/ CC BY-NC 2.0

Mayor McGinn’s new nightlife initiative has gotten some attention for its proposal to add more late night service. In the near to medium term, simply adding service is out of the question, and of course would trigger 20/40/40 restrictions, meaning 80% of the service would go to East and South King County. In general, new routes will be less productive than existing routes. Other sensible skepticism here.

However, planners can sidestep all of that by simply shifting dollars in the West subarea to beef up late-night routes. But does that make sense from a productivity standpoint? Answer after the jump.

First, there’s a road safety argument. Make it easier for late-night drinkers to get home without a car, and it’s reasonable to expect they’re less likely to drive while intoxicated. And of course, many night bus users are not partiers but people trying to get to jobs. I don’t know how to begin to quantify that, so let’s note the benefit and move on.

Metro’s most recent route performance report is from 2008, unfortunately before the last big round of route changes due to Link opening. However, it does illuminate some isolated instances where shifting dollars to night (defined as 7pm to 5am*, all days) might make sense.

Table 1. West Subarea Summary

First of all, the summary of the West subarea (table 1) indicates that night routes as a whole seriously underperform by all measures (page xv).

However, the elite night routes are very productive (see table 2 below)**. The performance metrics on these are more competitive than some low-performing West subarea routes both in the off-peak (9am-3pm weekdays, 5am-7pm weekends, table 3) and peak hours (5am-9am,3pm-7pm weekdays, table 4).

As someone whose days staying out late are long past, I’d personally lose out if daytime hours were shifted to 2am. But there’s a reasonable case to be made for it. If the Mayor is serious about the transportation aspect of the proposal, there are plausible asks he can make to the County.

Table 2. Most Productive West Subarea Night Routes
Table 3. Least Productive West Subarea Off-peak Routes
Table 4. Least Productive West Subarea Peak Routes. The 126 no longer exists.

* Unfortunately, the report doesn’t break out performance between evening and the wee hours, which I suspect is what the city is most interested in.

** Note that the top performer, the 8, has been substantially extended since this report came out, with unpredictable effects on performance.

37 Replies to “Late Night Buses”

  1. Thanks for posting that. I think some of the confusion of this argument is that everyone is talking about something different. I’m a big proponent of late night service on *Friday* and *Saturday* night. I certainly don’t know how many hours it would take but one last surge of service on those two nights at 2:15 or so for the highest performing core routes is what I’m talking about when I think late night service.

  2. I’m someone who would benefit from more late-night service. I’m not a partier either, but I work night shifts in Bremerton and live in South Seattle. I arrive in Seattle at 12:40 a.m. on the ferry, and I bike home. I’d much much rather take the bus (or light rail, of course) on Friday and Saturday nights, though, since it’s rather terrifying to ride through all the drunks driving around or away from Pioneer Square. I’ve been hit once and threatened (verbally and with intimidating vehicle positioning) more times than I can count. The service is spotty, especially that late at night, and particularly since they got rid of the routes after the light rail went in. It’s not at all calibrated for ferry connections, sadly.

    Sometimes I’ll pay the extra dough (and it is a LOT) to drive over on the ferry on Friday or Saturday nights for my safety and sanity, even though it’s less than a quarter-mile to work on the other side.

    So I’m excited about the potential here, and I know many other commuters during this risky time of night would benefit as well.

    1. Indeed, not all Night Owl riders are coming home from a night out. I use the 85 every Saturday to get to work. If I waited for the first 21 of the day, I would be at least 1/2 hour late. Also, during the week, there are a few people who use the 85 to get to work early and who also use it to connect to the first 560 in the WS Junction to get to the airport or to get to the first Link run from Stadium.

  3. This is some great information. It tells you quite a bit about the correlation between density and productivity when evening/night routes in the core outperform even some of our peak tunnel routes (hello, Route 301? -5.2?!?). Extending just these top performers by another hour (2, 3, 15, 49, 358 etc…), twice a week, is a well-targeted, relatively inexpensive use of resources.

  4. You make some interesting points Martin. But set against the likelihood of major service cuts in the next few years, I would rate the chances of expanding night service as low. It may also be more difficult and not as cost-effective for Metro to provide and staff more late night service.

