Resist the Urge to Start Small

Remember the Ohio Hub? Back in 2009-10, Ohio was planning on running five low-speed trains per day between Cleveland and Cincinnati and branded this exercise as high-speed rail called the Ohio Hub. The Republican victory in the gubernatorial election put it out of its misery (as unfortunately happened to the far better Florida project), but the idea of little facts-on-the-ground kinds of rail investment persists among American advocates who don’t understand how rail operations work. Now that there’s serious talk of infrastructure funding in the United States as part of a stimulus package, I’d like to explain, to prevent the debacles of the late 2000s from happening again.

The central conceit is that public transportation is not cars. It’s a different, more complex system. The road network has fewer moving parts – one just builds roads based on traffic projections. Public transportation has schedules, transfers, and equipment, all of which must be planned in coordination. “This junction gets congested, let’s build a bypass” works for road advocacy, but fails for rail, because maybe speeding up the trains by a few minutes doesn’t really help get to any timed connections and is therefore of limited value to the system.

Rail works when everything is planned together. This makes little additions not too valuable: a small speedup may not be useful if connecting lines stay the same, infrastructure investment may have limited effect on trip times if the rolling stock doesn’t change, etc.

The upshot is that it’s very easy to find 80/20 problems: 80% of the money gets you 20% of the benefits. In addition to examples of lack of coordination between infrastructure, the timetable, and rolling stock, there are issues with insufficient frequency. When frequency is low relative to trip time, the long-term elasticity of ridership with respect to service is more than 1 – that is, running more service makes the trains and buses fuller, as better service encourages more ridership. Thus, service with insufficient frequency will fail, trains and buses getting too little ridership to justify additional investment, whereas if initial frequency were higher from the start then it would succeed.

The Ohio Hub was one such example: five roundtrips a day, starter service. It makes sense to someone who thinks like a manager or a general-purpose activist: start small and build from there. But to someone who thinks like a public transportation planner, it’s a disaster. Already 10 years ago, Max Wyss in comments was warning that such service would fail – the original Intercity brand in Germany succeeded by running trains every two hours, with hourly service on stronger city pairs, often with timed transfers at junctions.

Regional rail projects suffer from a similar urge to start small. Peak-only service will invariably fail – the operating costs will be too high for ridership even if almost all seats fill. This covers just about every American effort at starting up new commuter rail service.

More fundamentally, the issue is that nobody likes failure. Insufficient, poorly-optimized service creates facts on the ground, but these facts don’t lead to any effort toward better service if people perceive what has been built to be a failure. If a handful of trains per day that average 70 km/h are called high-speed rail, then it doesn’t lead passengers to want high-speed rail; it leads them to avoid the train and conclude that high-speed rail is slower than driving on the freeway.

The passengers on such service may not be a great constituency for better service, either. If the train is very slow, then the riders will be the sort of people who are okay with slow trains. Older American railfans are filled with nostalgia for traditional railroading and openly say that slower is better. Such people are not going to advocate for modern high-speed rail, nor for learning from successful Asian and European examples.

Another group of people who ride trains and often advocate against better service is peak commuters on trains serving high-income suburbs. They are used to an adversarial relationship with the state; to them, the state taxes them to give money to poorer people, and they instead prefer hyper-local forms of government providing segregated schools and policing. Representatives of such riders engage in agency turf warfare, such as when state senators from Long Island opposed Metro-North’s Penn Station Access because it would use train slots into Penn Station that the LIRR believes are its property. On social media, people sporadically yell at me when I propose fare integration, on grounds that boil down to viewing any urban riders who would be attracted to lower fares as interlopers.

There’s an ultimate proof-of-pudding issue here. Americans have to a good approximation never seen a working public transportation system. At best, they’ve seen a megacity where people use the trains even though they are dirty and expensive to run because there is no alternative and construction was done 100 years ago when costs were lower. There is no coordinated planning; Americans do not demand it because only a handful of people know what it is, who are often young and have often lived abroad for an extended period of time, both of which make one less likely to be listened to in politics.

The result is that the sort of bottom-up activism people are used to is not useful in this context. In Germany it’s different – enough people have seen what works in Austria, Switzerland, and the Netherlands and know what to call for. But in the United States, it won’t work – the knowledge base of how to build reliable, interconnected public transportation exists but is too thinly spread and is the domain of people who do not have much political prestige.

It’s critical to then get things right from the start. Do not assume future activism will fix things. Half-measures are much more likely to lead to disillusionment than to any serious efforts to improve things to turn them into full measures. If the choice is between a high chance of bad service and low chance of good service, don’t settle for bad service and make a gamble for good service; bad public transportation is a waste of money and the general public will correctly perceive it as such.

273 comments

  1. Fbfree

    There’s an ultimate proof-of-pudding issue here. Americans have to a good approximation never seen a working public transportation system. At best, they’ve seen a megacity where people use the trains even though they are dirty and expensive to run because there is no alternative and construction was done 100 years ago when costs were lower. There is no coordinated planning; Americans do not demand it because only a handful of people know what it is, who are often young and have often lived abroad for an extended period of time, both of which make one less likely to be listened to in politics.

    I imagine that Seattle would be the US counterexample to this, where long-term planning, significant investment, and strong local advocacy all work together. It also shows that questionable decisions can still be made (e.g. excessive park and rides, transfer station design) but in their case, these failings are called out and considered.

    • Alon Levy

      Medium-size French and British cities that everyone familiar with German rail mocks for having terrible transit get a mode share of 15%, give or take. Seattle is at 10%. Can we please let go of the idea of Seattle as some kind of trailblazing success?

      • Eric2

        Those French/British cities are much denser on average than the Seattle metropolitan area… Seattle is doing very well for its density level (and also adding density very well by US standards).

        • Herbert

          To give a comparison, the 2020-2026 coalition agreement for Munich calls for “80% of trips to be made by non-polluting means”

          Given the laughable electric car penetration that means bike+ walk+ transit…

          • Eric2

            Munich is mostly closely spaced mid-rises, while Seattle (despite some recent change) is mostly widely spaced single family houses. Transit is basically impossible in the latter environment, the density of passengers per route is too low.

          • SB

            @Herbert have you ever visited suburban US?
            It is miles of low density single-family housing and they isn’t a magical button to change it overnight.

          • Herbert

            I mean there IS a magical button to turn vast swaths of suburban housing into ruins… In fact, the U.S. invented it.

            But that is probably not what you meant and also not desirable. But why not buy out the “property values” obsessed homeowners, tear down the stuff that was previously and replace it with more dense stuff?

            It’s not like Germany where the average household buys point something houses per lifetime. In the U.S. wealthy suburbanites buy and sell houses as if they were a commodity..

        • Alon Levy

          They’re also far slower-growing. You transplant Seattle’s growth rates into even a semi-competent place, like Canada, and you get Vancouver’s 20%-and-rising mode share.

          • Eric2

            Vancouver is much denser than Seattle, and has been for many decades. You can’t exactly undo Seattle’s sprawl overnight, so its mode share is still lower despite steady improvement.

          • Alon Levy

            Vancouver has been investing in TOD since the 1980s, yes – the Downtown high-rises are from that era, as is Metrotown.

          • Eric2

            And even the “suburban” parts of Greater Vancouver are much denser and more gridded than equivalent parts of the Seattle area.

          • RossB

            Vancouver just has a better public transportation system than Seattle. They have a “near perfect grid” (https://humantransit.org/2010/02/vancouver-the-almost-perfect-grid.html). The mass transit system largely covers the urban core, with the obvious exception of the UBC line. Suburban riders can find good connecting bus service, and know that once they are in the city itself, they can get pretty much anywhere.

            Seattle doesn’t have that. There are plenty of trips that require a rider to go way out of their way (https://seattletransitmap.com/app/). There are also lots of trips that are just really slow. The mass transit line is out of balance — it goes all the way to the southern suburbs, yet still barely gets across the ship canal. That particular station is difficult to access, even though it the only one serving the University of Washington. It is probably the worst possible location for a station — next to a sports stadium that is used maybe a dozen times a year, and not the campus or the hospital. Those from the hospital have to cross two streets to get there (https://goo.gl/maps/87yyvL5pnSR2192b9). Those from the campus can take a skybridge … wait for it … to a station 30 meters *under* the ground. So not only is there only one station north of the ship canal, it is a poor one at that.

            If you look at the census maps (https://goo.gl/maps/87yyvL5pnSR2192b9) or employment maps, the current line doesn’t make much sense. It is out of balance. To be fair, it will be extended north, but then what? Rather than add new lines in the city, it is largely focused on going north and south — all the way from Everett to Tacoma (an extremely long distance for brand new, very expensive rail). Yet other parts of the city — areas with high employment and population density — will largely be ignored. To be clear, the census data is a bit out of date. But the city itself has grown faster than the surrounding suburbs — both in absolute number as well as percentages. It has grown in urban clusters, making the current approach worse than ever. The problem isn’t lack of density inside the city — although of course this could be improved. The problem isn’t lack of investment. The problem is a substandard transit system that is the result of very poor planning.

          • Herbert

            I don’t buy the line that “the grid is the best transit system there is”. First of all, it is impossible to build a metro as a grid. Second, it’s not how density and destinations work, third, cities with rigorous grids (LatAm, U.S.) tend to have lower transit ridership than cities with non-grid growth patterns in Europe.

          • RossB

            “I don’t buy the line that “the grid is the best transit system there is”.

            I don’t buy stupid straw men arguments. I never wrote that. Stop making up shit.

            The point is, I can get just about anywhere quickly in Vancouver. This is not the case in Seattle. This is because Vancouver has a grid made up of both rail and bus service. It is just one of the advantages I mentioned, and it is clearly an important one. Arguing some bizarre abstraction about grids misses the point. It sucks when your bus trip looks like this, despite going from one densely populated part of the city to another: https://goo.gl/maps/9bGJriptgsTmdhCw7. That is a forty minute ride, while driving takes six minutes.

          • RossB

            >> So Seattle should speed up redensificiation…

            Unlike most of the U. S., Seattle’s growth has largely happened in areas that were already relatively dense. The city has grown faster than the suburbs. The only cities that are close to Seattle in terms of growth are the nearby eastern suburbs of Bellevue and Redmond, both of which have good existing transit (for a suburb). Furthermore, much of the growth has happened in “urban villages” or urban clusters, which again have better than average transit. This is not the way I would grow a city, but if you want to maximize transit usage, this is the way you do it.

            Yet still it lags Vancouver, and that is largely because Vancouver’s system is just better. It isn’t the density, it is the transit network.

          • michaelrjames

            Yet still it lags Vancouver, and that is largely because Vancouver’s system is just better. It isn’t the density, it is the transit network.

            Yes, but it just shows that transit and denser urbanism reinforce each other. In the US the only way to start that virtuous circle is to begin building transit. It’ll take time, just like in LA but eventually “they will come”.

          • Alon Levy

            Except that LA’s transit modal split is going down and there’s no TOD because nearly the entire State Senate delegation from the LA region voted against SB 50.

          • michaelrjames

            Except that LA’s transit modal split is going down and there’s no TOD because nearly the entire State Senate delegation from the LA region voted against SB 50.

            Yes, mysterious. But is it really going down or just a consequence of more exurban sprawl that skews the data in a relative sense? Still, don’t you believe as they fill out the network that it will eventually starting growing?
            Also plenty of densification going on in a wide area of “inner” LA. Santa Monica is actually approaching the density of SF … And the new lines are following the densification zones (which are typically linear).

          • Richard Layman

            the comments above don’t really distinguish between monocentric and polycentric transit systems (see the discussion in Belmont, Cities in Full). Basically, SF MUNI is monocentric, BART is polycentric. Polycentric land use and transportation form doesn’t have enough density of development and service to substantively move people away from an automobile-centric paradigm.

            While it wasn’t the intent of the system design, DC is fortunate in that the core of the Metrorail station (about 31 stations of ) function monocentrically for the city as a subnetwork of the otherwise polycentric system. The stations are more spread apart than MUNI Metro in the core, but it still works. It’s only improved by the addition of biking, car share, walkability, and to some extent, scooters and e-bikes.

            http://urbanplacesandspaces.blogspot.com/2018/06/further-updates-to-sustainable-mobility.html

            http://urbanplacesandspaces.blogspot.com/2018/07/dc-is-market-leader-in-mobility-as_16.html

            FWIW, the derogatory comment about the UK, while it’s true the rail network pales compared to Germany, it’s light years ahead of the US. You can get to lots of places via the railroad network, and all the major cities. Try doing that in the US. (Granted the UK is much smaller.)

          • michaelrjames

            FWIW, the derogatory comment about the UK, while it’s true the rail network pales compared to Germany, it’s light years ahead of the US.

            It’s funny that the pithy maxim (attributed to Churchill) about the US, of “eventually doing the right thing, after having tried everything else”, applies to the (post-war) UK about transit: finally they got around to modernising LU with signalling, trains (Vic & Jubilee lines), fare cards (copying HK), finally built their version of RER only 45 years late in CrossRail, and ditto about 40 years after Paris-Lyon they are (probably) building HS2.

          • Sascha Claus

            »cities with rigorous grids (LatAm, U.S.) tend to have lower transit ridership than cities with non-grid growth patterns in Europe.«
            Interesting. And certainly, this is definitely not because European cities are generally denser or because rigorously gridded cities are mostly found in terrain that’s conducive to sprawl, so there is no need to compare gridded, sprawling US cities with non-gridded, sprawling US cities or gridded, dense US cities to non-gridded, dense US cities.

          • Herbert

            Much of LatAm does not have sprawl in the U.S. sense because the kinds of people who live in suburbs in the U.S. live in gated communities in LatAm. And because a gate is expensive and a wall likewise, they cannot be über-wasteful with land, just wasteful. A similar effect is achieved by a city wall, which of course virtually every European city had at some point.

            At any rate, what Latin American cities do do is attract shanty-towns, favelas, slums and so on. Even where geography is not conducive to them. The fact that “urban gondolas” started in LatAm is not because the Austrian and Swiss companies wanted to expand beyond the shrinking winter market into places where there is no snow (though that might play a role in some harebrained gondola-schemes) but because the shanty-towns are often up in the hills and any mode of transportation that has to follow the contours of the surface is hampered by it. Gondolas can just fly above it.

            The example you cite is of course of a lack of bus service, not the superiority of grids to non-gridded cities. Now maybe some cities in East Asia are significantly gridded and attract huge transit riderships (the fact that nobody brought that up either way is telling), but I maintain my statement that a grid is not ipso facto better for transit absent other factors. In fact, at a first approximation, there seem to be reasons to believe a grid is worse for transit. As a first sanity check attempt at an explanation there is the factor that in most gridded cities all grid-connections are open to cars and a subway cannot offer shorter distances on many connections. Meanwhile in towns where some paths are curvy or some roads too narrow for cars AND other modes, other modes can provide shorter distances. The (very much gridded at least in parts) town of Erlangen has a high bicycle mode share in part because many of the gridlike streets are one-way for cars and open both ways for bikes. Add to that a river which has a bicycle-only crossing right where the biggest transportation demand is (and no car crossing) and it becomes explainable. Bus modeshare, however, is abysmal because with very few exceptions, most of them recent (a bus lane on one of the river crossings instituted in the last five years which gave an immediate ridership boost on the lines that use it) the bus has to take even more torturous routes than cars.

          • Herbert

            Compare this: https://www.google.com/maps/dir/Neum%C3%BChle,+Erlangen,+Erlangen/Arcaden,+Erlangen/@49.5914191,10.9777039,14z/data=!3m1!4b1!4m18!4m17!1m5!1m1!1s0x47a1f8b6d88b98f5:0x7700aa0c27684007!2m2!1d10.974494!2d49.5867285!1m5!1m1!1s0x47a1f8dc267ac6dd:0xf7a5e8f37523763b!2m2!1d11.0031585!2d49.5943468!2m3!6e0!7e2!8j1589371260!3e3!5m1!1e2

            To this: https://www.google.com/maps/dir/Neum%C3%BChle,+Erlangen,+Erlangen/Arcaden,+Erlangen/@49.5917416,10.9801621,15z/data=!3m1!4b1!4m18!4m17!1m5!1m1!1s0x47a1f8b6d88b98f5:0x7700aa0c27684007!2m2!1d10.974494!2d49.5867285!1m5!1m1!1s0x47a1f8dc267ac6dd:0xf7a5e8f37523763b!2m2!1d11.0031585!2d49.5943468!2m3!6e0!7e2!8j1589371260!3e1!5m1!1e2
            And the southerly route by bus is the one with the bus lane…

          • RossB

            Yes, but it just shows that transit and denser urbanism reinforce each other.

            Of course they do. But that has little to do with why Vancouver has much higher ridership. Nor is density likely to follow transit investment in Seattle, just as it rarely follows transit investment anywhere. Oh, there is the occasional suburb that suddenly goes from a sprawling low density mess to having a bunch of big buildings, but that is a small minority of the growth that occurs, nor does it lead to a dramatic change in the landscape. The growth that is occurring in Seattle, for example, has little do with transit, and everything to do with zoning and gentrification. Areas that are close to the middle of the city became safe (and popular) while lots of other places suddenly allowed growth. San Fransisco proper, for example, grew roughly ten percent — adding around 80,000 people — and the transit everyone uses is a slow as ever (as slow as any transit system in the country).

            In the US the only way to start that virtuous circle is to begin building transit.

            Sorry, no. Zoning is far more important. Vancouver did not have great transit when they decided to loosen the zoning rules, and allow accessory dwelling units (https://www.sightline.org/2013/03/07/in-law-and-out-law-apartments/). This accounts for the greatest difference in zoning between the two cities. Otherwise, they are both ridiculously restrictive, following the failed “Grand Bargain” principle that ultimately results in boring landscapes, high rent and mediocre transit outcomes (the antithesis of Munich — the example shown above).

            Put it another way. In a few years there will be a subway roughly here (although it may be closer to the freeway): https://goo.gl/maps/hbm6RTG8fX6Nbc4L9. There won’t be subway stops anywhere close to here: https://goo.gl/maps/rHcE52SHV2ptbSZy8, here: https://goo.gl/maps/41UdNnaornEEfea79, here: https://goo.gl/maps/7Ea2TaKjum7zeKSd9 or here: https://goo.gl/maps/2WBPh9f8sZq1v1HT9. Which place, in thirty, or for that matter, fifty years, will have the most people? I don’t know, but I would be willing to bet that the place where was a huge investment in transit will have by far the least amount of people.

            You have it backwards. Start with the zoning, then add quality transit. (Seattle needs to do both).

          • RossB

            While it wasn’t the intent of the system design, DC is fortunate in that the core of the Metrorail station (about 31 stations of ) function monocentrically for the city as a subnetwork of the otherwise polycentric system.

            Yes! That is the key. A strong monocentric system. Once you have that, you can add polycentric niceties on top. Of course it is a good thing that travel from the suburbs into the city is so handy, but only because the city itself has really good transit. Likewise, having a fairly fast and fairly convenient connection between Baltimore and D. C. is also great. But neither represents the bulk of transit trips in the city, simply because no one in Baltimore is going to take the train into D. C. on a whim, nor is anyone from Huntington. In contrast, if you are in the city, these sorts of trips happen all the time, just as someone in the suburbs thinks nothing of hopping in their car and driving somewhere.

            It reminds me of this great comment, written a while ago from a friend of mine: https://seattletransitblog.com/2013/02/14/news-roundup-geeks/#comment-292594

            Seattle is spending billions (54 to be exact) building a polycentric system, while ignoring the core, in a city that is basically nothing but core. Seattle is not L. A., nor is it Phoenix. Yes, it sprawls. It sprawls like crazy. But unlike Phoenix, there is a reasonable amount of density within the central core. Unlike L. A., there is no density outside it. It is really striking — you can see the northern city border of Seattle (proper) just by looking at the density maps: https://arcg.is/1W9WXz (again, keep in mind, Seattle proper has grown faster than the suburbs). In contrast, I can’t even find Phoenix in what is largely low density sprawl: https://arcg.is/1qC10n. L. A. is similar to Phoenix in that it sprawls, but the big difference is that it actually does have plenty of (relatively) high density pockets. The problem is, it is all over the place, from Encino to Santa Anna (and for a city that size, is never that densely populated). Seattle doesn’t have that problem. Seattle doesn’t need to serve its suburbs with anything more than commuter rail and good connecting service to the subway. Yet they will. They will spend billions pretending that Fife is the same as Ballard.

          • Coridon Henshaw

            RossB — Vancouver’s suburban bus service isn’t US-level bad but it’s far from “good.”

            Pre-COVID, many suburban trips required non-timed connections between infrequent bus routes that ran had on-time performances of 85% or worse. As a result, the average speed for many journeys was four times slower than driving.

            Post-COVID, many suburban routes have simply been shut down and probably won’t be coming back. Translink has been offsetting farebox losses by taking (stable) tax-sourced funding from suburban areas without providing any service in return. This will likely have consequences for suburban willingness to fund transit projects going forward.

          • michaelrjames

            RossB: “You have it backwards. Start with the zoning, then add quality transit. (Seattle needs to do both).”

            Perhaps you are correct. That is precisely what SB50 & SB827 were all about. But they failed. I suspect that as long as such measures appear (to pollies and joe public, certainly suburban joe) to be all encompassing only needing some vague connection to transit, it will always be perceived as a threat and never get passed. One would think–perhaps naively–that a more specific plan, say that showed all likely potential zones might allay some fears? Just a little. That is why I enumerated the maths on hi-density TODs and how they don’t need to be as giant and certainly not high-rise to succeed, and to support both walkable neighbourhoods and transit. It can add a million residents with disturbing very few existing residents.

            Incidentally, I don’t really get the arguments about polycentric versus monocentric, since regional rail like RER, or BART or DC Metro do both. Surely BART (esp. its original remit) was explicitly to do that, because of geographical constraints on its main centre?

          • Herbert

            If polycentric networks are BAD ™ then what is a town like Erlangen to do that is a fifth the size of neighboring Nuremberg (100k vs 500k) but has more people commuting N->ER than in the other direction?

            A system that only serves to get people to Nuremberg main station (such as Nuremberg S-Bahn) is going to leave the people who commute from Nuremberg to Erlangen wanting. Thankfully many of them are university students who’ll put up with shitty bus service, but for those who work at Siemens a new rail service is needed. And that’s exactly what http://www.stadtumlandbahn.de is doing.

            Is that a bad thing because it’s not monocentric?

            What about ring lines? Are they evil because they move some trips away from the center?

          • Alon Levy

            The thing to understand is, American cities that claim to be polycentric aren’t. They’re just weak-centered. Los Angeles, for example, has a ton of little job centers strewn all over the region, with 20,000 jobs here and 30,000 there. Seattle is fairly unique in having big suburban job centers, all around the same corridor between Redmond and Bellevue with a total of 160,000 jobs, but that’s over a fairly broad area (about 22 km^2), and the biggest job site, Microsoft, is 50,000 jobs. Serving all 160,000 jobs there is hard, because many tend to be on the wrong side of a freeway. In contrast, the broader area of Downtown Seattle including the university is 460,000, and a narrower 10 km^2 area is 300,000, and nearly all of the 300,000 are on the same side of the downtown freeways so a few train stations can serve all of them.

          • RossB

            RossB — Vancouver’s suburban bus service isn’t US-level bad but it’s far from “good.”

            Sure, but their urban bus system *is* good. It is far more important to have a very strong urban transit system than a suburban one. Build the best suburban transit system in the world and if your urban system is crap, very few people will ride it. Do the opposite, and you have something similar to Vancouver (at worst).

            Incidentally, I don’t really get the arguments about polycentric versus monocentric, since regional rail like RER, or BART or DC Metro do both.

            There is nothing wrong with doing both, but if you don’t serve the core, you will fail, miserably, as BART did. That is my point.

            Surely BART (esp. its original remit) was explicitly to do that, because of geographical constraints on its main centre?

