"Northbound Link Saturday Morning", by Oran

Last month’s weekday ridership (pdf) on Central Link was 21,774, up 8% over April and a new record.  The figures were 18,710 on Saturdays and 13,641 on Sundays, both records if you discount the large crowds on opening weekend.

Thanks to seasonal variations and other factors, it will be difficult to discern any trends until we have year-to-year comparisons for the entire Seatac-Westlake line, which should happen early in 2011. I’m not sure what else is to be said about these numbers that hasn’t been said in the monthly threads on the same subject.

Update by John Jensen: I thought I’d share a graph of the average weekday ridership thus far. Since the late December opening of the airport stop, we’ve seen a nice upward slope in ridership.

Cental Link's average weekday ridership has improved since the line's opening.

173 Replies to “May Link Ridership Another Record”

    1. I don’t think we will. There’s a piece in publicola today about how much empty office space there is in downtown, due to this recession.

      1. I’d guess that the LR line isn’t mature enough so that fluctuations in downtown office vacancy are the major source of ridership fluctuation.

      2. Edit: replace ” isn’t mature enough so that ” with ” hasn’t matured to the point where “

      3. Agreed. I’m sure ST didn’t factor 20% office vacancy rates into their ridership estimates! Nor should they have.

        Given the situation, it’s great to see Link doing as well as it is. And the economy eventually will recover which implies plenty of upside potential for ridership gains down the road as the vacancy rate moves towards something closer the historical norm.

        Also, I wonder if the existence of Link will tend to help bias the recovery (in vacancy rates) towards downtown. If I was a business looking for new office space in a recovering economy, I’d certainly consider access to Link when evaluating new office space.

      4. In conversations that OED has had with developers and business owners, transit access and pedestrian amenities at the job site are significant factors in their location decisions. Obviously, there are other considerations, but all else being equal, better access and nearby amenities do make a difference.

        However, at this point Link only serves a very small portion of the downtown employment shed. I don’t think Light Rail itself will be a major factor until North Link and East Link come online.

      5. My company relocated and this was a factor in choosing the location. Of over 100 employees, I believe only 6 people drive a car.

      6. message for Adam. If read correctly you mentioned that a major reason your company moved to downtown was because of Link. That’s a great story to hear. Does your employer offer a transit subsidy? Parking subsidy? Thanks!

      7. 25k by the end of July would be right in line with the steady growth seen for the first half of the year. However, one would expect a drop due to vacations and the UW being out. One can hope that the market for office space in DT has bottomed. If so any growth will help drive ridership. I doubt there has been much if any growth in YTD employment DT. In fact I’d guess that it’s been negative as leases expire and clients cut back on area or move to less expensive digs.

        Which all begs the question, where is the increase coming from? The effect of the service changes should be fully accounted for and the remaining bus routes don’t seem to be drained of riders so I’m thinking much of the growth must be from new riders? Given that employment DT has been flat to shrinking I’d further guess that these new riders are accessing DT for shopping/dining/concerts, etc. or as a transfer point to these types of activities outside of DT (Seattle Center, Eastside, UW for example). If so then that’s a bonus since it is filling in non-peak and reverse commute runs.

        It would be interesting to know how these ridership numbers (which I assume are estimates based on the sampling done by automated counters) correlates with the number of paid fares and ORCA taps. Technically, shouldn’t all ridership except fare evaders now be directly accounted for? If there’s a large gap then we’ve got either a problem with the estimates or a big problem with fare evasion.

      8. There are still a lot of flash passes floating around, but I agree that the gap shouldn’t be that large.

      9. I noticed that Sound Transit has been more active in telling passengers to tap before boarding and after exiting Link and Sounder trains with frequent audio and visual announcements at stations. They emphasize that users must do that even if their card is employer-provided. So I think people who have passes may not be tapping in and/or out as they should, which would depress ORCA counts and revenue.

      10. I’ve been noticing a lot more announcements…period. When the train is stopped for a long period of time in the bus tunnel, I usually hear the “traffic ahead” announcement and at the station I heard announcements all the previous weekend that the tunnel was closed. I believe ST is listening and is changing with customer input as much and as fast as they can.

      11. I noticed this on my last ride. I was glad that was the case. In fact, the driver at one point went over the PA system to explain the situation. Glad they’re listening to this blog. Also, I know Bruce say MB and such have these countdown clocks for another purpose, but at least they’re a step up from the useless “2 minute” warning system.

      12. Steady increase in airport traffic especially in the mid to late evening hours both to and from SeaTac station may account for some of the increase.

      13. That makes sense. Most people don’t fly all that often so there would be a learning curve (i.e. make a flight or drop off a relative and go, “gosh, there’s a train going to the airport now. I’ll have to try that next time.” Also, there is a seasonal increase in passenger traffic every year over the summer which should help level out the drop in commuters.

      14. I have to say, the STB staff are some of the most unrealistically pessimistic advocates I have seen.

    1. What about for days when the whether was poor? How about when there’s a convention in town? Keep in mind that Niles and others are manipulating the statistics to express dampened ridership. The idea being expressed is that taking the train somewhere is somehow less real than taking the train somewhere else. Would the commuter who works in downtown Seattle and gets stuck in game day traffic agree with this idea? What other trips do we classify as “real” vs. “purposeless”? The implication is simply wrong: a trip is a trip. What Niles does — manipulating statistics to get a desired outcome — is unscientific.

      Sure, Niles and others compare the average game-day ridership to the average non-game-day ridership (22,637 vs 21,067 in May, for the record), but never note how this exercise alters the underlying data substantially. Of the 20 weekdays measured, nine were game days. When one picks apart data like this, is one reaching meaningful conclusions?

      Probably not. For example, the game day difference was 1,570 in May but was more than double that in April (3,672) all while overall ridership still increased by 8%.

      Another example of the meaninglessness of game day data: May 25 had 21,054 riders and wasn’t a game day. The next day was a game day and had 21,526 riders. Nearly 21,000 people attended the Mariners game. Does that mean only 500 people take light rail to baseball games? I think it’s more likely that the correlation Niles is trying to create is unproven.

      1. Wow. Didn’t know you’d read so much into such an innocuous comment, I was just putting it up there before someone else did, since so much was made of it from last months data. I didn’t attach an opinion to the data, it’s just a simple calculation.

        I’m not sure if you meant it as a lecture, but I don’t really need a primer on what Niles is up to, I’ve been watching him do it for the last 16 years or so. :-)

      2. I was going to post about Mariners stuff anyway because it was (incorrectly) made to be a big deal a few weeks back, but I saw your comment so made it a reply. This is more of a pre-response to Norman and Niles than a lecture to you or anything like that. I guess it’s somewhat useful to know that arguing on “their” terms can lead to poor analysis on both sides.

      3. If anything, separating out the game day / non-game day numbers simply shows how minimal the game day factor actually is. 1,500 out of 21,000? That’s a 7 percent bump on game days; hardly the driving factor in ridership.

        I agree that a trip is a trip, and as far as congestion is concerned, a transit trip on a game day is even more valuable since the baseline congestion on those days is higher and thus every mode switcher makes a bigger difference.

      4. And if that entire bump came from the game then that means at least 5% of people at the game took Link two ways.

  1. We have to remember that this system is only 1 year old. I am sure you can go back to 1987 when the Vancouver and Portland systems were a year old and ridership on their 15 mile lines was still low. There weren’t any TOD’s around their stations, there was still bus duplication and people were still used to driving.

    In 20 years we’ll look back and be amazed at how far we have come. Though I won’t belittle these numbers…this is actually very good news as I see people are getting used to the new line, adapting and getting used light rail being here.

    1. We will look back in 20 years and be amazed at how far we have come, but most of those gains will have come from North Link and East Link, not the Rainier Valley segment. If it weren’t for the fact that Sea-Tac is at the end of the line, ridership would be even lower. In the long run the time penalty associated with traveling at-grade through the Valley will probably cost us more riders from South King than we pick up in the Valley.

      Hopefully, 20 years out we will actually realize this mistake and build a grade-separated bypass through Sodo like we should have done in the first place, then disconnect the Rainier Valley segment from the regional system and put in half a dozen infill stations (~1/2 mile stop spacing) so that people can actually walk to the line and we can triple the TOD opportunities. The Rainier Valley wants a continuous medium-intensity corridor, not a series of disconnected highrise nodes.

      This is what happens when you’re transit infrastructure does not align with your land use vision: poor ridership, wasted resources and angry communities. Light Rail isn’t magic; it only works when you do it right. Sadly, we haven’t so far.

      1. The idea I’m proposing would be classified as a “rapid streetcar” of sorts. It would maintain Link’s dedicated ROW and place stations about 1/2 mile apart. This is far enough apart to maintain reasonable travel times along the corridor, but close enough that the pedestrian catchment zones bleed into each other, creating a continuous corridor rather than a series of nodes.

        The other key to making this would would be substantially higher frequency. This is critical because with a medium density corridor, people often need to make lots of local trips up and down the corridor to access their various daily needs, often in a trip chain. Thus, high frequency is critical.

        Since the line would be disconnected from the regional system and frequency increased, it would never require 4-car trains. Infill stations could thus be reduced in size (and expense) to the length of 2-car trains, which would make it feel even more like a streetcar, though it would still be faster due to 1/2 mile (as opposed to 1/4 mile) stop spacing and dedicated ROW.

