Buses have many advantages over subways, especially over trains running at street level (since these can't go much faster than traffic). The main advantage of subways is ironically that they are inflexible - they can't be gutted in times of budget deficit.
And I would suspect this is how Toronto differs from the average American city. American city governments experience periodic budget squeezes that force politicians to gut every service which is gut-able and buses are always a logical target.
And this leads to both buses being abandoned due to unreliability and for the bus lines that do exist to, uh, worthless, run at too infrequent intervals to pick up significant ridership. And this lead to buses having a terrible reputation, reinforcing the cycle of poor ridership.
Many "street cars" have dedicated lanes, and some sometimes run in tunnels and bridges with no intersections with cars at all. So it's worth explaining what you mean up front.
If you check out YouTube you can find some freight trains that have to do that from time-to-time.
About a hundred years ago, when Toronto’s streetcar was implemented, they made their gauge slightly narrower than train gauge to prevent “trains” from running along streetcar tracks.
They pay for it to this day as it means new streetcars need to be designed for a gauge that nobody else uses.
Turning radii is also an issue. Possibly by design or not...
Pedantic correction: Toronto streetcars and subways are actually slightly wider than standard gauge. The rest of your point is unaffected.
The streetcars are not discussed much in the article, but they may be a large part of both Toronto's moderate transit success and its slowness to adopt real light rail transit.
Streetcars are seen as similar to buses but a bit faster, more comfortable and more reliable. Which suits them well for the busy downtown routes. But plans for suburban LRT routes have been dismissed as "glorified streetcars", even though they would be much faster due to less frequent stops and more signal priority. Since relatively slow downtown streetcars are the only surface rail people know about, they don't get excited about LRT.
Are you sure that was the reason they did that? I don't have any evidence, other than that I always assumed it was either cheaper or more practical to have narrower gauges in cities. It seems to be the case for most streetcar networks I've seen...
Buses have many advantages over subways, especially over trains running at street level (since these can't go much faster than traffic).
That's not necessarily so. Trams (streetcars) in cities, which have old tram networks usually segregate the lines from the parts of the street where privat traffic runs. Very often they also receive signalling priority and while blockage by cars can happen (especially on intersections) trams can run with a quite reliable time table.
I live in a city with one of the - arguably - best public transport systems at least outside of Japan. If anything runs in unreliable intervals (as you learn over announcemets while sitting in a tram) it's bus lines.
Now, if you just build some light rail on an existing road as often happens because public transport is chic nowadays then you run exactly into the problem you describe. But dedicated tram networks are very different and quite reliable.
I read this argument last time so the passage you quote actually take your argument into account. Sure, street level trains can go somewhat faster than cars but not a whole lot. Contrast that Elevated or underground lines, which can go much faster, have express trains and so-forth.
The OP makes a great argument that buses plus medium rail/subways/etc is much better combination to add to a city. I think that argument has to be made more loudly, for the other reason - US planners are bad, if there's a to force them to be better it should be taken. Maybe really smart people do light rail well but the US city doesn't deserve to be trusted to do that.
Houston, a city proud to be without zoning, and where being a "city planner" is barely a notch above a registered sex offender (I wish I were joking), has just voted for the third time to expand LRT transit, and expand Rapid Bus lines, and improve the quality of its bus stops, in a $3.5BB bond measure.
On the other side of things, Austin may get a third crack at approving an LRT system next year, but the last two times Austin voters disapproved of LRT. As a result they have a weird heavy commuter rail that runs on a strange schedule, goes almost nowhere, and does not serve the people well.
Buses are slower than a car by some amount. A well designed express bus system isn't much slower than however, for longer distances.
Sure an actual Metro-type system, like BART in the SF Bay area, Toronto (apparently) or the various marvelous French systems that exist, these are faster than a car. But you might not have experienced the worthless, pathetic hunks of junk that Americans build and call "light rail", these can manage to be slower even than buses and more expensive and less reliable. The San Jose light rail system is a prime specimen, a "transit system" that barely manages to be faster than walking.
Of course, when I lived in San Rosa, the bus system also had an expected arrival times identical to walking - I tested this often on Google maps in off-hours. Where I live now, the bus comes twice a day so it's a further world of impractical.
Anyway, in America, it's not just that not much money gets put into public transit. it's even just that cities were created around the car. It's that the state, civic fathers, the ruling/managing group, doesn't really think public transit is necessary so much of the money that's allocated can be passed to the rulers' friends since "no one" expects public transit to work.
Does the DC metro count as light rail? It's... not luxurious. Maybe new light rail is nice, but I've never ridden one. Regardless, I see the point the author is making. I'll probably look harder at the buses around here.
What exactly the difference between heavy rail and light rail is a matter of debate in some cases but I don’t think there’s any dispute that DC Metro or NY Subway is heavy rail.
I think heavy rail needs to meet FRA (Fed railway admin) standards. Including things like handling .22cal bullet standards against their glass and other crash tests.
DC metro is heavy rail. Light rail is generally single or double trainsets running in the street right-of-way or otherwise mostly at-grade intersections. Toronto's streetcars (but not the subway!), SF's Muni Metro, and Boston's Green Line (but not the other ones!) are well-known examples of light rail in North America.
Yeah. Light rail is cheaper and slower than a subway, and more expensive and faster than buses. A single light rail line may not serve as many people as an extensive bus network, but it may supplement an existing bus network better than a short subway line would for the same cost.
It depends a lot on the city, but in high-traffic areas, buses are extremely slow and unreliable because they have to fight through the traffic jams. Dedicated bus lanes get around this, but it’s very hard to recapture the lane from car traffic at that point since the car drivers will see it as being taken from them.
If you’re building a new rapid transit line with dedicated ROW, it then comes down to BRT vs LRT and the cost of BRT and LRT aren’t actually that different I believe. The buses are cheaper up front but the maintenance cost on light rail vehicles is supposedly cheaper (plus, they can use electric power and are cleaner than diesel buses). The article’s criticisms would apply equally to BRT though.
I could not agree more. Public transit in the Portland (Oregon) metro area was much better (easier to access, predictable, faster, weather tolerant) when they prioritized buses and had no LRT. Now that the city is all-in on LRT the bus system is neglected and under provisioned.
