Something you notice in Hollywood films as a non-US (Australian) watcher - a character hitting their absolute low point in the story is usually the scene where they are forced to catch a public bus. It's a bit odd watching from a city where catching a bus is a pretty mainstream transport option. Maybe positive representation for bus transit on screen needs to get on the progressive agenda #OscarsSoCar
I would like to add that Google Maps and Apple Maps (and others) have made navigating complicated bus lines much easier for the uninitiated. There are nice synergies to mass transit and mobile devices - they make the whole experience more pleasant (podcasts, music, audio books; noise canceling headphones are getting very cheap). While all of this is true of driving, driving is cognitively demanding and, I feel, minimizes enjoyment of whatever preferred media you consume. Using transit leaves me in a much more rested and thoughtful state than driving ever did.
Terrific article as always, but it seems to ignore the tension I suspect is always there between efficiency and equity (other than the discussion of coverage vs density). It seems that in many cases the express goal of politicians is to have some, any transit, however bad, available to people who cannot afford cars or two cars per family. Matt’s analysis seems to omit that constraint.
Like in many cases, it’s not that decision makers don’t understand what’s efficient. It’s just not what matters to them. I think they are ok with buses’ reputation as “poor mans transit” as long as it allows them to provide some service to the disadvantaged population. After all, if waiting 30 minutes is your only option, you’d rather have that than nothing at all.
This also can explain to some degree why stop-spacing is so close. Many people can walk a quarter mile to their endpoint, but for someone like my mom, who is overweight and has asthma and until recently was recovering from chemo, spacing stops that far apart would literally be prohibitive. Having spent many, many years on buses, I think health constraints are closer to a norm than an exception among lower-class riders.
Depending on the number of people with these sorts of mobility constraints, it seems that the city could provide better service by giving them free credits for paratransit, or Lyft, that can pick up and drop off door to door, while making the big vehicles do what they do best, moving large numbers of people from one hub to another.
Paratransit, in my city, is of absolutely terrible quality. Like impossible to book, and hours and hours late. Better (and likely cheaper) for the city to reimburse eligible people for wheelchair-accessible cabs.
Jarrett Walker points out often on his blog that Lyft/Uber haven't come up with anything that fundamentally does better than transit at the ridership job - but it does seem to be a major improvement on pre-app dispatched single-person rides, like taxis and paratransit.
Yeah, I have experienced the issue with wheelchair users boarding buses and said to myself that maybe people with disabilities should get free Lyft/Uber service at the city's expense. Wouldn't transit users without disabilities get a significantly better experience that would justify the extra tax money--to say nothing of the benefit to people who *do* have disabilities?
The concern with this might be that by unbundling service like that, planners/politicians would be able to starve the more-expensive-per-rider paratransit service of funding in order to improve service/win the favor of the median rider of big-vehicle few-stop services, similar to concerns around private/charter schools diverting voter attention from public schools.
In this case though the idea would be that the "coverage" service and the "ridership" service would have separate pots of money - they are currently bundled, which makes it easy to rob one to do the other, but Jarrett Walker advocates separating them institutionally, even if they are run by a single agency.
My question about this logic has always been that this makes sense if you happen to live close to the street where the bus line is (i.e. thinking about it in linear space where your main distance is between stops), but what about all the people who live perpendicular to that (i.e. two dimensional space where distance to the line is the big factor). Those people have to travel pretty far either way and I'm curious if distance between stops is really the deciding factor.
Was hoping someone would make this comment! I grew up in a city (Buffalo) with a pretty lackluster transit system, but pretty thorough coverage. My parents couldn't afford a car, and neither could many of our neighbors. I had to take really slow busses wherever I needed to go, including, ie, school, and trips that would be a 20 minute drive would routinely take 1h30m end-to-end. But it was always the case that if I was willing to show up 30 mins early to something and spend up to 2hrs getting there, then I could make it. And since I didn't have any other choice, that's what I did. Most of the busses I took were empty, but I needed them. That's even more true of the people who were travelling on these god awful lines for work, whereas, as a teenager, I was usually sucking it up to hang out with friends.
Seconding this comment. Transit planners understand the tradeoffs Matt has described, and many of them probably personally agree with his preference for ridership vs coverage and greater stop spacing. But the bus networks we actually have reflect the priorities of the elected officials on the governing boards of our transit agencies. They also reflect things like path dependence and the challenges of an FTA-mandated Title VI service and fare equity analysis, which can be a barrier to reallocation of existing resources and operating "like a business." Not to say that means we should just throw our hands up! But the constraints are real.
Yeah I was thinking something like this too. I know very little about transit policy, but the type of thing that Matt’s talking about sounds like poorer people will often end up getting screwed over? Like I know that, when there are just a few subway lines, the apartments get more expensive near the lines because of the added convenience of being so close to a transportation source.
So if you consolidate lines, eventually I would think the prices are going to go up along where the new lines are, making it harder for poorer people to afford to live closer to the busses and making them somewhat more difficult to access. Like you might end up getting a higher number of people on busses, but then making it much harder for a sizable population of people who would have ridden different lines before to actually use the busses, and that that group of people is disproportionately likely to be lower-income (after some time is given for prices to equilibrate).
Matt just shrugs this off with “they’ll have bigger yards” but idk that just doesn’t seem like a satisfying answer if you were relying on a bus to get to work that’s now moved 25 minutes away from your home. But maybe you can balance it so most people don’t get harmed too much and you still preserve the efficiency?
This is why in the long run this has to be paired with land use reform. Instead of expensive five-story apartment buildings next to the transit lines and fancy mansions behind them, you need ten or twenty story apartment buildings next to the transit lines and five-story apartment buildings a couple blocks away. And perhaps more relevantly, if there are currently some nice long straight roads leading to downtown that currently don't have much bus service because they go through low-density areas, you can put frequent transit routes along those roads and upzone them simultaneously to create more opportunities for low-income people to find something near a transit route.
The housing next to the transit routes would be new high-density construction and very expensive. It would require deep subsidies to provide even a small amount of affordable housing in those buildings. It might be more effective to provide middling-frequency feeder lines so poor people can get to the high-frequency transit trunk lines and so get around without a car.
Only the newest construction along those routes would have those characteristics. But if new construction is able to come regularly, then the existing buildings in those locations would provide the affordable housing, while the newest ones would be expensive - as opposed to the case when there are no new buildings, and all the existing ones are bid up to the prices that people would have paid for new buildings in those same places.
Paratransit is an interesting wrinkle here. 10 years of riding MUNI in San Francisco has impressed on me the significant tradeoff between inclusivity and efficiency when there is a large population of riders in wheelchairs.
At this point I’ve seen literally thousands of instances where a bus full of otherwise progressive and equity-minded types will audibly groan as we approach a bus stop where a person in a wheelchair is waiting, knowing that the process of deploying the ramp, onboarding the rider, and securing them in the specialized seatbelt can make the difference between being on time for your next bus transfer or clock-in time at an hourly wage job, and being late. Doubly so if the rider is also getting off before your stop, where the seatbelt and ramp process plays out again in reverse.
One might reasonably suggest increasing the number/frequency of buses, but this does nothing to alleviate the social tension (people on the buses with wheelchair riders will still be late unless they shuffle off the bus and onto the one behind them when it catches up) and in the case of buses like the 22-Fillmore, which are electrically powered by tethers on the roof connected to overhead power lines, the faster buses can’t pass the slower ones, and you end up with the very normal occurrence of two buses on the same route, one right behind the other, the first one full and the second one mostly empty, and everyone running late.
The social ideal here would obviously be having an efficient system where everyone rides together and doesn’t exacerbate the existing stigma of “otherness” that comes with disability/reduced mobility, but at this point I’m convinced it’s just not feasible, at least in SF, and some sort of on-demand paratransit or outright Uber/Lyft/Taxi subsidization for wheelchair riders is the better way to go.
(And don’t get me started on the impact of endless elevator outages on wheelchair-bound riders of underground MUNI and BART. It’s appalling)
This can be solved with a good dedicated level boarding stop using a Kassel Kerb. SF because of insane car owner NIMBYism in general has abysmal bus stop accessibility. https://en.wikipedia.org/wiki/Kassel_kerb
The new rapid bus line in Oakland (Tempo) has this, along with proof of payment, and it makes boarding so much faster. Probably 25-30 seconds on a normal bus vs 10 seconds or less on Tempo
It’s interesting that this is a prime case of planning by lawsuit. Disability activists argued that equity required that everything had to be accessible on principle, courts agreed with them on that, and so transit agencies were forced into massive investments on a schedule to meet this mandate, displacing other capital investment that might have been of more importance to the general public. This is how the law works, in absolutes, independent of planning or financial tradeoffs or, indeed, of anyone’s rights except those of the disadvantaged population involved in the lawsuit. The general riding public has no standing, so doesn’t get considered.
Now, sure, those elevators and bus ramps are useful to people besides disabled people, so the investment is not so narrow. But if the elevator isn’t working, disabled people can’t use the stairs, and given feasible limits on serviceability of the equipment in this high-stress environment, those people will not infrequently get trapped in the system. Resulting in more lawsuits in which judges will declare that the agency has an absolute obligation to spend unlimited funds to succeed at the impossible.
I wonder what value could have been obtained had those billions been spent on working-class jobs (taxi drivers) and wheelchair taxis instead of on massive construction and on equipment that has to be expensively maintained and monitored, and regularly replaced.
Partial solutions: higher bus stops for at-level boarding, "kneeling" buses that get closer to curb level, proof-of-payment systems that allow all-door boarding so that able-bodied riders can get in one door while the disabled rider uses another. As far as the audible groaning, it's public transit and disabled people are part of the public too. Do people huff and puff about the inconvenience of giving up a seat to a pregnant woman or elderly person?
SF does have kneeling buses, and this is much appreciated by older and less mobile folks, but doesn’t help much with wheelchair riders. There’s also ramped sections in the road medians for many of the streetcar line stops, which allows for at-level boarding of wheelchairs, but to my knowledge all of the buses have steps at the entrances so the deplorable ramp and lift is the only option for them. POP is also the default for all buses and above ground streetcars, which does help.
As for the groaning, part of the issue is just the sheer number of wheelchair riders on routes where homeless populations are concentrated. You just have to experience it to understand how frequently this happens. At a “normal” level of wheelchair ridership (relative to overall population stats) most people are gracious about the occasional inconvenience and happy to support a more inclusive public good, but when it’s every fourth stop, and they’re only riding for a few stops at a time before getting off, then anyone who can afford Uber or Lyft instead is going to migrate to that instead, which results in less funding for the system, which means fewer bus lines and lower frequency, which exacerbates the problem, and the public funding doom loop begins.
