Is it fair to have to pay more if your trip requires a transfer or connection? I’ve argued that it isn’t, but I also have an appreciation of the difficulty of eliminating these penalties. So when complaining about a fare penalty, try to understand the situation from the transit agency’s point of view. Not because they’re right and you’re wrong, but because you many need to help them solve the problem that it presents for them.
Definitions first. A connection penalty is any element of a fare policy whose effect is to charge people more for a trip that requires two or more transit vehicles than for a trip using only one, even when the trip is of exactly the same distance and when there’s no evident difference in class, amenity, or inefficiency to justify a surcharge.
For example, in the Sydney Morning Herald’s inquiry on Sydney’s public transit, I prepared this table showing the difference between rail fares and rail+bus fares for trips of exactly the same distance:
My basic view of connection penalities has always been this: The connection is not an added convenience for which the customer should pay extra. It’s an inconvenience required by the geometry of any efficient, legible, and frequent transit system. Transit systems must therefore do what they can to minimize this inconvenience and encourage people to make connections. Connection fare penalties are exactly contrary to this imperative.
As near as I can tell, most fare penalties for connecting arise from some mixture of four basic issues:
- Practical security problems around the physical tools used to record connections, such as the opportunities for fare evasion created by abundant transfer slips. If this is the primary reason for a connection penalty, smartcard systems will eliminate the problem and should lead to removal of the connection penalty.
- Laws and funding arrangements that refer explicitly to the cash single-ride fare. A key funding formula that governs Los Angeles County transit agencies gives these agencies more subsidy if they push their base fare downward. The formula is actually trying to reward ridership, but because ridership is hard to measure across multiple agencies, it estimates ridership as the fare revenue divided by cash single-ride fare. This penalizes agencies for raising their cash single-ride fares; it instead encourages them to invent other fees and fare penalties for reasons that have nothing to do with the real intentions of the formula. Again, smartcards, which will provide a more accurate and direct measure of ridership, should eliminate the need for such approximations and the perverse incentives that follow from them.
- A journalistic obsession with single-ride cash fare as the key measure of fare cost. The single-ride cash fare is easy to talk about, and journalists and other non-transit-experts tend to focus on it, producing soundbites such as “Los Angeles fares are cheaper than San Francisco’s.” Prevalence of these soundbites motivates agencies to try, for reasons of appearance, to keep their cash single-ride fares down. Again, this is just like airlines: If you become obsessed with reducing your base ticket price at all costs, you’ll have to invent more and more fees. Connection penalties are one fee of that type.
- Conflicts between the self-interests of geographically entangled agencies or operators. In multi-agency or multi-operator environments, the introduction of a free connection can come at the expense of the bottom lines of multiple operators, and tends to affect different operators differently. These problems are sometimes overcome through heroic negotiation, in which the operators agree to shift money among themselves in complex ways in order to present a simple integrated fare system to the customer. In other cases (as in Brisbane) the solution is to create a new meta-agency that sits over all the operators, pays them for operations, receives their fare revenue, and takes responsibility for a single region-wide or city-wide bottom line.
So it’s hard. It’s important. Good transit agencies are at least thinking about the problem, but it’s still hard. In multi-agency cases, for example, the problem may be impossible to solve without creating a meta-agency, with all the attendant anxiety about “adding another layer of government.”
What can a citizen do? One thing is to confront journalists and others who try to make simplistic comparisons based on cash single-ride fares. Journalists: if you want to do quick pieces about fares, look for average fares (fare revenue divided by ridership) rather than published base fares. Yes, this is harder because not all agencies can calculate ridership reliably, or (as in Los Angeles) they may actually be relying on fare revenue to get to a ridership estimate. Again, though, smartcards should eliminate that problem.
Finally, since I’ve now referred twice to smartcards solving the problem, let me be clear: Technology can solve information problems, but not ethical ones. Connection penalties based on the inadequacy of information-tracking tools should fade away once smartcards provide the necessary tool. But technology won’t change the political problem: to eliminate fare penalties without losing revenue, you will probably have to raise the single-ride cash fare to compensate. Effectively you’ll be admitting that you’ve been subsidizing some riders at the expense of others. You should expect all the blow back that we see anywhere when we tell people that we’re removing a subsidy on which they’ve come to rely.
That’s the real problem. Courageous leadership is the only solution for it.
The base cash fare thing has been a problem for at least a hundred years. The five cent fare was a huge deal back in its time.
