A few weeks back, Vancouver, BC-based urban planner Zach Shaner posted a comparison of bus transit service in Seattle and Vancouver. Shaner’s basic claim—that Vancouver’s bus service just works better than Seattle’s—is hard to dispute. A few years ago I reached the same conclusion, based on the fact that TransLink, BC’s transit agency, provided far more bus rides per capita than did greater Seattle’s.
But while I totally agree with Shaner’s diagnosis, I just don’t buy his explanation for Vancouver’s transit advantage. Shaner argues that three key factors vault Vancouver’s transit service ahead of Seattle’s: (A) more frequent bus service; (B) a single unified bus system rather than 6 separate ones; and (C) fewer hills.
While each of these factors is legit, at least in a limited way, I think Shaner is overlooking a root cause that’s far more critical than any of those: Vancouver’s vast advantage in limiting sprawl, and creating compact, transit-oriented neighborhoods.
I think the numbers back me up on this.
One easy way to see the importance of neighborhood design is to look at metrics of transit service efficiency. Look, for example, at the number of riders who board the bus per hour of service. The higher that number, the more riders a bus system can serve with a limited pot of money. TransLink reports that about 47 riders board the bus (numbers derived from p. 11 of this TransLink quarterly report), on average, during each hour of bus service. But according to the National Transit Database, the figure is just 33 boardings per hour in greater Puget Sound, counting all 6 bus agencies that service the region.
Now, I suppose it’s possible to argue that if only greater Seattle ran buses more often, its transit service would gradually get more efficient, with more and more passengers getting on the bus during each hour of bus service. Unfortunately, this appears to be wishful thinking. Yes, when bus frequency goes up, more people do ride the bus. But studies show that ridership tends to grow more slowly than does bus service frequency. The always-helpful Victoria Transport Policy Institute has the scoop, in a review of “transit elasticities“:
The elasticity of transit use with respect to transit service frequency (called a headway elasticity) averages 0.5, with greater effects where service is infrequent.
So, roughly speaking, a 10% increase in bus frequency nets a 5% increase in passengers. Economists call this a relatively inelastic relationship. And it means that, if buses ran more frequently in greater Seattle, then boardings-per-service-hour would likely decline, rather than improve to Vancouver’s levels. (Five percent more riders spread across 10 percent more buses means that each bus would be a bit emptier than it is today—which might be ok if you’re a rider, but a costly inefficiency if you’re a bus system planner with a limited budget.) With fewer riders per service hour, it becomes a severe financial strain to support high-frequency bus service.
That suggests that frequency of service isn’t at the root of Vancouver’s transit edge. Or, at least, it’s not an independent explanation for that edge. Instead, Vancouver’s high-frequency bus service is made possible by a far more fundamental factor: Vancouver’s more compact, transit-oriented neighborhoods put more jobs and people near transit lines. And this compact neighborhood design makes it possible to support high frequency service with high ridership per bus.
It’s kind of obvious if you think about it. When people live close together, a single transit stop can service lots of people at once. With stores and services close together, transit can link people to more of the sort places that they want to be, making cars less necessary. Less sprawl means that bus routes can service more people and more destinations with less driving. Compact, vibrant neighborhoods tend to have less surface parking than sprawling locales; and according to VTPI’s research, parking prices and supply are also major determinants of transit usage.
In short, compact neighborhoods create a virtuous circle for transit ridership. Metro Vancouver is far and away the most compact urban area in the Pacific Northwest; our analysis shows that no other city even comes close, really. And that fact gives greater Vancouver a huge transit advantage over far the far more sprawling Everett-Seattle-Tacoma corridor.
I certainly don’t mean to be harsh on Shaner. He’s got some good points. Still, it always strikes me as wrong-headed to start any analysis of a transportation system by looking at the characteristics of the roads and vehicles. More important is where the people are, and where they want to go. If you start by looking at people and places first, then you’ve got a much better shot at understanding the dynamics of how the transportation system works.
Bus Photo courtesy of Flickr user Stephen Rees under a Creative Commons license.
Skot Nelson
The lack of a single, straight freeway slashing through the city is likely a contributor here as well.Despite traffic jams and a glacial pace through rush hour times, the Seattle’s freeways provide a convenient route to get from point A to point B pretty much anywhere in Seattle. Buses don’t tend to use it (if at all? some do in Vancouver) making their routes longer and more circuitous.In Vancouver, drivers don’t have a freeway option, which makes getting anywhere an exercise in taking city streets complete with stop signs, traffic lights and other things that make trip times longer.Of course there’s no *one* factor that contributes to the difference between the two locations. I’m just mentally speculating that this might be a partial contribution. I cycle everywhere anyway…
Clark Williams-Derry
Agreed—Vancouver’s smaller road system means that cars aren’t such a great option for getting around—which also supports more compact development.
