Small Metros Aren’t Lean, They’re Underbuilt
Reece Martin does very good vlogs on public transportation, and has begun text-blogging more regularly, which I appreciate greatly. But a post of his from six days ago, talking about lean metros, misses a key aspect of short metro systems. He compares old legacy systems like Paris or New York or Berlin’s with newer ones, like Hong Kong’s, and points out that the newer ones are rather short relative to city size, saying that it’s a leaner, lower-cost way of doing things. But in fact, the reason we see such short metros relative to city size is not efficiency, but underbuilding, leading to overcrowding.
What’s a lean metro?
Reece divides leanness into two kinds. The first is the ability of some cities to build driverless metros with very short trains at very high frequencies, to save money on station construction. He gives the examples of Copenhagen and Vancouver. This is particularly common in Italy: Milan Metro Lines 4 and 5 have 50 meter long trains, the Turin Metro has 55 meter long trains, the Brescia Metro has 40 meter long trains. With this setup and with the generally low construction costs of Italy, even Brescia can afford a metro, in a city proper of 200,000 with a built-up urban area of 673,000 (and rising) as of 2011.
The second kind of leanness is just building fewer lines. He talks about Toronto’s system, with two main trunk lines, one branch line with a transfer to a main line, and a total of 70 km of length. He also brings up Hong Kong, which has, counting both proper metro lines and the two inherited commuter lines, around 212 route-km, with very high ridership. This can be supported through transit-oriented development, for which Hong Kong is famous. It can also come from strong bus-rail connections as in Toronto: a blog post from last decade that I can no longer find points out that York Mills has what looks like 14,000 weekday boardings on pre-corona numbers, despite low-density land use immediately surrounding it, because of the strong connecting buses on the Toronto grid, favorably comparing it with American metros like Washington’s.
Small station metros
I don’t want to criticize the Italian trend too much, but I do want to separate it from the other systems Reece calls lean. The issue with the style of construction used in Brescia is that it’s really good, if your city is the size of Brescia. Small-station, partly cut-and-cover driverless metros should be in the toolkit of metro areas of about a million people, in order to save money. Other tools should be heavily relying on legacy commuter lines (as in Zurich) and using trams if they’re available (as in Bratislava or Brno), and likely combining all three solutions when feasible (in fact Zurich has a large tramway network in addition to the S-Bahn).
In a larger city, such light metro lines are only useful in a very restricted set of circumstances. Singapore has short trains on the Circle Line – but the Circle Line has not been cheap to build, and its last section has been extraordinarily expensive. On a radial line, it’s a nonstarter. A large city needs the very high throughput of a driverless metro but also larger trains; those can be medium-length trains, like the 90 meter trains of Paris Métro Line 1, or longer, like the 138 meter long trains of the now-driverless legacy lines in Singapore, or the 200 meter long platforms of the Shinbundang Line in Seoul. If the line is too short, the city may find it needs to build another just for relief, as the area that was once thought peripheral develops.
Short metros
If a metro system is short, even if its trains are long, it’s not generally a sign of efficient construction in the city. It’s a sign of underbuilding and overcrowding.
Hong Kong has very high crowding levels, even with a system length that, counted properly, is not that unreasonable: the MTR’s total route-length is almost the same as that of the Paris Métro, which has 227 km, and its ridership is, on pre-corona numbers, slightly higher, 1.7 billion a year compared with 1.5 billion. Now, to be clear, Paris has very high ridership for the system’s size. I suspect the reason I’ve never seen overcrowding on the Métro is that the nature of Parisian job concentration is such that the lines that get overcrowded are ones connecting the suburbs with the city, that is, the RER and M13, rather than predominantly intra-city lines like M1. But the situation in Hong Kong is overall less one of leanness and more one of not being to expand as fast as it would like due to extreme construction costs, which are a strong contender for the world’s second worst, after New York’s. (Toronto is an even stronger contender.)
Then there are the developing-world metros that are just far too short for their city size. Hong Kong is a city of 7 million with a little more than 200 km of metro and commuter rail. Cairo is a metro area of 22 million with 100 km of metro. Cairene construction costs are high, but when, depending on how much one trusts dead links from 10 years ago, the city has the world’s highest rail ridership per km, it needs a lot more; that Cairo has 100 km of metro rather than 800 like Shanghai, a similar-size city in a country that, during its peak expansion, was about as rich as Egypt, is not about leanness but about the Egyptian government’s spending priorities.
For a middle-income country that wants to get out of the middle-income trap that Egypt is stuck in (or Brazil, home to the almost as underbuilt São Paulo Metro), China is a decent benchmark. So is Paris – France is rich but also, precisely because it’s rich, rather motorized by any developing country standards, leading to a modal split of about 43% public transit, 43% cars for work trips in Ile-de-France. Using these benchmarks, your city should have on the order of 30 km of metro and new commuter rail per million residents. If you have 4.5 like Cairo, it’s not efficiency, it’s total disinterest in the living standards of the urban population.
Pedestrian Observations 
Is Hong Kong’s metro usage more peaky than Paris? That can make a big difference to how crowded it feels.
I suspect the average trip length is a lot longer, i.e., ridership is a lot higher when measured as passenger kilometers instead of passengers. MTR obviously covers a much larger area, and serves trips comparable to those in Paris might be served by RER.
Correspondingly, ridership density should be a lot higher. While ridership density over time periods longer than an hour don’t exactly correspond to crowding, it should relate to crowding a lot more than just pure rider count per route km.
I think some reasone Hong Kong metro have higher crowdness despite similar ridership per distance of metro cobstructed could be because, a.) there are many mountains and waters in Hong Kong setting different area of Hong Kong apart, so regular commute distance for residents in Hong Kong could be larger than other cities, despite the high density, and b.) Higher modal split for short to mid distance transit trip on modes like buses, making metro ridership more likely to be longer trip (This is just speculation, I don’t have actual data to show this)
As for the length of HK MTR, I think it is more of a matter of policy direction than the cost of construction. After electrification of legacy rail into East Rail and the construction of three main urban lines in 1970s-1980s, almost all metro constructions in Hong Kong are primarily focused on serving newly developed communities instead of improve urban service, with the only exception of Shatin-Central Link and South Island Line East project and Island Line West extension in 2010s when there weren’t much newly developed area in HK. Which I think is contributed by the mindset of 1.) TOD being a mean to recoup cost and earn profit do not work as well in built up area of a city with little room for redevelopment, 2.) Other road based transit mode being privately operated and for profit mean enhanced rail transit in existing part of city would disrupt income and profit of existing transut business.
I mean they did also underbuild to save money. Before they went with the Modified Initial System the original plan had two lines crossing from Kowloon to Hong Kong Island.
Even now the MTR built ShaTin to Central Link with only space for nine cars, and abandoned Central South, in an effort to save money.
Building to new towns isn’t a bad strategy, per se, especially because Hong Kong needs whatever housing it can get and there’s no upzoning lever to unlock huge amounts of well-located housing in the same way as Toronto or even NYC (including Manhattan). Also, RE the ridership/mile:crowding: Hong Kong has only 8 lines (Disney and Airport don’t count)+1 HSR vs. Paris’s 14+3.5 RER+surface terminals. Parisian density might paper this over a bit, but Hong Kong is quite under built rail-wise.
Building to new town isn’t in itself a bad thing but they are virtually only building to new town. Which mean more and nore people from different new towns feed into the core network without improvement in core network capacity.
Government claim capacity on existing lines are improved by signal system upgrade and by relocating destinations to new towns, but there aren’t much room for signal system improvement beyond current service level I believe, and relocating destinations to new town as a mean to suppress traffic into urban area isn’t exactly something that work from experience of other cities I think.
“Using these benchmarks, your city should have on the order of 30 km of metro and new commuter rail per million residents.”
The Paris metro area is 13.125M (per INSEE’s new “functional area”/AAV in 2020. The metro is 227km, RER is 587km and Transilien is 1299km. Depending on if you count Transilien (is it not “new” commuter rail because it doesn’t serve the center?) this is 62km/M or 161km/M pop. Where do you get your 30km figure?
Also, shouldn’t area be a factor as well as population? The San Francisco CSS has 9.714M, BART is 212km and Caltrain 89-124km depending on if you count the long tail past San Jose that gets peak only service. That works out to 30.8-34.5km/M without including the fully underground portions of the SF light rail system. I don’t know anyone who would say that the Bay Area has an adequate transit system despite meeting your 30km/M benchmark.
The RER heavily uses legacy lines; if those are available, the metro area should have more route-km per capita. The RER’s greenfield length is much less than 587: the central tunnels are 17 km (RER A), 8 km (RER B + D), 1 km (RER C), 3 km (RER E, not including the under-construction extension), so 29 total; the greenfield branches are 21 km (Cergy), 15 km (CDG), 28 km (Marne-la-Vallée), 10 km (Evry), so 74 total. So it’s actually 301 km for Paris, around 25/million people and growing.