My other big beef with LRT is that cities are dumping hundreds-of-millions of dollars in to single use track - LRT is only used for moving people, whereas money spent on buses funds roads, which benefit all aspects of commerce. Also, on another, tangent, LRT is plagued by weather related woes - at least in Portland, if it is too hot or too cold, the trains cannot even safely run. Bring back the buses!
I haven’t been here long, but I can typically grab a bus everywhere that LRT isn’t. Like the article says, busses can take you to LRT stations and there are a lot of examples of stations bus routes feed into (Gateway TC, Loyd Center, Parkrose TC)
Interesting article, except for the mumbo jumbo about feedback loops.
Development is all about land and cost. Suburbia hit critical mass because road networks were building out in the postwar economic stimulus and cars were cheaper than traditional infrastructure. Taxes were low because everything was new.
The driver of today’s changes is also about cost. We are out of suburban land. If you look at new construction in suburban areas, say around Boston, you see new build outs on shitty parcels. Apartment buildings 50 feet from the interstate, hilltop shopping centers, etc.
At the same time, the existing suburban housing stock is old and creaky. I was born in the late 70s, my parents are in their 70s and their house is 40 years old... and basically depreciated. It’s not going to yield great value because it needs lots of work. That’s not an atypical story.
A vital city like New York is another great place to look for guidance from. As the old ethnic neighborhoods collapsed, they turned to slums. The slumlords sucked the value out, and left cheap shells that attracted investment.
It’s an incentives problem in the United States. In Japan, for example, the railways are private (so they need to turn a profit) and you usually have supermarkets and department stores on the stations. The stations are the hub. When you look at commuter lines into large cities in the US, you are driving hours to a parking lot with nothing but a platform near by.
As I understand it, railway companies in Japan are real estate companies in disguise. They own the land around the stations and make a significant chunk of their money from that. Real estate around train stations is quite valuable.
I think you mean several miles. Which is why the PLSS grid looks like a checkerboard out west. Alternating sections for the public and the RR companies.
It's legal and it's a model being increasingly adopted in the US, for better or worse. In the NYC metro area, both Port Authority and NJ Transit are visibly leaning on this model to drive revenue.
It's also a geographic, real-estate, and political problem. There's not enough density in enough areas with enough last-mile coverage for trains to work like other countries.
Many cities solve this chicken and egg problem by building the necessary density next to the train station. Hong Kong and Japan are the most famous examples, but it works in Singapore and large chunks of Europe as well (although Parisian banlieues may not be the best model of urban planning for other reasons...).
I have no idea why you think it has to be “twice as good” or what all that even means.
With that behind us, nobody’s proposing high speed rail from San Diego to Boston, just LA to SF and along the eastern corridor. And we can’t event have that. Let’s start small.
It means that unless the trains are moving at 400mph with perfect comfort, free wifi, and a price of $20, nobody cares.
LA to SF is a 1 hour flight as cheap as $50, with the same last-mile effort and total trip time as a train. Airlines are also elastic to meet demand without major capex.
So who's going to spend the 100s of billions to buy the land and build a line for a 2-3x slower travel option that will take a century to be paid back? California is actually trying to do this and has failed miserably because the land and infrastructure costs alone make the project infeasible.
LA to SF isn't a 1 hour flight. LA to SF is: you start at downtown SF, 45 minutes later you arrive an hour early at SFO for your 1 hour 30 minute block-time flight to LA where you then take a 45 minute taxi ride to your final destination.
That's a total of 3.75 hours end to end, assuming no fog hits you at SFO, as it does on the regular. That's within spitting distance of a rail link that allow you to jump off anywhere along the way. It's by no means 2-3X slower if you take into account the entire process, with no TSA, starting at your origin and ending at your actual destination.
For instance Ottawa to Montreal is a "17 minute flight" that takes an hour of block time, and an hour on either end, for a total of 3h, or you could take the 1h 30m train ride along standard-speed rail for $20 USD. Trains can easily be faster taking into account externalities, and much more pleasant.
They usually do in fact have free wifi, better comfort, and a $20 price point -- specifically because as you point out, they're competing against air travel.
Hundred of billions is again a weird, defeatist argument. California has failed because infrastructure in America is not about building infrastructure, it's about graft, and if infrastructure gets built along the way, that's fine too. That's the saddest part, honestly.
Honestly, this is kind of the textbook example of defeatism.
That's why I said "with the same last-mile effort and total trip time as a train".
Planes cruise at 500mph. Trains at 200mph will take 2.5x longer at best, assuming there aren't any stops along the way, and you still need last-mile transport to get to your destination. Trains are not going to be faster than planes at this distance, and that differential only gets worse as distance increases.
Meanwhile economic realities aren't "defeatist". There are much better things to spend $100B on than a slower alternative for a few people to travel between 2 specific cities. Trains may be the answer in the future, but there's a lot more that needs to change first.
Train stations tend to be located right downtown whereas airports, for various reasons, far outside. You can't exclude that delta.
SFO is in San Mateo, the next county over -- almost 15 miles away. Oakland airport is closer to 20 miles away. A minimum of 30 minutes by car, 40 minutes by Bart from downtown SF. On the other hand, the high-speed rail link would pick up at the Salesforce Transit Center (Embarcadero Station) at the corner Howard and Fremont, stopping at San Jose Diridon station.
Yes plaines are fast in the air, that's not a surprise. However, they're epically slow when parked at the gate while you clear security, epically slow as they taxi, dead stopped as they wait to take off due to ATC hold and weather issues, and similarly dead stopped when they're waiting for the previous plane to clear the gate at the destination airport, while you wait in line to get off, and while you wait for your bags, and while you wait for your taxi.
C'mon now. "Plane go fast" is just a small part of the story, and I'm an AA EXP.
Again, I think the infrastructure costs in the US tend to be massively overinflated due to graft instead of actual cost.
This conversation has gone off a tangent. The comparison isn't about trains being more comfortable than planes (that's obvious). It's about planes vs trains as a transportation option.
It's not viable to build a long-distance rail line given the same total time (even if more comfortable since people want cheap and fast), especially not at the cost of $100B+.
