How completely can buses come to resemble streetcars/trams? Paris has crossed one important line that North America still seems stuck on: off-board fare collection. Have another look at a new Paris bus:
Just forward of the back door, just above the advertising panel, is a green dot. It’s a button that opens the door from the outside, exactly like the buttons on Paris trams. You can board a Paris bus at any door, unless you need to pay cash, in which case you board at the front. If you board at the back, machines are just inside the door to either tag-on with your Navigo smartcard or validate your magnetic stripe ticket. This is a standard proof-of-payment system, enforced by roving inspectors, just like what North Americans see on light rail but never on buses.
If you’re used to dealing with crowded buses in any North American or Australasian city, you can probably sense how transformative this would be. First of all, if a bus is mostly loading (e.g. at the beginning of the line, where this bus is waiting to start), people can board all doors, so if there are three doors they can board in 1/3 the time. Second, front-door collection forces people to start at the front of the bus and then move back, and this movement itself takes some capacity. On Paris buses, there’s no need to move around in the bus, instead, people tend to stay around the door where they boarded, just as they would on a tram or subway.
So these buses are better for both capacity and comfort. Capacity, because the need for people to move around in the bus takes some space. Comfort, because while standing may not be ideal, fighting your way through other standees to get to the door is much worse. Note, too, that the whole North American crowded bus ritual, where the driver is constantly saying “please move to the back,” just doesn’t happen, so there’s one less opportunity for the customers to feel like cattle.
In the middle of the line, where you have people getting on and off, these two movements obviously conflict on each door. This looks pretty inefficient from the outside of the bus …
… but it’s still more efficient than forcing people to move forward or back through the bus as other fare collection systems do. After all, in a Paris 3-door bus, this situation is happening at all three doors, and if one door is moving better people will shift to it. The result is a process that naturally optimises the flow through all three doors, and thus minimizes the dwell time.
(Observing the situations, I wondered if it would ever be possible to train people to keep right as they enter and exit the bus. There is a small railing that divides the boarding path into two streams, and if these were used for opposite directions of traffic (at intermediate stops where both directions exist) this situation would move a little better.)
So why, exactly, is all-door boarding on buses so hard in North America and Australasia? The fare transaction experience can do a lot to form impressions of buses vs light rail. Nobody really likes that moment of being judged by the driver as they board, or the boarding queues that this moment requires. So if you can do proof-of-payment on light rail, why can’t you do it on high-volume buses?
They did proof-of-payment on busses in Moscow, at back in 1992 when I was there.
Vegas does all door boarding on their 2 POP lines. SF should go to all door boarding someday, but that might be forever.
Scott. Last time I rode Las Vegas MAX, about 3 years ago, it was all-door boarding but there were conductors on every bus. Defeated the purpose …
Translink does all-door boarding on the 99 B-line on Broadway and they did on the 98 B-line as well, before that got replaced by the Canada Line. City of Vancouver did a study before and after, showing a decrease in dwell time and total trip time. You can see the whole thing at http://vancouver.ca/ctyclerk/cclerk/20090217/documents/ttra2.pdf
All doors boarding is legacy mode in France and especially in Paris, where not long time ago you had the system of section (if your bus was crossing a section boundary you had to pay additional fare when already onboard). System of section disappeared not long time ago (un-enforceable). Some provincial cities, like Rennes, moved toward the more anglo system of front door boarding only: the main reason for it is fare evasion reduction.
What’s the compliance rate on all-doors boarding?
My understanding is that drivers object to mid-bus loading and unloading as they believe it is not safe.
Hyjal. That makes no senes. Rear door unloading is commonplace everywhere. Why is rear door boarding any less safe?
I agree that the many examples of rear loading show that it is entirely feasible.
But I have definitely heard that some drivers believe (note use of ‘believe’ to indicate that it is not factually based)that rear door loading is unsafe because it is harder for them to see whether people have finished getting on or off, and that, even though there are mirrors, their view could still get blocked by other passengers.