            No, BART was designed mainly to appeal to suburban drivers. It was not designed to serve the urban core. There is 175 KM of track. It was extremely expensive to build. Yet there is only one line in San Fransisco. There are only 8 stations in San Fransisco, 12 in Oakland/Berkeley. Could you imagine if the New York Subway system had 8 stations in Manhattan, and 12 in Brooklyn. It would be disaster, and no improvements in the LIRR, Metro-North or New Jersey Railroad system would make a bit of difference.

            Oh, and what geographical constraints?

          • michaelrjames

            RossB:

            Surely BART (esp. its original remit) was explicitly to do that, because of geographical constraints on its main centre?
            No, BART was designed mainly to appeal to suburban drivers. It was not designed to serve the urban core. There is 175 KM of track. It was extremely expensive to build. Yet there is only one line in San Fransisco. There are only 8 stations in San Fransisco, 12 in Oakland/Berkeley. Could you imagine if the New York Subway system had 8 stations in Manhattan, and 12 in Brooklyn.

            I am just so completely on a different page (and maybe stupid, certainly culturally blind to these arguments). Of course it is for suburban drivers–I keep saying it is in the name “BA”. It may have failed to build as much ridership as one would like (but in addition to whatever faults it has, how can you not lay a lot of blame on San Mateo and Santa Clara who blocked it from their areas for … 40 years? Hillsborough completely refused to have BART on their zone but luckily their suburb didn’t extend to the water, and the hypocrites got a BART station literally on their border.)

            There is nothing wrong with doing both, but if you don’t serve the core, you will fail, miserably, as BART did. That is my point.
            You think BART doesn’t serve the SF core? It goes thru the middle of it. I wonder what fraction of business would be within 800m of a station? That’s as far as Broadway (ie. most of financial district), and covers almost all the hotels district-Civic Centre and all of Soma including Moscone centre etc. OTOH the old CalTrain station is in the boondocks (literally, next to China basin?) and must be kilometres from all this (except Moscone?). I’m all for SF building a Paris-style Metro (though it is one quarter the density if similar surface area) but that was never BART’s remit.

            only 8 stations in San Fransisco, 12 in Oakland/Berkeley
            This is exactly Alon’s complaint about Paris’ RER. I just don’t understand it. For thru-running regional rail those number of stations seem quite fine to me! Again, it really means that you want the money spent on local Munis–fair enough–but wouldn’t you still want regional rail to link up the sprawled Bay Area? Unless you think existing CalTrain is adequate? (I’m perfectly in agreement that CalTrain and its ROWs–though ultimately on land controlled by SP I believe?–should have been the basis of such a system but there was total obstruction from the usual suspects that can’t just be magicked away.) I mean Paris’ RER lines intersect with almost every Paris Metro line so I don’t see what more one could wish (ie. other than more such Metro lines in SF). From my own limited time living in SF (but not using transit much at all–walking and cycling) I thought BART’s connections to Muni were pretty good, and it’s a small city. BART simply wasn’t created to make local transit (and the other counties had to be convinced they weren’t just being conned into funding SF) yet that seems the main complaint.

            Incidentally one can see how if the Marin link had been built, that would have covered quite a bit of west and middle suburban SF, but surely at even huger cost? I wonder if anyone ever seriously expected that to be built; it required double-decking the GG bridge! I suspect everyone knew it was an unrealistic fantasy and that is why Marin pulled out so early. (But I don’t know the deep history. Marin is richman territory and most like Hillsborough, possibly vociferously against paying taxes for transit; they may have hoped that “others” would use the transit so the roads/bridge were freer for them to drive. Hillsborough doesn’t even have sidewalks, and not a single bus route–on my decades old SamTrans map.)

            imagine if the New York Subway system had 8 stations in Manhattan, and 12 in Brooklyn
            A very inappropriate comparison on many grounds, so instead ask: how many stations within those boros does LIRR have?

            As to geography, surely the long linear sprawl around the bay is something a little special? NYC may be on islands but all affordably bridgeable or with population to justify lots of tunnels. Ditto Seattle. But the more appropriate comparison may be with the very long eastward sprawl of Long Island?

          • RossB

            American cities that claim to be polycentric aren’t. They’re just weak-centered.

            Yes, that is what I was getting at when I contrasted Seattle with Phoenix. If you look at a census map of Phoenix without labels, you can’t tell where the city. The same is true of L. A., even though it has a lot more pockets of density. Cities like that are extremely hard to serve. You have to spend a lot of money and do things really well before you get decent transit share. But like I said, Seattle isn’t like that. For that matter, neither is San Fransisco. While it too sprawls and sprawls, it is pretty clear where the center is, and where the near-suburbs are. That is just by looking at the census maps,employment data follows a similar pattern (but in my opinion are less important — not every trip involves work).

            In contrast, the broader area of Downtown Seattle including the university is 460,000, and a narrower 10 km^2 area is 300,000, and nearly all of the 300,000 are on the same side of the downtown freeways so a few train stations can serve all of them.

            But they won’t. First Hill (https://goo.gl/maps/xDgS5L55YXBqtqdx5) which is a major medical center, with tons of jobs and plenty of people, will not have a station. They were supposed to get a station with the first line, but then, when things got tough, they skipped it. Likewise, with the latest proposal it was considered too much effort. Instead the new line will have stations right next to old stations (someone here can feel free to defend that choice). This means Belltown (https://goo.gl/maps/XbjXpYrepD42P6KX7), for example, won’t have a station. Instead they will be busy building stations in Ash Way (https://goo.gl/maps/GikoMppu5zVRWGxcA), Issaquah (https://goo.gl/maps/XjbG4NzzLFJfLvx49) and my favorite, Fife (https://goo.gl/maps/uVbMZXDuJ23U3Hyy5). Those pictures aren’t the worst parts of those neighborhoods/cities. They are where the stations are being placed (next to the freeway, abutting a greenbelt, etc.) and most of all, very, very far from the (obviously neglected) urban core.

      • Fbfree

        From what I see, Seattle’s transit advocacy today is excellent, despite the complicated political environment in the US context. It hadn’t always been, thus they are starting at a disadvantage in comparison to some Canadian cities.

        • Brendan Dawe

          by contrast, Vancouver’s transit advocacy is dreadful. Lack of technically informed commentary that is very missing from other city’s conversations. Vancouver can be reduced mostly either transit-boosters who trust in Translink’s analytical and planning capacity or cranks who are mostly into cute trams

          • Sascha Claus

            Everything these evil Americans are doing to them? Short cars that have at most the capacity of an articulated bus, crawling in traffic where room for spearated tracks is available, …?

          • Brendan Dawe

            Usually some iteration of “rebuild all the BCER streetcar lines as they were in substitute for actual rapid transit” or “build a circumferential tram along a disused industrial line along the south side of false creek as a substitute for rapid transit” or “rebuild the tram to Chilliwack which is the one true transit as a substitute for rapid transit”

          • Herbert

            How did someone once put it?

            “A tram is better than no subway”…

        • Nilo

          In fairness I’m kinda into cute trams, too bad everyones just accepted the need for a Giant TBM under broadway. Just dig it up! It will suck for two years, but after that Translink can take the B99 money sinkhole into other bus service!

          • Brendan Dawe

            ya, I’m not one to disagree on cut-and-cover. the merchants will get it anyway from station excavation.

            But a point worth reinforcing is that the 99 is not that much of a money sinkhole – it’s operating cost ~$0.60 per boarding, quite far below the average far, and lower than the cost per boarding for the E&M skytrain lines. It turns out that operating buses that are supernaturally full all day long is operationally profitable It’s a resource hog, but it’s not a money sink.

          • Nilo

            Well at least you’ll have a lot more drivers to spend on bus service somewhere else!

          • Herbert

            If you can replace three buses with one tram, you can cut or put to better uses two workers. And there’s the added benefits that trams have a longer expected life, are cheaper to maintain (most of the stuff that wears down is cheap steel) and you’ll almost certainly get a ridership boost.

            So why not replace the busiest bus routes with trams?

          • Henry Miller

            Replacing 3 buses with trams can be good or bad, depending on if you can maintain frequent service. A anything every 15 minutes instead of a bus every 5 is a big loss. However if it is a bus every 3, or a bigger tram every 5, you are probably okay (note that mathematically this isn’t a 3 for 1 trade – what else does the tram give) .

            There are many details. If I have to spend several minutes getting underground instead of being at the stop the tram may need to be every 2 minutes to be as good as a bus every 5. Or… Insert every detail anyone has or will come up with, then figure out what works for the situation.

          • Herbert

            Trams aren’t underground. If they’re underground they’re subway or Stadtbahn or subway-surface etc.

            And running a tram in a Takt worse than 10 minutes during the day raises the question “why run trams at all?” Unless we’re talking about an interurban line…

            At any rate, bunching occurs somewhere around a five minute Takt for the bus. And a bunched bus isn’t a Takt but one full bus, two empty buses at three times the Takt interval… Replacing that with a tram might sound like worse service on paper but it’s giving riders one reliable vehicle that isn’t overflowing at reliable intervals. And then of course, one can change the Takt to a tram every five minutes…

          • fjod

            Herbert, I think you’re struggling to find a good translation for Stadtbahn because these services are conventionally called trams in English and many other languages.

          • fjod

            You’ve linked an article that refers only to Germany and Austria. Note that the more general article for premetros explicitly states that they’re trams.

          • Alon Levy

            Stadtbahn is literally subway-surface. Turn-of-the-century Boston and Philadelphia built cut-and-cover tunnels for some streetcars, running in a subway in city center and then branching on the surface. So did many postwar German cities (Frankfurt, Cologne, Dortmund, Essen…) and so did San Francisco. Tel Aviv is mistakenly building the Red Line this way.

            I bring up the branching aspect of it because it’s pretty important – the point of a subway-surface line is to have faster, higher-capacity service where there is the most demand, and slower, lower-capacity service elsewhere. So to concentrate service where it is needed, you should branch. If you’re not branching, like Tel Aviv and Vancouver, you’re better off building full driverless metro. Pure surface trams can also work but only in much smaller cities – Tel Aviv right now is the biggest non-US first-world metro area with neither a subway nor an S-Bahn.

          • Henry Miller

            Herbert, if you run the whatever (tram, subway, magic 500 passenger bus..) at 5 minute headways, then the less employees you need to run it the better.

            I think in the US most politicians will read your proposal as replace a bus every 10 minutes (already marginal frequency) with a tram every 30.

          • Herbert

            A tram every 30 minutes is no way to run a tram.

            The expensive parts are mostly the rails and to a lesser extent the rolling stock. Once you have that, the cost of adding service (i.e. labor and electricity) is cheap…

        • RossB

          I live in Seattle (and visit Vancouver on occasion) and my guess is that the transit advocacy is very similar. Inside the core of the city, they want transit, and are willing to pay for it. The problem is that outside that — in the suburbs — people are more fickle. Seattle has had its share of ballot box failures, but they learned their lesson. If you want to pay for a region wide transit system, you need to have the vote on a general election. That is because left leaning voters show up in the general election, while the often skip primaries and special elections. Seattle also had their last big ballot proposal right after they built what they should have started with — a line from the University of Washington to downtown. The agency undersold this, even though anyone with a brain knew it was going to popular. People were gaga over it, saying things like “OMG — it is so fast”. Ridership was much higher than the agencies (low ball) estimate.

          In contrast, the Vancouver proposal failed in part because Vancouver rarely votes on proposals. It is also a more mature system, and they had maintenance problems. There were plenty of people who voted against the proposal, just because they were pissed.

          Of course that is irrational, but voters are stupid (entire books have been written about that). I remember as a kid, watching as Seattle became one of the first (and the largest) cities to racially desegregate without a court order. At the time, most funding came from levies. The levies failed because (white) idiots were pissed about having their kid spend a little extra time on the bus, and actually learn a thing or two in the process about a major subculture (the most influential subculture in the world). So they doubled down, and decided to vote against the levy, thus ensuring that funds were cut, and good teachers would be laid off.

          The point is — voters are idiots. I don’t think there is that much difference between the cities in terms of their attitudes towards transit, it is just that Seattle knows this, and Vancouver didn’t (or simply had bad luck).

    • Nilo

      Yeah but Seattle proper has seen an almost 10 point shift away from drive alone within the last decade. Though a huge portion of that is walk to work, which obviously isn’t a sustainable source of growth forever. Similar drops have been seen in Bellevue and Tacoma

        • Eric2

          Because tolerable walking distance is very short (maybe 15 minutes for the average person) and you can only fit so many people in that range without 100-story skyscrapers.

          • Eric2

            Because people have multiple locations they need to visit – their work, spouse’s work, kids’ schools, family, friends, religious or other community meeting places, etc. The chance that you live within walking distance of all of these is pretty low.

          • Herbert

            How did people do all that before the invention of bicycles and railroads?

          • michaelrjames

            Because tolerable walking distance is very short (maybe 15 minutes for the average person) and you can only fit so many people in that range without 100-story skyscrapers.

            Fifteen minutes walking at a quite slow pace covers 1.5 km. But actually some transit experts say about 800m is the limit, so let’s see what such sizes can hold:

            In a circle of radius 800m at a density of 15,000/km2 (about half that of Manhattan or Paris) could house 30,000 people; as the Paris example shows this does not have to be high-rise but only 6-7 floors (and I have deliberately made space per resident at ≈2x so bigger apartments, even subterranean parking). It would be a totally walkable zone that itself would be a centre of jobs, with walkable schools (imagine that!), shops, clinics, cafes, restaurants etc.

            A TOD of radius 1500m, at the same density, would house 106,000 residents, ie. an extra 76,000 more than the inner 800m circle. This is similar in size to Paris’ very middle class 15th arrondissement that has 232,000 residents at density of 27,300/km2.

            No high-rise required, let alone 100 storey buildings. And consider some of the US’s sprawled big cities. Just ten TOD’s could hold a million residents–and have a lot of local jobs too–yet occupy just 70km2 which is, for example, just 2.5% of Houston’s 2,750km2. It doesn’t have to disturb 97.5% of current Houstonians but is a means of absorbing some of their predicted growth.

          • Eric2

            That’s great! In your 800m radius, at Paris densities and 6-7 floor heights, one could house 60k people. Basically a quiet town, with a small hospital and maybe a community college.

            Unfortunately though, we were talking about the Seattle area with a population of 4 million. To fit 4 million people in that radius, buildings would have to average 430 stories tall!

          • Herbert

            Or you could build like 70 of those “hyperdense circles” and link them with a couple of rapid transit lines…

            Which, oh yeah, is pretty much the proposal for “city of 15 minutes” by Anne Hidalgo, mayor of some small town in Northeastern France… P-something, what was the name again?

          • Vire

            I’m not sure how to specifically reply to Herbert, but on “why do people have to live far away from where they work?”: Do you want for people to live next to a metal refinery? Or a machines factory? Or a power plant? No matter how cities are developed there will always be necessity to put places of work farther away from where people live.

            What’s more, what if someone changes jobs and their new workplace is further than 15 minutes away (if we’re going with Hidalgo’s proposal as an example). Should they be expected to move every time their working location changes? I personally think this is a ridiculous expectation, so cities should always be developed to be dense, and even locally be developed as 15-minute cities, but there should always be strong transit connections and developing those connections should be one of the primary concerns.

          • michaelrjames

            Unfortunately though, we were talking about the Seattle area with a population of 4 million. To fit 4 million people in that radius, buildings would have to average 430 stories tall!.

            Now you’re being facetious. As Herbert points out, that is not what is being proposed or what modern TOD is. (Old TOD is where the entire city is like it, ie. Manhattan, Paris etc). These cities are still growing but in an ever outward spiralling sprawl with everything further and further apart, more and more driven miles etc. TOD is the perfect way to cope with some, possibly a lot, of this growth, and simultaneously build transit far out in that sprawl. The low-density sprawl is poor at supporting transit but if you have a TOD of ultimately 50k-100k or more, then voila. In fact such a TOD would also attract a certain proportion of those living in ‘nearby’ sprawl. It’s a version of Paris’ RER or Berlin S-Bahn or Tokyo’s regional rail etc. The model is often called “pearls on a string” with TODs spaced 3 to 5km apart, and like pearls amongst the low-density sprawl in between.

          • Nilo

            Herbert if you want to know how people dealt before transit, friend of the blog Jonathan English did a much better job than anybody here of discussing it.

            https://www.citylab.com/transportation/2019/08/commute-time-city-size-transportation-urban-planning-history/597055/

            Anyways you can’t fundamentally have everyone walk to a job because cities are job markets, and people aren’t going to move every time they change jobs. To take an extreme example, imagine a city where everyone is employed in a job next door, so their commute is zero. But when they need to find a new job there’s actually a relatively small chance that said job will be within a walkable radius. Thus the argument for job concentration, job density, and radial fast transit.

          • Herbert

            We’ll have a lot fewer coal and gas and so on power plants in the future…

            And of course not everybody CAN live next to their job, but ceteris paribus everybody WANTS TO live next to their job. If you could get the same house for the same price without the commute, who WOULDN’T take it?

            And we seem to be forgetting that work trips are only a small share of overall trips. Kids going to school for example can follow demand – instead of big box schools that are only accessible with motorized means, you can have the neighborhood school – at least if you have “one school for all” rather than the strange three-headed-cerberus that is the German school system in most but not all states…

            Of my daily trips that aren’t work related, there is none that takes me more than fifteen minutes one way. I mean I COULD go to the supermarket far away, but why would I? As a matter of fact, big box stores are horribly inefficient, as Walmart had to learn when they tried to expand to Germany and got beat to hell and back by Aldi. The “Aldi principle” seems to work even in the U.S. which in my ears is good news for countless “food deserts”.

            Of course the cutesy “eight floors with the ground floor for shopping” also enables the shopkeepers to have a stairs-commute and many of the shoppers can get what they need on a whim while walking…

      • Richard Layman

        It really helped that the light rail got extended to areas conducive to transit use, specifically Capitol Hill and almost to the UW campus (part of the next extension). As long as you extend outward (polycentric) it’s hard to generate significant mode shift. Reports from UMN Center for Transportation Studies find the biggest return in ridership comes in the core of the system. This makes sense, as the core tends to be the most dense, with the greatest possibility for proximate connections.

        – The Hiawatha Line: Impacts on Land Use and Residential Housing Value (Feb 2010, CTS 10-09)
        -Transportation as Catalyst for Community Economic Development (Dec 2007, CTS 07-07)

    • RossB

      “I imagine that Seattle would be the US counterexample to this, where long-term planning, significant investment, and strong local advocacy all work together. ”

      Seriously? Seattle is a poster child for mismanaged American transit. The rail system is extremely expensive, and full of dubious projects (https://www.thetransportpolitic.com/2016/04/06/youve-got-50-billion-for-transit-now-how-should-you-spend-it/). There is large scale investment in brand new (expensive) suburban rail, with stations way too close to the freeway. There is little in the way of far more cost effective urban infrastructure. Even when they decided to run an additional tunnel through downtown, they ran it essentially right next to the other tunnel, missing the opportunity to add coverage downtown. Stop space is terrible in the urban core — they were fixated on distance, not quality stations or connections. They ignored bus integration, as well as other basic ideas that are obvious from looking at their nearest neighbor (Vancouver BC) let alone overseas. They did borrow one stupid idea from Vancouver BC though — build things out of order. Just as Vancouver is dragging its feet building an UBC extension to the SkyTrain network, Sound Transit still hasn’t completed what will clearly be the most cost effective, popular section — the line from the University District (of Washington) to downtown. I could go on.

      This is all pretty obvious when you look at transit systems that work. If the transit agency simply asked independent consultants to come up with a plan it would be better. But the agency is made up of politicians who know little about transit. There is not a single transit expert on the board. The only person with any transportation expertise is the Washington State Department of Transportation head, who has no public transit expertise either (WSDOT largely focuses on building and maintaining highways in the state). They largely ignore the meetings, since the state capital is in Olympia, but the meetings are in Seattle (about 100 KM apart, with little in the way of fast public transportation between them).

      As politicians, they all have more important jobs (mayor of Seattle, or head of a large county). They don’t have time or interest in understanding the key features of good transit. Worse yet, the don’t realize this, and fail to delegate. That is because, as Alon wrote above, “The central conceit is that public transportation is not cars. It’s a different, more complex system.” If you think of the rail system like a freeway, then what Seattle is building makes sense. Of course you want it to be really long — so you can go farther. Extra stops are nice, but you can always work your way back. Except it doesn’t work that way, and the results — even after spending a fortune — will be terrible. The region will still be largely dependent on the buses, and many of them will continue to be stuck at traffic lights, if not traffic. Mode share will still be really low because most trips that people take will still take a long time with transit.

      • Herbert

        So Germany with few subways to begin with and short interstations (there are exactly two stops on the Nuremberg U-Bahn that are more than a km apart, I think similar things are true for Munich, Berlin and Hamburg) is doing a better thing than China or Mexico with long distances between subway stops?

        • RossB

          I have no idea. I would have to look in detail at the various cities to offer an educated opinion. From what I can tell, though, Mexico City has way more stations per distance than Link. It looks pretty ordinary, in that there are lots of stations within the core, as well as lots of overlapping lines. There are only big distances between stations as you go farther out, in what is one of the biggest cities on the planet.

          In contrast, Link follows the BART model. Not many lines in the urban core, and huge stop spacing for all but a few downtown stations (that in the case of Link, were built before Sound Transit had a chance to screw them up). BART is an American experiment, and like just about all American transit experiments, it failed. Ignoring the central core while building to distant, low density cities is a bad idea. Even the agency acknowledges that most of the ridership will come from the city (https://seattletransitblog.com/2019/01/30/link-riders-2040/) even though the big gaps between stations would tend to favor more distant riders. If they had more stations — and additional lines — in the city, it would be a much better system. Those in the suburbs would ride the bus a little bit longer to their station (which would be just fine, since the stations are all close to the freeway, and the freeways have HOV lanes). Those in the city save a considerable amount of time getting to their destination. Ridership from the suburbs might go down a little (although probably not that much, since suburban riders would have a lot more destinations within Seattle that are easily accessible). Ridership in the city, meanwhile, would go way up, and more than makeup for the loss of suburban ridership.

          Favoring distance over quality urban stops is common in America (BART, DART, RTD Light Rail) and it always ends the same way (with poor ridership and especially bad ridership per mile). This often results in the bad frequency ridership spiral, as agencies can’t afford to run largely empty trains deep into the suburbs. Large stop spacing and long distant rail may make sense for a city like L. A., which is huge, and has no major center. But it is ridiculous for Seattle, where density is almost exclusively within the city (and the eastern nearby cities of Bellevue and Redmond).

          • michaelrjames

            Dog knows I don’t want to inflame the old arguments but for its remit and sole reason for existence, BART didn’t fail. It is regional rail and hence the BA in its title. It is not SF Muni and was never intended to be. It was to provide alternatives to really long and slow car rides in the strung-out Bay Area, not to the inner city. You are really saying that San Francisco transit has failed but that has little to do with BART (except I suppose one could argue they have parasitized the precious transit dollars). That is certainly not to say BART could not be better etc etc. (And it was hamstrung by the withdrawal of San Mateo and Santa Clara in the beginning, the areas that could most use it.)
            Also it is madness to compare ridership on a per mile basis; naturally sprawl will always look terrible on that basis but it is totally at odds with the objective. It’s precisely saying that Paris should never have bothered building its (very expensive) RER system because its performance of ridership per mile will always be far “worse” than the Paris Metro–nevertheless it carries about 1.3 billion pax pa, some from more than 50km in the suburbs.
            This falls into the category of econocrats measuring what is easily measured then assigning great all-important significance to that metric.