To be blunt, people don’t huff and puff about pregnant women or the elderly because those groups don’t routinely make the entire bus late, and they’re also not routinely drunk at midday, soaked with their own urine, or screaming at things that don’t exist. SF is a somewhat unique problem in this space.
Transit users regularly cast a calculating eye on the elderly and the pregnant. The goal is to determine, if only on the scale of one’s own conscience, whether the elderly/pregnant person is spry enough to obviate any need to give up a seat.
I will admit I have some scruples about how to treat older people who dye their hair and dress in a style typical of younger people. Is giving up my seat an accidental insult? Would they prefer if I sat confidently and pretended to be fooled by their artifice?
I get up anyway since I don’t feel qualified to parse the ethics in that situation, but I’ve gotten miffed eyebrow raises for it enough times to make me question myself. It’s like the opposite of asking an obviously middle aged woman to show her ID when buying alcohol.
*Correction, not all buses have steps at the entrances (I misremembered since I’m usually boarding at the middle doors, and it’s been months since I’ve been on any of the buses due to covid) so there probably is room for improvement of at-level boarding procedures. Older buses have an extend-and-lift hydraulic system for wheelchairs that takes a long time, while the newer models have a fold out ramp that works decently, but in either case the total boarding time for a single passenger can be 2-3 minutes, and the same on the way out. Doesn’t sound like much, but when there are multiple wheelchairs getting on and off during your trip and missing your connection means waiting 15+ minutes for the next one, it adds up.
Is a seatbelt strictly required? I thought buses normally had a kind of rear facing padded back thing for wheelchairs to rest against, then they just apply their wheelbrake?
I’ve seen both (there is a wide range of bus types and ages in the system) and I’m not sure about the legal requirements—in practice it seems to be at the rider’s discretion, but in cases where the rider is unresponsive or seems confused, the bus drivers are usually pretty diligent about getting them fully strapped in.
Anecdotally, it seems that around big sporting events (World Series, Super Bowl, etc) MUNI puts a lot of older buses into service since there’s a not insignificant chance one might get torched.
Yesterday's post was all dry, technocratic wonkery about race, murder and cops. A big yawn, and my eyes glaze over.
So, in order to get people riled up and yelling at each other in the comments section, Yglesias throws us some juicy red meat transit-blogging. That gets the tempers flaring!
Try to play nice, everyone. I mean, it's a gut-punch, reading about European routing protocols and square footage of lane capacity and all. It's no wonder that people have a hard time sticking to facts and reasoned argument when the subject-matter is so inflammatory.
One major disadvantage of bus routes is that there is a very steep learning curve to learning route maps, bus frequency, and transfer mechanisms. Some systems (as well as Google Maps) now have trip planner apps that help with this but the process of just getting from A to B is still needlessly opaque. This system works well for people such as daily commuters who are willing to invest the effort but useless for tourists, out of town business people, and people with random destinations.
Some cities now have Circulator type buses which are frequent and run very busy closed loop circuits. They are often differently themed from commuter buses and may have different fee structures down to and including free. So as I read this article, buses should be more frequent, run fixed routes, and have fewer stops. Street cars. Congratulations, we have invented street cars.
No one is claiming the bus is anything novel and exciting. It's the same thing that they already realized in the 1880s would be valuable - a vehicle for multiple people to ride that goes down the busy street to all the major destinations, and is a few blocks away from most people. The advantage the bus has over the street car is it does it more cheaply, and with the ability to get around a stalled car or collision.
The streetcars versus buses debate goes back to the 1950s and 60s with many of these same tradeoffs argued. Buses were seen as more convenient and flexible and had the support of major bus manufacturers. And while the GM conspiracy has been debunked, favoring the car over privately run streetcar concessions killed this mode.
One of the reasons people liked buses better at the time was that they could pull over to the curb.
Original streetcar lines ran in the middle of the road - and stopped in the middle of the road. They didn't have any platforms or anything, people just got out and walked over.
This became a problem as car traffic increased.
Modern-built streetcar / light rail systems generally avoid this in favour of either running in the nearside lane, having side platforms out in the middle of the street, or having island platforms (most modern such vehicles have doors on both sides, whilst traditional streetcars generally only had curb-side doors like buses.)
But you can still see the old mode of running in systems that survived such as many of those in central and eastern Europe, and Toronto.
Running on the side of the street is actually terrible, and almost no new build systems anywhere outside the US do this, but yes platforms in the median of the street are the global standard. Got a lot easier once low floor trams became common, as you need a lot less "station" at your stop for that versus the high level step up models.
buses are better unless you're going to give the streetcar are well protected median right of way, and expect you need more capacity than an articulated bus can provide.
This may be true in a handful of US cities, but even Chicago has a bunch of routes that aren't filling a bus if they run every five minutes.
A standard part of a bus network redesign in the Jarrett Walker school (and he's pretty much the lead thinker on this) is creating a new bus network map that clarifies which routes are frequent. He also encourages bus networks to be designed as grids and transfers to be free to increase mobility. It's interesting stuff.
The main shortcoming of streetcars is that they're short.
modern streetcars actually can be significantly longer than buses. Their draw back is that unless you give them a privileged right of way, they're very easy to block.
If daily commuters are the bulk of riders/travelers, especially during peak hours, then getting them into efficient modes of transportation frees up road/rail space for other travelers. More commuters in buses means more cabs/cheaper rideshare/less traffic for tourists/etc.
Regarding fares, I’ve always appreciated the approach taken by German and Austrian cities I’ve lived in: everybody is required to buy a time-stamped ticket for a ride, but there is actually no mechanism to check whether you have done so. Fare enforcement is handled completely by “sting” operations, where suddenly five transit employees appear in your bus/train car/subway car, asking to see tickets. No ticket? Instant fine, of (at the time) 65 Euro. Most buy tickets, but many opted to chance it 100% of the time, and they thought they came out ahead, financially!
I don’t know if that would work in the States, but it’s my understanding that fare collection contributes a minor amount of income to transit operations anyway. I’m curious the extent to which fare collection or enforcement mechanisms are costly, and if there is ever a scenario in which we can agree that public transit is a public good that doesn’t need to run an operating profit, or operate like a business, to be a success. Its main function is to remove cars from roadways, which in theory is helpful to everybody.
A difficulty with POP (which I agree is better) is in America progressives have now gotten very skittish about anything that smacks of calling for more policing.
Alon Levy makes a pretty good case here that you should treat POP inspections as more like a loss prevention measure--the purpose is just to get your losses down rather than totally eliminate fare-dodging. https://pedestrianobservations.com/2019/11/13/fare-evasion/
Not sure this would work to sell POP inspectors to progressives though.
About those instant fines: what happens if the fined person refuses to pay the fine? Do they get arrested? Do they get whatever the local equivalent of a desk-appearance ticket?
So - we're going to the european airport from our hotel, and decide for reasons of virtue and cultural experience to take the Bus. We leave a solid 1.5 hours early just to make sure we have enough time. We get on the bus along with a bunch of people, and turns out two of these people - father and son - bought just one ticket between them. They get their tickets checked (not sure why their tickets were checked, ours weren't, might have something to do with their middle eastern origins). The dad gets asked to pay up a fine. Dad says no, just let me buy a ticket. Bus driver - nope, got to pay the fine.
So far so normal, if slightly racist seeming but what do I (brown American tourist who's feeling slightly uncomfortable at this point) know? Other people on the bus seem inexplicably upset with them. The next thing was amazing and might explain the upset other people - the driver actually made everyone other than the dad and son get off the bus. Called a replacement bus, which got there 20 min later (this after 20 min of arguing with ticketless dad on the bus, 40 min wasted time at this point).
Oh the replacement bus does not seem to be an express bus to the airport - it is going to the airport, but stopping for locals... When it finally gets to the airport, we have barely any time. Then, the final bit of excitement - (unlike the original bus) it actually doesn't even go to our terminal! It just stops for good at the first stop that could conceivably be called getting us to the airport. We have to take a shuttle to get out our terminal.
Typically I don't think these are enforced as criminal citations. So I suspect they might suspend or ban you from riding the bus/train but likely no arrest.
One unconsidered thing about free transit is that it would effectively transform the transit system legally into a public street and there would be no capacity to control misbehavior that would drive away riders but that the courts would consider to be civil rights. It would make the system unmanageable, and for this reason both the MTA and the NYPD have opposed it.
In Budapest we were told by our tour director that the subway tickets were rarely enforced. Being in a rush as well as cheap students we didn't buy tickets only to be stopped by very sketchy enforcement people in civilian clothes but with badges and ticket books. Needless to say the 65 euro fines payable on the spot cancelled the planned evening of drinking and we returned to our hotel poorer and upset.
In hindsight they were clearly targeting clueless tourists which we very apparently were. Unless it was all a scam which is also a possibility.
When I first visited Amsterdam, a friend living there told me they never enforced the ticket policy. Sure enough - on my first day, a couple of us got stopped by some very imposing Dutch police. I played the true dumb American, saying in my home, buses were free (it was actually true!). Amazingly they let us off without paying.
Proof-of-payment systems are generally cheaper, but they do have to be enforced correctly. The German systems work pretty well because the spot checks are legit random--the fare police could show up at any station at any time. This tends not to be the case in the US.
I had the unusual "privilege" of living in Camden, NJ while I was in law school and for about a year after. Camden is the southern end of the River Line, a light rail that runs up to Trenton, and one of the relatively few proof-of-payment transit systems in the US. I used it fairly often to visit friends in Central/North Jersey and New York via the NEC. had the even more unusual privilege of actually commuting on it, as I had an internship in Trenton my 2L summer and worked in Trenton my last six months living in Camden. Even after I moved away, I continued to use the River Line as I was living in New Brunswick and found it useful to visit friends still living in South Jersey. I didn't really stop using it until I moved to Philadelphia.
The upshot of all this is that I tended to ride the whole length (Camden-Trenton): either I was going from Camden to my job in Trenton (and back), or I was using the RiverLine to connect to the NJT NEC Line to/from points north. I gotta say, fare enforcement was pretty un-German. On the one hand, when I was commuting, I would catch the same train every day there and back. Both of these were peak trains which tended to be full at least at some point. And pretty much every time, the fare enforcement officers would show up at the same set of stations in each direction.
On the other hand, my non-commute trips--which tended to be off-peak--only encountered fare enforcement a handful of times over what must have been dozens if not hundreds of trips. When they did appear, it was typically at the same stations as during commute hours. I don't know if this actually helps NJT with its farebox, but I can say that a lot of people treated the River Line like it was free off-peak, and it kind of was. I didn't, but only because the fare was so low ($1.60 to go 30 miles? Yes, please.)