In the Bay Area it’s 212 km, yeah, plus around 13 for Muni Metro. So it’s comparable to Ile-de-France, yes. But there’s very little density of lines in the core, and too much expansion far out. On net, what the Bay Area needs is small expansions as measured by route-length densifying the core network: the Downtown Extension, a second Transbay Tube, a Geary subway, maybe an expansion of the Central Subway northwest if Scott Wiener gets to rewrite the city’s zoning code.
@Alon
That’s a peculiar way to measure RER length. The 587km is what is served by the RER, and it is constantly increasing, not usually by ‘greenfield’ construction but it all carries pax who don’t magically cross an extramural portal into a parallel-universe.
No single metric will be a fair comparison between such diverse cities and systems but one that might be better than others is pax per route km. At least it is a fairish measure of crowding. From Alexander (pre-covid):
City………Track………pax p.a ……pax/km
……………..km………….bn………..millions
Tokyo ……..310……..14.99.………10.49
HK…………..174……….3.9……………9.16.
Seoul………..331………3.46…………7.9
Paris ………..801………3.11………….3.7
NYC ………..373………1.87…….……4.6
London..……402……..1.72…………3.1
Obviously this is before the Elizabeth line which will bring some overall relief to the system but in the short term has increased complaints about crowding–because so many shifted to the Elizabeth line and they under-provisioned (short trains) the new line.
FWIW, Paris-M1 is a bit special compared to most other cities (but perhaps most similar to the Elizabeth line) in that it has an extraordinary density of heavily-used stations spread over its entire length, from both CBDs (ie. 9th + La Defense), the city hall, major entertainment zones (Bastille, Champs Élysées), retail precincts (Chatelet-Hotel de Ville; rue de Rivoli; Champs Élysées) and of course tourist sites (everywhere but esp. Louvre + Pompidou) and big interchange stations (Chatelet, Gare de Lyon, Nation, Etoile). It also is the only line that intersects all other 14 main metro lines and 3 of 5 RERs (RER-E soon). This means that its pax load is spread over a lot of the line, in shorter journeys between these various features and interchanges.
If I recall correctly, line 1 doesn’t intersect with 3 or 10. I think it’s line 4 that connects to every single main metro line + every single RER.
The Hong Kong number is 1.7 billion – you can onlt make it 3.9 billion if you include Shenzhen.
Right. My figure was directly from Alexander. It is boosted by his inclusion of the large cross-border traffic to/from Shenzhen. I’m not sure whether it should be included; they are pax that travel throughout HK on the transit system, and that former KCR line has been fully integrated into the Metro system, crossing the harbour and terminating at Admiralty with interchange with the 3 other lines there; it’s kind of turned into an RER line, so on that basis might be fair to include.
City………Track……….pax p.a ……pax/km
HK…………..174……….1.49…………8.5. (Metro lines only)
If you do Métro only, then Paris is something like 1.6 billion/227 km; it’s very busy. The RER has much lower ridership per km, but that’s always the case for long commuter lines, and those lines are still more crowded, is why I don’t like lumping the RER into metro ridership/route-km calculations. (Relatedly, the London Underground behaves in part like an RER, in the sense of having critically overcrowded central segments, and low-ridership tails in West London.)
You are undercounting London’s track kilometres and ridership. You should at least count the 167km of London overground and the 40km of DLR as well as their 189 and 92 million passengers respectively.
Certainly there is also a strong argument for also including Thameslink and the Elizabeth line – with a weaker argument to also include at least some of the C2C, South Eastern, Southern and South Western commuter services.
Isn’t 310km for Tokyo only including those few underground lines built by local government, ignoring all private/JR rail lines that these lines through run into, and also ignoring the like of Yamanote lines that take up metro functions for the city?
310km sounds like just Tokyo Metro + Toei Subway which carries only like 4 billion passengers per year. 15 billion passengers per year sounds like the entire urban/suburban rail network which is ~4700km long.
“The RER heavily uses legacy lines; if those are available, the metro area should have more route-km per capita.”
This makes no sense. First and foremost because in your article you say “Other tools should be heavily relying on legacy commuter lines (as in Zurich)” – what’s the point of relying on legacy commuter lines if you don’t count them as contributing to your transit line density?
Conceptually, why would a trip from Acheres-Grand Cormier to Chatelet count as a trip on valid rail investment but St Germain-en-Laye would not, just because A-GC first opened in 1989 while StG-e-L first opened in 1847? They are trips of about the same distance, using the same rolling stock, stopping at most of the same intermediate stations, and with identical infrastructure standards (its not like RER branch A1 is using rails and ties from the mid 19th Century). Again, this makes no sense.
If we were to look at the Metro this way you could say that Paris only has 48.5km of subway (M14 plus extensions since 1970) and discount the rest as “legacy tunnel.” Again, this makes no sense.
What’s more, the sections you are counting as valid would be useless without the sections you are not. RER A would not be one of the most heavily trafficked urban lines in the world if it was just 17km of tunnel in central Paris, many/most of those riders are coming to/from legacy routes in inner departments. Similarly the extensions to Cergy or CDG use legacy track before reaching the central tunnels, its seems foolish to count it as if a dozen isolated km in the suburbs would be valid for anything. Again, it makes no sense.
This logic also means that metro areas with existing line(s) across the center (i.e. Toronto) could never be considered to have an RER-style system because no matter how much you improved it (even if it offered better service or carried more riders than the Toronto Subway) it wouldn’t have any “greenfield segments.” Similarly you apparently would count the length of the <5> in New York since it is part of the “subway” but not the Harlem line of MNRR because the Harlem line is “legacy” – again this makes no sense since both lines were laid out as mainline track in the age of steam and they both use third rail EMUs at high frequency to Midtown today, it just so happens one line was absorbed by the subway while the other was not.
Further, 256km of your 301km is entirely intramuros, which means Paris proper has 121km/M pop, a far cry from 25km/M.
What matters is service, as @Michael alluded to. I can see not counting legacy lines that don’t have service to the center or do not offer reasonable frequency or all day/off-peak service, but when a route was originally laid out has no bearing on the service offered today or if it is contributing to the mobility needs of a metro region.
I would say the Bay Area needs much more that what you list. Van Ness/Mission is key N-S route in SF with bus ridership among several routes high enough to warrant a subway. Same for California St at least until the Richmond District, even though it parallels Geary. This should be continued as a subway running SW south of Market serving all of the TOD built there without Scott Weiner’s intervention (but also without the transit). The west end of the Twin Peaks tunnel should be continued as a subway to Park Merced. Central Subway to the NW was valid without upzoning based on bus ridership pre-Covid, maybe not now. BART could usefully branch to Oyster Point given how much office/lab space is being built there.
San Pablo/Intl Blvd is a N-S route for the east bay. At a higher scale the East Bay is also ripe for a Caltrain/commuter N-S service from Richmond to San Jose which would require a tunnel through DT Oakland and grade separation through Berkeley. The busiest bus corridor in the East Bay is Telegraph Ave, which BART skirts but doesn’t service, and which ends directly at UC Berkeley.
The wide boulevards on Santa Clara county are ripe for a coherent network of elevated lines along major corridors (El Camino, Stevens Creek, Alum Rock, etc.) to replace the wandering, inefficient VTA light rail.
Some of these corridors would be best served by “good lean” construction, but that’s fine, the geography of the Bay Area (with the Bay itself breaking the urban area into two long strips, and DT SF not in the geographic center) means that lines serving the secondary centers of the East and South Bay are in effect serving near-Brescia sized “metros” of a few-hundred thousand to a million people, with the total 9+M agglomeration focused on San Francisco, necessitating the high-capacity lines you listed.
All or most of that… Also more should be built (or should have been built) along the MacArthur Blvd/I-580 corridor in Oakland, but also a broad grid dependably linking from the hills to flats, or from both to the linear routes (given Oakland’s unusually cut-up street grids). BART, following the old Western Pacific RR corridor, often strays much too far from International Blvd. to be of much good to most East Oakland residents. Could tie in with a Telegraph or San Pablo Ave. network. And more BART infill stops in San Antonio/East Oakland, as BART to its credit studied some years ago but no action.
The Bay Area is in a weird position where nothing is going to make sense without upzoning, but just about anything can be justified if you assume upzoning.
In that context, I think it’d be shortsighted for Santa Clara County to commit tons of money to yet another new low-volume transit system–the NIMBYs who gnash their teeth at six-storey apartments are hardly going to let you build an El down the El Camino. If you’ve overcome that resistance, though, you’ve also cleared the way to start building enough housing to justify heavy rail (or at least, make it clearly shortsighted not to build a system that will support higher volume in the future).
