Photo by Oran
Photo by Oran

Most people think of Seattle as a world class, major city. Major cities have 24-hour transit service, as should Seattle. There is limited overnight service, with the Owl routes (82, 83, 84) being understandably derided because they are the most confusing. With the new hours from Proposition 1—approximately 230,000 annual service hours are left after the June improvements are implemented—we have a chance to accomplish this for the people who keep odd hours. My bias for evening and night trips is showing because I’m one of those people; this post is actually being submitted at around 4am.

Why do this?

As Mike Orr points out, transit use increases when it can cover more trips and more types of trips. Right now, our transit system is pretty good at covering regular commuting hours and trips taken while the sun is usually up. Service drops dramatically in most corridors after 7pm and especially late night. This is also the first type of service to take a hit when cuts are proposed since the “bang for the buck” in the reclaimed service hours is quite high when shifted to daytime service.

However, if we are to have “the best bus service we’ve ever had in Seattle,” ($) night service should be one of the first things to come back and be enhanced. Lots of workers are employed during these odd hours and we can accomplish a lot of trip diversions by giving them the tools to get to and from work. Other trips for activities in the evening and night periods can be made attractive to do via transit and without having to rely on an overpriced taxi or a potentially-surcharged rideshare company.

What to do?

In considering which routes to extend to 24-hour service, I looked at routes that meet these criteria:

  • Their span of service ends, at a minimum, around midnight. This is to avoid having to extend service beyond a few trips and because the routes with later service already tend to serve denser areas.
  • The route can simply have more trips added without having to divert in the middle of the night. By doing this, the confusing mess that are the existing Owls is avoided. (There are two exceptions below.)
  • Cover as much of the city as possible, with special emphasis on getting more service north of 85th Street, which is where the current Owls (except D line) end, and cross-town routes to avoid everybody having to go downtown.
  • Look solely at routes that can be paid for by Prop 1. This means no overnight service on routes like 255, 545, or 550 is discussed here.

Half-hourly service would be much better, but given that service hours are difficult to calculate, this proposal is cautious. If we have the hours, half-hourly trips would be much more convenient.

What routes?

These routes are, in my view, excellent candidates to add extra round trips as soon as possible. Adding these trips would also mean completely deleting the 82, 83, and 84.

  • 3S + 13N: Route 3 already stops after midnight and starts up around 5am both weekdays and weekends. Four trips each direction would be needed to for this corridor to have hourly overnight service.
  • 11: Like Route 3, Route 11 stops after midnight and resumes before 5am every day of the week. Weekdays would need three extra trips, Saturdays four, and Sundays five. Adding trips to 3 and 11 covers the 84 night owl but without the strangeness of “which trip goes to Madison first?”
  • 16: This route is like Route 11, down to even the same number of trips needed. It travels surface streets, would provide a link to Northgate, and covers most of the current 82.
  • 40: Hello, Ballard. Some would argue that D line is a suitable substitute for route 40 owls. I disagree, since 40 travels inside neighborhoods, through NSCC, and into the night life area of Ballard. Four trips on weekdays, five on weekends would get overnight service.
  • 43 + 44: These already run as both a coupled pair and deep into the night. Only a couple of more trips would be needed for cross-town hourly service covering Ballard, Wallingford, the University District, and Capitol Hill.
  • 48: The 48 is, as most know, the longest route that does not go via downtown. It is a vitally important link to neighborhoods not served through the CBD. With just four trips on weekdays and Saturdays, and five on Sundays, approximately 15 neighborhoods would gain legible Owl service.
  • 49: Much like the 43+44, the 49 covers a very dense, popular deep-into-the-night area of the city and only needs a couple of extra trips to have 24-hour coverage.
  • RapidRides C and E: Each needs one more round trip to have service every hour of the day. There’s really no reason not to, especially since the D does make these trips.

The odd duck but very needed:

  • Route 41: I think the service area for this route would be great to have but it comes with a huge freeway section in the middle that bypasses a lot of neighborhoods. To that end, I like William C.’s idea and so propose adding four trips every night to the 41 but routing them via surface streets instead of IH-5. Call it the “41Night,” if you will. It follows the 41’s routing in downtown along 4th Ave, goes up Eastlake to follow what would be the 66 along Roosevelt (we could also swap in the 70 if a routing closer to the center of the university is preferred, but UW already has an overnight service from the university) up to Northgate, then resumes the 41’s routing to Lake City.

At the pie-in-the-sky end of the spectrum, and at the risk of reestablishing a standalone Owl route:

  • Route 97: A last-minute brainstorm, but restricted to Rainier Beach since paying for the route to go all the way to the airport is likely outside the scope of Prop 1 funds. As a substitute for having 24-hour light rail service, run round trips on 97. Weekdays and Saturdays would need three trips, Sundays four trips. This accomplishes two goals: it links together the neighborhoods along the light rail route and it sets the precedent for 24-hour light rail service to eventually happen. I debated using the “as needed” route 97 versus route 8 but if 8 is going to get restructued–which it really should–we avoid confusion by having a single “only operates in the middle of the night but does it because light rail itself is closed” route.

I’m certain I have missed some areas and I’m definitely not an experienced trip planner so there are bound to be inefficiencies. Unfortunately, the route list is still downtown-heavy because there aren’t many cross-town routes that meet the “run later than most” requirement. Suggestions are definitely welcome, though I truly hope that the idea itself is worth supporting.

95 Replies to “Transit: Up All Night”

  1. As Joe says, commenting here to get notifications by e-mail.

    I suppose I should also point out: I comment on STB as “lakecityrider” but couldn’t figure out how to put my nom de plume in this post’s display name.

    And, finally, since I am up all night, I won’t see any of the potential replies until much later this afternoon.

  2. Three points.

    You didn’t mention the 7, which is an existing owl route; did you forget it or are you planning to delete it? Given Rainier Valley land use patterns, I think it’d be much more popular than your 97N, as well as having an already-established ridership.

    Also, I’d extend the 124 one more stop to the airport. Having some airport service is very important, and if (as people are saying in the 120 thread) Prop 1 measures the number of stops in Seattle instead of the number of service hours, it’s very possible.

    Finally, this – and any other plan for hourly or even half-hourly owl service – is highly reliant on timed transfers. The Ballard and Capitol Hill nightlife are only served by two routes each; anyone living off those will have to transfer to get home. (That’s the one advantage of the current 84: it connects Madrona to Capitol Hill with a one-seat ride.) Something significant will need to be done to enable transfers outside downtown, especially given the 48.

    1. The 7 didn’t get listed because it doesn’t need to be changed; it already meets the bar. I should have made that more clear. Also, if Seattle can pay the freight to run the 124 down to the airport, awesome, and that sounds even better than truncating it and trying for the 97.

      Transfers are a big consideration which is one reason why the downtown routes are actually kind of a bonus. Have them do like they do now; everybody runs downtown and meets at around 2:45 or 3:30, holds for 5-10 minutes to let passengers shuffle around, and then take off.

  3. Northeast Seattle is the most difficult to serve with a regular route, so let’s look at why. There’s a large undense area the size of Rainier Valley east of 15th and south of 125th. A route from the U-District can’t go simultaneously north to Lake City and east to Ravenna (65th), and we haven’t even addressed Laurelhurst/Sand Point at all. Instead of a 41-night, I’d suggest either replacing the 83 with the 72 (deleting Ravenna), or changing it to a 65th-35th-Lake City route (deleting 15th north of 65th). The latter would probably serve more riders and compensate for more hills.

