There’s a certain thread of argument in transit advocacy that is frustrating because it is totally factually accurate, and yet completely misses the point. The latest example is this report on American light rail by the right-leaning Manhattan Institute, called The Economics of Urban Light Rail: A Guide for Planners and Citizens.

America has way fewer light rail boardings per route-mile than Europe and Canada. However, Seattle does well, one of only 5 in the US to have more than 3,000. No surprise: the strongest indicators of success are frequency and density around stations, two standards Link currently meets.

There’s nothing here that someone who’s spent a week in our comment threads doesn’t know. To optimize for ridership:

  • Focus on close-in, urban areas, particularly over low-density areas but also far-flung TOD;
  • De-emphasize parking;
  • Don’t build highways through downtown;
  • Eliminate density restrictions around stations;
  • Don’t obsess over airport service; and
  • Go through the dense area, not on the cheaper right-of-way that’s sort of near it.

If regions actually implemented this agenda, there wouldn’t need to be an STB. Needless to say, Sound Transit and its municipal partners are largely failing at these criteria, even as they do much better than peer regions.

But the next level of enlightenment is to realize that not all virtually no stakeholders are trying to maximize ridership. The report almost gets there, blaming “route choices that are more politically expedient than economically efficient.” But there’s a certain hubris in “economic efficiency” as an obviously correct objective, while everything else is “political expedience.”

After all, “political expedience” is another phrase for “giving people what they want.” And what people want, unless they are just trying to not pay taxes, is high-quality projects in their communities rather than a distant downtown, or “social justice,” or minimizing “impacts.” In particular, minimizing impacts means spending money to not change things much at all, which is the opposite of “economic efficiency.” It’s one reason light rail money can’t simply be re-appropriated into high-quality BRT, as reports like this one seem to imagine.

Knowledge is great, and if this report somehow went viral we might get somewhat better policy outcomes. But this won’t go viral, and few people will read transit planning doctrine.

Rather than tell transit advocates that they should fight for something else (BRT) that is going to be disappointing, a more useful education would highlight some of the institutional factors that lead to the outcomes we’re getting. For example:

  • The state legislature yoked King County’s high-capacity transit fortunes to outer counties that demand capital projects but contribute negligible taxes to projects in the core.
  • A culture that requires public votes to build transit, but not other projects, and offers many opportunities for recall and other kneecapping ballot measures.
  • Zoning happens at the lowest political level, where specific interests are most able to overcome shared regional objectives.
  • Cities have permitting power over regional projects, giving them leverage to overcome system goals in favor of short-term interests.

There are veto points everywhere, so projects have to be as inoffensive as possible. As a large minority of voters will never approve a tax increase, that leaves little margin to be offensive, or disruptive, at all. And a project that can’t be disruptive can’t be transformative.

Fix that doom loop, and we’ll achieve European outcomes.

56 Replies to “Maximizing ridership is easy”

  1. “A culture that requires public votes to build transit, but not other projects, and offers many opportunities for recall and other kneecapping ballot measures.”

    We do vote on a lot of projects besides transit: Pike Place Market, school capital construction, renovating Seattle Public Libraries, Seattle Fire Stations have all had public votes in the last 10-15 years. Probably others that I’m forgetting.

    This year will have the Harborview hospital construction project.

    Transit has a larger voting area, at least for ST, but that’s a function of where it serves.

    1. Yeah I agree – the political process for large capital levies seems pretty healthy in Washington. I think there is more grounds for criticism with reoccurring votes needed to sustain transit O&M funding, but not other types of basic operating budgets.

    2. Not to put words in Martin’s mouth, but amend the sentence to “other [transportation] projects” and it makes more sense.

  2. The thing this graph tells me is that our forecasting methods are reasonably good. This is particularly true for systems over 2K riders per mile. The fact that systems are about half above and half below also demonstrates reasonableness.

    My observation is that the same cannot be said for cost estimation. It’s easy to point out when ridership is 30 percent off, but when capital costs are 30 percent off there is little complaining. Capital costs are also rarely low.

    I’ll be interested to see how the report deals with capital cost accuracy — which I think is a much bigger problem than our demand forecasting tools are.

  3. Great piece. Given that the institutional factors are unlikely to be changed anytime soon, the solution is, for better or worse, to build more tunnels in order to get the stations in the right locations in the middle of the dense areas. While the upfront cost is higher, the long-term benefit of having a Ballard station on 20th and a West Seattle station directly in the Junction will be worth the extra cost.

      1. This is exactly what the decision to spring for Link is/was.

        For decades Seattle has had almost all of the infrastructure needed to establish a robust regional BRT network that would serve as much if not more of the Link service area for a fraction the cost with properly managed HOV and express lanes and some added transit priority on city streets. But it was easier to ask for billions of dollars on three occasions to build the currently planned system than it was to ask for reallocation of existing space. Those billions have gotten us a handful of new and faster connections that wouldn’t have been possible with buses on reallocated ROW (downtown to Capitol Hill, downtown to Beacon Hill—it’s late; I know there are others, but not many) and strung together sequences of destinations that previously weren’t well linked. But they have missed creating the new transformatively fast crosstown connections that no amount of space reallocation for buses will ever create (“the 8”, Ballard-UW, even a third lake crossing). By and large the rail system is duplicating what the current regional bus system serves today and what an enhanced regional bus system might have been able to serve better for less $$$.

