A Tale of Two Cities: Vancouver, Seattle, and the Perils of the One-Seat-Ride

Apr 22nd, 2010

I hate to beat up on Seattle.  I love the Emerald City, and I have lived there quite successfully without a car.  But after living here in Vancouver for 4 months, I am unequivocally convinced that Vancouver offers far superior bus service.  Yet it isn’t immediately clear why this should be so.  Transit is everywhere in Seattle, and the five primary agencies that offer bus service (King County Metro, Sound Transit, Pierce Transit, Community Transit, and Everett Transit) cumulatively offer just as many annual service hours (5.86 million) as Vancouver’s TransLink (6.18 million).

Yet the qualitative experience of transit in the two cities is incredibly different.  In Vancouver there’s just a real je ne sais quoi; I really feel like I can go car-free, put on my backpack, and walk anywhere I want and take transit anywhere I want without planning any of my journeys.  The routes are intuitive, frequent, and they just work.  In Seattle, even though I know I’m surrounded by options, they somehow seem indecipherable.  I decided to search for reasons why.  And I want to disregard the most obvious reason:  the dearth of rail service in Seattle.   It is very obvious that SkyTrain anchors scores of bus lines, provides a backbone and decipherable structure to the network, and delivers complete reliability. But for the purposes of this post, on a bus-for-bus basis, why does Vancouver do so well and Seattle so poorly?   Daily bus-only ridership in metro Seattle – defined roughly as Snohomish, King, and Pierce counties – is roughly 530,000.  Bus-only ridership in metro Vancouver is roughly 825,000.  Despite offering only 5% more bus service hours, TransLink attracts 55% more ridership.  How does it do this?    After crunching some basic numbers, I have three answers: FREQUENCY, BRANDING, and TOPOGRAPHY.

FREQUENCY

Simply put, Seattle offers too much of the wrong kind of service.  While TransLink provides 215 bus routes, the 5-agency Seattle area provides 83% more, an astonishing 395 separate routes.  Of these, a whopping 37% are commuter routes without any mid-day service at all.  Only 13% of TransLink’s routes (29 total) are commuter-only.  With very few exceptions, in Vancouver you know you can catch any bus reliably throughout the day.  While Seattle offers the commuter market a wealth of one-seat rides to the downtown core, it does so at the explicit expense of the system’s intuitiveness and its ability to compete with the spontaneity of the car.

And the problem goes deeper.  Seattle has an unfortunate affection for the 30-minute headway. Even during peak hours, 50% of its routes run at 30-minute headways, compared to 30% for Vancouver.  And Vancouver does even better with its busiest routes.  During mid-day hours, Seattle’s bus service falls off a cliff;  only 10% of routes run every 15 minutes or better, 20% run every hour or less, and as already mentioned 37% don’t run at all.  At noon, 28% of Vancouver’s routes run every 15 minutes or better, and a couple dozen (such as 3, 5, 6, 8, 9, 99 , 145 etc…) run 8 minutes or better.  In short, Vancouver provides those along its busiest lines the ability to travel spontaneously, the crucial factor that takes modal share away from cars.

BRANDING

The structure and politics of transit governance in Seattle is complex and difficult to understand.  For some reason, counties are the political unit of choice for transit planning, leading to 20+ transit agencies in Western Washington.  Lack of coordination and excessive administrative overhead have led to duplication of services and poor regional connections.  The creation of Sound Transit, a competent and impressive regional body responsible for regional planning and light rail expansion, has helped matters greatly.  So has the implentation of the ORCA smartcard for fare payment, which has greatly reduced what used to be 300 different kinds of transit passes and transfer slips.