Proof-of-payment systems are pretty good, and they're especially good for buses where payment causes a lot of delays at each stop. A major snag is that in the American context, creating more opportunities for interaction with enforcement agents has a lot of downsides.
It seems to me that the best overall bus system would involve very cheap monthly cards -- $10 or something -- with a touchless interface.
In NYC, I've taken the bus thousands of times, and I'd say the average person takes like 5-10 seconds to get out their card and swipe it, which really adds up. It also makes for dynamic effects where since the bus goes slower, there are more passengers on any bus at any given time, and when the bus is at capacity it can take several minutes for everyone to get on and off. I know making the bus cheaper shifts ridership to it, so you would have to make the subway cheaper too. I think it's absolutely the kind of thing that would pay for itself in economic gains.
The system here in Boston has a tap card (among other things) and it's usually not a slowdown for people to tap on the way into the bus (though swiping a magstripe card in the same direction as you were walking might have been even faster). The big snag is the people who really, really don't want to load much money on the card, and are individually loading money onto the card each time they ride (the fare with a card is cheaper than just paying cash, so they have one). You could remove the discrepancy and just let people pay cash, or make a cash-value incentive for loading up a card ($25 of fares when you put on $20?), or have fare vending machines at every one of six thousand bus stops (ha! but they manage it in several European cities).
Yes, they add money on the spot (which the vehicle farebox lets them do). So they push a button, tap their card, put money in, push the button again, tap the card to load the card, then tap the card a third time to pay the fare with the money they just loaded.
As I understand it, most Central European and Eastern European systems use a system whereby you can buy a set of tickets in a strip but you must stamp or validate the one you are using before each journey or it is invalid.
Travelling with a non-validated ticket is the same as not having one at all, or else you could just buy a strip, and only "use" one on the rare occasions when a fare inspector was there.
I have to admit I instinctively dislike this system. It seems hostile to newcomers. I experienced this system in Vienna, and never would have realised if I'd not been told that merely buying a ticket did not mean I was good to go.
When I was traveling in Amsterdam the difficult part was remembering to swipe out on exit. If you didn't do it, the ticket wouldn't show as valid to the enforcement agent and it was going to be treated like you beat the fare. Fortunately people were kind in reminding us, and we didn't meet any Dutch enforcement agents.
Caltrain had an electronic system where you basically badge in at a station and then badge out as you leave. I usually had an easy enough time remembering to badge on, because you're kind of just standing around waiting for the train to come anyway, but when I got off, I was always thinking of my destination and frequently forgot to badge off, which sucked because I then got charged for the maximum possible fare.
Yeah I don't mind swipe out / in systems if combined with ticket barriers, but systems where you just have to remember seem unfriendly to me.
A lot of people seem anti ticket barriers but I think they're underrated as they make it hard to *accidentally* fare dodge, something people unfamiliar with a system may worry about, and thus be put off using it.
Here in Cleveland, we had a Bus Rapid Transit line that ran on proof-of-payment, but a judge ended up ruling that there was no non-racially-discriminatory way to execute what you're calling those "sting operations." (I rode that bus to work every day for a couple years, and I can confirm that the payment checks amounted to rounding up a bunch of black folks and pulling them off the bus to get ticketed for like $200).
After the ruling, they had to switch to everyone showing payment at the door, which crippled the line and made it much slower and much less convenient. I'm not even sure what to do other than cry.
That seems odd. If you required the inspectors to check everyone on the bus (and picked the buses and stops at random) that doesn't sound like it would be discriminatory
I went back to check what actually happened, and it turns out I was slightly wrong. The specific ruling was that checking for fare after the fact was an unlawful search without probable cause under the fourth amendment. So it was ruled specifically unconstitutional. The judge said that using non-law-enforcement personnel would eliminate the constitutional problem, but having police officers check was not allowed.
That seems like a very interesting case. I would not have guessed that having state actors, but not law enforcement, would be considered acceptable. (Also interesting that the article mentions a report on low police morale because they felt they were being used as revenue-raisers, which seems like a wider problem than just transport also).
And I suppose they were using police in the first place because it wasn't safe for non-police control agents to do it. Bus drivers in New York don't say anything about fare beaters now because drivers were being attacked so often that their union went successfully to bat on this issue.
It's true, a large minority of bus riders in New York just don't pay and drivers rarely hassle them about it. If you made the bus completely free, it would have much higher throughput, though it might slow it down for the average rider due to people using it in place of the subway.
I would guess that the court found that the people who used the buses disproportionately were coming from predominantly black areas, and so there was a "disparate impact" that meant that even if facially neutral, the policy was racially discriminatory since mostly black people ended up paying the fines. Something along those lines.
By that logic, police can't patrol high crime areas if high crime areas are predominantly black. That's pretty close to "abolish the police" logic and doesn't seem likely to stand up on appeal
Well, looks like my guess was off-base anyways. The "disparate impact" type of reasoning comes up in wide variety of areas though, and is quite controversial with the more conservative end of the political spectrum.
The spot enforcement method is used in New York City on Select Bus routes where the fare is paid on the street before boarding. In my observation there is still a lot of fare evasion, because the people inclined to beat the fare are also the people who have no intention of paying the fine, and there is no way to make them pay on the spot, not least because many of them have no money anyway.
Fare collection is a major cost. There is a huge associated infrastructure of turnstiles and bus fareboxes and fare media sales points and back-end processing, plus personnel costs of maintenance and enforcement. For the New York subway/bus system (back in the day at the MTA) if memory serves the operating portion alone was close to 30% of gross fare revenue and the capital contribution was billions of dollars. I don't think the needle has moved much in the current iteration.
Fare collection doesn't just cost in terms of all the equipment - it also costs in terms of travel time, since everyone who is currently on a bus has to wait 10-20 seconds for each boarding passenger at each stop, instead of just a couple seconds each. Speeding up the bus both helps the passengers already on the bus get where they're going faster, and (as Matt mentions in the original post) helps the bus complete its route faster so that the same number of buses can run the route 6 times an hour instead of 5 times an hour, decreasing everyone's wait time before boarding as well.
Very true, which is why off-board fare collection is an essential part of better-bus arrangements. And it definitely works from that perspective. When Select Bus was instituted on Second Avenue, the MTA did a very intensive enforcement to establish the baseline that you could actually have to pay a fine if you didn't pay the fare. It was labor-intensive, very unpopular and the politicians got involved, so the MTA backed off, and now people don't seem to care. Since this is the reality, I think you might as well make the buses free, especially if they would be effectively free through generous pass schemes and free transfers anyway. But keep the fare on in the subways, so you can kick out people who misbehave and have no evidence they paid the fare.
The Baltimore light rail system uses fare machines to purchase fares, and gives you a receipt. You are supposed to show the receipt if asked by the transit police that do periodic spot checks. Failure to produce proof of payment can lead to a fine or arrest. You can currently purchase an all day pass that covers you for a 24 hour period on all forms of public transportation. As of today's date, it's $4.40. $2.20 for seniors over 65 and disabled passengers. I've been asked to show proof of payment on more than one occasion. I remember in Berlin, Germany, the UBahn used the same fare machine system, and the receipts were time stamped, and good for a certain length of time. Here in DC, you have Metro, and faregates that use a SmartTrip card. Most of the fare jumpers are teenagers who just jump over the faregates and don't pay. Station Managers are not allowed to enforce the fare. Only Metro Transit Police have that authority. Since most of the kids ride for free anyway, because of school/transit fare programs, there is very little fare enforcement.
The Dallas Area Rapid Transit (DART) uses a system like that.
Even more, you can buy tickets on a mobile app that are good for 30 days and activate them when you board -- or just wait until you see somebody checking tickets and activate it before they get to you.
It's possible this has changed in the years since I rode the train (there's a pandemic on, after all), and I never took advantage of it, but the whole scheme is basically an honor system.
St. Louis does (or did) this on its light rail “metro link” except in certain circumstances where it’s too crowded to inspect (mainly after cardinals games). Always worked fine. They also had an out of towner tax by tripling the cost of tickets bought at the airport.
I'm sad to be the first to comment as I find this topic very compelling.
One substantive comment I have is on the max coverage issue. A problem with that is that the routes become confusing for most people that aren't frequent users. So you end up driving and paying the $10 to park out of fear that the bus will end up going to the wrong place.
Every time I find myself near a bus stop in any city I look at the bus stop signs and route maps and try to figure out the system. It is nearly always impenetrable and confusing on even where that bus is going, let alone how or where to transfer to a different lines.
Subways and streetcars get a bad rap for being expensive and inflexible but the ease of use and convenience are big advantages to them. For one thing, they are big and visible so there is a much greater psychological sense of permanence and reliability.
This reminds me of my experiment using the bus system in Reykjavik, which is great for such a small city. The maps and stops could be puzzled out, but try being on a bus to a place you’ve never seen, when you can’t understand the announcements and the names of the stops don’t fit on the signs!
Yea confusion about bus lines is a real problem. All the more reason to have fewer lines and fewer stops, but with more used buses with more frequent service.
For years I commuted via bus from my home in west LA to Santa Monica. And I hated it. Matt's is a great post about how some technical tweaks to bus systems could help improve it (on the margins at least) but unless there's a more systemic change, all these improvements will disappear into the void.
Frequency was not a problem -- during rush hour, buses came every ten minutes (or faster). This still didn't mitigate the frustration when you missed a bus by ten seconds: the ten minutes felt like thirty! Bigger problems were the variability. Three buses might stack up and all appear within five minutes, while at other times there might be twenty minutes between buses. This was a popular route (Blue Bus #7 for the cognoscenti) and was typically *very* crowded, as it served Santa Monica College and Santa Monica High School, so lots of standing and swaying, and intimate knowledge of strangers' bodies on a herky jerky bus which aimed for every pothole on the street.
But the worst thing that was even though you had given up commuting by car, you got to enjoy the benefits of rush hour traffic jams caused by such cars. You crawled. You sat at lights forever (and this even with the "extended greens" that buses were supposed to enjoy but couldn't because of too many cars in front of it). It got to the point that a mile or so from my destination I would just get out and walk the rest of the way.
And then the new rail line opened (the Expo Line). And that was heaven! A smooth ride, with a dependable ten minute transit from my home station (albeit a longer walk from my house) and Santa Monica. And I could usually sit down! Goodbye, bus. And (since the Expo line paralleled the bus route), it looks like everyone else agreed with me, as I noticed the buses now running mostly empty.
