You know the feeling. You hop on local route XYZ, hoping for a somewhat speedy journey to work, home, or play. Maybe today the bus will hit the jackpot. Then the familiar sound comes once every block, a soothing voice speaks the words “Stop Requested.” Sometimes it happens the instant the bus is pulling away from its previous stop. Your inner Hulk flares up. But then! A glimmer of hope – a new block without anyone grasping for the yellow cord! But, somehow the Universe knows. Someone boards at the next stop anyway.
It doesn’t matter if it’s rush hour on the 4 or catching a southbound 6 at 9 PM after a Thursday night Gopher football game (ok, both personal experiences). To make matters worse, you become increasingly aware of how many cars without yielding to your bus trying to re-enter the stream. You curse those exiting when you miss the next green by mere seconds (and sometimes see the same person walk by you out the window!). The whole ordeal certainly contributes to local routes that average well below 10 mph operating speeds. What’s to be done?
Background
Let’s get some basics of transit service, passenger behavior, and other items out on the table for those curious.
- A “walk-shed” is the geographic area that transit planners assume a bus/rail/gondola/zeppelin stop or station will draw passengers from. This is typically assumed to be 1/4 mile for bus stops (more on that later)
- A transit service plan requires a tricky balance between operational efficiency and access to adjacent land uses. Spread the stops out too wide and you’re serving fewer people and destinations within a given walk shed (but speeding up service). Place them too closely together and you duplicate coverage areas and increase trip times, but serve more people and businesses.
In Minneapolis and St Paul, most of our streets (outside the CBDs, some of SE Minneapolis, and perhaps a few other odd neighborhoods) were largely built out as streetcars serviced new neighborhoods (with the main goal of getting those people to each downtown). To minimize number of intersections crossed per mile traveled, the blocks were built as rectangles, twice as long as they are wide. Minneapolis is largely structured north-south and St Paul east-west (following the streetcar routes), but the blocks are the same: 660′ long (1/8 mile) by 330′ wide (1/16th mile). This isn’t abnormal – many neighborhoods of midwestern cities that boomed during the streetcar era share similar block proportions and dimensions.
Many factors play into this: culture, local climate, daily weather, how pleasant the walk is, how safe the walk is, and how desirable the transit service is you’re walking to. Many European transit systems already recommend urban, local routes use 400-600m spacing (link, page 5) – roughly 1/4 to 1/3 of a mile.
The Case for Quarter Mile Spacing
No transit planner worth their chops would be surprised by the information above. In fact, Metro Transit is doing a great job focusing on the arterial BRT program, which implements wider stop spacing (among other huge improvements) on many key urban transit corridors. The average spacing ranges from 0.3 to sometimes above a half a mile, which is too far to justify dropping the existing local routes from the corridors.
But what about those remaining local lines? What about the other (many) routes in our transit system? What if we could speed up their service without spending much from transit’s limited capital and operating budget? What if doing so could allow a bus to run more frequently without adding any operating costs (or new buses)?
Most bus routes today (outside of the CBDs with shorter block faces) have stops every long block (or every second short block) – or every 1/8th mile (660′). I propose we change this to a quarter mile (1,320′) standard. The former is certainly on the low end of most urban municipality standards, while the quarter mile is definitely on the high end (excluding those crazy Europeans, of course):
The Positives
While no bus in our system stops every block, every hour of the day (indeed, many low frequency, low rider routes go many blocks without making a stop), we can almost guarantee that consolidating stops will lead to faster run times for the popular ones. Research supports this – with studies showing an increase in bus trip speeds between 5.7 and 10 percent. The same studies show mixed reviews on impact to bus schedule reliability, with one showing no impact and another having a positive impact. This could mean 3-5 minutes saved off a typical transit trip (depending on length, etc).
In addition to improved service by trip time, Metro Transit could use the operational efficiency to improve route frequency. Rough hypothetical example: 10 buses (and drivers, and diesel fuel, etc) are required to run a route every 15 minutes, allowing them to run faster gives options. A 10% improvement in speed could allow Metro Transit to use just 9 buses for the same level of service, OR keep 10 buses operating at better headways – 5 buses an hour is better than 4, right? Thus, transit riders could see improved trip times AND reduced average wait times at stops without Metro Transit spending a dollar more.
Any operational gains that aren’t pumped back into the system through service improvements could instead go into stop amenities. Metro Transit maintains over 10,000 stops, yet as the Star Tribune excellently points out, they struggle to provide basic shelter at hundreds of shelters, many of them in urban locations. Heat, shelters, and better signage all cost money, and this provides a source.