      2. Capacity. If you truncate buses to Link as planned in ST2, you need the tunnel to be 100% rail to manage throughput. Otherwise, you are running a ton of buses downtown, and even with perfect bus lanes, you cannot match the speed of the tunnel.

        If your goal is to build the minimum amount of rail and go all in on perfect BRT, I suppose you could built a system that terminates at Northgate, Mt Baker, and Mercer Island. But then you are assuming a perfect political environment. At that point, it’s like arguing if only we had the perfect zoning & permitting rules for the past 50 years, we wouldn’t need to invest in affordable housing.

      3. The bus tunnel (now used by Link) was capable of handing a lot more buses. I don’t think the issue with the old approach was capacity. If we had spent all of our money on enhancing the bus system, in many ways it would be better. If nothing else, there would a lot more frequent buses. While Link is frequent, the number of bus stops that would be frequent would dwarf those of Link, as would the number of riders.

        But in other ways it would be much worse. Seattle actually has “close-in, urban areas” that are worthy of a subway. The UW-Capitol Hill-downtown corridor is extremely strong, with very good stops along the way. First Hill should be part of it, but even with just Capitol Hill, it is the type of area where a subway makes sense. Express buses (like the old 71/72/73) allowed for fast travel downtown, but skipped Capitol Hill. Local buses (like the 43) would get bogged down in traffic and bridge openings. Theoretically you could have extended the bus tunnel, but there would be little advantage to doing that.

        Bus tunnels are best when you have a trunk and branch system. UW to Capitol Hill is really not trunk and branch, and if you treated it that way, you would probably have to spend *more* than you did for rail (you would need a big entrance for the buses somewhere in the U-District).

        HOV freeway lanes work really well for trunk and branch systems as well. The main drawback is that they sometimes skip over interim stops (like the 71/72/73 skipped over Capitol Hill).

        Of course all the systems can, and should work together. There is no reason why you can’t have a bus tunnel AND a subway system. In general, it is a matter of picking the right tool for the job, and investing accordingly. For greater Seattle, here are a few areas:

        Northgate to downtown — Subway

        North of Northgate — Feeder system. Things get complicated because Northgate is not a good terminus for neighborhood buses. From a geographic standpoint, the train should be extended 145th (with a station at 130th). We would also need to add bus ramps to the 145th station. As is the case everywhere, change HOV2 to HOV3 on the freeway.

        Rainier Valley — BRT on both major corridors. Add more stops as well. Some of the buses would go out to Renton, while others go to SeaTac. You would have to add HOV ramps (and HOV lanes) for both.

        South Sound — Express trains to downtown and SeaTac (with more HOV ramps and lanes). You don’t gain much from the train — there aren’t that many people taking the train from Rainier Valley to Federal Way or Angle Lake. Those riders would do fine making a transfer at SeaTac.

        West Seattle — A clear case of a trunk and branch system where BRT makes the most sense. Make relatively cheap improvements to the West Seattle bridge, and travel between West Seattle and downtown never experiences congestion. Add a new bus tunnel, and then run frequent buses on most, or all of the West Seattle “branches” (Fauntleroy, 35th, Alki/Admiral/California, Alki Ave., Delridge, 16th). The big flaw with the train is that it doesn’t add anything. Unlike the main line, it doesn’t add any stops along the way. This means that BRT saves money from leveraging the freeways, while almost all riders get to their destination faster.

        Ballard Link — BRT for the same reason, although the case isn’t as strong. There are two major corridors in Ballard (15th and 24th) whereas West Seattle has several midsize ones. The problem with the train is that it adds very little over a bus tunnel and BRT. It will be faster, but 15th/Elliot is very fast, and has all the same stops. The bridge is a problem, but with a modest investment, buses could skip to the front of the line, which means that an opening would not involve as large a delay. The city could (and should anyway) work with the Coast Guard to extend the rush hour period of non-opening. Either way it is hard to justify the Ballard line based on the Ballard bridge. If so, then why aren’t we building a high bridge or a tunnel for Fremont and Eastlake? The point is, even with the issues with the Ballard Bridge, a BRT system (with a bus tunnel) would save a lot of riders a lot of time over the proposed train (at 15th) while saving a huge amount of money.

        Aurora/Phinney Ridge — BRT (classic trunk and branch).

        Ballard to UW — Subway. There are too many cross streets and the streets are too narrow to make surface BRT great between there (although I would certainly like to try). You need to dig a tunnel to get great speeds. A bus tunnel wouldn’t make any sense, because it really doesn’t branch on either side and it wouldn’t save you any money (you would need to build the entire tunnel anyway).

        Metro 8 — Subway. (Same Reasoning as Ballard to UW). Would not follow the street grid (like the existing 8) but would round the corner, perhaps with a station close to Seattle U. It wouldn’t loop around, but rather, end at Lower Queen Anne (considered by some to be part of downtown).

