Avgeek Joe/Flickr

University Link is coming in roughly one quarter. Expectations are high, not only for the revolution in travel time it will bring, but for the sheer volume of people expected to use the service. Sound Transit’s draft 2016 Service Implementation Plan (p. 122) projects average weekday boardings of 34,600 this year jumping up to 51,800 in 2016 and 60,500 in 2017, the first full year of service. In subsequent years, it will grow to 65,600, 68,400, 70,800, and finally 79,800 in 2021, the year when more North Link stations will open.

This is as good a time as any to remind everyone that comparing actual riders to projected riders proves nothing about if the model was valid, nor does it convince any real voter that their vote was or wasn’t justified.

The principal things that drive absolute ridership are the things that drive ridership systemwide: job growth, population growth, zoning decisions, fare levels, and the costs of driving (gas, tolls, parking). Transit consultants are not particularly qualified to predict many of these items, which will unfortunately overwhelm the technical merits of the line itself.

Moreover, as a rhetorical device appeals to projections utterly fail. Approximately zero people who voted for Central Link in the hopes for 40,000 a day would have changed their vote if the number was 32,000. Similarly, no light rail opponents suddenly become converts at 50,000. Within an order of magnitude, a project is useful and high-quality enough for a certain rough level of taxation or it isn’t. Tweaks to the ridership models or tax micro-concessions don’t change the politics.

What models are useful for is comparisons between projects, where the imprecisions from secular trends cancel each other out. Regrettably for pundits, there is no experiment where two versions of Seattle build different projects. At best, the region might build more than one of the possible projects, in which case network effects still complicate any assessment.

Going on about projections after they’ve completed their usefulness in alternatives analysis is a fun parlor game, but do not mistake it for thorough analysis or effective messaging.

72 Replies to “Ridership Projection Uses and Abuses”

  1. Pretty much like tunneling, isn’t it? Just as the most skilled engineers and crewmen can’t see through the ground, nobody can really see the future in ahead of either the boring machines or the project itself.

    Over the ages, the little moles have learned, and developed physically, to win a contest, although trains would have to be much smaller. Race would be great, wouldn’t it?

    But mainly because if Bertha or Brenda get stopped or wrecked, the attorneys would be busier than Medic One.

    While moles have never had either lawyers or first-responders. But for Government, tunnel engineering, and moles, there’s one ironclad rule: Learn from experience or die.

    Mark Dublin

  2. Rail ridership projections also depend on a number of assumptions about what will or will not happen to the bus network, which can sometimes radically make a difference. Consider the 130th St. Station, for instance. Make ridership projections under the assumption that the station would be served by a truncated route 522 serving Lake City->Station->Bitter Lake, and it looks good. Make ridership projections under the assumption that the current bus network is frozen in place forever, and the 130th St. station ridership projections look not so good. Not surprisingly, ST’s policies require that station ridership estimates not be based on “speculative” changes to the bus network, so the 130th St. station appears, at least on paper, to underperform.

    1. I agree. This is why the bus people need to talk to the train people as they are building the system. The ridership projections as well as simply the effectiveness of the system should be done with the cooperation of the bus agency (or agencies) while planning. This does make it a lot more complicated, but it gets you much better estimates.

      After all, you should be looking at overall transit ridership, not just rail ridership. Actually, you should consider all potential rides (including new ones as well as improved ones), but looking at overall transit ridership (increases) is a decent place to start.

    2. @assf,

      It is not ST policy that “speculative changes” to things like the bus network be excluded, it is in fact the Federal policy. The Feds usually disallow the inclusion of such things in ridership projections. I say “usually” because their are always special exceptions.

      The intent is fairly obvious, to avoid cooking of the books by local jurisdictions in their quest for Federal dollars. It is hard enough to compare various projects against each other with a common baseline, it would be impossible if the Feds let every jurisdiction choose their own baseline.

      So, ya, no cooking of the books. At least not if you intend to apply for Federal dollars.

      1. It’s perfectly reasonable to include changes to the bus network in your planning. You just have to get Metro to agree that they will make those changes if the station is built.

        Most cities don’t have this problem because *the bus and rail are part of the same agency* in most cities.

      2. Yep, exactly Nathanael. Thanks for clearly explaining something that should be obvious to someone who thinks for three seconds before making a comment.

        “Cooking the books?” You gotta be kidding me.

      3. If Sound Transit isn’t allowed to use “speculative” bus changes involving other agencies where do the crazy numbers projected for Lynnwood come from? Surely ST doesn’t believe ST Express service will be able to provide that level of feed? Those passengers for the most part won’t be walking to the station and there is only so much parking.

  3. One thing they missed is the projections for the 74x. Once the 71x,72x and 73x cease to exist the 74x, which is currently a trip of hang limbs out windows at own risk to ensure space, will have people bus top surfing for space once U-Link opens. Current 74 riders won’t relinquish there 1 seat ride and others will want to avoid the Husky stadium transfer to get to the ave.

    Let the fun begin.

    1. If LINK performs as it should, word should get around comparative speed with the 74. Regardless of traffic or weather. As a career trolleycoach driver, affection for my old trade could very well keep me aboard the 70. But it would take a powerful attachment to the scenic old Eastlake neighborhood to choose it over LINK.

