For those of you who just can’t get enough data, I have a present: Metro’s 2010 route-level performance data in Excel format. For ease of reference, I’m including it all here along with the same data for 2009 and, courtesy of Mike Shekan, similar data from 2008. If you don’t have anything that can read Excel files, open it in Google Docs and export it in the format of your choice. Here’s a chart of of the top 20 routes by annualized ridership:
Per Metro’s policy, this performance data is annualized data based on the 3rd quarter of 2010, same as Metro’s 2009 data. This means the ridership in this chart is simply the number of riders for that quarter multiplied by four, rather than the true ridership; similarly with all other numbers in this spreadsheet. I have a couple of posts coming up where I’ll drill down into this data and compare it to prior years. Also, note that the 2008 data is formatted slightly differently, and the routes are broken out differently, so some care is required when comparing that data to 2009 and 2010. Enjoy!
Update: I’ve modified this chart to correctly include RapidRide A.
“This means the ridership in this chart is simply the number of riders for that quarter multiplied by four, rather than the true ridership”
Why though?
So you can compare data between the years. Routes often change (in terms of alignment or service levels) within the year, but are virtually always constant throughout that one quarter of each year. The value of this data is thus in comparisons between different routes in the same year, or the same routes in different years.
Is there no fare on the Center Park route (688)?
What does the designation “lowincome” mean on this spreadsheet? Route 24 in Central Magnolia is listed as lowincome, which makes no sense to me.
it means the route goes through a certain number of census tracts which are considered low income by Metro’s geographic value analysis
Right, but the 24 and the 33 have basically the same route until they get to Magnolia. The 24 serves 28th ave, 34th ave and Viewmont. The 33 runs down Government Way to Discovery Park.
I can’t for the life of me figure out where the low income areas are on 28th or 34th ave.
I just noticed that this data seems to lack anything for RapidRide A or its predecessor routes. I don’t know why this is and I will inquire with Metro on Monday.
It’s there Bruce. Look under the 671, which is the numerical designation for RR-A :)
Ah – I filtered that out of my graph, figuring it was a custom route. My bad. I’ll redo that graph later.
Thanks!
Strange how the number one bus line – by a long shot – is somewhat redundant to the current LINK line. People always complain about how LINK doesn’t got anywhere yet, but clearly it goes to an area where many people rely on transit.
The problem is it should have run grade-separated along Rainier, instead of along MLK at-grade. It makes it so much less accessible to the population centers of that corridor. Even as a visitor from another neighborhood, Columbia City would be much more attractive if the stop was actually in the neighborhood center. That 5-10 minute walk has a huge deterrant effect.
Trust me when I say that if light rail ran along Rainier (or closer to it), ridership would be significantly higher.
Right on the money. Rainier is too narrow for at-grade, and elevated would have ruined the mountain view, so it would have had to be tunneled. That would have paid off enormously. It would have reached established neighborhoods like Columbia City, would still have had lots of TOD opportunities, it could actually replace the 7 bus, and would be way faster. It could have continued on to Renton instead of going to the airport (which was always a dumb idea). Also, think about this: if it weren’t for the at-grade portion of MLK, Link could have been a driverless system like in Vancouver and operating costs would be really low. What a wasted opportunity.
Before anyone mentions it, driverless trains absolutely could have shared the DSTT with buses. They are remotely operated and have safety systems that detect other vehicles. They just can’t go through intersections.
@zefwagner,
1. Link ridecount (I cringe at the use of inaccurate term “ridership”), at somewhere above 6,000,000 for 2010, is far and away a higher ridecount route than the 7, or any other route in the state. I don’t know if putting it under Rainier would have made its ridecount higher, but the airport is clearly the second major destination on the route, behind downtown. I’ve ridden Link after 9 pm several times this month, and it is full of people with luggage.
2. There still would have been a local route on Rainier between stations. Maybe the 9 would have sufficed, but I doubt the one-seat riders would have accepted that.
3. If the train that made a sudden stop when a car dropped out of the sky (after skidding down a hill in Tukwila and tumbling over a fence) during the blizzard had been driverless, would it have stopped in time?
I wonder! I assume that the (car) driver called 911 immediately, and that the 911 dispatcher would have contacted Metro, and Metro would have remotely stopped any trains on that section of track. We can assume probably a 2 minute lag for all that human interaction to take place and the train to come to a stop.
