The bus system has to make pretty aggressive tradeoffs in order to be useful to commuters. You can't make it stop at every block in order to make it take less walking to get to the bus stop. Similarly, you can't have parallel lines a block apart. You have to limit the service area so that you can afford to run busses every 5-10 minutes. And you'll want to improve streets so that busses can run down the center and pick up passengers from protected islands, rather than fighting with parking and turning traffic for space.
Like, I take the BART every day into work. It's fantastic. I walk for five to ten minutes to get to the station, a train comes every 5 minutes so I don't have to worry about the train schedule, then I walk another five to ten minutes to get to the office. If the bus experience could match that, it'd take over a lot of commuting.
Like, I take the BART every day into work. It's fantastic.
I walk for five to ten minutes to get to the station, a
train comes every 5 minutes so I don't have to worry about
the train schedule, then I walk another five to ten minutes
to get to the office. If the bus experience could match
that, it'd take over a lot of commuting.
For what it's worth, this is only the case if you live near a station with overlap (or live in SF itself). I lived in West Oakland and had a train practically every 3 minutes.
Then I moved to Berkeley. Commute in is every 15 minutes, 50/50 I have to take a transfer on the way in. Commute home is a nightmare: BART has a carriage shortage so they try to shave off cars from longer routes (such as the Richmond line). This means things get really crowded on occasion, and there's a point where a car being crowded begets more crowding, as it takes longer and longer to offboard/onboard people and therefore each subsequent stop has more people to cram on. A majority of the time on my way home when I get a transfer (as there seems to be a Richmond train from SF only for an hour span at the peak of rush hour), the trains are almost an entire cycle late, meaning I've got another 15 minute wait to see two trains come through one after the other.
The point of this is: the problems you raise with busses are also with trains/BART. Some of them are even worse for BART: you have to steeply limit the service area because putting down tracks is stupendously expensive.
This is less a "problem with trains" than it is a "problem with rail in the United States", or maybe just with SF.
Melbourne Metro, for example, has roughly the same amount of weekday riders as BART, 415,000 vs 446,000 respectively, with nearly 5 times(!) the amount of track, at 540mi vs 112mi.
This is, of course, excluding light rail services -- another 150mi of track in Melbourne with some 1,700 stops, and non-Metro heavy-rail services to the other nearby commuter hub cities like Geelong and Ballarat.
Berlin is, from my knowledge, more similar to Melbourne than to SF in this respect.
But y'know what the funny thing is? Everyone in Melbourne hates the bus service. Makes me wonder about what they think about buses in Berlin.
The problem with BART is that it runs 4 out of 5 lines on the same single stretch of track through SF. So even with the most modern signaling, there will still be 1/4th the maximum train frequency after the individual lines split. The Pittsburg-Bay Point line counteracts this during rush hour with trains that come just into the city and then turn around, but it's still only an average of ~7 mins between trains. Compare with the Victoria line in London which I believe runs 36 trains an hour!
Had the same experience when I lived over there too. Was pretty surprised West Oakland was as cheap as it was relative to anywhere North Oakland/Berkeley considering how much more access it had to everything.
The price of places under ten minutes from a BART station further out were sharply different to further out in the same area too. You could shave some time off by cycling but then you're the guy with a bike on the packed train.
Seem to recall 20+ minute waits for trains in Berkeley on weekends too (it'd be handy if they let you know this somewhere in the station before you pass the ticket point, but that's another thing), with a transfer required to get into SF?
SF is far from the worst area for public transport, but it'd never come near my best.
> BART has a carriage shortage so they try to shave off cars from longer routes
This is just crazy. Compared to the cost of new track some additional rolling stock should be almost insignificant. You couldn't come up with a better example of penny wise, pound foolish...
The rolling stock is actually a very significant cost of the system. In DC, they spent $886 million for Kawasaki to build them 428 railcars in one contract, which comes out to just over $2M per car (and a single train normally has 8 cars). According to Wikipedia, they now have 500 of these 7000-series cars in service; that's $1B.
First of all, those objections seem very context-specific. Buses work great already in plenty of cities without those problems. It seems like you’re looking for ‘how can buses be better than BART for my specific commute’, not ‘how can buses be better than driving or ridesharing for everybody who doesn’t have another mass transit option already?’
But to your point about stopping in every block: optimizing a bus route along a major avenue in a block-based US city seems like a similar problem to elevator scheduling in a high-rise building. I wonder if similar models to ‘sky lobbies’ and destination selection could be applied to get people more efficiently to their destination.
