King County Metro’s 2017 Annual System Evaluation is now publicly available.
One coarse performance measure in the appendices (pages 45-60) is weekday ridership on each route, comparing fall 2016 ridership to fall 2015 ridership. STB covered the March 2016 and September 2016 service changes that occurred during this period, as well as the March 2017 and September 2017 service changes that are not reflected in this evaluation.
Metro has roughly 180 routes, most of which saw daily weekday ridership shifts of 200 riders or less. The 50 routes listed below are the ones that had more dramatic shifts.
Some of the service improvements were funded by the Seattle Transportation Benefit District, approved by voters in November 2014. The sales tax and car tab for the TBD expire after 2020.
Martin recently laid out the cumulative investments the City of Seattle has made in additional bus service since 2014. Not all of these were made in 2016, and so would not affect the charts above.
In the future, it might be worth pulling out new and cancelled routes into separate lists. Their change to or from zero feels like it distorts the overall change of ridership on existing routes.
It might also help illustrate more clearly if the new routes are higher ridership than the ones they replaced.
Similarly, routes that take over for other routes have misleadingly large changes. The 372 gained 2900, but the 68 lost 2200. Likewise most of the gain in the 45 and 67 corresponds to loss from the 48 and 66.
The 62 is another example. It would’ve pulled from the cancelled 16 and maybe some from the 66.
My thoughts exactly. The 242 has been gone for at least a year, maybe more. 373 runs basically the same route, so it’s numbers are not surprising.
The 242 and and 373 are not the same route. They’re a mile apart, and the 242 went to Microsoft, while the 373 goes to the u district.
Route 26. Okay, so rebranding it from 26 to 26X justifies that we cancelled the 26 and created a new 26X?
Let’s do the math.
26, falls from 2800 to 0, drop of 2800, Fall 2015 to 2016.
26EX, rises from 700 to 2900, rise of 2200, Fall 2015 to 2016
The route, taken as a whole, actually falls from 3500 to 2900, a net drop in 600, Fall 2015 to 2016.
This change should be much lower on the list. No new route was created, a single route was simply modified. These even have the SAME ROUTE NUMBER!
Similar comment on 28.
I’m guessing that much of ridership on the old 26 (and 28) was the Fremont to downtown via Dexter section, which is now moved to the 62.
I think in order to really do a meaningful comparison, you almost need to aggregate by street, rather than by route. Even then, it doesn’t distinguish between people who travel along a particular street because they want to vs. because metro is forcing them to.
Yeah, it would be interesting to know if the new 26 and 28 have increased or decreased in ridership at the stops they actually still serve, but we don’t know. (Anecdotal evidence from a regular 28 rider–it seems about the same.
Looks like the 10 reroute to the Link station wasn’t too popular.
A decent part of the Route 10 service area is within walking distance of Link at Capitol Hill. Lots of former Route 10 riders simply walk to Link.
Any way we can assess what a route would have handled had it been kept? When Route 43 was canceled, it left a long corridor up 23rd Avenue without the direct connection its residents would have had to the hospital, the Broadway District, and Capitol Hill Station.
In addition, it would’ve been a permanently existing bus bridge between UW and Capital Hill. What would be first step in getting it back?
Mark Dublin
The 43 still technically exists, it’s just a skeleton of its former self with peak-only half-hourly service plus random other trips to/from the 44.
To U District
To Capitol Hill (not all go to Downtown)
Bringing it back would require a lot of advocacy + money. I’d rather see that money go to improving frequency & reliability on the 8, and to the connection between the 48 and 8 at 23rd & John.
The 49 also dropped, which doesn’t make sense until you realize that some riders switched to Link. It’s impossible to say how many 43 riders went to Link, much less how many preferred to, since there are so many other factors. How many switched to the 10 and are glad it’s frequent evenings? How may would still ride the 43 if it ran full time? `While it was a useful route, it was redundant along its whole length like the 4S is (with the 3 and 48). So it has to compete directly with other areas that would have no service or less service if the 43 were increased. The 43’s hours went into the 8, 11, and 48.
With the 43 still having a remnant of service, do you think it would have been better for the 10 to continue its old route that served more of 15th, plus Seattle U?