    I also thought about extending the 40 to Lake City (which some have argued should be done all day), but that loses the Lake City – UDistrict market, which is presumably larger, more direct, and better for multi-seat trips. Of course if we have enough money we could do both.

    Back to Laurelhurst/Sand Point. The simple answer here is to run the 75.

    Metro has positioned the 49, 7, and 36 as the favored routes in their districts (longer frequent span than neighboring routes), so I would stick to those. (Although longer term I’d restructure the 49 and favor the 43 instead.) The 11, well, it is further away from neighboring routes but Madison Park is lower density and affluent. I guess truncating it at MLK would be too much quibble about a short distance.

    “85th Street which is where the current Owls (except D line) end”

    E, not D.

    Route 97 gets into what kind of shadow should Link have. That affects more than nighttime. A bus on Link’s route has to backtrack or skip stations. If the 8+106 restructure goes into effect (unlikely because Metro staff said it was primarily compensation for cutting the 14), it would cover downtown to Rainier Beach minus Stadium and SODO, and the 43 would cover downtown to the U-District. An 8+106 could replace night service on the 7 and 36 but that would be highly controversial, and should probably be in the context of reorganizing the 7 if that ever happens.

    “night service should be one of the first things to come back”

    There’s no “back” to come from. There’s twice as many night owls as there were in the 1980s, and the only thing that’s been cut is the 280. The 82/83/84 were going to be cut in September but were saved by the city (prior to Prop 1). In the 1980s the night owls were 7S,81,82,83,84,85,280. Now they’re A,C,D,E,7,36,49,82,83,84,120,124,180; and honorable mention to the 44 and 150 for at least reaching 1:30am.

    1. The 75 would have been perfect but I discarded it because it ends too early. If it gets more service, that would make it an ideal candidate.

      Oops about the E. I had a 20% chance and botched it. :)

      As for night service “coming back,” that’s why I also said it should be expanded. Night trips after 7pm keep taking the hit whenever service cuts are proposed. The owl service we have doesn’t even really rank as “ok” to me; it works, but only if you’re really willing to push to ride the bus and invest a minimum of an hour and a half in the trip. Besides, it would be great to see the city and county take an interest in what happens after most people leave the office.

      1. I actually thought you had it with the D/E, since the D’s terminus is where 100th, 8th, 3rd, and Holman meet and you didn’t seem to consider the E a “true” owl route…

      2. I had actually just forgotten about the E when I wrote that section. Both D and E go north of 85th (though D just barely). E doesn’t run a trip at least once an hour every night so that’s why I put it on the list.

    2. Mike,

      I’m sorry, but “Laurelhurst/Sand Point” is not a night transit market. Folks living in Laurelhurst and Windermere have made their choices: they want an auto suburb. Sure, they ride the bus to work to save money on parking and the hassle of getting through Montlake, but they’re not going to ride at night.

      Wes didn’t propose it because the Metro planners have already noticed that they’re running Norman Portraits on the run in the earlier evening hours. Why compound the felony?

      I know you and Ross are concerned about complete coverage, but the city can’t afford it given the “starve the beast” attitude toward government that dominates in this state.

      1. The U-district, Northgate, and Lake City should have decent night and owl service. Wedgewood, Sand Point, etc. Don’t do a great job filling busses at times of day other routes are crush loaded. Perhaps that will change at some point in the future, but not any time soon.

  4. I like your list for the most part.

    A few random thoughts:

    The 48 seems a bit pointless without timed transfers, though I see the arguments for having it as an owl route.

    I’m not sure about the 16 I think an owl 5 might be a better idea.

    For. NE Seattle a combination of the 66 and 72 would provide pretty good coverage.

    Some additions might be the 5 mentioned above and the 73 which brings in Pinehurst and Jackson Park.

    With the E, 5, 72, and 73 you have 4 routes reaching near or at 145th.

    1. That said, while increasing night-owl service should be a long-term goal, I think we should focus first on the earlier parts of the evening that carry more riders, but are still vastly under-served. Lots of concerts and sporting events end between 10 and 11 PM, and it is all too common to see completely packed buses leaving downtown at these hours, but still running with only hourly frequency.

      When we are able to maintain 15-minute frequency on all the core routes to 10 PM, and 30-minute frequency on all the core routes to 1 AM, or at least midnight, then, I believe it makes sense to invest in more owl service between 2 AM and 4 AM.

      In the meantime, though, I am all in favor of making cost-neutral restructures of the owl routes we already have. For instance, the special owl routes can an should be replaced by owl trips on all-day routes. For instance, the version of the 73 that goes only to 65th St. could probably replace the 83, using the existing service budget.

      1. I disagree with putting cost neutral changes at the front of the list for anything at this point. We have 230,000 or more of annual service hours. My back-of-the-envelope calculations (thanks, Eric G.) say that we can do hourly owl service on the routes I listed for between 15,000 and 25,000 annual service hours. That’s about 10% of the hours left unallocated after the June service improvements.

        We should have both. 15-minute core route frequency until midnight or 1am and, in my opinion, 30- or 45-minute frequency on those routes for the entire 24-hour day. With prop 1, we can afford it and more.

      2. How much would the 15-minute core frequency cost, though? I agree with asdf2 that’s a vastly more important priority. And, there’re still some corridors which demand more service in the daytime. If there’s that 10% left over at the end, great!

      3. But is daytime really a vastly more important priority? This is service that doesn’t exist right now, at least not in a usable fashion. People were ready to cut the Owl 80s because they serve a virtually useless purpose. They are still almost useless, even though they were saved. There are needs all across the network, but I also point out that those of us who rely on after-7pm service and non-commute service keep taking the brunt of the cuts, especially as a percentage of hours of service being cut. At the same time, we pay the same taxes yet get vastly worse service out of it. Look at the drastic cut proposal that got averted; post-10pm service would have been gutted. The evening and night service periods and those riders shouldn’t be at the end of the line for additions.

        Absolutely the rest of the network needs investment and we have the hours to do it. (Everett Transit runs its entire system on 150,000 annual hours, for example.) I’m looking forward with great anticipation to the service additions in June. I also badly want it on the table that there are very supportive transit users who have pushed for and voted for expanded transit measures for a while and we need an upgrade, too.

      4. I agree, the off-peak service – especially the evening service – should be the top priority. What I meant to say was that frequent evening service (till 11 if not 12 or 1) should be ahead of a large investment in overnight service.

      5. Ah, in that case we are in more agreement. I would put more daytime service ahead of 15-minute frequency until 1am, for sure. By the same token, I don’t think that the number of hours possibly needed to run all night service is too much to ask. We can scavenge some from dropping the 80s and I believe it would do a good amount to increase support for transit in a community that can’t really use it right now.

        I’m trying to crunch the numbers now and will have a better idea of how many the hours the trips could take sometime next week. Probably should have done that to include in this post…

      6. The daytime level of service (especially on weekdays, when everyone is at work and can’t use it) is actually pretty good. Except for a few cases of severe peak overcrowding, it is evenings and Sundays that need the most improvement. 30-60 minute headways simply does not cut it, when you have lots of people getting out from large events at the same time. If nothing else, the limited number of people that can fit on a single bus makes it physically impossible for transit to carry more than a small percentage of the people home at those frequencies.

