or South Bellevue.) This argues that maybe the Harborview portion should be more frequent and maybe a different route from the 3/4 tails. Given that the south part of the 4 has so much less ridership than the James portion, I’m not sure how you can call it a nose rather than a tail, but maybe I’m misunderstanding you. There should certainly be a discussion of what are the optimal routes for trips originating from the 23/Judkins area, to both the north and northwest. That may involve keeping the 4 intact, or it may involve other routing adjustments we haven’t considered yet.

      1. The 550 also has this issue: a large chunk of riders get off at Mercer Island or South Bellevue.

        That said, the 550 still sees strong demand all the way to Bellevue. The 4 is a major trunk route to 23rd, and a neighborhood circulator afterwards.

      2. The nose would be either Judkins or Queen Anne. Why aren’t you picking on the Queen Ann tail/nose. Tail and nose just imply beginning or end of route, depending on how you want to say it. I hope you aren’t suggesting that this route serve just downtown and Harborview. There aren’t giant park and rides in the city and that is as it should be.
        Is this route in more need of examination in Judkins than on Queen Anne? Or maybe you could compare the tails/noses of other routes like the 10,12, 13 27, 30, 71, 76,17, 18,19, 24,31, 33 and many others. Judkins does have access north to the U and to Queen Ann on other routes, but not downtown. Also none has addressed the issue of access to the City offices. Neighborhoods have to be served. I’m just not sure why this one is under such scrutiny where I will assume ridership is as great and probably greater than on many other routes listed above.

      3. “Why aren’t you picking on the Queen Ann tail/nose.”

        The Queen Anne end of the 3 and 4 is Bruce’s next target and to not spoil it too much: they’re getting the same treatment as the Central District end.

      4. To Alex, isn’t all the way to Bellevue a lot longer way than 23rd to Judkins? Where the #4 turns into a workhorse headed to many major institutions. I have been on the 550 in the mid morning and early afternoon hours and was one of very few riders much of the way, for many more miles than the area covered from Judkins to 23rd. However, as a user of Metro I want the 550 to continue to go to Bellevue at that time, as occasionally I have business there and want to use the bus. I don’t just want my favorite route to work, I want my favorite routes to connect to many routes that will take me many different places. In order for my bus to be useful, it must be able to connect to many other buses or to the light rail.

    2. Joanna,

      Based on your last two comments, I think you may be misunderstanding the proposed change. What Bruce and Metro are proposing is *not* to reroute the 3 entirely down Yesler in place of the 27. Rather, it’s a much more targeted change, that would turn from Yesler onto 8th/9th, then continue up to Harborview, and turn onto Jefferson to resume the current route. Here’s a map.

      As far as your point about tails, I actually disagree with this statement:

      Also please acknowledge that it is reasonable that tails and noses will generally have fewer passengers and gather them as they move along the route and then lose them as they go to the other end.

      For some routes, generally low-traffic neighborhood circulators, this is true. But for others — especially well-designed high-frequency arterial service — both ends of the line are strong all-day demand generators. Look at Link, for example, where there’s a huge number of people getting on and off at both ends (downtown and the airport) all the time. The same is true for the 550, and the 43/49, for example. Aside from late at night, those buses are pretty much never empty.

      There’s a lot of value in designing a system such that these major corridors are served by high-frequency, high-capacity service, while tail residential neighborhoods are served with “right-sized” connecting service to those trunks. It means that you can put most of your resources into service where it’s needed and that’s efficient to provide, and avoid the added costs and complexity of running lots of empty buses to areas that don’t need the service.

      1. This is true for routes designed between two major destinations or park and rides, but not necessarily for the ones serving neighborhoods. Often even the ones moving between two major destinations will serve neighborhoods in between and ridership with rise and fall pretty drastically during the route, the #43 is probably an example. The #4 is an example of bus that moves through a lot of major to sort of major destinations, and serves neighborhoods at each end.

      2. There are times where I have not been clear on if Broadway or 9th was the connector advocated in this scheme between Yesler and Jefferson. However, some of my comments address both scenarios. And, then there are those here and on CD News proposing various different ideas.