There are arguments both ways on whether the 10 should have moved, but the 43 is not an argument for not moving because it barely exists. Pine Street to downtown has the 11 and 49 which each run every 15-30 minutes all day. John/Olive to downtown would have only the 45-minute 47 that doesn’t run at all evenings and doesn’t go east of Summit, and a few 43 runs a day (and fewer than it may appear because the off-peak runs turn at Broadway).
Even with the investments going into other routes, frequency of service in the corridor where the 43 and 48 overlap went down noticeably, especially on evenings and weekends. The transfer between Link and the 48 at Montlake Blvd is now the only connection from the 24th Ave E corridor to Capitol Hill and downtown, and the 30 minute wait for the 48 at the Montlake Blvd stop with no surrounding services in the cold at midnight isn’t a great experience.
The 48 doesn’t drop to 30-minute frequency until around 10 or 11 PM, by which point, the old 43 and 48 were also every half hour at best (maybe, worse). I also recall, Metro had an annoying scheduling quirk, where the 43 and 48 ran back-to-back, so even though, on paper, Montlake/23rd got 4 buses per hour, it was still, effectively, 30-minute frequency. FWIW, the new 48 does maintain 10/15-minute frequency later into the evening than the old 48 did.
As also mentioned, there are very few segments of 43 route that actually require the 48 to get to Capitol Hill. The one section that is too far north to walk to the 12 and too far south to walk to Link consists of low-density homes worth well over $1 million, and doesn’t have much ridership potential, especially late at night.
Finally, I’d like to mention that the period around midnight actually still does have some route 43 trips going from the U-district to Capitol Hill, as route 43 is the only way for route 44 trolley buses to get back to the base.
Metro’s 2025 plan restores a “43 on ADD” as David Lawson called it. It’s like the 43 coming from downtown but turns on 19th instead of 23rd, then east on Aloha to 24th, then northwest on Boyer/Fuhrman to the University Bridge and U-District station.
The report shows change in ridership as well as change in platforn service hours in the same tables. Oddly, Metro does show the obvious next column of ridership pernplatform hour, which would describe a route’s productivity.
Looking at only total ridership changes is not very informative when most of the biggest changes are a result of major service changes. Productivity is. Even though productivity is highlighted in the report, the change from a prior year is not.
Any calculation ever done about passengers who switched from bus service to LINK because the train is so much faster than previous bus ride? Nothing under-performing. Major passenger improvement.
Also, how much time does Metro give a route change for people to get used to the change, before the route is pulled? Sometimes it can take several years for the public even to notice a new route, let along make it a habit.
Another thing I can see happening is that the more LINK station opens, the more bus ridership it brings to the whole rest of the system. One more in a list of problems as long as a freight train with the system’s present fight to keep seven agencies revenue-separate.
Ridership system wide would improve if transit could take a palm-tree sized leaf from our most expensive and unfair competition, and let me decide which-system’s passenger I want to be when my emergency brake goes off.
Mark
It’s a shame to see the 31 on the list of reduced ridership lines since this is my family’s lifeline to our jobs from Magnolia to UW. It’s a chicken before the egg argument though. The line does not go late at all and East-West bus routes north of the cut are not at all prioritized and stuck behind suburban drivers getting to I-5 or 520.. It takes my wife 90 minutes to go the 7 miles between UW and Magnolia in the evening, while driving takes 30. We often just bike it (also significantly quicker) but it’s just not possible everyday. I really hope the city starts prioritizing buses on more East-West routes north of downtown.
In my experience UW to/from Magnolia on a bike is always faster than a bus and on par, if not faster than, driving (no to spent searching for parking or in transit from parking lots).
I agree. Besides that the fact that not running the 31 leaves no way into or out of Magnolia, except through downtown/Seattle Center, it also cuts the frequency of Fremont->U-district trips in half, leaving only the 32, running every 30 minutes. At least with the latter case, bikeshare down the Burke-Gilman is an excellent alternative to the bus (and usually faster, even on weekends), but the bus is still useful when it’s night or raining, and it would be nice if a two-mile corridor between urban centers could run more often in the evenings than every 30 minutes.