      7. Yes, the frequent network is actually pretty good and mostly goes to the places it should. The problem is its short span. The 40 southbound turns half-hourly right at 5:30pm as if it can’t wait to ditch the frequent schedule, and I can’t believe ridership precipitously falls then. I got caught in that when i took the 75 to Northgate and assumed the 40 would continue frequently until after 6 since it’s such a high-ridership route, but I ended up just missing the last frequent one.

        There’s something similar in the Shoreline peak-express routes. When I was drawing up feeders for Link a few years ago, I discovered that the routes are mostly in the right places for feeders and crosstown service, it’s just the span that’s too limited.

      8. Since 40 is one of the routes that hits all three items for improvement in June, I included it in this list because it should have a bunch of new service added rather shortly. Because, yep, 40 needs 15 minute service until at least 10PM…

      9. Mike, you forgot that a lot of the people that ride the 40 have Metro tearing their hair out going “Why aren’t they riding the D??? It’s so much more awesome ’cause it’s delayed even more than traffic on 15th would suggest to maintain an arbitrary distance to other buses, slogs through Uptown rather than take a coherent route, and misses the core of Ballard while stopping short of Northgate! AND THE BUSES ARE RED!!!!!”

      10. There is demand even at night out to at least 80th if not further.

        For a service hour neutral swap I think substituting the 72 for the 83 would be a better idea than the short-turn 73.

      11. “a lot of the people that ride the 40 have Metro tearing their hair out going “Why aren’t they riding the D??? It’s so much more awesome”

        Metro is glad that the 40 is so successful. It’s one of Metro’s biggest achievements, comparable to the 8N and 31/32. Many of the 40’s riders can’t take the D because it doesn’t go to Fremont or Northgate. Others take it because they live near 24th. Metro knew those people would take the 40 when it created the 40. That leaves only a minority percentage of people who live closest to the D but take the 40 instead — the ones Metro might be tearing its hair over. But Metro says the D is overall quite successful. So I don’t think it’s tearing its hair much, and it wants to extend the 40’s frequent span as soon as funding is available, which means June.

    2. I live on Greenwood Ave, so I would personally benefit from having a night-owl 5. Even so, I don’t see how Metro could justify running the E *and* the 5, while only running a single bus to all of West Seattle, and a single bus to all of SE Seattle.

      Even if we assume that Metro could somehow afford to run that much service, I would much rather have 30-minute service on a single route than 60-minute service on two parallel routes.

      1. The E is clearly the higher priority.

        I was mostly thinking of the routes the 80-series owls provided coverage on.

        It’s been ages since I was on an 82 but I seem to remember most of the riders got off by Fremont and the few who didn’t got off at Aurora.

    3. In order to be convenient, transit must have exclusive right of ways, be reliable, and above all frequent. Routes that serve the urban cores of Seattle should have the following head ways to be considered world class service :


      5am-7am: Every 8 minutes
      7am-9am : Every 3 to 5 minutes
      9am -3pm: Every 7 minutes
      3pm -8pm: Every 3 to 5 minutes
      9pm – 11pm: Every 9 minutes
      11pm -2am: Every 12 minutes
      2am -5am: Every 15 minutes

      5am to 9am : Every 9 minutes
      9am to 9pm: Every 5 minutes
      9pm to 1am; Every 9 minutes
      1am to 5am: Every 15 minutes

      1. That would be amazing. I’d not only sell my car and turn in my driver license (trading it for an Enhanced ID card), I’d get out an campaign for an income tax to pay for that.

        On the other hand, and it pains me to say this as the person who is the pie-in-the-sky dreamer in this thread, that’s sadly not gonna happen. It would probably cost double the current budget for the entire system.

      2. With the current network, yes.

        Which is a pretty good shorthand for why the current network sucks.

        Think about it: how many times have you waited for a bus up Capitol Hill. It’s nice enough out and you have enough time to spare that you can really hop on anything that comes, and walk from the closest place it gets you. 10, 11, 49, 43, 47, even the 2. Just get me toward where I need to go.

        And then it still takes you 15? (20? 30?) minutes to get away from 3rd Avenue.

        The world is full of transit systems like Alex describes. Most of them are in cities denser than ours, but many of them are in places less populous than us. What none them do is what defines us: fragment the fuck out of your service, until you can barely run anything well… not your core network, not even your gloriously branded “red standard” lines.

        When you say it can’t be done, picture the Capitol Hill conundrum. Only one subway station, but seven separate downtown-bound routes?

        Demand-math isn’t the enemy of gold-star transit levels. Nor are tax foes. The enemies are the Cruickshanks and Cullens of the world.

      3. How would you consolidate the Capitol Hill routes? Would it achieve Alex’s standard? Which areas would lose service, and how close are they to an alternative?

  5. These are all great ideas, and I hope Metro and SDOT pay attention. That said, I don’t like the idea of the 97. ST just has to shape up and run Link 24×7. Once Link reaches Redmond and Lynnwood they’ll be pretty close just with getting trains back to/from the yard, so why not do it now? As it is, a single train could serve Link at 90 minute frequencies at night.

    1. ST’s maintenance crews work on the track and systems during the 3 hours the line is powered down every night. Stopping now might void the warranty. But even after the warranty runs out, I want this piece of infrastructure to last as long as possible.

      If running night owl shadow buses is the price of getting maintenance done, so be it. But YES, that should be a high priority for graveyard-shift investment, given the large number of blue-collar graveyard workers there are in and around the airport.

      1. Would it make a material difference to single-track the line and run one train back and forth on it? Not being sure what maintenance ST is currently doing, I can’t answer for sure – but their periodic need to shut down the whole thing for even more maintenance makes me think it can’t be very much.

      2. Doing maintenance on one track while a train occasionally passes by on the other is not a job I would sign up for.

        We’ve talked this in circles many times, with no hint from ST that they are interested in finding a way to run trains 24/7. Indeed, the 2015 Draft SIP doesn’t even bother to mention night owl service as a long-term goal.

      3. I’m not aware of any other rail infrastructure in the country that requires 3 hours of maintenance every single night. If it really does require that, we should question the feasibility of all other ST rail projects.

        At some point, though, ST won’t have a choice but to run trains 24×7, even if it’s just for dead-heading back to the rail yard. Depending on the actual alignment, it seems it would take upwards of 45 minutes to get Lynnwood trains back to the Sodo rail yard, which would reduce track maintenance windows to around 2 hours.

        When periodic maintenance of the tracks is required, then we absolutely should have shadow buses, but it should be an exception, not the norm.

      4. …any OTHER rail structure in the country…

        I hope we follow more European and Asian models of how to do track maintenance. The US&A is not the industry leader.

        The last outermost trains will be returning to O&SFs toward the other termini.

      5. 24-hour Link was in one of ST’s long-term lists a few years ago. It may have been dropped or it may still be there somewhere.

        Can’t they just shut it down whenever there is maintenance rather than every night? That’s what the NYC subway does.

      6. Void what “warranty”. Are you saying that the track structure has a warranty? From whom, Tupak Mancini? If so, it’s worthless.

    2. Rather than a 97 at night, the thing to do is run “rail shadowing buses” at night, and also whenever the line is shut down for other reasons. This is done in cities around the world. “Owl Link” right down MLK is very legible. Not sure what to do about Beacon Hill.

      1. FYI, the 97 IS the route number for the rail shadow bus they already run when Link is disrupted or closed. This proposal would modify it to run only within city limits overnight due to funding constraints.