      3. The number of people on the 43 definitely decreases as leaves Montlake and then picks up again at Broadway. However, that is not as true for the #44, which used to be the other end of the #43. The #49 is often full coming out of the U District and almost empty by the time it gets to Broadway. In the evening it does not always fill up again to get downtown. During the day more use definitely begins again around Seattle Central to downtown. However, I think the neighborhood routes are extremely necessary to get to the link or any other express type transit. By the way to say low-traffic circulars is not fair. The 3, 4 and other buses like the 44 or even the #2, all have very high traffic and happen to also include service to a few neighborhood circulars.

  22. Buses also need places to rest and to turn around. Parking is already horrible around E. Yesler and 23rd and new construction is going up as I write. Ask anyone who has had a meeting at Douglass Truth Library there.

    1. Parking vs layover space can be a problem. Terminating the 4 at 23rd & Jefferson wouldn’t make that situation any worse, it would just mean that Metro could provide the current level of service to 23rd & Jefferson with fewer coaches underway at any given time, which will save a bunch of money, or that money could be reinvested in making service throughout Queen Anne/First Hill/Central District better.

      Look for a post about how this might pencil out on this blog next week.

  23. Ok, 23rd and E. Jefferson is Garfield High School, another major source of riders to downtown during the school year, but with no room for parking. Have you asked why after the freeway destroyed much of the Judkins neighborhood Metro bothered to string trolley wire to the location. I think it is to provide transportation for residents to downtown, which I would bet is the source of the unanticipated number of boardings at Walker. The number of riders seem to fill up the number of coaches. I want to see some analysis of this route compared to many that I listed above and again below.

    Again I ask if this route in more need of examination in Judkins than on Queen Anne? Or maybe you could compare the tails/noses of other routes like the 10,12, 13 27, 30, 71, 76,17, 18,19, 24,31, 33 and many others. Judkins does have access north to the U and to Queen Ann on other routes, but not downtown. Also none has addressed the issue of access to the City offices. Neighborhoods have to be served. I’m just not sure why this one is under such scrutiny where I will assume ridership is as great and probably greater than on many other routes listed above.

    1. Garfield is indeed an important transit destination. It is right next to the terminus of the 3 on First Hill, so trips between Garfield and Downtown would be unaffected by getting rid of the tail of the four.

      “Again I ask if this route in more need of examination in Judkins than on Queen Anne?”

      A fair question. Yes, absolutely, bus service on Queen Anne needs to be examined. My next post on this subject, as soon as I can type it up and Oran can make the maps, will be about that, and also about how much more frequent and reliable the service to First Hill, the Central District, and Madrona can be if we are willing to give up the long tails of the 4 in Judkins and the many long tails of the 2, 3 and 4 in Queen Anne.

      “Or maybe you could compare the tails/noses of other routes like the 10,12, 13 27, 30, 71, 76,17, 18,19, 24,31, 33 and many others. ”

      Routes 10 and 12: I already did these. I advocated abolishing the tail of the 12 that serves 19th Ave for precisely the same reason that I’m advocating abolishing the tail of the 4 to Judkins Park: because there is other bus service less than a quarter of a mile away, and that section of the route is not well used.

      Route 13: This will be covered in my next post.

      Route 27: Yes, the long tail of the 27 is very little used, but there is a crucial difference between that and (say) the 12. The 27’s is at the bottom of a huge hill, and the nearest alternative service is the 14, at the top of the hill.

      Route 30: Along with route 31, this route provides crosstown service to Queen Anne for a neighborhood which, other than peak-only service to downtown with Route 74, has only has one bus (the 75) which terminates in the U-District.

      Route 71: The tail of this route serves parts of Ravenna and Wedgwood which would not otherwise have any service within walking distance. This is not the case in Judkins or Center Park.

      Route 76 is a peak-only express that mostly serves to take the pressure off the overcrowded U-District routes.

      Routes 17 and 18 provide service to parts of Ballard whose nearest bus service would otherwise be more than half a mile away. This is not the case in Judkins or Center Park.

      Routes 19, 24, 31, 33: I agree that service in Magnolia is a big, confusing tangle of routes, and I hope to write about them at some point in the future.

      1. “Route 76 is a peak-only express that mostly serves to take the pressure off the overcrowded U-District routes.”