At peak time, with 7 minute headways, that’s pretty risky. I’d have to assume that any such system, to be 100% “safe” in that situation, would need a forward-facing proximity sensor (radar, ultrasonic, or whatever) with a range at least as far as the stopping distance of the vehicle. That’s not prohibitively expensive, hardware costs would be about the same as a few weeks worth of labor costs on a manned vehicle. I don’t know if it is done in any current driverless systems, though.
But there’s a couple stretches of track where we might not have visual line-of-sight that far ahead, anyway, meaning that a human operator still would not be “safe”.
“instead of going to the airport (which was always a dumb idea)”
You don’t know what you’re talking about.
Yes, going to the airport is a bad idea. It is serving a very particular type of demand in a way that slows down the trip both for people going to the airport (because it goes through Rainier Valley) and for people going through to Federal Way eventually (because of the jog over to the airport). Most of the people going to the airport get on downtown, and they would be better off with an express bus. Most rail systems use shuttles from the nearest station rather than veering off the main path (which they would have to do because airports are always far from other destinations). It doesn’t matter if lots of people use Link to get to the airport, as it’s the only option. They would still be better off with high-quality express buses. Look, what I’m saying is conventional wisdom among transit planners–airport links usually are not worth the cost in travel time or capital costs.
To address Brent:
1. Yes, ridership is higher than the 7 (of course!), but the point is a lot of people are riding a parallel route. Rainier would have clearly generated dramatically higher ridership–I’m surprised that would be a controversial statement.
2. Yes, there would be a local route, but it probably wouldn’t go downtown since the whole point would be to get people to the nearest station. The reason it is politically difficult to truncate the 7 is because it is a long and bleak enough walk to MLK to act as a deterrent.
3. Vancouver has had a driverless system for decades and their safety record is stellar. I’m not sure what your comment refers to, but driverless systems have a much faster response time than a human operator. That’s why Vancouver can run trains 90 seconds apart at peak times.
Going to the airport wouldn’t have been such a waste of time for Federal Way commuters had Tukwila not ruined the segment between Boeing Access Road and the airport. Highway 99 is a logical route for that segment.
The Rainier Valley will ruin the DT-FW segment more than the airport will.
Besides, route 574 already has strong ridership between Federal Way and SeaTac.
50% of Link’s on-offs come from the Rainier Valley segment, at the cost of making it slower for South King. Tukwila’s intransigence with respect to the 99 alignment makes Link slower for no discernible benefit.
Had ST had the extra billion dollars necessary to tunnel through the RV and grade-separate the rest of the system, I like to think that they would have built to U-Link/North Link instead. What ST built or didn’t build to the south is ultimately small potatoes compared to the Downtown-Northgate segment.
I’m convinced the airport wasn’t the destination, and express-bus trips weren’t the goal of Link. I think it was to tie together dense corridors, and the airport just happens to be near the north end of a dense corridor along International Blvd. The density doesn’t extend much north of 140th or so, nowhere close to Boeing Access Road – there’s really only maybe 1 more stop on 99 worth having, north of TIBS.
It makes sense to skip over all the single-family in the middle. And the major employment near Boeing is all around Marginal, pretty easy to serve from the existing alignment if we just added a stop or two.
The density really continues south along International Blvd, on the “South Corridor” alignment. That’s the destination, not the Airport. The airport is just a nearby bonus.
With proper ROW management (transit lanes, etc), an express bus should always provide somewhere near the fastest service. Trying to copy an express bus with rail is not a good idea, and in my opinion, only Amtrak should try to emulate that. For Sound Transit, “regional mobility” should mean more than “point A to Downtown”.
A city’s main airport is one of the most important destinations for rapid transit. No other location has as many daily visitors, or as many arriving without cars. That’s why Link’s ridership from SeaTac is so strong. A rail line also has psychological benefits for visitors. They don’t have to worry about getting lost or stuck somewhere, and it shows the city is serious about mass transit. That encourages them to visit again or do business here.
The original idea was to have Link on Rainier, but they found it to be too narrow and congested. A tunnel was probably already off the table by then. Also consider that the construction disruption would have displaced many more businesses and residents on Rainier than MLK.
Link could not replace the 7, just like it does not replace the 8.
