One of the more painful parts of being a transit advocate in Seattle is watching King County Metro — carefully, deliberately — load, aim, and shoot itself in the foot. Over and over again. This time, Metro is proposing to add two more stops to the E Line, as part of its final legislative package to the King County Council; a package which, once enacted, will enshrine the alignment and stops of the E Line in law. The two stops are among those currently served by the 358, on Aurora Ave at 80th St and Lynn St. If adopted, this change will further reduce the speed and reliability of a route which already has entirely too many stops and not enough bus lanes to be anything remotely resembling “rail on wheels“, as FTA Administrator Rogoff recently described RapidRide (presumably to a chorus of tittering laughter).

No, the smart thing — for both riders and taxpayers — would have been to go to the council with a package that removed the two superflous stops at 90th and 100th St, which would have given RapidRide E a fairly consistent stop spacing of roughly half a mile in the populated areas between Belltown and Shoreline. (The stop situation in Shoreline is even worse, but the City of Shoreline is the main culprit there — those stops were retained at the city’s request). That half-mile stop spacing is exactly what’s called for Metro’s Service Guidelines (page SG-13, #6), and that standard exists for a reason: transit that stops less is a faster and more reliable way to get around the city; and if the overall quality of the service (safety, speed, reliability, frequency) is high enough, riders are willing to walk further to the stops. Adding the 80th St stop will just compound the error of maintaining the stops at 90th and 100th.

The Lynn St stop, while not violating the half-mile standard, is just as bad as 80th, but in a different way. Looking at the ridership chart I compiled for my post on the origins of our current Aurora service, the current 358 stops on Lynn and Crockett stand out as having paltry ridership. This isn’t surprising when you consider that this stretch of Aurora is basically an unpopulated wasteland dug into the side of a hill. (There are lots of apartments on the east side of Aurora, but those buildings almost all face Dexter, which is an altogether nicer place to catch a bus); moreover, those stops are already very well served by Routes 5 and 16. Far more riders will lose out, and far more transit riders will be lost, due to the delay of making this stop, than will gain or be gained due to its presence.

It’s particularly galling to read about changes like this when other cities are using the idea of Bus Rapid Transit as more than just an way to get the FTA to pay for new buses. For example, San Francisco is moving ahead on the Geary and Van Ness BRT projects, projects which represent a plausible attempt to raise bus service on those streets to the level of SF Muni’s rail lines* — at a fraction of the cost. I realize those projects have budgets that are an order of magnitude larger than RapidRide, and I don’t fault Metro for not having the money to execute such ambitious plans, nor for not having all the money expected from the passage of Transit Now. I do fault Metro (and the King County Council) for lacking the gumption to make zero-capital-cost improvements to the design of their bus network that would drive ridership and reduce operating costs.

* I’m aware, by the way, that Muni has plenty of dysfunction of its own, and its rail lines aren’t exactly paragons of transit excellence.

107 Replies to “Metro Proposes Making the E Line Slower”

  1. The proper course isn’t for the council to legislatively remove bus stops. The proper course is for the county council to decline to legislatively determine bus stops at all, and pass that administrative power back to Metro.

    1. I suppose I should ask: Has the county council passed ordinances previously that specify bus stops, or is this the first time the county council will be exercising such draconian micromanagement?

  2. So we’re getting a bus with occasional 1/4 mile stop spacing (6 stops between 75th and 105th), at frequencies that are no better than our best local routes (7, 43, 49 etc), and we’re calling it BRT? So depressing. FTA should demand their money back.

  3. I like the Lynn stop because I don’t think we should have bus service on Dexter at all. There are literally 4 parallel corridors of north-south service here, on 5th, Aurora, Dexter and Westlake, all within half a mile of eachother. True, there’s a hill there, but people can walk up and down hills. You think that stop spacing < 1/2 is bad? How about route spacing < 1/2 mile?

    1. As has been pointed out to you before on other threads, there are good geographic reasons why there have to be so many closely spaced routes in this area. Between 5th/Taylor and Aurora is a cliff face. Getting between Westlake, Dexter and Aurora is a pretty steep climb for able-bodied people. There is no good access to Fremont from Aurora. These are legitimate reasons to run closely spaced routes.

      1. From Dexter to Aurora is a ~90ft climb. From Stewart and Denny to Stewart and Olive is a ~120ft climb, and riders of the 545 are expected to do that every day because there is no southbound 545 stop in cap hill like there is northbound in the morning. It’s a double standard to say that people on Dexter shouldn’t have to walk up a hill but people have to walk up a hill to get to Cap hill (and the 8 doesn’t count, it’s far too infrequent and unreliable to be used as a way to get up/down the hill).

        The lack of pedestrian access across Aurora is a good reason to make capitol improvements (in coordination with the E line), to improve access to and from Aurora to southbound routes and to Fremont.

      2. I’m sorry, but you clearly understand nothing about the built environment or topography in this area. It is not meaningfully comparable to the hike up Pine St, which is a steady, moderate grade, with good, wide sidewalks.