So they refuse to do it.
In many parts of the US, bus systems are associated with poverty; and thus with a greater likelihood of misbehavior–proof of payment systems are often viewed with skepticism. (I’m not agreeing with this line of argument; simply noting that it exists). Here in Portland, many critics of TriMet frequently complain about fare evasion on MAX (a notable and frequent critic of the agency recently made a fool of himself in the paper by claiming that proof-of-payment was unique to TriMet)–and would go apoplectic were it proposed for the bus system.
In addition, there are logistical issues with sending fare inspectors out on services which are sparse in both space and time–it’s easier for them to be effective on a high-volume corridor where they can ride back and forth between stations and screen a significant portion of transit riders passing by, then it would be for them to ride on hourly services to the suburbs.
Here’s a question: OTHER than the case of systems marketed as BRT (many of which do use POP fare collection, and are generally marketed differently), are there any bus systems which use POP on some routes, and pay-the-driver on others? A big issue with switching to POP is that it’s expensive and disruptive to do it at once–but doing it piecemeal may be confusing to riders.
“In the middle of the line, where you have people getting on and off, these two movements obviously conflict on each door.”
Tisk, tisk. Where are their manners? Everyone knows that you are supposed to wait for everyone to exit before you enter… wait, a second. Is that just a Japanese thing? In Tokyo (and elsewhere in Japan) this is standard procedure. Everyone behaves this way…. it is the expected norm.
I had always expected that this was a fairly common courtesy, but maybe it isn’t?
They do all-door boarding with proof-of-payment in Ottawa. Cash users board at front, passholders can board at any door.
And the VIVA buses in suburban Toronto have off-board ticket vending machines and all-door boarding.
“That would never work here” is what I assume is the standard line when someone would propose a “radical” concept like that.
You know what was even worse? I was in DC last summer, and my metro line was being bustituted. We all had to board the FREE articulated bus shuttle. The driver refused to open the back doors. WTF?
Switzerland also uses POP for all their bus lines, which includes their “train” buses, where a second car is pulled behind the bus that is unstaffed.
My experience in Europe is that if you are fare-checked and do not have a valid ticket you are heavily fined. Being a foreigner or tourist does not matter. Payment is often required immediately.
My experience in USA is that often warnings are given and no ID is collected. And if a citation is issued, you are now in the municipal court system where you can ask for mitigation or evidence or mercy… and evade a penalty. Thus the penalty for evading fare is minute.
Also USA fare inspectors tend to be uniformed.
It’s probably cost prohibitive for all but a very few routes.
Some huge cultural changes would be needed for POP to be effective other than on very selected corridors, bus or rail – and it is telling that systems like LA and Vancouver want to put gates on their rail. Unless the cultural changes occur, it won’t be possible to make POP effective on bus.
It’s because Paris leads and everyone else copies what Paris does!
It will come to Australasia and North America soon enough.
I think London still operates POP on some routes (the articulated “bendy-bus” routes) at the same time as having driver payment on others. However the vehicles were unpopular and are being phased out. POP was part of the problem, as the buses quickly got a reputation of being “the free bus” which anybody who didn’t want to pay used, and so was perceived as being lawless and a bit dangerous.
Partly this was the fault of the system though, as in deprived areas it is very difficult to top up the Oyster pre-pay card (as there are few shops with the facility) and if you were out of credit and had no change for the pavement machines you had to walk to somewhere where credit was available which was often not convenient. I didn’t pay on those buses several times (and I’m a very conscientious passenger) because it was made so hard to do so.
New commenter here. All doors boarding has obvious benefits, and if you are going that way it makes sense to do away with drivers selling tickets too. I can think of couple economical reasons why it is easier on rails though, and this is assuming the rail vehicle is consirably bigger than the bus.
First of all those “roving inspectors” probably expect to be paid. Given they spend most of their time roving between vehicles, the more vehicles you have to inspect, in order to enforce POP, more inspectors you need.