The 31 and 32 have been discussed a lot here (and especially on Page 2). There is definitely room for improvement, in my opinion. The 32 carries a lot more riders, but it is also fairly redundant. It is strange that we have asked so many riders from Northeast Seattle to make a really bad transfer to Link (at the UW) but don’t want folks to make a much better transfer from the D. Making matters (much) worse, the 31 doesn’t allow for a good connection between Magnolia and Ballard. It looks good on paper, but the stops on Nickerson are a really long way from the bridge.
There are a lot of ways of fixing this, but the simplest would be just make one route (call it the 31) but alter it. Have it serve Dravus, instead of Emerson. Interbay has way more people (now) and folks headed to Firsherman’s Terminal can take the 33 and walk a couple blocks. Run it as often as you run the combined 31/32, which means way more service to Magnolia (and Interbay). Those in Lower Queen would have to transfer, but with the D being frequent, and with other ways of getting there, this seems like a small price to pay. The functionality of the 31 increases dramatically, not only because it runs a lot more often, but works well as a connection from Magnolia to Ballard.
These are great options and I agree with you about the Emerson segment. I can count on one hand how many times in 5 years of taking the 31 I’ve seen someone get off at Fisherman’s terminal. Dravus is unfortunately too far away for most people in Magnolia (it’s a 30 minute walk for us and we actually live in Interbay) but connecting to the rest of Magnolia via Dravus makes a lot more sense. Taking the overpass towards Fisherman’s terminal causes huge delays during rush hour. I just really hope they don’t write off this line. Not everyone in Magnolia/Interbay lives in a huge single family home with multiple cars. It was actually the only neighborhood in the city where I was able to find an affordable place to live.
While this idea would definitely help getting from Magnolia to Fremont, I don’t think it would help Magnolia->Ballard much. You still have to transfer to the D-line, and you can already do that by taking the 24 or 33 to the D-line.
Magnolia->Fremont would be a significant improvement, but you’d have to weight the against the effect it has on Fremont->Seattle Center. Which makes me a bit weary because the ridership stats show more people making the latter trip today than the former. That said, if I were traveling between Fremont and Seattle Center, under your network, I don’t think I would actually make the D-line transfer in Interbay – I would probably just take the 62 to Dexter/Mercer and walk. The D-line just isn’t frequent or reliable enough to be worth transferring to, just to avoid a 10-minute walk. Especially with the wait being on an ugly, busy street, and with the walk under Aurora no longer the skinny sidewalk immediately abutting speeding cars that it was 10 years ago.
This gets quite complicated, but you could also do something where, instead of taking the same route the 33 takes, it instead replaces service currently provided by one of the loops on the 24. Making the 24 less loopy could result in some service hours being deployed into frequency.
I wonder how much of the 31’s decline is a result of problems with the 31, versus competition from other routes. Metro vastly increased the number of buses going down Steven’s Way due to the Light Rail changes, and a lot of students take any bus from east campus or U-Village to west campus, where the new dorms are. The 65/67 and 372 are now much more frequent than they used to be, so it could just be they’re soaking up more of the intra-U-district riders.
David, you’ve got a point about the 43 that pretty much closes the discussion. Enough time has gone for the people I want to serve make any effort at all to get their service back. When they decide they need it, the wire’s already there.
Curious. Is the 48 going to be wired south of Thomas Street? As for the 8, I wonder of its only remedy isn’t at least a queue-jump each end of its bridge over the freeway. Not a lot of hardware. Literally nothing to lose by setting that up.
Surprised aerial tramway from Capitol Hill Station hasn’t surfaced yet today. Though when Capitol Hill looks like the places where they do this in a non-mountain area, there’ll be conditions that mandate it: Tax-paying passengers living in Tiny houses all the way over the Hill to the Lake.
Mark
“Is the 48 going to be wired south of Thomas Street?”
Yes, that’s part of RapidRide 48. Which may be the 48/7 depending on which plan prevails south of Mt Baker. The 23rd Avenue road project, which is going in three phases, includes the infrastructure to string the wire on. Although SDOT may decide to buy battery buses rather than trolley buses; they said they’ll see what the state of technology is when they’re ready to buy them.
Metro and SDOT are working on improvements to the 8, though the biggest change has been delayed. Completion is expected in 2018.