Yet difficulties abound.  Each agency produces their own timetable literature, maintains different customer service centers, and offers online trip planners that offer insufficient information about connecting services (Community Transit has the best trip planner).  Regional travel remains remarkably opaque, especially for tourists.  Seattle offers no day-passes to tourists, and their well-intentioned Ride Free Area backfires, as confused customers sometimes have to pay when getting on, sometimes when getting off, or sometimes twice.  By contrast, TransLink offers a $9 day pass, offers automated next-bus information at each stop, has realtime GPS stop announcements at every stop on every route, and keeps fares to a simple 3-zone system ($2.50, $3.75, and $5.00).

However, Seattle’s in a bind.  Transit politics being what they are, with the faux-libertarian anti-transit activists always eager to stifle investment, Seattle’s compartmentalization has kept it alive.  A consolidation of transit governance would leave service levels even more vulnerable to changing political winds.  At least as separate entities there are more proverbial moles for the autophiles to whack.  The price is a chaotic framework with mediocre service.  But brand consolidation could be done without affecting the underlying funding of each particular route.  Why not take the Sound Transit name, make all buses carry its livery, fare structure, etc…  and distribute the proceeds to counties as necessary?  Make things more intuitive for the rider!

TOPOGRAPHY

Lastly, it is well-known that Seattle is just a physically difficult place to get around.  There are huge lakes, steep hills, ship canals, and an overall dearth of flat land on which to build grids or arterials.  As Jarrett at Human Transit succintly put it:

Nowhere in Seattle can you travel in a straight line for more than a few miles without going into the water or over a cliff.

While Vancouver has hydrochallenges of its own, the land itself is quite flat, especially in the densest areas. This makes grids function well, makes the city easy to understand, and allows a broader range of equipment to be used in all weather conditions (as opposed to Seattle’s blizzard nightmare of 2008).

For these and other reasons, Vancouver’s just got it.  Seattle can make up for probably 80% of what it lacks through visionary planning, and if it does, its physical quirks can be charming rather than infuriating.  Both cities have dedicated operators and planners, and both cities provide competent service, but Vancouver just continues to astonish me with an abundance of the one thing Seattle lacks:  ease.

Tags:
  1. Jarrett at HumanTransit.org
    Apr 24th, 2010 at 05:27
    Quote | #2

    Excellent post. Note that you’re comparing totals for the entire urban region, though your geography notes seem to focus on the core city. Seattle simply has a lot more suburbs than Vancouver, and suburbs tend to require more numerous and less frequent routes to cover all their nooks and crannies. Within the core city, Vancouver’s system is simple because it’s designed to work with the grid. Seattle planners understand grid principles but find them almost impossible to implement because of the obstructed geography.

    One way to assess SkyTrain’s impact is to note how many different kinds of express buses you’d have on Highway 1 if SkyTrain didn’t exist.

    • Zach
      Apr 24th, 2010 at 11:09
      Quote | #3

      True, I was comparing the entire urban region, and for fairness’ sake I probably should have included Abbotsford/ValleyMax into the analysis for Vancouver. But the agency separation of metro Seattle makes apples-to-apples comparisons difficult.

      For me Seattle’s rail expansion (especially between Westlake–>UW–>Northgate) will only be successful if it sparks a radical reorganization of bus service, including cutting commuter-only routes by 2/3 at least. As you’ve said before, transferring can be a blessing, and after North Link opens any north-south bus route within 5km of I-5 should probably become a frequent feeder service, such as are ubiquitous on SkyTrain.

      Thanks for commenting!

    • Mike Orr
      Apr 26th, 2010 at 12:25
      Quote | #4

      The main impediment to a grid-based system is not geography but protestors. People howl at any proposal to consolidate parallel bus routes to downtown. You tell them it will increase frequency to the neighborhood, but they don’t care about frequency, they just want a bus on their street to downtown.

      When light rail opened, the downtown route on its street (#42, MLK Way) was going to be removed, and an alternate route to Capitol Hill (8) extended to provide servce on MLK. But some people at a Asian certain community center flooded the City Council with calls, protesting that they wanted a one-seat ride from downtown. So the 42 was half-reinstated, running once an hour with limited hours on a shorter route. This cut into hours planned for an east-west circulator (39), which now runs every 45 minutes until 7pm. So the partial reinstatement duplicates service and prevents other neighborhood residents from getting to the train without walking a long hilly distance. But because they had existing bus service, they won out over residents who did not have existing service.