        Madison BRT — Surface BRT (too short and expensive to bother with deep tunnel stations).

        East Link — It is possible this could have worked as BRT, but I think rail was the right choice.

        Rest of the East Side — BRT. BRISK was a good idea, and for a lot less money, they could have added a lot more service as well as using the CKC for BRT (which would have been much better than what their building). If they wanted to spend even more money, they could add additional freeway improvements. For example, HOV ramps connecting southbound 405 to eastbound I-90 (and the reverse) for fast Issaquah to Bellevue travel, or southbound 405 to westbound 520 (for fast Totem Lake to UW). Even without those major investments, bus investments would have been better.

        It interesting to think of what that would look like. It would be something like this: The map is a bit cluttered, since it is has so many stations (which is a good thing). You can turn on and off various layers if you want to just look at a particular area.

        It doesn’t list all the surface BRT (or frequent bus routes) but you get the idea. You would have rail from 145th to downtown, as well as a split at the U-District, with a train going to Ballard (ending at 24th). Ballard and West Seattle buses would run unimpeded from traffic and traffic signals from West Seattle to Smith Cove, with only a handful of traffic lights (and no traffic) from there to Ballard. Aurora buses would be much faster, again running largely unimpeded from Green Lake to West Seattle. The Metro 8 would have much better stations in South Lake Union, with a direct connection to Capitol Hill and the Central Area. Rainier Valley as a whole has much better bus service, and it doesn’t take much longer to get downtown. East Link is still East Link, while the rest of the suburbs have feeder service and much better overall bus service.

        There would clearly be some winners and losers, but overall, a lot more winners. Rainier Valley doesn’t have the consistency of Link, and neither does Tukwila. Beacon Hill loses a lot — both the frequent connection to Rainier Valley, as well as the quick ride to downtown. That is why eventually, after all that is built, adding what we started with (a line from downtown to SeaTac) wouldn’t be a bad idea.

        But that would be about it. Even without South Link, you would have much higher ridership, while the cost would probably be lower. It would be much easier to get around the city — often by a combination of bus and rail (which is what most of us will have to use anyway).

      4. Why spend political capital when you can spend someone else’s money!

        I don’t think that’s relevant to Joe Z’s point. The Ballard Station at 20th is simply the best option. If you are going to run a train there, then you might as well spend a bit extra, and do it right.

        In contrast, the decision to build Issaquah Link is purely about political capital. They didn’t want to deal with the CKC, nor did they want to tell folks in Issaquah that a train to Bellevue just isn’t worth it. Or even that it would be cheaper to add an HOV ramp from westbound I-90 to northbound 405 (as expensive as that would be).

        For that matter, much of suburban Link seems to be based on the inability to change the HOV-2 lanes to HOV-3.

        But in the case of Ballard, there is really no cheap alternative that would accomplish the same thing, unless you are arguing for BRT. Running a train to 14th, or even 15th, is the same sort of expensive-yet-still-half-ass situation that hampers many light rail lines. It really isn’t much better than BRT (even BRT over the Ballard bridge) but it costs a lot more.

  4. “Don’t build highways through downtown“

    That’s a conclusion not borne out by the facts in the graph. There is plenty of freeway access in almost every major city regardless of demand by mile.

    I could see a slight modification to this conclusion — like “don’t build parking garages in downtown” to force parking costs up which then moves people into transit — but the highways by themselves don’t seem to reduce light rail productivity.

    1. There is plenty of freeway access in almost every major city regardless of demand by mile.

      Except, of course, that lonely little “outlier” dot in the extreme upper right-hand corner, “SF”.

      There are no freeways through downtown San Francisco, though “Yes, along the edge”. In fact, they tore down the one and only. Grant, it was going to fall down if they didn’t, but they could have rebuilt it. They even tore down another piece on the periphery of downtown, also damaged but to a lesser degree, in order to improve the neighborhood through which it passed.

      1. Boston is missing from the graph and it performs better than San Francisco (as noted in the article). Central Boston has two underground freeways (I90 and I93).

        One thing Boston and San Francisco do have in common is outrageously expensive and scarce central parking.

      2. Good point – odd that Boston is missing. Also similar to SF in that’s it’s a legacy streetcar line that gets great ridership b/c of a downtown tunnel + historical streetcar suburb density outside of the core.

      3. It is weird that Boston is missing. It would likely be right on that line, just a little bit above San Fransisco. It doesn’t have as much overall ridership, but you are right, the ridership per mile (which is what they measure) is better for Boston.

    2. If you listen to transit commuters and car commuters you hear “The cost of parking is too high to drive” more than “Congestion is too bad to drive.” People grudgingly drive through congestion if they have a free parking space at the end, but they don’t want to pay for $10 or $30 parking. Maybe because the cost is so an explicit payment at a single point, while the cost of driving in congestion gets lost in your gas bill and car maintenance and time spent and feeling like it’s still the best option.