          • Nilo

            Michael this is why nobody listens to you. BART is a failure even compared to its American companion in WMATA, much less compared to an international contemporary like the Munich S-Bahn. BART discarded all the core lessons from other S-Bahns: leverage extant rail lines for cheap ROWs, build parking only at the ends, and focus stops on islands of walkability. BART instead built almost immediately outside the core with massive amounts of parking (North Berkeley) focused on terrible hostile freeway alignments (Contra Costa Line, MacArthur station), and because of the gauge choice had to rebuild instead of reuse rail in alignments on the South Alameda, Richmond, and later SF Airpot Line.

            BONUS: The complete disinterest in using standard gauge, means incorporating Southern Pacific’s ROW (now caltrain) with its faster Peninsula and South Bay to SF ROW, that ends a mile short of BART’s four track mainline tunnel impossible! What a great idea.

            Please Michael people like Richard and I have lived in the Bay Area, don’t pretend the shit we have to live with is actually good.

          • michaelrjames

            Yeah, and you and Richard won’t listen to a word of different opinion. I didn’t say BART was wonderful but you completely ignore that it got built ( a big deal in the US!) and has reasonable ridership and at least is partly successful in its remit (despite all the obstructionism, some possibly from people like you who want to hold out for perfect, or those who think the money should have been spent on Muni) which is to serve the sprawl. It’s not bloody SF Muni, and part, if not all, of the reason for the peculiar choice of gauge was to avoid shit-fights with all the existing obdurate rail domains and their refusal to co-operate, like you see in the US. And certainly not to share its ROW with Muni or freight or shit existing regional rail–Caltrain. Despite all this it is a rare example of “successful” regional rail in a US big city in the modern car age.

            As I have pointed out, the ridership shows it (sure it could be better so go howl at the moon for perfect):
            BART: 175km at 674k pax/km compares reasonably with Paris RER lines C (754/km), D (763/km) and E (1,100/km), which is pretty remarkable given it serves American car-dependent low-density sprawl while suburban Paris is much denser. And NYC’s nearest equivalent, LIRR is far worse while having a much higher population catchment: 513km with 89.3m pax (2016) = 174k pax/km about 26% of BART.

            Try to get some perspective. For all its imperfection it simply isn’t the failure you make it out to be, or maybe ‘deserves’ to be (but since it broke all of your rules I guess that is a difficult admission to make, that it could even be a passing success). And at least its planners didn’t commit the sin Alon describes in this article: they didn’t think small or start small.

          • Eric2

            “BART: 175km at 674k pax/km compares reasonably with Paris RER lines C (754/km), D (763/km) and E (1,100/km),”

            So it’s worse than all of those, even though those mostly consist of preexisting rail with little capital investment, while BART consisted of massive new investment. And all these lines extend far out into the countryside, where they serve towns surrounded by forest (low overall density). And RER D is only half a line (it shares its core segment with RER B) so its figures should be doubled for accurate comparison. And IIRC not all RER trains run to the end of the line, so the the equipment+labor expense is lower than ridership per km would predict.

          • michaelrjames

            Yes, BART is somewhat worse than those RER lines (note A & B are another universe but they are exceptional even in world terms) but don’t tell me you aren’t a bit surprised (as I was) that BART isn’t that much worse? For C & D it is only 10% worse! Paris is a much bigger megalopolis (12.4 million) and is much denser (in reality a series of merged towns/TOD) with a long legacy of transit usage.

            It’s RER-E that is “half a line”, currently terminating at Gare St Lazare but soonish to be extended 50km west and take over some of RER-A’s western route. RER-D shares a short tunnel and track (≈2.4km, Chatelet-Nord) with B but the lines are otherwise quite independent. They don’t share platforms even at Chatelet or G. du Nord.

          • Herbert

            Have a look at some German S-Bahn stops… They’re not all the ideal, either.

            https://www.google.com/maps/place/Dresden-Zschachwitz/@50.9889722,13.842527,770m/data=!3m2!1e3!4b1!4m21!1m20!4m19!1m6!1m2!1s0x4709b893aa2c223b:0x12f6344bdba79cc9!2sHeidenau-S%C3%BCd,+Heidenau!2m2!1d13.8738134!2d50.969316!1m6!1m2!1s0x4709b8e70588bbcf:0xe4d677ae67c7c987!2sHeidenau-Gro%C3%9Fsedlitz!2m2!1d13.8945028!2d50.9609179!2m3!6e0!7e2!8j1589371260!3e0!5m1!1e2 This is a Dresden S-Bahn stop

            https://www.google.com/maps/place/Vach,+Bahnhof/@49.4993145,10.9651577,2674m/data=!3m1!1e3!4m5!3m4!1s0x47a1fe4a13c18d8d:0x6219f88f8aa02810!8m2!3d49.5176815!4d10.985821!5m1!1e2 this is a Nuremberg S-Bahn stop…

          • Alon Levy

            A couple replies re this subthread:

            1. I feel weird about that BART vs. WMATA comparison. The ridership gap is a lot smaller now (well, pre-virus) than it was when Christof Spieler first wrote about it in 2006, and the work trip modal split for transit is nowadays higher in metro SF than in metro Washington. In the 10+ years after Christof wrote that article, San Francisco had a lot of commercial TOD in and just outside the Financial District, while the Washington Metro instead collapsed due to poor operations, coming partly from the stress of the Silver Line but partly from generally poor management.

            2. The BART vs. RER comparison misses something important: in the extent of tunneling and urban complexity, BART is firmly RER A. It has even more newly-built underground stations, it has a longer overall tunnel segment, and it spent a lot of money to get into city center, if still less per kilometer than the RER A because BART was Downtown San Francisco’s first tunnel and the RER A was Chatelet’s sixth. It matters that for all this complexity, BART’s ridership per km is slightly less than that of the RER C and D, which were built on the cheap (no new underground station) and which Parisian railfans mock as not real RERs because of their poor service.

            3. That said, for the most part the problem with the original BART system isn’t the routing. There was and to some extent still is a problem in that East Oakland can’t get a train to both Downtown Oakland and San Francisco without reverse-branching, but in the grand scheme of things, that’s not BART’s main problem. The BART extensions are kind of bad, e.g. the Daly City route was reasonable but it was mad to extend it to Millbrae around the wrong side of the mountain when Caltrain is right there on the right side. But the real problem is parking lot suburban stations. Outside a handful of nodes like Downtown Berkeley, the density around BART is shit, the BART planners like the parking structures on the theory that you can fit more parking spaces than apartments in a prescribed radius, and there’s no integration with buses even though buses could serve these stations with timed transfers if local planners cared. North Berkeley is a travesty, Glen Park and Balboa Park are travesties, Daly City is somehow even worse, and even the Mission is underbuilt relative to service quality and the YIMBYs preemptively surrendered to local NIMBYs and excluded it from SB 50.

            P.S. No, you shouldn’t double the RER D’s figures, at most you should add up the RER B and D on their shared segment.

          • michaelrjames

            2. I have no problem with your description, but wonder about the implication that the long and huge Market street segment was an error. It was very costly but, like RER-A & B tunnels and the massive Chatelet-Forum excavation, hasn’t it proved its worth? It serves the whole major spine of San Francisco. I don’t know what alternatives there were–especially given it had to cross the bay–unless I suppose those who think Caltrain tracks and ROW should have been used. But that just leads us to the main contentious issue: the obstruction by those with power over Caltrain, not to mention the necessity (but reluctance of Caltrain) of removing all the at-grade crossings on its peninsula route, still a problem 50 years later. Plus apart from that, the Market street solution is still superior (in terms of serving SF; for serving the peninsula maybe arguable, but again remember San Mateo (at least critical areas within it) and Santa Clara were against BART’s plans. Then the issue of sharing the bay tunnel itself! You complain about RER A & B sharing a 2.4km tunnel in Paris but this would have been a much more serious operational compromise.
            I don’t know why the high cost is such an issue, for BART or RER-A, ie. from the perspective of today. It’s not like back then they were due to incompetence or corruption, simply very ambitious schemes. To me they have amply justified those decisions. BTW, the deep three-level Market street trench serves Muni exceedingly well too and I can’t remember if the city or Muni paid anything. I am supposing certain federal or state grants made it conditional; whatever, that was also a far-sighted decision and worth the extra marginal cost at the time, and probably by itself justified the route versus any Caltrain east alternative. In fact it is difficult to imagine that any ancient crayon maps of a future metro map of SF didn’t use Market street like how it eventually happened. It’s the natural choice just like Metro 1 used the central axe of Paris.
            Again, about the costs of RER-C and D, of course C needed very little expensive construction (just that bit, less than 1 km? of riverside tunnel to connect the existing west and east suburban rail already under the quays). RER-D actually needed an awful lot of expensive deep tunnel in Paris so it cannot have been that cheap (and probably why they chose not to duplicate the B tunnel to G. du Nord). But I am not even interested in the historic costs of any of these lines (it’s all sunk and ancient) but with performance, and despite complaints with both C & D they do the job, carrying probably about 350m pax (it was 300m 15 y ago) from the distant banlieus thru the heart of Paris. (C even does what you complain about the other RER lines: has lots (9) of stations within Paris … which slows it down!).

            3. Sounds right. I suppose it’s disappointing that those BART stations didn’t generate dense TOD around them but that too will be political and local zoning nonsense you get in the US, especially in the burbs. I don’t know the history but am sure there were huge battles about station location and that BART may have chosen “easier” options instead of fighting forever for better ones.

            Re your PS, I can’t see why you’d even do that (add B + D for the shared tunnel) since there is zero sharing of pax (or platforms).

          • Onux

            1. But WMATA still has 50% greater ridership on approx. equal trackage (626k vs 411k in 2019). If there is better overall mode share, remember SF has higher pop/job density in the core; more people rider Muni busses than BART trains. DC still has the better performing urban rail, with a better layout (DC Metro in fact has higher ridership than BART and Muni Metro combined).

            2. Michael, you are missing Alon’s point that BART built the equivalent of RER A tunnels and Chatelet, right along the best possible route through downtown, and yet gets worse than RER C ridership per km, despite RER C not going on the best route through the heart of Paris. Thus the failure of BART.

            2b. Connecting the Transbay Tube to Caltrain tracks wouldn’t have led to an issue in “sharing” the tube. BART trains would have simply through routed onto the Caltrain route, there would have been no loco hauled Caltrain trains to share with.

            3. I think the issue is routing, especially extensions, but including the original concept. BART misses high ridership areas all over: absent lines (Geary), following freight tracks instead of adjacent streets (Intl Blvd south of Fruitvale, San Pablo Ave in Berkeley), branching where it shouldn’t (in Castro Valley through miles of nothing to Dublin), not branching where it should (in El Cerrito to serve a continuous urban corridor to San Pablo). Parking structures are a non-sequitur, the potential riders from developing parking lots at existing stations is dwarfed by riders from existing development at stations that don’t exist. BART has an RER layout when the Bay Area needs an urban rail system. The Bay crossing is a difficult choke point, but every successful urban rail system follows a net system with crossing lines, even if just two (cruciform) or three (soviet triangle).

            3b. I don’t find the Millbrae routing terrible, the traditional route to SF from the South followed El Camino Real west of the mountain, and that is where the people are. Millbrae (or better yet a unified station slightly north at the airport) is the natural point for an urban rail system to end and meet RER, due to topography, distance from downtown, the existence of Caltrain, and the airport. East of the mountain is faster for direct service, both routes should be served.

            Michael: BART’s broad gauge had nothing to do with avoiding fights with existing rail operators, it was because the engineers thought it was necessary for stability in high winds crossing the Golden Gate Bridge (a line never built). Technically they were not wrong, broader gauges are more stable (Indian railways can double stack containers on flatcars, not well cars; for a given curve radius broad gauge can maintain higher speed).

            BART’s gauge is canard, however, nothing about it hurts performance or even cost of rolling stock. Toronto and Philadelphia use a non-standard gauge for trams and subways with no issues. Standard gauge is not superior except for being standard, i.e. interoperable. This would have helped if the Bay Area had built true RER, with a tunnel from Caltrain tracks across the Bay to Oakland freight yards. Then existing tracks could have carried trains almost everywhere BART does at less cost (BART rebuilt existing ROW for its outer network anyway; even in San Francisco some tracks follow old SF & San Jose RR ROW). They could in many cases carried them farther, particularly Oak-SJ from the start (East Bay “Caltrain”). The only major missing piece would have been the line/tunnel to Walnut Creek/Concord, where eventually following the freeway as built would be fine. Together with an actual urban rail system serving SF/Oak/Berkeley (see above) this would have been the superior choice for the Bay Area, if for no other reason than it would have connected SF directly to Sacramento for intercity rail.

          • RossB

            That said, for the most part the problem with the original BART system isn’t the routing.

            Yes, it is. Sorry, but you are focused on the trees, not the forest. To be clear — and this is Micheal’s point — BART did everything they set out to do. They built it. It goes really fast. Areas around the suburban stations have seen substantial growth. So what? This was an extremely expensive system to build and it has relatively few riders.

            More people ride the Muni bus system than ride BART. Let that sink in. BART is extremely fast. Muni is extremely slow. BART serves the entire region. Muni only serves a very tiny part of it. Yet more people ride the Muni buses than ride BART.

            Imagine if they had taken a traditional approach, and simply built a regular subway. There would be multiple lines in both San Fransisco and Oakland/Berkeley. There be way more stops in that urban core. For the exact same amount of money, you would have much better ridership not only on BART, but on the system as a whole! Keep in mind, despite its polycentric transit system nature (thanks Richard) most of the trips are within the urban core. Just nine stations between San Francisco and Berkeley account for half of all rider on/offs (https://www.bart.gov/about/reports/ridership). Even though BART serves the suburbs extremely well, and the core of the city poorly, most people use it within the core. Imagine if it actually did a good job serving that core. It would work better not only for those in the city, but those in the suburbs, since once they got into the city, they wouldn’t have to struggle with the slowest bus system in the nation.

            As for WMATA, it isn’t perfect. Leaving out Georgetown was stupid, and a good example of the political crap that the U. S. has to deal with. But it is by far the best U. S. system built from scratch, and head and shoulders above BART in every respect. Every failure is simply do to maintenance. The Prius is a fantastic car — arguably the greatest general purpose automobile ever built. But if you don’t put oil in it, it is useless. When both transit systems are running well, D. C. Metro has about twice the ridership of BART. Those in San Fransisco, Berkeley and Oakland just take the really slow bus (or tram) simply because they don’t have an alternative. They don’t have access to a real subway.

          • Nilo

            DJ, BART can’t regauge, because the rails have a different geometry than standard gauge ones, so you can’t just squash them together, and run Caltrain into it.

            Onux you’re half right, the actual natural end point is Colma, where the massive graveyards also signal an enormous density drop. Christof’s map here shows this with a huge white gap, and BART’s stations further south of Colma besides South San Francisco— Millbrae, SF Airport, San Bruno— basically avoid density. The Caltrain stops in those cities do a better job.

          • Nilo

            Also un-noted yet, but significant is BART spent a huge amount of money to build four very deep (almost certainly excessively deep) tracks under market street, except instead of reserving these for regional rail service, half of them were turned over to Muni for light rail operation. So you have Light Rail trains running in tunnels built for massive regional rail trains meaning an enormous amount of money was wasted on larger stations, tunnels ect…

          • FDW

            Nilo, BART and MUNI’s tracks aren’t all that deep under Market. IIRC, the BART platforms at Embarcadero are like 20-25m below street level, which is kind of reasonable given that the Transbay Tube is passing under an Active shipping lane. Now the upper tunnels are about the same size, but MUNI and BART’s don’t differ too much in dimensions. I agree that the platforms are a waste, which is why I would want to upgrade the stations to better take advantage of platform length.

          • michaelrjames

            Onux

            2. Michael, you are missing Alon’s point that BART built the equivalent of RER A tunnels and Chatelet, right along the best possible route through downtown, and yet gets worse than RER C ridership per km, despite RER C not going on the best route through the heart of Paris. Thus the failure of BART.

            No, it is a different conception as to what failure means. I mean we’re talking an American sprawled urban area occupied by American suburban car-drivers! And it achieves only 10% less than two (or three) of Paris’ RER lines! To me that is success on an unlikely scale while to you and others it’s failure! (Trouble thinking of an analogy but if my scientific career or publications were 10% poorer than a Nobel prize winners, do you think I or my peers would consider that a failure?)

            Michael: BART’s broad gauge had nothing to do with avoiding fights with existing rail operators, it was because the engineers thought it was necessary for stability in high winds crossing the Golden Gate Bridge (a line never built).

            All kinds of smoke gets generated on a contentious issue. Often the truth cannot be laid out in plain sight. To me this one doesn’t pass any logical test. I don’t believe it, and again it just comes down to “belief”. Marin county was the first early dropout in participation in BART so again it doesn’t add up. Another example is Sydney’s NW Metro (RER-like) which the incoming conservative government suddenly, without consultation with the experts who unanimously condemned it, reduced the clearance of its tunnels by just a few cm on some spurious grounds I can’t even remember. But everyone knew it was for one purely political reason: to make sure inoperability between the rest of the system’s duplex trains so they could semi-privatise it as an independent line.

            And BTW, just last week on Alon’s blog there was almost unanimous agreement (funny enough perhaps exempting Alon?) that different gauges for different systems was a minor issue, eg. even though Paris RER and Metro use the same track gauge, they never share trains and are designed and act as completely separate systems (of course RER couldn’t use Metro tunnels but Metro could use RER). Am I correct in that the TransBay tunnel could have been (still could?) be fitted with triple rails to be bi-gauge compatible? (Of course I am not saying that is true for long-range rail systems essentially doing the same thing; Australia had 7 states with something like 5 different mainline track gauges! Madness.)
            Hah, I just read the rest of your post and you agree:

            BART’s gauge is canard, however, nothing about it hurts performance or even cost of rolling stock. Toronto and Philadelphia use a non-standard gauge for trams and subways with no issues. Standard gauge is not superior except for being standard, i.e. interoperable. This would have helped if the Bay Area had built true RER, with a tunnel from Caltrain tracks across the Bay to Oakland freight yards. Then existing tracks could have carried trains almost everywhere BART does at less cost (BART rebuilt existing ROW for its outer network anyway; even in San Francisco some tracks follow old SF & San Jose RR ROW). They could in many cases carried them farther, particularly Oak-SJ from the start (East Bay “Caltrain”).

            Of course, all that is true and there is surely no one disagrees—either me or BART designers or backers—but the entire point was to avoid osbtructionism. You keep saying/implying I am wrong on that but every single thing I have read, both in SF & CA but nationwide in the US, screams the opposite. It is exactly why the NEC has its 240km/h-capable trains averaging whatever it is (120km/h?) for the whole route and bedevils almost all rail issues there. The only reason Brightline got built while the earlier HSR proposal went down is that it is by the same commercial entity that owns the freightlines and ROW that will be used (except to Orlando and guess what delays that will face ….).

            RossB:

            More people ride the Muni bus system than ride BART. Let that sink in. BART is extremely fast. Muni is extremely slow. BART serves the entire region. Muni only serves a very tiny part of it. Yet more people ride the Muni buses than ride BART.
            Imagine if they had taken a traditional approach, and simply built a regular subway. There would be multiple lines in both San Fransisco and Oakland/Berkeley.

            Again this confusion between regional rail and city metro. BART was an initiative to link the Bay Areas non-city counties to the centre and themselves. It was up to each city to build their own metros (Munis, whatever). So your complaint is that BART should not have been funded and the money should have gone to the individual cities/counties to build better Munis. But that doesn’t fly either (though not to say the cities should not have done it, but blaming BART is weird). On both ridership and strategy (fewer inner-city stations) it’s like you would say Paris RER should never have been built, and—I don’t know—its Metro should have been … extended into the burbs? And RER gets lower ridership than Metro, and you could claim even worse given that Metro serves about 2.4m residents (actually more since it does extend into the Petite Couronne) while RER’s catchment is the whole 12 million. Yet, most people consider RER an unqualified success, even those who complain about its imperfections.
            And on the same issue of city-Muni versus regional-rail BART, Nilo suggests that having one level of Market street given over to Muni is wrong. So presumably RossB disagrees since isn’t it precisely serving both with lots of stations in the most trafficked section of the city? Plus freeing the streets (some part of which have gone pedestrianised)! Am I so stupid to be missing some fundamental point. Isn’t it kind of like when Paris built that enormous pit of Chatelet-Halles that it served both future RER plus 5 Metro lines (one of which was realigned during construction)? Thus allowing efficient (more or less) transfer from local transit to regional transit all along those 8 km (or whatever) of Market. I would contend any designer would come up with essentially the same scheme (and I don’t think there is any meaningful distinction here between lightrail and metro rail).

          • Eric2

            “that different gauges for different systems was a minor issue, eg. even though Paris RER and Metro use the same track gauge, they never share trains and are designed and act as completely separate systems”

            In Paris, they didn’t ignore an existing RER gauge line and build an entirely new line on a nearby separate ROW at massive unnecessary expense.

            Also in Paris, both the RER and Metro are very busy and essentially use up all of the track’s capacity, which means there is no point in track sharing because the service complexity would degrade capacity. This is not the case on BART branches.

          • Onux

            RossB: “Even though BART serves the suburbs extremely well, and the core of the city poorly, most people use it within the core. Imagine if it actually did a good job serving that core.” This is the best possible summation of the issue with BART.
            Two thirds of BART’s ridership is between stations on the inner 44% of route miles. Imagine if the remaining 56% had been deployed in equally productive areas (or MORE productive, San Francisco has at least eight bus or light rail corridors with pax/mi higher than the outer portions of BART; four of them have higher pax/mi than the core of BART). BART in this case would easily surpass DC Metro in ridership, and could come just behind NY and Bos in pax/mi.

            Nilo: FDW is correct, the Market St subway is not very deep. The platform of the Central Subway at Union Sq (just 400 ft past where it goes under the BART tunnels) is only 70-90 ft below the surface.

            Marin was not the first dropout from BART, San Mateo county was (they already had the Southern Pacific line, which would become Caltrain) so Marin was ASKED to drop out by the BART board because without San Mateo’s tax base it was thought the system couldn’t support all the tunneling to get to Marin (San Mateo was going to use the SP ROW at grade). See:
            https://www.sfgate.com/local/article/Marin-County-BART-Golden-Gate-Bridge-study-14364699.php and
            https://www.marinij.com/2019/11/05/dick-spotswood-addressing-the-myth-that-marin-leaders-didnt-want-bart-here/

            Michael: “Again this confusion between regional rail and city metro.” No one is confused. We recognize the differences and we recognize regional rail should use existing infrastructure except where critical in the city center. That is what the RER did (did SNCF change to right hand running and 750V third rail like the Metro for RER?). BART did not do that. It built metro infrastructure in outer regions at great expense where existing infrastructure could have provided regional rail (and in the case of Caltrain does!), while neglecting ridership potential in the core.

          • RossB

            It’s like you would say Paris RER should never have been built, and—I don’t know—its Metro should have been … extended into the burbs?

            No, I’m saying it would be ridiculous to build the RER *instead* of the Metro. Seriously, if we are comparing designs, then the RER would have about a dozen stops in Paris (instead of the 33 it currently has) and people *within* Paris would get by with very slow trams, and very slow buses (the slowest on the continent). That, my friend is BART.

            Yes, that is what they set out to do. So freakin’ what? George W. Bush set out to invade and conquer Iraq. Mission Accomplished. I’m saying that was a really stupid idea, a waste of a lot of money and a lot of lives. Had he put the money into improving living conditions in the developing world, the world would be a much better place.

            The same is true with BART. It was wrong to spend so much money building a system that extends so far, when the transit system serving the core is so poor. There is no equivalent to the Paris Metro in San Fransisco, or pretty much most American cities (D. C. being an exception*). I don’t think you get it, probably because European and Asian countries just don’t do this. It is — my guess — another one of those purely American experiments that have failed. It isn’t unique to San Fransisco. It has happened in other cities (Dallas, Portland, Denver). They tend to be extremely long, have low ridership (meaning extremely low ridership per distance), and either slog through downtown, or under-serve it. They are, in essence, very expensive commuter rail systems without a local Metro to connect to.