  6. Wes, you said “I have no idea how to calculate exact service hours.”

    This is actually very easy.

    The old way is: Total the length of the trips you want to add (in minutes)/60 = Daily Revenue Hours, then multiply DRH by the # of days that variant operates in a year for Annual Revenue Hours. Aggregate the variants ARH for a Route Total and aggregate the routes for a System Total.

    The new way is:

    Cheers! and, great ideas!

    1. Thanks a lot, Eric. I’m playing with this site now and having a lot of fun at it. :) Sure beats doing the math by hand.

    2. Question – does Metro have to pay bus drivers a higher hourly wage to work in the middle of the night than the middle of the day? If so, that would be an argument for having more day service and less night service.

      1. In a similar vein are there restrictive work practices in the union contract that make it difficult to run that many all night buses efficiently. [My understanding is that this a problem with adding more Sunday service]?

        Are there overhead costs — e.g., adding more dispatchers — that make the new hours disproportionately expensive? Do we have safe layover with acceptable bathroom access, and which won’t result in neighborhood results: I certainly wouldn’t quietly acquiesce to a bus idling in front of my house at three in the morning?

        How do we know there’s actually demand for this service? While I agree with others up-thread that there are higher priorities, I’m certainly willing to give these runs a chance if the money can be found. To be specific, I have two questions. How long should we give these buses to attract customers? And, what performance metrics should we require of them?

  7. Major cities have 24-hour transit service

    This is just simply not true.

    Most Seattle-sized cities around the world — never mind Seattle-sprawled cities with Seattle-low density and Seattle-anemic transit modeshare — do not run 24-hour transit, though they may run late-night weekends.

    What they do tend to have, on the way to successful growth and car-free living, are cohesive and legible service networks with spontaneous-level frequencies, running the length of their normal service spans (5:30-12:30, or perhaps even 5:15-1:15).

    You won’t find a single example of a successful transit city with evening and weekend service as shitty as ours, and also with all-night transit that anyone electively uses. Screw the latter until you’ve fixed the former.

    1. “Screw the latter until you’ve fixed the former.”

      I disagree for two reasons. First is that we are fixing the former; maybe not perfect, but let’s deal with that in just a second. Seattle, in raw numbers, will have hundreds of thousands of new hours of service to deploy starting next year. Breaking out the “feasible with the buses and employees Metro has right now” wish list and doing all of it costs, according to published reports, about 43,000 service hours. That leaves 223,000 annual hours in the kitty. I’m working on a follow-up blog post, but my initial numbers are saying that we could spend 200,000 more hours of service on anything we want, put in all of the bus trips I’ve recommended here and still have a few hours left over. I’m not proposing that these trips be added in June–though I wouldn’t turn them down–but I do want the discussion on the table.

      My second reason is more personal: There are a lot of us who use this service, whether by choice or necessity, and work the odd hours at jobs to do tasks and provide that a lot of folks in this area and elsewhere find very useful. We pay taxes, too, but we keep getting the thin end of the wedge. Repeating what I posted earlier in the thread: evening and night service are always on the chopping block. Yes, more route numbers run a couple of token trips at night, but they’re not useful for actually moving around. The proposed cuts, now thankfully averted for the moment, would have slaughtered night service by 60%. We’re due useful, legible, and timely service just as much as the folks in Magnolia who have seen their service chopped to the bone or the overcrowded RapidRides coming out of West Seattle and Aurora. If we keep waiting for the rest of the system to be deemed “perfect” or even “ok, that’s good enough,” we’ll be waiting until light rail reaches Bellingham.

      Ultimately, we can afford it and so we ought to consider it.

      1. While I do not doubt your earnestness, nothing in your response addresses what I actually wrote. It is frustrating to be made the bad guy for having to point out your argument is baseless from its very first word; that each and every one of your premises is inaccurate; and that your resulting suggestion that Seattle “needs” or “should have” or is anywhere near capable of supporting 24/7 mass transit service is simply not credible.

        Most people think of Seattle as a world class, major city.

        Really? Who? In terms of aggregate cultural and economic influence, it is at best a second- or third-tier city just on this coast! One must defer to the generically broad “Metropolitan Statistic Area” (i.e. the entirety of King/Pierce/Snohomish) just to shoehorn us into the top 15 populated areas in just our country. Our city borders, containing miles of generally low-transit-demand suburbs within, manage to rank us an unimpressive 22nd… with the caveat that this arbitrary ranking is followed by cities (Boston, D.C.) built densely to their borders and with urbanity spilling over into seven or eight adjacent municipalities.

        A regional dominance disproportionate to our nominal size (thanks to being in a remote corner of the country), and the handful of influential industries and global business playesr rooted here, are barely enough to provide Seattle a “Beta-minus” ranking on the scale of Global Cities, but you’re kidding yourself if you think this gives us an inherent urban “gravity” that is a wart on pinky toe of cities like Tokyo, Shanghai, or São Paulo, or even of far more urbanized and infinitely more transit-enabled “peer” cities from Istanbul to Boston… all of which offer either extremely limited 24-hour service or none whatsoever.

        Major cities have 24-hour transit service.

        Again, simply false. And exponentially false for cities with sub-par effectiveness and modeshare the rest of the time. If you can barely get 18% of your city population to use the system at rush hour, and your system is too faulty to convince 92% of the people to bother with it at any other time, why would you expect more than a trickle of the already-tiny number of trips made in the middle of the night?

        Again, I dare you to find me a city with such a poor span of usable regular service that offers non-political overnight service (i.e. service with sustainable demand, and not driven by inertia and warped priorities, like ours).

        There are a lot of us who use this service.

        No, there aren’t. Again, that’s not personal — that’s just a fact! And for the reasons enumerated by ASDF and William C. above, your insistence on denying this fact leads you to ignore the massive opportunity cost of failing to provide service worth using in the rest of the off-peak hours.

        For every stubborn individual who will wake up at 3am rather than 4am to take transit to their crazy-shift job (rather than admitting to themselves that their situation is an edge case calling for a car), there are thousands of people who drive to their 3pm-11pm service jobs, because the 11pm transit service is shit.

        And again, the June hours don’t fix that. A frequent-service span until 10 is better than one that ends at 7, but it’s still pathetic by the standard of every one of those transit-enabled “world-class cities” you rave about. Heck, even RapidRide will continue to see 30-45 minute headways at night’s end in June.

        We can’t even admit to ourselves that defining “frequent” at 15 minutes puts us in a permanent 3rd-class transit state. But you want to send 200,000 hard-won service hours into perpetually-empty overnights?


        A magical thing happens to a city and transit system when service levels rise above the last-resort minimum: people start to default to transit. All the people, some of the time. Some of the people, all of the time. Demand and usage rise exponentially, reinforcing one another. People wonder how they ever lived without such ubiquitous mobility freedom.

        And then — only then — may there exist an overwhelming demand for real and usable levels of service into the overnight hours.

        STB is all too full of cart-before-horsing. 24/7 service is on the side of cart.

      2. To begin, d.p., I’m sorry for coming off as having made you the villain of the piece. That wasn’t my intention.

        I also think that I’m being unclear: I didn’t propose to send 200,000 service hours into the middle of the night. So far, my math says that I propose sending less than 20,000 service hours into the late night. The point I tried to make is that we can spend 40,000 to do the entire current wish list (that number comes from the City and Metro), spend another 200,000 hours to do anything we want during the other day parts, and then spend a net total of under 20,000 hours to have 24-hour service on key routes.