I understand that it's hard to get commuting professionals out of their cars (often I was the only one on the bus or the train). And I understand it's a chicken/egg problem: as long as people stay in cars, and light rail doesn't go everywhere you want to go (LA's big problem), you're never going to have the rapid, smooth and comfortable transit commute that might entice them to act otherwise.
LA has toyed with the idea of congestion pricing and the howls have filled the heavens. But that -- combined with serious funding of transit alternatives -- is the only way forward. It has to be carrot and stick. Given my neighbors' open-arms welcome of higher density in our SFH neighborhoods (yes: joke!), my optimism is in the basement.
Sounds to me like what your bus needed was a true dedicated bus lane. That can also cause howls, but probably less than congestion pricing, especially if the road was very multi-lane to begin with.
That's a great-in-theory type solution (and has been explored). But not only would it elicit the expected howls, it would also fail tests of social equity. For the relatively few people who take the bus and would profit from faster movement, the impact on the vastly more car drivers packed into now fewer lanes would be hard to justify. And the bus lane would be empty most of the time, given likely bus frequencies. Then, consider the amount of cheating by car drivers, the challenges of enforcement, and the likely breakdown of norms as people watch other, more selfish drivers pop into the empty bus lane.
Of course, if everyone were already on the bus, those negatives wouldn't apply. But then we wouldn't need dedicated bus lanes.
I’ve taken your bus as a tourist to get from Santa Monica to the LACMA area. It was very pleasant, more so than driving, but I can see how it would be infuriating for a regular commuter.
Agreed it's a hard problem. Though I think some of the cheating/enforcement issues are covered by my conception of a "true" dedicated bus lane: where necessary, put up a physical barrier (like a low, parking-bumper-sized piece of concrete or maybe a series of bollards, it doesn't need to be a full Jersey barrier). Of course, that would be somewhat more expensive than just paint and cause more howling. I'm also skeptical that reducing the number of available lanes will have a large impact on drivers' travel speed--when analyses are done, we're usually talking about very marginal differences (less than 5 mph speed differences or less than 5 minutes travel time differences), though it could be different in this particular case.
“you need to make some people mad for the sake of a better overall system”
This struck me as an insightful point. Does getting something (anything?) big done in our system require a Robert Moses or an equivalent like Baron Haussmann?
People like that did get big things done in a way that's hard to even imagine now, though they made mistakes. A good read about another bullheaded builder like that is Cadillac Desert, about Floyd Dominy who headed the Bureau of Reclamation for many years and built dams across the west.
I think that 10 minutes is really pushing it as far as "can just show up" timing on buses. That means the wait will add an average of 5 minutes to the trip. Additionally, buses have fairly mixed on-time performance. In Portland in 2019 the bus was on-time about 85% of the time, which means if you are a regular commuter you'd have an average of 1.5 late buses a week.
If you require someone to transfer once on a trip, they will hit a late bus (5+ minutes late) 3 times every week on average, and the wait will add an additional 10 minutes to the trip every day. There is also the risk of a much longer wait due to a bus totally not showing up (could add 10+ minutes to the trip).
This is quite a bit given the average commute in the U.S. is only 26 minutes. I think the frequency needed for people really to just show up without checking the schedule is probably about 7 minutes. The frequency needed for people to perform a bus -> bus transfer without worry is 5 minutes or less.
If you're at the level of frequency where people "show up and go" rather than paying attention to the timetable, then *late* buses don't matter, as long as the headway *between* buses stays the same.
That said, I definitely think that 6 or 7 minutes is probably better for this than 10 minutes. (I guess 7.5 minutes could also work, if you want 8 buses per hour.)
Unfortunately it typically doesn't work this way because if you have 10 minute spacing between the buses, one bus gets knocked off schedule by 2 minutes it will end up having to pick up more passengers (passengers who would have just missed that bus will now make it) slowing it down. This also speeds up the bus behind it because now you have fewer riders waiting because some were able to catch the bus before...
So the busses end up bunched together once one gets knocked off schedule by more than 2-3 minutes.
I've heard from Jarrett Walker's blog that good transit agencies try to notice when this is happening, and dispatch a spare bus to fill in that gap that got longer, and tell the one that is catching up to just dwell a few more seconds at each stop, so that they can manage headways. This does seem incredibly difficult though.
Not difficult at all, with GPS and cellular radio (which all buses in New York are equipped with). The same technology that gives riders the bus arrival times, also tells the bus command center about the bus spacing.
I think the difficulty is making sure you have a spare bus to do this, that it is based somewhere convenient, and that you understand the speed of traffic everywhere in the city well enough to get that spare bus to some stop in between the two buses where the gap has developed.
I think that one reason why bus systems emphasize coverage over ridership in many U.S. cities today is that they want to serve disadvantage communities, or at least not further disadvantage them. A bus line that went from the densest residential area to the densest commercial area is probably going from one rich part of town to another.
Furthermore, given the unique setup of American bus systems, the ridership / coverage tradeoff doesn't look like a tradeoff at first. The fact that buses are priced lower than trains because they are "worse" makes them the default option for low-income folks who can't afford cars. The infrequent service also makes them a default for people who have a low opportunity cost of time, i.e. are also more likely to be low-income.
Therefore, you get into a situation where transit planners say "We can't relocate service from a low-income region to a high-income one. The low-income folks both really need the bus AND really use it, while the high-income ones will just drive either way."
The response would be "If we invested in frequent bus service for high-income folks, we could build self-sustaining ridership and gain the rich folks' support for the overall bus system." Matt has said before that a service that is only for poor people can become a poor service. You'd need some vision and some appetite for risk to pull together the money for this, though, and some balls to take on the zoning nazis as well.
In my city, we have bus lines that go between relatively affluent and dense neighborhoods that until the pandemic were packed every morning. There was a push to expand service but there was pushback because it was “too expensive” and proposals to cut back on muchless used routes was a non-starter because they they served less affluent areas. If they’d added more service on the heavily used lines they probably would have gotten more ridership and could have covered a good chunk of the extra costs. But that’s what you get when you see transit not as a away of moving people around but as a sort of assistance program for low income people which probably makes it worse for everyone
If you can spend the money from low-ridership routes on getting low-income people a Lyft to the transit hub, and then repurpose the vehicles to make the major routes between transit hubs more frequent, then you can get around this issue.
I'd be a pretty early adopter for better bus service - I simply dislike driving (especially parking) and would be willing to have a slower commute to avoid it. A lot of the younger generation is the same, the "American love affair with the automobile" looks to have been generational and is fading. At least in cities and denser suburbs.
But there's a limit to how *much* slower people will accept, and 30 minute intervals with frequent stops just won't cut it. On major trunk lines! What is that even for? I guess it's just supposed to be a last resort for people who can't afford a car? But it would be so easy to be much better, it's extremely frustrating.
This is in the Northern Virginia suburbs, where we have well-funded local government, awful traffic that makes driving stressful, and connections to DC metro and commuter rail. If non-last-resort bus service in the suburbs can work anywhere, it ought to work here.
I wonder if its generational and going away, or simply that driving isn't as much fun as it used to be. Matt made the point in his book that the average driver spends significantly more time in traffic than they did previously. If your experience with driving was going for a fast drive, you probably think cars are pretty cool. If its sitting in traffic trying to get home from work, then you probably hate most of the time your in a car and don't like them very much.
Road capacity expansions have slowed to a trickle, while in most cities jobs have continued to decentralize, and car ownership rates have hit 800/1000 people (though that stopped increasing in the 90s?) Carpooling has also collapsed. Overall highway speed compared to vehicles on the roadway is highly non-linear, and most of the easy capacity increases have been built, and we've refused to look at pricing to control demand, so driving now really sucks.
The ideas outlined here are necessary to make the bus better, but not sufficient. In many places the bus is not enjoyable. It smells. It's not clean. People, disproportionally women, are harassed by strangers. If the bus is uncomfortable to ride, wealthy people will drive cars.
I would really be interested in an American perspective analysis of Toronto's transit system. I grew up there, it is heavily bus-based, has streetcars without dedicated lanes, terrible governance and a slew of white elephants, but it is widely regarded as working rather well, especially for a low-density city. I think one of the crucial aspects is that bus services almost universally serve to get people to the subway, instead of being intended for entire trips. Have any of the prominent transit writers done a comparison with similar US cities?
There was a great presentation about this at Transit Con earlier this year. When the Ontario government amalgamated the old City of Toronto with its suburbs, the TTC took a very different path than American transit systems and extended the city-quality bus routes into the suburbs. This created a suburban constituency for bus service. As you say in your comment, the linkage of these corridors with the subway lines at key transfer stations like York Mills was also an essential component of the system's success. Even though York Mills itself is a suburban mess, it has higher ridership than the Bethesda station in the DC area, which is typically held up as a great example of re-orienting suburban land use around transit. The reason for this is transfers to frequent bus lines.
As far as mixed-traffic streetcars, Toronto is is somewhat unique in the English-speaking world in that they preserved most of their legacy streetcar system. Melbourne is the only other city that did this in any substantial way (a few American cities like New Orleans and Philadelphia kept a handful of lines, but not a significant network). Matt is rightly critical of American cities that have built mixed-traffic streetcars recently, but it's unfair the levy that criticism at Toronto, which is simply running the system they inherited. In fact, the TTC knows that mixed-traffic streetcars are not best practice, which is why you are seeing things like car-free King Street, dedicated median lanes on the Spadina and St. Clair lines, and grade separation on the new Eglinton LRT. Ideally, Toronto would bring the King Street treatment to other major corridors with mixed-traffic streetcars, like Queen Street.
Something you notice in Hollywood films as a non-US (Australian) watcher - a character hitting their absolute low point in the story is usually the scene where they are forced to catch a public bus. It's a bit odd watching from a city where catching a bus is a pretty mainstream transport option. Maybe positive representation for bus transit on screen needs to get on the progressive agenda #OscarsSoCar
Indeed. An underrated aspect of the movie Speed is the community of riders and how they rally together in the face of adversity.
The socialist reading of Speed is under-explored for sure #ReleaseTheKenLoachCut
Ah, Hollywood. This, plus the idea that a bus in LA can go 50 mph continuously . . .
Growing up well-off in Auckland we called the bus 'the looser cruiser'
I would like to add that Google Maps and Apple Maps (and others) have made navigating complicated bus lines much easier for the uninitiated. There are nice synergies to mass transit and mobile devices - they make the whole experience more pleasant (podcasts, music, audio books; noise canceling headphones are getting very cheap). While all of this is true of driving, driving is cognitively demanding and, I feel, minimizes enjoyment of whatever preferred media you consume. Using transit leaves me in a much more rested and thoughtful state than driving ever did.