      This may also relate to the sensitivity to ST taxes and car tabs, because those are also explicit single-point payments. That gets to Martin’s point about how transit is funded differently than car infrastructure. The almost invisible car tax funds car infrastructure, and if it’s not enough the legislature adds to it without a public vote. But transit projects have to be explicitly voted for, which implies voting for a tax. And the base transit sales tax is inadequate and regressive, so localities have to supplement it with levies. And thanks to Eyman levies have to be renewed every few years or they expire. We don’t fund highway signs or entrance meters or the highway patrol by short-term levies but we fund transit projects and operations that way.

      1. Mike, maybe I’m wrong to address passenger mode-preference negatively: “Eventually, congestion will get bad enough that people switch to transit”, instead of positively: “Is it within transit’s capability to make itself good enough that people will ride it by choice?”

        Reason I keep asking to hear from transit drivers and other operating personnel. Because compared to the time and effort for politics to deliver favorable zoning, service quality is something that our people in uniform (including company office fashion) can create by their own individual and combined handling of their work.

        An organized “assist” that is also in the job description of ATU Local 587, which is there to do its members’ bidding.

        Mark Dublin

    3. The graph is only one part of the report. It shows the strong relationship between ridership, frequency and density in U. S. cities. The report (in a different section) mentioned that U. S. light rail tends to underperform even Canadian cities in part because we have those freeways going downtown (and some of them don’t). In the report, it also mentioned parking as a big factor as well. In short, the U. S. has lots of parking downtown, and lots of cities with freeway access to downtown as well (while lots of cities outside the U. S. don’t).

  5. Martin, thank you for this morning’s topic. As to Manhattan Institute’s politics, though, the fact that priceless New Electric Railway Journal was founded by ideologically Royalist Paul N. Weyrich proves that streetcars truly define the word “Conservative.”

    Favor everybody can do me: Am I right to define “Light Rail” as “Trains capable of regional speed, but also able to negotiate streetcar curve radii when necessary?” And also accept my apology for demanding oversize streetcars just to seize control from buses. The car-line my First Avenue Connector will be one third of will have to be self-financed by the linear business, arts and culture district it will be tailor-designed to serve. Up front.

    But behind it all, while so many things make prediction impossible, I really would like to see our transit future as one facet of our already-changing economy, in the hands of people now just turned voting age, that will more than restore our country’s health, prosperity, and freedom. Tempting to invoke Franklin Roosevelt, but pray God, could we please not owe our recovery to a war?

    And concentrate instead on some attributes of Roosevelt’s programs. One, much of his best, he himself knew he was “winging”, not calculating. And two, given his arts programs, Nordic Heritage should be able to gift Ballard with a breath-takingly-beautiful draw-bridge for Link which I now think is the perfect solution.

    Mark Dublin

  6. Some thoughts on the low performers:

    Cleveland actually as a really well designed rail system (I’m curious if the light rail metrics only cover the blue line) in terms of alignment and station area density for a city of it’s size. The problem is the frequency is poor and parking is abundant. This suggest the regression isn’t linear, and below a certain level of service the line curves down

    San Jose: I’m curious if the new direct connect to BART will help lift ridership closer to the trend line.

    Denver: A great example of close but not close enough to density. Denver went for the easy ROW and is often just far enough to be inconvenient. There is still a solid system, but it will require Denver to growth towards and around its rail station over time, which comes down to zoning. If Elitch Gardens ever gets redeveloped (Amazon HQ4?), Denver should return to the trend line. Denver needs a SLU-eque expansion of their downtown westward.

    On the high performers:

    LA has a good system that is diluted by long lines. New downtown tunnel will do wonders, but a great system will require more TOD, particularly along the gold & expo.

    San Diego and Minn – great examples that it’s more important to go where people want to go, rather than focus on being fast. SD’s big extension appears to serve UCSD well, but it’s long so might push it down a bit on riders/mile.

    1. Having ridden San Jose light rail for two weeks as a commuter on temporary assignment, I attribute their problem directly to the very slow Downtown mall. It was excruciating to have the last mile of my commute take over 10 minutes! The same is true for many of the other lower demand systems in the US.

      To that end, I don’t expect the new BART connection in Milpitas to help much. On the other hand, getting the Doentown San Jose BART tunnel opened will be a game changer.

      1. The Downtown is the most excruciating segment, but there’s an abundance of slow segment similar to Downtown, like the Tasman West segment. VTA has pretty much written the book on how not to build Light Rail. They got every single step wrong, and had the gall to ditch the mode when the corridors that actually had the demand for it came up.

        LA problem is that they’re cheating on operations like a lot of systems do, and this depresses ridership.

      2. I would put Portland’s Trimet MAX Lines in the same general category. But I understand they are attempting to close a few downtown stations for the sake of speeding things up down there. Hope to be able to visit again next year!

      3. San Jose’s rush hour ridership is like Link at 8pm. The problem is that most of it serves one narrow transit market: San Jose residents to north San Jose, Santa Clara, and western office parks. Santa Clara residents mostly can’t use it because it doesn’t go near where they live. Those office parks have very little around them that non-workers might go to, and in north San Jose the buildings are two stories. I used to be incensed about the height but I gather it’s capped by the airport flight paths. Still, that means it’s a bad place for a light rail line, and San Jose could have served its people better by putting it somewhere else.