        In the long run, a similar line built on similar principles should be run down Rainier Ave. In fact, completing this second line would restore the original streetcar transit system that the Rainier Valley was built around (two parallel lines, one on MLK, one on Rainier). See this map.

        The land use patterns and current zoning of the Valley reflect this pattern as shown on the Seattle zoning map. Notice how the commercial and multifamily zones follow the two streetcar lines shown in the streetcar map linked above. The Land Use pattern is already set, having been set by streetcars nearly 100 years ago. We should be leaning into and enhancing this pattern rather than trying to remake the Valley into something completely different.

      2. What is the time penalty for the MLK segment? I’d be willing to bet it’s no more than 5 minutes, which can easily be made up for by running more frequent trains. The difference in distance between the MLK route and a Duwamish route is negligible, and MLK from Rainier Beach to Mt. Baker only takes 10 minutes to cover. If you’re going to go to the expense of building a parallel light rail line, building it along the West Seattle ridge would benefit a lot more people than building a Duwamish bypass to shave 5 minutes off of trips from Federal Way.

      3. Not really, if you try to go from 15 min headways to 10 minute headways you incur a 50% increase in cost and you’re still stuck on the train for the same period of time; it only helps with transfer time and a more efficient way to handle that is to run on a schedule. Keeping that schedule becomes more difficult the more trains you have on the line.

        You’re right that it will be decades if ever before ridership warrants a Duwamish bypass. It starts to become a bigger deal the farther the line extends south. Hopefully ST will be able to figure out a way to run “express” trains from Federal Way with limited stops to keep travel time competitive with buses.

      4. “Hopefully ST will be able to figure out a way to run “express” trains from Federal Way with limited stops to keep travel time competitive with buses.”

        There’s a fantastically easy way to run express trains from Federal Way, it’s called the Duwamish bypass. It actually would not be that expensive. It’s about six miles worth of elevated track. Given that it would not need to be tunneled and you would need at most one additional station, possibly zero, you could probably get the whole thing done for $500 million or so.

        $500 million he says, like it was peanuts. ;-) To put it in perspective, that’s about 6 months worth of Sound Transit tax revenue.

      5. Any Duwamish bypass would have to contend with the Boeing Easements across E. Marginal Way. Rather than elevated you’d probably have to do a retained cut where the planes cross the road. The other option would be to follow the BNSF ROW down the East side of Boeing Field, but that means no Museum of Flight stop ;-)

        I have to agree with those who think such a beast really isn’t warranted, the added travel time of the MLK segment even with an added Graham stop just isn’t that much. I suppose it might make some sense if the ridership demands shorter than 5 minute headways on the South Line, but the capacity of the DSTT and North link would become the problem at that point

      6. Or elevate high enough to pull a 747 under the tracks. They’re really not that tall. No where near the height of the track at Southcenter. Or run it at grade since there’s zero cross traffic and not even all that much parking lot traffic from the west that needs to turn left across the street.

        I don’t see it happening though since even if it could be done for only half a billion dollars the overwhelming sentiment (and rightly so) would be to extend north and south. Bypass tracks however offer a have your cake and eat it too alternative and could even be incorporated with in fill station construction.

      7. The travel time between Sodo Station and Tukwilla Station on Link is 24 minutes. A grade-separated straight shot between those two stations is about 10 miles. Even using Link’s paltry top speed of 55 mph and some conservative estimates of acceleration and breaking, and assuming an additional stop in Georgetown, you could make it in 13 minutes. Beef up that top speed with some upgraded engines to the standard 70 mph seen in most rail rapid transit systems and make a few other improvements to increase travel speeds between the ID and SODO and you could cut that to 9 minutes. Skip the Georgetown station and save another 1.5 minutes.

        That’s a lower bound travel time savings of 11 minutes, not 5 and an upper bound of 16 minutes.

        People’s general commute tolerance is about 40 minutes, doorstep to doorstep. Factor in access and waiting time and you need to get in-vehicle time under 30 minutes if you really want to be attractive.

        As such, that 11 to 16 minutes makes a huge difference when you are traveling from South King County. 15 minutes isn’t so bad when that’s your whole trip, but when you’re talking about the difference between a 30 minute trip and a 45 minute trip (40 minutes vs 55 minutes with access time), then it makes a big difference.

        Furthermore, the real argument for building the bypass is not that it would allow us to fill in half a dozen additional stations along the Rainier Valley segment, which is necessary to put light rail within walking distance of more people and facilitate a TOD pattern that the community will accept. Adding five or six stations to the existing line along the Rainier Valley would add about 60 seconds per station. For a Rainier Beach to downtown ride, that pushes the total time from 25 minutes to about 30 minutes, which still remains within people’s general commute tolerance (though it’s right at the edge when you factor in access and waiting time).

        However, adding five minutes to a South King to downtown trip, which already starts at a minimum of 40 minutes under the current configuration, is pushing people too far. If you compare the extreme cases of adding stations in the Rainier Valley vs the fastest reasonable travel time for a bypass line, you are looking at a 20 minute gap, or the difference between commute times for South King residents in the vicinity of 25 to 40 minutes (right in the acceptable range) and 45 to 60 minutes (which is outside the acceptable range).

        Serving the needs of the Rainier Valley and the needs of South King County is impossible to do well with a single line. The Valley needs essentially a rapid streetcar. South King needs an express commuter system. Trying to serve both needs with a single line mean serving both areas poorly. Creating two lines allows us to serve both areas well.

        Of course that is more expensive, and as you point out there are other things we could spend that money on, but right now we have a multi-billion dollar system that is falling far short of its potential due to a misguided attempt to build a spork when we really need a fork and a spoon. I’d rather see us fix what we have and make it work before we spend another $5 billion putting rail on the west side of the city.

      8. I’d rather see us fix what we have and make it work before we spend another $5 billion putting rail on the west side of the city.

        I’d rather see a line to Ballard and West Seattle before we even begin to worry about a Duwamish bypass.

        Frankly the line South is more of a political beast than one driven by potential ridership. Just like Snohomish County and North Sounder South King needed to get its piece of the action.

        From a ridership perspective S. 200th should be the end of the line. Demand is really weak past that point.

        Compare this to a line going to Lynnwood where the projected ridership North of Northgate exceeds the entirety of East Link.

      9. “From a ridership perspective S. 200th should be the end of the line. Demand is really weak past that point.”

        The travel time between S. 200th and downtown Seattle are about 40 minutes. Of course ridership drops off south of that station. Ridership always drops off precipitously after 40 minutes. We’ve known this for decades. That’s exactly why the Duwamish bypass is so important. The 15 minute savings brings every South King stop inside the 40 minute threshold, while under the current system, they are outside the threshold, which is why low ridership is predicted despite there being the same population density as the Northgate to Lynnwood segment.

        “Compare this to a line going to Lynnwood where the projected ridership North of Northgate exceeds the entirety of East Link.”

        Travel times between Lynnwood and Downtown Seattle with ST2: 28 minutes. QED.

        “I’d rather see a line to Ballard and West Seattle before we even begin to worry about a Duwamish bypass.”

        I had dictatorial power to extract tax revenue from the entire region and invest it wherever I wanted to, I would have built Ballard to West Seattle, then extended West Seattle south to Sea-Tac and the Ballard line north to Northgate and built an east-west crosstown between Ballard and UW and a north-south bypass from UW to Mt. Baker Station, then another rail line up 99 and then half a dozen streetcars all before I even touched South King, but I don’t have that power and neither do you.

        We’re stuck with sub-area equity for now, so rather than writing off South King as having low ridership potential and sticking it with a system designed to fail, let’s do it right. The marginal cost of “fixing” both our South King line (bypass) and our Rainier Valley line (infill stations) is a tiny fraction of the cost of building to Ballard and West Seattle.

      10. Tony, I like your analysis however I’m curious what would you do for the significant population that lives between S 200th and S 348th? Further, it is complicated by the fact that these people are as likely to commute to Pierce County as they are to Kent Valley, Renton or Downtown Seattle.

      11. Tony,
        There really isn’t a good route from Downtown to the Airport and points South that will match the ridership of lines going North or East. The employment, residential, and retail density simply isn’t there compared to the other corridors.

        The reason I said stopping at S. 200th makes sense is the density also falls off a cliff further South.

        Sure there are a lot of people riding buses in SW King County but it isn’t anywhere what you would need to get say 50,000 riders per day.

        All that said, I think the line as planned needs to go at least as far as the Federal Way TC rather than stopping at Star Lake.

        As for your assertions about Rainier Valley I think you are wrong. Yes there is a small group of vocal NIMBY types in SE Seattle, but for the most part these are the same people who were behind “Save Our Valley” or fellow travelers like John Fox. They make a lot of noise but I really don’t feel they are representative of an entire quadrant of Seattle.

        I really don’t see where one would put “five or six additional stations” between Mt. Baker and Rainier Beach. The only two real nodes along the entire MLK stretch without a station are at S. Graham and S. Orcas. North of Columbia City and South of Othello topography, the street grid, and the powerline ROW limit the potential station walksheds quite a bit.

        For what it is worth I do support putting infill stations in. One between Graham and Orcas, the other at S. 133rd with a P&R and transit center.

      12. I am a transit user and fan, not a professional (and my many thanks to those of you who are). Reading Tony’s post, his points seem so obvious: you would design transit differently if the purpose was to stimulate development and serve a transit-reliant community than you would if you were trying to move commuters rapidly from the suburbs to the center city.

        “Light Rail isn’t magic; it only works when you do it right. Sadly, we haven’t so far.