VTA is doing a rapid transit corridor study that’s supposed to be released this month and the cities along Stevens Creek are in the middle of some kind of visioning process. It seems likely that without a new local funding measure and with limited upzoning, it’s going to be BRT instead of elevated rail. Cupertino is supposed to upzone its portion of Stevens Creek, but I don’t expect anything to be much taller than 6 stories.
The weird is that Milpitas, Sunnyvale and Mountain View have heavily upzoned their slow and wandering branches of VTA light rail while San Jose is neglecting its section of the trunk line along North First Street.
@John: “don’t expect anything to be much taller than 6 stories”
You don’t need any more to achieve the highest urban residential density in the world. Don’t argue for more/higher because it just creates more NIMBYism.
@Michael: “You don’t need any more to achieve the highest urban residential density in the world”
Problem is, e.g., in Cupertino, you’ll find roads <7m wide abutted by single family homes, and arterials 23m wide between blocks of those and shopping centers in acre-big moats of parking, and there are freeway interchanges taking up approx. 500m2, and freeways that are continuously 50m wide. Trying to reduce or get rid of any of THAT will cause NIMBYism, because no one here can conceive of life without a car; the catch-22 of “we have roads so we must drive, but we must ONLY have roads because we drive” is unnoticed.
That’s one of the reasons North American cities are so peaky; development is ONLY allowed in special circumstances. Therefore, ALL the demand must go into one plot, and so there exist economic pressures to build 20-30 story residential towers. I’m not saying the Haussmanian Parisian style of development isn’t worth shooting for, but you’re literally telling people to rebuild cities–in their entirety–to meet demand, and they just won’t. They’d rather have nothing of their own lives change, and have all the change where they don’t really need to deal with it.
And it’s still lower density than Korean/Chinese-style towers-everywhere; the math can’t be denied. The shortfall is because those places make the mistake of building hilariously wide roads and generous highways to accommodate car traffic that their transit systems make completely unnecessary, to say nothing of absolutely forbidding topographies making huge portions of land unbuildable. You can’t just look at numbers outside their context.
@aqaticko, building a 5 storey building next to a 2 storey building is much less oppressive and much less likely to upset the nearby homeowners than a 20 storey building would be.
The actual experience in the Bay Area is that these homeowners are about equally upset by high-rises, mid-rises, and duplexes.
@Tiercelet:
Zones. Around stations.
The maths is indeed simple enough. A circle with radius of 1.6km will hold 200,000 residents (plus all the retail and services) at Haussmannian density. A radius of 1.2km gives 113,000. There is no “peaky” zone in the world that even comes close. HK or Manhattan are not “peaky” but these are the only ones that come close; is that what you are trying to tell those NIMBYies what you are really planning for their precious acres?
HIMBY, not NIMBY.
@Alon:
They are BANANAS. We know. But actually they will be less enraged by some low-rise quality housing when it is built, especially a coherent zone of it, than by anything else. In fact if you look around, even in those brownfield redevo sites in SJ, actually there is more Haussmannian than hi-rise. And if one could get the stats I am sure you would find that it is doing the heavy lifting in the slow densification.
I reckon you’re both in denial.
To go Alon one further, the experience in the Bay Area is that these homeowners will virulently fight the construction of any expansion of transit within a 1.5 mile radius of their residence, on principle (the principle in question being a heady mix of racism and classism). The slightest deviation from car-centric SFH zoning is a call to arms.
It’s true; Imbaba, Lalbagh Thana, Ajeromi-Ifelodun–all mid-rise or lower.
How do they do it? Narrow streets and tiny, shared living quarters.
Now the narrow streets I’ve no objection to–it works great in Tokyo (and Kyoto, and Osaka, and Taipei, and a lot of other places with a better standard of living than the modern Bay Area). But if you want to combine high density and high living standards, you need high buildings. (With a runner-up alternative of grandfathered-in subdivided closets that still require an entire city uniformly lined with admittedly-lovely Hausmannian blocks.) Yes, you’ll get NIMBYs; but again, these are people who go feral if they see an eight-year-old car drive through their neighborhoods; they have long ago undermined any claim to having a point that society at large should respect.
@Tiercelet
This utter b.s. again!
That fits nicely into a TOD scheme. To get such a scheme enacted one can appeal to the main thing that interests Americans: their wallets.
As to the rest it is simply untrue. “Haussmannian” in an urban planning context means a mix of large boulevards and smaller, narrower streets & lanes, perhaps Woonerf. Provides excellent access, even for cars so those outside the 1.5mi TOD would still be able to access the retail within the TOD.
And, despite the snark, I am quite serious. It would work. However I agree that those entitled NIMBYs will still fight it tooth and nail, even though they would end up benefiting (their own property would increase in value and the closer to that 1.5mi radius the higher the value.
I expect that’s less to do with a desire for Haussmannian forms and more to do with not wanting to exceed the height where they can still do wood framing (or have to excavate for the parking garage).
@Michael
You say we’re in denial. It’s just one source, but Population around a point (https://www.tomforth.co.uk/circlepopulations/) says there are 8.5m people in a 20km point centered on Opera, and 13.5m people in a 20km point centered on Yongsan. A simple thought experiment: with Parisian geography/topography and roadway area coverage, would there be fewer or more people in a Paris that, in addition to Haussman blocks, also had 2-3 30-50 story towers every 1.5km’s or so?
@aquaticko
As I have said over and over on this blog, it is not in contention that if you built endless hi-rise one achieve hi-densities. The odd real-world fact is that you can’t find such zones with density greater than intramuros-Paris outside of Hong Kong and parts of Manhattan.
In fact what you speculate has already been done within Paris. The 13th arrondissement has about thirty 31-floor hi-rise apartment blocks that were built in an experiment beginning in the 70s but finished in the 80s. The 13th is the tenth densest arrondissement at ≈25,000/km2. The 15th which is similar in size is 27,000/km2 and is almost totally Haussmannian except for the pocket of BeauGrenelles which has some hi-rise (another 70s experiment that has never been repeated).
The point is that in the real world hi-rise≠hi-density. In which case why would you build hi-rise? Perhaps that is why Paris and even extramuros-Paris stopped building it 4 decades ago. The only pressure to build hi-rise residential comes from developers (who incontestably make more money from such a project) and starchitects and their clients. And the deluded who alway simplistically think it is the “easy” way to get density.
More to the point in this discussion I am sure Haussmannian type TOD in the Bay Area would be much more readily accepted than whatever mess you are proposing. I think it is hard to convince NIMBYs of any change at least partly because they don’t trust any change proposed by the usual suspects (developers and their proxy politicians). In fact it is happening, probably driven as much by the economics of building low-rise. The book Multi-Family Housing by Michael Crosbie has a bunch in the Bay Area to exactly address this issue. One is Cupertino Park:
It is perimeter-block (ie. occupies a whole block and is built to the sidewalk) and incorporates some aspects of mission style in having internal courtyards, and has striven to avoid the visual tedium of most American 5+1 blocks. In fact most of it is only 3 floors with a few places extending to 4 with the whole half-basement being car parking (with 3-4 floors above it); I reckon the design could have gone higher which would have meant 160 apartments. It doesn’t give the capacity but assume 300 residents (it appears to be mostly 1 & 2 bed) then the density is ≈49,000/km2, but of course that doesn’t include streets etc. But the point is that it certainly is a worthwhile housing density. It would take many square kilometres of such housing to make a real impact. If a 1.5 mile (2.4km) radius around a transit station was built to that scheme the 18km2 could have as many as 450,000 residents (at 25,000/km2 ie. about half the density of the Cupertino block).
I have a huge amount of respect for Alon and they have clearly had a big political impact in the UK. However there is definitely a need politically to make sure you don’t have too much opposition.
I am sure Borner’s is right that everything gets objections from the most unreasonable NIMBYs – but they are more likely to get a wider number of people in the community to comment unfavourably on a planning application if it is particularly unreasonable.
You can’t do a leafletting campaign through people’s doors to encourage people to comment on a planning application if a material number of people will actually support the project 😜.
Looking at the present isn’t informative without looking at the past. What is the density/population trend for these two arrondissements?
The 13th, which did the high-rise building you describe, was at:
1968: 22.1k/km^2 (before the building)
1975: 22.8k
1982: 23.9k
1990: 23.9k
1999: 24.0k
2009: 25.4k
An upward trend of densification during the period of high-rise building, then plateaus, with a net 14% increase over the period.
The 15th, which didn’t build up, had:
1968: 28.7k
1975: 27.2k
1982: 26.5k
1990: 26.3k
1999: 26.5k
2009: 27.8k
Loss of density, then plateau, and a late rally (visible in both arrondissements) for a net 3% decrease over the period.