      • Mike Orr
        Apr 26th, 2010 at 12:29
        Quote | #5

        AND there’s a frequent parallel route a few blocks east (#7).

      • Zach
        Apr 27th, 2010 at 15:31
        Quote | #6

        Regarding NIMBYism and bus reallocation, I’m cautiously optimistic about Link and Sounder’s future ability to restructure services. I’m most optimistic about ULink and North Link; I’m pretty sure that some routes such as 41, the 70 series, etc…will be happily truncated or eliminated. Where I’m worried about it is in Bellevue, South Seattle, Tukwila, Federal Way, and all along Sounder. It’ll be almost impossible to cut routes there, even if it improves the system greatly.

  2. Adrian
    Apr 24th, 2010 at 14:32
    Quote | #7

    Quite an interesting comparison, I especially love your analysis on the branding issue, and I agree that’s a bit of problem. I want to point out that Melbourne has their “metlink” branding system, which unifies all modes of public transportation and provides consistent signange, map information, and schedules for its customers. Something like this should be required if Seattle continues with their agency separation.

    • Zach
      Apr 25th, 2010 at 21:11
      Quote | #8

      Transit politics being what they are, outright agency consolidation is usually too much to hope for. That’s why I focused on branding…the rider doesn’t need to know how the beans are counted…but merely what services are available, where they’re available, and how often.

  3. EngineerScotty
    Apr 25th, 2010 at 07:25
    Quote | #9

    One thing missing from your analysis–and it’s a thing that Seattle has lots of (even though they are frequently crowded) and Vancouver has far less of:

    Freeways.

    Seattle, like most US cities, has a comprehensive and connected freeway network–making auto trips convenient. Topography keeps Seattle from looking like Phoenix or Atlanta, obviously, but two interstates penetrate downtown, as does Alaskan Way; numerous others serve the suburbs. It gets worse as you head south towards Tacoma; the south sound region has tons of them.

    Vancouver has very few freeway miles within the city limits (a short stretch of the TransCanada Highway near the eastern edge, and a few highway bridges from the south which turn into surface streets as soon as they cross the Fraser). If you want to get from, say, North Vancouver or Burnaby to Richmond or the airport, there’s no all-freeway route you can drive–none. And nothing resembling a freeway gets anywhere near downtown.

    A common factor in many great North American transit cities, especially places like Vancouver and Calgary which are not Manhattan, is less infrastructure dedicated to the automobile.

    • Zach
      Apr 25th, 2010 at 10:09

      Yeah. I live in Vancouver’s west side near UBC, and there’s not a freeway for 10km in any direction, and it makes for a blissful urban experience. I’ve been in Vancouver for 5 months and haven’t yet been on Highway 1, either as driver or transit rider. So even when Seattle builds Link Light Rail to Bellevue, Northgate, and eventually to Everett and Tacoma, it’ll be interesting to see if they can successfully compete against an overbuilt freeway infrastructure. My guess is that they’ll be pretty successful on the Seattle-Redmond line and as far north as Northgate and (maybe) Lynnwood, but that their eventual expansion in the South Sound region (with the mess of freeways… I-5, SR 167, SR 512, SR 18 etc…) will attract a far lower relative share of ridership.

    • Anc
      Apr 25th, 2010 at 14:54

      Has any city ever removed a Freeway once it was built?

      Probably just fantasy, but I wonder if one day Seattle will ever be able to get rid of I-5, turning 405 into the new I-5… *sigh*

      • Corey Burger
        Apr 25th, 2010 at 18:10

        Yep, lots of cities have. San Francisco’s Embarcado and Seoul’s Cheonggyecheon come to mind. The challenge is that every city has to overcome exactly the same arguments about traffic, etc., even though every city that has taken out a freeway has had exactly the same experience: nothing bad happens.