If you want to call it inflated prices because of graft then fine, but that means you need to solve that first. Which goes back to my original point that there are many other obstacles that need to be overcome before high-speed rail makes sense here.
You haven’t provided any data to show it doesn’t make sense to build a 300 mile long train link between SF and LA other than your gut telling you (wrongly) that travel time would be 2-3X longer when plenty of other examples all over the world show otherwise.
It doesn't make sense because they did a dreadful job of controlling costs. China pays between $17M and $21M per kilometer to build high-speed rail in challenging terrain, and even Europe only pays $25-39M per kilometer. [1]
California on the other hand was going to spend $100B for 840km (phase 1, beyond the central corridor on both sides) or $119M per kilometer. That's obviously 7X China's cost and 5X Europe's cost. Yes, that Europe, the one with the onerous regulations.
It's hard to point at California's staggeringly inefficient attempt as a reason why rail is bad.
A better starting point would be Via Rail's corporate plan 2017-2021 [2].
Montreal-Ottawa travel time is 13% faster by train than car, and 30% faster than air. Average fare is $48.50 CAD ($35 USD). For comparison, a flight would be $167 to $186 each way, 4X more expensive and 30% longer end to end.
Numbers are more comparable on the Toronto to Ottawa and Toronto to Montreal corridors, with train travel coming in approximately the same time end to end as a flight but 25% less expensive. This corridor is highly competitive and similar distance to SF/LA.
Rail does in fact satisfy your requirements for short trips.
Emperor Bezos from Elysium will instruct Count Musk from Mars to finally reinvigorate the barren wastelands of old Earth by laying hyperloops of chinese provenance across. The mutated apes will scream with joy.
For instance Ottawa to Montreal is a "17 minute flight" that takes an hour of block time, and an hour on either end, for a total of 3h, or you could take the 1h 30m train ride along standard-speed rail for $20 USD.
The even better comparison, while not in America, is the route from Tokyo to Osaka. The distance (506km) is almost the same as from LA to San Francisco.
The Shinkansen takes between 135 and 153 minutes. Granted, you have to get from Shin Osaka to the city center; an additional 4 minute train ride or you take the subway.
Let's not even get into the comparison of the travel experience between gliding in a quiet, serene manner on spacious seat past Fuji San with the experience you get in a middle seat with a 29" seat pitch in economy class of a domestic flight in the US.
Also, if you invest a few hundred yen into a bento at Tokyo station let's not even think about the comparison of the culinary delights awaiting you for the trip.
How fast is this train going? And you still need to get to/from the station. At 200mph it'll also be a 3-4 hour ordeal at least, and more if there are stops in the middle.
And you have wifi/mobile-net, comfortable seating, a table, power socket.
Now, that said, if the stations are not connected to the cities they are in, then that's a problem. And maybe this is a near unsolvable one, due to costs.
Typically, the commercial speed of high-speed trains is about 300-350km/h.
> And you still need to get to/from the station.
Train stations tend to be located in the city center, not 50km away.
> At 200mph it'll also be a 3-4 hour ordeal at least, and more if there are stops in the middle.
Madrid-Barcelona takes less than 3 hours, and starts/ends near the city center (check Atocha and Barcelona Santa).
Unlike airplanes, the bulk of the trip is spent sitting comfortably in your cozy chair with plenty of leg room, internet access, electric outlets, no noise or pressure fluctuation, and a nice landscape to enjoy.
This thread has gone off the rails. The point isn't that trains are more comfortable than planes. That's obvious. It's that a train is not currently viable compared to the existing airline routes.
The US and California are not spending > $100B to build this. Even with all that comfort, people want cheap and fast. Talking more about how nice trains are makes no difference to actually getting one built unless you have a plan on how to acquire all the money, real estate and political will to get it done.
The train is also slower than a simple speed and distance calculation would suggest.
The route won't be as straight as the route for an aircraft. Problems along the route (tight curve, noise limit, etc.) will create low-speed zones. Every town along the way will demand a stop, so the train has to slow for a stop before even reaching full speed.
I don't know the full list of requirements, but a route has been chosen, some number of federal approvals have been granted, some regulatory hurdles passed, and supposedly two federal approvals remain.
The route is easier (engineering and cost-wise), due to geography, than the California rail, or anything in the Eastern corridor, and to keep it even easier they're running it along existing utility corridors where possible to further reduce the amount of private land access they have to contend with:
That falls under "political" problem. Who's going to build it? Who's paying for it? Where are they getting the land? Is it existing rail lines or new ones?
The obstacles are known, they need to be overcome, and that's very difficult in the 3rd-biggest country by population and land area with an incredibly diverse set of national and regional governance and cultures.
There are some odd outliers such as Salt Lake City, which has pretty low density, but the transit system is well used. It's far from universal service but in the areas it serves, it is rather popular with a wide cross section of the population.
I know this will sound super cynical and overly pedantic. But it's not the Koch brothers anymore. I was by no means a fan, but David is dead. Any funding is on Charles. Or perhaps Charles and David's heirs.
From the article linked in the sentence referencing the Koch brothers: "But in 2017, the group received $110,000 in funding from a political arm of Charles and David Koch". Both were alive in 2017.
Spot on. I live in Minneapolis metro area, and commute to downtown Minneapolis via bus. The "express" bus from the suburb is not frequent enough and does not extend far so I can walk to my house after getting off.
Instead, I drive a good 10+ minutes to a park & ride, and get on the bus for the 45+ minutes ride to downtown. Also the buses could be electric and more comfortable, and have better lane management. Here in MSP, the transit buses can drive on the shoulder which are in terrible shape.
Oh, there was a discussion of an LRT to downtown St. Paul from where I live. We all heard the same arguments about "bringing the wrong kind of people" to our neighborhood. Instead, they are developing bus lines which is the better call, but have to see the ridership. Hope it works out.
One thing I'm surprised this article didn't mention is commute times. To get from Weston Station (northwest of downtown) to Union Station takes:
- 22-minute drive (or 1h in rushour)
- 1.25-hour bus ride
- 13-minute LRT ride
LRT (light rail transit) really is an excellent way to get into the downtown core from distances and in time frames that were previously only manageable by car.