    There might be another way. In the last state legislature cities got the ability to provide their own transit service within their limits if they wished. This was envisioned by Bellevue for their desire for a downtown circulator. But in Seattle it might be even better used for late night service. Seattle could bid out some routes and private contractors and even Metro could bid to provide them. The fare could be a premium fare for a premium service.

    1. Staff is a big question to me.

      Seattle actually has the ability to pay for some addition transit service–Bridging the Gap allows for “No more than 15%” of revenue to be spend on transit, but actual spending to date has been 8% according to the 2009 report:

      Of course, this might mean delaying some road or bike projects, both of which have exceeded their “no less than” required levy percentages.

    2. Well, if the city says, “please get rid of the 25 and provide more night trips in the city”, they represent both constituencies and Metro would probably strongly consider it, quite aside from the overall budgetary context.

      1. The 25 is such an interesting example.

        I’ve often spoken of my fear that Link, through excessive stop spacing even in the city, will provide nothing of use for urban dwellers on both ends of the transit spectrum: those forever stuck on lousy buses, and those anathema to using buses in the first place. I’ve even said that University Link would have been much better realized with a 15th-ish stop and a Montlake Library-ish stop.

        Recently, on a long walk from the Arboretum to Westlake, I strolled through the geologically isolated and willfully insular Mondlake neighborhood, most of which is served by the 25. Seeing the single family houses and getting a few hostile “interloper” glances from the residents, I could immediately tell that this was not a neighborhood that rides the bus… ever.

        Funny thing, though… the whole neighborhood is within a 10-minute walk of where I suggested a Link stop should be. If these people had been given the chance at a 10-minute walk + a 5-minute ride to downtown, every last one of them would have become a transit user. With no more need for the 25.

        As medium-to-low density as that area is, its population still numbers in the thousands. That’s the kind of opportunity that Link has abandoned with its “must go far without stopping” model!

      2. I actually meant “topographically,” but I was writing quickly (and there’s no editing feature).

        The Interlaken ridge and Portage Bay are the main barriers, of course. But there’s also a surprising in-neighborhood drop-off from 19th to 18th, which gives much of the neighborhood a “hidden” feeling.

        Still, it’s just one flight of stairs and a few blocks up to where a Link station probably should have been.

      3. A Link station there would have been less than a half-mile, or 7 to 10 minute walk for most people, from UW Station, serving a relatively tiny base of riders for hundreds of millions of dollars. I agree that the people in that neighborhood generally will not ride buses, but it is simply not worth it to spend so huge amount of money for so few riders. However, a streetcar along the length of the 48 corridor south of UW Station should certainly be looked at.

      4. More like 2/3 of a mile between the two. And it’s not just the people in the right at that 2/3-mile point — it’s the many thousands of people (and the multiple recreational destinations!) east, west, and further south from there, who would be within a mile of a Montlake station but are well out of walking range of the Husky Stadium station.

        Again, it’s not about counting individual residents block-by-block. It’s about creating a system that is comprehensive and provides flexibility of movement. That’s the only thing that turns non-riders into riders, and it’s what transit planners don’t seem to be grasping right now.

      5. And why does it have to be a “hundreds of millions of dollars” station or nothing? That false choice is the greatest pitfall of latter-day transit design. A less-used station can and should still be buildable with a smaller footprint and cheaper.

      6. Why “hundreds of millions of $”?
        Because a station south of the canal would of necessity be fairly deep, and would, like all Link stations would require escalators and elevators, would need platforms long enough to fit a 4 car train, etc, etc. Link is a project that will be in use for up to a century – it should not be built “on the cheap” ever.

      7. The depth of the tunnel under McGraw and 20th is about 200 feet below the surface. The depth of the tunnel under 15th by Volunteer Park is about 330 feet below the surface. That’s like the equivalent of building two First Hill stations! They couldn’t even get the First Hill station which is situated in one of the densest neighborhoods built. The grade of the tunnel from Capitol Hill to under the Ship Canal is over 4%. You couldn’t put two stations in there without major changes to the design. It would be an engineering challenge to modify the routing to accommodate two new stations and it is not going to be cheap.

      8. Oran and Lloyd,

        All of what you say is true. I didn’t mean “cheap” so much as “less exorbitantly expensive.” A station can still accommodate 4-car platforms and elevators without the need for majestic arched ceilings and giant mezzanines.