Jarrett, Re: “To eliminate fare penalties without losing revenue, you will probably have to raise the single-ride cash fare to compensate”
In most North American systems, a better solution would be to start charging by distance. The base single-ride fare is not too low for a short, 1 mile trip, but it is far too low for a 20-mile, cross-region journey. In Los Angeles it only costs me $1.50 to travel over 20 miles from Long Beach to Los Angeles on the Blue Line, while a short, rail-to-bus trip would cost $3.00 one way. And day passes are $6.00, which basically prices the average trip as $3.00, or one transfer.
Instead, Los Angeles Metro could have a $1.00 base fare for trips under 1/2 mile, plus $0.25 a mile off-peak, with higher peak travel fares. 20-mile trip from Long Beach to Los Angeles would cost $6.00, similar to a regional rail trip of the same distance. And at that price I believe Metro could get over a 60% farebox recovery ration from the Blue Line service (which costs $0.38 per passenger mile, with average trip length about 7 miles) assuming ridership only drops 25%
Brisbane pays operators on a per-service km basis. Connections are FREE. Go Brisbane!
“Effectively you’ll be admitting that you’ve been subsidizing some riders at the expense of others”
Yes, and many of those that would benefit are potential future riders, whereas those that lose are current riders. Why suffer bad publicity for a constituency that doesn’t yet exist?
Thinking about this in terms of Seattle’s smart card system (ORCA), I think the solution is pretty straightforward:
* Have a smart card system.
* Time-limit every smart card fare so that you can only be charged once every 2-3 hours or whatever the normal transfer/connection window is no matter how many vehicles you use in that window. I think this is already done in Seattle with ORCA with some exceptions for connections between particular services.
* If cash payment must be retained, offer only one cash option: an all-day pass, set at roughly a two-way single peak fare with perhaps a slight discount. In Seattle, that would be about $5.
* The system could be simplified further by simply charging the all-day fare to everyone using either cash or a smart card purse. So you pay $5 if you use transit any amount in a single day. You’d be charged once the first time you entered the system each day, and never again.
* You could still have monthly (and ideally, weekly) pass options that provided a slight discount. It might even be simpler to do away with passes entirely and just institute weekly and monthly maximums. So in the end, you’d have a daily fare ($5), a weekly maximum ($25, even if you ride all seven days), and a monthly maximum ($100, even if you ride every day in the month.)
In the case of ORCA, there’s a $5 charge for the card anyway, so it would be quite easy to just charge the $5 the first day and then on every successive day. The same cards could be used for both temporary single-day use (for tourists and infrequent riders) and for regular transit users.
For discount riders you could allow discount fares by providing youth/senior cards at TVMs and retailers, and force all riders who pay in cash to pay the full $5 fare.
The one drawback I can see with this kind of simplified system is that it provides far-flung suburban riders with a hidden subsidy, because they pay the same amount for services with much higher operational costs. On the other hand, such riders are often only going to use buses to get to and from work in a city center, while urban riders are taking shorter trips more frequently. It probably comes close to balancing out. And even if they do get a better deal in some cases I think there’s a value in encouraging suburban residents to use transit rather than driving.
The way Brisbane has chosen to do it is a bit different to other Australian cities:
* Free transfer – 2hrs or so
* Abolish all passes- monthly, weekly etc- gone
* Increase paper ticket costs to be more expensive than the smart card, and then abolish most paper tickets
* Come January, only single paper tickets (one way) will be offered and they will be more expensive
* Off-peak discount
* Capped fares on seniors card
The other side of the coin is that technology can create ethical problems. As much as I think ORCA in Seattle is great its introduced a penalty around transfers in the morning. It has a hard limit of 2 hours for a transfer. This generally works great, however if you used to pay a cash fare at say 8 or 9 pm you’d get an OWL transfer, which while opening up abuse makes some more sense because connections aren’t as well timed. Its especially a problem for people who take a really late bus, and perhaps make a connection that is 3 hours.
I just think in many ways ORCA is too simplistic in its fare implementation, although I’m not sure if the implementation constraints are for technical or business reasons…
The solution is trivial and simple and easy to realize. Proof of payment.
Why is that, what works in most of Europe very well, such a taboo for the English speaking world?
Forget smart cards! Forget transfer slips!
It is in the basic interest of a transit company and its passengers to have fast boardings at the stops. Without people getting their smart cards or wallets out and standing at the readers while others wait behind and block the door.