That example isn't really LRT, it's either commuter rail or an airport train, which (either way) only stops once during that 13-km trip. LRT normally involves much closer stops.
That could be getting at what confused me about the article. The author mentions the crosstown LRT favourably, and that will link up to the UP commuter/airport line I mentioned, adding another stop.
LRT linking to commuter rail lines, subway lines, and bus stops is something the author didn't really mention, but it certainly seems to be the strategy for that line in Toronto.
The author is against cutting funding for bus routes, which is reasonable. I guess I just don't see how that is also an argument against using LRT for arterial transportation into downtown cores. If the retort to that is "budgets", then we need increased funding for public transportation, not fewer trains.
If you are talking about the UPX, that’s not an LRT line. It’s just a regular train.
But I don’t think it’s a fair comparison based on routes. On the UPX line those are 1 direct stop apart. For comparison taking the subway would also be more indirect and longer, taking about 40 minutes I would guess.
Average wait time of 7 minutes if the trains follow the schedule exactly and don't bunch.
If things are allowed to drift randomly, you end up approximating the exponential distribution - which has an average wait time of 15 minutes for 4 trains/hour.
When traffic in a city is inherently bad. It sets the foundation for difficulty in solving the problem with buses (increase frequency, priority express buses)
In a high traffic situation, ETA gets adjusted from time to time. And buses will struggle to get in/out of stops and delays loom when accident occurs on the road.
Poor road planning and building alongside with massive amount of cars on the road is one of the major issue.
Maybe workplace hours should be adjusted as a whole or varying tolls based on high demand hours to avoid everyone competing for road space.
But I generally agree with the article that buses gets you really far. For a long time Singapore was relying on bus routes to move people around the city state, and progressively build subway lines as the demand and city planning go with it.
Interestingly, Toronto has pretty terrible congestion. I haven't lived there in a while, but IIRC "rush hour" is more like 3 hours (morning an evening). And (apparently) the bus network still does a decent job. Not sure how true that is, but I hope they at least added dedicated bus lanes wherever possible.
It's interesting to note that many SF residents find the T 3rd Street light rail to be slower and less frequent than the buses that it replaced.
I think subways work well because of their grade separation and extreme high capacity, but most surface rights would be better served by buses than light rail, and you can still create separated bus lanes for bus rapid transit instead of separated light rail if you have the money to spend.
1. I am not convinced the idea of commuting as "self-correcting mechanism" was presented well (or much at all). I was expecting to learn a bit more about this point.
2. Half of the points have a counter-example in both Houston and Dallas, where most of the LRT doesn't go to wealthy residential parts of town. Some of those rich parts of town are still actively fighting it coming near them, in fact.
Those two cities come to mind because I have experience living in both, and can compare with my experiences living in denser, older cities like Tokyo, Berlin, and Santiago, where a variety of networks were built for different reasons across different eras. Then again, even in Santiago, the the subway stops just before you get too deep into residential parts of Las Condes.
But I am also a non-expert, and at the end of the day, instead of going point-for-point and saying "Well, look at what Dallas did. Surely [blah, blah, blah]" there are two factors that weigh heavier on my mind: cost and historical political context.
Cost is foremost. While LRT is less efficient long-term, it is cheaper to build transit at-grade by about a factor of five (I read this long ago and forget where, or I would happily source). Histotical context, in that many grand subway systems get built in a time where, for the city to prove it was a top-tier global city, you demonstrated civic grandeur and engineering prowess by building a subway system. The prestige was used to push past the barrier of it being eye-wateringly expensive. Also while we have made these options safer over time, it has come at a (worthwhile, but non-zero) cost of more regulatory checks that have made the cost gap even greater, and harder to cover by dreaming of prestige
Both cost and context ultimately combine into political feasibility.
For a city like Denver, Dallas or Houston, if you ask the transit wonks who fight the good fight at the municipal level, I don't think you would find many that truly would choose LRT over other commuter rail, all else equal. But all else was not equal. Cost was the big barrier, and the context did not exist to cross that chasm, so they got what they could, when they could.
Houston only got its network because of a starry-eyed chance at the Olympics and/or Super Bowl. Originally it was just one line, then a plan for a 5-line network, of which 3 were built, but now possibly up to 4 again. I could write an entire essay on the mistakes made along the way, public and private, in that first line. Just look at the land-use (really the land-lack-of-use) in Midtown between 2000 and 2010 despite the LRT's presence. Yet the power of a network grows with every additional line, so once you at least have something, you might as well leverage that and add to it, rather than sell the public on a completely different approach.
In the end, I share the concerns of the author, but I do not see the narrative being quite as clean when I set it against the specific cases I am familiar with.
> and you’re the one who has to pay the cost of commuting distance
In Japan, the company pays for your commute, and the price of your commuter pass has no bearing on your pay. (There is a cap at most companies, and the average is 15170.8 yen per month (according to https://venture-finance.jp/archives/4982).)
(Unfortunately, as the company is forced to pay for your commute, the company is also allowed to dictate what route to take, which may in some cases mean less comfort. Neither you nor the company pay tax on this money, so misuse (and paying more than necessary could possibly be construed as misuse on part of the company) could in theory have bad consequences.)
I can't make the sense of the author's graphs. Is he plotting daily boardings on both the X and Y axes simultaneously? Aside from bus vs. streetcar, what do the varying colors mean?
For the “changing routes” problem, I like trolleybuses, which won’t be reconfigured and don’t smell like diesel. Real estate near the wires can get the usual infrastructure boost. Dual-mode trolleybuses can also extend service past where the wires are and you still get the advantage of not stopping to charge. The power stations can be reused for a train as well. Dedicated ROW in a city center is mandatory regardless of vehicle if you want reliable service.
Toronto seems like a bit of an unfair example when you compare sheer ridership because it’s bigger than all but five US cities.
This article is just BS. This is a line used vary often by fossil fuel advocates to dissuade investment in mass transit infrastructure. You see the fossil fuel lobby cannot say "screw mass transit, lets all drive" because after suffering couple of hours in traffic, every citizen will figure out that this theory is unworkable. So instead, they say "we don't need rail, we need buses, and we need to build "bus infrastructure" (which is just roads)." Then of course, the fossil fuel lobby can work behind the scenes to gut the bus budget and the "bus infrastructure" can be used for cars instead.