      3. He’s not referring to climbing up Pike or Pine which is a notable detour, but rather up Denny, which is quite a bit steeper. Depending on what part of the hill you live on you might choose something different. It’d also be better if it was easier to get onto southbound busses at Montlake from the freeway stop, but that stop is being removed in the near future. In all cases the amount of distance you’re expected to do that climb is longer, but you’re still expected to climb up a fair amount of the hill none-the-less.

      4. You run buses on Dexter for many reasons. Here are a couple:

        1. Dexter’s streetscape and overall built environment is more favorable to transit than either Westlake or Aurora. There’s a reason that Metro considered moving the 5 from Aurora to Dexter, despite the reliability nightmare of the Fremont bridge.

        2. Any bus running on Aurora basically has to either skip lower Fremont (a significant center of retail, jobs, and housing with excellent transit ridership) or make a ridiculous detour to serve it. Even a stop located right by the Fremont Troll would miss the mark by a bit. Dexter buses don’t skip much that matters. When you skip destinations you end up with inefficient, confusing service patterns.

        If Metro had to consolidate Westlake, Dexter, and Aurora routes onto one street we can only hope they’d choose Dexter.

      5. A stop by the Fremont Troll is exactly what is needed by a BRT serving Aurora. There are a lot of people who work or live in Fremont, and their bus options are really slow. A stop there would have to be carved out of Aurora, but if the buses are fast and frequent (which is the point of BRT) then they would be very popular. 36th and Troll Avenue (AKA Aurora) is within a quarter mile of most of Fremont. Yes, it is hilly, but people would walk that distance if it saved them time (especially in the morning, when folks working in Fremont would simply walk down the hill).

        I wouldn’t remove the other buses, mind you. I think this is the way a BRT (or rail) system is supposed to work. The BRT gets you very quickly and frequently to an area, but you might have to walk a few blocks. Meanwhile, the other bus will get you right next door to your destination, it just won’t get you there very often or very quickly.

    2. Bus service is not just for the athletic. 5th Ave N, Aurora, Dexter, and Westlake are literally different markets, like California vs Delridge vs Beacon vs MLK/Rainier. If the goal is to make more people willing to use transit for their everyday trips, you can’t expect people to walk up a steep hill every day or multiple times a day. Even if they can, they won’t. It’s an unreasonable expectation.

      The issue is how many destinations Aurora has. The number of apartments and businesses on Aurora is much less than the parallel streets, so the number of stops should be less. On the other hand, you can’t just think about the people going to downtown or 45th. What about people living on Dexter or 5th who are going to Oak Tree, Aurora Village, or Snohomish County? The 5 and 16 don’t do nothing for them. A Lynn Street stop may be more justifiable than the 90th Street stop, because it’s a larger gap to neighboring stops. Even if it may be the same distance-wise, it’s not the same in terms of how difficult it is to get to surrounding stops or how many destinations are in between.

    3. Beyond the hills, you realize there is literally no way to cross Aurora on foot other than at Denny, Galer, and Raye, right? There is no way for residents of Dexter (or of the east side of Aurora) to catch a southbound Aurora bus.

      The Aurora buses serve very few people. Ridership on the Dexter buses is extremely high. The Dexter corridor is the one that needs local service with reasonably frequent stops. Aurora should be for limited-stop buses that need to get to the north end quickly. If I were in charge, I’d remove most of the Aurora stops for all three lines, not just the E.

      1. Yes, this is why we need to spend money to improve the access to Aurora. There are certain immutable properties of the land that we can’t do anything about, but poor pedestrian access is NOT one of those. An investment of capital dollars today, would save money in the long run by eliminating the need for service on all 4 corridors simultaneously.

        Just because ridership on a bus is high doesn’t mean it makes sense in the context of the bigger system. I could have 2 routes 1 block apart that both have high ridership, but that doesn’t mean that’s the right thing when you could save even more money by consolidation, and then invest those services hours elsewhere where the system is underserved.

      2. I don’t think you could consolidate the service on Aurora and the service on Dexter, because they are serving entirely different purposes. At best, you would end up with separate local and express services on the same street. Yet Dexter (being more centrally located) works better for the local service serving that little wedge of land between the cliff and Lake Union, while Aurora (being faster) works better for the limited-stop service.

        If you tried to make all the service local on Dexter, you’d make the E and the 5 slow enough to be unusable for north-end riders (one reason I strongly oppose the idea of moving the already-slow 5 to Dexter). If you tried to put it all on Aurora, you’d lose most of the Dexter ridership, even if you put in a few more pedestrian bridges. Aurora is just not a fun place to be a pedestrian, and wouldn’t be without a billion-dollar rebuild.

        Consolidation only makes sense where… consolidation makes sense.

      3. You want to end service on an entire corridor, when Metro doesn’t have the stones to skip 2 stops?

  4. The particularly aggravating thing about RapidRide is the insistence of maintaining an on-board payment option, and the failure to provide stations at every stop.

    A problem with other lines (B, C, and F) is circuitous routing, but that’s not an issue here.

    1. I agree, Cheesewheels. I think that CT’s Swift did a great job with off-board payment, and I further think that KCM’s RapidRide needs to do the same.

      1. I’ve only used Swift once, but based on that experience I’ll concur: the dwell time is impressive. That should be one of KCM’s goals.