        Why? Why haven’t we done it right? Will North Link and East Link be “right”?” Sadly, I fear that we non-professionals who are so hungry for transit are only good at expressing our desire for more transit, not smarter transit. How do we help get smarter, more effective, more efficient transit?

      13. How do we help get smarter, more effective, more efficient transit? Aren’t we getting that right now? It may be slow, but we are getting it.

        I mean, it may not look like it, but there really isn’t any other US city that has so many transit projects coming down the line in the next 10-20 years. Not even Portland has a much as we have that has been funded. Dallas, Denver and Phoenix have had to curtail their light rail expansions and yet we are building an additional 35 miles by 2023. Houston, Minneapolis and Charlotte (the other LR newbie’s) have been talking about expansion plans for some time now, but no shovel has dug any dirt yet. LA still hasn’t got their LR to the airport, New York is just building the 2nd Ave subway 100 years later and look what it’s taking to expand BART to San Jose. Miami and Atlanta can’t expand their expensive metros and look at the mess Vegas has on its hands with the Monorail. And if you’re transit hungry…be glad you’re not in Kansas City, Austin, San Antonio, Indianapolis, Milwaukee, Cincinnati, Columbus or Detroit.

        This doesn’t even take into account the possible passage of ST3 around the 2016 time period which could result in another 30-60 miles of additional light rail. Sounder will be expanded down to Lakewood and I am sure ST3 will also include expansion of the system on the east side. There will also be expansion of the Seattle streetcar system as well, the First Hill line will start construction next year. Tacoma is talking about expanding their streetcar and Issaquah is ready to bring their streetcar back on line. The 6 Rapid Ride lines are all coming on line in the next 5 years as well. My mind is spinning with all the things going on here transit-wise.

        I realize that we here in the Puget Sound region have been transit hungry since we are latecomers to the mass transit ballgame, but obviously your hunger will be fed continually over the next few decades.

      14. Michael,

        You’ve simply reenforced Jennifer B.’s point. You point out how we are getting lots of new transit, but none of that guarantees that it will be good transit.

        Look at the planning that is being done for Rapid Ride: 1/4 mile stop spacing, on-board fare payment at half the stops, minimal provision for dedicated ROW and 10 minute headways at peak times, 15 minutes off-peak? That’s the bare minimum standard that we should expect for a poorly implemented local bus route. It’s not rapid transit by any definition.

        You are correct, we are investing heavily in transit in this region. The problem is, due to poor planning, we are going to spend billions on transit and then find ourselves terribly unsatisfied because the transit does not do what we want it to do. Then we’ll wait decades longer and spend billions more hoping that that will make transit do what we want it to do, all the while thinking that the problem is lack of quantity rather than lack of quality.

        In truth both quantity and quality are critical. I’d love to double or triple our current investment in transit. Even a perfectly optimized transit system will fail if it does not have sufficient resources, but a transit system that is resource rich but poorly planned will fail just as much.

        Furthermore, poor implementation is the surest way to undermine public support for future funding. Good implementation that delivers high public value for the money is the best way to convince the public to invest more.

      15. So, back to my question–what can we, the non-professionals, do to get smart transit?

        For example, will Metro listen to me if I tell them I support stop consolidation on routes that I don’t ride?

        How do I, a Link rider in the city, express my angst about suburban bus routes with nearly empty buses?

        How do I make sure that the new Link segments have properly spaced stops and use grade separation where appropriate, when I don’t personally know the answers to these questions?

      16. I think that if ST had tried to build a line any where other than Rainier Valley they would still be stuck trying to litigate their way past the NIMBYs who would be laying down in front of the bulldozers to protect their ‘hoods– just my opinion. Going through the Valley isn’t the best alignment to get to the airport, but airport to downtown traffic alone wouldn’t justify the cost of building the line, so we have a compromise alignment. Hopefully the City will recognize the potential of LINK and create the appropriate construction and land use codes to make the most of the line. If the day ever comes that ST wants to build an express line to the airport via Georgetown, the Valley line could be extended east to the Rainier Beach business district (proper) and then maybe on to Renton.

      17. “I think that if ST had tried to build a line any where other than Rainier Valley they would still be stuck trying to litigate their way past the NIMBYs who would be laying down in front of the bulldozers to protect their ‘hoods– just my opinion.”

        Yeah, all the NIMBYs in SODO, Georgetown and South Park would surely have blocked a straight shot to the airport.

        “…airport to downtown traffic alone wouldn’t justify the cost of building the line, so we have a compromise alignment.”

        True, but South King County Commuters + Regional Airport Traffic + 80,000 industrial jobs in SODO is worth the investment. I will admit that if we had planned on stopping at the airport, then the current configuration would have been fine, although we should do a few infill stations in the valley, but the plan was never to stop at the airport. The plan was always to go south to Federal Way or beyond, and if that was always the plan, then the line should have been built accordingly.

      18. True, but South King County Commuters + Regional Airport Traffic + 80,000 industrial jobs in SODO is worth the investment. I will admit that if we had planned on stopping at the airport, then the current configuration would have been fine, although we should do a few infill stations in the valley, but the plan was never to stop at the airport. The plan was always to go south to Federal Way or beyond, and if that was always the plan, then the line should have been built accordingly.

        The first problem is you aren’t likely to get many of those people working in the Duwamish industrial area. There is a reason much of the area has poor transit service and that the commute share for the area is rather poor. While there are a lot of jobs the actual job density is rather low other than right around SODO center.

        The second issue is the South King County Commuters just aren’t that much of a transit market. Sure plenty of people ride the bus, but the density just isn’t there to create the ridership a rail line would need.

        Of course you have to also consider the politics of how we got to where we are as well. If Rainier Valley had been bypassed for SODO think of all the cries of “economic redlining” we’d be hearing. SE Seattle also happens to be a fairly strong transit market as opposed to the Duwamish industrial area. The desire to provide economic justice, boost ridership, and some issues with the Boeing easements across E. Marginal Way are what pushed the line into Rainier Valley in the first place.

      19. cries of “economic redlining”

        Bingo, the rest of the distractions were just attempts to make the data match the pre-decided outcome (OK, City of Tukwila politics didn’t help either). With the direct route you’re serving not just the Duamish industrial area but Georgetown, White Center, Tukwila/Burien and it makes feasible a direct transfer point under the West Seattle Freeway. Not only would they have avoided the +/- billion Beacon Hill tunnel but would likely have been able to save money on the service facility and left more room for expansion to boot.

        Job density used to be quite high down around Boeing Field. But when you outsource parts and turn machine shops (like the GE turbine rebuild facility) into home repair warehouse space of course job density suffers. Decent transit might be the shot in the arm needed to reverse that trend. If not, well it’s still probably a better place for up scale mixed use yuppie condos than RV since it’s closer to DT and is an extension of that density rather than spreading it hither tither and yon.

        The irony is that after driving out many local businesses during construction the RV will end up with a few enclaves of upscale housing built up around the Link stations with under served areas along the rest of the route because, “well, they’ve got a train why do they need buses too?” This in place of what should have been high frequency local transit that could have transferred at Ranier Station (would Dearborn be a legitimate name?) and extended south to Renton and Tukwila Station. Tony was close with his spork example; I think it’s more like they needed a spoon and got a cork screw :=

      20. Frankly I think the Beacon Hill to Rainier Beach stretch is the best part of the whole damn line. High quality inner-city transit, what a concept! Federal Way can suck it. :-%

      21. “The second issue is the South King County Commuters just aren’t that much of a transit market. Sure plenty of people ride the bus, but the density just isn’t there to create the ridership a rail line would need.”

        Uh, no.

        First, it’s not population density that matters; it’s trip density. Park-and-rides are the primary means by which we expect South King commuters to interact with transit. Sound Transit uses a 2.5 mile radius for their park-and-ride catchment area, though I think they could easily use 3 or 3.5 miles. The maximum radius for a pedestrian catchment area 0.5 miles. Even if we assume that the Rainier Valley is 2 to 3 times denser than South King, the catchment area for a park-and-ride is 25 to 50 times the area of the pedestrian access zone. Thus, you have 10 to 20 times the potential trip generation for a South King stop than you do for a Rainier Valley Stop. Of course, we could have built park-and-rides in the Rainier Valley, but take that one up with the social engineers at City Hall.

        Secondly, there is political support in South King for the kind of TOD that would be well served by a high speed line with wide stop spacing: a very compact, high-rise node built in the immediate vicinity of the transit station surrounded by single family housing that accesses the station via the park-and-ride. This is also the land use pattern being proposed for Bel-Red, and it has been fantastically successful in Vancouver. However, this type of TOD, not to mention the tens of thousands of potential park-and-riders, will not happen if the travel times to downtown Seattle at UW (about 300,000 regional jobs) are too long.

        There is, by the way, no support for the land use pattern described above in the Rainier Valley. The support is instead for a continuous medium density corridor consisting primarily of low-rise (4 story) multi-family and mixed use tapering off to single family all within walking distance. This land use pattern requires stops every 1/2 mile and could be facilitated by adding five or six infill stations along the line.

        “Frankly I think the Beacon Hill to Rainier Beach stretch is the best part of the whole damn line. High quality inner-city transit, what a concept! Federal Way can suck it. :-%”

        Yes, what a great concept. However, that is not what the Rainier Valley segment is. By trying to be both a high-quality inner city line and a high speed regional line, Central Link has done both poorly, and ridership is suffering because of it. As I said above, we need a fork and a spoon, not a spork, or a corkscrew as Bernie said.

