While trying to score the Madison Bus Rapid Transit (BRT) design by the ITDP rubric, one must assess whether Madison Street is a top ten corridor for transit. I didn’t have a good answer off the bat. Intuitively, it seemed other corridors clearly have busier buses, but they are also longer. Adjusting for length by calculating ridership per mile brings perhaps offers a fairer comparison.

For Madison Street, it might be a bit of a photo finish, but it would snag the tenth spot in my list. If a portion of Route 11 and Route 60 were included, it’d do even better. But the 11 mainly serves points east of the proposed BRT line and the 60 only uses Madison briefly as it heads north/south.

Here is how Seattle bus corridors stack up by my math:

Top Ten Corridors (by Bus Riders per Mile)

 CorridorRidership/mileLength (miles)RoutesTotal Weekday Ridership
1Pike/Pine/Broadway3800510,11,4915,100
2Aurora Avenue260013E,5,26,2829,500
3Rainier Avenue190097,916,300
415th Avenue NW/Elliot Avenue18008D14,000
5Taylor Avenue/Jefferson Street17006.63,411,700
6Queen Anne Avenue17003.72,138,900
745th Street/Market Street17005448,300
8Beacon Avenue16006.73610,700
9Eastlake Avenue15005707,700
10Madison Street12003123,600

Some of the corridors are tricky to calculate because it is debatable whether some bus routes count toward their total since they veer off in different directions and may only use the corridor briefly. This is particularly pronounced in a diagonal street like Madison. Note that the ridership totals are based on 2015 Metro Transit numbers (starting page 64 here), except where new 2016 restructure numbers have been released.

Compare the top ten corridors (by riders per mile) map to the Move Seattle’s map of the next seven RapidRide corridors that will bring Seatte’s total to ten with the current RapidRides E, D, and C.

MoveSeattleLevy-Overview-Transit
The Move Seattle levy listed these seven corridors for improvement. (Move Seattle)

In case there is any doubt to the top ten, below are the next ten corridors by riders per mile. All of the seven RapidRide+ corridors (plus the existing C) are in the top twenty. It seems not an issue that the RapidRide selections are bad, but that some routes seem to be getting overlooked. Beacon Avenue’s Route 36 scored very well in riders per mile and serves an area with a significant low income and transit dependent population. Queen Anne anchors two top ten corridors (in riders per mile anyway) with Queen Anne Avenue and Taylor Avenue. Those two routes both go through Downtown and First Hill before branching out in the Central District.

 CorridorRidership/mileLength (miles)RoutesTotal Weekday Ridership
11Denny Way12006.687,700
12California Avenue SW120011.5C, 12814,300
13NW 85th Street/University Way11006.5457,200
14E 23rd Avenue10006.1497,100
15NE 125th Street/I-51000104110,000
16Westlake Avenue/Leary Way90012.54011,400
17Lake City Way/SR-52290013372, 522, 309, 31213,100
18Delridge Way800121209,200
19Dexter/Stone Way/65th Street7009.5626,500
20Roosevelt Way6007674,400

Another factor to consider is how bus ridership will interplay with light rail and streetcar ridership. The Link is the ultimate transit corridor hovering around 65,000 daily riders on its 18.8-mile length for about 3,500 riders per mile. Extending the Link to Capitol Hill and to Husky Stadium was great for the city, but it did cannibalize some bus ridership and the bus restructure altered patterns further. Extending the Link to Northgate will change the ridership landscape even further. Nonetheless most transit users prefer the reliability of light rail to erratic bus routes.

If Westlake Avenue included a portion of the South Lake Union Streetcar ridership, then it’d certainly climb the list. Westlake Avenue is also hampered in securing riders per mile by how long its bus route is. The Route 40 meanders all the way to Northgate Transit Center…via 24th Ave NW. Not an arrow shot. And there’s the larger issue of how to group the bus routes together in the first place. Obviously, one could make a case for different groupings. Regardless, the top twenty routes enumerated here represent crucial corridors for Seattle transit users and ripe avenues for service upgrades.

2 COMMENTS

  1. That’s interesting. It’s pretty hard to define a “corridor” in terms of the bus routes that we have today, and Madison exemplifies this: in your (eminently reasonable!) assignment of routes to corridors, the primary route you assign to “Madison” doesn’t go to the Madison Valley and a route assigned to “Pike/Pine/Broadway” does. It’s not obvious (unless you’re holding a Metro map, of course) why Pike/Pine and Broadway are a single corridor instead of two that intersect (this is even more true of a #40-shaped corridor, a #45-shaped corridor, or a #62-shaped corridor… or even a a corridor shaped like the #36, where it’s easy to imagine a local network oriented around an all-Jackson route and an all-12th/Beacon route — particularly if you forget that the FHSC exists).

    Roosevelt BRT combines the #67 north of the bridge with the #70 south of it while promising to make the whole thing faster than the old #66. If it delivers that speed/reliability promise it should transcend the existing corridors; if it fails to deliver, not so much. Madison’s promise is sort of similar, except that where the Roosevelt project is trying to take the more popular corridor just outside downtown (Fairview intsead of Eastlake in and north of SLU) and make it faster, the Madison project is trying to take the faster corridor (Madison instead of Seneca in First Hill) and make it nice. So if Madison succeeds as a route (probably mostly a matter of speed and reliability, maybe with some accessibility thrown in) the corridor will appear more prominent relative to Pike/Pine and Seneca.

    In some places the corridors are much more straightforward. For Ashland BRT in Chicago, the corridor is just Ashland. It’s long and straight and runs crosstown. If it was Seattle the north part of Ashland would be served by a route from downtown via Milwaukee, Desplaines, and some random couplet; the south part by a route from River North via Chicago Ave and Ogden; and a giant gap between the two.

    • Stop-level data would be nice to have for corridor analysis. Even with stop-level data you might have to get creative in some scenarios — imagine a time before routes #8 and #48 existed, trying to piece together a picture of transit demand from then-current travel patterns crossing the route. Similarly, even stop-level data wouldn’t tell you whether a Fairview-Boren route, or the Capitol Hill-Beacon Hill concept that’s appeared in some planning documents, would be the next #8 or the next #25. It might have done a decent job of estimating initial performance of a route like the #62. And it would let you do (wild, crazy) stuff like defining and measuring corridors downtown where half the bus network overlaps.

Comments are closed.