  5. I’m curious, why do the “add stops” people seem to win the arguments in front of the jury? I assume that the Council and the decision makers at Metro are (mostly) rational people who can understand efficiency and what puts the “R” in BRT. It seems a good opportunity to make transit more appealing to the wide middle swath is slowing being flushed.

    The FTA should absolutely demand their money back.

    1. Because a small, loud, extremely focused group (such as the few riders seriously inconvenienced by removing a stop) always has an easier time being heard than a vague general interest. It’s just a principle of politics.

    2. The political problem seems to be that, above a very low threshhold, Metro & the King County Council have little incentive to say “No” to a small group of people, even if it degrades the service for a much larger number of riders (or potential riders). People who want an added stop will lobby for one, but because the effect of adding each additonal stop is diffuse, there isn’t a constituency for limiting the number of stops. This is especially true with routes that haven’t opened yet, where people don’t know what they’re losing. So, over and over again, we end up with watered-down service that does little to attract new riders.

      The technocratic part of me thinks this could be solved by giving each route two explicit, published time “budgets”: one for average trip time and one for trip time variation. If adding a stop exceeds either of the route’s budgets, either by taking too long or making the route too unreliable, Metro would have to bring it back under the time budget by eliminating other stops (or potentially adding additional BAT lanes or signal priority, or eliminating on-board payment).

      This could be done both when modeling the route before it opens, and to reduce gaming the models, reevaluating the routes annually with actual data after the route opens.

      By making the tradeoffs explicit, this forces the county to design service that has the greatest benefits for the greatest number of people. If they have to say “No” to someone, they’ll (ideally) choose the change that inconveniences the least number of people. It also gives politicians the “out” of pointing to the time budgets and saying that while they wish they could add more stops, their hands are tied.

      The cynical/realist side of me is skeptical that politicians will want to limit their discretionary power, or give someone, somewhere a reason to possibly be angry at them. If they did implement time budgets, I suspect they’d allow for a myriad of exemptions or overrides, or make the time budgets so loose that it’s difficult not to meet them.

      1. The time budget idea sounds really great…it satisfies my need for objective criteria in these sorts of debates. I guess the tough part would be setting the allowable budget. Seriously cool idea, though.

      2. If people like the time budget idea, maybe someone who knows how to do political things could get the King County Council to legislate it. If we had time budgets, it hopefully create constituencies for faster, more frequent bus system. Right now, when we have debates about routing and restructures, there isn’t a much of a counterbalance against people afraid of loosing the long, slow, twisty routes that are the status quo.

        Right now, STB can ask people to email the KC Council or show up at meetings and advocate for shifting bus hours to faster, more frequent, more grid-like routes, but the reality of human nature is that they’ll be outnumbered at meetings by people with less abstract concerns. By setting limits, we make it obvious what the tradeoffs are, encouraging more people to speak up for better service, and giving politicians an incentive to approve service that works best for the greatest number of people.

        The most important part of implementing time budgets is that they be transparent and simple. You should be able to go the the Metro web site and easily find a table of how much time and unreliability each stop or segment adds to a route. Deciding whether a stop exceeds the budget should just be a matter of basic arithmetic. It might be possible to come up with a better measure of optimality, but the tradeoffs in choosing a route should be comprehensible to people who don’t have a math or CS background.

        I don’t know that the problem of setting the budget is insurmountable. Most people have some idea of how long a route should take, and how long they’ll accept. As a first go at setting a time budget, poll a bunch of riders and potential riders, throw out the outliers, and take the median or average. If a reality check is needed against overly optimistic estimates, you can maybe mix in travel time data from existing routes.

      3. The time budget idea sounds interesting — it might actually work as a tool for demonstrating to people why adding one more stop or one more diversion adds up to dysfunctional bus service eventually.

      4. Metro’s new performance metrics are a step in the right direction, and it is gradually moving service to stronger corridors. It may not be perfect, but that’s the reality of this county and the effect of democracy on it. I would rather see Metro going slowly in the right direction than stagnating, and it is doing so. The E is better than the 358, even if it’s not much better.

    3. Because the “add stops” people are advocating in favor of increased SERVICE. The entire point of service changes is to SERVE MORE PEOPLLE. This post will probably be delted as “off topic”.

      1. It won’t “serve more people.” It will serve the same people, but make the service slower for the vast majority of them in the interest of saving a small minority a walk of a couple blocks.

      2. Save your breath, David.

        Beavis truly cannot understand why buses five times slower than driving and 30 minutes apart are anything less than optimal transit service.

        He’s what we’re up against.

  6. Oh well, at least Metro can save some money on fewer buses serving the corridor.
    AM SB departures from Aurora TC has 22 trips or 8.1 minutes apart avg from 6a-9a.
    PM NB departures from Seattle to Aurora has 25 trips, or 7.2 min. from 3p-6p.
    Rapidride says 10 min in the peak, so Metro saves 11 daily trips just in the peak.
    This should really boost the riders/platform hour metric.

    1. I doubt Metro will be reducing the number of peak trips on the 358. Those peak trips are already routinely at crush load at the current headway. I believe the “10 minute” standard is a minimum — at least, it had better be.

      1. I was trying to be caddy. Promoting Rapidride sort of sounds like it’s a lot faster than current conditions and that the buses come more often.
        NOT.
        I’m so skeptical of announcements these days, it just slips out.