(per https://en.wikipedia.org/wiki/13th_arrondissement_of_Paris and https://en.wikipedia.org/wiki/15th_arrondissement_of_Paris)
One of these places–the one that built upwards–got denser; the other one saw its density decline.
This is not consistent with the story that density is best served by avoiding high-rises.
Similarly:
Without cheating by looking at the developing world–Yorkville is 60k people per square kilometer. Manhattan community boards 7 and 8 (UES and UWS respectively) are both 42k per square km. (vs. Manhattan overall at ~28k/km^2–so comparable to Paris, with these neighborhoods half-again or more that level.) If this doesn’t constitute the kind of “peaky zone” that @aquaticko was talking about, I don’t understand what the term means.
Look, it’s the same thing every time this gets discussed. You claim XYZ does not happen, people give lots of examples (…Eixamples?) of it happening, you give a bunch of reasons those examples don’t count, etc. It isn’t terribly productive. Suffice it to say I’m happy for you to have an aesthetic preference for the Parisian built environment–I agree, Paris is lovely to look at! But I do not find convincing your arguments for why your aesthetic preferences constitute universal laws of urbanism.
—
As for Santa Clara County and the SF Bay Area: the developments you are pointing to are not Haussmannian, or even TOD, in any meaningful sense–because of street design and parking. Have you actually looked at the Cupertino Park Center Apartments? These, like the other mid-rise developments that get built in the SFBA, are pedestal apartments, with the ground floor (and possibly a basement) devoted to cars, and presenting an unrelenting wall (or at best, metal grilles) to the street. There is no retail space, no services combined with the residences–further eliminating the prospect of walkability and reinforcing the car dependence. It’s not transit-oriented development; it’s just an extension of the corporate office park.
@Tiercelet
The 13th was densifying because it had large areas of old light industrial areas, ie. there was a lot of brownfield redevelopment. The point remains: it never reached the density of the 15th or half of all twenty arrondissements! Who is in denial here?
As to equivalent densities elsewhere, I always mean in the context of a whole city or at least a fairly extensive area (like Paris’s 87km2). The UWS and UES are at about 42-43,000/m2 but of course they are only ≈5km2. Paris-11 is also 42,000/km2 on 3.7km2. In fact I am not in denial about any of this. It is usually me who gives Barcelona-Eixample as a case of high density (over smaller area) but point out its various problems–this is part of the Goldilocks argument. In the current context, ie. TOD surrounded by suburbia, we are talking about smaller zones but … I discuss this further below.
Again, you’re just making the irrelevant argument (that I do not contest) that hi-rise may achieve hi-densities. Doh! Yet again you miss the point: you don’t need hi-rise to achieve those densities (and densities of 40k/km2 aren’t desirable either). A secondary point is that those hi-rise, or spiky, zones don’t extend over a whole city to maintain that density (except Hong Kong and Manhattan though of course the latter is not a “whole city”).
You’re also missing the point about Cupertino Park. I didn’t claim it as Haussmannian (indeed it is too low!) but to show that those Bay Area suburbanites already have a format amidst them that achieves adequate residential density. So clearly it is not just me that has “an aesthetic preference” for low-rise versus hi-rise or peaky urban form.
I actually said half that myself. The bottom floor is a half-basement which is for car parking and is why it is grilled. It is the English means of pushing the first floor of apartments up half a floor thus giving some privacy and security compared to being level with the ground. I said that they could have added an extra floor to increase density but no doubt this wasn’t done because it is in the middle of very low-rise suburbia (including low-rise office parks) and it was probably at the allowable limit. However if they build a TOD zone around BART stations then almost certainly they would add an extra floor (maybe another with setbacks but not higher or the internal courtyard effect would be lost) . The format is in fact fairly flexible–Cupertino Park does in fact apparently have a restaurant and some retail (on one side), and clearly could have more (at street level and facing the street) if it was in a TOD. Once you achieve this kind of density over a certain threshold area the rest (retail, restaurants, services etc) become autocatalytic, not when they are just an island in suburbia like CP.
My point was that it is a localised version of the European perimeter-block format. And acceptable. But doubtless, because of so many people like you, such a TOD would end up being peaky and quickly ugly and unacceptable just like most such American laissez-faire development guided not by any aesthetic or even practical urban principles, but by developers and corrupt local politicians (are there any other kind?).
Apparently you think you can convince the locals to build density a la Upper East Side, and you wonder why there is such strident NIMBYism? Instead, I say, show them how a Paris (or if you like Barcelona, or Vienna or Copenhagen) style development can achieve more than adequate density –and it won’t impact their own nearby suburban zone or their property values. To have even a glimmer of hope of achieving this you must have totally tight mandatory building laws, essentially Haussmannian; ie. where no exceptions to build either lower or higher. Otherwise the developers will destroy it, as usual. Indeed the TOD zone would have to allow endless objection to any deviation by NIMBYs (in the whole suburb), but at the same time zero ability to object if a development satisfies the code. Any exception, including by government, eg. for a campanile for city hall or a church etc, would have to be approved by referendum.
BART down Geary with a branch up to Marin was part of the original plans but never built. There should definitely have been a MUNI Light rail line down Geary and up Van Ness, preferably in a tunnel but it isn’t going to happen anytime soon. I agree with Tiercelet that spending money on a new system in Santa Clara because of NIMBYs.
SF’s Central Subway was original planned for 4-car 350ft / 100meter stations, but that was cut to 2-car trains with little to no option for expansion. This design definitely feeds into Alon’s concert that we built a small metro, but it was built too small.
One hopes that they can expand it down Lombard to south end of Golden Gate Bridge. That area gets so many visitors that the park is grid-locked, parking lots are closed, and just lacks any kind of connectivity. Tourists will utilize the route on weekends, while buses from Marin County can drop off passengers there on weekdays for a quick ride downtown.
[Arrgh the WYSIWYG comment thing is so utterly amateur and awful. Who writes this crap?]
Anyway, to the comment: so the solution to total, complete and utter unmitigated failure is … <i>more failure</i>? Correct? It works every time single for BART extensions (each of which blew out budgets and underperformed “projections” by double- and triple-digit percentages and for Santa Clara’s worse than worthless ilght rail joke, too.
As for forcing bus transfers to a useless line whose downtown stations are poorly located, endlessly slow access jokes, and barely serve the CBD, what are you smoking?
The solution to abject failure is prosecution, black-listing, termination, and unelection. It isn’t herp derp you said trust us this time for sure herp derp here’s more money go nuts herp derp.
@Richard
You are correct that a Central Subway extension would not be good for bus transfers from Marin, however, you are incorrect that an extension would be “more failure”. All the way back in 2011, before the tech boom that ended with Coivd, bus ridership on the various lines making up the C. Subway and its extension (basically bus routes serving Lombard, Columbus, Stockton, and 4th) saw ridership per mile of 4,500-5,000/km. Compare to BART’s single day highest ridership of 568k on 167km of track (back in 2012) for ~3,500 riders per km. Depending on how much you think ridership would grow with metro vs. bus, you could be looking at 9,000 riders/km. The failure wouldn’t be the extension, the failure is that the C. Subway line wasn’t fully built out to the Presidio from the start, and that is wasn’t finished a decade ago.
If you are going to yell about BART extensions to low density suburbs you can’t complain when someone suggests a rail project right in the heart of a high-density walkable part of the city with already high transit ridership.
Onux, you really have no idea of how wretchedly bad it is.
I do. I live in San Francisco.
Just because the 8/15/30/40/45 etc bus lines on a map seem to you as if they have something to do with a infrequent, unreliable, painfully slow in transit, with mind-bogglingly bad and slow station-street access and just a couple shit stations located where they’re pretty much useless, and just because you, as I, can tally up the ridership on those bus lines, doesn’t mean that the actual daily riders of the 8/15/30/40/45 bus lines want anything at all to do with the fucking useless joke of subway line that parallels but is worse than super-congested, zero-transit-priority (one guess why!) surface bus transit.
Sure, right, the solution to an outright and unambiguous fraud of a deep slow subterranean stupid toonerville trolley line failing to even pretend to begin to serve the densest section of transit traffic market is … [checks notes] … keep extending, extend to places with lower bus ridership than the core, stay the course, we can do this, just keep digging holes.
It’s a fucking disaster, it has no ridership, everybody who wasn’t on the take knew and predicted it would be a fucking low-ridership disaster, everything was bleeding obvious, yet somehow the lessons learned is… MORE! MORE OF THE SAME!
I just can’t even.
Anyway, small beer. Good luck with that industrial civilization thing.
[And wow again holy fuck is this WYSIWYG wordpress comment thing worse than worthless. Shit Javascript – perfect pairing with the End Times!]