      I should be careful, though, about constantly referring to Night Stalker episodes about the mad. blood-drinking Civil War doctor in the Pioneer Square underground. Only ones who need to worry are really cute girls-meaning all our lady passengers-and also hero reporter Karl Kolchak’s editor at The Seattle Times.

      So LINK will probably be ok, except in this case, the Editor is the one who’ll believe underground monsters. Also plesiosaurs biting the I-90 bridge in half.

      However, for trains and buses both, rooftop riders, including humans, goats, and chickens will add a wonderful cross-cultural element to the trip. Little charcoal-burning stoves in the aisles will also beat coin machines hands-down for curry.

      Unfortunately though, King County Medical Examiner’s office will have to face the fact that the regional electric rail network in Mumbai needs its own morgue.

      http://www.amazon.com/The-Lunatic-Express-Discovering-Dangerous/dp/0767929810

      After experiencing the world’s most dangerous transportation, which vast majority of world’s people use, author’s only completely wretched experience was Greyhound from SF to NYC. Ridership projections, positive or negative, are too awful to think about.

      Mark

      1. I doubt it; why would anybody give up a 1 seat ride for a detoured 2 seat ride that most likely will not save time. Chances are it will increase time once the alternative bus has to slog back up to the ave.

      2. @les,

        They will give up their one seat ride because the 2 seat ride on Link will be much faster and much more reliable. And with low Link head ways the transfer will be relatively painless.

        The opening of U-Link will be a seismic shift. Old ways of thinking are about to die.

      3. Cute girls learned to love commuter rail and fear no scary urban monsters via SEPTA and Reading RR “Shoppers Special” reduced off-peak fares. Best way to help UW Link reality satusfy the bean counters seems to be using it…. chickens and goats, more the merrier

      4. Trip Planner has 15 mins for 74x from Westlake to U-district with its nice clear express lane.
        ULink is listed as 8 mins. But by the time one adds 7 mins for a bus transfer and 10 mins to travel to the Ave you could easily spend 25 mins. Who is more the fool, the link rider or the 74x rider?

      5. People would take different routes depending on how far east they live. The 68, 372, and 65 will get pretty speedily to 55th, 65th, or 75th. If you catch the bus at Rainier Vista, you avoid half the campus crawl. Nobody will go up to the Ave unless they live near 15th. in which case it’s on the way anyway. The 373X will also be running. Link is also immune to traffic and road accidents, and it’s a train!

      6. @les,

        Nobody will be “shown the fool” for favoring one route over another. Such things are purely a matter of personal preference. In the morning my wife prefers the slow local to downtown over the fast express. The reason? It gives her time to talk to her mom in NJ and get mentally prepared for work. She certainly isn’t a fool, she just values other things over a fast reliable commute.

        That said though, the bulk of riders will prefer speed and reliability over a one seat ride. You did a good job cherry picking your end points, but most riders won’t be doing that. The bulk of the riders will see a faster and more reliable commute on Link, and that spells trouble for certain north end to DT routes.

        And the situation will only get worse when the buses come out of the DSTT in 2017/18 due to Convention Center construction. Expect some creative things out of ST then, and expect even more negative pressure on northend to DT bus routes.

      7. Les, that 74X is useless for people who go northbound in the morning and southbound in he afternoon. So it only serves part of the transit market.

      8. Mike Orr pretty much hit it spot on. If you live east of 25th, your route to Link would be the 65, 75, or 372, which will not go down the Ave. There are also people who live in UW dorms who could walk directly to Link faster than it would take to wait for a connecting bus and walk from the transfer point. Then, there’s the fact that even with the express lanes open, traffic volumes have increased considerably over the past few years. Today, even the express lanes now experience routine congestion from about Mercer St. to the tunnel entrance.

        Yet another scenario to consider is what will people do who are waiting on the Ave. for a #74, but see a bus headed to the UW station come by first. Would you continue waiting for the #74 that OneBusAway claims is due in 4 minutes (but might be wrong)? Knowing that when it finally comes, you’ll likely be standing and squeezing the entire ride? Or would you just hop on the shuttle bus that’s right here, right now, with empty seats available? Especially since, for the next 5 years, there will be absolutely no trouble for anyone boarding Link at the UW station to find a seat on the train.

      9. It should be noted that 74X will not be serving the AVE. Looking at the final approved routings, Route 74X will be using the Roosevelt/11th Ave NE couplet to provide some replacement service from the loss of the Route 66.

        Also, on 25th Ave NE, route 68 will no longer exist either. It will be replaced by Rt 372 (25th Ave NE) and Rerouted Route 67 (Roosevelt Way NE). My biggest concern is that Metro Staff at this time is planning to use the same stops on the 372, so some stops that exist for the 68 and 72, but not served by the 372 will lose their service. So, the 372 is not totally a replacement service for those routes, if those stops are skipped (however, if they added some stops on the 372, but also do some bus stop diets, it may work).

      10. adsf2 pretty much nailed it in his second paragraph. Right now, it is a moot point. Metro will pretty much eliminate the current set of frequent buses going from the UW to downtown, so people will be forced to take two vehicles (or walk a long ways) to get downtown. If you are on the Ave and see a bus headed to Husky Stadium, do you take it, or walk over to Roosevelt to catch a 74? My guess is most of the people will do the former. The 74 doesn’t even run after peak, so a rider won’t even have that option in the middle of the day. People will have to take the local (the 70) or transfer to Link.