        Oran is correct that the depth of stations make them inherently costly, but every cubic foot you shrink a station would nevertheless save a ton of money — and presumably decreases the size/power/imposition requirements for ventilation structures to boot. Not every station is ever going to attain the same boarding numbers as Westlake; building at an appropriate scale is vastly superior to building too few stations, period.

        That said, a prize goes to anyone who can provide a truly persuasive case that this:

        …having worked well for 100 years, needed to morph into this:

        …by 2010.

        (My overarching point remains that every potential station the powers that be — who, frankly, don’t seem to know the slightest thing about living in a city with working transit — deem unnecessary exponentially decreases the permutations of potential trips possible on the Link system. This will come back to bite Seattle!)

      9. The 25 is a pretty wonky route, but it’s not just Montlake, it also provides lifeline coverage to Lakeview Blvd and Laurelhurst, admittedly also filthy f’ing rich communities, and I preferred it to the UW campus routes to get to U-Village (no going any further south than 45th) until I realized how close the 30/74 gets. (Admittedly it took me a while to realize how close the UW campus routes get too – when I was getting home from track practice at Nathan Hale using the 65-to-67.)

  5. The biggest gap in night owl service is 85th to 205th. At least the 358 should run 24 hours like the 124/174. As it is, people in Burien and Bellevue Way have night owl service, while people in denser places like Northgate, Lake City, Oak Tree, and Shoreline don’t!

    1. Tell me about it–that gap is a remaining 60-year old vestige of the old city limits. I’d love to be able to take transit to the airport for flights at 7am or earlier–but it isn’t possible to get from either my home or the Northgate TC downtown early enough to catch the first or second Link trains. Trip planner actually tells me to leave at 12:30am and take 4-5 buses to accomplish this.

      I fly often enough (and others do too) that even a 4 or 4:30am bus would give us a choice. The first eastbound transcon flights all leave between 6 and 7am and it would be nice to take transit rather than drive/park, cab or Shuttle Express.

      1. Especially once North Link goes in they should have an hourly Link shadow between 1 and 4. Not to mention that they should run until 2:30 on Fridays and Saturdays.

      2. Tell me about it–that gap is a remaining 60-year old vestige of the old city limits.

        QFT. There’s really no reason why the 82 shouldn’t serve Northgate TC (just copy routing from the 16 and 5, possibly replacing the NSCC route with a simple Northgate Way run if there’s no one there that time of night) or the 83 shouldn’t serve Lake City (right now there isn’t even any service north of 80th east of Wallingford Ave – how about following 72/79 routing to Lake City and then 35th to the current 55th/25th loop?), unless that would jeopardize their ability to make the 3:30 departures, but the 83 is getting back to downtown at 3:17 while the 82 is getting there at 3:22, so the former can afford to add a few more minutes to the route (especially cutting the 20th and 65th legs). Admittedly timing is more of a problem for the 82, but maybe it can skip Queen Anne in one direction on the first trip too? The 83 doesn’t have so much time to work with that it would be easy to do both… Or perhaps one or two late 41 runs?

        (Aside: people attending U-District bars currently have to wait half an hour for the 83 to show up, and if they want to go to Capitol Hill they have a 45-minute wait for the 49…)

      3. Great ideas! Particularly (for me anyway) looping the 83 thru Lake City and back to 55th on 35th. I think the 41 should be beefed up a bit at evening and night anyway, until North Link reaches Northgate. That’s a trunk route if there ever were one and should be every 30 min until a couple of owl runs late at night. There’s (finally) some TOD at Northgate and the transit service should encourage use. Right now after 10 there are a couple of hourly runs, then nothing. Fun to walk out of a ball game at 10 and find out you have to wait an hour+ to get a bus back…kind of defeats the purpose and put me back into my car. :(

      4. Sounds like you could do with just a 41 leaving International District station at 10:42 (with corresponding bus leaving 130th and 35th at 10:18) – then maybe Metro could cut the weeknight after-Mariners special bus to Northgate, since it would be mostly unneeded unless the game went deep into extra innings. But good luck getting more bus service at all with Metro’s budget crunch… I imagine you’re disinclined to settle for the 72 that leaves ID station at 10:53 but is at least scheduled to reach 130th at the same time – less time waiting but no time savings and probably a more crowded bus until you’re out of the U-District.

        (Apropos of nothing, but how come the 11 timetables don’t include the Pine/Madison leg of the 84 that basically follows the same route outside downtown?)