        This follows my desire to not have the 80-series Owls cut. Why? Because if they were cut, they were never coming back short of a massive windfall. They continue to exist, thankfully, so they form a core to build around and, ultimately, cancel because they’ve been replaced by actual, functioning transit.

        You are dead-on that transit at 11pm is crap. I know this because I ride it, too. I think the 40,000 initial spend in hours will help. Heck, the remainder will almost certainly overflow into additional service between 5am and 1am because there will, almost literally, be nowhere else to put those service hours. But I also don’t want to make late- and overnight service wait for the conclusion of putting all of those hours to work. Add them at the same time. For example, as the 41 gets more service hours and additional trips that people can use, add the overnight runs, too. Don’t wait until the end when either all of the hours are just about gone and people are scrambling to fill last-minute requests or put it “on the table” as “something for later” and later never comes.

      3. I’m sorry for coming off as having made you the villain…

        It’s not that you made me the villain, or were even (nearly as) adversarial (as I can be). It’s that I feel forced to turn into this ogre just by virtue of pointing out on-the-ground facts inconvenient to the perpetual STB fantasy of unlimited transit at all hours to everywhere magically attracting riders.

        we can spend 40,000 to do the entire current wish list (that number comes from the City and Metro), spend another 200,000 hours to do anything we want during the other day parts…

        Now I understand your math even less than before. Are you suggesting that the entire near- and medium-term Prop 1 roll-out, as presently itemized by Metro, accounts for only 1/7 of Prop 1 funds? That seems a little far-fetched to me.

        under 20,000 hours to have 24-hour service on key routes.

        Given the length of your route list, this seems pretty far-fetched as well.
        Even at your hourly-service rate, which of course both reflects the weakness of demand and ensure that only a handful of the desperate will ever be on it.

        But I also don’t want to make late- and overnight service wait for the conclusion of putting all of those hours to work. Add them at the same time.

        Again… WHY!!!!!!?? The demand for 24/7 just doesn’t exist. Not even in our “densest” places. Not remotely anywhere else.

        This is the crux. We’re a spread-out city, and a still-small one too. We’re not a city that functions on a 24/7 basis with remotely the critical mass for mass transit.

        It really is that simple! Even with your attempt to “cover” much of the city, there just aren’t enough people with enough needs in the overnight hours to yield anything but empty buses! There aren’t enough people right on the corridors, nor enough people desperate to lose more sleep, walk far, and wait an hour for the privilege of moving at Metro snail-pace to a place they’d probably rather not be at the ass end of the timepiece.

        Graveyard-shifters are going to continue driving no matter what version of this plan you throw at them. And that’s really okay, because we’re not New York and we don’t need to pretend we are. But you want to throw a whole bunch of service into the void of uselessness, and that’s really shouldn’t be okay with advocates for growing transit in our city!

      4. Here’s where I’m getting the 40,000 hour figure:

        [T]he city would deploy its first 33,000 new service hours to 49 routes on a city priority list, because the buses overflow or are chronically late. That leaves a whopping 233,000 hours or so to fortify the citywide network, touching virtually every neighborhood…

        You have a point about my estimate of what the late night proposal will cost. 20,000 new service hours is coming in low if I’m doing my math right, but even at double that it still comes in at about what the city’s published “wish list” is proposing.

        There obviously is demand for overnight service because there is currently overnight service and there are demonstrably people on that service. Besides, ever since September, my and others here have seen the ridership patterns dramatically change, even on the much-derided 80-series Owls. They’re not just homeless shelters any more. Is there more demand? No one’s ever asked or tried it with sensible routes.

        To your point about grave shifters driving: We don’t have to (I don’t, though it leaves me stranded when I get to work until Metro wakes up again) so why not make the service more useful? Yes, hourly service is a beating. I’d like to have half-hourly but that would be a nightmare to propose so just getting full overnight service is worth the trade-off. Below this, Bjorn posted a transitmix map with timed transfers and sensible routes that would work quite nicely. There’s not as much demand as there could be–and, again, I can visibly see demand both from the riders on the few buses that exist and the massive number of taxis still running around at bar-closing-time–in part because the service sucks. So, let’s improve the service, both in the “regular” day parts and overnight. This isn’t light rail where putting it on the road will cost a few billion dollars. If it’s a bust when Prop 1 runs out, I’ll eat my hat and be completely wrong.

        No, we’re not currently New York City. Heck, we’re not even Chicago or Pennsylvania, or even Dublin (Ireland) with our level of transit and overnight transit investment.

      5. Ugh. You’re still refusing to address realities of scale, and you’re starting to embody my frustration with “STBlinders”.

        There obviously is demand for overnight service because there is currently overnight service and there are demonstrably people on that service.

        No! Describing the trickle of legitimate users of our overnight services as “demand” should be offensive to the hundreds of thousands of riders and the hundreds of thousands of wish-it-made-sense-to-riders who suffer our awful transit situation during the 19 daily hours of normal span.

        We’re not bringing the 61 back. That thing had about the same “demonstrated demand” as the night owls.

        We shouldn’t be prioritizing transit that is basically useless even for the trips it serves, whether that means sparsely-populated far-west Ballard or sparsely-needed 3:30 am.

        I can visibly see demand both from the riders on the few buses that exist and the massive number of taxis still running

        More straining. People grab taxis at the end of the night because they want to get home in 15 minutes rather than two hours, and thanks to our fundamental physical arrangement as a city, there will not be a critical demand for transit to make that possible at 2:30am every night of the week.

        There is enough demand to make transit service work at 11pm, which is why it is so incredibly offensive to offer terrible service that forces all of those people in to cabs at that hour too. And the might be enough demand on Friday and Saturday nights to extend the span of full-frequency operations on those two nights… as our denser peer cities invariably do.

        Again, the value here is in the extension of real, useful service, rather than peanut-buttering extended hours whose inferiority to car/cab will be insurmountable.

        Listen, I know that sucks for late-night or overnight workers the other five nights. But math is math and the chicken-and-egg of low-quality service and low density of demand simply dooms your plan to irrelevance!

        The city would deploy its first 33,000 new service hours to 49 routes on a city priority list,

        Okay, so this is as I expected. The 33,000 hours are just the immediate-relief trips. The actual frequent-network wish/intention list that I’ve seen, with many important routes boosted to full-time frequency (but only until 10, and often not on Sunday), and with the attention to the beginnings of network-structural improvements (C/D split, etc.) is the 200,000 hours.

        That hardly leaves “ample” reserves for squandering on 11s and 16s and 43s all through the night. After the 233,000 hours are spent, we’re still a long way from a transit network in the regular span that most people would feel able to rely on.

        Heck, we’re not even Chicago or Pennsylvania, or even Dublin with our level of transit and overnight transit investment.

        No, we’re not. And it’s because we aren’t those places physically.

        Horse before cart. Horse before cart. Horse before cart.

      6. d.p., thanks for putting the 200,000 hours in context. Do you have a link to the intention list?

        And I agree with you that better-than-hourly transit around midnight is more important than hourly transit on this somewhat-comprehensive map overnight. If the 200,000 hours still leaves obvious gaps around midnight, then I’d want those filled first. But if we can afford it without sacrificing more important needs elsewhere, I do still think some level of basic coverage overnight is a good thing to have. Maybe not this whole network, but at least something to the four quadrants of the city.