            This is precisely what Seattle is building. When all is said and done, ridership for the extremely expensive, extremely long subway system will be dwarfed by local bus ridership, just as it is in the Bay Area. I’m not talking about connecting bus service, either; I’m talking about direct, bus-only trips (or bus-to-bus trips) will make up the bulk of the transit use in the region, and those living with it will wish that the agency had spent the money building something serving the urban core instead.

            * Just to be clear, I’m not saying D. C. Metro is exactly like the Paris Metro (obviously not). But in both cases, the core of the city is well served by the subway. You can’t say that in San Fransisco, nor will you be able to say that about Seattle, even after it spends $54 billion and adds about 150 km of track.

          • Nilo

            Onux, maybe it’s just riding the subway out east, BART has always felt deep compared to that, especially the Mission Street stations. Yeah the obvious BART Phase one is some tunneling in Oakland and SF, ruse the SP ROW in Southern Alameda and San Mateo/Santa Clara (including the Bayshore Cutoff), and the ATSF ROW north of Berkeley. Phase two should have been Gaery, Contra Costa, Sunset in SF.

            Ross, quite a few European Cities of SF’s size when it started building BART don’t have metro and do regional rail and Tram/subway surface: Almost all of Switzerland; Germany outside of Munich, Berlin, Hamburg and Nuremburg. Regional Rail before metro isn’t the most outrageous thing. Munich a true champion of moving towards being more transit oriented over the past 50 years, started out with a complete S-Bahn (with one two track downtown tunnel), and a few metro lines, and only built out the U-Bahn afterwards. Finally Tokyo notably only really has “three” rapid transit style lines. The first two subway lines were built as an independent system and everything after that was build for regional rail through running. This is also true to a significant extent

            Also even Chicago, which actually has a subway system that really only serves the city, bus ridership is higher than metro ridership. Rapid Transit is the majority in NYC, Boston, and DC— the latter in part because the bus network is incredibly mediocre compared to the rail network.

          • Alon Levy

            Munich built the S-Bahn and U-Bahn simultaneously. Both were timed to open by the 1972 Olympics, and the U-Bahn actually opened a few months before the S-Bahn.

          • Nilo

            I guess my point was more that the Munich U-Bahn opened with 16 km and 17 stations for the olympics, which the S-Bahn opened with 360 km for the Olympics. Obviously Munich had a at that point I think pretty good legacy rail infrastructure, but it certainly seems like S-Bahn infrastructure was prioritized before U-Bahn infrastructure.

          • Herbert

            The S-Bahn is only now getting its “zweite Stammstrecke” (~second trunk). The U-Bahn has quite a few “Stammstrecken”…

            So how much of those 360 km were new tunnel? Because I’m pretty sure the 16 km were all new tunnel…

        • Nilo

          FDW, Muni trains are signifincalty narrower at 2.69 meters v. 3.175 meters. Though Muni trains are a bit taller.

          • FDW

            Nilo, MUNI’s train widths have varied, the older Breda’s and the PCC’s are like 2.9 meters, while the Boeings were more similar to the new S2000’s. As for the Mission St stations, they’re actually of similar depth to that of the lower Market tracks. As for what BART should’ve been: Phase I in the East Bay should’ve used International Blvd between DT Oakland and Hayward, Telegraph/Shattuck to connect DT Oakland and Berkley. DT Oakland-Richmond should’ve been San Pablo Blvd. In SF/Peninsula, Phase I should’ve been the Market-Mission Subway to Serramonte and a Market-Twin Peaks line like OTL. Geary should be phase I too, but it shouldn’t be interlined with either Market Line, rather I think it should be interlined with the SP corridor (Though in SF the line should be diverted to 3rd St.). A Second Phase in the East Bay would be DT-Oakland-Eastmont via Macartur, and DT Oakland-Walnut Creek via Broadway. In the West Bay, a Second Phase would be the “30 Stockton+8 Bayshore Subway” and a Market-Sunset Subway.

          • Nilo

            Yes BART’s tracks under Mission are definitely too deep if they’re as deep as the deeper set of tracks under Market.

            You’ve effectively tossed away the idea entirely of re-using extant ROW except on the peninsular line, significantly driving up costs, and ditching the ethos of an S-Bahn. SF local light rail should not have been part of the plans for BART. If SF wants to build out a subway surface system it should have done it without taking valuable ROW from Regional Rail. Commenting in detail on such fantasy proposals is a little silly, but I will say rail to Santa Clara should have been prioritized, because it was the fastest growing county both percentage wise, and I think raw numbers wise in the post-BART era. Good rail could have provided a growth mechanism especially at the southern end of the county that could have been dense, walkable, and avoided NIMBYs because there would have been few people there.

          • michaelrjames

            Onux: Michael: “Again this confusion between regional rail and city metro.” No one is confused. We recognize the differences and we recognize regional rail should use existing infrastructure except where critical in the city center. That is what the RER did (did SNCF change to right hand running and 750V third rail like the Metro for RER?). BART did not do that. It built metro infrastructure in outer regions at great expense where existing infrastructure could have provided regional rail (and in the case of Caltrain does!), while neglecting ridership potential in the core.

            I don’t accept the analogy with RER. What you are saying is that CalTrain is the version of regional rail that you (and Nilo below) find acceptable, and obviously San Mateo and Santa Clara. How do you think it compares to RER? I think the real reason is that they weren’t willing to invest to provide an exclusive ROW, ie. no at-grade road crossings. Clearly it was the cheapest possible option and it has remained unchanged to this day, and with CaHSR on ice, for the forseeable future. I don’t believe the RER has a single such crossing on its almost 600km of tracks. Of course all that track/ROW already existed and converting to RER mostly involved those very expensive tunnels across Paris (along with trains & signalling etc). Does any of the equivalent in Tokyo or Seoul, or any European big city have anything like the Caltrain peninsula line? I don’t think so. It’s more resembles any number of rural branch raillines. (Aside: I read that there is a train-vehicle collision once a day every day of the year in the US there are so many such crossings.)
            It’s almost certain that this was what caused BART to choose its particular strategy. No serious funding from SM or SC to build serious Bay Area transit. BART’s proponents may have ignored many lessons from overseas but they knew that CalTrain was not what any serious city would build in the modern era. As Alon commented, Munich (I think) simultaneously built its U- and S-Bahn. And Paris continued to build Metro as it built RER, eg. M14 and M7 extension, not to mention the huge M15/M16/M17 GPX underway.

            Clearly that is what SF and the BA should have done. Built both. In lieu of the willingness to spend that kind of money, you (and Nilo and RossB) suggest that that money should have gone to building a metro for SF. That may be a valid opinion, but what about the eastern counties? Incidentally it has been pointed out that they supported BART (or I suppose any viable regional rail plan) because of their lower SES and willingness to use public transit while the peninsular counties were more prosperous and their taxpayers historically won’t vote for a county measure to subsidise expensive mass transit. Hillsborough is the extreme version.
            Also when you or RossB say that “they” should have spent the money on SF’s metro, they are different sets of people. One can surely ask, if the money wasn’t spent on BART would have it made any difference to SF’s local transit? Would the “they” fought as hard for the money and political support to build it? Why haven’t they?

            Nilo: You’ve effectively tossed away the idea entirely of re-using extant ROW except on the peninsular line, significantly driving up costs, and ditching the ethos of an S-Bahn. SF local light rail should not have been part of the plans for BART. If SF wants to build out a subway surface system it should have done it without taking valuable ROW from Regional Rail. Commenting in detail on such fantasy proposals is a little silly, but I will say rail to Santa Clara should have been prioritized, because it was the fastest growing county both percentage wise, and I think raw numbers wise in the post-BART era. Good rail could have provided a growth mechanism especially at the southern end of the county that could have been dense, walkable, and avoided NIMBYs because there would have been few people there.

            Isn’t this a bit naive? Taking valuable ROW from Regional Rail. Market street is the main central thoroughfare thru SF and one wouldn’t expect the city or Muni not to have an interest in Regional Rail having it in exclusivity. Of course SF ratepayers were contributing to BART funding so they would want to protect their most valuable ROW for future city transit. I really can’t see how the outcome wasn’t a good one. In fact it is rather exceptional, and only possible via cut-and-cover then. Today it is possible by using very large diameter TBMs that can have two levels in one tunnel, thus at least 4 tracks.

            Onux, I read those news items on the Marin history. Interesting and thanks, but the thing is does it matter either way? And as I read the Dick Spottiswood piece I was wondering where the information was coming from. From his own memory? Unsourced, this wouldn’t make it past a Wiki editor. In fact one of the commenters, who claims to be a first-person witness wrote:

            (Lois) I was a child in Mill Valley when BART happened, and I always thought that my parents voted against it. Then I read somewhere that the GGB Disctrict prevened BART on the bridge for some political reasons, like they wanted their own busses to profit. Now Dick tells a different story. I wonder what the real story is? Can you cite sources, Dick?

            There is no doubt that the Marin supervisors were dead against it and arranged a negative engineering assessment saying the GG couldn’t safely carry a second level to carry the train; it is known that it is simply not true from a engineering perspective. But I would say that, given the congestion on the bridge, if the rich Marin drivers-commuters can’t get that extra level built then what chance transit? My point was that it would have been hugely expensive (though it is claimed the bridge option was cheaper than a tunnel) either way and was unrealistic. So it’s kind of moot whether the San Mateo withdrawal provoked the Marin withdrawal.

            On a personal note, I am not sure Marin really needs such a link. Whenever I visit there–I often walk across the bridge down to Sausalito for dinner in the evening and take the ferry back–I’d defy anyone with a pulse not to fantasize about living there. However it doesn’t include driving to SF every day, but it would involve using the ferry instead (or cycling); there is something magical about a ferry across one of the world’s great harbours. The greatest is probably the Manly ferry in Sydney and just like SF there are people who want to drive and moan nonstop about the bottleneck of the Spit Bridge (that also has to close to allow boats into Middle Harbour). The exact same people are adamantly against an extension of Sydney Metro there, because they go on about the riff-raff invading their idyllic Northern Beaches. Tony Abbott was the federal MP for Manly and he promised he would get a massively expensive road-only tunnel built all the way to the bridge, but his electorate kicked him out last election. And just like that extra level on the GG, the road tunnel to the pampered Northern Beaches is not going, and doesn’t deserve, to happen.

          • FDW

            Nilo, I’m talking U-Bahn here, not S-Bahn. That would be the Third phase, after the core lines are laid out. Now Santa Clara county grew the most after 1970 because it was the one place left in the Bay Area where sprawl still ran unchecked. In a more sane world, intenseification of development would’ve been prioritized for the areas with 30km of San Francisco. That means prioritizing areas like Central Contra Costa, Half Moon Bay, and Marin for connections.

            As for depth, while the situation under Mission isn’t the most ideal, honestly it isn’t nearly as bad potential Subway stations on Geary or Van Ness. For that matter, I would imagine that a properly built Rail system in SF to have a whole bunch of 30m+ deep stations.

          • FDW

            michaelrjames, you need to understand, the Bay Area is structured such that you need effective Local Rapid Transit for the Regional Rapid Transit system to work properly. BART’s decision to abdicate this role is a major mistake, especially how it’s obstructed local efforts to build LRT in the East Bay in the past.

            A Post-war SF Metro would naturally include the East Bay, which is so dense and urban that calling it a suburb is an insult. SF has 7 radial corridors and the East Bay 6, and several of those corridors would be 4 tracks to accomdate Regional service.

            It’s funny how everyone’s throws around comparisons to RER C, when the real RER C in the Bay Area would be a connection from the SP corridor to the Transbay Terminal via the Belt Line Railroad.

          • Richard Mlynarik

            you need to understand

            Good luck with that.

            Reality-unthethered Wall of Text incoming in 3 … 2 … 1 …

            connection from the SP corridor to the Transbay Terminal via the Belt Line Railroad

            That was exactly the 1987 “Peninsula Commute Service Interim Upgrade Study” proposal (complete with nice alignment drawings.) Under King Street and The Embarcadero (some partial realignment of large sewer box required), under Main Street between the Bay Bridge pier footings (plus a seven! track underground storage yard on the east side of the piers west of Spear Street), entering the Transbay Terminal side from the east end and terminating. A few hundred pages of analysis and plans. Ignored. BART contractors required and were delivered funding monopoly

            1987.

          • FDW

            Richard, I was actually talking about an aboveground connection via Beale St Spur to the upper level of the TBT. This would be in a Pre-WWII context. I say this because SP had always wanted a closer terminal to DTSF for peninsula services, and the TBT could’ve been the way to pull it off.

      • Nilo

        Eric2, we’ve pointed out why the gauge thing is an issue to Michael at least a dozen times. Maybe if we pointed out how much more expensive the RER A would have been had they tried to build it on an Indian Gauge he’ll get it? Otherwise he doesn’t seem willing or interested in listening.

        • michaelrjames

          Eric2, we’ve pointed out why the gauge thing is an issue to Michael at least a dozen times. Maybe if we pointed out how much more expensive the RER A would have been had they tried to build it on an Indian Gauge he’ll get it? Otherwise he doesn’t seem willing or interested in listening.

          You have? Don’t think so. I’ve been shouted at and insulted violently, but rational argument … nope. It’s apparently all too “obvious” to bother with.
          Obviously I don’t think arbitrarily choosing an incompatible gauge is a good idea, but it was precisely to avoid the toxic politics and osbstructionism you see in almost all urban domains in the US. The vested interests of existing ROW owners and even (or especially) local government is a big part of why the US has so little of the likes of BART or WMATA. It was also a kind of future proofing against various scenarios.

          As to expense, they were always going to lay new modern track so I doubt the claims about extraordinary costs. In fact they did reuse old disused rail ROW. But the real cost was the massive Market street cut&cover, which surely no one seriously objects to today? (If so, why? Just on cost? Get outta here.) And what about all those road crossings at grade on CalTrain tracks–that would have been an awful compromise, unless you are just waving a magic wand over the issue. And if so, how were you to convince Santa Clara (with their rights over that bit of CalTrain track) to agree? The political blocks were huge but they found a way around them, even if it meant several less than perfect decisions that you, 50 years later, can’t see your way thru the fog.

          I think someone other than me needs to do some listening … and thinking.

          RER or its sponsoring organisations () didn’t have to worry about any such issues because even SNCF (a potential blocker in the same way CalTrain and its sponsors were) and the various levels of government collaborated in its creation. Not that there weren’t battles but somehow the French managed to forge a plan that worked for Ile de France. It certainly cost a lot, as Alon describes, but I have no idea if anyone is seriously suggesting an alternative. Heck, Alon is unhappy they didn’t build more tunnel at even more cost! The cost was hugely controversial at the time but clearly the rationale for the scheme was compelling and persuasive. The fact that London finally built its equivalent kinda proves that doesn’t it–even if 43 years late and at even higher cost. In fact it is people like Thatcher who killed such projects; she visited BART (when she was Transport minister in the Heath government) and was appalled at its cost. That’s the kind of mindless opposition to spending money on transit one has to cope with in the Anglosphere. But hey, you’ve got Maggie on your side:-)
          The odd gauge of BART does reveal dysfunction but your problem is that you make the wrong assumptions as to exactly who and what was dysfunctional.

          • RossB

            The vested interests of existing ROW owners and even (or especially) local government is a big part of why the US has so little of the likes of BART or WMATA.

            You have it backwards. As I wrote up above, the US has several systems similar to BART. DART, for example (it even has the last three letters). Denver, Portland, San Diego — it really is the (bad) model for new systems in the United States. This is because of pull from the suburbs, not opposition. Seattle is a prime example of this. Imagine if Sound Transit hired a consulting firm — better yet, two of them — to come up with a way to serve the region, given the same budget they have now. They would, undoubtedly, take a look at the data, and come up with a subway serving the central city, with terminal stations that have good bus interfaces. The suburbs would be served by commuter rail (running on existing tracks) and buses connecting to the subway system.

            But that is not what they built. Instead, they simply asked each suburban city what they wanted. To a one, the all said “Rail!”. We want rail. Rail is cool. Make it fast, like the rail in the city.

            So, sure enough, from distant satellite cities — the size of Spokane (or half of it) — to all the suburban cities in between, there will be rail. It will be a failure in terms of value added. The ridership per dollar spent will be horrible. The ridership time saved per dollar spent will be horrible. It will barely move the transit-share needle, which will be altered more by (very inexpensive) improvements in the (overburdened) bus system, or changes in zoning (to allow more people into the city).

            All because suburban cities simply didn’t understand how transit works, and didn’t bother to hire someone who does. Actually, that isn’t fair to the city of Kirkland, a suburb of Seattle. They did hire a consultant. The consultant recommended they turn an existing (rail-to-trail) pathway into a BRT corridor. Send buses to all the various suburbs, while avoiding the congestion. Unfortunately, Kirkland was overruled by the rest of the board, so instead they will have a rail line from one large park and ride to another — none of which will actually help anyone get to Seattle, or to the urban center of the Kirkland (such as it is).

        • Herbert

          Except for the park and rides (which I think are the real issue, not the decision to build rail) what you describe is pretty much what Karlsruhe built. They are now building – at extreme expense – one tunnel in their downtown to increase capacity. But they are doing that after having built hundreds and hundreds of kilometers of rail into the suburbs – some of which on alignments DB was going to give up. In fact, they even revived a Heilbronn tram which had not previously existed for four decades (that they call it “Stadtbahn” and abbreviate it with an S is marketing)

          Compare that to Nuremberg which spent uncounted billions on turning 70 km of tram into 35 km of subway and 35 km of tram and a half-measure “S-Bahn” which is plagued by single track stretches, sharing tracks with regional or intercity rail and in some places is just calling a Regionalbahn at 60 minute headways “S-Bahn”. Heck if you ever ride the Erlangen-Nuremberg trains, you can see where the capacity crunch is worst (in both directions, mind you) but NIMBYs in Fürth successfully delayed the quadruple tracking of the line creating a non-Takt and ridiculously low frequencies.

          So who took the better decision?

  2. Brendan Dawe

    What I find so unconvincing about your example of DB Intercity Service is that it’s not like Germany didn’t have intercity trains in 1970. It is obviously an incremental improvement upon an existing intercity passenger services that have always operated in Germany, whereas Ohio hasn’t seen a daylight passenger train in generations and everything would have to be done de novo but for existing freight rights of way and a handful of Amtrak night train stops. That is was cancelled doesn’t really tell us much, after all the Florida high speed rail plan was also cancelled. Would it be better to have been more ambitious from the get-go? Undoubtedly, but there’s not really sufficient grounds to say that’s why it didn’t take.

    I can think of many projects that did start small, drove stakes, and have grown from there. For instance, the GO system in Toronto, which started a small number of trains on the lakeshore, incrementally improving over skeletal CN suburban service, which has steadily grown since, with new routes added a peak one-way commuter shuttles growing into all day affairs. It’s middle-class commuter constituency, rather than being a jealous anchor for bad service, have been a key political swing block driving the Provincial Government to invest in more and better service. Now,GO does suffer from many of the pathologies of North American passenger railroading, but it’s seen steady, incremental improvement and is set to see that go on into the future.

    Surfliner service in California, pathetic though it is, has also exhibited a similar pattern, with growth from the two-trains a day inherited from the ATSF to 13 trains per day today, plus commuter services, and a program of steady improvement going forward.

    I would like to believe that that big and ambitious plans will drive better political outcomes, but the stagnation or cancellation of big, ambitious projects like Florida or California High Speed Rail suggests that this is not necessarily the case. While I disagree with the certain type of rail advocate who says that ‘one must walk before one can run’, it seems that there’s thin grounds to assert that you must be willing to start with SBB-in-a-box to get anywhere

    • Benjamin Turon

      I agree with Mr. Dawe to a certain degree, sometimes getting something running is better than nothing, and America’s record for going big isn’t that great either. In fact, there is no high-speed rail in America, but services like the Cascades, Capitol Corridor, and the North Carolina services do exist. A once daily train like the Ethan Allen is popular, and a city like Saratoga Springs likely doesn’t need much more than several daily trains.

      Yet with that said, I think for a major corridor you need to dream, plan, and execute big – and most frustrating in the United States is how services languish, how investment is stop-and-go, decade after decade. You certainly see this with the history of the Empire Corridor, which did well 1975-85, fell flat in the 2000s, and then got a big dollop of investment after the 2009 ARRA.

      Frustratingly, you don’t see steady incremental progress to a clearly defined final goal of service level. My favorite examples are the ‘Downeaster’ and ‘Pacific Surfliner’. There is not good reason why these two corridors should not have seen the “Virgin Brightline” level of investment in infrastructure and trainsets. A few billion would get you to hourly service with trains running at 90 to 110-mph, with average speeds of about 65-mph, making the train a bit faster than driving, without the hassle of parking in downtown Boston or LA/San Diego.

      The big failure in America is really in planning and financing, the lack of vision and experience in the the entities that are supposed to be overseeing intercity rail service. Its part of the overall downsizing and lack of institutional knowledge in Anglo-sphere government. You see this elsewhere, like how I have pointed out before the difficulty that the US Navy and Royal Navy have in designing and ordering new warships, when the former ability to do that in house was out-sourced to private defense contractors. Similarly, when New York State did its intercity planning in-house at NYSDOT with a sizable staff, things happen competently. After this capability was axe, you got failure like the Pataki Ear HSR Program or inaction.

      Yes, America needs to think big, but first it needs the staffing and expertise to think big, and it needs to be willing to spend money.

        • F-Line to Dudley

          …because of supply-side shortfall. Until Albany-Renssalear sails WELL above that frequency threshold for the service pie that gets divided there, every destination outbound is left sucking for air. Saratoga Springs absolutely deserves that level of service, but it’s proportionally a much smaller piece of the Empire pie than Niagra so requires a rising tide to lift all boats proportionately.

          There might be some hope when the new dual-mode loco procurement (a triple-agency order for AMTK Empire + MNRR + LIRR) gets bid out to RFP later this year. And then when the Amfleet-replacement coach order due to bid out any month now thrashes through the first-up national/Northeast Regional pool, NYSDOT will have the largest-share pick of all the statie options. The unit prices will never be better given the quantities involved, so if they don’t load up for bear on expansion units this time around it’s going to be self-inhibitingly more expensive to do later. Loco triple-procurement starts at 75 replacement-level units…80-90 if MNRR displaces its small Harlem or Harlem + ConnDOT straight-diesel fleet for the shuttle services with the same dual-mode make for the sake of maintenance scale (LIRR’s shuttle fleet is large enough that it’ll probably stay separate).

          From there it’s up to sometime- transit czar Cuomo to explain why he is (or more likely isn’t) going to max out the options on both orders for large Empire Service frequency increases. If the loco triple-procurement doesn’t start at 90 on the base order and expand out to triple digits on the exercised options, you won’t have any hope of service increases enough to hit any meaningful past-ALB frequency thresholds. TBD on the coach fleet shares, but similarly if their quanties hew to replacement-level you can be sure that service increases are getting laughed at from behind the Gov.’s closed door. I guess watch for PR or lackthereof, since he’s eminently predictable in how he delivers good news and non-delivers bad news. if he calls a press conference from the ALB platform they’re probably leaning to (multiple asterisk-laden) expansion units. If he sends a hapless spokesflak: somebody else’s problem to explain why you’ll never get nice things. Same as it ever was.

          • adirondacker12800

            It’s faster to drive to Rensselaer then it is to take the train from Saratoga Springs. People drive to Renssalaer instead of taking the train from Saratoga Springs. They have to fix that before lots of people start using Saratoga Springs. I doubt there is enough demand for once an hour. Not at two half hours from Rensselaer to Manhattan.