        That’s missing the main purpose of an express and why people ride it. The 76 exists not because the 71 is overcrowded but because people in Wedgewood don’t want to spend fifteen minutes going through the U-district when it’s neither their origin nor destination.

      2. Re Magnolia, the Metro budget-cutting spreadsheet several months ago suggested replacing the 33 (East Magnolia) with a peak-only route, and replacing the 19 and 24 with a vaguely-specified route from Ballard – east Discovery Park (the “ARTS” Indian cultural center) – Central Magnolia – downtown. (I’m not sure if west Magnolia — Viewmont Way — would be abandoned completely or if the peak-only 19 would be kept.)

        I rode the 24 one Saturday to take a look. The 24 makes a huge zigzag on 28th W, 32nd W, Magnolia Village, and Viewmont Way W. (Central and west Magnolia.) There are steep hills on and these parallel streets, so half the passengers rode from one part of Magnolia to another. If 28th is abandoned off-peak, it would require some moderate hill-climbing but not extreme, at least from my limited outsider’s viewpoint. 32nd goes way up high but it’s all big-house residential, so I suspect there’s little transit demand except by commuters. Deleting the 24 would lose all-day access to the southwest entrance of Discovery Park though, which is the side with the bluffs.

        The 33 covers the east side of Magnolia (Thorndyke-Gilman). If it’s eliminated, the 31 would still serve Thorndyke, and the new route would have to go on Gilman to get to Volunteer Park’s east entrance/ARTS. That would leave a 5-block gap in the middle if the bus takes Emerson rather than Dravus, but there’s only a few businesses in that area.

      3. My point here is that all this analysis has to be done at once, not one route at a time. Otherwise, it seems like your are cherry picking the routes to pick on. While the #3 may say First Hill for those specific routes, Garfield, of course, is not on First Hill, but in the CD on 23rd and Jefferson. I was mainly pointing out that there in not bus parking available at 23rd and E.Jefferson. Next door to that is the Garfield Community Center. Of course, I didn’t name them all, only the ones with which I am most familiar. If in your opinion the Judkins crew can transfer from the #48 or the #8 to downtown buses why would it be bad for another neighborhood have to transfer in the U District? And, then there is the odd little section of the 14 that serves Capitol Hill. They may all have a reason for existing. A thorough analysis has to be done for all before agitating to change one.

      4. “My point here is that all this analysis has to be done at once, not one route at a time.”

        Metro cannot restructure the entire Seattle bus network at once. There is a huge amount of staff time and public outreach that would have to go into changes of this scale.

        Is is true that all the parts of the 2, 3, 4, 13 system need to be looked at at once. I am introducing this subject by talking about the long tail of the 4, because I’m just a guy writing on a blog, and I don’t have time to type up 20 page posts covering every detail of the restructure at once, and if I did, no one would read them. Rest assured that I have more posts on the way covering other parts of the city’s bus network, and (as has been pointed out multiple times in this discussion) that I have written similar posts about Route 12, and also I have written a similar post about Route 16. I am not singling out Judkins park or Center Park in any way.

        “While the #3 may say First Hill for those specific routes, Garfield, of course, is not on First Hill, but in the CD on 23rd and Jefferson”

        Those routes currently signed “First Hill” terminate on the loop at 23rd & Jefferson, as shown on the map above. Those buses begin and end across the street from Garfield.

        “If in your opinion the Judkins crew can transfer from the #48 or the #8 to downtown buses why would it be bad for another neighborhood have to transfer in the U District?”

        Three responses here. First, because the 30 runs only every half hour, and to work well, transfers need to be between buses that run frequently. Half an hour isn’t frequent enough to make that transfer convenient. The 48 and 8 run every 15 minutes (’til 8 PM for the 8, 10 PM for the 48) and the 3/4 currently run every 7.5 minutes during the day and every 15 minutes ’til midnight.

        Second, because the 76 is reasonably well used. This is not true for the tail of the 4. If those buses through Judkins were full of people, we wouldn’t be having this conversation. Instead, they are carrying a handful of people.

        Third, people in Sand Point have no other service to downtown than the 76. People on the tail of the 4 have a raft of (faster) one-seat ride options to downtown, including the 7, 14 and 27. So their situation is different to yours.