    2. Some of the trips peak trips are temporary WSDOT-funded trips due to the DBT construction, which will directly affect Aurora between Denny and Mercer.

  7. Looking at the Rapid Ride line from Ballard, which basically follows the 15 local – not the 15 express has made me refere to Rapid Ride as simply “The Ride”. This latest development simply confirms the name I have been using.

    Feel free to use it.

    1. Part of my morning commute involves taking either the 245 or RapidRide B south from Overlake TC for a mile or so along 156th. In schedule and in practice the 245 is consistently the faster choice for this trip. Now that the 245 is showing up on the real-time signs (or the signs and off-board readers are dead) the only advantage the B Line has is the higher frequency. I may borrow your idea and start calling it “Ride B” instead.

    2. The point of Rapid Transit isn’t to cover long distances quickly by skipping over destinations in a corridor on a freeway or mega-arterial, it’s to consolidate ridership along a corridor by serving its important destinations with a simple service pattern. If you skip LQA you need more complex service patterns to connect it to downtown and will probably neglect to connect it to Ballard at all. Skipping stuff is what gives us our current motley group of routes on I-5. Serving stuff will make our transit system simpler, more flexible, and more powerful.

      1. That is patently false…it is rapid ride not frequent arrival ride. The point is to lure people to it…it is supposed to be as fast or faster than a SOV…

      2. @Doug: The terminology isn’t straightforward. Deal. What you’re talking about is an express bus. Express buses have limited utility and efficiency.

      3. I believe that most people expect rapid and express to mean essentially the same thing – fast. That isn’t what you are proposing. Local versus express is what you are suggesting…the best way to make the bus line rapid is to cut the number of stops… That, I believe is Bruce’s point (and a good one I might add).

      4. I actually have no problem with making the Queen Anne deviation — many riders use those stops. If tons of riders used the stops at Lynn St, I wouldn’t have a problem with Metro stopping the bus there either. Metro’s failure on the D Line is in failing to go in to bat for bus lanes and queue jobs on Queen Anne Ave and 1st Ave N.

        The point of BRT is to emulate those attributes of rail systems that make them successful: make buses faster by spacing stops out more widely, especially in areas with very weak ridership, but serve the heart (rather than the periphery) of densely-urbanized areas that generate lots of ridership. Missing 1st & Mercer would, on net, reduce the utility of the D Line, because it would miss the dense heart of Uptown.

      5. Rapid and express are fast in different ways. Speed, by itself, doesn’t drive ridership. This is pretty uncontroversial.

        I support ~1/2-mile stop spacing for RR routes in north Seattle along with improvements to the pedestrian environment. I don’t support skipping LQA entirely on Western. And I think we should have a plan to add a stop to the Aurora route above the Fremont Troll.

      6. @Al – We don’t need the D-line making the Queen Anne deviation to serve lower Queen Anne. We already have the 1, 2, 3, 4, 13, 18, and 32 for that.

        What is the point of delaying everybody going from downtown to Ballard just to have one more bus serving that same corridor?

      7. We absolutely should be connecting all of our urban centers with fast, frequent, reliable transit.

        But instead we have RapidRide.

        Since RapidRide has to work within the existing road network, we must consider the goal of connecting all of our urban centers against the impact on travel time and reliability for the majority of origin-destination pairs.

        In Metro’s new service guidelines is a simple formula for making these decisions. It’s (through-riding passengers * time differential) / (boardings and alightings along deviation) <= 10.

        Using interpolated times from OneBusAway for the 15 and the 24 travelling between Elliott & Mercer and Denny & Warren, the time differential is 3 minutes. I believe this to be a conservative estimate – as Bruce pointed out in an earlier post, the new signal timing at Elliott & Mercer can result in a long delay to turning traffic.

        This would mean that 3.33 people would need to ride through Uptown for everyone who gets on or off there for the deviation to not meet Metro's criteria. I don't have data, but based on my experiences I think this threshold is met.

        Furthermore, if travelling between Uptown and downtown, you would be better off just catching the next trolleybus that comes along. The schedules aren't particularly coordinated, but they still add up to provide less time between trips than RapidRide's pathetic 15-minute off-peak frequency. Thus I feel the real question isn't how many people board/alight in Uptown, but how many travel between Uptown and Ballard/Interbay.

        Pioneer Square is also an urban center and significant destination, but RapidRide C will not serve it until 2016 (if at all) because doing so would make it impossible to use SR-99 to reach West Seattle, and any other path through SODO would be slower and subject to delays either from game-day traffic or crossing the BNSF mainline at grade (to reach the West Seattle Bridge from the E-3 busway). In that case travel time and reliability trumped connectivity – why not here?

      8. The 18 is already connecting lower Queen Anne to Ballard and is planned to operate almost as frequently as the RapidRide D-line. There is very little marginal value to people making such trips to justify the service delays it entails.

      9. @asdf: The 18 is moving to Westlake in September and will no longer serve LQA.

      1. No, but in the world we’re used to where almost nothing runs more frequently than every 30 minutes (except possibly during rush hour), 15 minutes all day 7 days a week, suddenly seems very frequent by comparison.