        It would have been a very interesting experiment to start with a system much like it exists now and see which one people prefer. I think it would be the opposite. The buses going downtown are at least as frequent as Link and you avoid the transfer. Lots of people would take Link, but I doubt very many who are headed to upper campus or the Ave. But no one will ever know, since the era of fast, frequent bus service from the heart of the U-District (the Ave) to downtown is about to be over.

        I disagree with the other statement, though. The dorms are not very close to Husky Stadium. It is about a fifteen minute walk, not counting the time spent getting to the platform. It is a two minute walk from any of the dorms to the bus stop on Campus Parkway (where those buses stop). I’m sure some people will walk (just because people avoid taking an extra bus trip) but they won’t be saving any time (quite the opposite). If you are at the far west end of the UW hospital, then you might hop on the bus, but you will likely just walk (half mile bus rides are rare). If you are on campus anywhere near the fountain (or closer) then you will definitely walk to Link.

      11. I had been hoping for a U-Link opening by mid-February (better, the end of January). This should have provided enough time for operation of U-Link with the current bus network to reach a sort of (temporary) steady state (even better if METRO could, if it proved necessary, find enough change under the couch cushions and enough drivers to run a little local enhancement along the 44 and/or 48 routes between the “greater” U-District and the UW-Stadium station). I expect this would give us much invaluable data on the effect of U-Link on travel patterns, as well as a little bit of advance notification about how well the METRO reorganizations are going to work (I think they will be a success, but am not altogether certain).

        Alas, from what little I seem to be hearing, this won’t happen.

    2. The 74X is getting more trips for exactly the reason les says. “8/9 trips” will go to “10/11 trips” in March. If more than that is needed, Metro doubtless kept a few hours in reserve in case demand is higher than expected somewhere. I had always assumed the 74X was little used because the 30 is, but Metro’s stats show it to be one of the highest-volume peak-only routes

      The argument for keeping the 74 and 76 is that they avoid possibly degrading service until Roosevelt Station and U-District Station open. Since Metro has chosen to keep them, it only makes sense to increase their service to meet demand.

      1. The retained but amputated route 71 will add an extra bus coming eastbound from 15th :-( … But hopefully short lived when Wedgwood sees 71 goes east to go west to go east again to the train anyway. Whereas the 65 is doubly frequent and direct.

        Students are telling me that the campus Rainier Vista stop coming sooner than March even, but the slight walk back from Garfield Pl to Rainier Vista is inconsequential .

        Walk the Rainier Vista to and from route 65 now to UWMC

        Mike Orr’s point about train bypassing I5 issues was clear to route 64X riders the day the latest expansion joint popped

      2. That was my take, Mike, until I remembered what Warren mentioned. The 74 won’t go on the Ave. The 76 won’t go to the U-District. Other than the 70, there will be no one seat ride from the Ave to downtown. Other than the 70 and 49, there will be no one seat ride from Campus Parkway to the downtown.

        There is a cut off point where taking a shuttle to a train makes more sense. This cut off point has been moved very far north because of the restructure. For many people, there really will be no choice between a fast one seat ride (which exists now) versus a shuttle bus to a train. Either they have to walk a very long ways to the fast bus (the 74, which doesn’t run during the middle of the day) or they take the shuttle (or walk) to the train. If you are by any of the dorms right now (between Campus Parkway and 40th) and wanted to go downtown, you would walk less than two blocks to the bus stop. That distance will go up considerably, adding about five minutes to the walk. What is true for the dorms is true for folks on the Ave or upper campus.

        The 74 will pick up additional riders because the old 66 is going away. Along with that, there won’t be express buses going down the Ave. So someone on Roosevelt who used to wait for the 66 will now take the 74. Someone around 12th, who used to walk over to the Ave and catch an express bus to downtown will probably walk the opposite direction, to the 74. These numbers will be significant, but not huge. Nothing like the number of people who would take, say, the 73 if it was still running as an express under its current routing.

        The 76 is pretty much irrelevant when it comes to U-Link. I don’t think anyone thought people on 65th and Roosevelt would be forced to shuttle to Link any more than people on Northgate would be forced to shuttle. They didn’t change the 41, so I’m not surprised they didn’t change the 76. Service was added just because they had a little extra money. Frequency was added (I’m guessing) mainly because of the section close to Roosevelt, which I assume has very high ridership now. I think you could make the argument for increasing frequency on the 41 which I know has huge ridership, but I think since that area didn’t change much as part of this restructure, they didn’t add service there. There may be some politics involved here (those who have to learn new bus routes have increased frequency).

        The one bus that I think might get a bit crowded is the 70. It will run more frequently during rush hour, but I think midday is when it will be a bit more popular. During rush hour, this bus is pretty slow. It is much slower than an express (e. g. 73X). Meanwhile, Link will be running very frequently. For many, the fast route downtown will be to take a shuttle bus to Link. But during the middle of the day, things reverse themselves. Link will not be very frequent, making the transfer that much more time consuming. Traffic isn’t so terrible, so the 70 can get downtown at a decent clip. If they ran the 70 every 8 minutes or so in the middle of the day, I think it would be very popular. My guess is that it will be popular, but simply be a “first bus” bus. If the 70 is expected along any minute, then you take it. But if not, you take one of the other buses headed to Husky Stadium, cross the street, walk down the stairs and wait for the train.