      22. Even if we assume that the Rainier Valley is 2 to 3 times denser than South King, the catchment area for a park-and-ride is 25 to 50 times the area of the pedestrian access zone. Thus, you have 10 to 20 times the potential trip generation for a South King stop than you do for a Rainier Valley Stop.

        That statement assumes unlimited capacity in the parking lot, which is unlikely.

      23. “Park-and-rides are the primary means by which we expect South King commuters to interact with transit.” “there is political support in South King for the kind of TOD that would be well served by a high speed line”

        The two are almost mutually exclusive. All of BART’s park-and-rides are sprawl machines, allowing people to live further and further from the city without having to pay the full penalty of their commute (worse – the region is paying for their sprawl in the name of transit!). Also, none of BART’s park-and-rides have any sort of real TOD built up despite the decades and decades of time that have elapsed since they were built. Why? Because a park-and-ride station attracts car users. Although it’s great to imagine people parking, getting out of their cars and shopping, in reality they’d rather just stop somewhere else on the way with parking closer to the front door of the store.

        Also, parking takes up space for TOD. Sure, I can imagine building parking vertically, with 20 floors of parking under a station with retail and housing above. But in reality, that’s too expensive. Stations are built surrounded by parking (see: Tukwilla Station), leaving people to get off the train and walk through parking lots to get to TOD – good luck with that.

      24. Tony,
        I disagree about the support for density in Rainier Valley. So far we’ve heard from the same small group of people who were behind “Save Our Valley” and John Fox. Sure they’ve made a lot of noise and drowned out those who would support more density around the SE Seattle Stations.

        You really don’t need 30 story condo towers to make a system like Central Link work. As was shown during the debate on the TOD bill modest upzones can take the walkshed around a transit station from Seattle SF density to the density needed for a walkable neighborhood that generates lots of transit demand. NC3-65′ along the arterials in the station area gets you most of the way there. That said I’d like to see higher limits right at the stations say 85′ or even 120′. Even so a lot of density could be added in all of the station areas simply by filling in the parking lots and vacant lots with what is allowed under the current zoning.

        One final point, even if one were to double transit usage in SW King County you still wouldn’t generate nearly enough ridership to make up for the loss of the SE Seattle stations.

      25. Bernie,
        The trend in manufacturing in developed countries is to make more with fewer workers primarily due to increased automation.

        Sure we could have built a line along E. Marginal assuming those Boeing workers would always be there, but then we would have a line with stations serving the relative handful of workers there now.

        There may be 80,000 workers in the area from Royal Brougham to the City Limits and from I-5 to West Seattle and Delridge. But remember this is a huge area making up a healthy chunk of the land inside the Seattle City limits. Compare this to the UW or downtown Seattle the job density is much higher and much easier to serve with transit.

      26. The comparison isn’t to DT and UW. The line is going there. The comparison is the number of jobs and potential for large employers along the Duawish vs Ranier Valley. Sure automation has reduced the need for labor but why did Boeing move production out to Fredrickson and Corporate to Chicago? I think it’s got a lot to do with, Seattle, King County and WA State in general not making jobs a number one priority. The Duwamish is unique in having access to the Port, Rail, nearby affordable housing and it’s already zoned industrial. The potential in RV is a few mixed use condos.

      27. Hmm. Boeing moving corporate to Chicago had a lot more to do with McDonnell-Douglas folks not wanting to be in Seattle and with getting senatorial support for the refueling contract than it did with WA State’s jobs policy. Aside from the idiotic B&O tax, WA State’s actually a pretty good place to do business, especially if you need educated workers.

        That said, comparing jobs between the Duwamish and RV misses the other end of the commute. Aside from a relatively small pocket in Georgetown, there’s no housing between SODO and Tukwila — a Duwamish-based starter line would have virtually no riders who didn’t park at TIB. That’d be insane.

        The Duwamish bypass may make sense once the system is more mature and has more residential connections. But to say it would have made the right initial route is just crazy.

      28. That’s not true. TIB is a perfect transfer point for people living in Burian, White Center and Renton. Far greater than any of the RV stations which aren’t really teaming with people (if your arguement is they will be when development comes then the same would apply all along highway 99 and in Georgetown). Without the expense of tunneling under Beacon Hill the line would have extended far South of the airport. It also would have made practical a transfer from the West Seattle freeway to DT.

      29. “TIB is a perfect transfer point”

        It still is.

        “Without the expense of tunneling under Beacon Hill the line would have extended far South of the airport.”

        The S. 200th extension is costing the same as the Beacon Hill tunnel. S. 200th is a giant parking lot on one side and a trailer park on the other. Unless they build a giant P&R (which I thought you were against) ridership there will probably be pretty low. Metro has shown before that they don’t have the guts to do a proper job of altering bus service to feed into the stations.

        “if your arguement is they will be when development comes then the same would apply all along highway 99”

        Tukwila didn’t want Link on Highway 99.

        “It also would have made practical a transfer from the West Seattle freeway to DT.”

        Sodo Station is 1/4 mile from the Spokane Street viaduct. Metro wouldn’t pony-up the dough to build a transit ramp from the busway to the viaduct.

      30. $300M was just for the tunnel, add in the cost of the underground (way underground) station, station art, filling voids… Then you’ve got another mile of mostly elevated track going east/west to complete the connection from SODO to Mt. Baker. Then you’ve got more elevated track at the other end to get back to the west side of I-5. When you add all that up it’s more like $600M. And it’s pretty pathetic that ST needs $300M to go two miles (Over the U.S. as a whole, excluding Seattle, new light rail construction costs average about $35 million per mile. – wiki). Additional cost savings could have come from moving the MF to cheaper land south that didn’t require elevated track.

        You’re right that Metro doesn’t seem willing or capable of providing feeder service to rail but I’m just pointing out that the potential with Burien, White Center, and West Seattle is as good or better than RV. But ridership wasn’t the reason RV was anointed.

        Ramps from Spokane St. to the bus way are a different animal than a connection at Spokane St. You wouldn’t have to run them up the bus way which then ends up putting the buses DT anyway.

      31. That average cost is brought down considerably by systems like Portland and Salt Lake City where light rail has been built almost exclusively in existing ROW. It’s pretty much the norm in the States to go the cheap and easy route of re-using railroad ROW and building in the median of freeways, regardless of what effect it has on ridership, access to stations and future development. You have to look at construction costs of lines with similar profiles (tunneled, elevated, whatever) to see if Sound Transit’s costs are in line or not. And is $300 million a lot for two miles of elevated track and a station? It’s a lot of money to me, but in the world of concrete and steel it doesn’t seem like that much anymore. The new braided ramps between 405 and 520 are costing nearly as much, and that’s what, 1/2 mile of concrete?

      32. According to ST’s project estimates Agency admin is $12M, Final Design is $40M and ROW acquisition $35M. So before we even get to the concrete and steel we’re more than $8M a mile north of the US average. Actual Construction cost is only 54% of the budget. There is however a whooping big $56M contingency fund, almost 20% and well above the difference between their high and low estimates padding the total. WSDOT’s budget for the braid ramps is $227,559. It’s more like a mile of two lane multi level roadway (NE 8th to 520) in each direction and includes rebuilding the NE12th Bridge (bike lane, yeah!).

      33. “WSDOT’s budget for the braid ramps is $227,559”

        Think you’re missing three zeroes!

      34. “I disagree about the support for density in Rainier Valley. So far we’ve heard from the same small group of people who were behind “Save Our Valley” and John Fox. Sure they’ve made a lot of noise and drowned out those who would support more density around the SE Seattle Stations.”

        This is how I interpret it as well — though I’m up on Beacon Hill, so maybe the zeitgeist is different up here. But the people who are doing the anti-density stuff are mostly the same folks who were involved in Save Our Valley, etc; the general opinions that seem to be expressed by most others are reasonably supportive of increased density and development along the Link line, though of course many folks do still worry about losing open space and their SF housing, and being priced out of the neighborhood. People in SE Seattle are desperate for more retail and services that other neighborhoods in Seattle routinely have and SE Seattle does not, and smart development down here will bring that.

      35. I think it was courageous of them to try to use Light Rail to spur TOD in the Rainier valley. I think it is too soon to deem it a failure. I think there does need to be significant changes though. I would suggest they need some form of circulator bus shuttle between the stations and the business districts along Rainier. Additionally there should be frequent local bus segments along MLK.

        I think the train time is too long between Sea-Tac and downtown. I would like to see a bypass from SODO to the airport link near south Boeing field. But that shouldn’t preclude a keeping the line through Rainier valley. I also think they should re-consider a link to SouthCenter area or even to downtown Renton. Perhaps the some intervening station between Rainier Beach where passengers could switch to a Sea-Tac bound train where the Rainier valley train continues to Renton.

      36. It may not be a circulator bus shuttle in the traditional sense, but the #8 bus goes past the Mt. Baker, Columbia City, Othello, and Rainier Beach Stations (and points in between). It has 15 minute headways most of the day on weekdays/Saturdays and 30 minute headways mornings/evenings/Sundays.

      37. Believe it or not, some people would rather not have to walk the 5 to 10 blocks between MLK and Rainier Ave. at Columbia City or Rainier Beach stations.