      7. On the one hand late night transit isn’t well patronized because those who need/want to be somewhere make other arrangements. On the other hand the demand is pretty low as can be seen simply by looking at the number of cars and pedestrians on the street.

        I agree the first priority should be improving all-day, weekend, and evening frequencies. After that the earlier part of the night (before midnight) should be next.

        That said there is a valid public policy objective in keeping the transit network at least as functional as it is currently at 11 PM until after bar close. As people do go out and drink on weeknights I wouldn’t limit this service simply to Friday and Saturday. For legibility reasons it really shouldn’t be anyway.

        While ridership wouldn’t be spectacular if the 11 PM service levels were extended until 2 AM I suspect you’d find a lot less of a drop off than the current numbers might lead one to believe.

      8. “For every stubborn individual who will wake up at 3am rather than 4am to take transit to their crazy-shift job (rather than admitting to themselves that their situation is an edge case calling for a car),”

        People with night-shift jobs are, with few exceptions, those who can least afford the expense of a car. Because if they could get a day shift they’d take it. If you’re working at some warehouses, all new hires are on night shift for the first year or two until they get an opportunity to switch to a day shift.

      9. Mike,

        Life is full of hard choices. When you endlessly privilege cases so “edge” that they amount to a handful of people per route over the entire proposed overnight service period, you do real harm to the many, many, many more people deprived of usable service at other times.

        Social justice involves transit as good as it can be to as many places as it can make sense and for as broad a span as can be justified. Buses to Licton Springs at 3am don’t meet that definition any more than subways to Federal Way.

        Anyway, we’re talking about the city here, i.e. manageable distances. I’m way too cheap to cab or Lyft, but since I rarely drink to drunkenness I have car2go as an option, which, in the no-traffic hours, costs barely a dollar or two more than bus fare. (In fact, I’m so cheap that I do the strategic-gas-refill thing, and rarely pay for car2go at all.) Those who cannot or will not own cars have access to ever more reasonably-priced options that do not involve fingers-in-ears denial of immutable facts.



        If we were even a 50% busier city, I might agree. But if you’ve ever been out late on a weeknight, you’ll notice that patrons dwindle to a trickle far earlier than on weekends, to the point where the only beneficiaries of your much-extended service would be the employees. But to offer service to every bar employee would demand extending until 3am every night of the week, which is practically all night; and even if every one of those employees took those late buses, you’d still be shy of your critical mass of need.

        And you won’t capture most of those late-nighters anyway. The later it gets, the less patience anyone has with waiting on street corners and delaying their arrival in bed. That applies to bartenders as well as the very last of the patrons. Especially on weeknights, it barely matters what kind of service you offer in the wee hours; most people won’t want to wait 5 seconds for it.

        There’s a reason that mid-sized cities experimenting with late-night transit, and those who have had partial extended-service programs for decades, all focus their energies toward the weekend. It’s simply where the demand is real and justifiable.

        That these same cities offer full-frequency service during the regular weekday span, until midnight or 12:30 or 1, recognizes the multiplicitous quality of weekday evening urban activities: dinners, movies, plays, hangouts, working late. Multiplicity drives useful transit. Late-night Tuesday binging doesn’t.

        Sorry to invoke the veto power of “elsewhere precedent”. But precedent exists for a reason.



        I’m glad we’re in agreement on the point of priority. Peanut-buttering service hours around the clock is really the same as peanut-buttering them around the least useful daytime services, opposing service rationalizations and claiming that the needs of the very few outweigh the very purpose of mass transit, which is to enable mobility for the masses.

        I won’t say that Seattle will never be a city that can support all-night transit. But there will need to be a tectonic shift in how transit-enabled we become as a city before that will happen.

        We aren’t a city where transit-as-primary-conveyance cuts across all cultures and classes, much as advocates would like to believe otherwise. We need to start advocating from a solid footing upon the facts, even when those facts are harsh: we’re not a megalopolis; we’re not European in form or function; our unrestructured network is inefficient for getting around at any time of day; West Seattle is fucking sparse and the “densification” there is negligible; Everett is not about to double in population; 70-mile subways don’t make a lick of sense; tech drones and their livers are not a long-term economic strategy.

        And just because you roll buses around the city all through the night, does not mean people will come.

      10. DP, you’re making an either/or of evening vs night. We need both, and of course evening has priority on limited funds for the reasons you mentioned — there’s a huge latent market for evening ridership that won’t/can’t use it under the current schedule but would use it under a frequent schedule with a longer span. But at the same time we can’t just dish out service hours strictly according to ridership tiers: no night service until evening service is perfect. Because if you did that, peak and daytime service would have the highest claim and you’d have no evening service either. You have to “subsidize” evening and night service somewhat to get them functioning — everywhere in the world does that. Do you think the St Petersburg metro breaks even at 11pm with 5-minute service and 10 people on the train? No, but they recognize that you need a long span to serve the largest cross-section of the population and make transit work the best it can.

        So we can give night service some hours to make it marginally better, but not the full funding above while evening service is incomplete. My outline reflects the need, not the funding capability, because that’s where you have to start from” what do you want? That’s how I assessed Wes’s outline too. I don’t expect Metro to fill this complete outline in June at the expense of evening service. But maybe it can do one or two parts of it, and make a long-term goal of the rest. Obviously, we also need more 2-dimensional density too.

        Another thing is that the evening problem may be solved by June. I have been pushing for “RapidRide frequency”; i.e., 15 minutes until 10pm every day on all frequent routes. I think Metro is about to do that, at least Monday-Saturday. That would get rid of the 30-minute dropoff in the early evening and eliminate this issue. I don’t think 15-minute service has to be extended to 11pm or midnight or 1am before we can add any hours to night owl: 10pm is a reasonably good step for now, and much better than 7pm (or 5:30pm on the southbound 40).

      11. d.p.

        As a bit of a night owl and somone who often works late hours I’ll say some weeknights seem to have nearly as many people out as a Friday or Saturday.

        In any case part of my thinking is based on how many people I see on the last runs of some routes, especially those serving the. U-District and Capitol Hill. Sure it isn’t anything close to weekday peak, but it is much better than the OWL service.

        I’ve had occasion to take a late RR A/124 a couple of times and I’m fairly supprised at how relatively well patronized those routes are during late night hours.

        Clearly the classic OWL routes are broken and make almost every mistake possible to discourage any but the homeless from riding them. Sure ridership is never going to be great at those hours but where the whacky OWL routes have been replaced with extra runs on all-day routes the number of people riding them as transportation has gone up.

        Don’t get me wrong I wouldn’t invest a huge amount of service hours after 11 PM and especially after 1 AM, but I think it is worth devoting some service hours at those times to improve on what we currently have.

      12. I am very much a night owl as well. Have you seen some of the crazy timestamps on my comments?

        There is simply no comparison between 1am on a weekend and 1am on a weeknight, be it the Pike/3rd transfer point, in our minor activity centers, or at the busiest intersections of Capitol Hill. The weekend is an order of magnitude busier.

        There are many weeknight 10pms that can look similar to weekend 1ams. And that’s exactly my point: “transit you can trust” depends on usable service levels throughout the regular span, and not on variable edge surges like at last call.

        I don’t think 15-minute service has to be extended to 11pm or midnight or 1am before we can add any hours to night owl: 10pm is a reasonably good step for now.