Cid just mocked me the other day for writing blog posts in the WYSIWYG WordPress editor and not in Emacs as God intended.
Nothing against WYSIWG! But this is just ampersand less-than semi-colon bad ampersand greater-than semi-colon bad. Or, as one (ahem) might have said it circa 1990, pre-LaTeX, {it bad}
PS It does so happen that I know {it quite a bit } about Emacs. Sadly enough. But you probably knew that.
Hey, I never know whether someone is on the Emacs or vi side of things.
I was the first non-Stallman person to do work on “free software”, before “GNU” existed, most particularly including post-PDP10/20 Emacs written in C. And then on RIP Lucid Emacs/XEmacs, because, oof, RMS. So, yeah. Highly profitable. Great success, as with all else I’ve endeavored.
The Central Subway extended, together with the line on 3rd, would connect, in order, UCSF Medical/Benioff Children’s, Chase Center Arena, Caltrain 4th & King, Moscone Convention Center, Union Sq Shopping District, Chinatown, North Beach and Fisherman’s Wharf (plus a few other stops). Hospitals, train stations and shopping districts are always large ridership generators. Arenas and convention centers generate large peak loads (and arenas are busy year round, unlike stadiums). Chinatown, Downtown, and N. Beach are three of the five densest neighborhoods in SF. Moscone, Downtown(Cable Cars), Chinatown and Fisherman’s Wharf are 4 of the 5 biggest tourist draws in SF (if your end of line Presidio station is close enough to the GGB you get the #1 draw too).
There are stations literally at the Caltrain terminus, at the Convention Center, and right onto Geary St at Union Sq. You have a strange definition of “pretty much useless” station locations.
All of these things add up high ridership potential, and you saying ‘mind boggling’ or ‘useless’ or ‘joke’ over and over won’t change those facts.
“actual daily riders of the 8/15/30/40/45 bus lines want anything at all to do with the fucking useless joke of subway line”
You can take your “everyone would rather ride the bus than a subway” argument and try to sell it somewhere people might listen, like a “World’s Finest Transportation Professionals” conference. The ~5km of subway on Market – with the same terrible light rail vehicles, no in-fare connection to BART, awful delays from interlining – got more than double the highest ridership 5km of bus you could find in SF, even stretches with multiple limited and express overlays. Muni’s schedule averages 34kph from W. Portal to Embarcadero and C. Subway does 20kph from Brannen with part of that at grade, two speeds no bus on SF streets has ever done even in the stretches with red carpet lanes.
If there is a reason people are still riding the 30/40 etc. its because they are getting on somewhere like Fillmore or N. Point where the subway isn’t because it hasn’t been extended there. The Lex Ave Subway wouldn’t get great ridership if it came to an abrupt stop at 50th St.
I’m not sure if that was ever considered, but it would seem to be cheaper for Central Subway to pass ABOVE the Muni tracks under Market Street. The mezzanine would have to be shrunk/altered, but the resulting cut-and-cover construction would have been cheaper.
Of course, the main goal is to make the train more accessible. There might not be a mezzanine above that station, but who cares if a short flight of stairs is all it takes to get on your train.
@Martin
“it would seem to be cheaper for Central Subway to pass ABOVE the Muni tracks under Market Street.”
There isn’t enough room. People are about 2m tall and can make do with a 3m or less ceiling, trains are about 4m high. In SF you also have the fact that Muni uses street running light rail vehicles that need space for overhead wire. There isn’t enough distance between the Market subway vault and the invert of the roadbed above to fit track ROW, even tough there is enough room for a mezzanine where people can walk.
The depth issue for the Central Subway is overblown. Depth is only ~25m going under the Market St subway. This isn’t particularly deep, particularly for the condition of passing under another line, I believe that London Underground average station depth is around 20m. Moscone Center station is only 18m deep. Chinatown station is deeper at 36m because of geology, but again this isn’t extreme. NYC has four stations between 30-50m, the deepest Underground station is also 50+m, and we haven’t even talked about crazy soviet examples in the 80-100m range.
Alon has correctly stated that in an ideal history Muni and BART would have been built side by side at the same level, allowing transfers across two island platforms (including a ‘wrong-way’ transfer at Civic Center so people coming from north Mission St could transfer cross platform to head west towards the Sunset, and vice versa). This would have placed the C. Subway (and future crossing lines, *cough* ahem, Geary) at the level of the current BART tracks. This would have been ideal, but four vs three flights of escalator travel isn’t a big deal.
In general I think that depth and station access issues are overblown. Once when being a tourist in Paris I saw a sign for the RER when trying to get from Ile de Cite to the Eiffel Tower. Steps, corridor, turn, corridor, steps, turn, etc., etc. and eventually I reached a train platform. It wasn’t until I saw a map later that I realized I had walked down and UNDER the Seine then back up again to reach RER C. If you are taking a transit trip you are expecting some walking along the way.
I think a better way to count km is, how many run mass transit service. So a train every 10 minutes off peak (or maybe every 15minutes or 6 trains per hour if it has express service). Probably should be grade-seperated.
”I think a better way to count km is, how many run mass transit service.”
I would generally agree with this, with the caveat that it might be worthwhile to set requirements on location (ideally all lines will through run the center, or at least a secondary center) and speed (a network of streetcars no faster than buses simply won’t attract or carry as many people as heavy rail, even though it might boast impressive track-km statistics).
I have long concluded that all US cities need a network of slow transit for the “last mile”, and plenty of express routes that are easy to transfer to so I can get across a city. Our cities are too sprawled to rely on only slow local transit, and only express transit isn’t close enough to anything to get riders (though we do have dense cores where express can get good riders)
@Henry
Yes, it’s because “they” don’t believe in taxation. By “they” I mean the 1% or maybe 5% because they are the only ones who earn enough to pay any tax if it was increased. However, that is the very reason HK-MTR is a model worth looking at by others who don’t believe in tax like the US. The land value-capture that the government (as proxies for the big companies) allows MTR to use to help self-finance the system is a kind of compromise between levying more taxes versus giving it a slice of those property companies’ usual domain, ie. outrageous property profits from development rights. It works quite well, in that those property developments above and near new stations have allowed MTR to contribute 50% of the capital cost of building new extensions. I’d suppose Alon would say that it would be better if HK managed to build the Metro at lower cost; sure but I’d say that is also part of whatever compromise government made to get a modern Metro built. Those big companies pretty much run HK government. It helps that the road option is simply not viable.
Given all that it is a remarkably successful system. It is one of the few big-city systems I have no complaints about cf London, NYC, Tokyo, Shanghai, Beijing etc. Efficient, inexpensive, easy to use and comfortable (if it is crowded more than Paris I can’t say it impacts one negatively, not like Tokyo or mainland China).
The problem is that construction costs in Hong Kong are so high, thanks to its early adoption of privatization of state planning to consultants (“FILTH” – Failed in London, Try Hong Kong), that the costs to the state there per km, net of real estate profits, are high as well, higher than in France or Germany, let alone Spain or Italy or Turkey. The low taxation in Hong Kong and Singapore comes from having no social safety net (except housing, at low quality in HK) and, through very high immigration levels in Singapore, a low dependency ratio. High immigration levels are available as a strategy for Europe and the US, but maybe not at Singaporean levels – Singapore is 44% foreign-born, and if the US and Western Europe get there from the current level of ~14%, it means having to find 230 million additional people worldwide who are willing and able to go be migrant workers. It’s probably feasible but barely, and the extreme right rioters here are already pissed off enough at 16% and the minister of justice just said warned that it would be difficult to ban AfD.
I’m not sure it is the same thing at all. And I think you still have bought into the quite false claim that MTR itself is a private entity only working to maximise profit for shareholders (and I’m not going into that yet again). Not to mention Flyvbjerg’s extremist negativity on any megaproject though I note that in his first book there is no mention of HK’s many projects. (He has criticised the HSR project but that is really a PRC project forced on the MTR.)
I mean, isn’t it likely that HK always had a very lean public service due to its very low tax income? Having said that, I recall from general histories of HK it had a strong public works department, and like all HK government (until perhaps recently?) was renowned for competence and lack of corruption. Until the 70s there was not much government funded infrastructure, then from the transformative and proactive governor MacLehose (1971-1982), there was an explosion of it in highways, cross-harbour tunnels, huge land reclamation, massive docks (though maybe that was all private?), schools, universities and the first two Metro lines and of course massive public housing.
Was there a big difference in the cost of those first two Metro lines completed by 1980, and later ones? And was there any change in the way the projects were managed?
In any case, surely cost is only one issue for such projects, a not unimportant one but getting such projects completed and that meet their operational goals is more important. The difference with the UK and US is stark, in that with all their psychodramas and no matter how much money spent/wasted they seem incapable of designing and completing such projects.