The Drawbacks
Obviously, there’s a downside. Consolidating stops means serving fewer people (and businesses) within the same walk shed:
With stops every other block (green dots) and an assumed willing walk distance of 1/4 mile, we certainly leave some areas out that used to be “in” (blue). Furthermore, we add walk distance to a fairly significant chunk of land that was closer to an “every block” stop:
About 14% of the land area that used to be served now has a 2.5 to 5 minute longer walk, and a whopping 36% of the transit’s captive area must walk between 0 and 1/8 mile further (the red line being the most affected) – about 0 to 2.5 minutes.
However, we know that land use patterns have consolidated along the transit’s route over time. More businesses are at the legacy streetcar stops than several blocks away. Apartments, duplexes, row homes, etc are more common along the arterials than several blocks away. A “no GIS database” rough estimate would say 80-90% of riders would see marginal changes in walk distance.
Additionally (and despite the walk/wait time cost discussed earlier), the studies linked above show that ridership held steady after consolidation (Portland) and actually grew (Seattle) despite other system routes seeing ridership decline. I would wager a bet that faster and potentially more reliable service could 1) convince those who must walk further to continue to ride, 2) draw new riders from within the walk shed, and 3) make new (car-lite/free) development near stations more attractive.
Conclusion
Stop consolidation could be an extremely inexpensive (they’d still have to go remove those signs, print new schedules, market the changes, etc) means of improving service on every urban bus route. When done in combination with other (slightly more expensive) improvements such as all-door boarding (recently highlighted by Bill), better stop placement, signal preemption (or at least holding a green), we could see much needed improvements to bus speeds for very little cash.
Yes, there are other factors to consider, most important among them being those who an extra 3 minutes walking might truly be a challenge (disabled, elderly, etc). As boomers age this certainly becomes more of a challenge. We should evaluate the capability of Dial-a-Ride services to adequately serve these individuals, perhaps on a pilot route (and I define “adequately” as “equal to or better than current trip times”).
But if you believe the system-wide benefits are too great to ignore, share this with any high-ranking Metro Transit employee or Met Council councilmember you happen to rub elbows with at fancy dinner parties (I jest). OR, you can get serious and make comments on Metro Transit’s Transportation Policy Plan – there’s a hearing on September 17th. Additionally, Metro Transit’s Service Improvement Plan is in development and should be ready for comments in November. Share your thoughts, and use this post as a reference!
I agree, and I’ve been complaining about this for a while — it’s one reason I’ve gone from a bus/bike combination to giving up on the bus entirely.
I think it comes down to a question of how one views buses. In American society, they’re considered a last-ditch resort for the very most unlucky, and consequently, they are designed for lowest-common-denominator service: they have to serve the very most needy customer at the expense of everyone else. On the other hand, of course, they’re supposed to compete with cars, for which there is no such expectation.
Such a comment kills my bleeding-heart liberal street cred (and I’ve made similar comments about bike infrastructure built for 8-year-olds). But I think we need to ask ourselves if we’re serious about this transit thing, and by extension this density thing. And if we are, we need to make the bus the most appealing to the most people. That means stopping less. They just desperately need to be faster.
As Alex points out, making them faster actually creates savings. Without doing the math, the first thing I’d do with that money is invest it in transportation programs for folks who truly can’t walk 1/8th a mile.
1. I agree wholeheartedly. This has long been a passion of mine. Metro Transit has way too many bus stops (over 10,000 in the system) and as you note, we are unable to provide basic stop amenities at the vast majority of them. Even very busy bus stops go without shelters, schedules, etc., despite qualifying for them.
There is simply no way Metro Transit will ever be able to afford more than a basic pole in the ground with so many stops to maintain. Reducing that number by 50% would change things significantly. We might actually have a shot at adding some nice heated shelters at busier stops, possibly even with real-time boarding info and off-board fare collection at the very busiest. Another frequent complaint people have about bus stop maintenance/safety is snow removal. With half the number of stops in the system, we might actually be able to afford snow removal at busier stops outside of downtown and other existing maintenance districts. Would improvements to shelters and winter maintenance be enough to compensate for making *some* people walk *a little* further? I strongly suggest that it is.
2. As you stated, Arterial BRT overlays are never going to happen for many routes, most of them in fact. Will they be forever doomed to 1/8-mile stop spacing with a basic pole in the ground? I certainly hope not. 1/4-mile bus stop spacing is a great idea, and would bring many more riders and revenue to the system. I really like the idea of doing it on a pilot route, to get feedback from riders and weigh that against complaints from people who have to walk a little further. I suspect the results would be overwhelmingly positive. Route 4 is actually an excellent candidate for this pilot, because it’s fairly frequent and doesn’t overlap much with other routes (aside from downtown).
This is kind of obscure (for non-mathematicians), but the circle vs diamond walk-shed “radius” is a great example of something called a “norm.” It is basically just a way to generalize the concept of size to a few basic essential properties. People (mathematicians mostly) study what we can know just based on those few properties that can then be used for any norm.