        I emphatically disagree. In cities where “frequent” means 7-10 minutes and “ramping down” means reducing to 15-20, then it’s okay for service to ramp down at 10pm.

        But in Seattle, where idiots still think four times an hour on the heaviest lifters is A+ service, and where we let frequency fall off a cliff rather than ramp down, declaring 10pm a “reasonably good step” means continuing to forgo the business of anyone who gives a shit about their time, and accepting that these people will drive by default if there’s even a chance that they will stay out past that point in time (i.e. not especially late by anyone’s standard).

        Again, a push for full-span frequency could be aided by a categorical division between “core” and “not core”. Metro has proven it can’t stagger interlines properly in the evening — just look at the 2/12, or the 28/40, coming through half-hourly but at the exact same time — so the solution is designated routes that send most everyone in the direction of their desire, but come so often that you won’t even have time to complain it isn’t your “ideal” one-seat route.

        But that only works if it works through the entire service span. Falling off a cliff at precisely 10:00 is unacceptable, and doing so in service of low-priority skeletal service at 4am would be the height of backwardness.

      13. *sigh*

        Minneapolis goes to bed even earlier than Seattle does; one of the earliest-sleeping cities in the US.

        Due to public outcry, the light rail lines in Minneapolis run all night. (At extremely reduced frequencies.)

        It is well worth having a skeleton network of core routes running at night. They should be pretty much the same routes as the core routes during the daytime (none of this confusing “night route” stuff) and they should be running less frequently — but at least once an *hour*.

        What’s going on here is that a number of people are worried about getting completely stranded after “the last bus” or “the last train”. If there is always another bus every hour, this worry goes away.

      14. “Again, a push for full-span frequency could be aided by a categorical division between “core” and “not core”.”

        It sure would. They did that pretty clearly in the Twin Cities; it’s obvious what the “core” routes are, and they actually chose them appropriately.

      1. Use HTML tags. Put text to be bolded <b>inside these</b>

        (I hope this works right; if you see ampersands, it didn’t.)

      2. I’m verbose. If I don’t use mark-up for emphasis and to visually break up run-on thoughts, I find I am even more likely to be misinterpreted.

  8. Two possibilities for cross-town adds: The oft-discussed “8N”, from Queen Anne to the CD (4-5 trips for hourly service); also, some kind of service on the 50 (this is more problematic, but is also 4-5 trips).

    1. I would put service on the 8N, but not to make it available around the clock (if that’s what you meant). With the 43 and 11 already mentioned, I think the 8N is mostly covered.

      The time that the 8N does need more frequent service is between 8 and midnight, 7 days a week, and existent service into 2 am on Fridays and Saturdays. On weekends, the number of cabs — visible, traditional, medallion-cabs (retronym!) — going up / down Denny alone tells you that people would take the 8 if only the 8 were still running when they planned to get home from merriment in Belltown / Cap Hill.

      For so many other trips, the 8N is a great enhancer: the bulk of the time for Crown Hill / Phinney / Fremont -> Cap Hill / “the east” trips is getting through Belltown, waiting for the correct uphill bus, waiting for everyone to load on to the correct uphill bus, etc…

      Being able to transfer on Denny lets you cut all that out, since the 8N slices through a lot of the coverage of the Cap Hill buses. If they came ever 15 minutes instead of 30 later in the evening, a lot more of these transfers would be feasible. After rush hour, the 8N actually has a shot at keeping actual 15 minute headways instead of the comedic bunching that it is so known for in the daytime, so the service hour deployment would be efficient.

      [the (aerial) G-Line would also solve this but I suppose we are trying to be realistic][though think of the terrible jokes it would enable, much less the mobility!]

  9. I’ve read comments in the past from night owl bus drivers who say a good portion of their after midnight clientele are homeless people looking for a warm place to nap. If that’s true, that only a fraction of the riders on night owls are workers, and most people are just aimlessly riding around the city and napping, expanding late night service would be a waste of money.

    I also disagree that most people think of Seattle as a world-class city. Yes, most major cities in the world do have 24 hour service. But Seattle isn’t a major city. It’s a sleepy, boring, go-to-bed-early provincial town, which some people mistakenly believe is a world-class city because we have a couple of corporate headquarters located here. Think of the busiest intersection in downtown Seattle at 5 PM. 4th and Pine would be a good spot. Now, what’s going in that area at 3 AM? Are any shops open? Are there many people on the street? No. It’s completely deserted and everything is closed. Decades from now, when Seattle becomes a 24 hour city, then yes, let’s have 24 hour bus service, but until that time, it’s a waste of money.

    1. If you want a world-class city, look to Bellevue. Bellevue has Sam. Need I say more?

      Well, I’ll say more anyway.

      The 80-series routes are useless as transit. Getting rid of them would raise some howls from shelter advocates, but, really a straight E-Line coming every half hour, all night, would be a godsend for people trying to find hotel rooms, or get to their graveyard shifts at the hotels, 7-11s, and cash advance shops. The lack of minimally-frequent all-night service on the E may be what is keeping some 7-11s and cash advance shops from staying open all night, depriving us of additional sales tax revenue.

      Most importantly, for those needing a warm place to slump over, the E is a far more comfortable ride than the 80s. Good riddance to the 80s, when they finally go away.

    2. I can’t speak to the generalization that most world class cities have 24/7 transit, but I think–in relation to ST’s Link ‘maintenance pause”, that it has pretty good company, eg. Paris and London. In both of those metropoles there are reasonably useful Night Bus services.

      The anecdote that the night busses are primarily used by the homeless does not conform to my own anecdotal information, based primarily on occasional early am or late night trips on the 84 and on the old 174 night schedule going out to SeaTac Airport. 84 was not heavily used, but I saw the same ‘non-homeless’ riders boarding each time I was on it. And 174 was clearly full of folks heading home–or heading to jobs.

      1. I realize I did not make it clear that the Underground and the Metro in London and Paris do stop running at night. There is no sadder sight than dashing down the stairs to the platform to see the doors close on the last train.

  10. As a Lake City resident, do you think the 41 covers enough of that neighborhood to be sufficient?

    Any consideration for the 5?

    I think it’s important to not think about numbers but not treat them equally in all cases. An important consideration is the cost of a non-ride.

    If we fail to add a needed bus at the crowded peak, the cost is ~5 minutes and less personal space (not that bad) and possibly pass-ups (worse).

    If we fail to add 15 minute service until 10, the cost is an extra 10-20 minute wait .

    But if we fail to add all night service, the cost is a several hour wait, or an expensive cab ride, or a drunk trip home, or waking up a family member to come get you, or not taking that shift at work. The lack of service comes at a much higher cost-per-rider not served.

    1. The first two cause enough harm to the usability of the transit network as a whole that they significantly depress modeshare and driving hundreds of thousands of daily trips back into cars.

      The latter impacts only the infinitesimal number of regular users who happen to work and live precisely adjacent to an all-night one-seat line, and who are either desperate enough or patient enough (or delusional-transit-bloggy enough) to waste unfathomable hours of their life, rather than get a car like any rational person with understanding of this city’s lack of critical activity density and land-use deficiencies.

      A system that accepts being the “loser cruiser” at noon and the system of last resort at 11pm — and that is still what we have; the Prop 1 hours are but a tenuous first step towards being anything else — will simply run empty in the overnight! Throwing money there is wasteful and stupid.