The lobbyists then say "Buses are for ordinary people, railroads are for the rich, blah blah blah". Usually, this line completely ignores the fact that buses and railroads cost about the same in most places.
To his credit the author does address this fact here, but his analysis is shallow and downright silly. He says, "yes they cost the same, but railroads are usually in more expensive neighborhoods." Well, yeah we live in age when high density neighborhoods are expensive and getting pricier all the time and Light rail is most suitable for high density neighborhoods. So that is not surprising.
And then he complains about how light rail makes home prices go up. Yeah, well that just means it is successful. When a city service is successful, it makes the area it serves more desirable which people from less desirable areas want to move there which makes prices go up.
The response to that should not be to make everything less desirable to ensure affordability, but to learn from the successful and desirable places and make all parts as successful and desirable. Thus, eventually there won't be enough rich people to move in, and you will have just a nice affordable livable city. Would you be in favor of setting half a city on fire in order to make the other half more affordable to live in?
His graphs show that he is trying to mislead readers. Why do you think he made the color of bus lines in Toronto so similar to the color of light rail in Toronto? Or why did he color certain high usage bus lines that prove his point in bright red, and certain low usage bus lines that do not prove his point in much darker shades, and then in the legend he only said that the bright red lines are bus lines. Or why did he split the most popular Toronto street car line in two? He said "I separated out the 504A and 504B so that they’d fit on the graph; otherwise it’d break the y axis." They would not only break the y axis, buddy, they would also break the entire premise of your article.
If he had a dot that represented a light rail line that is much twice as high as any of the bus lines, then his premise that light rail is unnecessary because it can be handled by a bus line kind of goes out the window. But that is the reality.
He correctly says that light rail and buses work best in combination with buses being feeders for light rail lines. This increases the ridership of both buses and light rail. But that fact leads to the conclusion that both light rail and buses must be built. This is what most cities try to do nowadays. Instead the author draws the absolutely wrong conclusion that light rail needs not be built.
So yeah in conclusion this is a typical professionally made manipulative troll article. It has a couple of nuggets of self evident truth to get some authenticity and not look like a complete troll, it tries to harness a lot of pre-existing anger about an issue (high housing prices) and disingenuously channel that anger towards his issue (less light rail), and it has some graphs that are very carefully designed to be confusing and misleading to the max.
Thank you for raising these points because no one else has, in this thread or the previous one (linked in another comment).
I especially want to emphasize: the "streetcars" that the author equates to buses in his graphs and his reasoning are by all measures light rail. Toronto has a mix of subway, LRT, and buses that seem to work together. This totally undermines his entire argument, not to mention makes his article dishonest.
Exactly what I was thinking thank you. The author also goes on about the suburbs having a negative feedback loop, but clearly history shows that suburban sprawl kept growing furled by larger motorways and parking lots.
Some thought deserves to go to other cities outside of North America also, Melbourne for example has an incredible tram network, supported by bus lines and a heavy rail network for faster transport more remote suburbs
I totally agree. A specific example I found insulting was Portland, OR. He uses it as an example for his argument, but the reality is that our LRT is still supported by a vastly larger bus network. I live a 15 minute walk / 5 minute drive to the nearest train station, but there are still at least 5 or 6 bus lines closer. Additionally, the bus network still seems well funded. Not only are busses relatively clean and modern, they also support all the latest mobile payment options like ApplePay and the iPhone’s new mobile card thingy that works with RFID just like the trains do.
Streetcars in Toronto are in fact light rail, doesn't that invalidate this entire argument?
One can argue some LRT implementations have the same limitations as buses, but you can't call them buses, point to their success, and say that means buses are the only answer.
It's clear that Toronto's mix of subway, LRT, and buses are successful, and perhaps that's what should be emulated.
This person has it all wrong. Suburbs were created artificially, and it just doesn’t matter. Mixed income neighbourhoods are the desired result of all development. The pockets, rings, and regions of classist housing is the problem, not its location or how we navigate them.
Yes, the feedback loops is true, blah blah blah, but author needs to dig deeper.
What? Light rail is really crap. 45 minutes from Sunnyvale to San Jose. It's slower than biking. It's covering like 6 miles in that time. 9 miles an hour? Really?
98 comments
[ 2.9 ms ] story [ 163 ms ] threadAnd I would suspect this is how Toronto differs from the average American city. American city governments experience periodic budget squeezes that force politicians to gut every service which is gut-able and buses are always a logical target.
And this leads to both buses being abandoned due to unreliability and for the bus lines that do exist to, uh, worthless, run at too infrequent intervals to pick up significant ridership. And this lead to buses having a terrible reputation, reinforcing the cycle of poor ridership.
Why would trains at street level be limited in speed by traffic? Trains in the UK run at street level and regularly go over a hundred miles an hour.
There are also "bus only" lanes, except the buses can go around blockages sometimes.
If you check out YouTube you can find some freight trains that have to do that from time-to-time.
About a hundred years ago, when Toronto’s streetcar was implemented, they made their gauge slightly narrower than train gauge to prevent “trains” from running along streetcar tracks.
They pay for it to this day as it means new streetcars need to be designed for a gauge that nobody else uses.
Turning radii is also an issue. Possibly by design or not...
The streetcars are not discussed much in the article, but they may be a large part of both Toronto's moderate transit success and its slowness to adopt real light rail transit.
Streetcars are seen as similar to buses but a bit faster, more comfortable and more reliable. Which suits them well for the busy downtown routes. But plans for suburban LRT routes have been dismissed as "glorified streetcars", even though they would be much faster due to less frequent stops and more signal priority. Since relatively slow downtown streetcars are the only surface rail people know about, they don't get excited about LRT.
That's not necessarily so. Trams (streetcars) in cities, which have old tram networks usually segregate the lines from the parts of the street where privat traffic runs. Very often they also receive signalling priority and while blockage by cars can happen (especially on intersections) trams can run with a quite reliable time table.
I live in a city with one of the - arguably - best public transport systems at least outside of Japan. If anything runs in unreliable intervals (as you learn over announcemets while sitting in a tram) it's bus lines.