See the picture on the right here:
http://en.wikipedia.org/wiki/Norm_(mathematics)#Properties
The diamond radius is an example of the 1-norm, where you measure size by summing distance N/S + distance E/W
It was neat to see something from my math education show up like this.
Thanks for the comment. You’re right, of course, and actually there has been research done to see what walk shed methodology works best: http://www.its.berkeley.edu/publications/UCB/2011/VWP/UCB-ITS-VWP-2011-5.pdf
They find that it doesn’t really matter if you pick one catchment area over another (the gains in half mile vs quarter mile are marginal), and that the shape is also less important (ie doing a full-on network path bs just assuming a general shape). Planners should just use something simple that works with the data they have (a norm) instead of spending tons of time modeling something else for marginal gains in accuracy.
I used to ride the #3 bus between the U of MN and Saint Paul’s North End. There are THREE stops at the Como Student Housing cooperative all within .2 miles of each other. (https://twitter.com/BillLindeke/status/510099009638522880)
It’s so ridiculous.
I challenge anyone out there to top that.
No problem. Hilariously enough, it’s the 3 again (and the former 16 and 50, but that’s behind us now). Westbound on 4th St. through downtown Minneapolis, it stops at 2nd AND Marquette AND Nicollet AND Hennepin. 3 stops in .16 miles from 2nd to Nicollet, 4 stops in .25 miles from 2nd to Hennepin. Interestingly enough, the eastbound route skips the Marquette stop, and moves the 2nd Ave stop to halfway between 2nd and 3rd, so going back wasn’t so bad.
Joey, I actually think those stops (also used by the 7, 14 and the old route 94) are necessary, given how much they’re used to make connections downtown.
That one’s pretty bad, but there are many situations like that on east-west routes in Minneapolis. As Aaron Isaacs has commented here on streets.mn, apparently most bus stops were still unsigned until the early 90s. *The 1990s.* While that explains the color scheme on the bus stop signs, it still seems completely insane that our bus network was that informal during my lifetime. After bus stop signs went in at every corner, it wasn’t until stop spacing audits years later that removed most of the bus stops that were spaced an absurd 1/16th-mile apart on those east west routes. As they say, every bus stop has a constituency, so not all of those 1/16th-mile stops have been eradicated yet.
Hennepin Avenue downtown, in the northbound direction, is still really bad. There’s at least 2-3 stops I would eliminate between Loring Park and the river. Same goes for Nicollet Mall, which we have the perfect opportunity to address with the coming reconstruction project.
I’ve got one. A few years ago I used to take the 61 going past highway 280. Riding the 61B, which does a loop up Industrial Blvd was always excruciating. I think there’s like 20 stops in the whole 2-mile loop.
I can’t remember how it actually went, but I think the eastbound bus might even pass some stops on Hennepin twice because the loop ends west of where it begins.
What drives me nuts in the 6 southbound. Stops on the south side of Franklin and north side of 22nd. Also the south side of 28th, uptown transit station, Lagoon, and 31st. 28th and Lagoon should go, as should 22nd.
Let’s make some maps, people. Let’s have a contest for the worst bus stop spacing in the Twin Cities!
The topic is central to any discussion of transit efficiency and effectiveness. We do have a practical example for consideration. The former No. 16 bus had 69 potential stops for boarding/disboarding, and took 45 minutes to travel the 10.45 mile route, on average, according to official figures. The former No. 50 bus had 31 or 32 potential passenger stops (depending on direction) and took 40 minutes on the same route. These figures correspond well to my personal experience on those lines. The actual temporal difference was slight; perhaps the No. 50 benefited psychologically in terms of perception of a smoother experience. Need I compare, once again, the Green Line with its difficulty meeting the 48 minute schedule over its 11 mile route with only 23 stations?
Wow. That is amazing that eliminating 38 stops (55%) saved only 5 minutes. But I guess 10 minutes a day for the 100+ trips a day would really add up.
I think you may be juicing the schedule times. My personal experience (fairly limited and back in college) also says time was a bit longer than that. Mike Hicks’ recent post highlights the schedule time differences between DT to DT transit options (before the 16/50 were changed): https://streets.mn/2014/05/30/green-line-travel-time-update/
At 4 PM, an eastbound 16 from Nicollet to Cedar/5th (Central) was scheduled to take 64 minutes. The 50 was scheduled to do it in 54. That’s a 15% reduction in trip time by removing ~50% of the stops.