        • Aaron W.
          Apr 25th, 2010 at 19:06

          While both freeways, neither Harbor Drive in Portland nor the Emarcadero in San Francisco were Interstate Freeways like I-5 or even I-405 in Seattle. A more apt comparison would be to compare them to the Alaska Way Viaduct in Seattle.

          • EngineerScotty
            Apr 26th, 2010 at 09:44

            The removal of Harbor Drive was fairly uncontroversial, as it was functionally obsolete (nowhere near modern standards even in the 1970s) and had been rendered unimportant as a regional highway link with the construction of I-5 and I-405 through downtown Portland.

            While this is a pretty macabre thing to say–the best hope for freeway removal seems to be an earthquake. :(

          • Kamala Rao
            Apr 26th, 2010 at 11:01

            My research found that the Embarcadero carried just as much traffic before it was removed as the Seattle Viaduct does today (about 100K vehiclies per day for both) – and Cheonggyecheon carried nearly twice as much (170K vehicles/day), so I think its a fair comparison.

          • j steve
            Apr 27th, 2010 at 10:51

            Considering the battle over just the viaduct, I can’t see the political reality of tearing down I-5 within the city over the next 50 years.
            I will say, at least the “option” of simply tearing the viaduct down and adding more transit options was out there and discussed, even if it was in false terms. 20 or 30 years ago, that wouldn’t have even been a glimmer in the eyes of the city to do.

      • Zach
        Apr 25th, 2010 at 21:14

        The MOST we can ever hope for is for I-5 to be lidded Yesler Way to Denny Way. If I’m really feeling generous we MIGHT someday be able to co-opt parts of I-5 for HSR, using the express lanes to Northgate and then either median or side-running from there.

  4. Dave
    Apr 25th, 2010 at 09:56

    I think you make a great point about 30 minute headways. Even for commuters who know and rely on that line, a 30 minute headway can be terribly inconvenient. If you miss a bus, your whole day can be thrown off. I’d rather have fewer routes and walk a little farther to a stop knowing the bus comes every 8 minutes.

    This post also reminded me of the streetfilm on increasing the distance between bus stops: http://www.streetfilms.org/making-muni-faster-and-more-reliable-through-bus-stop-consolidation/

    I understand that bus service is a “last mile” kind of service, but if we start to treat it more like rapid transit, I believe you’d see a lot more riders.

    • Zach
      Apr 25th, 2010 at 10:01

      Absolutely. 30-minute headways are fine for the fixed-demand commuter market, but we just can’t take modal share away from cars until we focus on off-peak frequency and personal spontaneity.

  5. thettransitfan
    Apr 25th, 2010 at 13:01

    Awesome article Zach. I didn’t realize how different Seattle’s bus system was to Vancouver’s in terms of frequency, especially midday frequency. The most frequent routes in Vancouver are generally the trolleybus routes which replaced the streetcar lines in the 1940s and 1950s. Headways are generally very good until late night hours. Improved frequency between 7pm and 2am is something that Vancouver still needs to work on. Not sure how Seattle compares.

    EngineerScotty makes a good point – an absence of freeways in Vancouver has definitely helped here. One result has been long-term continuity for our transit routes. We’ve basically just been adding routes over the past 50-60 years. Because the urban fabric wasn’t ripped up to accommodate freeways, our travel patterns weren’t disrupted either. If you get on the #10 in Marpole today heading downtown you’ll follow exactly the same route that you would have in 1910 on the streetcar.

    Looking forward to more transit articles – you should get on Twitter if you’re going to be posting regularly!

    • Zach
      Apr 26th, 2010 at 11:33

      Seattle’s actually much the same. Its old streetcar routes still roughly define many of King County Metro’s busiest routes. The difference, as mentioned, is frequency.