      2. You have to start somewhere. Jarret Walker has said that transit agencies need to define “frequent” somehow and make frequent-transit maps, even if the definition is less than optimal. It gets people to realize that frequency is something they could have and is a significant benefit to them. Then they will demand it more, and be more willing to reduce peripheral routes to strengthen the frequent corridors.

      3. He also calls 15 minutes a sub-basement minimum for that. Even with the “Ride” lines in place, that our “frequency map” is sparse to the point of useless.

  8. My head hit the desk when I read they were adding these two stops. I don’t know much about the environment around the 80th St stop, but there really is no defense to the Lynn St stop. According to Bruce’s chart, hardly anyone used it.

    1. The stop at 80th doesn’t have anything super important to pedestrians at it. Here is a google street view looking north at the intersection from a little to the south of it.

      https://maps.google.com/maps?q=80th+%26+Aurora,+Seattle&hl=en&ll=47.686077,-122.343128&spn=0.009288,0.015085&sll=37.0625,-95.677068&sspn=44.52365,61.787109&t=h&hnear=Aurora+Ave+N+%26+N+80th+St,+Seattle,+King,+Washington+98103&z=16&layer=c&cbll=47.686536,-122.344422&panoid=FZYq4JT_GUufkqM0j1NcjQ&cbp=12,0,,0,0

      I kind of doubt that there’s a huge need for transit access to the store selling car racks, or the paint store. I’m just saying, I think most folks currently using the stop at 80th would probably adjust to the stops at 75th or 85th.

      Also, I actually asked Metro to eliminate my stop, at 125th. The stop at 130th and at 115th would be fine. They do get feedback asking for stops to be eliminated to improve speed, they just don’t listen to it.

  9. AFAIK, BRT was indeed defined upwards in the final transportation bill that just passed (HR 4348, §5302, (2)), stipulating that BRT must have >50% reserved ROW. Because of this new definition, Rapid Ride-style ‘BRT’ is effectively precluded from future federal funding, right?

    1. There are two levels of federally-recongized BRT in the bill. RapidRide fits the lower level but not the higher level.

      1. Yeah that is my understanding too. If I remember correctly the Fed will fund 90% of up to three “full” BRT projects a year and the other funding stream will fund “corridor” improvements that would be the RapidRide level. The relative size of these two funding streams I’m not sure, but either way I think it means less funding for RapidRide type projects.

  10. I live by the affected 80th stop. I always get off atte one before because the bus gets stuck at the light. It doesn’t make sense to retain it. It’s not even well used now.

  11. Doesn’t adding more stops add to the cost of creating BRT? I believe there shouldn’t be any stops on Aurora between 46th and Mercer. This route should be used to carry people from North Seattle/Shoreline (relatively) quickly. If the two or three people on each bus want to get to South Lake Union from Aurora Village, then those few people can transfer to the SLUT when they reach downtown.

    Also, I don’t agree with it, but I can understand why Shoreline would want such close stop spacing. Since they’re not going to get much with LINK, at least this gives them something with public transit.

    1. Actually, Link is going up I-5 in large part because the City of Shoreline opposed Link up their part of 99. They didn’t want an elevated train tearing up their expensive landscape-urbanist facelift of Aurora (which is, admittedly, quite nice as suburban six-lane arterials go). As I noted in the linked articled above, Shoreline is holding these stops as a bargaining chip because Seattle has failed to provide full-time BAT lines on its section of Aurora, as Shoreline has done on its section (at not-inconsiderable expense and effort). Seattle’s BAT lanes will be peak-period peak-direction only, benefiting only commuters into Seattle.

      1. Oh come on, guys. Disagreeing with Bruce isn’t “off-topic”. I state AGAIN that this article is mis-titled, and should be titled “Metro Adds Service”. There is no reason that E or F-line service should necessarily be spaced “evenly” if use isn’t. For crying out lout. You guys are better than simpply deleting posts you disagree with.

  12. I think this line could benefit a lot from a local underlay, at least from 85th to Shoreline. Expecting BRT to live up to expectations and meet coverage needs for a 12 mile corridor with +/-10k daily rides or more is not realistic. A 30-minute all-stop local from Aurora Village to Northgate, UW, or Ballard might give you the flexibility to stop at 75, 85, 95, 105, 130, 145, 155, 165, 175, 185, and 192 and skip the rest in that segment.

    The Lynn stop doesn’t actually bother me that much, that still brings the number of stops from 3.5 to 2 per direction between Denny and the ship canal, not bad.

    1. You don’t need to piss away millions of dollars a year on underlay for half-mile stop spacing. You need the balls/ovaries to stare down the tiny number of people who don’t want to walk five blocks to the next nearest stop.

      Local underlays are almost always a terrible idea. What you are describing is similar to what existed on aurora before the 358 restructure. Abandoning that setup in favor of the 358 significantly improved ridership and rides/plat hr.

      1. This is particularly true because Aurora is mostly pretty flat. Major portions in north Seattle lack sidewalks but all it takes to fix that is a suitcase full of cash and another delicious jar full of balls/ovaries to stare down recalcitrant land owners. Unlike the mile-plus stop spacing of Swift, half-mile stop spacing is probably fine.