      38. Well, I do believe strongly that there needs to be a station between Graham and Orcas…it IS needed! However, people not wanting to walk to stations? They’re just going to have to get used to it. As I have said countless times here on this blog, go to any city in Asia or Europe and people are used to walking up to a couple miles to get to a metro station and they are quite used to it.

      39. “However, people not wanting to walk to stations? They’re just going to have to get used to it.”

        Actually, they’re not. In America, people have choices, and one of those choices is to drive. This people-can-suck-it attitude will not capture riders, but it is a great way to waste billions of dollars on boondoggles that no one uses, thus undermining support for future transit investment.

        “As I have said countless times here on this blog, go to any city in Asia or Europe and people are used to walking up to a couple miles to get to a metro station and they are quite used to it.”

        In Asia, population densities are 10 times that of Seattle. Europe has 3 to 5 times our density. Personal income is lower; vehicle ownership rates are lower; they have far fewer miles of freeway per capita; parking costs a fortune if it is available at all; and gas costs $5 to $10 per gallon. In both Europe and Asia, the cost of driving, both financially and in terms of convenience is far higher. When you’re competition is weak, you can afford to take a people-can-suck-it attitude, but in America, people have choices.

      40. Plus there are a lot more thinner Europeans and Asians because they walk to train stations more.

      41. You are disregarding one of the main reasons people might be unwilling to walk at those stations — perceived safety issues. As a small female, I wouldn’t do that walk at night alone. I’m not particularly worried during the day, though. At Rainier Beach, you have to walk past the Chief Sealth trail crossing where there aren’t enough eyes on the street and it is very unsafe feeling at night.

        The Columbia City walk to Rainier on Edmunds is only 3-4 blocks, and does feel safer, but I’d still be a bit uncomfortable there at night because there’s not enough foot traffic at that time. I have still done it, but I worry.

        It’s not just laziness that makes people not want to walk to a station sometimes.

        Having said that, I walk to/from Beacon Hill Station in the dark all the time. Maybe it’s just that it’s my own neighborhood so I feel more comfortable there, but it also seems slightly better for “eyes on the street,” at least, the route that I would follow to get there is.

      42. Tony, I think you’re seriously underestimating the scale of redevelopment that is going to happen in RV over the next ten years.

      43. Want to place a bet right now?

        I’m going to say no more than 30,000 new housing units within walking distance of all four Rainier Valley stations combined, and that is a generous upper bound. We’ll probably get 1/2 that.

        30,000 is also about the number that lie within the pre-TOD park-and-ride catchment area of a single South King station (assuming we actually built enough spaces). Factor in TOD (which will only happen if travel times are fast enough) and you can expect the potential ridership at a typical South King station to double.

      44. 30,000 units in ten years? Seattle’s 20 year goal for the entire city was only something like 45,000 units.

      45. First off, my numbers were off by a factor of 2. I was thinking population, not housing units. I’m prepared to bet today that we won’t see more than 15,000 new housing units in the Rainier Valley, and probably much less.

        Secondly, 15,000 is an extreme upper bound. I really don’t expect there to be anything near that level, but Jason Mitchell suggested that I was “seriously underestimating the scale of redevelopment that is going to happen in RV over the next ten years.”

        My point is that even if we get 15,000 new units in the Rainier Valley, which we won’t, ridership will still be poor relative to North Link and East Link, and compared to the potential ridership we could have had in the Rainier Valley and South King if we had done it right. So unless Jason Mitchell expects that we will get substantially more than 15,000 new units, his suggestion that my assessment stems from Central Link an underestimation of TOD potential does not hold up.

        FYI, for comparison, New Holly is 1,400 units. You would need three New Hollies at each of the four Rainier Valley Stations in order to get to 15,000, and that would still amount to failure.

      46. Tony,

        For someone arguing that we need quality transit not just quantity, I’m concerned by your ridership-numbers-above-all-else attitude. Park-and-ride lots are a mistake. Yes, they greatly increase ridership. But they absolutely induce sprawl. We need a TOD model for transit, making stops into urban centers. Each new rider from a low-car, dense area can be an environmental and quality-of-life win.

      47. Tony the Economist, I concur with Matt the Engineer that your obsession with ridership is curious, but—just for laughs—what ridership numbers out of Central Link would satisfy you?

        Also? Can we clarify exactly what population gain you’re willing to bet we will or won’t see in RV over the next ten years? I’d like a concrete number. And what exactly are you willing to bet? I’m told by people who don’t cook that I’m quite the cook and I am happy to make you dinner if I lose.

        – Jason the Mitchell

      48. If the SODO bypass track is built to the airport, the Rainier Valley line could be extended to Renton. Its a straight shot on MLK heading south directly into downtown Renton.

      49. It would certainly be possible to build a bypass line later and turn the MLK/Beacon segment into a streetcar. Pretty much all extensive rail systems have reconfigured their routes later, switched stations to different routes, or abandoned stations. In Manhattan, trains go through neighboring tunnels when their main tunnel is under maintenance.

        Grade-separated lines have one major advantage which is rarely mentioned. Not only can they run full-speed without stopping for lights, but they don’t have car/ped collisions. All those five or six accidents would have never happened, and Link’s reputation would be higher.

  2. A point to note: The chart would look a little less dramatic if the vertical axis actually had zero at the bottom of the scale, rather than 12,000. The graph could give the visual impression that ridership has dramatically increased, by 3x or 4x, which is not quite the case. Ridership has increased, which is good, but less dramatically than indicated on the chart.

    1. That’s a good point. I’ve re-done the graph so the vertical axis starts at zero.

  3. Martin,

    Does ST release figures about how many boardings take place at each stop? I’d be curious to know how much of the ridership is being driven by Sea-Tac and the park-and-ride at Tukwilla and how much is being driven by the in-city stations in Beacon Hill and the Rainier Valley.


    1. I believe they release stop level data in the annual Service Implementation Plan. Link will probably be the same if the data is available.

  4. Great news, I still think we will have to wait to see huge increases until UW and East Link open up. I have seen some small differences on NB I-5 flows during the morning rush on the flow maps where the parallel to Boeing Field no longer is congested. However, the West Seattle Bridge to DT is still congested. I think once BRT is in place to West Seattle, that should help relieve I-5 congestion heading into DT. I will probably go over and see what an afternoon is like riding Link to get a first-hand look of what traffic is like in the coming days.

    1. The purpose of transit improvements should not ever be to decrease congestion on the “free” motorway. That can only be done by accurate cost-to-society pricing.

      Transit improvements are for the increase of mobility, both for the portion of the population that thinks it can own and operate a motor vehicle and for those who have figured out they really can’t.

      1. Exactly. It’s very frustrating to see folks like the Washington Policy Center making all this noise about congestion when the very studies they cite show New York to be the second most congested city in the country. Of course, New York also has a subway system which provides 7 million rides a day.

  5. http://www.flickr.com/photos/viriyincy/4632560071/sizes/l/in/pool-624040@N24/

    I like this picture from Oran at the top of this post. Two comments and a question regarding this picture:

    1) The people whose faces you can see look pretty glum. This is the normal expression I usually see on passengers on Link trains. I have to laugh when people write about passengers riding Link trains with “smiling faces.”

    2) Large pieces of luggage are always “stowed” on the floor (when they are not put right on seats), usually in aisles. You can see in this picture how just two large pieces of luggage block the aisle. These two pieces of luggage also reduce the capacity of this car by about two, since each piece of luggage takes up close to one standing spot. So, whatever “capacity” figure you want to use for Link cars, subtract about one person for each large piece of luggage. And there are generally always some pieces of luggage on just about every Link car.

    3) For those who claim that there are a lot of people who refuse to ride buses, but who will ride light rail, would you please point out those people to me in this picture? Who of the people on this train are people who would refuse to ever ride a bus? I would like to be able to recognize those people. Because, to me, these people look just like the people I see on Metro and ST buses.


    1. In the above comment we learn that taxpayer money should only be spent on projects that make people smile with glee (except highways), that no travel should be allowed if one has luggage with them (except in single-occupancy vehicles), and that Norman has the unique ability to look at someone through a photograph and determine what transportation modes they frequent.

      1. On the contrary, I can’t tell what modes people would use. However, there are people who claim that at least some people who ride Link would never ride a bus. I’m just wondering who those people are, and how we can recognize them. I don’t ever see anyone riding Link who looks like they would refuse to ride a bus. But some people claim this is true.

        And how do you read that I believe people should never take luggage? I just noted that there is no place to put large pieces of luggage on Link trains, except in places where passengers could otherwise stand or sit. Therefore, capacity on any Link train is reduced as the amount of luggage on that train is increased.

      2. Can you point out exactly where you can store luggage on a bus? Either that or reduce your capacity figures for buses.

        Also, please describe how we can recognize people like you who wouldn’t take light rail.

      3. Norman, I’m sure the straw men (“are people who claim”) will come out of the wood-work to defend their opinions in day. But back here in reality, I note that 1) you cannot tell who would or would not ride a bus based on their looks, 2) rail bias is an actual observed phenomena that is separate from your ability to decide for people what modes they prefer based on looking at them.

        As for capacity arguments — what’s your point? The marginal cost of increasing light rail capacity is small: run another train at a higher frequency. The marginal cost of widening I-5 is so tremendous that it is simply not a feasible option.

      4. As for capacity arguments — what’s your point? The marginal cost of increasing light rail capacity is small: run another train at a higher frequency.

        Actually until you reach the limits of the system you can increase capacity by adding more cars to each train too which saves the cost of an additional operator.