        When the line was new I took it north to Tazman Drive, which I think was the northern terminus (the east-west segments hadn’t been built). It was the middle of nowhere. Tazman Drive looked like a new street with no buildings around it yet. So it failed to serve the Santa Clara residential areas, and the buildings that eventually came are towers in the park. Later in the mid 2000s I had a couple events at the convention center and stayed south of there near Great America. From there you can’t even find a supermarket without taking the train several miles, and restaurants are almost as hard. And walking, you have to walk half an hour to reach even one isolated restaurant or convenience store. (I did eventually find a Safeway plaza east on the Montague Expressway, a 20-minute walk east of where I was staying, which itself was a 45-minute walk from the convention center. And the north-south bus was half-hourly, or hourly on Sundays/holidays.)

      4. Let’s not forget the huge free parking lots that a San Jose light rail rider has to walk through to get from the station platform to buildings in North San Jose, Sunnyvale or Mountain View.

      5. In San Jose, parking lots play an important role for pedestrian mobility. Many streets do not have sidewalks, but you can walk through parking lots to avoid getting run over by cars on the street. There are also cases where cutting through parking lots can save half a mile or so of walking compared to following the streets.

      6. And, of course, I can’t talk about Santa Clara transit without mentioning one of its worst abominations. There’s a street near a cluster of Google offices where somebody came up with the brilliant idea of building the light rail tracks where the sidewalk should go, to avoid hindering car capacity on an extremely wide street. The fencing around the tracks also prevents you from cutting across to the office buildings on the other side.

        The pedestrian detour to get around the stupid tracks is about 15 minutes, while cars get to drive the direct route on the super wide street with no sidewalk.

      7. The worst Santa Clara abomination to me is the half-mile wide blocks. It takes a full 10 minutes to walk one block, with only one building in the middle. When I walked to said restaurant and convenience store, it took 30 minutes to walk 3 blocks, passing only 3 buildings. It’s like that at Old Ironsides, further south, and southeast at least.

        The Safeway plaza is at Montague & Agnew, where the blocks are just as large, and Agnew Street has a very long stretch without intersections, like a country road. The area across from the Safeway plaza and north of it looks like it intends to be a New Urbanist development. It has close-together houses like in the Issaquah Highlands, but along one street instead of a 2-D grid. Across from Safeway is a huge apartment complex. Thankfully the housing is near the Safeway plaza, so they at least got that part right. But the apartment complex has a huge setback like at 212th & West Valley Highway in the north Kent industrial district. There’s a nice waterfall in the space, but anybody walking to the Safeway would have to spend five minutes walking around the open space to get to it. And the nearest other jobs are a mile away.

      8. I will say this about Silicon Valley: as car centric as they are, their bike network is actually surprisingly decent. The arterial streets are pretty good about having bike lanes and, unlike here, they don’t have on street parking, which means the bike lanes aren’t in the door zone. They also have a surprisingly good network of off-street, Burke-Gilman-style trails, including one going right to the Great America convention center.

        My last visit there, I decided to check it out and rode right out of the San Jose airport to my company’s Sunnyvale office on a Bird scooter, then switched to riding Spin bikes around from hotel to convention center, and back to the airport again.

        While things are generally too far apart to walk, they are surprisingly close enough together to make getting around not terrible on a bike. The transit stinks; fortunately, the weather is decent enough that if you have a bike, you don’t need it.

    2. LA has a really solid transit system (I didn’t even rent a car when I was a tourist for a few days in the pre-virus era, so long ago…).

      People automatically assume LA = car. But there are a lot of transit riders and some good investment in bus service along with the rail. I was very pleasantly surprised.

      1. Alex [the other one] in 1992, used my own Metro driver’s vacation time to visit Los Angeles to check out plans to wire a bus route from Downtown LA to Santa Monica. Which ended up, I’m afraid, a casualty of the repercussions over the beating by LA police of a man named Rodney King. Good things like that don’t happen anymore.

        But from the minute I drove my rental car out the gate, major surprise was how polite and skilled the average LA motorist was. When it’s again possible, pro-automobile groups around Seattle ought to sponsor week-long driver-training excursions down there.

        Blue Line to Long Beach, a lot better than impressive. El Monte Busway also exemplary. Remember, this was 1992. When COVIDIA flies away, really is time to go see the additional light rail lines built since. Town needs a lot more transit- positivity credit.

        Mark Dublin

      2. From what I understand, LA has a Seattle-sized area with a solid transit system. That’s different from the entire city or metropolis having it.

      3. The L. A. region has very low ridership for its density. Some of that may be due to the region being so huge, without a very strong core. But given the fact that L. A. was mentioned several times in this report (for what not to do) I think you can make the case that L. A. has not been spending its money wisely. It is quite possible that a smaller rail system along with lots of bus improvements (and especially an increase in frequency) would lead to a lot more ridership.