Now, if you just build some light rail on an existing road as often happens because public transport is chic nowadays then you run exactly into the problem you describe. But dedicated tram networks are very different and quite reliable.
The OP makes a great argument that buses plus medium rail/subways/etc is much better combination to add to a city. I think that argument has to be made more loudly, for the other reason - US planners are bad, if there's a to force them to be better it should be taken. Maybe really smart people do light rail well but the US city doesn't deserve to be trusted to do that.
https://www.khou.com/article/traffic/metro-houston-approves-...
On the other side of things, Austin may get a third crack at approving an LRT system next year, but the last two times Austin voters disapproved of LRT. As a result they have a weird heavy commuter rail that runs on a strange schedule, goes almost nowhere, and does not serve the people well.
Sure an actual Metro-type system, like BART in the SF Bay area, Toronto (apparently) or the various marvelous French systems that exist, these are faster than a car. But you might not have experienced the worthless, pathetic hunks of junk that Americans build and call "light rail", these can manage to be slower even than buses and more expensive and less reliable. The San Jose light rail system is a prime specimen, a "transit system" that barely manages to be faster than walking.
Of course, when I lived in San Rosa, the bus system also had an expected arrival times identical to walking - I tested this often on Google maps in off-hours. Where I live now, the bus comes twice a day so it's a further world of impractical.
Anyway, in America, it's not just that not much money gets put into public transit. it's even just that cities were created around the car. It's that the state, civic fathers, the ruling/managing group, doesn't really think public transit is necessary so much of the money that's allocated can be passed to the rulers' friends since "no one" expects public transit to work.
What exactly the difference between heavy rail and light rail is a matter of debate in some cases but I don’t think there’s any dispute that DC Metro or NY Subway is heavy rail.
If you’re building a new rapid transit line with dedicated ROW, it then comes down to BRT vs LRT and the cost of BRT and LRT aren’t actually that different I believe. The buses are cheaper up front but the maintenance cost on light rail vehicles is supposedly cheaper (plus, they can use electric power and are cleaner than diesel buses). The article’s criticisms would apply equally to BRT though.
My other big beef with LRT is that cities are dumping hundreds-of-millions of dollars in to single use track - LRT is only used for moving people, whereas money spent on buses funds roads, which benefit all aspects of commerce. Also, on another, tangent, LRT is plagued by weather related woes - at least in Portland, if it is too hot or too cold, the trains cannot even safely run. Bring back the buses!
I think the 'red transit lane' is a step in that direction:
https://www.oregonlive.com/commuting/2019/10/paint-it-red-po...
Thank you but no thank you. We need less money spend on inefficient roads, not more.
Development is all about land and cost. Suburbia hit critical mass because road networks were building out in the postwar economic stimulus and cars were cheaper than traditional infrastructure. Taxes were low because everything was new.
The driver of today’s changes is also about cost. We are out of suburban land. If you look at new construction in suburban areas, say around Boston, you see new build outs on shitty parcels. Apartment buildings 50 feet from the interstate, hilltop shopping centers, etc.
At the same time, the existing suburban housing stock is old and creaky. I was born in the late 70s, my parents are in their 70s and their house is 40 years old... and basically depreciated. It’s not going to yield great value because it needs lots of work. That’s not an atypical story.
A vital city like New York is another great place to look for guidance from. As the old ethnic neighborhoods collapsed, they turned to slums. The slumlords sucked the value out, and left cheap shells that attracted investment.
https://en.m.wikipedia.org/wiki/Checkerboarding_(land)
The US doesn't just need trains as good as Japan, it needs them to be at least twice as good for the investments to make sense.
With that behind us, nobody’s proposing high speed rail from San Diego to Boston, just LA to SF and along the eastern corridor. And we can’t event have that. Let’s start small.
LA to SF is a 1 hour flight as cheap as $50, with the same last-mile effort and total trip time as a train. Airlines are also elastic to meet demand without major capex.
So who's going to spend the 100s of billions to buy the land and build a line for a 2-3x slower travel option that will take a century to be paid back? California is actually trying to do this and has failed miserably because the land and infrastructure costs alone make the project infeasible.
That's a total of 3.75 hours end to end, assuming no fog hits you at SFO, as it does on the regular. That's within spitting distance of a rail link that allow you to jump off anywhere along the way. It's by no means 2-3X slower if you take into account the entire process, with no TSA, starting at your origin and ending at your actual destination.
For instance Ottawa to Montreal is a "17 minute flight" that takes an hour of block time, and an hour on either end, for a total of 3h, or you could take the 1h 30m train ride along standard-speed rail for $20 USD. Trains can easily be faster taking into account externalities, and much more pleasant.
They usually do in fact have free wifi, better comfort, and a $20 price point -- specifically because as you point out, they're competing against air travel.
Hundred of billions is again a weird, defeatist argument. California has failed because infrastructure in America is not about building infrastructure, it's about graft, and if infrastructure gets built along the way, that's fine too. That's the saddest part, honestly.
Honestly, this is kind of the textbook example of defeatism.
Planes cruise at 500mph. Trains at 200mph will take 2.5x longer at best, assuming there aren't any stops along the way, and you still need last-mile transport to get to your destination. Trains are not going to be faster than planes at this distance, and that differential only gets worse as distance increases.
Meanwhile economic realities aren't "defeatist". There are much better things to spend $100B on than a slower alternative for a few people to travel between 2 specific cities. Trains may be the answer in the future, but there's a lot more that needs to change first.
SFO is in San Mateo, the next county over -- almost 15 miles away. Oakland airport is closer to 20 miles away. A minimum of 30 minutes by car, 40 minutes by Bart from downtown SF. On the other hand, the high-speed rail link would pick up at the Salesforce Transit Center (Embarcadero Station) at the corner Howard and Fremont, stopping at San Jose Diridon station.
Yes plaines are fast in the air, that's not a surprise. However, they're epically slow when parked at the gate while you clear security, epically slow as they taxi, dead stopped as they wait to take off due to ATC hold and weather issues, and similarly dead stopped when they're waiting for the previous plane to clear the gate at the destination airport, while you wait in line to get off, and while you wait for your bags, and while you wait for your taxi.