        Still you are right, even if each link train only could hold as many people as a 40′ bus the cost of running more cars more frequently is still much less than trying to widen I-5 would be.

      5. There is no need to widen I-5. Why do you keep making this false assertion? You can greatly increase the capacity of I-5 by adding buses — no need to add lanes!

        Is this too complicated for you? If you replace 3 SOV’s per minute in one highway lane with 1 bus per minute, you go from 3 people per minute in that highway lane to up to 90 people per minute in that one bus. The same lane would be carrying 30 times as many people per minute!

        Highways are not nearly at capacity for PEOPLE. Replace some SOV’s with buses (or even vans), and you have much greater capacity withOUT adding any lanes!

        2,000 SOV’s per hour = 2,000 people per hour in one highway lane.

        100 buses per hour = up to 9,000 people per hour in one highway lane.

        Is this too complicated? JUST ADD BUSES! You don’t need more highway lanes to increase highway capacity, or ridership.

      6. You are correct Norman, that if we removed 2000 SOVs per hour from I-5 this would solve bus congestion problems and leave a lot of room for many more buses. However, removing SOV traffic is easier said than done. Politically, it is far easier to build new ROW for trains (which many other cities have done). In fact, given the relatively small number of bridges over the ship canal one could argue that University Link is worth it for the additional tunnel route alone.

        Also, while a few billion might seem like a big number it’s fairly tame in the big picture of regional expenditures over decades. Highway projects easily cost billions, hundreds of millions for replacement bridges are needed, and so on. Seattle’s Bridging the Gap levy will spend over $350m on local road maintenance projects alone, and that’s extra from SDOT’s regular budget.

      7. Why are you talking about highway projects? I am talking about adding buses to existing highways — not building more lanes. I thought you understood this.

        How about this: try adding more buses to routes which are regularly full. See if that alone does not take some SOV’s off those roads. Why do you seem to believe that just adding more buses will not, by itself, remove cars from highways? It seems to me that ever time Metro or ST adds hours of bus service, ridership goes up. Is this not true? If it is true, where do these additional riders come from, if not from out of cars they used to drive?

      8. You are correct in the sense that we don’t have a capacity problem with roads. We have a peak capacity problem. Metro ridership has increased from roughly 100 million in 2000 to 126.9 million in 2008. If you figure most people take a round trip that’s about 13.5 million “round trips” or about 43,000 cars off the road. King County during this time has grown from 1,737,034 to 1,884,200. That’s 147,000 new people. Divide by the average household size or 2.39 and multiple by say 1.7 cars per household (that’s a guess, if someone can find that stat let me know) and we’ve got 105,000 more cars. Transit isn’t keeping up and simply adding more buses (or trains for that matter) won’t change that. Only dedicated ROW, be it lanes for buses or grade separated rail will provide relief (for those who can and choose to use it).

        You have to add peak capacity. That’s why a fair comparison has to be building new lanes, increasing HOV limits (3+, 4+, ?+) or rationing (odd even days based on license plate number, etc.) or onerous tolls or sky high gas taxes. Just increasing buses won’t eliminate (or even dent) the peak capacity issue. We’ve been doing that and it only mitigates it. It can’t eliminate it and it can’t provide a clear alternative. You’re trying to compare apples and kumquats.

      9. Just to be clear–I was talking about existing road and bridge maintenance projects, not new lanes. Some of the proposed highways do have new lanes, but that wasn’t my main point. As far as I know there are no serious proposals to widen I-5 or make 15th Ave NW an interstate or anything like that.

      10. Not to mention the buses are getting stuck on I-5 also. The only way to keep the buses moving is to separate the HOV lane from the regular lanes of I-5 – just adding buses does not cut it at this time.

      11. Rail tracks, bridges, etc. have to be replaced periodically, too.

        We have not really tried significantly increasing bus capacity on current highways. I am talking about at least doubling current bus hours on heavily-utilized routes.

        For example, on the I-90 bridge, there are currently about 50 buses per hour in each direction during peak hours, giving a capacity of about 4,500 people per hour on buses. Double that to 100 buses per hour, and you get a capacity of about 9,000 people per hour — an increase of 4,500 people per hour.

        If most cars on the I-90 bridge are SOV’s, then a lane with 2,000 cars per hour is carrying no more than about 2,000 to 3,000 people per hour. By increasing bus capacity by 4,500 people per hour, you have the capacity to eliminate thousands of SOV’s/hour from that bridge.

        Are you saying that removing thousands (or even 1,000) cars per hour from the I-90 bridge would have no impact on congestion on that bridge?

        Kaleci: you fail to understand my main point: adding buses reduces the number of cars, thus relieving congestion.

      12. To carry 9,000 people per hour you could instead run Link and do it with about a tenth as many operators, and no diesel.

        Meanwhile, I’m interested in raising taxes to pay for bus service, and you’re not.

      13. Norman: Okay – let’s assume traffic goes down on the freeway so the buses maintain a reasonable speed. How do you expedite their exit from the freeway and through the streets of Seattle? The can’t go through town at 55 miles and hour and the stops can only accommodate so many buses an hour.

    2. In regards to your second point-everyone has easy access to a door in spite of the aisles having luggage in them. I remember how hard it was getting on or off the 194 at the airport with or without luggage. Add standees and it really got difficult.

      Frequent service on Link means I am now able to get to the airport by 9AM on a Sunday and back home as late as midnight. Much better service for me anyway.

      1. BRT-style buses with three wide doors, and off-board payment should be just as easy to board and exit as Link trains.

        Buses could run on the same schedule as Link, and for a lot less money.

      2. What for? We already have a great light rail line that serves the airport! Ha ha ha.

      3. And takes about 10 minutes longer than the 194, plus a 5-minute walk from the Link station to the terminal.

      4. “And takes about 10 minutes longer than the 194” Except when it doesn’t. Ah, fond memories of being stuck in traffic on the 194 when I should have been on a plane.

      5. If you are flying Alaska or United it is a damn long walk from the bus stops on the South end of the baggage claim level to the check-in counter.

        If you don’t like the station location blame the TSA and the Port of Seattle.

      6. Exactly. Thought exercise BRT and actual BRT are often separated by a large chasm. I don’t think you can throw a bus onto a carpool lane, pin on an “Express” sign and call it BRT.

      7. SWIFT is good enough, and cost about $30 million for a 17-mile line.

        Central Link cost about $2,600 million, or 87 times as much as SWIFT.

      8. How do you define “good enough?”

        I don’t think “good enough” will get us to where we need to be in terms of mobility…especially if population projections pan out.

      9. And, “good enough” will never get you from the U District to Downtown in 6 minutes during rush hour (or any hour for that matter).

      10. “…will never get you from the U District to Downtown in 6 minutes…”

        Especially since Norman is on record as opposing dedicated ROW for BRT.

      11. People on this blog have written over and over how buses between the U. District and downtown are often crammed to the gills. If people fill buses to capacity, then, obviously, buses are “good enough.” If buses were mostly empty, you might be able to make some sort of argument.

        Regarding SWIFT: SWIFT was carrying about 2,660 weekday riders in February, while Link was carrying about 21,774 in May.

        So, Link was averaging around 8 times as many weekday boardings as SWIFT. But Link cost about 87 times as much as SWIFT!

        At $30 million, SWIFT cost about $11,280 per weekday boarding.
        At $2.6 BILLION, Link cost about $119,408 per weekday boarding.

        So, Link’s capital cost was about TEN TIMES as much per weekday boarding as SWIFT.

        I say that SWIFT is good enough if it carries 1/8 as many people as Link for 1/87 the capital cost of Link.

      12. In my view, “good enough” buses don’t mean that we can’t plan and built even better trains and buses. Expensive in the short term, but faster and can move far more people per operator. I’d love to see more services like SWIFT as well.

      13. Except they have no level boarding which makes a huge difference for those with luggage, strollers, bikes, and especially people in wheelchairs.

      14. Not true for bikes. luggage or strollers. They all board SWIFT buses easily. For wheelchairs there is a ramp.

      15. Never as easy as a Link train. No wide gaps to cross, no height differentials, no ramps, on all 4 double-stream doors. Link can load multiple wheelchairs, bikes and people in the same time it takes to just deploy and stow one ramp.

      16. SWIFT stops are already undergoing construction to install bumpers because the level boarding has been a problem. While this cost is still not as much as light rail, you were inaccurate to claim that SWIFT does not have a problem in this area.

      17. It’s not just SWIFT buses. People regularly bring strollers, and luggage on Metro buses with no problems. Bikes do not go inside buses, other thsn SWIFT buses, and nobody has any problem getting a bike on a SWIFT bus. Saying there is a “huge difference” is nonsense. People do this all the time on buses all over the city.

      18. How long does it take to deploy and stow a ramp on SWIFT? About 20 seconds? That is about the shortest dwell time at any Link station.

        On SWIFT buses, there are 2 other doors where people can enter and exit while the wheelchair is using the ramp at the third door. I would imagine that a stop for a SWIFT bus where a wheelchair boarded or exited would not take much longer than 20 seconds. If you have information that deploying and stowing a wheelchair ramp on a SWIFT bus takes a lot longer than that, please share it.


      19. Using your numbers that means at least 20 seconds of dead time for Swift’s front door where it is impassable. It doesn’t include the time for the wheelchair user to roll on and secure him/herself correctly or be secured by the driver which takes much longer. In the same time, a wheelchair or up to 4 wheelchairs would’ve rolled on to Link and gotten him/herself in place with time to spare.