  6. Chad N
    Apr 25th, 2010 at 18:57

    Great post! I’ve noticed the same qualitative difference in bus transit between the two cities and its great to see actual data confirming it. Vancouver’s advantage in bus service hours would become more clear if you compared regions with similar population. King County alone (leaving out Pierce and Snohomish Cos.) has 1.9 million, while Vancouver metro has 2.1 million. Since most Sound Transit bus service is inter-county, comparing Metro Transit (King Co.) directly to Translink would be more relevant.

  7. Paul C
    Apr 25th, 2010 at 23:09

    Another factor is the higher average cost of housing in Vancouver. It has forced more people to give up the car and take transit or if in a family maybe only having one car for one spouse while the other spouse takes transit.

    • Zach
      Apr 26th, 2010 at 11:31

      Agreed. My $1,150 rent doesn’t get me much. A mere 450 sq. ft., and I’m lucky to live in a rent-controlled flat!

  8. The Zurich Model « Voony's Blog
    Apr 25th, 2010 at 23:22
    #26
  9. Online CNA Classes
    Apr 25th, 2010 at 23:26

    Terrific work! This is the type of information that should be shared around the web. Shame on the search engines for not positioning this post higher!

  10. MikeP
    Apr 26th, 2010 at 07:52

    Some others have mentioned this, but the single most important factor, certainly before topography, is density! Vancouver proper has 13,817.6 people/sq mi while Seattle has 7,136/sq mi. It is even worse for metro areas:
    Vancouver 2,116,581 ppl/1,111.4 sq mi = 1,904/sq mi
    vs. Seattle 3,407,848 ppl/8,186 sq mi = 416/sq mi

    That is why Seattle has worse transit.

    Thx Wikipedia

    • Zach
      Apr 26th, 2010 at 11:29

      Your point is definitely valid, but I’d be interested to differentiate the density numbers a bit more carefully. While Seattle doesn’t have anything analogous to the residential density of Vancouver’s West End, I’d bet if we exclude the West End then city-proper density in both cities would be roughly equivalent. Ballard, the UDistrict, and Capitol Hill feel exactly like Vancouver’s Broadway corridor to me, and I’d bet that Vancouver’s vast swaths of huge single-family homes (Quilchena, Kerrisdale, Marpole, etc…) match or exceed the low densities of Seattle’s Magnolia.

      In short, what sets Vancouver apart, density wise, is the West End. What sets its metro area apart is the clustered density around SkyTrain stations in Burnaby and New Westminster. Seattle’s rail build out should allow it to play catch-up pretty well…if the NIMBYs don’t block the critical upzoning.

      • Paul C
        Apr 27th, 2010 at 22:51

        I used to think the same thing that the much higher population density in Vancouver city proper was attributed to the high density in the downtown core. A few weeks ago though I was checking out the census tract for the area I live in in east Vancouver. The census tract is the smallest area that the census people look at has an area. larger areas would be the city the metro region the province etc. Now my census tract is about 1.14 sq km. If you were to drive around the area the majority of what you would see is detached homes. There are a few apart buildings but nothing major. And everything is under 4 stories high. So most people would assume the density level couldn’t be all that high. Even I would of thought the same thing. What I found out was there are approximately 13,900 / sq mile. Which brings it around to what the average for the whole city is. So even without the downtown core the other areas are dense in their own way. Another thing I discovered was apparently 40% of the people live in apartments. Now I know there are not that many in my area. So that leads me to think that those people are living in secondary suites or basement suites in the houses.

        The downside though to comparing density levels. Is what exactly is being counted as the area. Is all of that area habitable or is some of it a case where no one could even live. Example is in the Metro area of Vancouver. If you look at the border in the north shore. You will notice that it crosses through the watersheds. Point being that no could even live that high up those mountains so even though we are counting it we could never increase the density in that area. Even Seattle has the same thing there are areas that you can’t live on. Another case is the inlets or in Vancouver’s the farm land that is not to be developed.