C'mon now. "Plane go fast" is just a small part of the story, and I'm an AA EXP.
Again, I think the infrastructure costs in the US tend to be massively overinflated due to graft instead of actual cost.
It's not viable to build a long-distance rail line given the same total time (even if more comfortable since people want cheap and fast), especially not at the cost of $100B+.
If you want to call it inflated prices because of graft then fine, but that means you need to solve that first. Which goes back to my original point that there are many other obstacles that need to be overcome before high-speed rail makes sense here.
And like I said before, California is actually trying to do this and completely failed: https://www.vox.com/policy-and-politics/2019/2/15/18224717/c...
California on the other hand was going to spend $100B for 840km (phase 1, beyond the central corridor on both sides) or $119M per kilometer. That's obviously 7X China's cost and 5X Europe's cost. Yes, that Europe, the one with the onerous regulations.
It's hard to point at California's staggeringly inefficient attempt as a reason why rail is bad.
A better starting point would be Via Rail's corporate plan 2017-2021 [2].
[1] http://www.globalconstructionreview.com/sectors/why-china-ca...
[2] https://www.viarail.ca/sites/all/files/media/pdfs/About_VIA/...
Montreal-Ottawa travel time is 13% faster by train than car, and 30% faster than air. Average fare is $48.50 CAD ($35 USD). For comparison, a flight would be $167 to $186 each way, 4X more expensive and 30% longer end to end.
Numbers are more comparable on the Toronto to Ottawa and Toronto to Montreal corridors, with train travel coming in approximately the same time end to end as a flight but 25% less expensive. This corridor is highly competitive and similar distance to SF/LA.
Rail does in fact satisfy your requirements for short trips.
(editypo)
The even better comparison, while not in America, is the route from Tokyo to Osaka. The distance (506km) is almost the same as from LA to San Francisco.
The Shinkansen takes between 135 and 153 minutes. Granted, you have to get from Shin Osaka to the city center; an additional 4 minute train ride or you take the subway.
Let's not even get into the comparison of the travel experience between gliding in a quiet, serene manner on spacious seat past Fuji San with the experience you get in a middle seat with a 29" seat pitch in economy class of a domestic flight in the US.
Also, if you invest a few hundred yen into a bento at Tokyo station let's not even think about the comparison of the culinary delights awaiting you for the trip.
How long does it take to/from each airport, and how long do you take to go through security, check-in and boarding?
A 1-hour flight easily becomes a 3 to 4 hour ordeal.
Now, that said, if the stations are not connected to the cities they are in, then that's a problem. And maybe this is a near unsolvable one, due to costs.
Typically, the commercial speed of high-speed trains is about 300-350km/h.
> And you still need to get to/from the station.
Train stations tend to be located in the city center, not 50km away.
> At 200mph it'll also be a 3-4 hour ordeal at least, and more if there are stops in the middle.
Madrid-Barcelona takes less than 3 hours, and starts/ends near the city center (check Atocha and Barcelona Santa).
Unlike airplanes, the bulk of the trip is spent sitting comfortably in your cozy chair with plenty of leg room, internet access, electric outlets, no noise or pressure fluctuation, and a nice landscape to enjoy.
The US and California are not spending > $100B to build this. Even with all that comfort, people want cheap and fast. Talking more about how nice trains are makes no difference to actually getting one built unless you have a plan on how to acquire all the money, real estate and political will to get it done.
The route won't be as straight as the route for an aircraft. Problems along the route (tight curve, noise limit, etc.) will create low-speed zones. Every town along the way will demand a stop, so the train has to slow for a stop before even reaching full speed.
https://www.texastribune.org/2019/05/15/texas-bullet-train-a...
I don't know the full list of requirements, but a route has been chosen, some number of federal approvals have been granted, some regulatory hurdles passed, and supposedly two federal approvals remain.
The route is easier (engineering and cost-wise), due to geography, than the California rail, or anything in the Eastern corridor, and to keep it even easier they're running it along existing utility corridors where possible to further reduce the amount of private land access they have to contend with:
https://www.texascentral.com/alignment-maps/
The obstacles are known, they need to be overcome, and that's very difficult in the 3rd-biggest country by population and land area with an incredibly diverse set of national and regional governance and cultures.
Instead, I drive a good 10+ minutes to a park & ride, and get on the bus for the 45+ minutes ride to downtown. Also the buses could be electric and more comfortable, and have better lane management. Here in MSP, the transit buses can drive on the shoulder which are in terrible shape.
- 22-minute drive (or 1h in rushour)
- 1.25-hour bus ride
- 13-minute LRT ride
LRT (light rail transit) really is an excellent way to get into the downtown core from distances and in time frames that were previously only manageable by car.
LRT linking to commuter rail lines, subway lines, and bus stops is something the author didn't really mention, but it certainly seems to be the strategy for that line in Toronto.
The author is against cutting funding for bus routes, which is reasonable. I guess I just don't see how that is also an argument against using LRT for arterial transportation into downtown cores. If the retort to that is "budgets", then we need increased funding for public transportation, not fewer trains.
But I don’t think it’s a fair comparison based on routes. On the UPX line those are 1 direct stop apart. For comparison taking the subway would also be more indirect and longer, taking about 40 minutes I would guess.
UPX runs every 15 mins.
If you time it right, great, but if you don’t...
If things are allowed to drift randomly, you end up approximating the exponential distribution - which has an average wait time of 15 minutes for 4 trains/hour.
In a high traffic situation, ETA gets adjusted from time to time. And buses will struggle to get in/out of stops and delays loom when accident occurs on the road.
Poor road planning and building alongside with massive amount of cars on the road is one of the major issue.
Maybe workplace hours should be adjusted as a whole or varying tolls based on high demand hours to avoid everyone competing for road space.
But I generally agree with the article that buses gets you really far. For a long time Singapore was relying on bus routes to move people around the city state, and progressively build subway lines as the demand and city planning go with it.
I think subways work well because of their grade separation and extreme high capacity, but most surface rights would be better served by buses than light rail, and you can still create separated bus lanes for bus rapid transit instead of separated light rail if you have the money to spend.