It appears to me that the main cost in HK is high inequality. (Arguably such projects, especially inexpensive transit, are an anti-inequality measure.) However that is a much broader socio-political problem and without big changes, any money notionally saved on infrastructure would not be magically redirected towards solving these issues (eg. affordable housing). And the UK and US share that anyway (the US has only Singapore & HK below it on the developed world’s inequality index).
I am reluctant to put HK in with the UK because they seem a million miles apart. (There’s an argument that HK was really a Scottish colony not a British or English one.) During my 20+ years in Europe when I often flew to the other side of the world often stopping in HK for a day or two or sometimes just 6 hours, I witnessed the continuing transformation of the territory. They continued building their Metro into a very impressive system including the 37km line to the new airport, and of course HKI was a marvel which was the world’s biggest infrastructure project at the time, and actually seemed a bargain (two major rail lines, freeways, one of the world’s biggest suspension bridges, a Newtown for 250,000 and etc for about US$20bn. The main drama was from many naysayers who said it couldn’t be done in the timescale (attempt to complete before the 1997 handover, missed by only months but by then China had accepted the project).
Recently they completed HKI’s third runway after less than 10 years construction (again involving massive land reclamation). True, it did cost about the same as the original airport + transport-links project. But complete and functioning as designed.
“He has criticised the HSR project but that is really a PRC project forced on the MTR.”
Actually not. The Hong Kong HSR was proposed by mainlanders and was built to the standards of China’s national rail. Yet the project itself was executed by the MTR, while most of the consultants as well as contractors involved weren’t even from mainland China. Hence it still counts as an MTR project, and the MTR is the one to be blamed for the poor management and skyrocketing costs.
The highest priority part of HS2 is being built.
Yes it’s annoying that it ends in the midlands, but the most congested section in London to Milton Keynes which is being fully bypassed.
The thing is that a bug part profit from development property is actually something Hong Kong government usually charge. Like, for the land of Tuen Mun Swimming Pool which was a swimming pool and will later become a MTR station with residential on top, the Hong Kong government usually charge property developers 3 billion USD for the right to turn that 6 hectare of land into residential zoning. But Hong Kong government decided to “subsidize” the Tuen Mun South extension by waiving this fee for MTR. That’s the anount would have needed to be paid by other property developers if they want to build properties on them. I don’t think it is common in rest of the world for there to be billions of expense charged for new housing construction that can be waived to enable cheaper TOD construction?
@Phake Nick
[Perhaps you mean HK$3bn for 6Ha …? Even that sounds too high, even for HK.]
It is a trade off between one part of government and another (ie. MTR) to facilitate critical infrastructure. Then the development opportunity enables MTR to make money in lieu of the usual property developers, and perhaps more importantly to make recurrent income from the project, ie. future income to offset both MTR operating costs and future capital costs for new build.
It is relatively straight forward land-value-capture as recommended by any number of experts on how to fund public infrastructure, ie. to recover some of the increase in property values that is brought about by the public transit. HK is in the singular situation because the state owns all the land, and releases/sells it to private interests by leasehold only. The UK could only hope for such a situation for HS2 or CrossRail but were obliged to pay market rates to private landowners. To help fund CrossRail/Elizabeth line a special “business-rate supplement of 2p on non-domestic properties with a rateable value of £55,000 or more created £4 billion for the project, nearly as much as the government is providing.” (The Economist, Nov 2013)
Where they can they are adopting the HK approach:
It’s not a lot of land but has peculiarly high value due to the proximity to transit.
In the Ile-de-France there is a versement transport which is a hypothecated urban regional payroll tax levied on the total gross salaries of all employees of companies of more than 11 employees. The rationale is that these businesses benefit from the transit system for their workers and their customers. It can provide up to 40% of RATP’s operating costs.
it is HK$24.2 billion that are waived for the Tuen Mun Area 16 land as subsidy to MTR for financing the extension. The area is supposed to enable development of 4.28 million square feet property and thus the waived land premium cost translate to about HK$5,600 per square feet of constructed area after the property is finished.
I don’t know what any of that means. It sounds one of those theoretical calculations rather than what a developer would pay upfront for a 6Ha site. If it includes other government charges (income), eg. eventual sales tax on apartment sales then the government still receives that. You need to give credible references.
The original point remains, that regardless of the actual value to the government from selling the land to private developers, they have transferred that value to another government entity (MTR) to facilitate critical public infrastructure. A perfectly valid and good strategy.
It should have been used by the UK government to get HS1 built. They had given all kinds of development concessions to the PFI (PPP) they entrusted to get the thing built (at no public expense, a continuing Thatcher fantasy re public transport) but it still collapsed and required government rescue. Of course in the end it was still the private developers who made truckloads of money from all the development opportunities opened up by building HS1 thru Stratford and terminating at St Pancras-Kings Cross. In the HK model it would be HS1 that made most of that money plus a perpetual rent income to subsidise operations. Paradoxically it may well have met Thatcher’s dictum of “no public expense”!
Also compare what happened to Hudson Yards. MTA sold its air rights of the 11Ha site to a master developer for $400m. But over the subsequent 2 years the developer received $434m in tax exemptions and as time rolled on:
That doesn’t include the $2.4bn to construct the subway extension and station. The master developer will retain a lot of apartments that provide rental income. The argument is that the city benefits from such a development however there are alternative models, like HK’s, and it would probably not have resulted in a gated community for the rich.
https://www.info.gov.hk/gia/general/202309/05/P2023090500575.htm this is the government announcement regarding the financing arrangement for Tuen Mun South extension and others. See also https://www.oecd-ilibrary.org/sites/5b45f2f0-en/index.html?itemId=/content/component/5b45f2f0-en for the overall arrangement of such land premium charges.
It is the amount developers would actually have to pay upfront. For example, Google search indicate a property called Novo Land in the same district but less central location see its developer acquired 0.02 million square feet of land for additional development of 0.1 million constructable area, and the property developer paid government 0.5 billion HKD for converting those farmland into residential in year 2020 after the developer acquired the farmlands themselves, aka roughly HK$5,300 per square feet.
Anywa, what I mean is that this give MTR a significant cost advantage and profit margin that wouldn’t be possible if not for such cost that would have been charged against other property developers. And thus I am not certain about MTR’s TOD model can be expected to fund expensive rail project in countries that government didn’t initially charge as much for allowing residential construction, and thus permitting MTR to earn a much higher profit margin than normal property developers
Well, it is a bit of a fantasy figure and probably inflated to suit various political purposes. I am also unsure what it really means:
It is being used as how much the government is “funding” the MTR … and presumably a guide to how much they expect the MTRCL to recover from those developments to offset total government cost (the government still has to provide direct funding to MTR to get stuff built). As I said I have no problem with any of that.
And yet again, the amount is not at all the critical thing, which is that it went to MTRCL and not to the usual suspects of rapacious private developers who own and run Hong Kong. But who are smart enough to know that these transit projects are required for their own prosperity; one cannot build massive housing projects so far from anywhere and expect them to work without transit. Singapore does the same if a different funding model (but not fundamentally different in that they build most housing and the proceeds of its sale helps fund the public infrastructure).
As to your second link, that just details what I mentioned briefly, ie. HK leasehold system. Another example that works well is the Australian Capital Territory (like DC, the federal zone that contains Canberra city). The private neo-feudal version that still persists in the UK is the reason leasehold land has such a toxic reputation.
it didn’t seems like the government is providing any extra direct subsidy beyond that land, the cost of construction of Tuen Mun South extension is only 15.8 billion HKD being a short elevated extension
That “subsidy” isn’t cash today but likely development profits that only arrive years into the future.
Michael, as Alon has mentioned before the French actually paid a premium on land to private owners and that was good as it kept the overall project costs down by reducing opposition.
Certainly for HS2 farmland really really isn’t that expensive to buy – but certainly our leaders could have shown a lot more imagination to keep costs more under control.
Any premium will be quite limited, especially once a project is designated a project of national strategic significance (“declaration d’utilité publique”). And it hardly compares to:
All of this sounds like yet another reason not to make everyone who lives along the route hate the project by having no stations.
Put in some reasonable intermediate stations in the countryside, spend a sensible amount of money on local charities etc, and then I am sure you wouldn’t have any issues finding a farmer who was prepared to let you run the electrical cables across their land without it costing too much.
UK land ownership being what it is, I suspect the largest obstructions to land acquisition cost control are the non-resident titled gentry who own the land rather than the people who live and work on it. Local payoffs don’t work on absentee landlords.
The UK can’t bring its rail infrastructure out of the 1800s because it can’t bring its social structures out of the 1600s.
@Coridon Henshaw
I’m filing that for future deployment …
I wonder if it could be snappier, like “Win95 = MacOS84” ?