          • Herbert

            You don’t have to have demand to fill the train at every stop. In fact if you have that kind of demand, you’re not running enough trains.

            You just need enough on the entire line for the train to make back marginal costs with “make back” being a vague term subject to political considerations like developing tourism or whatnot…

          • Benjamin Turon

            I don’t see where that demand would come from for hourly intercity trains at Saratoga Springs. For NYC-Albany? Yes, definitely hourly service would a boon. For NYC-Buffalo? Maybe, certainly if you build a high speed rail line across the state. If you upgrade the existing line (3 tracks at 90-mph on CSX) then I think bi-hourly would be good. For Saratoga Springs, I would consider a early morning down and late evening up train, that would of course serve Schenectady and a new Albany-Colonie Station on Central Ave.

            Of course a regional DMU/Fuel-Cell-MU might be OK for hourly service to Saratoga in the somewhat distant future, but a lot of TOD would need to be done at the Saratoga Station, and park n’ ride for Ballston Spa and Burnt Hills, and likely such a service should terminate in downtown Glen Falls, using the existing rail spur, and not Saratoga Springs. If you get that D&H bke train back, you could get to Lake George. The Capital District had a extensive electric interurban rail system, stretching from Hudson to Warrensburg, but it was all gone by WWII.

            Frankly electric cars will work well for most trips for most people in a suburban region like the Capital District, and incrementally increasing private electric car ownership is a much easier transition then rail service that will miss 99% of all people. For transit, I would upgrade the CDTA bus network and build TOD around the bus stops, build more sidewalks and paths, make it so you can walk for many things. Adding sidewalks makes bus usage more attractive.

          • F-Line to Dudley

            That reply doesn’t even rise to “Well, at least I tried” level of staying on-topic. Was it not completely transparent that ALL of NY State is hamstrung by the underwhelming frequencies to Albany? So, yes, it is a moot point to talk about destinations beyond ALB when ALB itself is left lacking for frequencies…which is why the discussion immediately pivoted there. Doesn’t counterpointing, you know, actually require offering a substantive counterpoint??? Or are we just in a mood to ‘sound’ contrarian today without actually being so?

            And who was suggesting that park-and-ride intermediaries are the be-all/end-all of Capital District transit demand? No one. Saratoga Springs is in the CDTA bus district; to/from NYC is a trip that could be done totally car-free for a large share of the potential audience were it not for the totally unusable frequencies. So where did this become about the drive to Renssaler lot? Not everybody who is hitching a ride for an intercity trip can get to Rensselaer easily by bumming favors from friends/family/colleagues/Uber. The untapped demand is more local than that. Saratoga Springs was brought up in the first place because 100.00% of its demand can’t be met by park-and-riding from Renssalear…no matter how ‘easy’ it supposedly is to make that drive if you have the ability. But despite that distinction flying completely over head, it should’ve been completely implicit upon first mention that stiffening of ALB frequencies is a prerequisite for anything/everything related to Cap District-demand. Saratoga Spring’s frequency problem is an outflow of Albany’s frequency problem.

          • adirondacker12800

            For NYC-Albany? Yes, definitely hourly service would a boon.

            Albany already has hourly frequencies. Or did. The trains ocassionaly sold out but that was a function of not having enough cars. Until they make taking the train from Saratoga Springs faster than driving to Rensselaer, people who own cars will be driving to Rensselaer because it’s faster to do that. It’s why Rennselaer is so busy and the other stations in the Capital Region have lousy ridership numbers.

            Click to access NEWYORK18.pdf

          • dejv

            Sascha Claus, the Zierke site is quite dated at 15 years since last update. The last Talent 1 he uses as an example could be delivered 7 years ago IIRC as it didn’t conform to now effective crashworthiness regulations (EN 15227).

            IIRC, Siemens and Bombardier gave up on DMUs after crashworthiness and emission rules tightening so Alstom is the only large European manufacturer still actually making them and Stadler has them just as an offer (Polish manufacturers such as Pesa or Newag do produce some as well but they’re really heavy). There’s other crashworthiness rules tightening coming in few years so you can expect the offer to reduce even more.

    • Alon Levy

      In Toronto, the RER was driven by forces outside GO; the engineers at GO fought against it, and at the end of the day there have been so many constituencies wanting to wet their beak that costs have gone completely out of control.

      • john

        But the GO engineers were only defending their territory. You understand how it is, Don Levyone…

    • Herbert

      The pre-IC long distance trains were not in a Takt and did not have connection hubs. Introducing the IC sped up many many trips without the trains necessarily being that much faster (200 km/h top speed)

  3. Henry Miller

    Most proposed train routes are easy and cheap to test without building a single track. Just buy some buses and run them with good service: signal priority, stop only at your proposed stations. If you cannot fill busses all day with 5 minute or less headways you shouldn’t be thinking rail in the first place.

    The above is only a first approximation. In the real world cites tend to be build around geographic features (almost always a river) that doesn’t have roads where you would want the train to cross. Studies can determine how much ridership you would have, but that train isn’t going to reach its full potential without a network of frequent service so there is no point in the study until you have the rest of the network in place with buses.

    Too many think trian first instead of just getting something working now which the bus can do. Get something fast to prove demand first. Most city bus system in the US are terrible. Until people who are forced to try the bus once (that is their car broke) can’t honestly say they wouldn’t be imposed on if they had to ride the bus you don’t have the network to make trains work either. (unless you have the money to build a trian grid every half mile across the whole city in a year, if you can pull that off I’ll believe the train doesn’t need a functional bus network)

    Once you have solved the first and last mile local problem you can ask about high speed rail between cites. However until then you high speed network will suffer because it doesn’t go anywhere on either end (except for a parking lot and rent a car since those are the only things that can work – see airports)

    • Alon Levy

      That doesn’t work unless the train is following an arterial road, which is only the case for a handful of urban corridors. In regional service, a bus is completely useless – it has to use freeways, which can’t easily go into intermediate cities and make stops. In intercity service, it’s even worse, because decent low-speed trains average 100-130 km/h and buses don’t.

      • Herbert

        Plus even a one for one replacement of bus with rail will see a jump in ridership. People just like trains. It may be “irrational” but so is humans liking flowers and we’re not condemning that as some “silly conceit we should grow beyond”…

        • RossB

          Like frequency, my guess is it is highly dependent on the amount of time spent on the vehicle. Consider this scenario: A bus comes by, and you know, for a fact, that the train will be there in a minute. Assume that the bus gets you there one minute sooner (they both travel at the same speed). If the trip takes an hour then you definitely take the train — there is more room to stretch out, etc. On the other hand, if the trip takes five minutes, you just hop on the bus.

          Rail preference is a slight thing, and needs to be considered along with everything else. Consider this example. Seattle built a bus tunnel about forty years ago. They added rail to it thirty years later. About a year ago, they expanded the rail system, and kicked out the buses. Oddly enough, ridership within the tunnel (downtown to downtown stops) went down. This surprised a lot of people, because the train is just as fast (if not faster). The problem was frequency. With the old system, you stood on the platform for seconds — a bus or a train would be by shortly. With the new system, you could be standing there for anywhere between six and ten minutes. In contrast, the buses (now on the surface) had frequency in the seconds (as before, when the drove in the tunnel). The combination of low frequency and deep boor stations was more important than the speed advantages and more comfortable ride of the train.

          The point is, if they could only afford to run the train every six to ten minutes, then kicking the buses out was probably a bad idea. Fortunately, in a few years they will be able run the trains more often (3 -5 minutes is my guess), which should be adequate to get more riders.

          • Herbert

            If you can afford to run a bus every few seconds (which to me sounds like a recipe for bunching) they can afford to run trains that frequently…

          • Eric2

            Once on one of the blogs, there was a real life transit planner saying that in their experience, rail gets 1/3 more ridership than bus if they are equal in all other ways.

            I imagine with Seattle, there was also an effect that the light rail follows a single route, while the buses that routed into the tunnel followed many routes. So if your origin wasn’t along the light rail line, it wouldn’t help you and you would continue taking a bus downtown.

          • Richard Mlynarik

            Rail preference is a slight thing, and needs to be considered along with everything else

            But trains!

            The combination of low frequency and deep [bore] stations was more important than the speed advantages and more comfortable ride of the train.

            This is about to be demonstated, at a cost over well over 1.5 billion US dollars, by the San Francisco “Central Subway”.

            Of course its failure to carry even a fraction of “predicted” ridership will be attributed to COVID-19. (The abject ridership failure of BART extensions promoted, “designed” and built by the exact same contractors got attributed to “9/11”.) They never even both to explain “budget” blowouts of factors of two or three — nobody cares, it’s “transit”, Jake. Idiot “advocates” will greenwash and shill and vote for it, rent-seeking public-agency-controlling contractors keep promoting and profiting from the worst projects at the highest costs. It’s a successful model. No way out.

            But trains!

          • RossB

            “If you can afford to run a bus every few seconds (which to me sounds like a recipe for bunching) they can afford to run trains that frequently…”

            Obviously. That has nothing to do with my comment.

            “Why would you say rail preference is a “slight” thing?”

            I guess you didn’t read my comment, but decided to comment anyway. Nor did you bother to look up the word “slight”. In this context, it means “small in degree” (the first definition when I looked it up). Rail preference is minimal — it ranks below frequency, speed and walking distance. My guess it is there with “window seat” and less important than being able to sit. The U. S. is riddled with really poorly performing trams, largely because someone, in full Harold Hill mode, sold them on the idea (https://www.youtube.com/watch?v=v4z_9NcIJXI). OK, that is for a monorail, but same idea.

          • Herbert

            When Nuremberg extended its tram four stops from Thon (already a bit suburban but still pretty dense) to “am Wegfeld” (literally a piece of fields and grass next to a big box store that most people can’t buy at [Metro]) they replaced the pre-existing bus stops with tram stops.

            They predicted a rise in ridership, otherwise they wouldn’t have built it. They got 8650 new daily riders after having predicted 4100 to 7500

          • RossB

            The study is from 1989 (seems like just yesterday). To quote the study:

            The data do not provide explanations for this phenomenon, but other studies and reports suggest that the clearly identi-fiable rail route; delineated stops that are often protected; more stable, safer, and more comfortable vehicles; freedom from fumes and excessive noise; and more generous vehicle dimensions may all be factors.

            OK, let me break those down:

            Clearly identifiable rail route: This can be solved quite easily be logical routing (i. e. having the vehicle go straight). Interestingly enough, Seattle has one of the worst, most confusing rail routes imaginable. Oh, you can that the train goes a little but further (as one would guess a bus would) but then it doubles back on itself, because (guess what) Seattle has hills.

            Delineated stops that are often protected: Fair enough, but that suggests a greater overall investment in infrastructure surrounding trains, which itself could contribute to the entire difference.

            More stable, safer, and more comfortable vehicles: A bit outdated, I would say. Buses are pretty stable, and certainly safe.

            Freedom from fumes and excessive noise: They got them things called electric buses. Seattle has had them for years.

            More generous vehicle dimensions: Definitely a factor, assuming that you really can make the streetcar wider than a bus (often you can’t).

            I’m not saying there isn’t a preference, but I would say it is a lot less than what stated for a modern (relatively quiet) diesel bus, let alone an electric one.

            I’m also saying — with great confidence — that the preference isn’t likely to effect ridership for short trips, but could for long ones. I ask you — since I’m guessing you have a preference — would you wait an extra minute for a train for a ten minute trip? Would you wait an extra five minutes?

          • Herbert

            I can give you the “worst case for rail” from my personal experience and I still preferred rail. I lived https://www.google.com/maps/place/51%C2%B001'03.6%22N+13%C2%B046'32.2%22E/@51.017667,13.773412,458m/data=!3m2!1e3!4b1!4m14!1m7!3m6!1s0x4709c614704dc7f7:0xb16f21c40f5974a7!2sOtto-Dix-Ring,+01219+Dresden!3b1!8m2!3d51.0198398!4d13.7738504!3m5!1s0x0:0x0!7e2!8m2!3d51.0176675!4d13.7756062 hereabouts for a while. As you can see, the bus stop is closer as you can see on http://www.dvb.de the bus ride to downtown is shorter than the tram ride. And I still preferred the tram and took the longer walk time and time again.

            And as I have pointed out above the four stop extension of tram line 4 in Nuremberg from here: https://www.google.com/maps/place/N%C3%BCrnberg+Thon/@49.473347,11.0642911,18.5z/data=!4m13!1m7!3m6!1s0x479f56329031e22b:0xf22b6aa8c892466f!2zVGhvbiwgTsO8cm5iZXJn!3b1!8m2!3d49.4802931!4d11.0708667!3m4!1s0x479f5633ff65d5c5:0x1d8df74ae40ba60a!8m2!3d49.4731748!4d11.0649447 to there: https://www.google.com/maps/place/Am+Wegfeld/@49.4930701,11.0513869,473m/data=!3m2!1e3!4b1!4m5!3m4!1s0x47a1fd8b3c584065:0xf72ce4fcd84ae78b!8m2!3d49.4930701!4d11.053580
            9 caused an unexpected increase in ridership. I mean they did expect an increase in ridership, but what they got was even more than they expected. As for “one seat rides” – prior to the extension bus line 30 coming from Erlangen would go to the old terminus, now it goes to the new terminus and then turns toward the airport. So if coming from Erlangen and going to Thon, one loses the one seat ride but if coming from south of Thon and going to one of the four new tram stops (exactly the same stops as the four old bus stops) one gains a one seat ride – any trip from the north beyond Thon or from the south beyond “Am Wegfeld” was a two seat ride then and is a two seat ride now. As for schedule speed, I can’t tell you that, but the difference can hardly amount to more than a minute.

            So where did the increase in ridership come from?

            And as for electric buses: They are limited in range, more expensive and (I have ridden on some) they still lurch about because rubber on asphalt is less smooth than steel on steel…

            And another thing: In its newest procurement of vehicles, Nuremberg took particular pains to replace their old subway cars which would be two married pairs (without the possibility to walk through the whole length) with one four car unit where one can walk through the entire vehicle. This of course is an obvious advantage of rail over bus – once can walk and look through a vehicle of considerable length, which increases perceived safety.

      • Henry Miller

        Don’t mix regional service with local. However regional service in the US isn’t worth talking about without first getting local service working. Most people in US cities cannot reasonably use transit. Even suburbs could support half hour service if you could get somewhere. When the library is a 10 minute drive assuming all red lights, and transit is 45 minutes nobody will take transit. They won’t ask if it is safe to send their kids alone.

        Once you have a local system in place you can start building a regional system. Until the regional system won’t get many riders because you still need a car to get there so why not drive?

        Sure if you want hrs only you can get some riders if you start big enough, people who live near your stations (and in cities that have some okay transit). But starting big means starting local which can be small.

        • Richard Layman

          I think this is a bit doctrinaire. You need to plan at different scales simultaneously, although what is possible regionally differs by metropolitan area. E.g., the regional opportunities in DC, Maryland and Virginia are significantly greater than the five county area of Detroit.

          http://urbanplacesandspaces.blogspot.com/2011/01/without-right-transportation-planning.html

          an example for Maryland:

          http://urbanplacesandspaces.blogspot.com/2019/10/a-transformational-projects-action-plan.html

          I plan on writing a series about multi-state rail planning, based on how JR is set up in Japan, with six districts. There is overarching connections between regions, and district, multi-prefecture, prefecture, regional, and local service involving a variety of operators. (If you don’t know about the NHK World tv show “Japan Railway Journal” you should watch it — it’s also available online.)

          I see 6+ districts in the US, with overarching, multistate, state and sub-state planning.

          Or how the VVs do it in Germany, but on a bigger scale:

          http://urbanplacesandspaces.blogspot.com/2017/03/the-answer-is-create-single-multi.html

          • Herbert

            You can build a system with TOD at the same time.

            They didn’t call it that, but that is exactly what East Germany did. They built the line (usually tram or S-Bahn in one case the eastern U5 extensions in Berlin) first so that even the construction workers could use it and then they built the housing.

            Study after study shows that people are most likely to change their mobility behavior after big changes in their life – a car breaking down, a child being born, marriage, a new job, losing a job, moving. Most of those are not within the control of government (and shouldn’t be) but moving to a certain extent is. And if you give people the transit before they move somewhere, chances are, you’ll get better mode share. But if you try and build transit into a neighborhood where “everybody drives”, you’ll have to do a lot to convince people to consider transit…

          • Car(e)-Free LA

            This is a good point. So far as I’m concerned, the biggest issue is NY/NJ/CT and the MTA and PATH being separate. Regional rail integration is one thing, but there isn’t any reason except state-line politics that’s prevented Hudson, Bergen, Essex, Union, and Passaic counties from being laced with subways. Montclair isn’t much further out from Manhattan than Jamaica is.

          • adirondacker12800

            there’s a big swamp between Jersey City and Newark.

    • michaelrjames

      Henry, that is completely the wrong mindset, and why the US doesn’t have good passenger rail. It’s a “let them eat cake” attitude by people who have zero intention of ever using the bus (and probably rail but …). And then, if a bus service is established they say that the route is sufficiently served, especially as a bus route will never build the same patronage as a rail service so the argument is reinforced by the “patronage doesn’t justify rail”. Of course politicians love this because buses are so ‘cheap’. However as I have said many times on this blog, it turns out buses may be the most expensive form of mass transit for the very reason that in the Anglosphere at least, it is used to delay proper (rail) transit for decades and decades; from when it was much easier and cheaper to when it is more difficult (NIMBYs, political opposition) and vastly more expensive. That’s the real impact of buses.
      Plus what Alon wrote.

      • Herbert

        If you want to really get a bus that “does what rail does” you need its own right of way. And even if you have that, the bus still emits stinking exhaust, jerks about and has low capacity…

        Why not build proper rail to begin with? Especially in times of near zero interest rates…

        • michaelrjames

          If you want to really get a bus that “does what rail does” you need its own right of way. ….
          Why not build proper rail to begin with?

          I agree, and what you describe is BRT (if it doesn’t have ROW it can hardly be bRt). The Brisbane BRT does have its own ROW except for a few places. It has its own lanes across the main bridge into its underground tunnel thru the CBD but they were too miserly to give it a ROW at both ends of the bridge so it has to cross regular streets and typically this ends up being a bottleneck at peak hours.
          And that is why such cities build a BRT to begin with: budgets. However I have read that the designers did make all the newbuild compatible with light-rail if they wanted to convert it in the future.

          As to pollution, yes, though the whole fleet is LNG. Until the end of the 60s the city had trams and electric trolley buses.

          • Herbert

            Shutting down BRT for conversion to rail won’t help ridership…

          • michaelrjames

            Herbert: “Shutting down BRT for conversion to rail won’t help ridership…”

            They don’t need to shut it down. The conversion would be one of the easiest possible. And they already do running repairs without stopping service (by closing one side and diverting buses to the contraflow side for a few hundred metres). But night work would get most of it done (I’m thinking of the way Paris built T3 across its 30 or so major radial arteries without blocking traffic at all.
            However you may be right in that it probably scares the pollies.

          • Herbert

            I don’t see how you can lay track in a single night. Especially not along the whole stretch.

            Certainly not if you wish to have “fancy” stuff like Rasengleis.

            And piecemeal conversion leads to trips that aren’t one seat in the interim…

          • michaelrjames

            It’s not the whole stretch but the point is that what is laid that night can be driven on by the time traffic is ready to use it next a.m. It uses pre-cast slabs–it’s either the LR55 system or something very similar. I can’t find the bit about the traffic but I distinctly remember reading how it was remarkable that none of the 30 crossings (of major radials leading out of Paris across the Peripherique) were ever blocked during peak times.

            http://www.railwaygazette.com/news/features/single-view/view/paris-t3-forms-a-green-corridor.html
            Paris T3 forms a green corridor
            01 June 2007

            Construction has involved the use of pre-cast concrete panels at the 35 road crossings along the route, installed complete with rails. A typical crossing requires between four and six panels ranging in length from 6m to 8m, each weighing between 20 and 25 tonnes.

      • RossB

        Oh come on. You are saying that the reason the U. S. hasn’t created a good rail system is because they have spent a tiny amount of money on a really bad bus system? If you can’t afford to run the bus any more than once an hour, do you really think they are going to spend a ton of money on a subway? That is silly. America has a poor rail system for the same reason it has a poor bus system: poor planning along with lack of interest and investment.

        • michaelrjames

          I didn’t say that. I’m saying that providing a bus service, poor or not so poor, is no substitute for rail. And that too many people like Henry and politicians and the car lobby delude themselves that it is “good enough” (for something they never plan to use themselves). It is that mentality that retards serious transit in the US; more than actual money which is a secondary issue. If it weren’t for that mentality there would be more money etc.

          But as Alon said, even with its best mass transit. like NYC subway, it is sh!t compared to the rest of the world.

          • Henry Miller

            You can provide a great bus system now for large parts of most cites for less than the cost of a rail system. See the topic here, a rail line is not a system, my tiny city of half a million would need 300 miles of rail to have a great system which would take years to build even assuming infinite money and competence in design, or we can put buses on the roads we have and have a great system in a few weeks (what is the lead time to order 400 buses and train drivers?)

            Once you have a great bus system upgrade to rail where it makes sense. Some existing roads are near a freeway and can hop on that to cross the river. Others are farther away, and could be made better if we upgraded to a train (or brt) that included some sort of crossing.

            Get the system working now with what you have, then upgrade to better. If you won’t invest in a good bus system then you won’t follow through on a rail system over the next 30 years

          • RossB

            I didn’t say that. I’m saying that providing a bus service, poor or not so poor, is no substitute for rail.

            OK, that is different than what you said. You said that bus service is “why the US doesn’t have good passenger rail”. I’m saying that is ridiculous. If anything, the existence of good bus service should, if anything, lead to good rail service. The problem is that the U. S., in large part, has poor bus service. Do you really think otherwise? Do you really think that the U. S. has invested too much in bus infrastructure and service? Seriously?

            Regardless, let me go back to your latest statement.

            I’m saying that providing a bus service, poor or not so poor, is no substitute for rail.

            That too is ridiculous. A good bus system is much better than a bad rail system. The streetcar lines in Seattle perform very poorly, mostly because they follow a ridiculous route, get stuck in traffic, and are thus slower than a bus (the buses often get stuck in traffic, but they don’t get stuck if a parked car is a few feet into the street).

            The U. S. is full of bad rail systems, some of which cost plenty of money, and have middling ridership. It is a matter of picking the right tool for the job. Quite often — given the low density, major existing freeway infrastructure and trunk-and-branch nature of U. S. trips, the right tool is a bus. There are a lot of American cities that would love to have the bus system of Brisbane, but instead have really bad, really slow rail, that they bought because the didn’t quite understand the demographics of their city. (https://humantransit.org/2009/11/brisbane-bus-rapid-transit-soars.html).

    • Herbert

      If your last paragraph was right, beetroot stations like Limburg Süd, Haute Picardie TGV and so on would never be successes. The opposite is the case.

      • Yom Sen

        Are they really a success?
        Haute-Picardie had 376k passengers in 2018, around 1000 per day. Not that bad, but not a tremendous success. Amiens has above 5M passengers per year.
        Champagne-Ardenne-TGV and Lorraine-TGV were built at the same time.
        Choice was made to build Champagne-Ardenne station very close to Reims, connect it to TER network with trains to Charleville or Châlons and extend Reims tramway to the new station. For Lorraine-TGV, it was chosen to build it in Louvigny between Nancy and Metz where it was cheaper and with good road access instead of Vandières above the busy regional railway Nancy-Metz-Luxembourg.

        If I remember correctly, initial plan was for both to be served only by “Province-Province” TGV, like Lille-Strasbourg and Bordeaux-Strasbourg + Paris-Frankfurt for Lorraine-TGV while all trains to Paris would go from Reims, Nancy and Metz central stations.
        In 2018, Lorraine-TGV had 583k passengers and Paris-Frankfurt trains don’t stop there anymore.
        Champagne-Ardenne which serves a much lower population had 969k passengers.