      2. The city has funding for sidewalks on the remaining sidewalk-less parts Aurora, I’m told, but they won’t be done in time for RapidRide’s opening. I’m pretty certain there are sidewalks at least as far as 105th, though, so I can’t imagine lack of sidewalks being a good reason to keep 1/4 mile stops between 46th St and 105th.

      3. I agree that an underlay would be a poor performing coverage route but maybe that’s the price for the stop spacing you want. I’m not sure there’s a lot of precedent out there for half-mile spacing without local service. Interstate Max maybe, but it has east-west service that provides local coverage.

      4. I am all in favor of decreased stops, but I just want to respond to Bruce’s comment. Sure it is just 5 blocks between stops, but remember, most people will already be walking a few blocks just to get to the stop. Most people don’t live on Aurora, but probably a few blocks East or West of it. If I live in an apartment on 92nd and Fremont, then it is already about 5 blocks just to get to 90th and Aurora. I certainly wouldn’t want to walk to 85th or 100th to get to a stop that is going to get me downtown just two or three minutes earlier.

      5. Remember, if you live a few blocks away from Aurora, you don’t have to walk to Aurora, then walk along Aurora to reach the nearest stop – you can instead walk along your neighborhood street until you’re North/South aligned with the bus stop and not have to touch Aurora until you’re practically at the stop. While this way may not decrease the total distance, it does make the walk far more pleasant and probably a bit safer too.

  13. I think those stops at Lynn/Crockett serve a useful purpose. For someone traveling from the Dexter corridor or on top of Queen Anne to somewhere along Aurora Ave N, walking up/down the hill to these stops, while difficult, is often a heck of a lot quicker than taking the 3/4/26/28 downtown and then transferring to the 358 there.

    I have also used these stops when traveling from Fremont north to a destination along Aurora. The nearest 358 stop to lower Fremont is at 45th Street, a mile walk uphill. It’s not exactly a pleasant walk either if you walk along Aurora, and walking a block over from there just makes the trip that much longer. Instead of doing that, I’ll take the 26/28 south from Fremont and walk up the huge staircase to transfer to the 358.

    These stops are ones that I don’t use often, but the trips where I use them are ones that I would be likely to drive instead if those stops did not exist.

    1. “These stops are ones that I don’t use often, but the trips where I use them are ones that I would be likely to drive instead if those stops did not exist”

      And that is the problem. Metro appears to be serving your limited needs and ignoring the needs of the vast majority of people further up the line for efficiency. Every stop ads deceleration, cash fumbling, and acceleration time. In the case of RapidRide, we really should be picking up a lot of people to make it worth the stop.

      1. Well… each additional stop adds deceleration and acceleration time, but shouldn’t cash fumbling time scale roughly linearly with ridership?

        (Not that the poorly-utilized Lynn stop is the solution to Eric’s dilemma, which is not only his dilemma, but that of many people — transit connections between lower Fremont and the neighborhoods due north of it are abysmal. The solution to that is to send the 5 down Dexter so there’s actually bus service between upper and lower Fremont on its major arterial, and to build a station along Aurora near the Fremont Troll).

      2. A stop right above the troll would help the Fremont situation, but not the Queen Anne situation, where transit connections to the north are just as bad. Metro’s original proposal to run the 3 down to SPU instead of its current spiraling dead-end on top of the hill would have been a slight move in the right direction there, but going anywhere north of the ship canal would still likely require going south to downtown to catch the route to the area you actually want to visit.

      3. “but shouldn’t cash fumbling time scale roughly linearly with ridership”

        Good point. Although, RapidRide tends to not have as much as an issue with cash payment slowing down service since folks can generally board all three doors. But at less used stops, there is not as much as a benefit to 3 doors or off bus payment.

        Go watch a RapidRide coach load at Crossroads or 140th during the rush hours and you’ll see what I mean. In the time it takes 2 or 3 cash payers to fumble for their fares, 20-30 folks with ORCA cards or transfers will board and get their seats in back. However, when I’m at a different stop with no ORCA reader at the curb, everybody stands in line and waits for the cash fumblers which slows down service far more than the cash payment at the more crowded stops.

        You may think the solution is to install more ORCA readers but they cost money and even as crazy as I am about off-bus payment, there is a balance to be struck.

  14. Has anyone considered that the fact that Link will operate parallel to this section with almost no stops means that the E is / should be designed less as a competitor to the light rail and more like a limited stop route? Why wouldn’t you take your crosstown bus to the rail line and ride it if you wanted to go south in a hurry? As young and healthy people we always forget that a lot of public transit riders cannot walk 5 blocks to a stop, and as it is extremely costly to put them on paratransit we would like them to use the regular bus stops. I’m pretty young and healthy and yet if you make me walk up the hill on Denny Way every day to catch a bus I’m going to drive, and I am far from the only person. Likewise there are many riders – elderly and disabled – who will be shut off from bus service if we do not provide enough stops.

    1. “Why wouldn’t you take your crosstown bus to the rail line and ride it if you wanted to go south in a hurry?”

      In order for that to work, we would need much better cross-town service than what we have today. For example, suppose you live in Greenwood and want to go downtown. You can take 48->Link with a transfer at Roosevelt or just take the 5. With wait time, the 48 portion alone is likely to be almost as long as the 5, even if Link arrives instantly and travels to downtown instantaneously.