      3. “But during the middle of the day, things reverse themselves. Link will not be very frequent,”

        It will still be more frequent than the 70 or any other route except the 49. 8-10 minute headways mean 4-5 minute average wait time.

      4. Right Mike, but with the 70, you don’t need a transfer. That is my point. Imagine you are in the U-District headed downtown and you see a 70 and a 45. At rush hour, the 70 will be bogged down getting to downtown. But the 45 has a bit of extra walking (from the bus stop to the station). There is also the wait for the train. But this is tiny (on the order of a couple minutes). So taking the 45 seems like the right choice.

        In the middle of the day this reverses itself. The 70 will run a lot faster, and the time spent waiting for the train goes way up. I’m guessing a lot more people will take the one seat ride then. But I agree that if you see a 45 by itself, you will probably take it, since the 70 won’t be along that often

    3. “why would anybody give up a 1 seat ride for a detoured 2 seat ride that most likely will not save time.”

      Some people will; some people won’t, We can assume that half the people will; that’s probably more accurate than assuming nobody will. Two different routes will also increase total ridership, because everybody lives in a particular location so many houses north or south of the existing stops, and some people will find the new route more convenient than the old one. We don’t have to spend a lot of time worrying whether Link will save a few minutes or not, because that’s not the only factor people base their choices on. Some people don’t like a long one-seat slog from downtown, through intersections that are always prone to backups and stoplights; they’re only doing it because there’s currently no better transit option.

      1. If you assume the second clause to be true (most likely will not save time) then I think the first one follows (no one will take a two seat ride), with one exception. The big question is whether it will or won’t save time. In some cases, yes, in some cases no. But as I said above, because of the restructure, the dividing line has been moved far to the north. To make matters more interesting, the only fast one seat ride to downtown is the 74, which is made even faster. The express buses that traveled on the Ave were fast once they got on the freeway, but getting to the freeway was not especially fast. The bus looped around a bit. Now that loop is gone, and getting from say, 45th and Roosevelt to downtown will be very fast (south on Roosevelt, right on 42nd, get on the express lanes).

        Which gets me to the exception. It depends on how far people are willing to walk. If you are on Roosevelt, I think you take the 74 without question. But if you are on the Ave, I don’t know if you want to walk four blocks, even if it is faster. That is where personal preference comes in.

        But if there is no difference in walking and total time than I think the number of people who prefer a two seat ride is very, very small. The opposite is true, though. There are plenty of people who stay on a bus to avoid a transfer, even if the transfer would save them five minutes.

        The big question is how much time is saved. I think it varies quite a bit, day by day, but in general, I think if it is faster to get to the 74 than a shuttle bus, then it is faster to take the 74. But frequency matters and reliability matters. The shuttle bus and Link should be a lot more reliable.

        As mentioned above, it is largely a moot point. The 74 will make 11 trips a day. That is a very small number. People will take Link because for the most part, they really have no other choice. Well, except the the 70, but that bus is very slow.

      2. “’why would anybody give up a 1 seat ride for a detoured 2 seat ride that most likely will not save time.’

        Some people will; some people won’t, We can assume that half the people will; that’s probably more accurate than assuming nobody will.”

        That’s an interesting point, Mike. I’m thinking of options like that as well when Link opens; currently my commute on the 11 in the evening is generally jammed and of course has to stop at every single stop to board/deboard passengers (I don’t believe the ones I’m normally on ever pass a stop until – sometimes – 33rd). It not rarely can take 45 minutes from 4th to Madison Park and is uncomfortable to boot for most of that trip. I’ve walked it occasionally and taken about the same amount of time, hill and all.

        When Link opens I may well take the train up the hill, which will be a much faster and more comfortable ride, and then transfer to the 11 (or 8 if I feel like walking the last mile). I may have to wait a bit, but the most unpleasant and slow part of the ride will be skipped. There’s also the opportunity to eat/shop etc. on the Hill before catching the onward bus.

        The 11 is fine in the mornings and I doubt I would ever change to the train then, but I can see doing it every evening despite having a one seat ride on the bus.

  4. The article is correct as to “real” voter concerns. When it comes to transit, it’s *options* not specific head counts that we need to produce, and that’s what produces actual public good.

    “Unreal” voters and naysayers on transit will always hide behind nit-picking things like actual head counts vs projections

  5. I generally agree that ridership forecasts are sensationalized. I think there is a collective blame to place here. I offer a solution.

    Transit ridership is highly variable because it is only a small percentage of overall daily travel. Further, most rail system usage can vary widely because of the feeder or competing transit network, the modes of access, the reliability of mixed-flow transit options, the power of employee subsidies for either fares or for parking and all sorts of other complex variables. That’s in addition to land use assumptions, projected roadway congestion and ease of parking near stations as well as standard fixed items like base fares, average parking rates and household sizes! Unfortunately, some small group of people decades ago decided that average conditions are enough for public consumption and grant applications. Rather than study variability, modelers keep introduce new ways to make the average mathematics more complex; the new PSRC model is an extreme example that continues this obsession of averages.

    It was suggested in an ST audit a few years ago that ridership forecasting should be presented to be more variable than everyone presents it. For comparison, we are fully accepting of 70 percent chance of rain for tonight but we’re never told that ridership has a 70 percent chance of being attained in 2017! Given the data now available on things like riders using Orca cards, on traffic volumes, and on inputs like work-at-home behavior that varies by day, we are constantly proving that demand can vary 10 to 15 percent from one day to the next!