Second is that the cost of fare evasion depends on the marginal cost of extra passengers, and this tends to be lower on rail; that is rail systems more often have excess capacity, and if that is taken up by people who wouldn’t ride at all if they had to pay for it the cost is zero. If the system is capacity constrained it is the opposite, not only will fares be lost but additional expenses will be generated by extra service needed.
That said, IMO it is a good idea in general on buses too, but in practice depends on the bus service on question. To further complicate matters, in real citiess benefits probably exceed the costs in some routes but not in others, but having all doors boarding on some routes only could be highly confusing to passengers.
BTW, something I would appreciate pointers on, has anyone does calculations on the direct and indirect costs of fare collection? Ticket machines alone are suprisingly expensive, then there is all the background infrastructure needed, sales desks, inspectors, etc. not to mention indirect costs due to slower boarding times for example. I am pretty sure there are some heavily subsidized systems where it might actually be more economical to just let people ride for free.
@ Teme. Good questions all. Generally, I advocate all-door boarding and alighting, with proof-of-payment fare collection, only on major frequent services that routinely run articulated buses. These are still smaller than the largest European trams, but not much smaller than a typical North American light rail car. There's some case to be made for doing this only on "rapid" services which stop less frequently, as it's easier to catch fare evaders if they are trapped on the bus for a couple of minutes. The logical place to start is in agencies that already have these kinds of services separately branded, such as Los Angeles with its Metro Rapid bus product.
Your other questions are good and I invite other commtenters who may know more to provide references. There certainly are times when the cost of fare collection exceeds the revenue it generates, but this is usually not the case on major urban systems. Small-city and rural services are more likely to be in this situation.
Proof-of-payment was the standard practice in the small German city (population 125,000) where I lived in 1986 – although it did switch to “pay the driver” after 8PM every night. I never took any of the many less-frequent regional routes – I don’t know if they had proof-of-payment or not. But I think in addition to “cultural” issues, whether real or imagined, another reason we’re not seeing proof-of-payment in the USA is because there are just too many bus stops – double or triple the amount you would see in a typical European bus system. We can’t even properly maintain shelters, signs, and schedules at all our bus stops – imagine the challenge of maintaining a ticket machine at every stop too.
Has anyone done a comparison of fare evasion rates on POP systems between US and European systems?
When I review a fare evasion rate in a US system, I always wonder whether there is enough fares being checked to create a stastically signficant sample size. Do systems in Europe have similar reporting problems?
The POP systems ob buses in the US are primarily focused on BRT lines like Las Vegas, Orange Line in LA and the MAX line in Salt Lake City. These lines have TVMs at the station boarding areas to allow you to purchase a ticket prior to boarding. There is no doubt that POP does speed up the operation of the buses.
Muni in San Francisco has talked about expanding the POP to the bus system. But it always boils down to how much revenue you will lose versus any improvement in service. That’s why it is important to have sound data and analysis before trying to expand POP systemwide.
We can’t even properly maintain shelters, signs, and schedules at all our bus stops – imagine the challenge of maintaining a ticket machine at every stop too.
Which is why onboard TVMs is a prerequisite for POP on busses.
I'm actually not keen on on-board TVMs. The Parisian system is that if you need to buy a ticket you board at the front and deal with the driver, but multiple ride tickets and passes are priced to encourage uptake, so not many people need to do this. I have some acquaintance with the on-board TVMs on Melbourne trams. There's only one on each vehicle, so using it requires a lot of movement within the vehicle, raising the same issues in a crowded situation. And let's face it, TVMs have lots of problems, and if you don't encounter the problem until you're already on-board then you have an excuse for evasion and it's easy for real evaders to take advantage of that.
The on-board situation is always going to be cramped if not crowded. Much better to put TVMs at major stops, and have the driver handle things (exact fare only, into a safe, as in the American style) for the few people who still need to pay case.