      In shoreline, the cross-town service gets even worse. The 510 and 511 buses offer very fast service along I-5 to 145th St. today, but if you want to use those buses to get to shoreline, you have to walk all the way to and from the I-5/145th St. station because, outside the peak, there’s no cross-town service to speak of.

      “I’m pretty young and healthy and yet if you make me walk up the hill on Denny Way every day to catch a bus I’m going to drive”

      On the contrary, the more you walk up Denny Way, the better in shape you get, and the easier the walk becomes. You just have to do it. I consider the opportunity to get exercise during my commute to work via transit an important reason why I don’tdrive.

      1. I agree with your last paragraph here. Back when I was living in Lake Forest Park, I would take the 358 to the 331, but after a while I started taking the 41 to the 347 or 348 and walking from North City to home, about 1.5 miles. Once you get used to it, you enjoy it more.

        Heck, even now, if it’ll be a while before my bus arrives, I’ll walk along the route toward my destination and catch it at a later stop.

    2. Yes, Link will provide the express layer that’s currently missing. The main problem is the 11-year gap until the Lynnwood Extension opens. I’m not worried about crosstown service: it’s still possible that Metro will rise to the occasion when Link opens. Shoreline could also explore options like non-Metro van shuttles — Metro may get more willing to allow these as it converts its own fringe routes (Vashon, Enumclaw) to “alternative service delivery”.

      The issue is that, for now, the 358/E is the gateway between north Seattle and Snohomish County. Therefore it has long-distance responsibilities that the 5 and 28 don’t have. It would be best to just extend Swift to downtown and let the E be like the 101. But Metro has no interest in doing that, unfortunately. In the meantime, I’m glad that the downtown-to-Aurora Village time has decreased from 60 minutes to 45. I’m hoping that RapidRide will bring it down to 40 or 35. That’ll at least be something. And we mustn’t forget about the 15-minute frequency evenings and Sundays. That’s a significant contribution that RapidRide is bringing.

    3. If Rapid Ride is going to be the same as the 358–stopping at all the same stops, running with a similar frequency–then why on earth are we spending all this money turning it into Rapid Ride?

      We have really poor cross-town bus service in the north end. If Rapid Ride won’t get me downtown any faster than the 358, I’d rather they spent the money on frequent, reliable bus service to Northgate and the future Link stations. But since we’re doing Rapid Ride, then for pete’s sake, let’s make it rapid.

  15. If I had my way, I’d have all routes for KCM, ST. and CT have their stops no closer than a radius of .25 miles.

  16. Jarret Walker of Human Transit has some interesting takes on stop spacing, both in the US and in Europe. Aside from the fact that ‘one size does not fit all’, the evolution of stops on early streetcar lines were competing for business with pedestrians, so they stopped very often. Now, we’re competing with automobiles for customers, where trip time becomes the driver for new business. Somehow, Metro and many North American transit agencies didn’t get the memo. People will walk a little further for a faster trip on the bus, rather than ride the ‘Milk Run’ every day.

    1. Yes, that’s the problem. Transit is crippled because agencies don’t put in enough limited-stop and express service, and when they build light rail they put in too many stops (Seattle is an innovator here), so people notice that bus/rail takes twice as long as driving and drive instead. To compete with cars, and to compensate for the walk between the station and people’s actual destination, transit needs to narrow the gap between transit’s speed and driving.

      1. Again, no.

        Denver and Dallas and their ilk have already “innovated” suburban-focused light rail with massive gaps between stops and little utility for anything but rush hour commuting.

        The result had been pathetic ridership.

      2. And Jarett disagrees with you, too, repeatedly describing the 1/2- to 3/4-mile “sweet spot” for rapid transit service and the 1/4- to 1/2-mile sweet spot for local bus service.

        You keep confusing eliminating milk runs with advocating for service you can’t walk to. Your conflation couldn’t possibly be more destructive.

      3. “Transit is crippled because agencies don’t put in enough limited-stop and express service”

        As long as limited-stop express service runs all day 7 days a week in both directions, I could not agree more. That is why I am such a big fan of Sound Transit. If the wide stop spacing means I have to bike a couple of miles to reach it, so be it – it’s well worth it.

      4. “Denver and Dallas and their ilk have already “innovated” suburban-focused light rail with massive gaps between stops and little utility for anything but rush hour commuting.

        The result had been pathetic ridership.”

        However, Denver and Dallas had good reasons at the time: there were massive gaps between developments. Close spacing is useless when you’re stopping at vacant lots. Denver is getting some infill now — and it’s always needed the Colfax Streetcar.

        In Seattle, with natural hills and water boundaries, you don’t have the “flatland” development pattern of Denver, and closer spacing makes more sense.

      5. Though to be fair, once you start heading towards Shoreline, there sure is some of that “flatland” development pattern.

      6. Dallas’s problem is that there are so few walkable areas that almost all stations have little walkshed, and the feeder buses are half-hourly. That’s not the rail line’s fault, it’s a land use problem. (The other problem is that the downtown segment is on surface streets so it’s slow like MAX.) In Denver I’ve only seen a few stations but they looked similar.