In most POP systems you have most riders with a monthly pass. The few left who have to pay can do that either at ticket machines or at ticket machines on board of the buses. No tickets should be sold by the driver, the machines accept cash (change given) or bank/credit cards. That makes boarding fast, and most people (who have a monthly ticket) can simply leave it in the pocket. Access to the system is easy, you use it without thinking too much.
And when a single ticket on paper has a time code on it and you are allowed to use it for 2 hours within a city or 3 hours including the suburbs the problem of transfers is solved.
From time to time you have ticket inspectors (usually in plain clothes) who can check a paper ticket with a time code very easy, as well as a monthly ticket.
Not everything must be done with electronic gimmicks just because you can do it. A casual user, maybe a tourist, does not need to get a smart card. And daily users have their monthly ticket – sold also at ticket machines.
Discounts for senior citizens are usually expected by the politicians, but I do not see a reason for that. In Germany the average senior citizen has more money available than younger people under 40. So it is not even fair. Also you will not find gas stations giving a discount on gas for senior citizens or supermarkets on chocolate. Discounts for senior citizens follow the stereotype of public transit being some kind of welfare for the poor. Theoratically it should be better solved by the government giving people in need enough people.
And off-peak-discounts – even if the marketing and sales people kill me now – they are nonsense.
No one plans his trips within a city basing on discounts in public transit. If you are so price sensitive you will ride your bike. I doubt that someone rides an overcrowded bus at 7:30 in the morning if he has not a good reason for it.
An off-peak-discount makes – from the point of a company – only then sense when you keep the price for a ticket during off-peak-hours and increase the fare during peak hours. Than it is only a good excuse to get more money from your riders. But it makes travelling only more difficult for the public transit user. The car has always the same price and being in a crowded bus is already enough as a “peak hour punishment”, you do not need to charge extra money for it. It is an illusion to think you could get a better distribution of the riders over the day with this.
In addition to my last post:
There is one sentence where the word people should be read as money, you will know what I mean.
And: the best off peak discount is already that you have (hopefully) a good chance for a seat and less body contact with other riders.
Who do you think will – lets say – tell his dentist, he can only come after 10 in the morning cause then he gets a disount of 10 cents on his bus ticket? You only make your system more complicated, more than driving a car.
@TransitPlannerMunich – the price of a car is different at different times of day.
If you just stick to what you’ve gotta put in, a car is more expensive during rush hour because it will take you more time, as well as more gasoline due to lower speeds and extra idling.
Also we do use proof of payment in the states, however it isn’t always practical for every system. Installing and managing ticket machines at all the possible places to board transportation would be unworkable.
I rather like the system they use in Perth. The system here is based on zones, with higher fares the more zones you are travelling. Tickets for 3 zone trips are valid for 2 hours, more than 3 zones for 3 hours. You can make as many transfers (or even a return journey) as you like within the time limit. So long as you board your final transit vehicle within the time limit, your ticket is valid for the whole trip.
There is also a short trip fare which is single trip only for trips of ~2km or less. There are reduced fares for seniors, unemployed, children and full-time students. The seniors fare is only available outside of peak hours.
About 5 years ago, they implemented a contactless smart card payment system. There are two ways to add value to the card – a cash payment at any card salespoint (ie nearly all newsagents, and all train and bus stations), or via automatic direct debit. If you refill through cash, your fare is 15% off the cash fare, and automatic refill is 25% off the cash fare. You have to validate the card on boarding and alighting each transit vehicle (or for trains, entering and leaving the station). I have stood and timed the boarding of a peak hour bus, and 60 passengers boarded in less than 3 minutes, including 3 who purchased cash fares from the driver (and the 3 minutes would have been 2:30 without the cash payers……) The average boarding time for those with cards was ~2 seconds per person. The cards have a small aerial embedded, so you only have to wave your purse/wallet near the reader for it to register.
The smart card system calculates the cheapest fare available for your travel as you tag on and off each trip. If you forget to tag off, you get charged for the longest trip possible within the system.
@Nicholas Barnard:
Such ticket machines can easily be installed in every bus:
http://upload.wikimedia.org/wikipedia/commons/6/61/Fahrscheinautomat-MVG-Bus.JPG
You can pay with cash or a bank card. Change given.