@Coridon Henshaw, the local payoffs for HS2 are basically non-existent. It’s the national payoffs that are bigger. And rich people who live in London will definitely benefit from HS2.
@Michael, there’s a class system in Australia, New Zealand, France, Germany, the United States etc. As with any group of people with informal power ours is in many ways the least powerful because everyone is aware it exists.
Matthew you have to remember that Michael get’s his British culture from the Guardian which is like getting your US new from the WSJ.
Its easier to believe the UK is a “Feudal” relic that is responsible for all the evil in Aus- I mean the world, than the actual story here that British Socialism failed and created this mess with the 1947 system in Transport and Land use.
The best agricultural land in 10,000 quid per acre. That’s not where the costs for HS2, its in the ordinary homeowners of rural and peri-urban England leveraging the worst planning system in the world to ruin everything. And they have been the ones to make sure that England doesn’t have a competent procurement state in general, supporting both Labour’s homeowning activist class and the Tory Essex Man anti-professional ideological systems. And if you don’t believe me go read the planning application websites. Local elderly are a more vicious group than an absentee landlord.
The Squirearchy is dead because agricultural wealth earns so little. The fantasy is needed to avoid Michael realising the problem with Britain is that Xenophobic priveleged greedy boomers like him are in charge.
1600’s social order FFS that’s the stupidest thing one could say. 1600 the average British person was a tenant farmer, 90% of women were illiterate, it was an Anglican-Christian monoculture and the politicians were all biologically related to each other. Only 25% of adult males had the franchise, 0.01% went to university. Oh and there were no substantial immigrant groups either (we have the Parish records).
UK gini is high by W.European standards, but low by global standards particularly once you adjust for wealth. 65% of household own their own home. 100% of the adult citizen population has the franchise etc and we have a large population of immigrants. Which does rather change the social order.
Also the 1800’s railways thing is stupid to. Its that we’re stuck in the 1950’s which makes sense since the structures that worked in pre-1947 rail (TfL and Southern) were crippled because they were making everybody else look bad.
UK (or England) has changed more in the last 100 years than the US.
If the UK is a negative case*, its because of people like Michael and Henshaw, xenophobic, reactionary-conservative boomers whose civic identity is based on irresponsibility and hate hidden with bullshit leftist rhetoric. That is the 1947 UK land use/transport system which was designed to destroy metropolitan England.
*UK is a colossal disappointment, I want rid of it. But “not as good as France/Sweden” does not a failed state make.
Certainly in the UK the transit activists have left a lot to be desired. If they had campaigned better I think we would have better transit infrastructure.
Ridership per km seems a solid output measure, but honestly, I think there might be too many variables to really discern being underbuilt vs. efficiency, at least as so far discussed here.
Not to mention, how much ridership per km is “too much”? I know that there were some standards used for justifying Tokyo’s 80’s/90’s investments in both the subway and commuter rail system, but insofar as these are standards that reflect “culture”–i.e., what people are used to–it seems like they must shift over time, at least up to physiological constrains, or down to a line no longer really being feasible/useful. No one’s ever going to find a level of crowding that restricts breathing or much in the way of movement tolerable for long (and ideally, people of all social class should find riding a metro tolerable, if not outright preferable, and there’s no denying that social class will affect how much crowding is too much), and no one’s going to (or should) keep running a metro line that regularly has few enough riders on it.
Even then, from a system-wide perspective, it seems like there’s not likely a plausible “too much ridership” issue, at least up until the point that overcrowding deters ridership growth. That’s a “no one goes there; it’s too crowded” situation. I have to imagine that the overcrowding on Tokyo’s lines in the 80’s was getting close to that situation, but it was during that time that developments in Shinjuku and Shibuya started taking the heat out of Chiyoda; a good thing, since there are 5 lines in the ~1.5km from the palace to the shoreline. I think the question that this leads to, to flip the lead, how many metro lines in an area are enough to suggest that developing new centers of activity is more productive? At what point do we say the solution is not building more metro service in an already-developed area, but rather to develop more in an area which is underdeveloped relative to its already-existent metro service?
I’m still struck by the past emphasis on monocentricity as the ideal city form for transit, and the implications of that for the sort of megacities that have sprung up since Tokyo laid that template, and which are likely to continue to form in South Asia and across West-East Africa.
From what I can tell nobody figured out how to design a non-car transportation system for a city or metropolitan with a lot of job sprawl. You can have a really big center that can seem polycentric like Tokyo or Paris but you can have a lot of job sprawl like the San Francisco Bay Area or the metropolitan areas of Florida has because people need to make too many transfers and if thats the case, they would prefer to drive because it is at least direct from your house to your job. You don’t need to get on BART to go to Montgomery than catch a bus or Muni light rail to get to Mission Bay or something more complicated and time consuming.
I’m guessing that using transit for leisure or fun is similar. You need to have stations that have a high connection of leisure options within general walking distance or people are just going to drive. The exception seems to be for big sporting events. People are willing to do multiple transfers to avoid the hassle in those situations.
For leisure I am pretty sure for trips into London or Paris that the public transport share is pretty damn respectable.
I would have thought if the last mile services ran every 15 minutes or so and connected together nicely that you could also see a decent public transport for more rural destinations too.
I suspect that Cornwall has better public transport share with its half hourly service on at least some of its branch lines than other similar places – but not quite sure how to get the data to prove it.
Same with Tokyo and Osaka, Kobe, and other big cities or metro areas in Japan. Those are noodle networks though rather than hub and spoke. Going from New Jersey or Long Island to the hipster scene in Williamsburg is a lot more annoying by transit than car.
There is no place to park in Williamsburg. And execrable traffic for miles and miles around it.
“Going from New Jersey or Long Island to the hipster scene in Williamsburg is a lot more annoying by transit than car.”
This *should* be pretty simple to solve. Through-run LIRR and NJT, infill stop at Queens Plaza, and take the G to Williamsburg.
The M Train goes to Williamsburg. It stops a whole block away from Penn Station. Uses the same concourses as PATH. The G train doesn’t go to Queens Plaza and is unlikely to go Queens Plaza. To send trains from New Jersey to Long Island or vice versa both fleets would need to be replaced.
Designing a transports system for a city with job sprawl isn’t hard. What is hard is paying for it.
Because most trips are not to the center you need to figure out how to get people around the city. You need not only those in/out lines, but a lot more circumferential lines, since that is what people will be using to get to their job that is along the next in/out line. Most trips will involve 3 transfers, so you need to focus on stations designs that assumes transfers.
The typical trip to work should look like: ride local line for 7 minutes to the local suburban station. You then transfer (3 minutes), to an express line to the next suburban station (5 minutes), then a cross platform transfer (2 minutes) to a different express train (5 minutes), then a 3 minute transfer before a 5 minute local trip to work. Total time is a reasonable 30 minutes, and the as a crows flies distance is 15 miles. Getting back home is similar, but you need to take different lines to get that cross platform transfer.
As I started with, I can plan the above type system on paper easily. However it is horribly expensive and so isn’t worth trying to go in more depth. (real cities also tend to have geographic features that make getting all the cross platform transfers impossible for many) trips) Which is why this is likely to remain a comment on a blog forever. Though if you build it across an entire metropolitan area you should get >50% mode share, and the total cost to the residents is much cheaper in the long run than the current everyone has a car and many large highways trying to handle the traffic, so I guess if you can figure out how to build it – go for it.
The plan Alon and most transit advocates prefer is in-out lines to downtown, and everyone gets what they need along their in-out line to downtown. Circumferential lines are still important, but only after all the in-out lines are built, and then they are just relief for the few people not going in/out and so you need much less of them. Most stations do not have a circumferential to transfer to.
In fact the Bay Area is somewhat unique, essentially being a linear city built around the bay. Hence the original vision of BART to circumnavigate it and link all those zones with one line. (The main sprawl outside this zone, ie. Richmond etc was the first to be served by BART.)
Tokyo isn’t polycentric, its just got a huge downtown that stretches across the Yamanote plus the western bits of the Sumida river. Tokyo is a story of letting your CBD expand with highly developed shoulder-station nodes like Shinjuku and Shinagawa.
Failure to do this properly is a big problem for Japan’s other major urban areas. Particularly Osaka with lack of through-running for Nankai and Kintetsu to get to Umeda, the wasted dud lines like the Sennichimae line. The upcoming Naniwasuji line is 50 years late. The result has meant Osaka’s real estate prices are too high for its relative productivity compared to Tokyo and Nagoya. Nagoya too has problems given how weak its centre because of lack through-running services to Sakai (it needs NE-SW narrow guage lines Naniwasuji-Crossrail style connector).
The lesson for South Asian cities is to properly intergrate their legacy lines as the spine on their office/high density CBDs. Africa outside SA doesn’t have that opportunity.
Bad spamfilter :(.