    The forecasters need to mature their perspective and basic documentation to include variability in discussing projections, like every other scientific forecasting done like weather, medicine, flooding and even earthquake likelihoods. (At least, the recent ST3 studies are now presenting ranges but even these ranges are pretty small.) How?

    – First, it’s going to take repeated pressure to retrain forecasters, managers, elected officials and the public to ask and answer “what’s the probability of reaching that” and to expect that from all forecasts.

    – Second, it’s going to take an expectation of “full disclosure” (date prepared and summary data) of the land use forecasts, the roadway and nearby bus service assumptions and other behavioral data.

    One linguistic baby step is to state “50-percent chance of minimum daily ridership” rather than “average ridership”. It may sound silly or awkward, but it’s a more honest portrayal of what forecasts mean.

    1. Another really useful measure would be to have transit-knowledgeable people ride the system. For detailed numerical counts, I think buses have had passenger counters for a long time.

      But a person can note not only when and where passengers start to act uncomfortable. And assess passengers’ general mood sense of how the ride is going.

      This would be excellent work for operators on “light duty”. Meaning injured seriously enough that they can’t perform ordinary transit work, especially driving, but are not contagious or in need of bed-rest.

      Mark Dublin

    2. Why is it worth doing this? There’s no practical advantage to mathematically exact ridership estimates. Measuring actual ridership after the fact is even more exact and real-world relevant. We only need ballpark estimates that are accurate enough to decide whether to build the line. The ballpark estimate will be somewhat close to the exact estimate; the difference is one minus the other. The issue is how large or significant the difference is. The capacity needs to accommodate at least the normal spikes (e.g,, weekday commuting on a high-use day). Whether daily ridership fluctuates below that is immaterial. At worst you’d have some empty runs. But empty runs aren’t always bad: they have to be evaluated against the benefit of transit availability (=maintaining frequency) and the savings of deleting a run (maybe little compared to disrupting an 8-hour shift).

      1. I agree. I think people forget the point of the modeling. Off the top of my head, I can think of several reasons, but here are a couple:

        1) Determine the appropriate type of service. For example, if ridership estimates show that Roosevelt HCT (https://seattletransitblog.com/2015/05/25/roosevelt-hct-is-underway/) will carry 100,000 people, then we shouldn’t build BRT. We should build light rail, even if it is surface light rail. It really doesn’t matter if it carries 80,000 or 120,000. Either way, rail would make sense if demand was that high). [Side note: streetcars wouldn’t help, because our streetcars have the same capacity as our buses]

        2) Comparing options. This includes things like routes as well as stops. This should include the entire transit network (as mentioned above — https://seattletransitblog.com/2015/11/07/ridership-projection-uses-and-abuses/#comment-658332). This allows you to spend your money wisely.

        Like most modeling, it doesn’t have to be exact to be extremely valuable. It is also remarkably accurate. I’m not sure what the models estimated for the current light rail line (Central Link) but my guess is that it was close to current ridership. I’m pretty sure it wasn’t 60,000, which is what the estimate was for just a section from Capitol Hill to Henderson Street (http://community.seattletimes.nwsource.com/archive/?date=20010525&slug=sound25m). So despite pretty much all of the factors that are difficult to predict that should boost ridership, the numbers aren’t close. So despite Seattle being the fastest growing city, despite the Rainier Valley having its share of that growth and despite airport travel growing like mad, it isn’t enough to skew the models very much. If Link was at 80,000 riders (or 20,000) I would say that the modeling sucks. But it isn’t. It is pretty much spot on, from what I can tell.

    1. Thanks for recognizing that we have three station openings this next year — Angle Lake as well as U-Link (two stations). The author doesn’t identify the upcoming Angle Lake opening as part of the growth.

      Given the adaptability and temporary nature of college students, I suspect that the UW station will attract its riders most quickly — probably within a few weeks or months. Ridership for Capitol Hill and Angle Lake will probably grow more slowly at first but they will be showing more sizeable percentage growth in 2018 to 2021.

    2. This is fascinating. I’m curious, John, if you’d be willing to spell out for posterity how you operationalize comparative measures of “significance”? The most obvious would simply be a ridership count, but I expect even you aren’t delusional enough to predict higher ridership figures. So if it’s not ridership, how do you measure significance in this prediction?

      1. Angle Lake is a unique stop for LINK for many reasons.

        It is the first absolutely non-urban stop.

        It is the first stop where parking will not be restricted by the surrounding area.

        It is the first stop that makes intraurban destinations available by rail to the nearby residents.

        It is the first stop to which a significant number of non-Seattle-residents will drive to in order to commute on.

        It is a stop more akin to a Sounder station than a traiditional LINK stop.

        It is significant because it might introduce transit to a class of people who would be unwilling to board a typical bus.

        It has the potential to reshape not just the area close to the station but in a wide radius around the station.

        It has the potential, as part of the mid-point in the Sea(ttle)-Tac(oma) poles to create a brand new center of population

      2. And after all that, the ridership at Angle Lake will still be anemic compared to the urban stops like Cap Hill, UW, and U-Dist. And some of the ridership at Angle Lake will actually be scavenged from TIBS and not new ridership at all.

        Angle Lakes best feature though? It is an end-of-the-line station, and such stations always draw well because they are the first entry point for everyone from beyond the terminus. But Angle Lake will lose that status at the next system expansion southward.