        “And, then there is the odd little section of the 14 that serves Capitol Hill.”

        Another little used and largely redundant part of Metro’s network that I have advocated abolition of. Like the 4 from Jefferson to Dearborn, this is a legacy from the 1940’s conversion from streetcars to trolleybuses. Also, like the tail of the 4, it probably made lots of sense when it was created, but it has since been made redundant by new routes in Metro’s network.

      5. Actually, the 18 could take over the Magnolia route. It would just be replacing the Seattle detour with a Magnolia detour. That would keep service from north Ballard to Fisherman’s Terminal intact (it may otherwise be deleted or truncated due to RapidRide), and would perserve a (slightly-longer) one-seat ride from 24th NW to downtown, while simultaneously subsuming the 24 and its service hours.

        The main difficulty I see is that if it really wants to go to the door of the cultural center, I don’t see how it would avoid backtracking and adding 2-3 minutes in the middle of the route.

      6. “people in Sand Point have no other service to downtown than the 76.”

        The 76 does not go to Sand Point; it’s a steep hill away. (The southbound stop is closer and less hilly, but the northbound stop you have to walk up and over the hill, which makes for a 30+ minute walk to Magnuson Park.) The 74X goes peak-hour from Sand Point to downtown. The 30 goes to the U-district/Fremont/Seattle Center. The 75 goes to the U-district and Lake City/Northgate/Ballard.

      7. This is in response to your response. Restructuring analysis has to be done for the whole city before you pick on the 2, the 3 and the 4 or 14

        Many in Juckins have to catch and other bus to get to the 7, and especially the 14 and 27.

  24. A further Post from the some of actual users on CD News:
    Route #4 proposal
    I am a resident of Center Park Apartments, a low-income public housing community owned and operated by Seattle Housing Authority and am also an elected officer with the Resident Action Council (RAC), a grass-roots coalition of Seattle Housing Authority residents. Because of this, I am adamantly opposed to ANY reduction in service on this route for the following reasons:

    1. This route serves two destinations that generate a considerable number of transit users–Center Park, an apartment community specifically designed to accommodate people with disabilities; and the Lighthouse for the Blind, an employment center for people with vision impairments. This route is relied on by people who live and work in both of these locations–many of whom have NO OTHER MEANS OF TRANSPORTATION available to them.

    2. This route also serves the RAC office and location where our meetings have been generally held. While there is transit service on Rainier Avenue South (two blocks to the west) and on ML King Jr Way itself, it is NOT easily accessible, due to the need to travel those two blocks (or climb a fairly steep hill). In addition, the Rainier Avenue service going southbound also requires passengers to cross Rainier, a rather busy arterial. Although there is a signalized intersection near the southbound bus stop, the pedestrian light generally does not allow enough time for a person with limited mobility to safely cross the street.

    There have been several attempts and reducing or eliminating Route #4 service south of 23rd and Jefferson—all have been adamantly opposed by riders in this neighborhood….n fact, when METRO had to curtail service on the route due to construction on the I-90 freeway in 1984, a number of people at both Center Park and the Lighthouse complained enough that METRO temporarily “solved” the issue by “creating” a temporary route (which operated until they diverted another transit route to provide the replacement service). While the service was somewhat adequate at the time, it is NO REPLACEMENT for the Route #4 service that is currently available, because this route serves many of the medical facilities Center Park residents need to get to on a regular basis, such as Harborview, and Swedish (Providence and First Hill. It also provides easy access to several major shopping destinations at 23rd and Jackson. It is for these reasons that the Route #4 service be left AS IS and not changed in any way.

    1. Residents of Center Park have the option of walking down Walker St (which has sidewalks) to Rainier & Walker (crossing Rainier at the crossing at Hill St) which will allow them to catch the 7 to downtown, the 48 to the commercial district on 23rd, or, during the weekday, the 9 to First Hill. Alternatively, they can walk to MLK and catch the 8 to the same commercial district, as well as Capitol Hill and the Seattle Center. Absent Route 4, residents of Center Park will still have possibly the best bus connectivity in the South End.