The 7 remains a high-ridership route because many origin-destination pairs are on Rainier, and because Link is further from those east of Rainier, and because there are just so many transit riders in southeast Seattle.
Great. The 7.
The most successful line is the one that LINK obsolesces.
If Link obsolesces it, why does it still have fantastic ridership? ;)
Because most of the people living along Rainier don’t have computers. So, they take the 7 to the Columbia City Library to do their telecommuting. ;)
Do your numbers include weekend ridership?
In short, yes.
The “Annual Rides” number, for instance, is (I believe) all the boardings in the 3rd quarter of 2010, multiplied by four. Similarly for the revenue hours etc. Derived quantities like rides/rev hr are computed using the same numbers.
I’m in data heaven and “getting stuff done around the house” hell. Too.. Much… Data…
A couple of quick observations:
. My favorite piñata of a route, the 219: 5% fare recovery or $392K annually to operate vs. ~$19k in revenue – ouch. The difference would buy a LOT of taxi rides or subsidized Shuttle Express trips. A large portion of the folks I used to give rides to on that route could easily take the 240 instead, BTW.
. The 212 fare recovery numbers should improve with the coming shakeup by folding in the best producing parts of the 225 and 229 but without the relatively unproductive tails of those two routes. I’m sure there are lots of examples like this.
. The 671 fare recovery numbers seem low, especially when you compare to the 15 (RR D). I suspect converting routes to RR might actually drop fare recovery numbers in the short term as you are massively increasing service. I’m sure WPC-types will be happy to jump on this “inefficiency” and totally ignore the improvement in quality and long-term growth potential.
I believe someone at Metro described that route to me (in jest) as “running ten times a day from East Nowhere to East Nowhere”. It’s perhaps a paragon of bad route design.
Even in sixth place ridership on that route is disappointing compared to some of the routes it’s behind. Fourth place might be more acceptable.
Beware of breaking the pinata routes. Some of those are the only lifeline to the nowhere they serve. (As a South Parker, I live under the fear that Central City transit nerds will write my neighborhood off.) I’m reasonably confident that Metro’s February cut list avoids this tragic mistake, and gives every existing constituency continued service, if fewer options and less frequency.
Is there an explanation of what the “KeyType” column means? Some I can figure out but “TB”?
I thought TB stood for TurnBack (variations of the route that don’t run the whole distance), but I could be wrong.
What’s the 15 TB? (the one scheduled route with > 100% Farebox Recovery/Operating Expense) The only 15s I can find on the schedule that turn back before the whole run are at 5am, which can’t be right.
If it’s a trip that would otherwise be deadheading, it wouldn’t take much ridership to get full farebox recovery.
Who knew how many variants of a route they are – I ride the 7 regularly and never knew it had so many variants!
Someone who knows this better than I – how are route changes handled in this data? I.e. Off peak the 49 turns into the 7 and vice versa.
What’s SH? I think with help from above and obvious ones I have figured out the rest.
SH I would assume means Shuttle.
Route changes are handled as separate routes, so when the 7 is through-routed with the 49, boardings during the 7 part of the route are charged to the 7 etc.
Is there any consensus on what single number best represents a route’s ridership or success?
Total ridership? Fare revenue? Passenger miles? Riders per revenue hour? Riders per passenger mile? Average load? Or any of the others?
Or would some figures work best for local routes while others work best for express?
Best is a subjective word.
Not really. If you look at the finished 2009 report, for which the ’09 spreadsheet is the raw data, you’ll notice that routes are rated within subareas and guide times by a factor which represents how well they compare to the average in all of those measures. This is by no means perfect, but if you had to come up with only one number, something like that would perhaps be the best. Again, note that those values are only valid for comparison within the same subarea and guide time.
http://metro.kingcounty.gov/am/reports/2009/2009-RtPerf.pdf
A holistic sampling of many different productivity factors.
Well, Bruce, congratulations. Your seemingly innocuous post meant for us data geeks has now been turned into a context-free slam on Metro planners in Publicola!
http://publicola.com/2011/08/08/some-of-metros-highest-ridership-routes-slated-for-cuts/
I offered a retort in the comments. I just can’t believe they would use this data to try to make it seem like Metro is just trying to scare people. What’s so hard about pointing out that the 43 has parallel routes?
Thank you for posting this data.