Terrific article as always, but it seems to ignore the tension I suspect is always there between efficiency and equity (other than the discussion of coverage vs density). It seems that in many cases the express goal of politicians is to have some, any transit, however bad, available to people who cannot afford cars or two cars per family. Matt’s analysis seems to omit that constraint.
Like in many cases, it’s not that decision makers don’t understand what’s efficient. It’s just not what matters to them. I think they are ok with buses’ reputation as “poor mans transit” as long as it allows them to provide some service to the disadvantaged population. After all, if waiting 30 minutes is your only option, you’d rather have that than nothing at all.
This also can explain to some degree why stop-spacing is so close. Many people can walk a quarter mile to their endpoint, but for someone like my mom, who is overweight and has asthma and until recently was recovering from chemo, spacing stops that far apart would literally be prohibitive. Having spent many, many years on buses, I think health constraints are closer to a norm than an exception among lower-class riders.
Depending on the number of people with these sorts of mobility constraints, it seems that the city could provide better service by giving them free credits for paratransit, or Lyft, that can pick up and drop off door to door, while making the big vehicles do what they do best, moving large numbers of people from one hub to another.
Paratransit, in my city, is of absolutely terrible quality. Like impossible to book, and hours and hours late. Better (and likely cheaper) for the city to reimburse eligible people for wheelchair-accessible cabs.
Jarrett Walker points out often on his blog that Lyft/Uber haven't come up with anything that fundamentally does better than transit at the ridership job - but it does seem to be a major improvement on pre-app dispatched single-person rides, like taxis and paratransit.
Yeah, I have experienced the issue with wheelchair users boarding buses and said to myself that maybe people with disabilities should get free Lyft/Uber service at the city's expense. Wouldn't transit users without disabilities get a significantly better experience that would justify the extra tax money--to say nothing of the benefit to people who *do* have disabilities?
The concern with this might be that by unbundling service like that, planners/politicians would be able to starve the more-expensive-per-rider paratransit service of funding in order to improve service/win the favor of the median rider of big-vehicle few-stop services, similar to concerns around private/charter schools diverting voter attention from public schools.
In this case though the idea would be that the "coverage" service and the "ridership" service would have separate pots of money - they are currently bundled, which makes it easy to rob one to do the other, but Jarrett Walker advocates separating them institutionally, even if they are run by a single agency.
My question about this logic has always been that this makes sense if you happen to live close to the street where the bus line is (i.e. thinking about it in linear space where your main distance is between stops), but what about all the people who live perpendicular to that (i.e. two dimensional space where distance to the line is the big factor). Those people have to travel pretty far either way and I'm curious if distance between stops is really the deciding factor.
Was hoping someone would make this comment! I grew up in a city (Buffalo) with a pretty lackluster transit system, but pretty thorough coverage. My parents couldn't afford a car, and neither could many of our neighbors. I had to take really slow busses wherever I needed to go, including, ie, school, and trips that would be a 20 minute drive would routinely take 1h30m end-to-end. But it was always the case that if I was willing to show up 30 mins early to something and spend up to 2hrs getting there, then I could make it. And since I didn't have any other choice, that's what I did. Most of the busses I took were empty, but I needed them. That's even more true of the people who were travelling on these god awful lines for work, whereas, as a teenager, I was usually sucking it up to hang out with friends.
Seconding this comment. Transit planners understand the tradeoffs Matt has described, and many of them probably personally agree with his preference for ridership vs coverage and greater stop spacing. But the bus networks we actually have reflect the priorities of the elected officials on the governing boards of our transit agencies. They also reflect things like path dependence and the challenges of an FTA-mandated Title VI service and fare equity analysis, which can be a barrier to reallocation of existing resources and operating "like a business." Not to say that means we should just throw our hands up! But the constraints are real.
Yeah I was thinking something like this too. I know very little about transit policy, but the type of thing that Matt’s talking about sounds like poorer people will often end up getting screwed over? Like I know that, when there are just a few subway lines, the apartments get more expensive near the lines because of the added convenience of being so close to a transportation source.
So if you consolidate lines, eventually I would think the prices are going to go up along where the new lines are, making it harder for poorer people to afford to live closer to the busses and making them somewhat more difficult to access. Like you might end up getting a higher number of people on busses, but then making it much harder for a sizable population of people who would have ridden different lines before to actually use the busses, and that that group of people is disproportionately likely to be lower-income (after some time is given for prices to equilibrate).
Matt just shrugs this off with “they’ll have bigger yards” but idk that just doesn’t seem like a satisfying answer if you were relying on a bus to get to work that’s now moved 25 minutes away from your home. But maybe you can balance it so most people don’t get harmed too much and you still preserve the efficiency?
This is why in the long run this has to be paired with land use reform. Instead of expensive five-story apartment buildings next to the transit lines and fancy mansions behind them, you need ten or twenty story apartment buildings next to the transit lines and five-story apartment buildings a couple blocks away. And perhaps more relevantly, if there are currently some nice long straight roads leading to downtown that currently don't have much bus service because they go through low-density areas, you can put frequent transit routes along those roads and upzone them simultaneously to create more opportunities for low-income people to find something near a transit route.
The housing next to the transit routes would be new high-density construction and very expensive. It would require deep subsidies to provide even a small amount of affordable housing in those buildings. It might be more effective to provide middling-frequency feeder lines so poor people can get to the high-frequency transit trunk lines and so get around without a car.
Only the newest construction along those routes would have those characteristics. But if new construction is able to come regularly, then the existing buildings in those locations would provide the affordable housing, while the newest ones would be expensive - as opposed to the case when there are no new buildings, and all the existing ones are bid up to the prices that people would have paid for new buildings in those same places.
Paratransit is an interesting wrinkle here. 10 years of riding MUNI in San Francisco has impressed on me the significant tradeoff between inclusivity and efficiency when there is a large population of riders in wheelchairs.
At this point I’ve seen literally thousands of instances where a bus full of otherwise progressive and equity-minded types will audibly groan as we approach a bus stop where a person in a wheelchair is waiting, knowing that the process of deploying the ramp, onboarding the rider, and securing them in the specialized seatbelt can make the difference between being on time for your next bus transfer or clock-in time at an hourly wage job, and being late. Doubly so if the rider is also getting off before your stop, where the seatbelt and ramp process plays out again in reverse.
One might reasonably suggest increasing the number/frequency of buses, but this does nothing to alleviate the social tension (people on the buses with wheelchair riders will still be late unless they shuffle off the bus and onto the one behind them when it catches up) and in the case of buses like the 22-Fillmore, which are electrically powered by tethers on the roof connected to overhead power lines, the faster buses can’t pass the slower ones, and you end up with the very normal occurrence of two buses on the same route, one right behind the other, the first one full and the second one mostly empty, and everyone running late.
The social ideal here would obviously be having an efficient system where everyone rides together and doesn’t exacerbate the existing stigma of “otherness” that comes with disability/reduced mobility, but at this point I’m convinced it’s just not feasible, at least in SF, and some sort of on-demand paratransit or outright Uber/Lyft/Taxi subsidization for wheelchair riders is the better way to go.
(And don’t get me started on the impact of endless elevator outages on wheelchair-bound riders of underground MUNI and BART. It’s appalling)
This can be solved with a good dedicated level boarding stop using a Kassel Kerb. SF because of insane car owner NIMBYism in general has abysmal bus stop accessibility. https://en.wikipedia.org/wiki/Kassel_kerb
The new rapid bus line in Oakland (Tempo) has this, along with proof of payment, and it makes boarding so much faster. Probably 25-30 seconds on a normal bus vs 10 seconds or less on Tempo
Didn't know it had been deployed in the US. Good stuff!
It’s interesting that this is a prime case of planning by lawsuit. Disability activists argued that equity required that everything had to be accessible on principle, courts agreed with them on that, and so transit agencies were forced into massive investments on a schedule to meet this mandate, displacing other capital investment that might have been of more importance to the general public. This is how the law works, in absolutes, independent of planning or financial tradeoffs or, indeed, of anyone’s rights except those of the disadvantaged population involved in the lawsuit. The general riding public has no standing, so doesn’t get considered.
Now, sure, those elevators and bus ramps are useful to people besides disabled people, so the investment is not so narrow. But if the elevator isn’t working, disabled people can’t use the stairs, and given feasible limits on serviceability of the equipment in this high-stress environment, those people will not infrequently get trapped in the system. Resulting in more lawsuits in which judges will declare that the agency has an absolute obligation to spend unlimited funds to succeed at the impossible.
I wonder what value could have been obtained had those billions been spent on working-class jobs (taxi drivers) and wheelchair taxis instead of on massive construction and on equipment that has to be expensively maintained and monitored, and regularly replaced.
Partial solutions: higher bus stops for at-level boarding, "kneeling" buses that get closer to curb level, proof-of-payment systems that allow all-door boarding so that able-bodied riders can get in one door while the disabled rider uses another. As far as the audible groaning, it's public transit and disabled people are part of the public too. Do people huff and puff about the inconvenience of giving up a seat to a pregnant woman or elderly person?
SF does have kneeling buses, and this is much appreciated by older and less mobile folks, but doesn’t help much with wheelchair riders. There’s also ramped sections in the road medians for many of the streetcar line stops, which allows for at-level boarding of wheelchairs, but to my knowledge all of the buses have steps at the entrances so the deplorable ramp and lift is the only option for them. POP is also the default for all buses and above ground streetcars, which does help.
As for the groaning, part of the issue is just the sheer number of wheelchair riders on routes where homeless populations are concentrated. You just have to experience it to understand how frequently this happens. At a “normal” level of wheelchair ridership (relative to overall population stats) most people are gracious about the occasional inconvenience and happy to support a more inclusive public good, but when it’s every fourth stop, and they’re only riding for a few stops at a time before getting off, then anyone who can afford Uber or Lyft instead is going to migrate to that instead, which results in less funding for the system, which means fewer bus lines and lower frequency, which exacerbates the problem, and the public funding doom loop begins.
To be blunt, people don’t huff and puff about pregnant women or the elderly because those groups don’t routinely make the entire bus late, and they’re also not routinely drunk at midday, soaked with their own urine, or screaming at things that don’t exist. SF is a somewhat unique problem in this space.
Transit users regularly cast a calculating eye on the elderly and the pregnant. The goal is to determine, if only on the scale of one’s own conscience, whether the elderly/pregnant person is spry enough to obviate any need to give up a seat.
I will admit I have some scruples about how to treat older people who dye their hair and dress in a style typical of younger people. Is giving up my seat an accidental insult? Would they prefer if I sat confidently and pretended to be fooled by their artifice?