      We cannot provide fixed-route bus service to the doorstep of every disabled person in the city. There has to be a reasonable expectation that bus riders can walk a few block to their bus. For those people who cannot do that, Access is the appropriate service.

      1. I can’t remember what the crossing there on Rainier S. is like. I know that in the area there and further south one of my pet peeves is how dangerous crossing Rainier Avenue S. can be and that used to be true even at lights on MLKjr Way. However, that situation may have changed on MLK. There are certainly areas there around S. Judkins where a want to a bus headed downtown would not be possible without the #4. All I ask is that the analysis of these situations be done all at once.

      2. I think we should be improving the safety and accessibility of sidewalks and crosswalks in the area to make accessing high-frequency bus services like the 7 and 8 easier for all users.

        “All I ask is that the analysis of these situations be done all at once.”

        With over 200 routes in the Metro system, that’s a lot to ask for from one man writing these posts in his spare time. Please be reasonable.

        If you’re looking for something resembling “all at one” analysis, take a look at the details of the 600,000 service hour cut proposed by Metro staff. Indeed, routes like the 14 Summit, 2 & 4 on Queen Anne, 4 Judkins Park, among many others, were proposed for deletion, with resources from those deleted routes given to nearby routes to strengthen the transfer based bus network.

  25. The route also serves many social service providers. Also I think that buses that serve medical institutions will always be a little slower in loading and unloading. While there are disadvantages to living in an area with so many institutions it should at least drive various choices intransportation and respect for it.

    1. Great, Joanna. You’ve now written 47 posts about the “need” the 4’s tail serves. Problem is, there’s an average of 6 people on the bus!!

      Blows my mind that people can keep talking about many agencies along the route, and the many uses the route has, and the many who rely on it exclusively, when all of those “many”s add up to six people.

      Illogic does not build you a transit system. Or a city. Or a society.

      Grow a brain.

  26. Seriously, Joanna. Do you actually not realize just 23rd/Jackson to downtown alone takes three times as long on the 4 as on the 14.

    If you want downtown, 48/8 + 14 is already better. If you want First Hill, 48/8 + improved 3 would be infinitely better.

    Only the dumbest human being on earth would ride this current route end-to-end. And I’m done with Seattle pandering to the Dumbest Human Being on Earth Lobby.

  27. d.p., calling me names won’t keep me from asking questions. I’m not sure why you felt you had to resort to name calling. I would not have been shocked to find out that it took a little longer on the #4. However, your statements are totally untrue. From 23rd and E.Jefferson to 3rd and Pike the #4 route takes about 20 minutes and the #14 from 23rd and and S. Jackson to 4th and Pike takes 21 minutes. What part of downtown are you talking about. I think you could add 2 to 3 minutes onto the #4 time to account for the space from S. Jackson.

    When they both leave 3rd and Union coming to the CD, the 14 takes about 17 minutes to 23rd and S. Jackson and the #4 is around 15 to 17 minutes.

    Perhaps you were thinking how long does it take the 14 to get to Jackson and 12th or 9th. In order to make your case or have an accurate measurement you have to pick a similar spot for the end point, as well as the beginning spot. Where are you trying to go. By the way the #4 won’t take you to S. Jackson and 5th Avenue South.

    1. I wasn’t trying to get personal, but I’m also unwilling to maintain the socially-acceptable veneer of polite circular argument.

      You’ve now spent 26 posts arguing against fact and against logic.

      Fact: The 4 is slow and unreliable, much to the detriment of the First Hill-Jefferson corridor that gives it nearly all of its ridership.

      Fact: The government offices “served” by the 4 are all within two blocks of 3rd, and accessible from 3rd via a system of escalators/elevators thoughout business hours and beyond.

      Fact: The average bus on the 4’s tail has about six people on it, representing a tiny fraction of total bus usage in its own corridor. This suggests that most people in its service area already use the other, better options available… options which wasting money on the 4 prevents from receiving further improvements!