This is not a terribly impressive report as far as the actual narrative goes. Additionally, I find the analytics sorely lacking, as many other commenters have previously pointed out. (It reminds me of a company’s annual report after a down year in which they throw a lot of fluff into the opening pages.) I certainly hope this was done in-house and not produced by an outside consulant.
I found this the most interesting part of the narrative section of the “Executive Summary”. Sadly, there was no further discussion to follow….
>>Seattle Investments
Metro and Seattle work together to plan and implement additional service funded by Seattle’s voter-approved November 2014 Proposition 1. In accordance with the contract between Metro and Seattle, Metro is in the process of assuming funding for some of Seattle’s investments. As Metro assumes funding for service, Seattle may add more service hours at its discretion, in
coordination with Metro.<<
I hope we hear about the details shortly.
So, as someone who travels in the service area of the former 71, 72, and 73, I’ve been trying very very hard to be positive about the change while constantly being frustrated by the quality of the transfer at UW station. I think that this report bears out the quality of that transfer for those three routes – the net loss from the old routes doesn’t equal the increase from the 372, 373 or 71.
I get very frustrated about this because that area is rapidly graying, and got used to depending on transit over the last two decades. While it’s okay on paper to have a half mile transfer, in reality, this is really burdensome to mobility challenged people.
The furthest the transfer distance gets is to the 372/75 on Stevens Way, and that’s 0.3 miles according to Google, not 1/2 mile. It’s relatively flat, and both an escalator and elevator are available to get you up the steepest part. For the 73, which stops on Pacific, the walking distance is much less than that.
I have also confirmed experimentally that it is possible to get from the 372 stop at Stevens Way onto the actual train (not the station entrance) in as little as about 5 minutes, although the 5-minute figure does require some serious hustling (plus an Orca card to avoid messing with the TVM’s).
Boy, I hear you, Asdf. I’ve done all of the measurements that you describe invludeing the measurement of time (and thanks for calling me out on my .2 mile exaggeration – it was just that). I really wanted to hear that the transfer wasn’t depressing the quantity of people using transit in NE Seattle.
But the point was less about the nature of the transfer, and more about the reality that the buses serving that area of town that got rerouted to UW station all saw downturns in boarding.
We have to do the best we can with limited resources, and the UW wasn’t the best partner when it came to station access. But my main point is that we could have done way, way better.
-wb
(And because I’m like most STB people and love rehashing the details, I’m includin a Page 2 to this response with practical details).
it’s a *very* illegible transfer. Last week at Daly City Bart, I encountered a sign that showed me a list of buses that berthed to my left and a list that berthed to my right. At UW station, I feel like I’m constantly directing people to where they can find their bus.
And, I don’t know about you, but I feel that I’m in the peak of my years, so it’s easy for me to hustle. That’s not the case for everyone who uses that transfer, of course. I’m surprised at the number of times I’m traveling with luggage or something that makes me realize what it’s like to be less than able.
I am still ticked that I have to walk to and from Stevens Way from the # 372 and it takes more then the 5 minutes you mention. It is also up an incline from the station to Stevens Way. Add that walking time my trip to and from downtown is not any faster then what the # 72 provided.
I really miss the 71,72,73 and looking forward to Roosevelt station opening and an end to this ridiculous forced transfer at husky stadium. It’s added about 30 minutes a day to my commute, but I just recently realized the impact the reroute seems to be having on the Ave businesses. The Ave has struggling for years with homeless presence, high end competition from u village and the ongoing shift from brick and mortar retail to amazon. But last Saturday I visited the u bookstore and some other u district businesses for the first time since my old commute. I’m probably not the only one who used to frquently stop at Ave businesses.
Don’tutry to defend the indefensible. Don’t flack for Metro. They should not have done the re-org until U-District and Roosevelt opened with — we can hope, though it doesn’t look too promising — better transfers.
The old 71/72/73 got most of their ridership, not from north Seattle, but by people going from the Ave to downtown. Those riders are mostly walking to Link, so change or no change, they’d be “lost” from a Metro perspective, anyway. Many of the remaining U-district->downtown riders are taking the 70, which again, superficially looks like a loss, if you only look at the 71/72/73, while ignoring the fact that ridership on the 70 went up nearly 50%.
Yes, the walk up the hill is closer to 10 minutes than 5, but if the bus isn’t coming for another 15 minutes anyway, what difference does it, anyway? And, the time you have to wait for a bus is much less than it was under the old network, and would have been had the 71/72/73 continued going downtown.