I get up anyway since I don’t feel qualified to parse the ethics in that situation, but I’ve gotten miffed eyebrow raises for it enough times to make me question myself. It’s like the opposite of asking an obviously middle aged woman to show her ID when buying alcohol.
*Correction, not all buses have steps at the entrances (I misremembered since I’m usually boarding at the middle doors, and it’s been months since I’ve been on any of the buses due to covid) so there probably is room for improvement of at-level boarding procedures. Older buses have an extend-and-lift hydraulic system for wheelchairs that takes a long time, while the newer models have a fold out ramp that works decently, but in either case the total boarding time for a single passenger can be 2-3 minutes, and the same on the way out. Doesn’t sound like much, but when there are multiple wheelchairs getting on and off during your trip and missing your connection means waiting 15+ minutes for the next one, it adds up.
*deployable ramp, ugh
That's interesting, thanks for the explanation! I didn't realize that disorderly wheelchair users made up an unusual share of the population in SF...
Is a seatbelt strictly required? I thought buses normally had a kind of rear facing padded back thing for wheelchairs to rest against, then they just apply their wheelbrake?
I’ve seen both (there is a wide range of bus types and ages in the system) and I’m not sure about the legal requirements—in practice it seems to be at the rider’s discretion, but in cases where the rider is unresponsive or seems confused, the bus drivers are usually pretty diligent about getting them fully strapped in.
Anecdotally, it seems that around big sporting events (World Series, Super Bowl, etc) MUNI puts a lot of older buses into service since there’s a not insignificant chance one might get torched.
The title of this post should have been "Build Bus Better".
Yesterday's post was all dry, technocratic wonkery about race, murder and cops. A big yawn, and my eyes glaze over.
So, in order to get people riled up and yelling at each other in the comments section, Yglesias throws us some juicy red meat transit-blogging. That gets the tempers flaring!
Try to play nice, everyone. I mean, it's a gut-punch, reading about European routing protocols and square footage of lane capacity and all. It's no wonder that people have a hard time sticking to facts and reasoned argument when the subject-matter is so inflammatory.
One major disadvantage of bus routes is that there is a very steep learning curve to learning route maps, bus frequency, and transfer mechanisms. Some systems (as well as Google Maps) now have trip planner apps that help with this but the process of just getting from A to B is still needlessly opaque. This system works well for people such as daily commuters who are willing to invest the effort but useless for tourists, out of town business people, and people with random destinations.
Some cities now have Circulator type buses which are frequent and run very busy closed loop circuits. They are often differently themed from commuter buses and may have different fee structures down to and including free. So as I read this article, buses should be more frequent, run fixed routes, and have fewer stops. Street cars. Congratulations, we have invented street cars.
No one is claiming the bus is anything novel and exciting. It's the same thing that they already realized in the 1880s would be valuable - a vehicle for multiple people to ride that goes down the busy street to all the major destinations, and is a few blocks away from most people. The advantage the bus has over the street car is it does it more cheaply, and with the ability to get around a stalled car or collision.
The streetcars versus buses debate goes back to the 1950s and 60s with many of these same tradeoffs argued. Buses were seen as more convenient and flexible and had the support of major bus manufacturers. And while the GM conspiracy has been debunked, favoring the car over privately run streetcar concessions killed this mode.
https://www.vox.com/2015/5/7/8562007/streetcar-history-demise
One of the reasons people liked buses better at the time was that they could pull over to the curb.
Original streetcar lines ran in the middle of the road - and stopped in the middle of the road. They didn't have any platforms or anything, people just got out and walked over.
This became a problem as car traffic increased.
Modern-built streetcar / light rail systems generally avoid this in favour of either running in the nearside lane, having side platforms out in the middle of the street, or having island platforms (most modern such vehicles have doors on both sides, whilst traditional streetcars generally only had curb-side doors like buses.)
But you can still see the old mode of running in systems that survived such as many of those in central and eastern Europe, and Toronto.
Running on the side of the street is actually terrible, and almost no new build systems anywhere outside the US do this, but yes platforms in the median of the street are the global standard. Got a lot easier once low floor trams became common, as you need a lot less "station" at your stop for that versus the high level step up models.
buses are better unless you're going to give the streetcar are well protected median right of way, and expect you need more capacity than an articulated bus can provide.
This may be true in a handful of US cities, but even Chicago has a bunch of routes that aren't filling a bus if they run every five minutes.
A standard part of a bus network redesign in the Jarrett Walker school (and he's pretty much the lead thinker on this) is creating a new bus network map that clarifies which routes are frequent. He also encourages bus networks to be designed as grids and transfers to be free to increase mobility. It's interesting stuff.
The main shortcoming of streetcars is that they're short.
modern streetcars actually can be significantly longer than buses. Their draw back is that unless you give them a privileged right of way, they're very easy to block.
If daily commuters are the bulk of riders/travelers, especially during peak hours, then getting them into efficient modes of transportation frees up road/rail space for other travelers. More commuters in buses means more cabs/cheaper rideshare/less traffic for tourists/etc.
Regarding fares, I’ve always appreciated the approach taken by German and Austrian cities I’ve lived in: everybody is required to buy a time-stamped ticket for a ride, but there is actually no mechanism to check whether you have done so. Fare enforcement is handled completely by “sting” operations, where suddenly five transit employees appear in your bus/train car/subway car, asking to see tickets. No ticket? Instant fine, of (at the time) 65 Euro. Most buy tickets, but many opted to chance it 100% of the time, and they thought they came out ahead, financially!
I don’t know if that would work in the States, but it’s my understanding that fare collection contributes a minor amount of income to transit operations anyway. I’m curious the extent to which fare collection or enforcement mechanisms are costly, and if there is ever a scenario in which we can agree that public transit is a public good that doesn’t need to run an operating profit, or operate like a business, to be a success. Its main function is to remove cars from roadways, which in theory is helpful to everybody.
A difficulty with POP (which I agree is better) is in America progressives have now gotten very skittish about anything that smacks of calling for more policing.
Alon Levy makes a pretty good case here that you should treat POP inspections as more like a loss prevention measure--the purpose is just to get your losses down rather than totally eliminate fare-dodging. https://pedestrianobservations.com/2019/11/13/fare-evasion/
Not sure this would work to sell POP inspectors to progressives though.
About those instant fines: what happens if the fined person refuses to pay the fine? Do they get arrested? Do they get whatever the local equivalent of a desk-appearance ticket?
So - we're going to the european airport from our hotel, and decide for reasons of virtue and cultural experience to take the Bus. We leave a solid 1.5 hours early just to make sure we have enough time. We get on the bus along with a bunch of people, and turns out two of these people - father and son - bought just one ticket between them. They get their tickets checked (not sure why their tickets were checked, ours weren't, might have something to do with their middle eastern origins). The dad gets asked to pay up a fine. Dad says no, just let me buy a ticket. Bus driver - nope, got to pay the fine.
So far so normal, if slightly racist seeming but what do I (brown American tourist who's feeling slightly uncomfortable at this point) know? Other people on the bus seem inexplicably upset with them. The next thing was amazing and might explain the upset other people - the driver actually made everyone other than the dad and son get off the bus. Called a replacement bus, which got there 20 min later (this after 20 min of arguing with ticketless dad on the bus, 40 min wasted time at this point).
Oh the replacement bus does not seem to be an express bus to the airport - it is going to the airport, but stopping for locals... When it finally gets to the airport, we have barely any time. Then, the final bit of excitement - (unlike the original bus) it actually doesn't even go to our terminal! It just stops for good at the first stop that could conceivably be called getting us to the airport. We have to take a shuttle to get out our terminal.
So - yeah, what happens is deeply disturbing.
Typically I don't think these are enforced as criminal citations. So I suspect they might suspend or ban you from riding the bus/train but likely no arrest.
How would they do that? The entire point is they aren't controlling access to the transit system, you can't ban people.
I assume if you don't pay now, some kind collections process commences.
How do you collect from people who don't have any money and no credit rating to lose?
One unconsidered thing about free transit is that it would effectively transform the transit system legally into a public street and there would be no capacity to control misbehavior that would drive away riders but that the courts would consider to be civil rights. It would make the system unmanageable, and for this reason both the MTA and the NYPD have opposed it.
Here’s a real-life “why you can’t have nice things” situation
In Budapest we were told by our tour director that the subway tickets were rarely enforced. Being in a rush as well as cheap students we didn't buy tickets only to be stopped by very sketchy enforcement people in civilian clothes but with badges and ticket books. Needless to say the 65 euro fines payable on the spot cancelled the planned evening of drinking and we returned to our hotel poorer and upset.
In hindsight they were clearly targeting clueless tourists which we very apparently were. Unless it was all a scam which is also a possibility.
When I first visited Amsterdam, a friend living there told me they never enforced the ticket policy. Sure enough - on my first day, a couple of us got stopped by some very imposing Dutch police. I played the true dumb American, saying in my home, buses were free (it was actually true!). Amazingly they let us off without paying.
Ha, this is exactly right for Austria too - if you see a sketchy looking person board the train it’s almost always someone checking for tickets.
Proof-of-payment systems are generally cheaper, but they do have to be enforced correctly. The German systems work pretty well because the spot checks are legit random--the fare police could show up at any station at any time. This tends not to be the case in the US.
I had the unusual "privilege" of living in Camden, NJ while I was in law school and for about a year after. Camden is the southern end of the River Line, a light rail that runs up to Trenton, and one of the relatively few proof-of-payment transit systems in the US. I used it fairly often to visit friends in Central/North Jersey and New York via the NEC. had the even more unusual privilege of actually commuting on it, as I had an internship in Trenton my 2L summer and worked in Trenton my last six months living in Camden. Even after I moved away, I continued to use the River Line as I was living in New Brunswick and found it useful to visit friends still living in South Jersey. I didn't really stop using it until I moved to Philadelphia.
The upshot of all this is that I tended to ride the whole length (Camden-Trenton): either I was going from Camden to my job in Trenton (and back), or I was using the RiverLine to connect to the NJT NEC Line to/from points north. I gotta say, fare enforcement was pretty un-German. On the one hand, when I was commuting, I would catch the same train every day there and back. Both of these were peak trains which tended to be full at least at some point. And pretty much every time, the fare enforcement officers would show up at the same set of stations in each direction.
On the other hand, my non-commute trips--which tended to be off-peak--only encountered fare enforcement a handful of times over what must have been dozens if not hundreds of trips. When they did appear, it was typically at the same stations as during commute hours. I don't know if this actually helps NJT with its farebox, but I can say that a lot of people treated the River Line like it was free off-peak, and it kind of was. I didn't, but only because the fare was so low ($1.60 to go 30 miles? Yes, please.)
And amendment: Yes, the River Line has the full Central European buy a ticket/validate a ticket system.