      Even your cited numbers are misleading. Sure, they’re scheduled similarly in the inbound direction… but which was is usually on-time, and which one never is? Check the outbound schedules, in which they try to compensate for the uphill slowdown by suggesting 20 minutes (to Jefferson, add 3 minues to Jackson) for the 4, while scheduling 18 or 19 for the 14. In this direction, the 14 is again more or less as advertised, while the 4 is routinely 10 minutes off-schedule any time of day and 15, 20, 30, up to ∞ off-schedule at rush hour. (And the 14 is hardly perfect either, with its slow 3rd S.->2nd Ext. detour to Jackson. Both routes offer many opportunities for improvement.)

      BTW, though I didn’t cite the 27 as a potentially faster 2-seat ride — it’s just too infrequent most times of the day — it routinely shoots from Pike to Yesler/23rd in 14-16 minutes. And half of that is just the slog through downtown. It provide a pretty great example of how east-west connectivity could function if Metro stopped kowtowing to every loudmouth who wants to “keep everything the way it is” and got serious about improving east-west speed and reliability!

      In one of your posts, you suggested that the southbound 7 bus on Rainier is not accessible because it requires crossing the street! Don’t you think it would be easier, and cheaper, and wiser to fix the crosswalk/signal-timing issues than to keep running an entirely separate bus route for the rest of eternity!? When you expend all your energies arguing to keep one thing exactly the way it is, you miss opportunities to use those energies toward improving something that could offer much greater value and benefit.

      Which leads to the bigger question of why you’ve spent all your time pulling out every possible defense, no matter how strained, for every inch of the current route and operations. You can’t possibly believe that the 4 works well as a transit line as it is. No one could possibly believe the 4 works well as a transit line as it is! Knowing that funds are limited, knowing that you can’t create better options when expensively malfunctioning options siphon the funds to do so, what could possibly be gained for you or your neighbors by defending this one?

      For personal reasons, I take offense when people appoint themselves “advocates for the disabled” on matters like this, when what they’re really doing is infantilizing the disabled community. Yes, physical impairments do create different needs that must be considered and addressed. But the disabled no more benefit from a slow, infrequent, limiting transit system than anyone else does. To tell someone that substandard transit — transit that actually limits the places that can easily be reached and the way the city can be used — is necessary “because of their disability” strikes me as incredibly “othering” and offensive.

      1. I would add that yesterday I walked around the city and county building complex on James between 3rd and 5th, and noticed a (perfectly flat) skybridge connecting all three. Therefore, from the stops at 3rd & James, a disabled person has completely step-free access between all of those buildings.

      2. That’s assuming non-employees can use them, and you don’t have to go through the metal detector to get in the building. I think the skybridge is used to take jail inmates to the courts, so they may not want the public using them, especially if it allows sketchy people (just barely out of jail) to use them too.

      3. Mike, the wheelchair-accessible routes (including King County’s tunnel under 4th and a tunnel under 5th to access Seattle Municipal Tower) are fully accessible to the public, as shown on this map:

        I don’t know about the metal detector question. I could see the tunnel access being outside of elevator and thus security control. Even if metal detectors are an obstacle, it’s still no reason to maintain a fruitless bus detour that requires wheelchair lifts (an equal obstacle) and lets off on non-flat surfaces regardless.

        If they ever get around to building anything in the Great Hole next to the Pioneer Square tunnel escalators, it will be required to have public through-access and will not need courthouse-style security controls.

        Bruce, the skybridge is indeed a non-public perp-walk. Also, I’m pretty sure it rests on the roof of the KC Administration Building without actually connecting to it.

  28. Here’s an outside-the-box thought.

    Something like *half* the traffic on this route is from James and 3rd to 9th and Jefferson. Even more of it is one or two stops off from that.

    Now, this signifies one of two things:
    (1) Two overlapping traffic patterns, one from James and 3rd to the east end, and one from 9th and Jefferson to the west end

    If it’s #1, this calls for breaking the route up into two routes. Although the “#4 unique tail” does seem to be not worth it… so we’d want to look at the numbers for the #3 in order to see what to do.

    (2) Lots of traffic which is just trying to get across I-5.

    If it’s #2, this is the patented situation for a point-to-point gondola.

    1. Let me be Bailo for a second…

      How about we “just” put a lid on I-5 from Mercer to Jackson?

      1. :-) Actually that’s been proposed a lot, hasn’t it? People really don’t like sunken expressway “traffic sewers”, do they?

Comments are closed.