        Angle Lake will do Ok, but it won’t be a game changer

      3. Are you serious? Every bus rider who parked at Federal Way Station is going to drive the extra mile to get on here.

        Students are going to use the heck out of this.

        I can seven see UW students without much money living in Kent Valley apartments and taking advantage of the UD stations to save hundreds in rent.

        Then there’s the sports fans. For every fan who took LINK from Tukwila to Seahawks or Mariners there will now be five more who will feel safer they can get a free parking space.

      4. “Every bus rider who parked at Federal Way Station is going to drive the extra mile to get on here.”

        Extra six mile.

        It would be great if Angle Lake really takes off as you predict, and especially if a medium-density neighborhood grows up around it. Maybe you see something I don’t see since I’ve never lived in south King County. But I doubt it will have that major effect. I fail to see how driving six miles to Angle Lake is much different from driving eight miles to TIB, enough to transform south King County’s travel habits.

      5. “I can seven see UW students without much money living in Kent Valley apartments and taking advantage of the UD stations to save hundreds in rent.”

        They can already take the 180 to SeaTac or the 150 to Seattle. How is one mile from 176th to 200th going to make that much difference to them?

      6. Mike Orr, exactly. Especially since the hundreds in rent would be offset by extra hundreds in car ownership costs!

    3. UW Station will grow immediately because there are 50,000 students and staff at UW, they fill more buses than any other destination besides downtown, and their express bus routes are being cut. Capitol Hill will get an immediate boost because there’s a lot of pent-up demand: people that have waited decades for express-level transit. But the transformation won’t be as thorough because the alternatives are closer to Link’s quality. The difference between taking Link from downtown to UW is vast compared to the 71/72/73X or 43, 49, or 70. The difference between taking Link from downtown to Broadway is not as much compared to the 10, 11, 47, and 49.

      There’s another factor that will drive ridership increases too: the new one-seat rides available on Link. The 71/72/73X go only to downtown. There are no express buses from UW to Capitol Hill, Rainier Valley, or SeaTac.

      Angle Lake will be bigger than UW only in Bailo’s dreams. He keeps saying that Kent buses will be rerouted to Angle Lake, but Metro has never given any indication of that. So therefore he’ll have to drive from Kent to the Angle Lake P&R. It’s unclear that this is a big deal, or that many Kentians will follow him when they still have the option of driving to Kent Station and taking the 150, or driving to TIB and taking Link. It also requires backtracking to get from the east side of I-5 to Angle Lake Station. That could be alleviated with a new overpass, but the powers that be haven’t indicated that’s an option either. Angle Lake is of only temporary importance until Kent-Des Moines station opens. That will be much closer to Kent, and in-line for a fast BRT on KDM Road if that ever happens.

      1. Pretty much the only reason for Angle Lake Station to exist at all is that the parking at TIBS isn’t enough to meet demand. Sound Transit could have simply built the Angle Lake Parking Garage next to TIBS and kept the train terminus at the airport, and it would have done about the same amount of good.

        On weekends, the parking garage at Angle Lake is going to be virtually empty, and even on weekdays, it’s likely to not fill up more than about halfway. One interesting idea Sound Transit could use to make some use out of their investment on weekends would be to allow Friday-Sunday (or Saturday-Monday) parking in the Angle Lake parking garage for a $10 fee. Even with a $5/round trip Link ticket, it would still be much cheaper than nearby private lots, and they would only be taking up a space for one weekday at a time.

        Also, the airport->Angle Lake section of Link would probably be mostly empty, which means lots and lots of room for luggage. Everybody could simply stack their suitcases on a seat next to them and nobody would notice the impact.

      2. I think you underestimate Link for Capitol Hill. Yes, Broadway to Westlake will an improvement but not ginormously so. Everything south of there, though, is a 2 seat bus ride. University may be close enough to Westlake to just walk, but trips to pioneer square and the ID will be a dramatic improvement.

      3. It should be noted that Angle Lake Station was the terminus station in Sound Move. It would provide for more parking spaces for weekend events. My suspicion is that UW may try to get rid of the Federal Way (and perhaps the S. Renton P&R) Husky Stadium shuttles when Angle Lake Station Opens, since LINK will directly serve Husky Stadium for next year’s football games.

      4. Yeah, I agree, Mike. It should be fairly easy to figure out how many people will take Link to the UW from downtown. Start by figuring out how many people take the bus there now. I think it will be bit more that that number. Some will take it because it is faster, while some will now avoid it because it is slower (it depends on where you are headed). But I don’t think transit ridership will change that much, because even if it is a lot faster or slower, it isn’t going to play that big of a role in your decision. Bus service is generally very good right now and parking is very expensive. I just don’t see that big of a change in overall transit ridership for that trip. One exception are folks in the south end of the Central Area who take the 48 as part of a two seat ride to the UW. If you now take the 3 east, then take the 48 north, you will likely take the 3 the other direction and take Link. But this only makes sense now if you are headed to the Husky Stadium area (e. g. the UW Hospital). You could probably take the number of people who get off close to Husky Stadium and assume at most half of those people will now take Link.