And Link is much more reliable. 6 minutes to downtown. The 71/72/73X from Campus Parkway was unpredictably 20-45 minutes depending on traffic. I used to reverse-commute on it every day, and two or three times a week it was ten minutes late and/or ten minutes longer than scheduled. And it got caught in a 20-minute traffic jam on I-5 or Stewart Street or the DSTT at least once or twice a month. All that is gone with Link.
So much about Link is designed for young people with perfect knowledge of where to go. Consider that there were only a few hidden signs that indicated which direction trains were going prior to 2016, or that ST is spending tens of billions on extensions and stations but won’t address that many existing stations only have up escalators, or that there is an obvious shortage of TVMs st the Airport Station. That’s on top of bad or no signage for connecting buses and shortage of pick-up zones in Seattle. Thank God for ADA or it would probably be even worse!
I’m really saddened when I see a culture that board members and senior staff won’t try to be more vocal about solving basic user problems — and instead either ignore or make excuses for many obvious system problems. Given about how many leaders are over 50, it makes me wonder if they’ve ever regularly ridden on the system that they are supppsed to oversee.
+1 – young(er) person here and I find the design disappointing, especially for a newly-built system.
I can navigate easier on transit systems in cities I don’t even live in anymore, or even in places I’m visiting as a tourist for the first time. Heck, in Oslo I managed to figure out a service change poster written entirely in Norwegian without asking for help. And yet sometimes I still mess up on Sound Transit.
Signage is still less clear than it should be. Finding the right exit is not obvious.
ORCA readers are located haphazardly and there are still not enough of them.
Nearly constant loud announcements on relatively unimportant topics in the DSTT.
The insistence on using “station” in every mention of a station.
Commenters have raised these issues, and many more, often on this forum. These are not insurmountable problems to fix.
Most of the former 71/72/73 riders were between downtown and the U-District. The tails were always low. A lot of the riders from the U-District have shifted to the 70 (which I would not have predicted, and I avoid), and others have dispersed to a variety of routes to UW Station: not just the 71. 72, and 73, but increased service on the 44, 45, 48, 65/75, 67, and 271. Those coming from the 71/72/73 tails also have the opportunity to transfer to the 70; I don’t know if any of them do. And there’s more peak-hour and shoulder-hour runs on the 76 et al so that may have absorbed some people.
In my opinion it’s this diversity of new routes between downtown and the u district that makes the situation so bad. The bus tunnel used to provide a single location to catch efficient trips between downtown and the ave. And the ave has lots of amenities that make a 10-20 minute transfer pleasant. But now there is no longer a single downtown hub for frequent/fast trips to the north… unless you want to deal with husky stadium.
From my office I get to gamble every day. which is more efficient? Walk 6 blocks for a chance to catch a 76/316, or 2 blocks for a chance to catch the 64/70, or 3 blocks and down some stairs for a chance to catch the link/74. Combined and on paper their headway’s are at least 2x the old 70s, but The lack of a common origin and surface routing make them much less useful.
Link alone has the frequency of the 71/72/73. The others are not intended to provide extra frequency but to serve different transit markets; they just happen to intersect various parts of the U-District. Pacific Street is so saturated with buses it’s like 3rd Avenue: never wait more than a few minutes for an intra-downtown trip. And four routes go to 65th if that’s of interest to you, although they pair up differently at different points along the way. Link is heavily better if you reverse-commute and don’t have the express lanes or peak-express routes. And the situation is much worse for those who have only one 15-minute or 30-minute route like our friends on the 74′ s tail. Pacific Street is also a shorter walk than Stevens Way.
I also have about six ways to get to work, all if which converge on two routes but there are several ways to get to them, including some by Link. But I am just glad I’m not limited to one route like I’ve been in the past, especially one infrequent route. Recently I found a shorter way that does not involve Link so I,’ve been using that, with a built-in 15 minute flat walk exercise. When Northgate Link opens things will get better.
The transfer at UW station is a dog. Not well signed, long walks and over-engineered (although is there a ST station that isn’t over-engineered – maybe Stadium and Sodo).
Hopefully the situation can be improved when Northgate Link opens.