  6. You make it sound like it would be a bad think that 80% of new service would have to go to the east side. Right now there is (afaik) one route on the eastside after 12:30– the 280. On the other hand there are a ton of busses that run in Seattle at night. So yes, maybe routes on the eastside wouldn’t be as productive (because they’re lower density), but tell me how I’m supposed to get anywhere to or from Redmond after 12:30?

    1. I don’t see any reason to assume that new transit hours would be used to add night service on the Eastside.

      1. That comment is based on this in the original post:
        “In the near to medium term, simply adding service is out of the question, and of course would trigger 20/40/40 restrictions, meaning 80% of the service would go to East and South King County”

        This makes it sound like it’s a _bad_ thing that 80% of service would go to East and South King County. All I’m saying is that there is almost 5 continuous hours every day where the vast majority of the county has literally no service, and adding service to those areas is a good thing.

      2. Your hitting the question on the head. Do we continue assigning service hours based on politics and a notion of geographic “fairness?” Or do we add bus service where it will attract the most riders and get the most bang for the taxpayer’s buck?

      3. The whole point of this post is that you don’t have to solve larger equity issues to support late-night ridership.

      4. But what if I think we SHOULD solve the large equity issues. Fundamentally, public transit is not distributed purely based on usage. The reason we have hourly service in the more rural areas is because of this–we agree that there is some baseline level of service that everyone deserves access to. This is why public transit is not a for-profit venture, it’s funded by the government and taxpayers precisely because we believe that people should be able to get around w/o a car, not because we believe we can get more usage.

  7. I would assume any increase for Seattle would have to be Seattle funds contracted directly to Metro for the service, rather than just asking Metro to increase service. Those funds could come when budgets return to normal, or it could come from some new tax or fee tied in some way to Seattle Nightlife under the proposal being made.

    There’s a few routes that already run into the 12:00am and 1:00am hours — usually on 60-minute frequency at that point. It might not be too costly to simply identify some of those key routes and add an additional bus run or two and a couple more hours of driver pay.

  8. The lack of a Link shadow route dampens the ability of airport employees to take Link when it is running, instead of driving their cars. I wonder if Metro or ST has attempted to measure the number of riders they are missing out on because of the 3-hour hole in service.

  9. I don’t think the lack of a Link shadow has much of an impact on transit usage for airport employees.

    Airport employees are generally split between the Airline Jobs and the other jobs at the Airport, ie TSA, food service workers, and other vendors. Most vendors work most of their employees during the day because that’s when there are the most passengers and they tend to use daytime transit service. However, most Airlines schedule employees to work Day or Night shifts. Day shifts start very early, ie before 6:00 AM, and there is very little transit service at that time of day and there is very little ridership. Night shifts usually end around after all of the red-eyes are out, usually around 12-2, and ridership is virtually nonexistent.

    Don’t forget the Port of Seattle provides Airport Parking for employees and many companies will subsidize the monthly fee of $20 a month. It’s much cheaper and the frequent shuttle between the parking lot and the terminal makes it very compelling to choose single occupant commuting. The 3 hour hole also existed before Link opened up. The 194 and 174 didn’t run continuously as far as I remember

    1. ST has addressed the airport employee issue in the last 2 shakeups, There is a trip now (Since February) leaving the Stadia station before the DSTT opens that arrives at the airport shortly after 5am. (there is one earlier than that, but it starts at Beacon Hill.)

      I do not see how the Mayor (a city entity) can change Metro’s (a county entity) route and scheduling by decree. The county has a revision schedule that rotates around the region annually. (I don’t remember where they are this year.)

  10. One option for better late night service is to stretch out the last few runs.

    Most routes have a standard evening frequency, such as every 30 minutes, and they run at that frequency until the last run, say at 12:30am.

    A few routes however, such as the 120, run every 30 mintes until 10pm, then every hour til 1pm, with a final run at leaving downtown at 2:19am.

    More 2am bar-closing service could be added without any additional service hours by speading this service pattern througout the system.

    1. Of course the question is, which way would get more riders for those given service hours? The information in this article is basically worthless for determing that. Lumping ridership from “7pm to 5am” doesn’t tell us a thing about how many people ride buses after 11 pm or after midhight, or after 1 am, etc., which is what the real question is.

Comments are closed.