        My guess is there are similarities with Seattle. Much of the ST2 expansion is great, and definitely worthwhile. But a lot of the city — and certainly the region — lacks decent bus service. Given the importance of frequency, it is a big mismatch. I’m not sure if we should focus on bus service *instead* of expanding Link, but if we are going to spend a bunch of money on Link, the least we can do is spend a bit more on bus service. My guess is L. A. is in the same boat.

      4. LA Metro’s rail program has a complicated history. They had something that ST didn’t — aTitle VI lawsuit! That’s because buses in their poor minority neighborhoods were packed to the gills while the rail program was energized to serve places like Pasadena first over East LA. Since the ruling, voter-approved funding has been more considerate of this issue.

        If ST3 had this scrutiny, I believe it would have been different. The subarea funding pot tracker in itself is discriminatory. Then, within each subarea the projects usually go to benefit middle and higher income residents over low income ones. ST logic is currently so bent that the ST3 project equity discussions are about construction impacts rather than who gets served with a station — because that was fixed into the referendum.

    3. My perspective on Denver having lived there for a few years:

      1. They build light-rail VERY cheaply because they make zero effort to serve any dense neighborhoods within the city of Denver. The lines are built along freeways to save cost and mostly only serve suburban commuters going to downtown.
      2. The Denver airport is uncommonly far from the city center and they built a light-rail line out there. To make matters worse, this line doesn’t even get used much because there are faster bus options which go directly from the airport to the city center or to large transit centers. Denver traffic is not terrible so these busses don’t really get slowed down by traffic unless you go at the height of rush-hour.

      1. 1. Mostly freeways … the W line seems like an interesting route with some opportunity if zoning allowed.
        Freeway running is a bummer but the real problem is the terminus in Union Station rather than downtown. Either downtown needs to grow west or they need a better way to serve downtown directly … the shuttle is fine, but not good enough with pair with rather slow light rail routes. With a real terminus, the LRT lines should do OK with good feeder service.
        2. Airport is heavy rail and shouldn’t show up in these metrics

      2. @AJ

        The new W line illustrates exactly my point. “6th Avenue” in Denver is a 6-lane highway. Take a look at google street-view if you don’t believe me. I also have serious doubts about Denver’s desire or ability to implement any sort of TOD around the Pepsi Center or Mile High Stadium. These are stadiums surrounded by HUGE parking lots and are adjacent to a massive highway (the 25)

        Union station may not be completely ideal, but it is a short walk from quite a lot of jobs, and Denver for it’s credit is making a strong effort to develop dense TOD around Union Station. Also, not all rail lines terminate there, much of the system terminates closer to downtown at the Convention center.

        Again, though, this illustrates my point. Take a look at the three lines terminating at the Convention Center. They take off south and follow a massive highway as they make their way to the suburbs. Just a little ways to the east of this highway is some of Denver’s most dense and vibrant neighborhoods (Capitol Hill, Lincoln Park, Cheesman Park, Washington Park, South Broadway). These lines could have served all of these Denver neighborhoods but they chose to use a mega-cheap freeway right-of-way instead.

      3. I’ve only ridden the W on the 13th Ave section. Yes, after Union Square it follows the freeway, but there’s a good 5 mile segment, mostly along 13th, that has everything but the zoning to make for a great transit corridor.

        Yes, LoDo is excellent TOD. I’m just saying the next step would be to replicate that west of Speer. That would better leverage the existing LRT infrastructure.

        East of downtown, looks like high quality BRT on Colfax should be sufficient to serve most of those neighborhoods … would have been nice to fund that with the original FasTracks, but I don’t think there was good BRT anywhere in NA when FasTracks was proposed so I don’t blame them.

    4. San Fransisco is another example of why it’s more important to go where people want to go, rather than focus on being fast. Muni is extremely slow, but runs through a highly dense area. To be fair, it is fast when it really needs to be fast.

  7. One of the problems with the graph is not every mile costs the same to build. The lower performing systems often operate segments in slow Downtown surface environments. I can see why some cities save on capital costs for this reason.

    The graph probably understates the bigger problem of slow surface segments in Downtowns: the cost to operate the slow trains. It’s a double whammy — both deterring riders (apparent in the graph) and adding operating costs (mostly labor) to the system (not apparent in the graph).

    It’s a tough dilemma to resolve.

    1. It is interesting that they have such a reliable relationship while focusing on only those two factors — frequency and density. I would think that speed also plays a part.

      But I think speed also matters in comparison to the alternatives. A system that is fast but runs by the freeway doesn’t save a lot of time in the middle of the day. But maybe systems like that weed themselves out. You aren’t going to have too many high density areas right next to the freeway, simply because the freeway takes up so much space. By avoiding the freeway, you choose the higher density areas. In some cases (like Capitol Hill) these are also fast. In other areas (like much of Portland) they aren’t.

      I do think it is interesting that Seattle’s system is very much like Portland’s — sitting just above the expected ridership. That suggests that speed really isn’t that important when it comes to ridership. That seems counter-intuitive, to say the least. I don’t know of any study that shows anything different though. I would assume as a city makes a route faster, ridership grows.