Michael D. Setty
There are two factors that explain Vancouver’s greater success with transit than Seattle’s. I think density and neighborhoods are a secondary factor.First, Translink operates much more frequent local services that create a much greater network effect than in Seattle. That is, local bus services in Vancouver provide much more frequent connections in a network that allows easier travel to much of the region outside downtown. Secondly, Seattle’s transit system dedicates a very large share of its resources to less productive express services, which generally leads to longer headways, particularly in the suburbs. Vancouver has much less express bus service compared to Seattle, with most longer distance travel by transit occuring on the rail transit trunk lines.The “elasticity of service” may only be 0.5 when service operates every 15 minutes or more often, but it can be much higher when potential connections are considered. Seattle ridership and productivity suffers compared to Vancouver because Seattle generally tries to provide direct services where comparable situations in Vancouver have much more frequent local services feeding extremely frequent rail trunk lines. So far, Seattle is still providing many far flung expresses that parallel the LRT line, even though it would be much cheaper per passenger mile to feed as many riders as possible to the LRT line (which would be much cheaper to operate per passenger mile at higher volumes), and allowing buses to make more trips during each peak period than they can when operating express.
jesse
Why compare Vancouver to Seattle? The only thing they have in common is their geographic proximity. If you want real transit solutions, pick a city with an exceptional transit system, not one that happens to be close by.
Bryan
i think the general attitude about transit in the US is very different than at least here in Vancouver. one thing, we don’t see riding the bus as degrading, or as a symbol of lower economic status. from my understanding in the US on average people use the bus as a last resort, when they cant afford a car, or have lost their license. maybe because of the car representing american freedom? I bet on average there’s going to be more bus riders per capita in Vancouver than most american cities (this may include other major canadian cities) just because of the status symbol effect. Vancouver is fairly young as a city and developed quite closely with its transit system, from my understanding many european cities have a similar attitude, possibly because their transit system developed before most people had cars of their own.
Jeff
Why no mention of the 80/20 rule that is in effect? Is it any wonder that ridership is low when such a high percentage of bus traffic needs to support outlying suburban/ex-urban communities? The tax payers should be up in arms over such an inefficient use of tax payer dollars.
eldan
There’s something to all of these, but I think there was also a design decision that Vancouver’s transit services made a long time ago which has served them well. They just don’t have as many bus routes as we do. It has the drawback that there can be longer walks to the bus stop, but the huge advantage of concentrating service, which allows headways to be shorter. The trouble is, now that Seattle’s suburbs have all these bus routes, it’s political hell to combine any of them and make a more concentrated system.
Eric Doherty
Ya sure, the neighbournoods and density in Metro Vancouver help. But how did they get that way? Was it that our car dealers and property speculators didn’t want to create freeways and sprawl everywhere?No. People fought like hell for better transit and against freeways in the 1960s and early 70s. And they are still fighting – even in Delta and Surrey.e.g. http://www.livableregion.ca/blog/blogs/index.php/2010/05/26/video_aamp_media_waterfront_park_not_watThings don’t just happen, people get together and make them happen.
Zach Shaner
It’s not so much that I discounted density, but rather that I wanted to look for ADDITIONAL reasons. I explicitly left out the impact of SkyTrain and zoning laws in order to compare things on a bus-for-bus basis. Your analysis is surely correct that Vancouver’s structural transit advantage is due the high densities of the West End, the Broadway corridor, and TOD around SkyTrain stations. But I was also focusing on the end-user experience, and frequency and branding are very important to that experience.Thanks for the feature!
Clark Williams-Derry
Eldan & Jeff –I think you both have VERY good points. Jeff – Agreed, 80/20 is dopey—a terrible way to apportion service. But Seattle (or, really, the “West” subregion in Metro’s way of dividing up the world) still has about 60% of the region’s bus service. And that hasn’t changed much for a while, since the 80/20 rule is mostly about divvying up new service.Eldan – Part of the reason that Seattle has more routes is that it’s serving a more dispersed population—thanks in part to 80/20, and the mindset behind it. I look at the Metro Vancouver transit system map (http://bit.ly/9jxDYc) it’s hard to find housing that’s more than 5 blocks’ walk from a bus route. Seattle’s similar, though it appears to me like there are more “holes” where it’s a long or inconvenient walk to get to transit (http://bit.ly/9POfm4). But when you’re trying to put buses near where the people are, you need fewer routes in Vancouver. (For comparisons, the populations of metro Vancouver & King County are similar.)Eric -Absolutely right. Vancouver’s density isn’t an accident. Well, some of it’s just luck—it’s hard to build houses in mountains. But it’s not just the lack of highways, but also the Agricultural Land Reserve—which I think people underestimate when they think about Vancouver’s urban form.
Clark Williams-Derry
Zach -Fair enough!My sense is that there are parts of Seattle that offer Vancouver-level frequency. Rush hour buses from downtown to the U district & Northgate are the ones I’m most familiar with. I’m sure there are others—buses from downtown to Ballard, maybe some of the buses towards the south end.So the difference, perhaps, is that Vancouver has more nodes & corridors of density that support frequent service and a better user experience.
Liz
I am suspicous that the statistics on average elasticity of transit use do not bear much relationship with specific realities. My belief (although I have no data to prove this) is that in neighborhoods with appropriate density, frequency of service has a much higher impact (probably a factor greater than 1x) on ridership. I also have to believe that it is not a linear relationship; in other words there is likely a “tipping point” at somewhere closer to 5 minute headways where MANY more people are suddenly willing to take the bus because it actually becomes dependable and convenient for them relative to other choices like jumping in their car or grabbing a taxi. And I’m willing to bet that just about anyone is willing to stand in the rain for less than 5 minutes, but very few will do so for 10 minutes or more.