        • Herbert

          Interesting that you don’t mention Limburg Süd & Montabaur at all. Even though those two communes have the highest share of BahnCard 100 owners in Germany…

          • fjod

            That’s not surprising: two small towns with disproportionately high connection to intercity trains attract a disproportionate number of people who ride intercity trains a lot. Is their ridership level actually more successful than stations with a similar level of service in more densely populated areas?

            Also: Henry refers to stations at ‘either end’ of a line, not to intermediate stations (which have much smaller marginal costs). The German government didn’t decide to connect two small towns to HSR of their own merit, it did so because they lie between Cologne and Frankfurt. If those cities didn’t have developed urban public transport, the case to connect them would’ve been much worse.

          • Herbert

            The federal government decided to let the train stop there because the prime ministers of the states in question said “otherwise we will delay the project”.

            Btw, the same line also served Siegburg instead of Bonn.

            Oh and among others the “United Internet” (owner of 1&1 and the likes) is headquartered in Montabaur right next to the ICE stop…

            On a related note, what is your opinion about regional train service to the Ilmenau-Wolfsberg thingy on the Erfurt-Nuremberg HSL? https://tools.wmflabs.org/geohack/geohack.php?params=50.685855_N_10.997529_E_dim:2000_region:DE-TH_type:railwaystation&pagename=Bahnhof_Ilmenau-Wolfsberg&language=de&title=%C3%9Cberholbahnhof+Ilmenau-Wolfsberg (coordinates for input into your favorite mapping app, as fetched from de-wp)

          • Alon Levy

            I… don’t have strong opinions? I don’t know how important Ilmenau is in the region or how much travel demand there is from there to Erfurt and beyond. Presumably such regional service would slot into the Erfurt-Nuremberg line to have a short transfer at Erfurt to trains to Berlin, but I don’t know how big Ilmenau is so I can’t tell you whether it’s a good use of operating resources.

          • Herbert

            Ilmenau has a bit under 40 000 people on an area of a bit under 200 km² (60% of which is uninhabited forest) It also has a small university with a bit over 5000 students…

          • fjod

            I don’t deny that, given the right kinds of stimulus, intermediate stations on HSR could occasionally stimulate significant growth. I just think there are a litany of examples where this has been done poorly in underpopulated areas (Haute-Picardie, Lorraine, East Midlands, Puente Genil, Noorderkempen etc.), to the point where the cost of slowing the trains down to stop at the station isn’t even worth the benefit of the station’s existence. And Ilmenau, just in terms of site (at the top of a hill, far from connecting transport corridors) looks like one of these.

            I think also perhaps the Montabaur case is different because it sits at a sizeable micropolitan area (40k inhabitants) that was previously very poorly served by rail. It’s now maybe overserved, but you could argue that given the choice of either overservice or underservice, the former is better. It also has bus and local rail connections, so avoids the worst of the ‘beetfield’ criticisms. Perhaps it’s more comparable to Ebbsfleet than Haute-Picardie – although Ebbsfleet has been underwhelming in its economic impacts when compared to Ashford, the other HSR station in Kent.

          • Herbert

            An overtaking station had to be built at Ilmenau regardless. It was decided to “build it in such a way that platforms can be added later if needed”. At the time of its construction, Thüringen’s CDU led state government did not wish to pay for regional train service to this station. Now Bavaria has ordered local trains to serve the Erfurt-Nuremberg corridor. Adding the stop in Ilmenau should be doable if people want that…

          • Yom Sen

            I did’nt mention Limburg because I don’t really know it and don’t have stats…
            Montabaur is definitely not a “beetroot station”. It is quite close to the town center and is served also by regional trains. Limburg Süd looks more like a beetroot station, I think it’s the only one in Germany, correct?
            Inconvenients of beetroot stations are:
            1. They are far from the city centre which is not only a favourite destination for many people but also a hub for urban transit
            2. They are not connected to the regional railways, meaning that you have to run parallel bus services just for the station with most people coming by car anyway, so you cut any network effect.
            Obviously, point 1 is not a big advantage in the case of a small town like Limburg and the station is actually just 3km from the centre. Point 2 does not concern the town itself and seems less critical here since no major city would benefit from this.
            Advantage of being outside the city, better access by car and parking, is probably more important in this case.
            As their names indicate, main purposes of Haute-Picardie and Lorraine stations are to serve the whole regions and not the small villages nearby (and indeed beetroot fields in the case of Haute-Picardie…) The fact that they both miss for just a few km the connection with existing railways linking the largest regional cities is really a pity.

          • Herbert

            Montabaur had an old railway station at a different location which it decided to shut down with the opening of the new one. Limburg decided to keep both stations open. At any rate, much of what you see near either station is new development.

            And stations that are not exactly in the heart of the town they allegedly serve were par for the course in the 19th century railway boom. Sometimes they built a branch to the city itself, sometimes they just hoped for development to occur near the station and sometimes the station was eventually given up anyway…

          • Sascha Claus

            Limburg and Montabaur are special (like everone else 😉 ), because they are in commuting distance to both Frankfurt and Cologne (and Bonn), making them a place to live for couples where one works (or wants to work) in each of those cities.
            The “highest share of BahnCard 100 owners in Germany” are probably many of those commuters.

          • Sascha Claus

            Regarding a possible Ilmenau-Land station, you can already get from Erfurt to Ilmenau city in exactly one hour, so a transfer at Erfurt to a train and another transfer to a bus won’t knock anybody’s socks off.
            If direct trains from Berlin and Munich stop there, that might result in an uptick for tourism; it is not totally unknown in Germany to be picked up at the station by your host.
            But if it ever happens, that would be more of a political thing to ‘support a region that’s left behind’, which caught a lot of media coverage and poltical rhetoric in the last years.

          • Herbert

            At any rate, I think it was good they built it in such a way that a stop is no major headache in the future…

  4. michaelrjames

    There’s an ultimate proof-of-pudding issue here. Americans have to a good approximation never seen a working public transportation system. At best, they’ve seen a megacity where people use the trains even though they are dirty and expensive to run because there is no alternative and construction was done 100 years ago when costs were lower.

    Yes. It’s similar in Australia though because we are so well travelled there is an sizeable constituency for rail, and an overall favourable polling for HSR despite its cost. It’s really only a certain type of reactionary conservative politician & business type (led by a certain media tycoon) who are so rabidly against it. And they’ve turned it into a joke: “if there is new chatter about HSR linking Melbourne-Sydney it must be election time.” Other than it never being done, my fear is that a miserable half-hearted effort will be done that kills the idea for another 40 years. I think there are deliberate attempts to undermine HSR by these means, including to keep inflating its cost, or to downgrade its service until it is useless (whereas our distances demand the fastest HSR possible).

    The CaHSR or HS2 debacle hasn’t been good for HSR enthusiasts here. The rescue from that is that increasingly we look to our northern neighbours and China, Korea, Japan, HK, Singapore. The current Labor leader has been a long term enthusiast and is talking about it in the context of post-covid stimulus, but I can’t see this conservative government ever doing such a thing. And the neolib economists are saying spending that kind of money is the last thing needed. However people are less convinced by those austerity and budgetary arguments since this govt pulled out a massive $300bn money dump when they needed to.

    Ever so slowly, the tide is turning. The cities’ roads are getting worse and worse, and the attempt to ‘fix’ them such as the WestConnex tunnel complex in Sydney is so expensive and then the tolls to use them will be so expensive, and everyone knows that those tunnels just shift the locus of the congestion from one point to another. Meanwhile the NW Metro and SW Metro (ultimately to join to form our first RER/CrossRail like service) are being built and will help build confidence.

    In the US one would think that the LA-Las Vegas route might be a good first HSR line, with its guaranteed ridership and no need for your own car once in LV. Problem is the route is difficult and thus more expensive. It would be a sub-2h trip so near perfect–with the scenery it would become a drawcard itself. But will LV survive covid-19? Well, a HSR line could be sold as one means of revitalising it! It is also a route that lots of Americans from all over the nation, and all walks of life (not just bizoids or elites etc) would get to experience so I reckon it would be transformative of opinion and politics. If it was done right.

    • SB

      Does the ALP has concrete HSR plan or is it “we are planning to build it after we win?”

      • michaelrjames

        Does the ALP has concrete HSR plan or is it “we are planning to build it after we win?”

        I think it was the Rudd government (or Gillard?) who commissioned AECOM to make a report on it, which included routes and how to reserve ROWs now (one can only laugh hysterically at that; I mean an Australian government spending money and political capital in advance of anything real being done!). But the report was widely criticized both by other transport experts and on comparisons with European HSR, ie. on the high cost $120bn, funny that almost a decade on that doesn’t seem quite so ludicrous, most of which was due to about 100km of tunnel and outrageous underground marshalling yards under Sydney’s Central Station. The environmental group (with a considerable enrolment of real experts) came out with their own report (link below) which some consider a better starting point. IMO, the likes of AECOM are a big part of the problem.

        http://bze.org.au/hsr-go
        ZERO CARBON AUSTRALIA HIGH SPEED RAIL – REPORT LAUNCH
        Posted on 13 Mar 2014

        • michaelrjames

          Off-topic but I can’t help posting it.
          Julia Gillard (Australian PM 2010-13) has just been appointed Chairman of the Wellcome Trust (perhaps just “Wellcome” these days?) which at the time I was funded by them (’90s, Oxford) was the world’s largest medical research charity. Research budgets bigger than many OECD countries, including probably Australia’s (which is close to the bottom of the OECD, per capita, including getting cuts under Gillard! even Labor don’t value R&D). I don’t know, but Gates is possibly bigger these days?
          Incidentally you have probably seen Prof Sarah Gilbert on tv news about the leading covid19-vaccine candidate from Oxford (Jenner Institute); she began as a post-doc at that same WT funded genetics institute. The Jenner Institute was created as a near neighbour and has as director Adrian Hill, a young star at the WT institute back then. One of the many examples of “seeding acorns” that bear fruit of unimaginable wealth decades later (what is the cost of a vaccine to restore the world’s economy? As the advert says: priceless.)

          Sorry about this intrusion, but I like to educate you transit geeks a tiny bit about the wider world 🙂
          Joking aside, who could have guessed back then that this work might lead to making transit safe to use again? Sarah’s work has been on a malaria vaccine but in the process they developed a particular approach (for very difficult evasive parasites) that can be quickly transposed to different antigens etc.

          • Tonami Playman

            I Noticed the Y connection to Canberra will cause reduced frequency between Canberra and Sydney or Canberra and Melbourne. I don’t know if a route without a Y has ever been studied. I created a possible alternative that allows through running and locates Canberra station North of the city next to the Phillip Avenue LRT station and the Thoroughbred park. It will require a 2.5km tunnel to go under the hill at Lawson park.

            https://www.google.com/maps/d/edit?mid=1AwmT-VBkaQlGgvy26rnLrvaeyzTKq81_&ll=-35.26460778262493%2C149.12980160113284&z=12

            It should add roughly 70km and about 20mins to the total length of the Melbourne to Sydney route. This will bring the total length to about 922km and estimate trip time from 3:21 to 3:41.

            In this alignment, all trains will stop at Canberra en route to Sydney and the LRT provides connection to the city center. It is quite a detour, and adds time to the Melbourne – Sydney trip, but it saves having to tunnel 50kms west of Canberra which might not bring that much time savings relative to the cost of tunneling.

          • michaelrjames

            [Sorry, managed to post to the wrong part of this thread, this is the correct one.]

            Of course the current plans (BZE use the same Y for Canberra as AECOM) have a substantial km penalty in that Y which means most trains will bypass Canberra. Many will say so what? It seems no way to overcome the geography.

            My only fantasy solution is that western route, from Wagga-Wagga directly east to the centre of Canberra. At least this saves on km of track and on travel time. Just the little matter of 50km of the Brindabellas in the way. However I have wondered if some of this can be conquered by combination of cuttings and higher grades for HST for the approaches, so not a base-tunnel but a halfway-up tunnel. Or maybe a cutting all the way. Or a funicular-HSR! I have no idea but was hoping these hills were typically Australian and, well, mere hills worn down to almost nothing by 150 million years of weathering. I wonder what Bradfield would have done? (There is a story that he got the elevation wrong by about 10-20m for his scheme to divert the Burdekin waters inland (to drought-proof inland Oz), which made the scheme impossible–in its day it had to rely on gravity to traverse the coastal range. I doubt this, considering he is the engineer who did the Sydney Harbour bridge-2 rail track, 2 tram tracks, 8 traffic lanes, in the ’30s–and would have given Sydney a world-class commuter rail network had they not stopped after building the first third of his plan.)
            Or take the WSP/PBQD strategy: cost to at, say $10bn, build one quarter of it before adjusting to the true cost ….
            Nah, seems a lost cause.

    • Herbert

      To give just one example of a maybe slowly turning tide, the original plan for Canberra called for a tram. The first line opened in 2019, nearly a century late…

      • michaelrjames

        That’s because Walter Burley Griffin designed our Frank Lloyd Wright city of Canberra in the golden age of trams. But, it took about half a century for the government to get serious (post-WW2) about building it and of course by then it was a car-based world. I haven’t been to Canberra since they opened the new tramway but I imagine it is very good. The city arterials are very wide and can easily handle them. I hope they build an entire network.
        But the conservatives were hysterically against it, including Murdoch national rags which had a kind of fatwa on it. Predicting it would be a white elephant and its cost was a scandal.

        FYI, Canberra has a Hare-Clark multi-member PR system which means progressives (Greens + a better type of Labor) have a lock on its government. The national political correspondents have to live in Canberra and it is partly or mostly this that makes the conservatives frothing crazy. The city is overwhelmingly middle- and upper-brow professionals and thus progressive (or loony green-left as they would say). Even the so-called Liberals (=conservatives) there are a progressive flavour which the electorate demands.

        I’ve never lived there or spent much time there but it’s only people who don’t live there who make those jokes about the capital. Last time I was there, after a day of business (grant writing), I was sharing a beer at an outdoor patio of a pub with Peter Doherty who did live there for many years. It’s where (at ANU) he did the work that eventually won him a Nobel Prize, and he said that living there was quite special because it lives up to its description of “bush capital”, very leafy and very Australian feel to it. It is the sole Australian city not on the coast.

        BTW, the first HSR should be built between Sydney and Canberra because it is only 285km. Then Melbourne to Canberra (650km) to make the Sydney-Melbourne link.

        • Herbert

          There are plans for expansion, but whether they survive Covid and the next election is anybody’s guess…

          A high(er) speed link to Canberra would be a good idea already because the airport is notorious for fog…

          • michaelrjames

            The airport is another obstacle because it is privately owned and would put up a ferocious fight against HSR, supported by some of the pollies who have to fly there from all over Oz.

          • Herbert

            I mean privatizing airports in the first place is a dumb idea, but you could buy them out or give them a small share of the railway…

            After all, English canal owners (despite being wealthy and well-connected) did not stop the railway boom for long…

          • michaelrjames

            I mean privatizing airports in the first place is a dumb idea, but you could buy them out or give them a small share of the railway…

            Don’t get me started. Privatising monopolies like airports was one of the dumbest things. Even the US doesn’t do that (to its big international or any strategic airports). Now when they fail in the market, we’re supposed to bail them out?
            I have thought that perhaps some kind of compromise could be reached with the frothing conservatives by experimenting with operating franchises, a bit like the Brits (yikes!), with two on each line/route. One could be offered to Canberra airport and maybe one to Virgin Australia (in the process of being rescued from bankruptcy). In fact with a 1h15m link to central Sydney (only 45m to western Sydney) they could try to steal some of Sydney’s air traffic; it would be faster than arriving at the future Sydney West airport and catching whatever shit transport options they’ll have for the first few decades (probably a $100 taxi or $35 coach). However, just like in the UK, I would expect it wouldn’t work so after 5 years or whatever they would go bust and it should all revert to the state (without compensation).
            You see how far I am willing to compromise to get some HSR built?

    • Alex Jago (@alexjago51)

      I have to say, when it comes to Aussie HSR my opinion is largely “If the Feds want to spend about $100 billion on passenger rail, can they start with the continuing catch-up on the lost half-century of investment in our cities?”

      But also yes, Sydney to Canberra is the logical first leg for HSR.

      • Alon Levy

        Ugh, Sydney-Canberra means paying like half the cost of Sydney-Melbourne and only connect to Canberra, much less than half the size of Melbourne.

        • michaelrjames

          Sure, but it’s the shortest HSR route so quickest to get built before a change of government or change of faint hearts, and cheapest (though the Sydney part is the most expensive per km). Once built, the rest would flow.
          Canberra is on the route Syd-Melb. It’s one third of the full route, which is twice what Paris-Lyon or Tokyo-Osaka. The scale (and cost) seems to defeat the imagination or courage of Australian politicians. We’ve got to make that first step easier.

          • df1982

            Sydney-Canberra would punch above its weight relative to population because it would connect the nation’s financial capital with its political capital. There is already a hell of a lot of commuting back and forth between the cities but nothing at present is really satisfactory (car/bus takes 3.5 hours, train 4 hours with a station in the middle of nowhere, and flying 1 hour, but it’s a pain to do so for a sub-300km trip and the fog makes it unreliable).

            It is also a manageable undertaking because you would only have to go to Sydney’s outer perimeter and then use existing tracks to get to the CBD. Doing so would still allow you to have 1 sub-1.5 hour service with 2-3tph.

            Sydney-Melbourne is much more challenging because, at a distance of around 800km there are very few compromises you can make while still keeping the trip under 3 hours there. Also the topography means it’s near-impossible to continue southwards from Canberra (without a 50km base tunnel), and so a branch just past Goulburn is the likeliest option.

            By the way Canberra Airport’s owners were offering to tip in a bit of financing to a HSR line… but only if the Canberra terminus was located in the airport. This would be a disaster, however, since there is a perfectly good approach that can get the train station adjacent to (what passes for) the Canberra CBD. So hopefully the kind offer is respectfully declined.

          • michaelrjames

            It is also a manageable undertaking because you would only have to go to Sydney’s outer perimeter and then use existing tracks to get to the CBD.

            Apparently Sydney doesn’t have enough tracks to do that, hence the proposal for 60+km of tunnel from western Sydney! I tend to agree that you don’t want to compromise the last bit, and you can’t share with commuter lines or freight lines, but that is an extraordinary large fraction of the total cost. I forget if BZE suggested an alternative cheaper means. The tunnel solution is a lazy fallback these days (for transit and road) because politicians are too faint-hearted to contemplate the disturbance of above ground construction. I’d like to see a serious estimate of viaducts above the existing rail–get the Chinese to tender since they seem so good at it (though at the moment it seems we are frenemies!)
            Indeed if I was paying for that much tunnel I would rather it was on the southern route direct to Canberra (and it would save on length of track and of course time) but agree, not likely. Equally I wonder if anyone considered old fashioned cuttings (can’t be that high, this is Australia … flat as…) with the consideration that HS trains can manage much greater gradients. (Don’t I recall Alon saying this was the French approach to avoid tunnelling?)
            However I also strongly believe all trains from both Melbourne and Sydney should stop at Canberra, versus the AECOM plan which has a Y about 20km north of Canberra.

          • Car(e)-Free LA

            It really isn’t that hard to get into the core of Sydney with minimal tunneling. This route would require 5.5 km of tunnels and get trains from Campbelltown to Central Station in about 15 minutes. Even at Anglosphere construction costs, it shouldn’t cost more than $4 billion AUD.

          • michaelrjames

            This route would require 5.5 km of tunnels and get trains from Campbelltown to Central Station in about 15 minutes. Even at Anglosphere construction costs, it shouldn’t cost more than $4 billion AUD.

            Maybe that was what BZE suggested too. I’m a bit too lazy to go searching thru their large report, maybe later. Did you look? I gave the link earlier. I think they estimated $80bn versus AECOM’s $120bn. One thing, in addition to the amount of tunneling, was the mind-boggling cost of additions to Sydney’s Central Station–this is what happens around the Anglosphere; witness the stupendous cost of renovating St Pancras, or what is proposed for DC’s Union Station, or SF’s TransBays/Salesforce Terminal (which at least was all new & tunnels etc.) There seems no real reason why a HSR station (addition to existing stn) suddenly becomes a multi-billion dollar affair, and in fact I doubt the mods to Gare du Nord for the TGV cost anything much at all (except Eurostar requirement of being a passport controlled zone). Engineers and rail geeks were bemused by the grotesque gold-plating nature of it.

            But at the time I recall someone who had local rail knowledge saying that the track was a real problem because the existing ROWs were too little and already stressed carrying south-western pax commuter trains and freight, and a HSR really needed its own. It’s not like Paris or London etc with their absolutely massive rail legacy from the 19th century, as well as Sydney being so geographically constrained on all sides (water and mountains). On this I think I cede the issue to the rail experts, which is not to say it couldn’t be done cheaper. (I’m for viaducts above existing tracks and some straightening on some bends (requiring the thing politicians hate: property resumption and a NIMBY storm) etc. Not necessarily to achieve 350km/h but to keep the speed up highish most of the way. The thing is that HSR is quieter than whatever existing trains and freight use those tracks today, and perhaps with clever engineering viaducting (with HSR on top) could even quieten down the existing trains. That would also “solve” the Central Station issue because instead of deep excavation (under existing train station!) it would all be above (though there is much exaggeration about the cost of doing that too…).

            At the end of the day I would live with the cost (hah, as if anyone cares …) but it comes down to where one wants to spend that kind of money (what’s $20 to $40 billion!), and I can think of far more productive things ie. associated with the HSR plan. Example, I want to create a new city on the mid-coast of NSW, perhaps next to an existing town (Coffs Harbour is often considered a good spot but of course you’d have Russell Crowe as NIMBY-in-chief because his getaway retreat is there for the exact reason that it is a sleepy quiet place, and far enough/close enough to Sydney!) and so need the feds to pre-emptively buy up about 20km2 of land to ensure it gets done the right way, ie. not the usual endless Australian sprawl and (like with BART come to think of it) get out of the utter obstructionism of local mayors, councillors and their NIMBY rate-payers (and a conservative state govt if they are in power). The locals will fight for a HSR station but then bitch endlessly at the noise, crowds and development it brings!.

      • Untangled

        Probably just me but my ideal initial Stage 1 Australian HSR route won’t be Downtown Sydney to Downtown Canberra, it’ll be more like an outer suburban Sydney to outer suburban Melbourne route. It’ll likely cost roughly the same as the former (roughly though) since it avoids expensive urban bits while giving much more people a big starting taste of HSR on a useful route (rather than to Canberra) and eventually extending it to the downtowns.

        • michaelrjames

          t’ll be more like an outer suburban Sydney to outer suburban Melbourne route

          That’s exactly the sort of half-arsed non-solution a lot of us fear! So most people in those cities spend more time trekking to some HSR station way out in the exurbs than on the fast train itself. They’d still fly.
          It is of course the hard bit of HSR, and for example why the Dallas to Houston HSR does exactly as you say—though the difference is that the highway systems of both Texan cities just possibly might support it. (More likely extensions to central Dallas and Houston would eventually be built but probably not with private capital.)

          Hasn’t the debacle of the half-arsed NBN taught you anything? As Tony Windsor remarked about the NBN: Do it once and Do it right.

          • Untangled

            I’m not gonna say I don’t fear it too, but I had similar fears about Sydney Metro building out to the suburbs first without a commitment to build to the CBD (that would by my only criticism when they decided to go Metro). But that CBD bit is now happening now and that is dampening a lot of my fears about HSR finishing in the suburbs in the first stage.

            As for whether people will still fly on a “half-arsed” stage 1, yes they would. Plenty of business travellers would but given there’s also a huge, miserable budget airline market in the corridor, I would say there would still be a sizable market for the half-arsed HSR to capture before the CBD connection is completed.