      1. I think it’s that speed is just one of many factors, like station location, frequency, and span of service. Often, decisions to improve speed can result in deterioration of other parts of the ridership recipe. For Portland, a downtown tunnel would be an immense improvement in speed and capacity, but will likely results in a significant station diet. So some trip pairs improve, others degrade. This isn’t an argument against a downtown tunnel, it’s just to note that if you focus just a speed, you’ll have issues elsewhere. If you can truly improve speed and degrade nothing else, ridership should improve.

  8. Emergency clarification: Message should’ve read: “It’s good that things like that do not happen anymore,” but in a Twitter- literal world, sarcasm’s too ambiguous to be advisable.

    This posting’s accurate and valuable. But for maximizing ridership, I’m wondering if our most powerful pro-transit motivator might not simply be the traffic jam. However long the gas-glut lasts….we’re out of road-building room. A moving train, or BRT, might be increasingly persuasive seen through the window of a stuck car.

    Especially if occupants of child-safety seats have ever had their first train ride.

    Mark Dublin

  9. Martin;
    That’s a set of painful damn truths I happen to agree with. Somewhat. From most to least.
    First thing I’d change is the, “A culture that requires public votes to build transit, but not other projects.” We should have a vote on highway projects/gas tax increases. But a Seattle vote should be a vote and not tampered with by say Skagit voters.
    Second is the zoning issues. Sadly we’re going to have to have more state & possibly federal constraints on local powers to zone. It’s how we overcome structural racism & ableism in zoning, not to mention keep buffoons from building overpriced homes next to Naval Outlying Fields & then complaining about the sound of freedom to KING 5.
    Third, I think “Cities have permitting power over regional projects” should be somewhat limited. I mean after all regional projects are regional – and since we can’t get Seattle to densify we need Sound Transit 2 & 3 projects.
    Fourth, I am all for an elected transit board. Either directly or comprised of all the mayors or all the county councilmembers. You should be able to vote and 100% know that person reps you on your regional/county transit board.
    Fifth and finally, I think what people want is a cost-effective transit network that is reliable and connects them to where they need to go that is clean & time-competitive with the car. That’s why I support bus lanes, not just sexy light rail.

  10. Having some stations where there is relatively low density is not terrible IF you don’t force that lower density to remain the case in perpetuity with zoning restrictions and parking minimums. The same way a new highway out to the countryside will spawn sprawl development, rapid transit will spawn density IF it is allowed to. The problem is we both avoid building in high density (“political expediency”) AND limit development where we do build rapid transit stations. There is absolutely zero reason for any single family zoning within a 15 minute walk from a rapid transit station. Period! Tell the cities and NIMBYS if the want to nitpick where the station goes to “mitigate impacts,” they’d better be willing to accept more density. If that means “gentrification,” well that’s a choice you have to make. Maybe put the station where it’s already more dense instead. We’re investing way too much taxpayer money not to!

    1. “Tell the cities and NIMBYS if the want to nitpick where the station goes to “mitigate impacts,” they’d better be willing to accept more density. ”

      That depends on getting pro-urban county councils and state legislature, because only they have the power to tell the cities this. Without that, the city leaders would have to decide to do it themselves. And they’re often afraid of single-family areas’ voting power. (A force which they themselves created.) And the EIS process and city permitting power have a status-quo bias. Stanchions or trains or trolley wires are a negative impact that must be mitigated. This is a one-way mitigation, not a two-way equal compromise.

    2. Having some stations where there is relatively low density is not terrible IF you don’t force that lower density to remain the case in perpetuity with zoning restrictions and parking minimums.

      From the report:

      By contrast, larger light rail systems that stretch into low-density suburban areas tend to underperform … Two species of overexpansion in U.S. cities deserve special mention. First is an overemphasis on serving transit-oriented developments. Many cities have seen new developments on “New Urbanist” principles: apartments with mid-rise units and a mix of commercial and residential development aimed at satisfying most residents’ daily needs without having to drive. Many of these developments are also transit-oriented, to allow for travel outside the development, such as to downtown jobs. Because of strict zoning laws in developed areas of cities, these developments often must be built miles from established downtowns. As such, transit-oriented developments frequently disappoint. New, isolated developments are rarely large enough to be self-contained or offer the amenities of true city centers. Residents who want to travel to specialty stores or jobs not readily accessible by the existing transit network—and in typical low-density U.S. cities, this is almost all of them—will need to own cars. Once they own cars, there’s no reason not to use them for all trips, especially if zoning policies guarantee copious parking.

      In other words, TOD stands a decent chance of success in someplace like Rainier Valley (where it is happening) but is unlikely to happen in distant suburbs. Furthermore, if you put the station close to the freeway (which is common for our expansion) it runs into all sorts of problems (again, quoting the report):

      Light rail lines along freeways are undesirable for several reasons. First, the freeway takes up much of the land accessible from rail stations on foot. Second, because freeways are convenient to access by automobile but unpleasant to live near, they tend to be surrounded by lower-value land uses. Third, trips on a light rail line that runs alongside a freeway are competing directly with the region’s fastest car trips.

      In summary, a lot of these places just don’t make sense for rail transit, as they will never have the ridership to justify it, no matter how the zoning changes.