Also, as I discovered a few weeks ago, the transfer is unavailable during Husky football games. The streets were open and buses were everywhere (game-specific shuttles), but the regular routes aren’t serving the stop next to the station. Signs there tell you to walk a half mile to board the routes that normally stop there.
Don’t reorganize your whole bus network around this transfer if you’re going to take it away a half dozen times a year.
And the Pacific Street shuttle comes every 7 minutes. It replaces at least five bus routes running a total of 18 times per hour off-peak. Metro saturated Pacific Street with bus runs to mitigate the gap between UW Station and the U-District, and they’re replaced by one 7-minute shuttle whenever there’s a game? And there’s no other way to get from most of northeast Seattle to downtown?
The shuttle is definitely a joke, and I doubt it runs anywhere near 7-minute frequency, getting stuck in traffic without dedicated bus lanes. Practically speaking, your options are to either walk or to avoid Husky Stadium completely and take the 70 downtown, or the 49 to Capitol Hill.
I avoid northeast Seattle when there’s a shuttle, and the once or twice I went there I took the 49;
Tlsgwm, what if it takes longer than the official length of time between reports to know what they’re telling you? Of course it’s right to issue the report. But before acting on the statistic, it’s a good idea to understand when you have enough information to make a decision based on what you’ve got.
Mark
It would be nice to have an extra column in the charts summarizing the changes that have occurred to the particular route. Something like:
A) Increase in service.
B) Decrease in service.
C) Change in routing.
D) Change in routes nearby (this would include Link).
This would be helpful for those who can’t remember the changes that occurred since last time. Looking at the chart, some of these are obviously altered routes. The first two are brand new (so they get both an ‘A’ and a ‘C’). I think the 67 is the same thing. I’m not sure about the RapidRide runs — did they get an increase in service? When did the C and D split up?
It would be interesting to see which buses simply had more riders, even though there was no major change to it. Theoretically, every bus (or train) can influence every other one, so this is would be a judgement call. But some areas might just be growing, or people may be choosing to ride certain buses over others, based on traffic, frequency (the E comes to mind).
It would also be nice to look at these changes on a map. Ideally, you would base it on Oran’s map (https://seattletransitmap.com/app/). I would start by simplifying that map — make all the bus routes the same color. Then change the color of routes that have changed. Have a different color (or a different type of line) for routes (or part of routes) that no longer exist, or are new, or have more service or less.
Another alternative is to be able to flip back and forth (which is what the old map had). I think with this information we would have a much better idea as to why these ridership changes are occurring.
+1 I think your ideas would be very helpful not only to us people out here trying to digest the data but to the planners themselves.
All of this is speculation without stop specific boarding numbers. The better way to look at the data would be to consolidate corridors – for instance UW to downtown or Fremont to downtown. Gather all routes that traverse these corridors to understand if the overall numbers have changed…if they end up being the same or declining then the reroutes were all for not. I expect that there is an overall increase in transit use though…Finally, taking those numbers and then breaking them down by corridor would really help to understand the impact that the change shave had.
IIRC, wasn’t one of the rationales behind splitting the old 48 into the 45/48 to provide greater levels of service on the now-48 southern portion, where presumed demand was greater. That doesn’t appear to be the case–the northern half has higher overall ridership, and is stronger on 5 out of 6 efficiency metrics. Did they really just get that wrong? Why, I wonder?
I think that’s rather obvious: Link opened. A certain number of 48 riders were riding Between UW and Mt Baker Station to board Link. Since Link has higher frequency (and 23rd Ave now crawls most of the day between Madison and Jackson, making buses even slower), using the bus to make that trip segment is pretty unattractive in comparison.
As the commenters have clearly shown, measuring rides per route is deeply flawed when the network was restructured. A more appropriate measure would be rides per platform hour comparing the two networks. Most of the routes on the increase and drop lists were changed significantly in March or September 2016. With U Link, many routes were changed. With SE Seattle, Routes 106 and 107 were revised, Route 124 improved, Route 9 reduced, and routes 7, 14, and 36 lost riders to the duplicative Route 106. Some interesting ones were not directly related to U Link or SE Seattle. The E Line and routes 40, and 44 got more trips. I bet they also improved in productivity. Route 40 was sped up on Westlake.