1. I am not convinced the idea of commuting as "self-correcting mechanism" was presented well (or much at all). I was expecting to learn a bit more about this point.
2. Half of the points have a counter-example in both Houston and Dallas, where most of the LRT doesn't go to wealthy residential parts of town. Some of those rich parts of town are still actively fighting it coming near them, in fact.
Those two cities come to mind because I have experience living in both, and can compare with my experiences living in denser, older cities like Tokyo, Berlin, and Santiago, where a variety of networks were built for different reasons across different eras. Then again, even in Santiago, the the subway stops just before you get too deep into residential parts of Las Condes.
But I am also a non-expert, and at the end of the day, instead of going point-for-point and saying "Well, look at what Dallas did. Surely [blah, blah, blah]" there are two factors that weigh heavier on my mind: cost and historical political context.
Cost is foremost. While LRT is less efficient long-term, it is cheaper to build transit at-grade by about a factor of five (I read this long ago and forget where, or I would happily source). Histotical context, in that many grand subway systems get built in a time where, for the city to prove it was a top-tier global city, you demonstrated civic grandeur and engineering prowess by building a subway system. The prestige was used to push past the barrier of it being eye-wateringly expensive. Also while we have made these options safer over time, it has come at a (worthwhile, but non-zero) cost of more regulatory checks that have made the cost gap even greater, and harder to cover by dreaming of prestige
Both cost and context ultimately combine into political feasibility.
For a city like Denver, Dallas or Houston, if you ask the transit wonks who fight the good fight at the municipal level, I don't think you would find many that truly would choose LRT over other commuter rail, all else equal. But all else was not equal. Cost was the big barrier, and the context did not exist to cross that chasm, so they got what they could, when they could.
Houston only got its network because of a starry-eyed chance at the Olympics and/or Super Bowl. Originally it was just one line, then a plan for a 5-line network, of which 3 were built, but now possibly up to 4 again. I could write an entire essay on the mistakes made along the way, public and private, in that first line. Just look at the land-use (really the land-lack-of-use) in Midtown between 2000 and 2010 despite the LRT's presence. Yet the power of a network grows with every additional line, so once you at least have something, you might as well leverage that and add to it, rather than sell the public on a completely different approach.
In the end, I share the concerns of the author, but I do not see the narrative being quite as clean when I set it against the specific cases I am familiar with.
In Japan, the company pays for your commute, and the price of your commuter pass has no bearing on your pay. (There is a cap at most companies, and the average is 15170.8 yen per month (according to https://venture-finance.jp/archives/4982).)
(Unfortunately, as the company is forced to pay for your commute, the company is also allowed to dictate what route to take, which may in some cases mean less comfort. Neither you nor the company pay tax on this money, so misuse (and paying more than necessary could possibly be construed as misuse on part of the company) could in theory have bad consequences.)
Toronto seems like a bit of an unfair example when you compare sheer ridership because it’s bigger than all but five US cities.
Actually the metro area pop. is only beat by LA IIRC, so your point is even understated
The lobbyists then say "Buses are for ordinary people, railroads are for the rich, blah blah blah". Usually, this line completely ignores the fact that buses and railroads cost about the same in most places.
To his credit the author does address this fact here, but his analysis is shallow and downright silly. He says, "yes they cost the same, but railroads are usually in more expensive neighborhoods." Well, yeah we live in age when high density neighborhoods are expensive and getting pricier all the time and Light rail is most suitable for high density neighborhoods. So that is not surprising.
And then he complains about how light rail makes home prices go up. Yeah, well that just means it is successful. When a city service is successful, it makes the area it serves more desirable which people from less desirable areas want to move there which makes prices go up.
The response to that should not be to make everything less desirable to ensure affordability, but to learn from the successful and desirable places and make all parts as successful and desirable. Thus, eventually there won't be enough rich people to move in, and you will have just a nice affordable livable city. Would you be in favor of setting half a city on fire in order to make the other half more affordable to live in?
His graphs show that he is trying to mislead readers. Why do you think he made the color of bus lines in Toronto so similar to the color of light rail in Toronto? Or why did he color certain high usage bus lines that prove his point in bright red, and certain low usage bus lines that do not prove his point in much darker shades, and then in the legend he only said that the bright red lines are bus lines. Or why did he split the most popular Toronto street car line in two? He said "I separated out the 504A and 504B so that they’d fit on the graph; otherwise it’d break the y axis." They would not only break the y axis, buddy, they would also break the entire premise of your article.
If he had a dot that represented a light rail line that is much twice as high as any of the bus lines, then his premise that light rail is unnecessary because it can be handled by a bus line kind of goes out the window. But that is the reality.
He correctly says that light rail and buses work best in combination with buses being feeders for light rail lines. This increases the ridership of both buses and light rail. But that fact leads to the conclusion that both light rail and buses must be built. This is what most cities try to do nowadays. Instead the author draws the absolutely wrong conclusion that light rail needs not be built.
So yeah in conclusion this is a typical professionally made manipulative troll article. It has a couple of nuggets of self evident truth to get some authenticity and not look like a complete troll, it tries to harness a lot of pre-existing anger about an issue (high housing prices) and disingenuously channel that anger towards his issue (less light rail), and it has some graphs that are very carefully designed to be confusing and misleading to the max.
I especially want to emphasize: the "streetcars" that the author equates to buses in his graphs and his reasoning are by all measures light rail. Toronto has a mix of subway, LRT, and buses that seem to work together. This totally undermines his entire argument, not to mention makes his article dishonest.
Some thought deserves to go to other cities outside of North America also, Melbourne for example has an incredible tram network, supported by bus lines and a heavy rail network for faster transport more remote suburbs
One can argue some LRT implementations have the same limitations as buses, but you can't call them buses, point to their success, and say that means buses are the only answer.
It's clear that Toronto's mix of subway, LRT, and buses are successful, and perhaps that's what should be emulated.
Yes, the feedback loops is true, blah blah blah, but author needs to dig deeper.
1. If the road traffic is congested, the solution is bus lanes.
2. If the bus lanes are congested, the solution is light rail rapid transit.
3. If the light rail is congested, the solution is a subway/train line.