  11. In order to optimize ridership, stop focusing on home based work trips. Successful high capacity transit ridership is not based on home based work trips. Get away from the mindset that home based work trips generate most of the ridership because they do not.

  12. From the report: “The light rail systems in Seattle, Minneapolis, Houston, and San Francisco, which are all among the best-performing in the United States, have no segments, or only short peripheral segments, along freeways”

    Hey, just give us time! Eventually we will make all the mistakes every other city has, and then some. The report does make a critique that includes Seattle, along with Houston and Phoenix:

    In Seattle, finally, Sound Transit has proposed extending the light rail line south from its current terminus at SeaTac airport to the suburb of Federal Way, 20 miles from downtown. The extension lacks the virtues that made the existing Seattle light rail network relatively successful. The planned 7.8-mile, $3.16 billion light rail line—at $400 million per mile, an inordinate construction cost for an above ground line—will run along-side Interstate 5 through low-density residential and retail areas with poor pedestrian accessibility. All three cities should reconsider or halt plans for expansion.

    By the way, the Federal Way extension is by no means the worst extension. More than anything, this reads like a fairly extensive explanation why ST3 is largely crap.

  13. Ridership is now dependent on safety due to the pandemic. Prior to that, getting you to where you want to go and reliability were factors. From south Everett, I cannot reach the regional transit connection just 4 miles away via transit, for Everett Transit has foolishly chosen to only serve it with a route that goes from downtown Everett, where almost all of the City Councilmembers live, no coincidence, but also where the other major regional transit connection is. Thus, to get to the Eastside is a four bus charade of hoping no transfers are missed. Having several self-serving transit bureaucracies instead of one centralized overseeing provider is part of the problem. For reliability, finishing the northbound onramps at 164th & I-5 would improve reliability for decades, especially now with Everett Link surely to be delayed and Sound Transit unlikely to complete the short, Northgate-like segment from Lynnwood to 128th/Mariner/Green Line BRT separately. This simple project, whose design is probably sitting in some Indiana Jones-like warehouse, is almost a level paving proposition, with no overpass needed. More complicated would be improving the throughput at every major interchange along I-405: 527, 522, 520, I-90, 900.

  14. Northgate-like segment from Lynnwood to 128th/Mariner/Green Line BRT

    I would like to know more about what you mean by that. It may be that you are thinking the same thing I’m thinking:

    One of the big things I would like to see is making the Ash Way HOV ramps accessible from the north. This would lead to a big times savings, and it looks very cheap ( It looks to me like they had this in mind from the beginning. This would improve routes like the 201/202, that go along Ash Way, then to the Ash Way Transit Center, then on to the freeway. It would also improve express routes, like the 512, which go on the freeway between Everett and downtown. When that bus gets truncated at Northgate, it means that once the 512 gets on the freeway and into the HOV lanes, it never leaves them until it gets close to Northgate. South Everett, Ash Way, Lynnwood and Mountlake Terrace can all be served via HOV lanes *both directions*.

    Ideally you build the same structure at 128th that exists at 164th. That is more expensive that simply completing the work on 164th, but no more expensive than the original 164th work. It looks very similar. There is plenty of space in the median, and the park and ride sits a little ways from the overpass. A bidirectional overpass would improve both those routes. The 512 would stop in 128th, with very little cost to through-riders. the 201/202 could get to Mariner without leaving the HOV lanes. It would serve Mariner, then Ash Way (the street) then the Ash Way Transit Center, then get back on the freeway to get to Lynnwood. When Lynnwood Link gets here, the two routes are combined, since they could easily serve the same stops.

    Both of these projects would be relatively cheap, and make a huge difference for a lot of riders.

  15. “The state legislature yoked King County’s high-capacity transit fortunes to outer counties that demand capital projects but contribute negligible taxes to projects in the core.”

    I don’t think that’s remotely true.

    1) If you remove Sounder, all subareas have been net contributors to North King. Yoking Central Link to the greater region has been an immense financial benefit to Seattle by managing cash flows and lowering borrowing costs. Part of the reason Pierce and Snohomish are so insistent about finishing the Spine is because they are keenly aware they have been bankrolling their share of Link since the beginning, particularly since the ST2 vote.

    South King is the other net deficit subarea, but that is mostly about South King’s relative poverty and makes good sense on equity grounds. The other 3 subareas pre-pay and only get their “fair share” at the end of the plan.

    2) Nearly all of the regional assets are in Seattle. This is sound policy, but it also means that if Seattle wasn’t yoked to the rest of the region, they’d have to pay 100% of the capital and O&M cost for facilities like the OMF and the old and new tunnels. Subareas contribute to regional assets based on their share of projected ridership. If Seattle was going alone, suburban riders contributing fares but nothing in taxes.

    To build Ross’s plan of miles of subways in Seattle and BRT in the suburbs would require either significantly higher taxes in Seattle or a massive redirection of suburban tax revenues into Seattle.

  16. Note this is very different than saying the subareas are spending their money poorly. If you think TDLE is trash and Pierce should have a BRT network instead, fine. But that doesn’t change King County’s HCT project portfolio or timeline.

Comments are closed.