        The connection between that one part of Capitol Hill and the UW is different and a bit trickier. That connection is made much faster. You can start with the number of people who take a bus to that area (by looking at the numbers for the 43). But it is also possible that a significant number of people drive that way today. Those people may now take Link. I would guess that you could take the number of riders who board in a five block radius of the station on the 43 headed towards the UW and be very close to the number of people who will now take Link between those two stops. I would maybe double that number. Add 50% for those who get out of their car, and another 50% for those who take a feeder bus (instead of the two buses it takes now — again, likely via the 48). That is probably being a bit generous, but we aren’t talking huge numbers (less than 8,000 people take the 43, and my guess is less than half ride from the Husky Stadium to Capitol Hill).

        The trip from Capitol Hill to downtown is also hard to measure. The buses that go close to the station and go downtown (10, 11, etc.) don’t go into the tunnel. The tunnel stations (that have existed since the 1990s) are very popular and have been since the 1990s. Connecting them to a new area (Seattle Central College) is a very big deal. This is where you will see some huge improvements. For example, from Pioneer Square to the college now involves two slow buses. Now it would involve one fast train. I could see the train taking away many of the riders from not only the downtown buses, but from a bus like the 60. Depending on the time of day, taking the 3, then the 60 is the fastest way to get from the south end of downtown to that part of Capitol Hill. Once CHS opens, it will never be. So you could add a few riders from the 60 (who depart close to the college) onto that number as well.

        But I don’t think that will lead to a huge increase in overall transit ridership. Those who travel between those places probably take the bus anyway. Which is why ridership is not necessarily a very good metric for judging a system. Ridership on one mode (a train) is a very bad metric, since a restructure to that mode (e. g. cancelling a popular bus route and replacing it with a slow streetcar) will not increase overall transit ridership. But even overall transit ridership is not great. Overall ridership will give you an idea if the improvements are getting people out of their car, but it won’t tell you if the system is simply better. Someone who takes the bus every day from their apartment near the station to their job in Pioneer Square (or vice versa) won’t be a new transit rider. But their trip will improve enormously. That is very valuable, and the type of thing that is worth measuring if we want to determine whether the changes make sense or not.

      5. @Warren,

        I suspect you are right about the UW wanting to get rid of the south end Husky shuttles next year – there just isn’t as much need with link running to within a 100 ft or so of Husky Stadium.

        Any idea what the ridership is on the south end shuttles. My *guess* is that Link could handle the demand pretty well, but it would be nice to know for sure. Afterall, one 4-LRV Link train is equivalent to an awful lot of buses.

      6. “I think you underestimate Link for Capitol Hill. Yes, Broadway to Westlake will an improvement but not ginormously so. Everything south of there, though, is a 2 seat bus ride.”

        True, I miss the one-seat ride on the Pine-3rd routes. It used to be 14 and 11/125, and the 7/49 evenings & Sundays. I used to take the 14 from Chinatown and Little Saigon, and the 11/125 to the library. Then the 14 was severed and the 11/125 broken, and the 7/49 will be severed in March; that’s the last one. But Link going to lower downtown is just part of the vast number of new one-seat rides Link will add.

        “It should be noted that Angle Lake Station was the terminus station in Sound Move.”

        Was it really? I thought the terminus was SeaTac.

        “My suspicion is that UW may try to get rid of the Federal Way … Husky Stadium shuttles when Angle Lake Station Opens,”

        That’s so the tail wagging the dog. ST would not make decisions merely to eliminate a Husky shuttle ten days a year. It’ll be more significant when the 197 is deleted. When will that be?

      7. “But Link going to lower downtown is just part of the vast number of new one-seat rides Link will add.”

        That won’t help me though since I live halfway between Westlake and Capitol Hill. I’d end up going to Westlake anyway. But for those who live near Broadway or on the east side of it, it will be a bigger benefit.

      8. @Mike Orr

        “The tail wagging the dog”

        I don’t think ST made any decisions at all about U-Link based on the Husky Shuttles. They represent such an incredible small percentage of total UW Station ridership that their contribution will be totally insignificant. And usually special event ridership is excluded from ridership forecasts anyhow, so it really didn’t figure in.

        However, from the POV of the UW the Husky shuttles do represent both a significant cost and a significant headache. The UW will certainly want to get rid of the shuttles and move as many people as possible over to Link.

    4. It should be fairly easy to figure out how many people will ride the Angle Lake station. There is very little in the way of density there. There will be very few people who walk to that station. So that leaves the people who drive and the shuttle buses. How big is the parking lot? That answers the first one. I have no idea how the buses will be tailored to work with the new station, but my guess is that you really won’t have a huge number that go there. As has been noted many times, taking a light rail line — even one with the enormous gap between Tukwila and Rainier Beach — just doesn’t make sense from the suburbs. Never has, never will. Which means that either you make some very unpopular restructures (i. e. get rid of the express buses so that people will have a longer commute) or you are really benefiting only those that want to get to Rainier Valley. There are a few people that want to do that (I’m sure) but not huge numbers. I’m guessing that the popularity of this station will closely match the number of parking spaces.

  6. Above discussion proves my point about predictions. The more knowledge and precedent they have behind them, the more accurate they’ll be.

    So since our experience with this situation is essentially zero, let’s let U-LINK run for awhile, and see how its bus connections work out.

    If serious problems develop, it should not be too hard to make adjustments. It’s not like the 74X is on trolleywire, let alone rail.

    One adjustment I’m pretty sure will definitely happen is the the Route 43 will be restored to daily all-day service, since wire is already there, meaning that change will be no big deal at all.