    2. The 5 is too close to the E to consider. Our peer cities have them spaced a mile apart, so that would argue for the D, E, 16, 72 (or whatever), 65, and 75; as well as the 44, 48, 40, and a 130th route. So the existing night owls and regular routes are in the right places more or less but their span is too weak. A contrary factor though is that our peer cities have 2-dimensional multifamily/row-house density throughout their night-owl areas. So maybe we can’t justify the 65 and 75 and 40 and 130th route, but maybe we can do something in between like the 71/65 route I suggested above.

      But none of these gets you to route 5 unless you want to delete the D night owl. In that case you should extend the 5 to Aurora Village, which would make it different from its day route, and raise the question of why not just use the D in the first place, since the whole point of RapidRide is to market it as the primary route in the area.

  11. On a related note: Philadelphia experimented with 24-hour weekend train service (in addition to its preexisting 24/7 bus service) last summer, combined with an advertising push targeted at bar clientele. There was such great demand from service workers that they’ve made it permanent.

    1. Philadelphia is also the nations fifth largest city with a rail system that seattle could never build. The regional rail system has 153 stations with dozens of heavy rail rapid transit stations. Nearly one million daily riders use the system. Seattle is far from this reality.

  12. In map form:

    I attempted to keep about 95% of your idea the same, but I did make some minor changes to save hours when practicable. Some modifications I made, in roughly geographic order:

    C+120: The C takes 30 minutes, Route 120 takes 20 minutes to reach 26th and Barton. These two routes are interlined on Barton Street, saving one C-Line coach.

    A+124: Together, these routes create a 90 minute cycle, allowing for hourly service on both routes with only three buses. I removed the nighttime deviation into Tukwila Station due to a lack of other connecting routes at this hour.

    7, 36, Link: Excess time exists on both Routes 7 and 36 for further southern extensions. I’m not sure how to optimize this area while remaining faithful to your plan.

    41+67+70: These three routes are combined into one route on a 120 minute cycle. I detoured this route over to 15th Ave in order to create a timed connection point with other routes.

    43+44: True to your idea I kept the 43, but this route doesn’t have much of a purpose. UW-Downtown travel is already covered by two more direct routes, and coverage to Capital Hill isn’t expanded substantively.

    D: This route could tightly fit into a 60 minute cycle if the Queen Anne deviation were eliminated (at least at night). Care would be needed to ensure Route D is interlined Downtown with a route with ample recovery time on the southern end.

    Timed connection points exist in a few locations:
    Downtown (3rd and Pike/Pine): All routes arrive at :00, and depart at :10. Some interlinings may change from daytime counterparts, but I haven’t looked into interlinings at this level of detail yet.
    UW: All routes in all directions arrive at about :32 and depart at :38. The location of this connection point needs more detailed examination.
    Northgate: It should be feasible to for Route 16 to arrive a few minutes before Route 70 (daytime 41) departs further north, and vise-versa from 70 to 16. Unfortunately, Route 40 is too long to be timed with the tail of 70.

    1. Are you rebranding the 120 into a RR route or using a non-RR coach for the night owl service as all door boarding isn’t available overnight anyway? Otherwise, you are muddling the brand, as weak as it may be.

      Also, not specific to your comment, I saw someone upthread mention that the 40 should get overnight service because, among other things, it serves NSCC. No education center that has no on campus housing needs to be served any later than normal service hours. In fact, skipping NSCC and going directly along 105th/Northgate Way on the north and along 92nd on the south would save time and make more sense.

      1. For the handful of late night trips that would have a C+120 interline, I’m not concerned about running the “wrong” coach on the wrong route. What few regular customers are riding probably won’t notice or care the bus is red instead of blue/green or vise versa. The first-time customer segment who benefits the most from route-specific liveries are either riding during the day or are too drunk to notice or care about the bus livery. (This is why I’m not a fan of route-specific liveries; operations suffer for minimal marketing gain).

        Does NSCC hold any late night classes?

    2. Last class (at least Winter ’15) gets out at 950pm. Security stays until 1130pm so, theoretically, that’s when the campus closes. Normal span of service would be enough for even the professors who are staying late after class.

      The question of the coach on the C/120 was more a matter of the feds having a problem with Metro running a non-RR coach on a Rapid Ride route or vice versa. It may be necessary to not call it the C to avoid any issues.

  13. For a city that likes to pretend we will grow around transit (if everyone wants to go downtown or u district….), service after 7 in this city is a complete joke.

  14. I’d suggest a “Lake Night Link” feature route that follows most of the current and planned Link stations in Seattle. It could also include a stop at the Airport and maybe TIBS.

    I’d suggest this for a number of reasons:
    1. The public will already know the Link alignment so if they need a late night bus, they will know where the transit stops are and where they can go on transit.
    2. The airport has late arriving planes (including delayed planes) as well as early morning departures. Many workers need to report so early that they can’t wait for Link light rail to open.
    3. The stations Downtown, on Cap Hill and around UW all have some level of late night activity, including some 24 hour stores within walking distance.
    4. Perhaps Seattle, ST and Metro could come up with a joint subsidy arrangement.
    5. As Link opens, there won’t be a need to change the late night service structure.

    1. I guess I should add that I expect that this will be run with buses. While train service would be nice, I certainly understand the need to have the tracks inoperable for maintenance at night.

    2. I agree, and have you covered in the plan. :) I would like to see the route 97 “Link Shuttle” run for overnight Link trips.

      1. I somehow missed that, lakecityrider! Where would Route 97 have its northern terminus? I’m suggesting that it goes to UW if not further – like Link itself will in a few years.

    3. Night owl on the 97 is one of those things that may look good on paper but would probably not be as useful to people as the 7, 36, 49, 124, etc.

      1. None of those are a single-seat ride to the airport, though. 124 is close but you’d have to transfer to the A at Tukwila Station. I seem to recall that 124 used to be single-seat to the airport, but maybe that changed when the A started running.

      2. At least for Prop 1 money, we can’t. The 124 has too many stops outside the city.

  15. I don’t mean this to be advocacy for late night service, but there does seem to be increasing pressure for it around the country. In the Bay Area, BART cointracted with AC Transit to run additional San Francisco-Oakland-East Bay service after midnight. Many people would like BART to run after midnight, but they maintain adamantly that they can’t do it. so they’re paying for more AC buses.

    An interesting thing happened in Philadelphia. Responding to pressure, SEPTA started running the subway lines all night on weekends. This was thought of as service for the millenials, party goers etc. But it turned out that the largest user group were service workers with late night shifts.

    It’s not the only question, but I would definitely ask the question “What set of transit changes best facilitate car-free living?”

    1. I do not disagree with the need to ask that question.

      Which is precisely why I argue that “perpetually mediocre service in the regular span, plus lip service to as-yet-unproven demand in the midweek overnights”, will not achieve that.

  16. I wish Route 60 had the same hours on Saturday that it has on Friday (i.e., until around midnight). Like the 48, it also touches a lot of neighborhoods (Capitol Hill, First Hill, Beacon Hill, Georgetown, West Seattle/White Center. Many of these places also have nightlife, etc. that people want to get to (and get home from) on a Saturday night),

  17. I just want to say how much I agree with this whole post. I’ve recently been thrown onto a 5am shift in downtown and realized I literally didn’t have a single bus route available to me that would get me to work at 5am. My bosses aren’t giving me any wiggle room to catch the 41 that gets to downtown at 5 20 either.

Comments are closed.