      20. 20 seconds of dead time for one door on SWIFT is inconsequential. Twenty seconds is easily enough time for everyone else on the bus who wants to exit or board to do so through the other 2 doors.

        I would say the average dwell time for Link at Westlake is around 40 seconds, southbound. Link trains often sit at stations for well over 20 seconds, sometimes to wait for “runners” to make it onto the train.

        Even if a wheelchair added a few seconds to a SWIFT stop, so what? How many wheelchair boardings are there on average on each SWIFT trip? You are whining about something that is not even a factor.

        By the way, wheelchairs on SWIFT can probably get in position and be ready to go by the time the ramp is stowed, most of the time. SWIFT drivers have written that virtually no wheelchair users ask drivers to secure them — they all use the passive system.

      21. Regardless, wheelchair boarding is much easier and faster on Link than on any bus. It can be done simultaneously for up to 4 users without impact to normal dwell time (20 seconds).

        SWIFT drivers have written that virtually no wheelchair users ask drivers to secure them — they all use the passive system.

        Good to hear that. Say it to Jeff Welch’s face for me please, three times.

      22. I just timed a motorized wheelchair user board a brand new Metro bus (#6854 low floor 60′) on the 106. 40 seconds from the first beeps of deployment to when the ramp is completely stowed. Securement happened after stowage (30 s).

        When that guy exited: 10.2 s deploy, 5.7 s exit, 15.3 s stow (asap), total 36.2 seconds not including door open and close time.

        Yor numbers are way off.

    3. Overhead racks on trains would be a nice addition. I see them in nearly every interior photo of urban Japanese trains. It won’t block advertising and adds capacity.

    4. I bet there are a lot of Sea-Tac Airport passengers that ride Link, that wouldn’t have otherwise ridden the 194 or 174. This is probably especially true for out-of-town people who trust fixed guideway transit over regular bus routes. I know this is anecdotal, but there are quite of few people in my office that have given up driving to the airport (and parking) for Link…especially for short business trips.

      1. I’m one of those people who only rode the 174 once, and never again. I cadged rides, used Shuttle Express or a car service until last September.

      2. My cousin and her husband were two of them. They rode from the airport to Othello where his son lives. I didn’t even know about it until afterward, as I didn’t know where he lived. They would not have ridden a bus for fear of getting lost or having to wait an hour somewhere. And there was no bus from Sea-Tac to Raniner before Link, so they would have had to go downtown and transfer.

    5. Amateur psychoanalysis of a photo of LINK passengers seems a little inappropriate, but if they look a little glum, maybe we should assume they’re tired after a long flight.

    6. Norman,

      Link more or less replaced routes 42 and 194, and has about double the ridership. There are your people who won’t ride buses. Anecdotally, if you’ve never met anyone like that you don’t get out enough.

      As regards capacity, your constant exertions to somehow reduce the capacity of a Link car are well documented. Of course, cargoes of all kinds also impact the capacity of other modes of transportation, too.

      Meanwhile, in all the highway-vs.-rail comparisons, you’ve never once turned your marvelously skeptical mind to the car capacity estimates, which assume perfect car operation. One person drives like a moron, and capacity collapses. A few more cars that optimal enter the highway, and capacity collapses. But for some reason you aren’t interested in that. How come?

      If you really think anyone is convinced that rail was a mistake because the riders in this photo aren’t smiling you are absolutely delusional.

      1. And there are 194 and 42 trips not served by Link, and it’s actually a bit more than double. What part of “more or less” did you not understand?

      2. I didn’t understand, for example, you leaving out between about 10% and 15% of all boardings on Link (2,000 to 3,000 boardings per day in May) are trips which never leave the bus tunnel. So, are you saying that only the 42 and 194 buses could have taken those riders on trips within the bus tunnel? None of the people who ride Link between Westlake and International, for example, would be riding on of the buses still operating in the tunnel, if Link were not there?

        I didn’t understand you leaving out the 174. People used to ride the 174 all the want into downtown Seattle. Now at least some those people are forced to get off the 174 at the Tukwila Link station, and transfer to Link. Are you claiming that Link is not carrying any passengers who used to ride the 174 all the way into Seattle?

        What about all the other bus routes that now terminate at Link stations, forcing bus riders to transfer to Link to complete their trips. I don’t understand how you can ignore all those bus routes.

        Why don’t you list all the bus routes which have been changed to “feed” Link stations, instead of taking bus riders to their destinations on buses? I don’t understand why you neglect to mention all those former bus riders who now have to take part of their trip on Link.

        Maybe you could explain these omissions to me.

    7. The gentleman wearing the light blue jacket, facing away from the camera, refuses to ride buses. He will however, as evidence in this photograph, ride light rail.

    8. Ah, if I can make a cultural observation… Seattle has strong historical ties to Scandinavian immigrants whose culture I observe to be… staid… ( Characterized by sedate dignity and often a strait-laced sense of propriety; sober. See Synonyms at serious.)

      1. So right on the spot…ride the Tunelbana in Stockholm in February and you will see the frowniest people in the world, who I’m pretty sure love their trains (and will pay a pretty big premium to live within a few blocks of a station).

      2. Haha. Indeed, it’s true. (Native Seattleite here, partially descended from Ballard Scandinavians who came here a century ago.) It explains a lot about Seattle, including our habit of avoiding jaywalking!

    9. 1) Have you seen people driving in their cars on the way to work? How about those on the bus? They are all pretty glum too.

      2) Please also subtract the same figure for buses.

      3) Unfortunately I don’t have the power to recognize people’s preferences just by looking at a picture of them. Either you have that superpower or believe that people that look the same are the same. Either one is an interesting piece of information.


      1. In other words, you can’t see anyone on this Link train who looks different from bus passengers, either.

    10. Out of all the baseless and ridiculous reasoning that you have given for a one year old rail system’s “failure,” This is probably the most ridiculous. Come on, people riding link don’t look happy? Yeah, compared to all those people driving on I-5 who are just ecstatic.

  6. http://communitytransit.blogspot.com/2010/03/swift-vanpool-ridership-rise-as-overall.html

    Swift, Highway 99 ridership
    · Nov. 29, 2009, Swift begins service with 1,523 boardings that day
    · December , 1,699 average weekday boardings
    · January, 2,367 average weekday boardings
    · February, 2,660 average weekday boardings

    Does anyone have any more-recent ridership numbers for the SWIFT bus route? This is all I can find. It shows that ridership on SWIFT buses rose from 1,699 in December ’09 to 2,660 in February of 2010 — an increase of 963, or 57% in two months.

    Are SWIFT ridership numbers from more-recent months available? If not, does anyone here ride SWIFT regularly, and have any sense of whether ridership has increased since February, or not?

  7. Very Nice Gains the last several months. Could STB post the station boardings (ons/offs)for the same month? It would be interesting to see where the the riders are coming from and sure to spark lots of discussion over WHY.

    1. The “where” would be very interesting. I would think it would be good for planning too.

  8. Someone check my math but I think there are about 233 Link trips on a weekday. Divided into the ridership numbers you come out with 93 people per train over a weekday service period. For the first quarter ST only reported 61 boardings per trip but ridership is up almost 50%, ~14k to 21k and they don’t breakout weekday service so this makes sense. Obviously ridership isn’t constant during the day but neighter is is the frequency of service so, if things are being efficiently run the passenger loads should be relatively balanced. The other factor is not everyone is riding end to end so the same seat might represent two or more riders on the same trip. I guess we’ll have to wait for a report of passenger miles before we can know what value to assign to a boarding. For example someone parking at TIB instead of Wally Park is pretty low in societal value.

    One thing I can’t figure out is the ratio of “Boardings per Revenue Vehicle Hour” vs “Boardings per Trip”. If we approximate a trip at 1/2 an hour I’d expect the per vehicle hour number to be roughly twice the boardings per trip. But in the Q1 ridership report it’s almost the reverse 39:61. Are they counting each train as two vehicles? That’s sort of what it looks like from the Sounder numbers but it’s still hard to reconcile.

    1. I would guess that they’re counting each train as two vehicles. Also note that the trains lay over at each end of the line, so I think it’s really more like one trip per hour, not two.

      1. I thought about that but layover time counts toward platform hours not revenue hours which is what they specifically call out.

  9. Link light rail is VERY COOL and getting cooler by the month. The trends are pointing up. I will bet Norman a half case of (warm) Old English 900 24 ouncers that this positive trend will continue for one more year, at least…

  10. If you look at the raw data chart, ridership increases as you get to the later dates, which should mean more good news to come.

  11. Anecdote: two coworkers and I were talking about getting to the airport. They’ve both taken Link now and were familiar with DC Metro, BART, etc from various trips. However, neither had ever taken the 194 bus. They had no real explanation, just something about never sure which one to take. It’s a weird thing, but I think people have some sort of irrational fear that a bus will suddenly switch routes or something but a train seems more reliable and familiar.

    1. Lots of buses show up at the terminal. Buses go lot’s of places which is actually a big plus. But people see a train and immediately realize, hey… that goes DT. OTOH, my wife, who actually flies a lot on business, just noticed, “hey, there’s a bus that that goes directly to Bellevue.”

      1. I’ve spoken to a number of people that have expressed a strong disdain for buses. I’ve heard comments like, “not safe”, “it’s for poor people”, “too slow”, “dirty, smelly”. Some of these people would consider taking a train.