I know very little of transit in USA and even less in SF, but I find looking only at the cost side of things (fare evasion) strange. The benefit ofcourse is that it speeds up vehicles and improves reliability (less delays due to crowded stops) which should decreases operating costs and at the same improves service which should mean more passengers and thus more revenue. The latter is a benefit assuming more passengers generate more revenue then added costs of course. The benefits are what I was after with indirect costs of fare collection, it is difficult to quantify but not IMO impossible.
I don’t know if Singapore ever officially switched to POP, but even after its buses were 100% POP, it took a while before the city switched to ezLink and another while before it placed card readers at bus stations.
There’s actually a better argument for POP on buses than on trains, which is that at high crowding levels, faregates are cheaper than POP.
Various notes on SFMuni :
1) In the distant past SFMuni had fare collectors that stood ON THE SIDEWALK at the rear door, held it open, accepted payment, and issued transfers before you boarded. This was at key stops in the downtown (e.g. Sutter and Market) during the afternoon commute. This was prior to the development of the Fast Pass. Their equipment included a counting pod w/ pull ball on the bottom, a change dispenser (pre-dollar fares), and a hole punch for marking transfers.
2) The Fast Pass has saved SFMuni a lot of time on the fare collection part of the process – just flash and head on back. It also generates a small amount of interest due to the resulting early collection of fares. Unfortunately, I can’t find a link to an article or study that confirms these points. It’s something I read in a newspaper article years ago.
3) POP on the SFMuni is a muddle. The proof part of POP applies everywhere. But all-door boarding only applies to the LRVs except when an SFMuni employee is at the back door(s) acting as a gatekeeper. Of course, one will find a driver being practical at a high load stop and doing a Nelson (remember Copenhagen ?).
4) The ninety minute rule may get modified after San Francisco transitions to the Clipper card so as to cope with various outages. One possibility would be to have the fare inspectors (but NOT the drivers !) tack on sixty minutes to compensate for the delay. Something a hard ninety-minute rule doesn’t handle well is a multi-transfer ride that’s infrequent-frequent-infrequent. If there’s a missed run on the final leg then the rider may have to pay a second fare. And getting compensation for that second fare would be way more costly in time than the 2nd fare.
5) SFMuni needs to roll out their own Wi-Fi network, They have the real estate and the conduits. This would make their communications (voice, location, Clipper) a lot more reliable. It’s a pity the mayor’s plan for a municipal network was so lame.
While only the bendy buses in London have POP, a lot of routes served by regular buses require off-bus fare collection in central and busy areas. At some stops, you have the situation where some routes require off-bus payment while others don’t (indicated by a yellow background rather than a white background behind the route number on the signs listing the bus routes that stop there).
It’s a solution that falls between the two – entry is only via the front door, but it’s fairly quick as most people are just swiping their Oyster cards, and those that aren’t are giving the driver a paper ticket. (No transfers to worry about – they aren’t available, which is annoying). Only further out, along the less busy parts of the route, does the driver need to worry about taking money and giving back change.
The vast majority use Oyster cards as it is a lot cheaper – around half the price. But I’ve noticed elsewhere – heavily tourist oriented Quebec City, for example – where even if they don’t have smart cards, there is still a very advantageous price difference in using prepaid tickets rather than paying cash. (In Quebec City, you can buy single tickets at convenience stores, so it’s not that it’s a bulk discount, simply an incentive to speed up boarding).
Just to clarify this sentence – “even if they don’t have smart cards” – by “they”, I meant that the transit authority has not yet adopted smart cards (rather than it being the case of individuals opting not to use them).
Which is why onboard TVMs is a prerequisite for POP on busses.
What Jarrett said – there isn’t enough room for all that pushing and shoving to get to it.
Plus, I’ll use any excuse to cut the ridiculous number of bus stops that characterize every American bus network in half, or more.
I understood POP was harder to enforce in countries where ID cards are not compulsory. France, Germany and many of the European countries where POP is pretty standard have a legal requirement for every person to carry ID.
In Australia, UK and NZ ID cards are not mandatory and there are I understand legal limits to what inspectors can do if they find fare evaders on a bus, other than make then get off the bus.