Proof-of-payment systems are pretty good, and they're especially good for buses where payment causes a lot of delays at each stop. A major snag is that in the American context, creating more opportunities for interaction with enforcement agents has a lot of downsides.
It seems to me that the best overall bus system would involve very cheap monthly cards -- $10 or something -- with a touchless interface.
In NYC, I've taken the bus thousands of times, and I'd say the average person takes like 5-10 seconds to get out their card and swipe it, which really adds up. It also makes for dynamic effects where since the bus goes slower, there are more passengers on any bus at any given time, and when the bus is at capacity it can take several minutes for everyone to get on and off. I know making the bus cheaper shifts ridership to it, so you would have to make the subway cheaper too. I think it's absolutely the kind of thing that would pay for itself in economic gains.
The system here in Boston has a tap card (among other things) and it's usually not a slowdown for people to tap on the way into the bus (though swiping a magstripe card in the same direction as you were walking might have been even faster). The big snag is the people who really, really don't want to load much money on the card, and are individually loading money onto the card each time they ride (the fare with a card is cheaper than just paying cash, so they have one). You could remove the discrepancy and just let people pay cash, or make a cash-value incentive for loading up a card ($25 of fares when you put on $20?), or have fare vending machines at every one of six thousand bus stops (ha! but they manage it in several European cities).
What's the snag -- you mean they try to add money on the spot? Or they try several cards?
Yes, they add money on the spot (which the vehicle farebox lets them do). So they push a button, tap their card, put money in, push the button again, tap the card to load the card, then tap the card a third time to pay the fare with the money they just loaded.
As I understand it, most Central European and Eastern European systems use a system whereby you can buy a set of tickets in a strip but you must stamp or validate the one you are using before each journey or it is invalid.
Travelling with a non-validated ticket is the same as not having one at all, or else you could just buy a strip, and only "use" one on the rare occasions when a fare inspector was there.
I have to admit I instinctively dislike this system. It seems hostile to newcomers. I experienced this system in Vienna, and never would have realised if I'd not been told that merely buying a ticket did not mean I was good to go.
When I was traveling in Amsterdam the difficult part was remembering to swipe out on exit. If you didn't do it, the ticket wouldn't show as valid to the enforcement agent and it was going to be treated like you beat the fare. Fortunately people were kind in reminding us, and we didn't meet any Dutch enforcement agents.
Caltrain had an electronic system where you basically badge in at a station and then badge out as you leave. I usually had an easy enough time remembering to badge on, because you're kind of just standing around waiting for the train to come anyway, but when I got off, I was always thinking of my destination and frequently forgot to badge off, which sucked because I then got charged for the maximum possible fare.
Yeah I don't mind swipe out / in systems if combined with ticket barriers, but systems where you just have to remember seem unfriendly to me.
A lot of people seem anti ticket barriers but I think they're underrated as they make it hard to *accidentally* fare dodge, something people unfamiliar with a system may worry about, and thus be put off using it.
Here in Cleveland, we had a Bus Rapid Transit line that ran on proof-of-payment, but a judge ended up ruling that there was no non-racially-discriminatory way to execute what you're calling those "sting operations." (I rode that bus to work every day for a couple years, and I can confirm that the payment checks amounted to rounding up a bunch of black folks and pulling them off the bus to get ticketed for like $200).
After the ruling, they had to switch to everyone showing payment at the door, which crippled the line and made it much slower and much less convenient. I'm not even sure what to do other than cry.
That seems odd. If you required the inspectors to check everyone on the bus (and picked the buses and stops at random) that doesn't sound like it would be discriminatory
I went back to check what actually happened, and it turns out I was slightly wrong. The specific ruling was that checking for fare after the fact was an unlawful search without probable cause under the fourth amendment. So it was ruled specifically unconstitutional. The judge said that using non-law-enforcement personnel would eliminate the constitutional problem, but having police officers check was not allowed.
https://www.clevescene.com/scene-and-heard/archives/2017/11/02/cleveland-judge-has-ruled-rtas-proof-of-payment-fare-enforcement-method-unconstitutional
That seems like a very interesting case. I would not have guessed that having state actors, but not law enforcement, would be considered acceptable. (Also interesting that the article mentions a report on low police morale because they felt they were being used as revenue-raisers, which seems like a wider problem than just transport also).
And I suppose they were using police in the first place because it wasn't safe for non-police control agents to do it. Bus drivers in New York don't say anything about fare beaters now because drivers were being attacked so often that their union went successfully to bat on this issue.
It's true, a large minority of bus riders in New York just don't pay and drivers rarely hassle them about it. If you made the bus completely free, it would have much higher throughput, though it might slow it down for the average rider due to people using it in place of the subway.
I would guess that the court found that the people who used the buses disproportionately were coming from predominantly black areas, and so there was a "disparate impact" that meant that even if facially neutral, the policy was racially discriminatory since mostly black people ended up paying the fines. Something along those lines.
By that logic, police can't patrol high crime areas if high crime areas are predominantly black. That's pretty close to "abolish the police" logic and doesn't seem likely to stand up on appeal
Well, looks like my guess was off-base anyways. The "disparate impact" type of reasoning comes up in wide variety of areas though, and is quite controversial with the more conservative end of the political spectrum.
The current administration is doubling-down on this disparate impact concept in all areas.
Sigh . . .
When I was briefly working in California, CalTrain (SanJose <--> San Francisco) worked this way. I assume it still does, but this was ~15 years ago.
As a temporary visitor (only working there for a few months) I was always worried that I'd mess up and get fined but in general it was smoother.
It still worked that way 5 years ago when I was regularly commuting on CalTrain.
it still does
The spot enforcement method is used in New York City on Select Bus routes where the fare is paid on the street before boarding. In my observation there is still a lot of fare evasion, because the people inclined to beat the fare are also the people who have no intention of paying the fine, and there is no way to make them pay on the spot, not least because many of them have no money anyway.
Fare collection is a major cost. There is a huge associated infrastructure of turnstiles and bus fareboxes and fare media sales points and back-end processing, plus personnel costs of maintenance and enforcement. For the New York subway/bus system (back in the day at the MTA) if memory serves the operating portion alone was close to 30% of gross fare revenue and the capital contribution was billions of dollars. I don't think the needle has moved much in the current iteration.
Fare collection doesn't just cost in terms of all the equipment - it also costs in terms of travel time, since everyone who is currently on a bus has to wait 10-20 seconds for each boarding passenger at each stop, instead of just a couple seconds each. Speeding up the bus both helps the passengers already on the bus get where they're going faster, and (as Matt mentions in the original post) helps the bus complete its route faster so that the same number of buses can run the route 6 times an hour instead of 5 times an hour, decreasing everyone's wait time before boarding as well.
Very true, which is why off-board fare collection is an essential part of better-bus arrangements. And it definitely works from that perspective. When Select Bus was instituted on Second Avenue, the MTA did a very intensive enforcement to establish the baseline that you could actually have to pay a fine if you didn't pay the fare. It was labor-intensive, very unpopular and the politicians got involved, so the MTA backed off, and now people don't seem to care. Since this is the reality, I think you might as well make the buses free, especially if they would be effectively free through generous pass schemes and free transfers anyway. But keep the fare on in the subways, so you can kick out people who misbehave and have no evidence they paid the fare.
The Baltimore light rail system uses fare machines to purchase fares, and gives you a receipt. You are supposed to show the receipt if asked by the transit police that do periodic spot checks. Failure to produce proof of payment can lead to a fine or arrest. You can currently purchase an all day pass that covers you for a 24 hour period on all forms of public transportation. As of today's date, it's $4.40. $2.20 for seniors over 65 and disabled passengers. I've been asked to show proof of payment on more than one occasion. I remember in Berlin, Germany, the UBahn used the same fare machine system, and the receipts were time stamped, and good for a certain length of time. Here in DC, you have Metro, and faregates that use a SmartTrip card. Most of the fare jumpers are teenagers who just jump over the faregates and don't pay. Station Managers are not allowed to enforce the fare. Only Metro Transit Police have that authority. Since most of the kids ride for free anyway, because of school/transit fare programs, there is very little fare enforcement.
The Dallas Area Rapid Transit (DART) uses a system like that.
Even more, you can buy tickets on a mobile app that are good for 30 days and activate them when you board -- or just wait until you see somebody checking tickets and activate it before they get to you.
It's possible this has changed in the years since I rode the train (there's a pandemic on, after all), and I never took advantage of it, but the whole scheme is basically an honor system.
St. Louis does (or did) this on its light rail “metro link” except in certain circumstances where it’s too crowded to inspect (mainly after cardinals games). Always worked fine. They also had an out of towner tax by tripling the cost of tickets bought at the airport.
That's how it works on the MAX in Portland OR (or at least how it did when I lived there)
I'm sad to be the first to comment as I find this topic very compelling.
One substantive comment I have is on the max coverage issue. A problem with that is that the routes become confusing for most people that aren't frequent users. So you end up driving and paying the $10 to park out of fear that the bus will end up going to the wrong place.
Every time I find myself near a bus stop in any city I look at the bus stop signs and route maps and try to figure out the system. It is nearly always impenetrable and confusing on even where that bus is going, let alone how or where to transfer to a different lines.
Subways and streetcars get a bad rap for being expensive and inflexible but the ease of use and convenience are big advantages to them. For one thing, they are big and visible so there is a much greater psychological sense of permanence and reliability.
This reminds me of my experiment using the bus system in Reykjavik, which is great for such a small city. The maps and stops could be puzzled out, but try being on a bus to a place you’ve never seen, when you can’t understand the announcements and the names of the stops don’t fit on the signs!
Yea confusion about bus lines is a real problem. All the more reason to have fewer lines and fewer stops, but with more used buses with more frequent service.
For years I commuted via bus from my home in west LA to Santa Monica. And I hated it. Matt's is a great post about how some technical tweaks to bus systems could help improve it (on the margins at least) but unless there's a more systemic change, all these improvements will disappear into the void.
Frequency was not a problem -- during rush hour, buses came every ten minutes (or faster). This still didn't mitigate the frustration when you missed a bus by ten seconds: the ten minutes felt like thirty! Bigger problems were the variability. Three buses might stack up and all appear within five minutes, while at other times there might be twenty minutes between buses. This was a popular route (Blue Bus #7 for the cognoscenti) and was typically *very* crowded, as it served Santa Monica College and Santa Monica High School, so lots of standing and swaying, and intimate knowledge of strangers' bodies on a herky jerky bus which aimed for every pothole on the street.