        It’s ironic that where I have been living in Chicago, the buses are often cleaner than the trains and at least in the part of the city where I was living it was much more convenient to take an express bus than the train. Gets you downtown in half the time. It’s also interesting that in some cases, its much quicker to take the train for shorter distances than the bus.

      2. “I’ve spoken to a number of people that have expressed a strong disdain for buses. I’ve heard comments like, “not safe”, “it’s for poor people”, “too slow”, “dirty, smelly”. Some of these people would consider taking a train.”

        This is just the sort of ignorance that bus haters try to perpetuate.

      3. But… buses *are* too slow. Generally.

        They ride in the same traffic that cars do, but have to make frequent stops. If you’re gonna be stuck in traffic anyway, to a lot of people, you might as well do it in your own car and have more comfort and control.

        This is not rocket science… it’s a logical response to the way buses work around here.

        Now, proper BRT could go a long way toward changing those beliefs, but we aren’t really going to get proper BRT in Seattle, as far as I can tell.

        (Incidentally — one of the reasons I dislike riding most buses is that the frequent stops and starts give me motion sickness. I don’t know why — it didn’t happen when I was young, but started as an adult. I was getting nauseous on every bus ride to/from work. It got old fast. It doesn’t happen on trains — or long-distance buses — so I will choose a train whenever possible. Buses were a lot more enjoyable when I could ride them regularly without wanting to puke. Wish I knew why it happened.)

  12. Let’s put these numbers in perspective: Vancouver’s Canada Line, which connects downtown Vancouver to YVR airport, is 14 miles long. This is the same length as Central Link. The Canada Line is pushing 100,000 daily riders, while we are celebrating 22,000.

    You might say that’s not a fair comparison, that Vancouver is denser than we are and they don’t have a freeway, or that our line is challenged by the fact that there’s a giant dead zone between Rainier Beach and Tukwilla. All these excuses are just that: excuses. We picked the corridor; we picked the technology; we picked the station locations and we set the zoning criteria.

    By the way, the Canada Line traverses about 4 miles worth of single-family zoning along Cambie St. Those 100,000 daily riders are pre-TOD. By the time we have crawled our way up to ST’s original ridership projections of 42,000 daily boardings, the Canada line will probably be up to 200,000 assuming that Vancouver embraces TOD as aggressively as they have on all their other Skytrain lines.

    This is why I call Central Link a failure, because compared to our neighbors to the north, it is. Vancouver is a convenient comparison because they are so close, but by global standards, Vancouver is a sprawling mess. Hop across the pond to Copenhagen (1.9 million people) or Stockholm (2 million people) and you can really see what a successful transit system in a mid-sized city looks like.

    We need to be setting our sites higher. We can make excuses, but the main reason that we have a light rail line that has only 22,000 daily riders is because we’ve lowered our standards so far that we actually think that’s good.

    1. What lazy dishonest comparisons you make. Your 42K number is a prediction for 2020. Densities for Vancouver and Seattle are apples and oranges. As you point out Copenhagen and Stockholm are several times larger and more dense than Seattle. And they’ve had over 100 years head start on us.

      1. Well, here’s a surprise for you.

        [Source: Wikipedia]

        Population (July 1, 2009)”Latest estimate: Seattle has 617,334 people”. Puget Sound Business Journal. http://www.bizjournals.com/seattle/stories/2010/06/21/daily9.html?ana=e_du_pap. Retrieved 2010-06-22. [1][2][3]
        – City 617,334 (US: 25th)
        – Density 7,136/sq mi (2,755.2/km2)
        – Urban 2,712,205
        – Metro 3,407,848 (US: 15th)


        Population (2010)[2]
        – City 530,902 (04-01-2,010)
        – Density 6,016/km2 (15,581.4/sq mi)
        – Urban 1,181,239 (01-01-2,010)
        – Metro 1,899,427 ((04-01-2,010) 34 closest municipalities)


        Population (2009-12-31)[2]
        – City 829,417
        – Density 4,411.8/km2 (11,426.5/sq mi)
        – Urban 1,252,020
        – Urban Density 3,318.4/km2 (8,594.5/sq mi)
        – Metro 2,019,182
        – Metro Density 309.7/km2 (802.2/sq mi)

        So, your figures for Stockholm and Copenhagen are actually their Metro populations. The actual city population of Copenhagen is almost 100,000 LESS than Seattle’s population. Seattle’s Metro population is OVER 1.4 million more than than Stockholm’s. Stockholm’s city population is about on par with Detroit or San Francisco. You will notice that the densities of the in city populations of Copenhagen and Stockholm are about double that of Seattle.

        Except for cities like New York, Boston, Baltimore or San Francisco, you will not find European style densities in American Cities. The die is cast and while there are trends towards increasing densities in American cities, it will be generations before more American cities begin to approach the European pattern. Indeed, in most American cities, the demographic trend is the reverse of Europe. In many American cities, the wealthy live in the burbs and the poor live in the inner city. In many cities in the world, the reverse is true.

        I think it is too harsh to call the Central Link Light Rail a “failure”. It is a starter line with some ambitious goals. One, it got done. It isn’t perfect but almost nothing worth doing is perfect. It expresses the goal of changing commute patterns. It serves an area that for generations was the forgotten quarter of the city and creates opportunities for renewal, and transit oriented development. It unleashes vast creativity in people to think up new possibilities — hence the activity on this blog and I might say a very receptive management team at Sound Transit that takes the time to interact with enthusiasts/transit nerds (I dare say it has been the involvement of this community that impacted the alignment choices for EastLink and is presently influencing the Trolley Bus decisions in Seattle.)

        Vancouver BC has always had more power to shape it’s design. The very fact that freeways end on the edge of the city (Well, Trans Canada 1 does sorta skirt the North Vancouver area) makes that city/metro very ripe for Rail rapid Transit. Seattle, as have most American cities has been a car oriented city. It had the benefit of a better than average bus system and a populace with a liberal mindset willing to consider changes in lifestyles.

        It is a good thing that we have both Vancouver and Portland nearby to compare and to “vie” with. The success of the Canada line is something to study. I see the population that the Canada line serves as similar in composition to Southwest King County e.g. Federal Way, De Moines, Burien, Highline etc.

        So, here’s to a glass half full.

      2. Most of the Canada Line ridership has come from people who were already riding the bus, very little of it is new ridership. Every express bus from south of the Fraser River has been truncated at Bridgeport Station. Not only that, but the 98 B-Line, which the Canada Line replaced, had nearly 20,000 boardings per day by itself. I’m sure if we truncated every bus from South King County and Pierce County at TIB and every bus in the valley at Mt. Baker we’d have some pretty impressive ridership numbers on Central Link too.

      3. Well, if there was the duwamish bypass then if transferring riders to Link would actually save time then I’d be all for it. But it’s doubtful that either would occur.

      4. In the case of the Canada Line people weren’t given a choice, TransLink just did it. They knew they needed to leverage their investment in the Canada Line and that in the long run transit service would be improved because the truncations allowed increased bus service (increased frequency) in the Richmond and Delta areas.

        BTW, the Canada Line is actually slower than Central Link. The 9 miles from Richmond to Waterfront takes 25 minutes, for an average of 21.5mph. Link takes 36 minutes to cover the 15.6 miles from SeaTac to Westlake, for an average of 26mph. The Canada Line averages over 100,000 daily boardings, even though it is slower than Link and requires a bus transfer for most riders. I think it’s obvious that it’s more than travel time that influences ridership, and the real value of a “Duwamish Bypass” is dubious.

      5. “Duwamish Bypass” isn’t going to happen. Not in the next several decades or so. We’re stuck with what we have. Maybe it was the right choice, maybe not. A learning experience (hopefully) either way. The delay is something we have to deal with. Forget extending farther south? Create express bypass trains? Whatever, it has to be a reaction to what we have and how that works.

      6. “Forget extending farther south?”

        Maybe, but then you have to start thinking about the people who will be served by light rail in south King County and north Pierce County whose destination isn’t downtown Seattle. In my opinion, downtown Seattle to S. 272nd is about the limit for what you can realistically do with light rail. Beyond that it might be better to start looking at how to make it easier for people to get to the Sounder stations in the valley. Sounder travel times beat driving and light rail at any time of the day.

      7. Actually, I’d like to see a station at about S 133rd – so the 150 could end there.

      8. Oh – and for extending south: I believe it would be a good idea for better serving a Tacoma commuter market from South King County. Just sayin’.

      9. Don’t forget that Vancouver has a very frequent grid bus route network serving all of the arterials between False Creek and the Fraser and the Expo Line and UBC. These routes all act to feed the Canada Line (and the Expo line as well). In addition there are significant employment centers located along the Canada Line Corridor particularly at Broadway.

        As you point out another factor is Translink got rid of all competing transit service. I’m sure you are right that central link ridership would be much higher if every route going between Downtown Seattle SE Seattle and S. King County forced a transfer to Link.

      10. I think the other big issue is network effects. The Canada Line connects into a broader network which radically increases the number of destinations. Central Link will look much better when it plugs into more important stuff.

      11. “I think the other big issue is network effects. The Canada Line connects into a broader network which radically increases the number of destinations. Central Link will look much better when it plugs into more important stuff.”

        I agree. Canada Line is an addition to a bigger, more useful existing system. Central Link doesn’t have a wider rail network to plug into. It’s an apples/oranges comparison.

    2. A Vancouverite told me nobody gets on or off at the Cambie stops; they’re mostly coming from further away.

Comments are closed.