Of course, comparing a local vs express situation to consolidating the same number of passengers to fewer stops isn’t a 1:1 comparison. A 6 on Hennepin stopping half as often needs longer dwell times per stop (unless we buy better buses and implement off-board payment, etc)…
It must be a nice world you live in, where the end-to-end travel time on the 16 was a reliable 45 minutes. My experience was that 45 minutes would hardly get you to Raymond, if you left downtown Minneapolis at rush hour. I rode it frequently during 2008-2011, and it was always a lurching, crowded, achingly slow ride, except for very late at night. The 50 was far better, the Green Line is on a completely different level.
about an hour most times, and probably 80 minutes at any time close to rush hours
Yup. No way the 50 was only 5 minutes faster than the 16. That’s not even true according to the scheduled times, let alone the painful reality. And don’t forget that the 50 was noticeably safer and more peaceful than the 16. Weekly boardings on the 16 were about 3x the number of boardings on the 50 each week, but the number of incidents reported to Metro Transit Police on the 16 was over 10x the number reported on the 50.
I’m on board with this (pun intended). Also, it’s conceivable that reducing the number of stops and marginally increasing the the frequency of routes globally in the network could have some significant reductions in transfer wait times. In my experience, transfers and the corresponding wait time deter a lot of people from taking the bus. It’s especially frustrating when buses get off schedule and you miss your connecting bus and then have to wait 15+ minutes for the next bus.
This is a very thoughtful piece (although I’ve never felt angry when people pulled the cord at each corner).
One very important part of the puzzle you did not consider: disabled riders. We constitute 1/4 of the population, and I would argue that we are at least worthy of a mention.
There is an option that is door-to-door for those of us who need the service, but it costs approximately five times as much as the regular bus costs us, and it’s not nearly as efficient, environmentally.
Especially during the winter when walks are shoveled erratically at best and curbs cuts are never dug out, asking us to go farther to catch the bus is actually a rather significant hardship. Even during warm months, some of us can walk, crutch, or roll quite well for a block or so but after that it becomes exhausting.
Haddayr – Agreed! And if 25% of the “GenPop” is disabled I’m guessing amongst public transport “frequent riders” the % is even greater. I’d hate to encourage MORE isolation for people (especially in winter). The door-to-door service option is not only expensive, isn’t it also fairly unreliable in terms of when they actually show up? If making bus stops farther apart pushes MORE people to MetroMobility, isn’t it likely that MetroMo will be even more stretched? Then there’s safety – Maybe there’s some middle ground here but please consider the safety aspects too when thinking about taking out stops on busy streets.
and then add in all the people who are temporarily disabled, or have little kids with them and/or heavy objects. Not to mention the chronically impatient who get tired of waiting for a bus and just start walking – I’ve had the bus stop for me mid-block when I miscalculated how far I could get before the next bus should arrive. It’s awesome.
I generally get off the bus (the 14) anytime it stops within a few blocks before my house, but when I had a toddler with me, or had gone grocery shopping? I want the stop in front of my house.
Most of the really slow routes have an express option during at least part of the day. Let the rest of us have the local all-stop routes!
The figures I gave for the former No. 16 and No. 50 are (were) the official schedule figures. No surprise if the schedule may have gotten a little ragged sometimes during rush hour. But I can truthfully tell you from personal experience and at various times of the day including afternoon peak times, that the No. 50 wasn’t much faster than the No. 16. I think a greater difference might show up with a lots of stops/fewer stops comparison on a route that has less street traffic or runs in the suburbs. David W.’s remarks highlight the psychological factor, with the No. 50 having been the more pleasant ride. By the way, waiting to transfer when schedules aren’t met is worst on a windy day in the winter! And one thing that’s been overlooked regarding the reduction of University Avenue bus service when the Green Line was introduced is the diminished service for those with disabilities. The State Council on Disabilities never responded to my inquiries, although they received the certified letters I sent when regular mail went unreplied..
Again, I think you’re very much juicing the numbers there. Mike’s post was based on the scheduled trip times throughout the day. This report: https://metrotransit.org/Data/Sites/1/media/pdfs/central/report/ExistingConditionsReport.pdf#page=26 shows travel times between 46 and 50 minutes for the 50, 60 and 63 minutes for the 16.
But, even if your travel times are right, that’s still an 11% reduction in travel time – right in line with the research noted in the post.
I am fully supportive of this. Spacing bus stops every other block will not even require people to walk much more. Currently, if I live mid-block on a parallel side street, I could go either left or right and reach one of two bus stops equidistant away. Spacing every other block means that I only have one choice, but it is still essentially the same distance. The further away I am from mid-block, the more my walk to the bus stop would increase or decrease. But, in this scenario the most it would be is an extra half-block.
I just have one word: bikeshed. Put quality bike racks at the very least all the stops of major intersecting routes. Supplement that with better infrastructure connecting to those stops/stations.
I support quarter-mile bus spacing!
Which means that I oppose Metro Transit’s plan to eliminate the Saddest Bus Stop in the World (the southbound 4 at Groveland). The last I heard, they’re planning to get rid of it, without replacing it, as part of the Hennepin-Lyndale redesign. That would mean a half-mile gap between bus stops from Vineland to Franklin. That would be really bad for me and all the other apartment dwellers who live by Hennepin, north of Franklin.