But the worst thing that was even though you had given up commuting by car, you got to enjoy the benefits of rush hour traffic jams caused by such cars. You crawled. You sat at lights forever (and this even with the "extended greens" that buses were supposed to enjoy but couldn't because of too many cars in front of it). It got to the point that a mile or so from my destination I would just get out and walk the rest of the way.
And then the new rail line opened (the Expo Line). And that was heaven! A smooth ride, with a dependable ten minute transit from my home station (albeit a longer walk from my house) and Santa Monica. And I could usually sit down! Goodbye, bus. And (since the Expo line paralleled the bus route), it looks like everyone else agreed with me, as I noticed the buses now running mostly empty.
I understand that it's hard to get commuting professionals out of their cars (often I was the only one on the bus or the train). And I understand it's a chicken/egg problem: as long as people stay in cars, and light rail doesn't go everywhere you want to go (LA's big problem), you're never going to have the rapid, smooth and comfortable transit commute that might entice them to act otherwise.
LA has toyed with the idea of congestion pricing and the howls have filled the heavens. But that -- combined with serious funding of transit alternatives -- is the only way forward. It has to be carrot and stick. Given my neighbors' open-arms welcome of higher density in our SFH neighborhoods (yes: joke!), my optimism is in the basement.
Sounds to me like what your bus needed was a true dedicated bus lane. That can also cause howls, but probably less than congestion pricing, especially if the road was very multi-lane to begin with.
That's a great-in-theory type solution (and has been explored). But not only would it elicit the expected howls, it would also fail tests of social equity. For the relatively few people who take the bus and would profit from faster movement, the impact on the vastly more car drivers packed into now fewer lanes would be hard to justify. And the bus lane would be empty most of the time, given likely bus frequencies. Then, consider the amount of cheating by car drivers, the challenges of enforcement, and the likely breakdown of norms as people watch other, more selfish drivers pop into the empty bus lane.
Of course, if everyone were already on the bus, those negatives wouldn't apply. But then we wouldn't need dedicated bus lanes.
I admit this is a hard problem.
I’ve taken your bus as a tourist to get from Santa Monica to the LACMA area. It was very pleasant, more so than driving, but I can see how it would be infuriating for a regular commuter.
Yep. Rush hour, 2x/day, every day.
Agreed it's a hard problem. Though I think some of the cheating/enforcement issues are covered by my conception of a "true" dedicated bus lane: where necessary, put up a physical barrier (like a low, parking-bumper-sized piece of concrete or maybe a series of bollards, it doesn't need to be a full Jersey barrier). Of course, that would be somewhat more expensive than just paint and cause more howling. I'm also skeptical that reducing the number of available lanes will have a large impact on drivers' travel speed--when analyses are done, we're usually talking about very marginal differences (less than 5 mph speed differences or less than 5 minutes travel time differences), though it could be different in this particular case.
“you need to make some people mad for the sake of a better overall system”
This struck me as an insightful point. Does getting something (anything?) big done in our system require a Robert Moses or an equivalent like Baron Haussmann?
People like that did get big things done in a way that's hard to even imagine now, though they made mistakes. A good read about another bullheaded builder like that is Cadillac Desert, about Floyd Dominy who headed the Bureau of Reclamation for many years and built dams across the west.
Yes, or at least a move in that direction, away from anyone in the local community being able to object to anything and get it stalled.
I think that 10 minutes is really pushing it as far as "can just show up" timing on buses. That means the wait will add an average of 5 minutes to the trip. Additionally, buses have fairly mixed on-time performance. In Portland in 2019 the bus was on-time about 85% of the time, which means if you are a regular commuter you'd have an average of 1.5 late buses a week.
If you require someone to transfer once on a trip, they will hit a late bus (5+ minutes late) 3 times every week on average, and the wait will add an additional 10 minutes to the trip every day. There is also the risk of a much longer wait due to a bus totally not showing up (could add 10+ minutes to the trip).
This is quite a bit given the average commute in the U.S. is only 26 minutes. I think the frequency needed for people really to just show up without checking the schedule is probably about 7 minutes. The frequency needed for people to perform a bus -> bus transfer without worry is 5 minutes or less.
If you're at the level of frequency where people "show up and go" rather than paying attention to the timetable, then *late* buses don't matter, as long as the headway *between* buses stays the same.
That said, I definitely think that 6 or 7 minutes is probably better for this than 10 minutes. (I guess 7.5 minutes could also work, if you want 8 buses per hour.)
Unfortunately it typically doesn't work this way because if you have 10 minute spacing between the buses, one bus gets knocked off schedule by 2 minutes it will end up having to pick up more passengers (passengers who would have just missed that bus will now make it) slowing it down. This also speeds up the bus behind it because now you have fewer riders waiting because some were able to catch the bus before...
So the busses end up bunched together once one gets knocked off schedule by more than 2-3 minutes.
I've heard from Jarrett Walker's blog that good transit agencies try to notice when this is happening, and dispatch a spare bus to fill in that gap that got longer, and tell the one that is catching up to just dwell a few more seconds at each stop, so that they can manage headways. This does seem incredibly difficult though.
Not difficult at all, with GPS and cellular radio (which all buses in New York are equipped with). The same technology that gives riders the bus arrival times, also tells the bus command center about the bus spacing.
I think the difficulty is making sure you have a spare bus to do this, that it is based somewhere convenient, and that you understand the speed of traffic everywhere in the city well enough to get that spare bus to some stop in between the two buses where the gap has developed.
It's true that sometimes the bus bunching is so bad that you run out of "gap" buses. Rush hour short-turns can also help, though.
yeah they also have time stops which ideally will "re-set" the spacing.
I think that one reason why bus systems emphasize coverage over ridership in many U.S. cities today is that they want to serve disadvantage communities, or at least not further disadvantage them. A bus line that went from the densest residential area to the densest commercial area is probably going from one rich part of town to another.
Furthermore, given the unique setup of American bus systems, the ridership / coverage tradeoff doesn't look like a tradeoff at first. The fact that buses are priced lower than trains because they are "worse" makes them the default option for low-income folks who can't afford cars. The infrequent service also makes them a default for people who have a low opportunity cost of time, i.e. are also more likely to be low-income.
Therefore, you get into a situation where transit planners say "We can't relocate service from a low-income region to a high-income one. The low-income folks both really need the bus AND really use it, while the high-income ones will just drive either way."
The response would be "If we invested in frequent bus service for high-income folks, we could build self-sustaining ridership and gain the rich folks' support for the overall bus system." Matt has said before that a service that is only for poor people can become a poor service. You'd need some vision and some appetite for risk to pull together the money for this, though, and some balls to take on the zoning nazis as well.
In my city, we have bus lines that go between relatively affluent and dense neighborhoods that until the pandemic were packed every morning. There was a push to expand service but there was pushback because it was “too expensive” and proposals to cut back on muchless used routes was a non-starter because they they served less affluent areas. If they’d added more service on the heavily used lines they probably would have gotten more ridership and could have covered a good chunk of the extra costs. But that’s what you get when you see transit not as a away of moving people around but as a sort of assistance program for low income people which probably makes it worse for everyone
If you can spend the money from low-ridership routes on getting low-income people a Lyft to the transit hub, and then repurpose the vehicles to make the major routes between transit hubs more frequent, then you can get around this issue.
This is my favorite Slow Boring yet!
I'd be a pretty early adopter for better bus service - I simply dislike driving (especially parking) and would be willing to have a slower commute to avoid it. A lot of the younger generation is the same, the "American love affair with the automobile" looks to have been generational and is fading. At least in cities and denser suburbs.
But there's a limit to how *much* slower people will accept, and 30 minute intervals with frequent stops just won't cut it. On major trunk lines! What is that even for? I guess it's just supposed to be a last resort for people who can't afford a car? But it would be so easy to be much better, it's extremely frustrating.
This is in the Northern Virginia suburbs, where we have well-funded local government, awful traffic that makes driving stressful, and connections to DC metro and commuter rail. If non-last-resort bus service in the suburbs can work anywhere, it ought to work here.
I wonder if its generational and going away, or simply that driving isn't as much fun as it used to be. Matt made the point in his book that the average driver spends significantly more time in traffic than they did previously. If your experience with driving was going for a fast drive, you probably think cars are pretty cool. If its sitting in traffic trying to get home from work, then you probably hate most of the time your in a car and don't like them very much.
Road capacity expansions have slowed to a trickle, while in most cities jobs have continued to decentralize, and car ownership rates have hit 800/1000 people (though that stopped increasing in the 90s?) Carpooling has also collapsed. Overall highway speed compared to vehicles on the roadway is highly non-linear, and most of the easy capacity increases have been built, and we've refused to look at pricing to control demand, so driving now really sucks.
The ideas outlined here are necessary to make the bus better, but not sufficient. In many places the bus is not enjoyable. It smells. It's not clean. People, disproportionally women, are harassed by strangers. If the bus is uncomfortable to ride, wealthy people will drive cars.
I would really be interested in an American perspective analysis of Toronto's transit system. I grew up there, it is heavily bus-based, has streetcars without dedicated lanes, terrible governance and a slew of white elephants, but it is widely regarded as working rather well, especially for a low-density city. I think one of the crucial aspects is that bus services almost universally serve to get people to the subway, instead of being intended for entire trips. Have any of the prominent transit writers done a comparison with similar US cities?
There was a great presentation about this at Transit Con earlier this year. When the Ontario government amalgamated the old City of Toronto with its suburbs, the TTC took a very different path than American transit systems and extended the city-quality bus routes into the suburbs. This created a suburban constituency for bus service. As you say in your comment, the linkage of these corridors with the subway lines at key transfer stations like York Mills was also an essential component of the system's success. Even though York Mills itself is a suburban mess, it has higher ridership than the Bethesda station in the DC area, which is typically held up as a great example of re-orienting suburban land use around transit. The reason for this is transfers to frequent bus lines.
As far as mixed-traffic streetcars, Toronto is is somewhat unique in the English-speaking world in that they preserved most of their legacy streetcar system. Melbourne is the only other city that did this in any substantial way (a few American cities like New Orleans and Philadelphia kept a handful of lines, but not a significant network). Matt is rightly critical of American cities that have built mixed-traffic streetcars recently, but it's unfair the levy that criticism at Toronto, which is simply running the system they inherited. In fact, the TTC knows that mixed-traffic streetcars are not best practice, which is why you are seeing things like car-free King Street, dedicated median lanes on the Spadina and St. Clair lines, and grade separation on the new Eglinton LRT. Ideally, Toronto would bring the King Street treatment to other major corridors with mixed-traffic streetcars, like Queen Street.
Yes. https://www.bloomberg.com/news/articles/2018-10-10/why-public-transportation-works-better-outside-the-u-s