    Mark

    1. 74x is already a serious problem and has been for a long time yet nothing has been done. Now it will only get worse.

  7. I disagree that technical merits of the line itself aren’t as important as such points as the price of the fare. People are unwilling to pay vast sums, but some of the lines in Europe I have examined have a relatively high fare. Their technical details, however, make them very preferable to alternatives.

    Then again, some of the technical details where Inlive leave a bit to be desired. Upzoning without avoid line gets you the Portland Pearl District: lots of density with paltry transit use by comparison due to the quality of the transit line.

    For Husky Stadium Link, my big concern about the technical details continues to be getting to the stations. All of the zoning in the world isn’t going to help if people can’t get from/to their destinations to the stations very well.

    1. Lets try that second paragraph again.

      Some of the technical details where I live leave a bit to be desired. Upzoning combined with a line that lacks the technical details gets you Portland’s Pearl District, with all the attendant parking issues and low transit use.

      1. I agree. All of these play a factor, of course, but the technical details of a system make a huge difference.

      2. Oh, and as I said above, that is just in terms of ridership. Ridership is not the key point here. Improved mobility is. East L. A. has fairly high transit ridership because they can’t afford alternatives. To focus efforts elsewhere (in an attempt at only improving overall transit ridership) would be stupid. Even in areas where people can afford a car, it makes sense to improve things. If you live downtown, you have fairly good transit everywhere. But it could be better — a lot better, which is why building grade separated light rail makes a lot of sense.

  8. FILLING UP SEATTLE: Here’s a little thought experiment for a Sunday morning puzzle. Suppose you want to fill downtown Seattle each weekday with workers for the day shift in 2050, so thinking big, maybe 1/2 million workers each morning.
    LRT (3 min headways at 600/train), with 3 lanes entering the CBD would bring about 36,000 riders per hour, so that’s 7%.
    Buses (10 sec headways @ 50 riders each) with 6 HOV lanes entering downtown Seattle would bring another 108,000 riders each hour, or 22%.
    Autos & vans (2 sec headways at 3 seats per vehicle) with 20 GP lanes entering downtown would bring another 108,000 riders if they were all full, so another 22%.
    So, rough numbers, with all modes filling all seats for one hour would bring 1/4 million workers into downtown Seattle each hour.
    As was pointed out in a recent Advanced Transportation Technology Conference in Mercer Is, we don’t have a lack of lane miles, just a utilization problem. Uber, Lyft, Metro rideshare products, buses and trains (Link and CR), all working together using creative solutions to fill empty seats are the low hanging fruit here.
    Your thoughts..

  9. Only 9,000 additional riders when Northgate opens? That seems way too low. Are the 41’s going to continue? They alone carry double that daily. Won’t the 7X peak expresses be truncated at Roosevelt?

    It seems way too low.

    1. I agree. I just wrote a comment stating that their modeling seems to be pretty good, but this seems ridiculous. The 41 won’t continue (or course it won’t). That right there is about 10,000, if not a lot more. The improvement in service (especially in the evening) will be significant, which will attract riders. That doesn’t count every other connection (Capitol Hill to Northgate, UW to Northgate, Capitol Hill to Roosevelt, etc.).

      As mentioned above, the modeling should include an assumed bus restructure, especially since restructures will be a lot easier after this one. I think everyone agrees that this restructure was a stretch, because getting to the UW station is a lot easier than getting to Husky Stadium. A reasonable restructure, along with things like Roosevelt HCT (which looks like a given right now) should result in a very big increase in transit ridership.

  10. I respectfully disagree that measuring actual use vs projections is useless. The implication seems to be that anyone who cares about ridership numbers that much is a died-in-the-wool War on Cars soldier that would have voted against a transit measure no matter what ridership was. Judging by the conversations I have with my father and brother (both well-meaning, generally liberal people that live in the suburbs and drive everywhere due to work and lifestyle preferences), the sight of half-full buses and light rail cars reinforces negative stereotypes they already get from their co-workers and shitty media like local TV news and KIRO Radio – that mass transit is just for poor people and urban sophisticates like myself and that transportation is a zero-sum game in which more buses, light rail and bike lanes means less lane space for them.

    The best way to counter (rational) people’s anecdotal impressions is with hard data. I’ve routinely brought up how great Central Link has been performing compared to its 2011 projections and had them concede that people will use transit that is accessible and reliable. Sure, projections are totally optics, but optics matter in politics.

  11. I agree that models are very useful as a analysis tool to compare between alternatives. I disagree with your statement “comparing actual riders to projected riders proves nothing about if the model was valid”.

    From the modeler’s perspective, it does exactly this! And they should be using what they learned about what actually happened versus what they thought would happen to improve their next forecast. For example, a forecaster in the 1950s might have gotten their employment forecast wrong due to the vast number of women that began entering the workforce after then (http://www.pewsocialtrends.org/2009/10/01/the-harried-life-of-the-working-mother/).

    From a general public’s perspective, well then yeah, who cares if some 20 year old model was wrong by 10% or even 100% or more? The decisions were made and the projects built. At least with infrastructure it’s not like we’re gonna tear down a bridge that only gets 20,000 daily vehicles instead of a forecasted 50,000.

    My quip is that if these models are always way off despite very closely approximating the major future inputs (housing, jobs, aggregate travel behaviors, other transport), then the decision makers should lose faith in the models.

Comments are closed.