          • michaelrjames

            No, it’s really the wrong way around and risks it having low patronage in the early years. Then the nay-sayers and the road (& aviation) lobby would scream blue murder and say that no further taxpayers money should be spent on this white elephant that Australians have shown they are not interested in using. It would be different if it could use existing (slow) ROW to trundle into Central (as Car(e)-Free LA suggested) but apparently that can’t be done (or the relevant authorities/owners won’t agree to it; remember this will be a federal project but all the ROWs are owned by the states). That’s more or less what Lyon-Paris did initially for the last 10 or 20km (?) stretch into Gare de Lyon until it was upgraded; I’m sure the TGV’s raging success in that first year swept away any objections that the sceptics and bean-counters had.

            I’m afraid your suggestion is indeed like the half-arsed NBN, the so-called MTM, Mixed-Technology-Model, ie. some fibre but whatever old crap from the node to the premises even if it is 60 year old copper wire. During this lockdown there has been a lot more realisation of running such an inadequate system with more and more people working from home and netflixing etc. Remember that it was actually Tony Abbott’s fiendish plan to wedge his competitor Malcolm Turnbull who was a supporter of Labor’s NBN but, as a minister was forced to enact the inferior system to show he could be a team player while Abbott was (briefly) PM. For that kind of intra-party factional bullshit we got a crap and still v. expensive broadband. It was supposed to be a lot cheaper to build but wasn’t any cheaper so that the telco retailers are complaining of NBN’s charges (because this government is obsessed on showing it has a good ROI and can be privatised). Naturally it will turn out much more expensive because the original plan will cost even more and will be done some day.

            The plan should be to build–as quickly as possible to avoid political reversals etc–a HSR line that will be an uncontested immediate success, just like Paris-Lyon was. Canberra really is the best choice because it is the shortest line to connect cities and would have the traffic, and then it would be madness not to build the link to Melbourne.

          • Herbert

            Anecdotal but when I tried to buy a ticket for the Turkish high speed train when they still terminated in the nowhere of Pendik, I was told “there are no tickets for any train today or tomorrow – they are all sold out”. Of course TCDD for some reason did not then run trains all that frequently, but it shows that while having a good station location is important, it is not a hindrance to having okay ridership…

            And the line has since been extended to what can be called “Istanbul proper”…

          • Untangled

            I can’t see why the train can’t trundle to Central from where the HSR section ends. You’ll have to work out how service paths but I this shouldn’t be too difficult with a bit of organisation. At the Sydney end, with the Bankstown Line being taken off the existing network soon, that’ll open up plenty of paths for access to/from Central to Wolli Creek on existing tracks. Then on the East Hills Line after Wolli Creek, paths shouldn’t be a huge problem either though you’ll ideally also build 4-tracks all along the corridor (and then later build a fast tunnel to avoid the line). At the Melbourne end, it’s a bit tricky since the gauge issue but the ARTC-controlled Albion standard-gauge corridor is still a good option (which is in need of an upgrade regardless).

            As for the NBN comparison, as I said earlier with another comparison, I earlier was little worried the metro rail line proposal finishing in a suburb when there wasn’t any extension commitment, but given what has happened since I can be confident the same will happen for this almost-there HSR line. I’m not sure if the NBN is the right comparison because there is a broadband war in one party, but rail, on the other hand, doesn’t seem to suffer from this factional war to the same extent, there isn’t really a rail war. The Federal Libs doesn’t seem averse to funding rail projects, like the Melbourne Airport Rail Link. Even the Victorian Libs had a state HSR plan (granted there were flaws) in the last state election and the NSW Libs also has a state HSR policy. And for what it’s worth, personally, I don’t think the MTM is even the worst part about the NBN, *ahem* CVC.

            By all means, build it as quickly as possible. If we’re serious though, I can’t see why the urban bits can’t start construction before the opening of the initial almost-Sydney/Melbourne stage to lock in the construction and defray the low patronage/blue murder/reversal risk. Your argument is also based on the assumption that an almost-Sydney/Melbourne HSR would be less well received by the public in terms of public opinion and patronage than Sydney CBD-Canberra CBD, but I don’t believe in that will be the case, given just how insanely busy that air corridor is, I would bet that even the almost-there HSR would steal just enough air passengers for it to have more patronage than Sydney CBD-Canberra CBD HSR.

            But yeah don’t wait for the initial section to be completed and completed before starting the next section, start it as the previous section winds down. Even if you went for the Canberra option, I wouldn’t want to wait for it to open and then wait for it to be declared an “uncontested immediate success” before starting the Canberra-Melbourne section, why wait on that risk at all? Start the next section, remove the risk entirely and build it faster. That said, there’s certainly also a risk that the project could be cancelled beyond the initial stage regardless of the initial route option taken or the public opinion/patronage after opening, but at least we’ll get a decently fast Sydney-Melbourne train (even if it trundles to the centre) if we choose that option, instead of just a super-fast Sydney-Canberra train. It’ll be madness with either one of them sure but I would take one over the other.

          • michaelrjames

            I think you just to take seriously the toxic politics re all these transit projects. Just like with BART which irritates the purists or who believe the money wasted on it should have built SF Metro. You’ve perhaps forgotten who controls 70% of media in Oz (that American bloke who also owns the WSJ & Fox etc) and how rabid and no-holds-barred and dirty they will fight. So many so-called Liberals (conservatives) are either in thrall to him or terrified of his papers not supporting their re-election, they will destroy stuff they might otherwise have supported. Like the NBN (because of his Fox Australia interests Murdoch didn’t want competition that a fibre NBN would nurture). Even Turnbull, hater of Murdoch, couldn’t resist and venally went along with the stupid plan to destroy it. Being a self-confessed tech guy he didn’t have any excuses.

            If the politicians come to build such a thing then they must be as crafty as possible, and as clever, and most of all fast & strategic. Random hope and “she’ll be right mate” won’t cut it. The Australian political cycle is one of the shortest in the world and doesn’t ever give them long (and recently PMs even less). Again look what happened to the NBN–and for completely f’d up reasons. Or even the Opera House which luckily had gone past the point of no return when the conservatives got in power and tried to fk it up (and did that to the acoustics which hadn’t been built) out of pure ignorance and spite.
            The first HSR to be built needs, citing Master Jobs, to be insanely great.

          • Untangled

            tHeY rUiNd NbN™ sO sAmE lOgIc mUsT gO 4 hSr 2.
            And yes, a Sydney-Canberra HSR all the way in would be insanely great. Equally, I don’t see why an initial Sydney-Melbourne HSR with existing rail to access the city centre can’t be insanely great either (like how you cited what Paris-Lyon initially did as a good model with “raging success” even in “that first year”, I mean why bring that last-mile connection option up as a model at all when you disavow it now but whatever). I fly actually Sydney-Melbourne quite often as well.
            I won’t go into the BART SF city-limit metro vs Bay Area regional metro war.

          • michaelrjames

            I don’t see why an initial Sydney-Melbourne HSR with existing rail to access the city centre can’t be insanely great either (like how you cited what Paris-Lyon initially did as a good model with “raging success” even in “that first year”, I mean why bring that last-mile connection option up as a model at all when you disavow it now but whatever).

            I only just realised something that I suspect a Sydneysider would have got straight up. I think you were talking about a station in Western Sydney, yes? But the HSR route doesn’t go via the west, it goes via the South-West which has the only clear (low) path, ie. halfway to Woolongong before it turns inland. But anyway I am just going on what some train experts claimed re tracks on the route into Sydney. I am not familiar with the route and will now have to go and check on the BZE report which presumably compares theirs to the AECOM route?

            However if it is true that there could be “spare” track (that was also politically available and irrevocable, sigh!) then just maybe that is viable. It’s still a long way from desirable and risk future jeopardy but at least there would be HSR for 3/4 of the route. And assuredly it would be a considerable amount cheaper to build. Also I doubt there is much spare capacity anywhere in Sydney’s rail network so after a few years all parties would probably be interested in remedying it (well, maybe not, Australian politicians prevaricated on the second Sydney airport for a mere 40, or is it 50, years.

            My original worries remain. I do not want, under any conditions or pseudo-guarantees, a HSR that stops in Western Sydney. That would be a white-elephant and immeasurably stupid, and of course why I worry about it since our politicians have proven they are entirely capable of doing such things. Especially state versus federal and even Western Sydney versus Eastern! Some goon would say the Eastern pax could just catch a taxi, you know a mere $100+, maybe $200? Or provide a $50 bus that takes 75 minutes.

          • Untangled

            The BZE report route Route S1 (via East Hills) for Sydney is basically close to what I would like to see in the initial stage. Then later build the either BZE Route S3 or the AECOM option for Sydney (both of which are basically long, massive tunnels to Glenfield/Liverpool although the exact alignment differs). The East Hills BZE S1 potion has paths free even in the peak-hour today with even more paths opened once the Bankstown Line is removed (since it has to share tracks closer in today) and redirected into the new metro line. Adjust the approach into Central/Redfern a little and it should be fine. The new HSR section would stop in the Western suburbs (or south-west if you want to call it that) then it would just use existing tracks to the Central for the first few years or so before a dedicated connection. Same with Melbourne CBD connection (and maybe also Canberra spur would be pushed back too).

            I mean there could a station in Western Sydney in addition to the one at Central, at an existing station ideally, maybe either Glenfield or Campbelltown to start off with. There could also be a station in Melbourne’s North as well, maybe either the Airport or Roxburgh Park.

  5. Gok (@Gok)

    A lot of new US projects have jumped straight to frequent-ish non-peaky service and still have poor ridership. The Miami/Palm Beach train runs hourly. The Denver regional trains have 15 minute headways. SMART runs about 2 trains per hour which is about as much as a regional train that only stops at suburbs is going to be able to support. I’m sure you’re going to say that there are deeper “real problems” with them but even doubling service frequencies on those projects seems unlikely to increase ridership. The bigger issue is “starting small” in the sense of building a service in isolation, as opposed to in coordination with (say) TOD and anti-driving sin taxation.

    • Herbert

      Anti driving taxes risk to run into too much opposition. But TOD should be doable, right?

    • Amchadacela

      Brightline suffers from poor connectivity with local public transit, local governments and FDOT refusing to fund improved public transit connections to and from Brightline, and that they could not build out to Orlando in their first phase. A three stop 80ish mile corridor is similar to what was described in the blog post starting too small.

  6. Pingback: Thoughts on a Northumberland Public Transport Plan – Hogg1890 – Blog
  7. df1982

    I would say that the original Florida project was as lacking as the Ohio Hub, not because service standards weren’t great, but because in isolation Tampa-Orlando is not an ideal HSR market. The distances are too short (and the urban areas in question too sprawly) to make rail enough of an improvement over driving that it could be a sufficient patronage driver, particularly since the Orlando station was going to be at the airport rather than downtown.*

    With Brightline doing Orlando-Miami, however, I would think the project starts to stack up, and apparently Brightline is looking to use the same corridor to extend service to Tampa. Given that Florida East Coast already owns tracks all the way to Jacksonville, the state could end up with a decent T-shaped network incorporating Miami-Orland-Tampa-Jacksonville, all within HSR-competitive times.

    *Not that I think this was the wrong decision. On balance, an airport station with a light rail connection to downtown is probably the best solution for Orlando. A station directly in Disney World is also a must.

    • Herbert

      I hope Brightline doesn’t run out of money before their rail service enters the profit zone.

      I’m sure they can read “busiest air routes out of Florida” as well as I can and already have ideas for MCO-ATL. Of course when serving Atlanta one needs a stop that’s better situated then the current Amtrak station. Probably one at Five Points if possible and one at the airport. If Five Points isn’t possible, a stop at the airport at least gives them MARTA access and enables codesharing with Delta…

      • Amchadacela

        The problem is Brightline had to load up with debt to even get their first phase Maimi-West Palm Beach done, and fund their second phase. I do think in theory that Orlando to Miami section could make a small profit, but with all that debt they have to pay off, it might be too much to overcome. They have also been losing money on their existing service. The only way I see Orlando-Atlanta HSR happening if FDOT, GDOT, and the federal government pay for the infrastructure and the service could then be franchised out or operated publicly. With Republicans dominating FDOT, GDOT, and the federal DOT, I do not see it happening anytime soon.

        • Herbert

          Debt is not a problem if you can service it. The problem is that they still haven’t started Orlando service, so they’re still not able to make money…

          • michaelrjames

            Debt is not a problem if you can service it.

            The trouble with businesses that have a very high debt is that they are susceptible to any kind of event that disturbs their cashflow. That is why the airline industry is in such trouble. Globally they hold some of the highest debt to asset of any industry and it can only be serviced by that regular high cash flow … which suddenly became close to zero by April.

            Curious that F-Line thinks Florida East Coast will be happy to subsidise passenger rail with its freight profits. Maybe, but only if the expected property speculations pay off bigtime and ontime.

          • F-Line to Dudley

            Michael…that’s exactly what FEC is banking on. They’re invested in the property growth and make out like kings if that goes well. But, as mentioned, they’ve padded their existing profit margins already by getting a brand new 90 MPH freight speed railroad out of the deal. FEC has long operated on precision-scheduling practice that’s prioritized having fast track and time sensitivity on its mainline schedules, so they are one of the extremely rare pre-existing fits in the freight world where the pax investment in Brightline speeds actually widens their margins to meaningful degree. So they’re also not stacking the deck totally to home-run swings on real estate for paying it off. It’s doing them a solid right now today while their freights are still running at full schedule during the COVID shutdown. Obviously the major risk/reward is tied up in coattails from Brightline growth on its surroundings, but they aren’t standing still either at slowly amortizing the up-front track investment in Brightline from the enhanced fast freight margins.

            Grupo México obviously sees a portfolio more diversified than just the Brightline risk for billions it sank into buying FEC from its private-equity predecessors 2-1/2 years ago. As owners of Mexico’s primary Class I, they see big $$$ in spanning the upgraded FEC with Ferromex RR’s Texas interchanges for cross-Gulf services involving an intermediary like Norfolk Southern. And since the Class I’s are always horse-trading for buying/selling lines…more buy opportunities for carving out a self-owned midsection of that Gulf corridor. Brightline is an adorable little pawn compared to the chess pieces Grupo México is hustling across the board, and they’ve now got Virgin’s money directing the primary day-to-day at Brightline so it doesn’t have to occupy much of their mindshare.

            This isn’t a unicorn operation. The portfolio involved is diverse enough–and FEC’s incumbent position *specific* enough to that type of fast freight ops–that there was indeed a risk-managed fit here where you could hardly pick out a similar fit anywhere else on the North American continent. It’ll justify itself whether the real-estate coattails fire on all cylinders or are half-cocked. That’s merely the difference between them living large in 20 years off the proceeds or simply very slowly amortizing their initial investment with incremental gains. Either way the biz backing here is so extremely diversified an outright backfire is long-term improbable.

            And that of course is why Branson can afford to stick his neck out with the much riskier Vegas venture. Brightline is metastable over the deep multi-decade long-term from outright business failure. Challenges abound in real-time, and medium-term trying to alligator-wrestle “Florida Man” politics for better built-out transit connections…but the diverse interests backing the corridor grants Virgin the luxury of multiple decades to get it right. Branson also managed to insert himself into the FEC’s sale by getting their outgoing owners, Fortress Investments, to increase their stake in Virgin Trains USA while they were in the process of simultaneously cashing out of FEC. And that in turn provided some of the extra financial backing for Virgin’s XpressWest acquisition 9 months after the Grupo México deal. It’s all related. Virgin transferred its maximal risk to the Vegas corridor as a direct outflow of easing its deep long-term risk profile on Brightline. One doesn’t happen without the other. Again…you’ll never find another potential North American business comparison quite like FEC/Brightline and the sources of big bucks that specific corridor attracted as a function of right place/right time. So it’s accurate to say that this probably isn’t a reproducible template for something similar anywhere else without a whole ton more up-front risk. But where…and amongst whom…it did take place, the shoe happened to fit and project to wear pretty well over the long haul.

          • Herbert

            Once one is running one profitable passenger rail line, one can borrow against the earnings to build the next…

            And so on and so forth…

        • F-Line to Dudley

          Brightline will be fine. Florida East Coast is the largest Class II freight carrier in North America…so much larger than the next-largest that they expend considerable lobbying resources getting the Feds to adjust-up the cutoff between Class II’s and Class I’s so they can dodge the extra regulatory overhead of being re-classified as a Class I. They now have the longest corridor on the continent where >80 MPH freight speeds are doable (other than them it’s pretty much just P&W on NEC trackage rights in Connecticut), and that’s padding FEC’s margins considerably for time-sensitive freight and lower operating costs. And they’re awash in new sources of private equity capital from their 2017 acquisition by the Mexican mining conglomerate that owns Ferromex RR (Mexico’s largest Class I RR with U.S. interchanges throughout Texas & Desert Southwest). That corridor is a mega revenue producer right now this second even with the pax trains temporarily idled. FEC didn’t get into whole Brightline venture blind to “Florida Man” state politics and all the cycle-to-cycle volatility in the state house snatching defeat from the jaws of victory on transit initiatives. Their financial backing of Brightline was a deep long-term investment banking on slow/accident-prone political turnover eventually delivering the connecting services over multiple decades. The freight revenue and new freight owners across the Gulf larger than they are are part/parcel to the amortization plan for the Brightline investment.

          Plus on the other side of the coin you’ve got Richard Branson, who’s never shied away from dumping a personal fortune into his hobbies. Which right now is swingier to trains than air travel because of the structural dump the airline industry (including his own) just took that are going to take years to sort out. Plus Virgin Galactic has kind of been lapped by recently by SpaceX and Blue Origin in the public funding sweepstakes for private manned spaceflight, in part due to the major redesign SpaceshipTwo had to undergo after its fatal training wreck…so he’s in kind of laying low in maintenance mode targeting the zero-G tourist excursion market until the gov’t contracting opportunities loosen up. And Hollywood studios are taking a bit of a knee as the streaming wars wage on and now it isn’t doable to launch any blockbusters in front of crowds. It’s not coincidental timing that Branson-land resources started getting reassigned to Brightline and DesertXPress when the trending moved flatter in his other holdings, and now world events have put a deep freeze on things.

          Still anybody’s guess how the long-term picture is going to play out. I’m pessimistic overall at “Florida Man’s” staying power keeping FDOT a perennial transit headcase for the deep long-term. But the biz fundamentals backing Brightline are strong enough that they’re capable of waiting out an ice age’s worth of intrastate dysfunction. The biz plan was predicated on that stability from Day 1 being able to annually dwarf Brightline’s ebbing and flowing for the first decade-plus of service. Nothing that’s developed since–especially with FEC’s newly enriched coffers–has called that into serious question.

          • F-Line to Dudley

            Well, that’s Branson Fun Bux at work. He’s hedging his own fortunes on that route, but the relative certainty backstopping the Florida venture for the long-term does make it easier for him to stick his neck out on Vegas. The pieces fit together for Virgin Rail USA such that Florida’s inherently herky-jerky fortunes at the sub- decade level are not a supra- decade concern and Branson can fix eyes on other prizes. California giving them two issuances of tax-free bonds for the in-state construction to Victorville since start of the calendar year (sadly buried amid all the COVID news) is the most unabashed substantively good news that project–under Virgin or its umpteen vaporware predecessors–has ever gotten. So while they’ve got a long way yet to go, this right now is the most ‘real’ it’s ever looked from behind the starting gates. The other biggie about L.A. Union proceeding to design-build for run-thru service at long last is also huge, because that clears the decks at the congested terminal for the enhanced local traffic management badly needed to plug future HSR and more general-purpose intercity slots into there. It means Metrolink finally has its starting canvas for chunking together run-thru service patterns at major frequency increases and to start reforming its clunky ops portfolio. Which in turn is what’ll plug the project pipeline for big investments on the last mixed-traffic leg into L.A., electrification, and the like. Also still a long way to go, but L.A. Union run-thru is the biggest kickstarter catalyst for working that whole local bucket list down so Vegas-L.A. has a fully usable not-too-slow final connection west of Victorville.

            Also should be mentioned as quasi-footnote that Amtrak has pegged the 1997-eliminated Desert Wind as one of its Top 3 most-wanted LD trains ripe for a revival, and has stated for the record that this line greatly increases the likelihood they’d consider the revival. So that’s some modest additional backing that would probably come in quid pro quo form from the Feds as a rewarded perk for Virgin surviving the build gauntlet to ribbon-cutting day. The DW did Chicago-Salt Lake City on same route as the California Zephyr, then SLC-Vegas-L.A. currently un-served by anything except Barstow, CA to L.A. after the Southwest Chief picks it back up. It’s a relatively easy one to reinstate since it gloms onto parts of two of the more financially robust LD’s that could merit additional frequencies, especially if the new Victorville-Vegas alignment is so much more cut-rate on the schedule for re-backfilling the missing SLC-California section.

            So all and all, it’s looking comparatively rosy for Virgin West. At least in terms of its whole history to-date as a proposal struggling to “break the third dimension”. They’ve got a few more regulatory hurdles to go, but the funding pipes are starting to assemble such that real design-build is realistic possibility if they don’t hit any major speed-bumps (COVID thus far not being one of those with the recent good news out of Cali).

          • Brendan Dawe

            What are the other two most-wanted LD trains?

            And what would an Amtrak-over-Virgin-West look like? Superliners at a 100 over Victorville-Vegas and then on to UP? New equipment at 125? Salt Lake-Vegas with transfer?

            Are we assuming that the High-desert corridor happens?

          • adirondacker12800

            Why would anyone who isn’t an extreme railfan take a slow train when there is a fast train? Put a 100 mph train or even 125 mph train on the tracks the fast train catches up to it.

          • Herbert

            Mixing speeds is also a guaranteed way to ruin capacity. Which is why the new lines in Germany happen in the first place. The speed is usually a nice co-benefit, but the main reason they get built is that a 200 km/h train and a 120 km/h train cannot be easily mixed on the same tracks.

            So you build a new line for a 300 km/h train so that the 120 km/h trains get more capacity…

          • Brendan Dawe

            No doubt, and if they’re running full frequencies than a once-a-day Amtrak isn’t going to fit.

            That said, I ask because:

            1) It’s Amtrak. They’d do such a thing, and they do such a thing with 177 km/h LD trains on 217 km/h track every day
            2) It’s America. They’d do such a thing, and they do such a thing with 177 km/h LD trains on 217 km/h track every day
            3) It’s Express West. It’s only planned to 240 km/h anyway (unless virgin has changed this?) and even if it’s built it won’t have any wires or trackage rights over the mountains to go into actual population centres until some far off future where Palmdale AND the high desert corridor have been built.
            4) It’s a once-a-day Long Distance train. People don’t ride them to go fast.

            In summary, F-Line suggested this was being looked it. That doesn’t mean it’s a good idea nor does it have to mean it’s a good idea

          • Car(e)-Free LA

            The Express West plan is absurd. Victorville-Las Vegas at 350 km/h should be incredibly easy to do and the trip should take an hour, tops. Also, building this without a plan to actually get to Downtown Los Angeles ASAP is ridiculous.

          • adirondacker12800

            Amtrak finally got rid of the ancient dining car and baggage cars that were slowing down the long distance trains. They can use the local tracks for half the trip between Washington D.C. and New York. They don’t get in the way of the faster trains if they are on a different set of tracks.

          • F-Line to Dudley

            Relax…the Desert Wind was one-a-day and that’s all they saw in the revival, because it would give the Zephyr overlap a second frequency where it was seen as most needed.