>"You can't make it stop at every block in order to make it take less walking to get to the bus stop"
You absolutely can, that's how the main bus lines in Copenhagen operate. Some of the stops are only a couple hundred meters apart, and during most of the day the schedule isn't fixed, it's just signed as 5-8 minutes between buses.
The main A and C bus lines criss-cross the city on bigger streets, usually parallel to each other for a lot of stretches, and occasionally on the exact same streets for short stretches. In addition to this, there are a number of lines that service smaller streets on longer intervals, with the stops a little further apart.
Of course we also have the S-trains and Metro, which step up in distance and grid size, compared to the A and C bus lines.
It works. I rarely have to walk more than a couple of minutes from a bus/train stop to get anywhere I'm going in the Copenhagen area.
And this is with stops basically every other block, buses every 5 minutes in rush hour on the main lines, occasionally parallel lines (where it makes sense) and no artificial limits on the service area.
Plenty of cities (Chicago, in my experience) have parallel lines 1-2 blocks apart.
For the record, I take the transbay bus to work most days, and it's a similar deal (plus I can almost always get a seat). Slightly less convenient with the schedule (every 10 minutes instead of 5)
It sounds like you're talking about downtown Chicago. There are a lot of bus lines that converge in that area, along with every single train except the yellow line, because the city is hyper-centralized so that's where everyone's commuting to.
In most the city, buses just run along the major thoroughfares, which typically means they are spaced 4-ish city blocks apart.
That's exactly the problem, buses can never match that. They can't have too short a route or too fast an interval, it doesn't work when you have huge variance due to personal cars, lights etc. impeding them.
Citylab's push here is in response to South American cities doing exactly that with great success[1]. Initially developed in Curitaba, follow on initiatives in Bogota have been very popular[2]. Basically, treat the bus like a metro. Give them their own lane. Make sure the platform is level with pickup. And you've got a massively cost down subway system. For an inspiring documentary you should check this[3] video out.
I recall reading an article a few years ago about Bogotá's transit, where the mayor (I think) said something very interesting. Among other arguments, he made a moral argument about the dedicated lane for transit. Roughly, there's 40 or 50 people on the bus, vs 1-2 per car, and you get, what, 3? 2? cars in the space of a bus. He found it completely unjustified to let the convenience of those 45 people be outweighed by the convenience of the 4 or 5.
I'm not sure that it stacks up perfectly logically, but it's an interesting way to think about it. And it looks like they have a pretty good system.
> Roughly, there's 40 or 50 people on the bus, vs 1-2 per car, and you get, what, 3? 2? cars in the space of a bus. He found it completely unjustified to let the convenience of those 45 people be outweighed by the convenience of the 4 or 5.
That's only true if the lane is as full of busses as other lanes are full of cars.
> That's only true if the lane is as full of busses as other lanes are full of cars.
1) The density of people per unit road area is about 10 times greater than with cars. A bus lane only needs to carry 1/10th as many buses to come out ahead (and a bus occupies considerably less space than 10 cars).
2) The relevant metric isn't people (or vehicles) per unit road area. It's the lane's flux of people per unit time. On that metric, an free flowing carpool lane will beat a congested regular lane every time (even when the carpool lane appears empty to the layman).
You can do your own back of the napkin calculation, but vehicles have to packed roughly 5 to 10 times denser in the congested lane to achieve the same vehicle flux per unit time.
That means a free flowing carpool lane (>=2 people) only needs about 1/20th the vehicle density per unit area to come out ahead of a congested single-occupancy lane.
It's just no contest with an uncongested bus lane: Buses will win on a people per time metric with only 1% as many vehicles per unit area. To visualize this, 100 closely packed cars (remember, they are not touching!) is about three city blocks long.
Yes, but if the bus is faster, more people will take the bus and so you will add more buses. In places like Quito with dedicated busways, there are buses basically non-stop. Buses can follow one another better than trains too because you don't have the signal system reducing headways.
For the record I'm a huge fan of busses and highly skeptical of trains. I'm sure that there are some places in the world where the Mayor's statement is true. Those places are rare.
In a downtown corridor during rush hour? Sure - I could totally see a lane full of just busses. However, most dedicated bus lanes won't be that busy. This isn't to say that there aren't good reasons for dedicated lanes - just that the reason the Mayor gave was at best hyperbolic and at worst maliciously deceptive.
> more people will take the bus and so you will add more buses.
In a well run business that would be true.
In the first world, most busses are heavily subsidized. This means that the more busses that are run, the more money the system loses (not strictly true, but roughly).
What you say should make sense, but it doesn't for important institutional reasons.