Its happens. I just thought I’d screwed up posting it and gave up.
Any opinions on what I said?
@Borners
At 23.59.59 on 18 January 2024 the Spammer went critical and did what any intelligence, artificial or organic, would do:-)
The subway systems of Seoul and Sao Paulo are interesting contrasts. Both mega cities started to build their metros at around the same time. Both countries were military dictatorships when metro building started. Seoul kept building steadily and has one of the most extensive metros in the world. Sao Paulo did not and has a system that really doesn’t serve a big chunk of the area or the population. This is why Seoul has an annual ridership that is nearly two billion while Sao Paulo is around 1.1 billion according to Wikipedia despite Sao Paulo having a more populous central city than Seoul.
<i>metros should be in the toolkit of metro areas of about a million people</i>
There aren’t enough people in most North American ones. Living in one now, Albany’s Combined Statistical Area is just over a million, people are not going to wait for a bus to take them to the train to get on another bus. There is no traffic to speak of and there is plenty of free parking. There are work destinations along existing tracks that might support rush hour service with park-n-rides out in the suburbs. …. I see the webcams of the horrendous traffic on the afternoon news. It’s not horrendous.
I lived in and around Newark, New Jersey most of my life. People, who own cars, use the buses when it’s difficult and or expensive to park. That overlaps with using the train to get to Manhattan. Most places in North American don’t have Manhattan a few miles away. Or even Downtown Newark.
Calgary and Edmonton would be good examples for cities of this size. (not perfect, but good) Both metros (well LRT). Stockholm was also smaller than this in the 1950s when they started building a metro.
The important thing to note here is that this doesn’t happen overnight. Build something small but good this year then next year build something more. In 20 years you will have a good network and see changes in the city form. Keep building – by constantly building you can learn what works and bring costs down. (I don’t know how to handle the politics of this though)
You can kind of live without a car if you live in Hudson County or Newark, NJ. PATH offers just enough stops to make it doable and there is mixed use developments. However, this really limits you in many ways and if you want to go to a METS game chances are it is easier to drive from NJ to Queens than take NJ Transit or PATH into the City and get on the 7 to Flushing. Probably faster to even in really bad traffic.
The LIRR goes to there too. They run extra trains for events. Extra trains will run for other events at other venues. They let car owners use the buses and trains. They do in places where traffic is bad or parking is difficult/expensive or both. Most place with smaller populations don’t have those problems.
Somewhat related, but why isn’t ridership density (passenger-kilometers per route kilometers per time) used more? It seems like it would make numbers like ridership vs length for systems that serve regions and trips of different natures easier to compare, e.g., while ridership per km of Paris Metro is very high, ridership density of the central segments of Paris RER is certainly a lot higher.
Ridership density is used a lot within Japan, but manages to be computed when Japanese organizations research overseas railways for benchmarks and comparisons, so the information is out there somewhere.
The problem is, the more you built and the more areas you serve your passenger/km per route-km goes down.
For individual travellers exits are measured on the RER but they aren’t and cannot be measured for the Metro (except those going Metro->RER) so they’d have to guesstimate it.
the information of ridership density is used a lot in Japan especially when they want to justify closing rail lines or reducing their service, or see the risk of that. It isn’t exactly relevant when one instead try to expand rail service.
Why doesn’t it work both ways?
Knowing a service will be clearly profitable is a good reason to open it too right?
You can only have the ridership density data after service already exists.
Also, high ridership doesn’t mean profitable, if the operation cost is too high ir fare is not high enough, then even high ridershio service can get cut.
And also, a service being profitable doesn’t meab it is worth opening either, especially if it involve overcoming physical barriers be it track capacity limit or gap of existing track that need construction of new tracks, the construction/acquistion cost could be too expensive to pay for, even if they can be expected to recoup from fare revenue in the long run
Japan has relatively good data on this because the expectations that lines be profitable. That’s reinforced with the other regularly used statistic which is daily per-station ridership (other systems especially the Anglos love pretending they don’t suck by usual annual ridership).
That’s also a product of the side businesses of Japanese rail operators, they need estimates to justify/calibrate their investments in a station/rail corridor beyond the railine itself.
The UK would definitely benefit from including passenger density as a measure, mostly because it would reveal that non-Southeast services are subsidy hogs to the point it cancels out the SE’s greater share of the capital budget. The UK government likes having the system illegible to itself and the public. Indeed the lack of good understanding of where non-Tfl rail services get such bad cost performance would be really useful. Whatever structure of railway industry moves to after the franchise model needs to solve that cost problem to get us to SBB/OBB financial probity.
Part of the problem is that TfL has no idea why it makes a profit since its just something they’ve always done.
I’m not sure if Hong Kong fits into this discussion neatly. It just doesn’t strike me as a system that took shortcuts, whether you are talking about rolling stock design, station design, or its signalling systems.
The MTR’s trains are some of the longest in the world, and if I recall correctly they were equipped with full width gangways from the outset, a pioneering outlier among the world’s metros when it opened in 1979.
It’s electrified at 1.5 kV with overhead collection, more akin to a main line or regional railway.
The stations are massive, occupying a large footprint with sprawling mezzanines and multiple entrances spread out over the neighborhoods in which they are located. All the more remarkable considering Hong Kong’s density and terrain. The junction stations are bi-level with island platforms for ease of cross-platform same direction transfers.
Hong Kong’s geography and terrain may have been a limiting factor when work on the ‘Modified Initial System’ began in the 1970, but that has certainly not been the case in the current millennium with the expansion that has been been underway, which started even before the British handover to the Chinese.
@Subutay Musluoglu
I agree.
Beginning with such a bold vision for the first two lines setting the pattern. I haven’t seen any histories about this but it seems a lot of credit must go to MacLehose who presumably appointed good managers. It would be good to know how it was financed because the UK Colonial Office (or whatever) historically was adamantly against spending much on such ‘extravagances’. Somehow I doubt he was allowed to run up government debt to finance it, but I don’t know. And, unlike the LU, to charge very reasonable fares which was obviously necessary.
Here are some glimpses of how it changed in the golden years, from <i>A History of Hong Kong</i> by Frank Welsh, 1993:
<blockquote>Total government expenditure was increased by over 50 percent between 1970 and 1972, and has continued to increase steadily since then–more than doubling in the five years previous to 1992 …
… but the spending was no panic response to an unacceptable situation. The planning effort involved was huge, perhaps without parallel, involving as it did the reconstruction of a transport system, rehousing half the population, and providing acceptable levels of education and health care, all in sort order. New skills needed to be introduced into the civil service, and the existing methods employed there sharply improved.
Even though rebuilding in Hong Kong proceeded at an uncomfortably rapid pace, the complexity of the construction necessitated protracted planning studies. These were put in hadn’t during the Maclehose years …
… When the MTR was opened in 1980, connecting Central with Kowloon it was as a result of a compromise between the Hong Kong preference for private enterprise and the impossibility of running such a system without public funding. It is proving not inexpensive for the public purse, with equity injections of nearly $10,000 million from that source demanded in the last three years, but to anyone who has experienced the New York or London systems the Hong Kong metro is a vision of a different world–clean, quiet and reliable.
‘Throwing money about’ in other areas has produced dramatic effects in Hong Kong. To most of the world it may seem like a Utopia, and not only to the developing countries. Millions of Americans might wish to live under a system where no one is deprived of adequate medical treatment through lack of funds. … Specialist high-technology medicine is not as widely available as in the United States or the United Kingdom, but the generally satisfactory standards are reflected in the statistics for mortality and most morbidity, which are generally better than in either of those other two countries. And no statistics can adequately reflect the mental security given by freely–and quickly–available medical care. </blockquote>
Welsh reminds me of the considerable contribution of the (Royal) Hong Kong Jockey Club: “.. in 1991 the Jockey Club contributed some £500m [≈HK$10bn then?] … which amounts to about 15 percent of total government income from all sources ..” That doesn’t include its own spending, much of it directed to the poor of HK, but also the JC built the HK University of Science and Technology (HKUST) plus IIRC a large part of all university research grants in HK.
Concerning your point about short metros and small stations, is there some trade-off study dealing with a cost comparison between building a metro network with long trains (and corresponding longer stations to accommodate them) and an alternative where a network with shorter stations but wider trains is chosen? I know that using the floor area of a train is not exactly the best metric regarding transport capacity, but assuming, as a mere exercise, that a metro network with trains 90 meters long and three meters wide would have the same capacity of trains 60 meters long and 4.5 meters wide, wouldn’t it be cheaper to build a network with wider trains, given that the cost of stations seems to increase more with their length than with their width? Is there any analysis available addressing this? I understand that wider trains are not currently the industry standard, but it seems to me that construction costs are a bigger constraint on getting a metro network off the ground than vehicle acquisition costs.