            You realize the slow-speed traffic “mixing” Vegas-Victorville will by default involve about 3x as many hi-rail pickup truck inspection moves per day NOT on the public timetable…for purposes as mundane and varied as checking for rail heat kinks during desert heat waves and clearing dead animal carcasses off the tracks…than it will ever involve revenue Amtrak trains on the public timetable. MOW vehicles trawl every HSR line in the world between slots very definitely not going 300 km/h, and we don’t freak out about that fact because their presence is beyond inocuous amid the churn of high-speed schedules. One single LD revenue train doesn’t have any more impact on the works than the MOW schedule…but for the fact that existing on a public timetable at all makes it something over-tempting to think way too hard about. This is what track crossovers are for. The HSR trains will be taking the crossovers on given segments any which way on any which schedule when some poor sap in the hi-rail is scooping up a pungent half-eaten antelope kill underneath a gaggle of circling buzzards. Are we going to complain about THAT slow-speed mixed traffic too? Of course not.

            FWIW…4 routes got formal planning studies done after the 2008 Passenger Rail Improvement Act. DW and Pioneer (alternating Denver/Salt Lake & Seattle/Portland endpoints) were two big ones, since they operated in conjunction with Zephyr frequency increases and start out from a point with decent farebox recovery. Those are still the two most oft-cited. North Coast Hiawatha (Chicago-Seattle far north via ND/MT) was the third; that study came and went without any gained momentum and can now be considered D.O.A. Fourth one I can’t remember, but was a Florida service (either another “Silver Service” flavor or an umpteenth attempt at studying a Chicago route) and had a lot of routing complications to sift through, so also failed to gain legs.

            I dunno…take the existence of LD’s with whatever grain of salt you may. It’s a thing that exists and a few of them do alright enough revenue for existing where they exist, so as a thing that exists where it does decent business I guess there’s a couple route revival proposals that clear the threshold of being net-plus for coming back into existence. It’s not going to clog Virgin’s tracks and is an implicit blessing of fed support that can lead to another small funding release, so Virgin’s not going to say no to Nevada officials who pitch a stinking one-a-day DW trip on their track. As explained, it is completely and utterly non-impactful to the HSR schedule because it’s a myth that slow things aren’t already occupying track segments on daily basis on any ‘dedicated’ HSR line. This one just happens to be one single lash-up of 5 Superliners carrying people and baggage instead of a hi-rail disposing of rotting antelope carcass. Same difference if LD’s just aren’t your cup of tea. The high-speed train is blowing past on opposite track via the nearest crossover either way.

          • Ronald S.

            As the spouse of an ex Brightline (covid 19) manager I can tell you that the hub bub about Virgin/Brightline is all but dead. Brightlineis losing name recognition every day they dont resume service here in South Florida. They were bleeding dollars because they put senior managers from the hospitality industry in charge of a transportation company. For one example,they, operations management along with the predominantly hospitality managed marketing dept were giving away almost 30k in alcohol monthly with the attempt at increasing ridership. We all are actually appalled that Fortress Investment group mismanaged Brightlines finances. Personally I’ve never seen a company start then suspend business for 2-3 years then attempt a “comeback”. Its a strategy that defies logic. They should cut back on the freebies, drop prices a tad and operate just to keep their name and prescence out in the public eye.

          • michaelrjames

            Alas, that doesn’t improve anything, costs more and imposes time penalties.
            Of course the current plans (BZE use the same Y for Canberra as AECOM) have a substantial km penalty in that Y which means most trains will bypass Canberra. Many will say so what? It seems no way to overcome the geography.

            My only fantasy solution is that western route, from Wagga-Wagga directly east to the centre of Canberra. At least this saves on km of track and on travel time. Just the little matter of 50km of the Brindabellas in the way. However I have wondered if some of this can be conquered by combination of cuttings and higher grades for HST for the approaches, so not a base-tunnel but a halfway-up tunnel. Or maybe a cutting all the way. Or a funicular-HSR! I have no idea but was hoping these hills were typically Australian and, well, mere hills worn down to almost nothing by 150 million years of weathering. I wonder what Bradfield would have done? (There is a story that he got the elevation wrong by about 10-20m for his scheme to divert the Burdekin waters inland (to drought-proof inland Oz), which made the scheme impossible–in its day it had to rely on gravity to traverse the coastal range. I doubt this, considering he is the engineer who did the Sydney Harbour bridge-2 rail track, 2 tram tracks, 8 traffic lanes, in the ’30s–and would have given Sydney a world-class commuter rail network had they not stopped after building the first third of his plan.)
            Or take the WSP/PBQD strategy: cost to at, say $10bn, build one quarter of it before adjusting to the true cost ….
            Nah, seems a lost cause.

  8. Jacob Manaker

    I’m not sure your examples support your headline. If anything, they seem to support the conclusion that it’s better to do something small correctly than do something large poorly. For example: the same round of funding that was to support the Ohio hub also was to fund low-speed trains in North Carolina, the Cascades, Wisconsin, Florida (as you mention), California, Illinois, and Michigan. At European construction costs, the entire pot of money would have been enough to convert the Cascades to HSR; given how much people crow over Seattle’s modicum of transit use, converting the Cascades to HSR would almost certainly have developed a HSR constituency and expertise that might have avoided the CAHSR cancellation.

    The analogue in Europe is the piecemeal construction of intra-national HSR lines, which you’ve (soundly) castigated, because they inhibit international travel. But large areas of the US aren’t likely to see true HSR any time this century, simply because they have no large cities — for example, the Great Plains, Mountain West, and Appalachian South. With these large “dead zones,” HSR systems in the US are much less likely to link up.

    At the city scale, I’m not convinced that constructing an entire S-Bahn system is the minimum. In fact, your post on network effects suggests the opposite! A line or two through the least auto-oriented areas of the city may then make additional branches off the trunk through auto-oriented areas reasonable, when they wouldn’t pass muster in their own right. Unsurprisingly, constructing the auto-oriented lines simultaneously with the system core, rather than giving the city a chance to densify first, seems to be a common flaw with US transit plans.

    • Herbert

      You don’t have to build the entire network at once, but if your “starter line” is a stub, you’ll have problems…

      • RossB

        I think it depends on the stub. As I wrote below, consider the example of Seattle. They passed an initiative to build a very long line from Everett to Tacoma. They soon realized they couldn’t afford anything close to that. Some suggested they just go from the University of Washington to downtown Seattle. Instead they went from downtown Seattle to the airport. This is about 25 kilometers — I don’t think anyone would call it a stub. But it isn’t what they should have built first. Ridership was OK (by U. S. standards) at around 40,000 a day.

        What they should have done is run a line from the UW to downtown. It would have only five stations north of downtown, along with the four existing downtown stations that were part of the old bus tunnel. That would have been a much smaller line (at around 8 km). There would have been fewer stations. But ridership would have been much, much higher. When they added only two of the stations north of downtown (Capitol Hill and the one by Husky Stadium) ridership basically doubled. Only a few of those are riders connecting to the southern end. Most are riders going between downtown, Capitol Hill and the UW. Adding another station in First Hill as well another couple stations by the university would have much higher ridership than it had for years. It would have been far more cost efficient, which means that frequency could have been increased sooner (again, leading to more ridership).

        The point being no one would call the former a stub, but lots of people would call the latter one. Yet the latter one would have blown away the former in terms of ridership. I don’t see any problems created by that stub — in fact I see the opposite. Having such a short line would have forced the agency to focus on the stations, and making sure they are done right (or exist in the first place). In contrast, a relatively long line lead to cutting corners, and misplaced (and missing) stations.

    • Matthew Hutton

      I’m sure within the next 15 years you will be able to do London to Beijing by HSR, and almost certainly London to Singapore. That’s china’s belt and road initiative.

      • Eric2

        Definitely not 15 years. Maybe 50. At the moment Russia doesn’t even have plans for HSR further east than Kazan.

      • jcranmer

        You’ll have to do a gauge change to Russian railroads, and either go through Kazakhstan and Xinjiang to reach Beijing, or use the Chinese East Mainline going the long way around Mongolia.

        Singapore is even worse: you have to transit through Iran and Burma, as well as cross the Pakistan/India border. That’s some political resistance to cut through.

        • Eric2

          I think the idea with Singapore is Moscow-Beijing-Guangzhou-Hanoi-Bangkok-Singapore. Iran/Pakistan/Burma are much harder to traverse in terms of terrain, population distribution, and politics.

          • Herbert

            Iran is building some lines they claim to be high speed rail…

    • RossB

      I agree with all of your points, Jacob. More than anything, it is a matter of doing things right, even if it is small, rather than going big, but poorly.

      I can think of several examples, and they go back to what Alon wrote. You have to consider the entire system. Consider a new commuter rail system, with six trips a day (each way) for a couple hours peak direction. At 20 minute frequency it is less than ideal, but if the trip takes 45 minutes to an hour, quite reasonable. Yet only half a dozen trips is starting small. So is not building any park and ride lots (heresy for much of suburban America). But the key is everything else. Good feeder bus system, and decent bus service in the middle of the day, even if the buses run every half hour. That way, people can get to the station, and get home in the middle of the middle of the day. During rush hour, riders get a time advantage from the rail. Outside of rush hour, the bus is fast enough. This is likely to be successful out of the gate, if the corridor is solid. You can add additional service as well as park and ride lots later (if at all). If it isn’t successful, then starting bigger is just as likely to fail.

      Likewise, the same is true with subway systems. A relatively short (but fast) line, with lots of stops is likely to succeed, and have plenty of riders. It forms the basis for further expansion (new lines or extensions).

      The main thing is not doing things that are likely to fail. If you can only run a commuter train once a day, then it probably isn’t worth it. If you can only build a subway line two miles, it is also probably not worth it. Likewise if the train is stuck in traffic or at traffic lights. The main thing is to avoid making cheap compromises that will forever weaken the system.

      I can think of examples of this in Seattle. There are two new commuter rail lines, called North Sounder and South Sounder. Both involved buying the rights to use the rail. They both started small, but South Sounder was very successful, while North Sounder wasn’t. Additional trips were added on South Sounder, but ridership didn’t go up that much, because like a lot of commuter rail, service was heavily oriented towards peak service. There are plans for additional park and ride lots. I happen to think that is a waste of money (for reasons I won’t get into) but I can’t argue with the order they are building things. In contrast, North Sounder has been a bust. Very few people ride it. Any investment (in additional trains or bigger parking lots) would be putting good money after bad. Starting small (but not too small) was the way to go.

      Then you have the subway system. From the beginning, the focus has been on connecting the satellite cities in the area — a line from Everett to Tacoma. This is a dubious plan, and unfortunately, it has lead to lots of very poor decisions. Stations are poorly placed — or skipped. Ridership is OK, but not great. If, instead, they would have focused on the route from the U-District to downtown — with all the requisite stops — ridership would have been much better for years, even though the line would have been much, much smaller. It really wasn’t a matter of starting small, but a matter of starting small in the wrong way. Rather than starting with a short distance line, they cut corners with stations (or simply left them out).

      I think there is nothing wrong with starting small (although you can obviously start too small). The key is looking at the entire network, and considering whether the project — by itself, with no further investment — will actually be successful.

      • Herbert

        The marginal costs of more frequent service are much lower than the marginal costs of new infrastructure…

        • RossB

          Yes, but that nothing to do with my comment. Commuter rail in the U. S. typically involves leasing time on a privately owned railroad (e. g. BNSF), not building new infrastructure. Added frequency *is* the infrastructure.

          • Alon Levy

            No, the vast majority of American commuter rail by ridership, and I think also by track length, is on trackage owned by the operator or by another public agency (e.g. Amtrak).

          • michaelrjames

            No, the vast majority of American commuter rail by ridership, and I think also by track length, is on trackage owned by the operator or by another public agency (e.g. Amtrak).

            Which assuredly explains why all those decades ago BART chose to make sure they remained unentangled, at least physically (if only political disentanglement was as ‘easy’ as choosing a different gauge) with CalTrain or Muni and their web of sponsors and clients and agendas.

          • RossB

            Huh, I stand corrected. Fair enough. That is interesting because, according to Wikipedia … Outside the Northeast Corridor and stretches of track in Southern California and Michigan, most Amtrak trains run on tracks owned and operated by privately owned freight railroads. So Amtrak itself doesn’t own most of their own track. Anyway…

            Anecdotally, I looked at Chicago. The most popular commuter rail in Chicago is owned by BNSF. Canadian National and Norfolk Southern own other lines, but as you suggest, Metra itself owns some lines. That would make for an interesting chart — who owns the commuter rail lines.

            My guess is that you are right, most of it (by ridership) is owned by a public agency. This would make sense, simply because the Northeast (like all things transit) dominates in the U. S.

            But I also wouldn’t be surprised if most of the new commuter rail lines are leased, which is why they don’t run more often. I don’t know what the plan was for the Ohio Hub. I looked it up, but couldn’t find out (admittedly, some of the links were out of date, and I got tired of tracking them down via the Wayback Machine). The plan was to both speed up freight and passenger service. I’m pretty sure someone else owns it, so either they were going to buy it, or enter into some sort of agreement.

  9. smooth indian

    Interestingly the Ohio hub plan was developed by republicans in the early 2000s. They had an elaborate plan with potential train timetables to introduce high speed or medium speed rail throughout Ohio.
    The thing about public transport in the US that bugs me is the often poor frequency. Often bus or train services outside of peak hours run at ridiculous once an hour or once every 90 minutes. There is also neglect or complete lack of suburban and regional train services. With modest investments in track, signaling and stations Ohio can have a very useful regional train system linking Columbus with Cleveland, Toledo, Cincinnati, Dayton and Athens. Similarly Cincinnati, Columbus, Dayton and Toledo can run useful commuter train services using small capacity trainsets. However hostility from politicians and apathy from voters have prevented any progress on this front.

    • Herbert

      It’s ridiculous that Kasich is celebrated as a “moderate” when he killed a train project thought up by republicans…

          • DJ

            This is sort of thinking I can’t grasp. Why there isn’t next iteration of service design: “OK we need this # of trainsets and crews for desired service, but they’re gonna be poorly utilized. So how can we sensibly expand service to get them fully utilized?

            And BTW 3C corridor is exactly a case where starting small is feasible, you should be able get all-day 2-hourly service with just 5 trainsets of GE Genesis + 4-5 Amfleets + one cab car and with 6-7 consist, one could also do some hourly services in morning/afternoon. That’s quite small and it would be useful enough to drive further expansion IMO.

          • Andrew in Ezo

            @DJ- my understanding is that institutionally, passenger railroading in the U.S. is stuck in a 1930s, steam railroading mindset, where quick turnarounds and the practices and equipment that facilitate them are just not common (or nonexistent). Case in point- the recent announcement for the Amfleet replacement stated that new cars will be “capable of bi-directional operation”, which is incredible considering we are already 20 years into the 21st century (FWIW JNR installed user rotatable and reclining seats on its top rank intercity stock more than 60 years ago, I’m sure other contemporary railway systems also had such). In the U.S., some loco-hauled trains still need to be turned (!) at their terminal or nearby coach yard to get the seats in the right direction- this of course requires a dead-head non-revenue movement and shoots turnaround times to hell. There are also archaic maintenance and brake check procedures that slow down turnaround times.
            Ideally, you want to have 20 to 30 minute turnaround times at terminals, which gives enough time to unload the train, get basic cleaning done, re-crew including loading restocked refreshment trolleys, and have the next group of passengers board.

          • adirondacker12800

            If you want bidirectional trains you have to ask for bidirectional trains. … like the ones they have…

          • DJ

            @Andrew in Ezo, not completely, e.g. they use cab cars at least since 1960s.

          • Herbert

            There is some of the “nostalgia for the good ole days of steam” in Germany, too. Like people who say “Why should we build a new tram line when the old (single track, in a horrible state, not electrified, no train run in four decades) branch line is still there”. Of course this is partly a political ploy to stop the tram line…

  10. Amchadacela

    As sad as this blog post makes me feel, I know you are right. It feels like a lot of American commuter rail and light rail projects since 1980 meet what you defined. SunRail being dysfunctionally planned with infrequent service, not enough bus rapid transit/frequent bus service connecting into it, and using clunky diesel equipment really set it back. It is tough to argue as a transit advocate to put more resources into a poorly planned struggling service. But for Florida, it is so tough to get any transit money because of Republicans. Even if Orlando Metroplan came up with a world class rail and bus plan, our Republican Governor and Conservative leaning house and senate would oppose it.

  11. Reedman Bassoon

    The headline of this topic was “Resist The Urge To Start Small”. California High Speed Rail went big. It’s still big. It is spending big money. Nothing to show for it (other than well-paid construction workers, contractors, and civil servants) and no plan to operate a rationally-functional transit system.

    • Alon Levy

      Yeah, and starting small would not have changed anything, because NorCal-to-SoCal rail travel requires a tunnel between Los Angeles and Bakersfield and minor investments in Amtrak lead to Amtrak-level ridership.

      • Herbert

        Actually they have construction to show for it, which is farther than any hsr in the U.S. has gotten thus far.

        Eventually they’ll have enough track to start test runs. And that’ll draw eyeballs…

      • df1982

        The problem is that they started inside-out, when they should have gone outside-in. If you had built starter segments from San Jose-Merced and LA-Bakersfield (neither of which have rail connections of any kind at the moment), then at the very minimum you would have decent regional connectors at each end, and Amtrak-grade tracks crossing the middle section, which could then be upgraded to a full HSR line if more funding can be found.

        That way you would at least draw the socially depressed Central Valley cities into the economic powerhouses of LA and Silicon Valley. The present way of doing it creates a high-speed line connecting small-scale cities with each other, which won’t have nearly the same benefits.

        The irony is that it seems that this was done to placate Central Valley voters, when in fact they would have benefited much more from staging it the other way.

        • adirondacker12800

          They should had the Vogons do it. They have much more advanced technology.
          …they had the paperwork done for that section which they could then file for all the money available because the Republicans blew up the economy again. If they had tried to get funding for anything else they wouldn’t have had the paperwork available and somebody else would have gotten the money.

        • Richard Mlynarik

          The irony is that it seems that this was done to placate Central Valley voters, when in fact they would have benefited much more from staging it the other way.

          Nope, this is the way WSP/PBQD and allied consultants always work.

          Start at the far end or the middle. Create a stranded “investment”. Too big to fail.

          They’re doing it with their BART extension “to” “San Jose” right now — start digging in Santa Clara, and when they “unexpectedly” run out of money or into “unexpected” technical difficulties there will be “no choice” but to find more money and continue — after all nobody wants a tunnel that connects to nothing, do they? They do it with all their projects.

          As for where California HSR should have started, yes, pretty much right. Tracy-Fremont-San Jose and Sylmar-LA-Anaheim to get started, then Sylmar-shin-Bakersfield, Redwood City-Fremont, Tracy-Stockton then Tracy-shin-Bakersfield straight shot to connect the dots as directly and with as few roadway/railway/etc crossings and impacts as possible, then Stockton-Sacramento. Would and easily could all have been done by now, at something much less than half their cost.

          Instead, WSP/PBQD always goes for maximum public impoverishment, maximal self-enrichment, decades of trough-feeding, and off-the-charts levels of rent-seeking fraud.

          • anonymouse

            I don’t think it was done to placate voters, I think it was done to placate the politicians who were the only ones putting constraints on the planning process. And the problem there is, the actually important sections (the mountain passes) run through empty land where there aren’t any politicians. That’s why, when talk started about phasing construction, the thoughts were starting in the Bay Area, or in LA, or in the Central Valley. Because that’s where the population, and the politicians, are.

          • Richard Mlynarik

            The politicians don’t “put constraints”.

            The engineering consultant tells them what constraints to put, and they do as they’re told.

            I watched it happen, in real time, at a dozen city, county and regional agency board meetings, in real time in the early 2000s all around the Bay Area. It’s cheap and easy to arrange. The people who suddently discover “issues” that need “placation” turn out to be really cheap cheap cheap to buy, and the payoff (for example, WSP/PBQD fully controlling a state and county agencies and all its contracts, including hiring itself as prime contractor) is measured in the billions of dollars.

            We’re talking investment payoffs measured in the factors of many hundreds, or even thousands. It’s super easy to put “constraints” just where you want them for that sort of return. Good business!

    • Car(e)-Free LA

      That’s…not exactly true. I’m certainly critical of some CAHSR decisions (Pacheco over Altamont…seriously?), but at the very least, there shouldn’t be a problem opening a San Jose-Bakersfield line in 10 years and getting it to San Francisco whenever the DTX actually happens. Connecting the Bay Area and Central Valley by HSR will absolutely be successful, no matter how flawed the planning process was. Eventually, the line will get to Southern California. There isn’t any reason to believe project completion is in serious jeapordy.

  12. Michael

    There’s the fundamental issue that the US directs transportation investment toward growth & the places with the most hope of being good transit cities have an entrenched anti-growth populace.

    You could probably design a fantastic subway network in, say, Cleveland, built at globally reasonable costs, upzone everything around it… but it would never get a dollar of federal funding, because the city has zero expected growth in the next 50 years. What we’re left with is cities that don’t want to grow, and cities like San Antonio or Tampa, which best case, become clones of Los Angeles.

      • Michael

        Maybe something salvageable with the legacy assets, but the point is, there’s no network. Here’s cleveland and munich shown at the same scale, with transit networks. We’d have to build 5 new lines in Cleveland, that would just be beginning to be a comprehensive system.

        We probably should.

        http://mapmerizer.mikavaa.com/#12;41.46506040099805;-81.68165755810546;48.1351253;11.5819805;cleveland;munich;roadmap;false;true;false

        And by the way. Cleveland is technically a much bigger urban agglomeration by most metrics than Munich.

        • Alon Levy

          The European metrics are usually tighter than the American ones, rather like MSAs and not CSAs. The internal German metrics are wider. For example, Upper Bavaria, which also includes Ingolstadt, has 4.7 million people. Including Ingolstadt in the Munich region is kind of weird, but then so is including Canton and Akron in the Cleveland region.

          • Herbert

            “Metropolregion” is a marketing term and possibly a ploy to abolish the states (which have dumb boundaries in many cases) without technically abolishing the states. Note that the “Nuremberg metropolitan region” by and large aligns with an irredentist definition of “Franconia” (minus the Hessian and Baden Württembergian bits) plus a bit of Upper Palatinate. And “Metropolregion Berlin Brandenburg” looks suspiciously like a backdoor attempt to make the failed 1990s Bundesland merger happen.

            So only someone stuck with a machine translator would take those as a basis for any statistics on “greater city x”…

          • Michael

            Agreed. The UN statistic puts the Munich 1.538M versus 1.763M for Cleveland. Obviously, looking at google earth, Cleveland sprawls across a 80+ km diameter while Munich mostly fits into about 25 km. The land use patterns are very different, but that’s the chicken & egg problem.

            Cleveland has nothing that approaches a comprehensive transit network, so everyone with the economic & physical means chooses to drive. And in a pretty rich country, that’s nearly everyone.

          • Alon Levy

            I really don’t think UN urbanized area statistics work very well for Germany. German metropolitan areas are not contiguous – there are small rural gaps between cities and their suburbs.

          • Herbert

            Yeah, because the closest translation for urban sprawl “Zersiedelung” is an unqualifiedly negative term…

    • Herbert

      America simply gave up on its “rust belt”. Germany did not.

      The result is Trump gaining votes in the rust belt… And yeah NRW is currently governed by Armin Laschet, but I think people in Germany’s most populous state are smart enough to throw out the idiot at the next state election…

        • Herbert

          East Germany – outside of Saxony – never had much heavy industry to begin with. And the most frequent complaint I hear out of NRW is “they are spending more to gold plate the east while our cities are falling apart”…

  13. No Body

    Why does the subway in New York work so
    well? Drivers in SC are not educated and cause many, many bad accidents. How did they pass their driver’s license?
    You Republicans looking for a place to bury bodies
    New road work not yet graded is an effective
    Burial ground. End pollution forever, NO MORE
    COAL. Limit emissions from BMW also -8 wheelers. Use electric trains and quit lying for even 5 minutes.
    Commutes are not necessary by cars. Save everyone now. I know the sound horse manure makes when it hits the trail.

Leave a Reply to Car(e)-Free LA Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.