> In the first world, most busses are heavily subsidized. This means that the more busses that are run, the more money the system loses (not strictly true, but roughly).
So what? Transportation infrastructure for single-passenger vehicles is also subsidized in the US (the gas tax doesn't come anywhere near paying for roads, for instance), and it gets more expensive with more users too (in the form of increased wear on roads requiring more maintenance, or the need to expand roads to accommodate increasing use). But we do it anyway, because transportation infrastructure is a public good. I've never understood the inclination that public transit should stand alone among transportation modes and operate without subsidy.
>In a downtown corridor during rush hour? Sure - I could totally see a lane full of just busses. However, most dedicated bus lanes won't be that busy.
These are precisely the places where dedicated bus lanes and BRTs exist. BRT is called bus rapid transit because there is a planned frequency for these buses - therefore a need for consistent, uninterrupted rights of way.
>In a well run business that would be true.
In the first world, most busses are heavily subsidized.
Only because highly popular bus routes (and train routes) subsidize unpopular routes that exist for the public good and not to make money - neighborhood routes, night buses, etc. In the US, routes are highly subsidized because of a lack of density compared to Europe or Asia.
> In the first world, most busses are heavily subsidized.
Public transport makes it possible for more people to use the roads it makes it faster for people that have to go by car. So what you call subsidized is actually a way to activate capital sunk into expensive infrastructure and make it more effective.
> I'm sure that there are some places in the world where the Mayor's statement is true.
In Europe this is predominantly true (maybe not the UK outside London). There's still massive traffic jam problems but buses certainly during rush hour are full and special lanes are provided for them.
I've seen this same argument about buses only taking up the amount of space as a handful of cars, but I think it's based on an incorrect assumption. Namely, it's only true in a static picture. If you look at the dynamic picture, cars and buses in motion, then buses take up a lot more "room." They're slow to accelerate. They stop frequently. They block traffic when they stop. They're huge, and people can't get around them as easily as they can around cars. They impede traffic in all sorts of ways that cars don't. Buses might still have an advantage, but I don't think it's as large as is claimed. (Separate peeve: it's always assumed that buses are nearly full and cars are nearly empty.)
What you bring up stems from two things, infrastructure needs to prioritize busses, and second comparing busses to single car does not work. Since we are obviously talking about rush hour. At a traffic light in a busy intersection where I live there will be max 15 people in cars, 50-400 people in busses, and about 100 more on the streets walking and bicycling. I think 5 times more people in busses is a pretty good advantage.
I don't disagree (infrastructure should prioritize, or at least accommodate) buses when they're used. But it currently does not. Also, your example seems to be from a dense, urban center, but of course not everyone on those buses started in the city; a lot of them had to get there from somewhere else, and there the advantage is less pronounced the case for dedicated infrastructure harder to make.
I would love to see separate lanes for buses and cars. In Hawaii, where I live, there was a proposal for an elevated, bus-only, viaduct that ran the length of a major freeway. This would have enabled buses that run as frequently as trains and are unimpeded by rush hour traffic, and are cheaply and easily scalable with demand. Instead we're getting an elevated train that has disrupted traffic already with its construction, that has blown through its budget with no end in sight, will be stopping short of its original end point due to cost overruns, and is only expected to reduce parallel car traffic by 2%. Ugh.
They did this where I lived in Sweden too. The city wanted a tram line, but couldn't afford it. So they built a bus line. Bought massive triple bendy busses. Even used tram signals at the stoplights. Worked great.
Same in Finland(Helsinki) as well. Its called the joker line, huge busses that go every 8-10 mins. Now theyre actually trying to replace it with a fast tram.
Similar in the Netherlands, plans for a tram line, turned out to be to expensive, got changed to bus line with dedicated lanes. Even though it was built to make an eventual transition to a tram easy, this seems to work better in hindsight than dedicated trams.
Don't know what it was like years ago, but these days it's a mess and universally hated by Quito residents. The Trolley bus is dirty, crowded and a haven for pickpockets and thugs. The schedule is completely irregular now and it often stops running early for no reason.
Bummer. Maybe it was bad 20 years ago, but from the perspective of a kid who could get all around the city at age 15, it was fantastic. We did also hitchhike by jumping into the backs of trucks too... those were the days.
No I just spent 6 months working there and studying Spanish. Loved the city in general but it does have very real safety issues. I was never robbed myself (I'm 6'3" and scary looking) but every single other person I know who has lived or visited Quito was robbed at least once, usually multiple times. My local Spanish teacher was robbed 4 times in 3 months.