        As for what brought about more people using the bus in Vancouver. I don’t think it is one thing but a combination of things that worked together.

        Jarret Walker touched on it in his Human Transit blog. I’m not sure about Seattle and its lay out. But in Vancouver proper. Your anchor points or starting points for the bus routes other than the southern ends. Have very high ridership. Example every bus going west bound on a east west route. Either ends up Downtown or at UBC. You are almost guaranteed without any difficulty of getting packed buses in the morning going in that direction. The eastern side of those routes most likely starts at a sky train station another good anchor point. In the north south routes. Almost all of them go through downtown. So any bus going north in the morning will be busy and any bus going south in the afternoon will be busy. The low point is really the south side of the north south routes. This layout led to a very easy way for people to either get downtown or get to UBC or go east to the Skytrain.

        The big decsion to not build freeways in downtown helped by making people realize hey you can drive in if you want. But your not going to get a quick easy route. And we are going to make you wait at those lights whether you like it or not. Also as I mentioned before the higher house prices had a major factor. Although I do feel the higher house prices, the no freeway, the higher density. All had a factor in helping each other out. Which one started it is hard to say.

        So in the end it wasn’t just one decision it was multiple little things that happened over time. The layout of the bus service was good with the almost perfect grid. And the other factors led to more people living in the area that choose to use the bus because they didn’t like the traffic or just couldn’t afford to drive. :)

        Ack I’m long winded :)

  11. uberVU – social comments
    Apr 27th, 2010 at 08:38
    #33
  12. JB
    Apr 27th, 2010 at 08:52

    Hi Zach,
    I agree on almost all your analyses of the transportation system in both Vancouver and Seattle. I was very confused to understand the bus system in Seattle when I was there so instead of using it I walked all over the city so I won’t get lost.
    There is another important issue that you can add to your research which will explain a bit more why the bus use in Vancouver is more frequent. In my humble opinion it is the ‘mixed-use’ nature of Vancouver in general (Downtown and suburbs around the main corridors) (mostly Residential condos with Commercial, Offices even educational institute in the same complex) comparing to the single use (Commercial and Office building with very few residential) of downtown Seattle and the sprawls of residential use in the suburbs where car is still a big part of the American lifestyle, I was scared when I looked from my hotel window to see the mean highway that goes through the city.
    I also agree with your suggestion of Branding, it’s defiantly a good start for making things simpler.

  13. Jeff Welch
    Apr 27th, 2010 at 10:01

    Good article. The conclusion that “visionary planning” will make up for the bulk of Seattle’s deficits compared to Vancouver however misses what was hinted at earlier in the piece: the lack of political will to change.

    As you said – anti-transit faux-libertarianism is an issue, but that only scratches the surface. Patchwork taxing districts all demanding an equitable share of the transit pie, as well as the motive for millions of dollars worth of members of the $100K Club needing to perennially justify their existence aren’t factors of “visionary planning”, but of the politics of individual and regional financing.

    37% of Seattle routes operate only at peak times? Guess what percentage of Metro drivers are Part Time Transit Operators. Yup. 37%.

    • Zach
      Apr 27th, 2010 at 15:34

      Re: 37%. That’s quite a delicious coincidence.

  14. Jeff Welch
    Apr 27th, 2010 at 10:02

    Sorry – post got truncated.

    Last bit notes that 37% – the number of routes operating during peak times only – is the same percentage as the amount of Part Time Metro drivers.

  15. Kari Watkins
    Apr 27th, 2010 at 17:05

    Great post, but you are missing the critical difference in your numbers. Some folks are on the right track with population – Vancouver is much denser. Translink’s 6 million annual service hours cover a service area of about 700 sq miles (1800 sq km). KC Metro alone covers 2000 sq miles – and I believe your analysis included Pierce, Community and Everett to get the 6 million annual service hours in greater Seattle. So those same service hours are spread over a MUCH greater area.

Comments are closed.