The situation is a bit different on some Australian rail systems where transit police have police powers. Legally it may also help that rail operations do not operate on a public road but rather a rail corridor with its own relevant acts or parliament.
It was many years ago I recall being told of legal issue so can’t back it up with any links.
When I first lived in London 20 years ago, there were lots of scary notices about fare evasion and fines up to £1000 (or some such – the minimum, first-time fare was around £50, written in smaller print). Now that pretty much everything but the bendy buses are gate controlled, they’ve calmed down a lot – penalty fare is still £50 but no warnings about scary max fines. If the inspector believes you are wilfully dodging the fare (rather than making an honest mistake), they can issue a summons. Not sure if the inspectors are part of the British Transport Police – you will sometimes see them backed up by ordinary police though.
Details on TfL’s policy here: http://www.tfl.gov.uk/tickets/14436.aspx
In my local area, the relatively short buses have two sets of doors. The problem is that the second of doors reduces seating (and standnign) space, but are needed beauase the front doors are too narrow to allow simultaneous boarding and alighting.
For Paris, I suspect some stickers labelled “keep right” (in French) would help things enormously.
POP and ID cards… the UK national rail system operates on a de facto POP system, but there is no requirement for you to carry ID. However, if you are asked for ID and you have something offical with your name on it, then you have to show it. Given most people these days have credit/debit cards, it’s quite unlikely that a fare dodger would be completely unidentifiable.
Alex, not all European countries require people to carry ID on them. Germany requires people to have ID, but not to carry it with them. Finland has voluntary ID cards. Denmark does not have any ID cards at all, other than passports and drivers’ licenses, and does not require people to carry them.
PARIS is a great international icon place with fabulous nightlife..So moving to Paris a city of love…!!!!Great post!!Fantastic vehicles!! MOVING TO PARIS
Perth (Western Australia) has central city “CAT” buses which are free on three routes, and these board at both sets of doors…. I know it is easier since the bus is free, but my next door neighbour is a bigwig for the public transport authority here, and he mentioned that the smart card ticketing system now has around 80+ percent of the “market”, and that they are opening up their ideas on having things like all-door boarding on the buses, especially in certain busy areas.
Perhaps some people remember that Tri-Met in Portland Oregon once had this sort of system (during the period when they first got some “bendy buses”) but abandoned it after a while – though I don’t know why.
@Dan. Proof-of-payment was tried briefly in Portland in 1980, and was abandoned within a year or so as I recall. I think it was a mistake to do it on the entire network, including many routes that were not especially frequent or crowded at that time. As a result, inspectors had a lot of time on buses with nothing to do, or waiting for buses, and the overall credibility of enforcement was just not there.
Proof-of-payment really makes sense only on busy services where inspectors can check a lot of fares fast. That's also the context in which all-door boarding and alighting saves substantial time.
Jarret. Those railings are in doors primarily to prevent boarding of strollers where there isn’t enough place for them inside.
Chris. I don’t know specifically about Russia, but in Czechoslovakia – another country of former soviet block – the conductors were gradually replaced by POP in late 1960’s and early 1970’s.
Hyjal. The real safety problem is if impaired person boards in the rear part of vehicle and driver pulls out of stop before he/she sits down. Buses are short enought to allow to see all doors in the mirror safely.
Scotty. I do remember some lines that run partially POP and partially with driver-collected fares but unfortunately, I don’t remember how the driver-POP transition was done.
Ad ID cards – fare evaders don’t carry them (or credit/debit cards) with them even in countries where it is mandatory. In such case, the inspectors hand the evaders to the police, who will identify them.
“if [the railings] were used for opposite directions of traffic (at intermediate stops where both directions exist) this situation would move a little better.”
Are you sure? After all, in case of symmetric boarding/alighting, the time needed is exactly the same, except it is easier to board the bus when there are the fewest people (i.e. all have alighted). And – as you mentioned – it wouldn’t be beneficial for asymmetric stops.