>They can't have too short a route or too fast an interval, it doesn't work when you have huge variance due to personal cars, lights etc. impeding them.
Lots of cities have found ways around this. Express buses with dedicated rights-of-way in Pittsburgh, for example. Or Chicago's 146 Outer Drive Express and J14 Jeffrey Jump routes are also good examples.
Too often the biggest problem with buses getting around isn't the buses. It's cars driving, turning, or parking illegally in bus lanes.
There will never be enough parking. You might also consider that the 20 m^2 land for a parking spot in prime locations like San Francisco are easily worth >$1000 per month and using that for free parking is part of why your rent is so high.
Close...the infrastructure to focus on here is still transit; if the buses were sufficient, the parking issue would be solved simply by fewer cars on the road.
If there were no free surface parking every road would magically gain two lanes, e.g. for buses or bikes. Or you could use all that space to increase density. See for example how it works in Tokyo.
Bus rapid transit. Dedicated, sometimes separated, lanes. High throughput stations. Carefully designed light systems, etc. If you do it right you can get a system that is nearly as good as a subway at a fraction of the cost. A lot of cities in South America, for example, have proven that it works already.
I went to school in Ottawa, where they have a 'transitway' [1] that is basically a series of bus-only highways, though there are some parts that are only dedicated lanes.
I wasn't a daily user but it was decent enough - along the route I used to use most often a bus came every 3 minutes during peak times.
They can do bus-only lanes though. Ideally, also toll the roads high enough so that the traffic is never backed up. (And, even more ideally, have no toll when when the road is below capacity.)
Buses works great as a middle-class transit option in lots of places. When I lived in Barcelona I used the city bus as my primary method of short- to medium-distance transit, as did, I believe, the majority of middle-class residents.
Taking the bus doesn't have to be the terrible experience it has devolved into in many US metro areas.
You have a point, even though I disagree in the long view. It is a bit of a chicken-and-egg situation. Making transit really good is very hard until cars are deprioritized. And until transit is really good, deprioritizing cars screws a lot of people over.
In my city, a new project was started that did precisely that. Dedicated bus lanes for almost the entire route, bus stops spaced about a mile a part, and 8 busses per hour.
It received massive criticism when it was constructed, as there already was a functioning network and the project was quite expensive compared to other bus lanes. It did become a massive succes, as its a viable alternative to the train for commuters due to the shorter travel time.
They are in the process of replacing the diesel busses with electric busses, first on the short routes and in 2020 on the longer routes.
Buses can be adapted far easier and cheaper to changes in transportation needs that any light rail system. even if you toss in road maintenance as its already performed for cars it is still cheaper to operate than rail. buses can also reroute easily if there are problems with their route and a broken down bus or fire doesn't block the route for other buses.
plus BART has about ten billion dollars in maintenance backlog.
the only rail system in the US which works well is NYC and we get horror stories about it regardless
This does not work if you want to speed up routes in rush hour, because people will demand stops at every block. Usually people are not alturistic enough to get off in bulk if they have the option to stop everywhere. Busses work great even in low density environment if you build infrastructure that is good for busses. Bicycles + buss makes it posible to serve quite large areas fast and simple.
> This does not work if you want to speed up routes in rush hour, because people will demand stops at every block.
Actually, this is exactly how Marshrutkas work, at least in Sofia/Bulgaria. They do not stop on "every block", they stop only if they are not full enough or someone wants to get off. But normaly, this persons tell their destination early enough to the driver, so he or she can plan better the few stops. Nowadays most people prefer the Metro over the Marshrutkas though.
London is doing a decent job of covering your points:
> You can't make it stop at every block in order to make it take less walking to get to the bus stop.
How big's a block? London often has bus stops a few hundred metres apart even out in the sticks.
> Similarly, you can't have parallel lines a block apart.
London often has multiple routes serving the same area in addition to parallel routes nearby.
> You have to limit the service area so that you can afford to run busses every 5-10 minutes.
My most frequent bus (53) goes from Central London to Plumstead Station (~11 miles) on a 6-12 minute schedule during the day. That's not uncommon for London.
> And you'll want to improve streets so that busses can run down the center and pick up passengers from protected islands
Slowly happening over London - couple of stops near me have had their inset bus stop bays pushed out into the road to block traffic when the bus is stopped.
Like, I take the BART every day into work. It's fantastic. I walk for five to ten minutes to get to the station, a train comes every 5 minutes so I don't have to worry about the train schedule, then I walk another five to ten minutes to get to the office. If the bus experience could match that, it'd take over a lot of commuting.