        Isn’t Dallas building a streetcar, implying that they’re designating an area for closer stop spacing and denser growth? That should answer DP’s concern without forcing longer-distance travelers to spend an hour on a train running at 13 mph.

      7. I know, Nathaniel, and I don’t begrudge the fewer-than-10,000 commuters per line (the systems each have only 60,000 weekday one-way boardings over 3 lines)who are now lucky enough to park-and-ride to work while avoiding the worst of the traffic.

        The similarity to Link is that both systems basically avoid serving the city entirely, and shoot straight for suburban nodes with little-to-know regional appeal. They don’t offer useful service *between* the various places people actually want to go, and so they are relegated to glorified commuter rail for a tiny minority.

        It’s the same model that Mike claims Seattle is “innovating”.

        Mike, Dallas’s streetcar is a tiny New Urbanist exercise, much like the SLUT and roughly as useful. It’s no replacement for rapid transit that you can *use*.

      8. “The similarity to Link is that both systems basically avoid serving the city entirely”

        The people going to UW, Northgate, Capitol Hill/SCCC, and the stadiums would beg to differ. It serves the top one, two, and three trip markets in the city, as shown by the number of existing parallel express-and-local routes, 7-minute frequency, and overcrowded buses. Unfortunately it addresses Aurora only partially, and it doesn’t address Ballard and West Seattle or Queen Anne much at all. But that’s not the same as “avoiding serving the city entirely”. In fact, one of the most common complaints about Rainier Valley is it serves too much of the city and should have gone through Georgetown. That doesn’t sound like Dallas DART.

      9. People go to the U-district at all hours, and there’s strong evening ridership to Capitol Hill, and weekends to Northgate. It’s not all commuters or even mostly commuters, if you define commuters as “peak hours”.

      10. The people going to UW, Northgate, Capitol Hill/SCCC, and the stadiums would beg to differ.

        Unless they’re going to UW, Northgate, or SCCC from anywhere other than downtown or those three places!!!

        Admit it, Mike, Link will be an easier way to the UW for someone with a car to park in Lynnwood then for someone on top of Capitol Hill.

        That is not “serving the city.” It’s totally fucked up!

      11. You know what, Mike? Sorry to harp on this, but let’s switch our example to First Hill.

        Regardless of the history of the First Hill Link station or its deletion, the *result* is going to be the precise manifestation of what you always advocate — mile-plus rapid transit gaps and an ultra-local overlay (in the form of the Firsl Hill Streetcar).

        And you know what? It will be terrible! First Hill to UW (a trip that happens all the time) with nearly twice the travel time as Lynnwood to UW (a trip that happens once a day at most).

        THAT is what happens when you skip over the city, so that no one can access your rapid transit on foot.

        Denver and Dallas are slightly more extreme examples, but it is the same horrendous mistake.

  17. Funny…80th and Aurora was my bus stop for a year. There is no reason to keep it. 85th works just fine.

  18. This is a misuse of taxpayer money. The City of Shoreline has a history in botching their transit efforts. With Sound Transit express buses, they got nothing. Okay, two stops south of city limits, only serviced off-peak, the SB stop too far from the nearest Park & Ride, a tiny one at that, for all but the able-bodied and weather-averse. With Rapid Ride, rather than seek underlying service – if stop coverage is what they’re so interested in – they further neutered the “rapid” in RapidRide, akin to when Metro had a limited-stop #360 that took as long as the #358 to go end-to-end. With light rail, the Interurban Trail provided a great foundation for it, yet they pushed for an I-5 routing…and won. Presently, they’re lobbying for a station at N. 145th, a street they don’t fully own, where there’s over 30,000 vehicles/day to the east, a traffic nightmare for decades, no development planned along, and a location subject to environmental issues and costs (Thornton Creek), while being down the hill – and just 15 blocks away from – a N. 130th station that has no alternative, i.e. that location’s likely a “go.” What was sad and shocking were that officials like Seattle Councilmember Conlin and Edmonds Mayor Earling were willing to cede their decision making as (regional) Sound Transit boardmembers to Shoreline officials.

    @Bruce Nourish I disagree with your notion that a local underlay is a terrible idea. All one has to do is look to the north, where Community Transit’s #101 provides closer-stop spacing that their Swift/BRT does not, and they operate in harmony. Of course, I’ll concede that RapidRide’s speed and reliability-killing features: on-board fare payment, traditional bicyle and wheelchair loading, might make it a terrible idea, however. If not, on Aurora, the City of Shoreline could have lobbied for moving something like the #345 and/or #316 over to cover parts of Aurora (instead of duplicating parts of the #346), which would’ve covered northern Seattle as well, at least 115th northward.

    1. Outside of people with serious mobility issues, who is going to wait for an hourly 101 bus (with no real-time OneBusAway info, so who knows when it’s really coming) to avoid walking or biking 1/2 mile to a much more frequent Swift bus? For me, I would say if the walk to the nearest Swift stop takes too long, the solution is to take a bike, or even a skateboard, not wait for the 101.

  19. Is there a pedestrian route from the stop at Lynn up onto Queen Anne Hill? Or just down towards Dexter via that pedestrian underpass?

Comments are closed.