Funny enough I was during the summer for some days in Tiflis (Tbilisi), Georgia. Even in the Caucasus they have now ticket machines installed in all city buses (behind the driver!). Simple ones, but you put money in, get a printed ticket, a POP ticket. You board at all doors and present your ticket only to ticket inspectors in case they come. The driver is only there to drive the bus.
The system is not perfect yet, especially regarding transfers, but it is a beginning and a base for a proper POP system.
When they can do it, why can any other city in the world not do it? One ticket machine per bus is not so expensive as you might expect, it might even reduce your operating costs:
I can tell you a secret: we do it cause it does make sense from the economic point. The less time a bus stands at a stop, the faster the boarding is, the less amount of vehicles you need to operate a line. And also it is easier for the passengers, you get more of them. Even if some of them don’t pay, over 95% will pay the fare, especially when they have a monthly ticket.
Yes, it works. How much of the operating costs of an American public transit company are payed with the farebox revenue? Even in New York?
In a country like Germany it is typically well above 70%, in Munich 100% of operational costs are covered by that what is paid by the passengers and about 95% of all costs including investment. Having an easy system for ticket, boarding and transfers is vital for the economic stability of such a system. How else could you run an urban bus line with 13 km length, transporting 10 million passengers a year (or over 30,000 daily) and that with just 11 vehicles running on that line? When every passenger must pass through the front door, produce a ticket or smart card, show it to the driver or to a reader, we would end up to need more vehicles, drivers etc. and have higher costs. Efficiency is the key, and comfort for the rider a result of it. And also the trip is then faster for the rider, what also increases the attractiveness of such a line.
And about your statement of a car being more expensive during peak hour cause of more time needed and extra gasoline for extra idling: Also you spend more time in public transit during peak hour, but the difference is that in your own car it is comfortable, with the temperature you prefer and with a nice seat. The extra gas for idling is nothing that is on the mind of a typical commuter. But to pay more for a peak hour ticket will be obvious for everyone – in a negative way. Why do supermarkets not charge extra 10% during times when many people come to shop and they need extra staff?
@numbat: when the boarding of 60 passengers would need 3 minutes here, I could forget all stability of a line here. What a nightmare, even if it positive from your perspective! This is the reason why we have now articulated buses with 4 doors and boarding at all doors simultaneously. We do not have buses for idling and standing at the stops, we want to have running buses. Time is money, also for a transit company.
Just short calculation:
If you have a bus line with 10 km lenght in every direction, 30 stops in each direction and at each stop in average 5 boardings and a headway of one bus every 10 minutes, you can save the operating costs for one bus already when every passenger gets on the bus in average 2 seconds faster. By extra doors, boarding at all doors, better bus stops, low floor vehicles, POP, priority at traffic lights etc. you can reach much more. And of course a clever bus stop spacing. So your operating costs can go down easily AND you can attract more passengers. When you forget old rules and traditions. It pays off.
Hmmm…trying to figure out how the farebox is supposed to know how far your trip is going to be? I think paying a transfer fee makes more sense then being charged full fare each time you get on a bus, making 2-3 bus changes extremely expensive in comparison!
Transit companies can change bus routes, giving some people a single seat ride who used to need to connect and forcing other people to connect who used to have a single seat ride. (Normally, with proper planning, the number of riders impacted are low relative to total system ridership, but on a real number basis may be significant in larger systems.) From a social/ethical standpoint, therefore, connections should be free (or, perhaps, free for a “short” connection). Zonal fares should be charged for longer trips.
Issues: 1. If zonal boundaries are fixed, the person who travels 1/2 mile, but crosses a zone boundary, pays more than a person who travels four miles but does not cross a zone boundary. Sophisticated smart card systems can address this; but riders have to be thorougly trained to swipe both on boarding and on exiting.
2. Proof of payment is an alternative to smart cards; but it is extremely difficult to administer in a zone based system. For example, a person has a one zone monthly pass or one zone fare receipt and is riding in the second zone of a two zone bus route when a POP inspector appears. Did the passenger board in the second zone or did the passenger override and board in the first zone? (The fare receipt could be stamped with location of purchase; but there are some issues with this, also.)
3. In really large urban areas [or areas where bus and rail fares are set at different levels for the same distance (usually because buses provide the frequent stop service when rail or light rail provide the express service)], transfers between modes often raises equity issues.
4. Most transit agencies currently count bus passengers on an unlinked trip basis. If smart cards were to allow the count of passengers on a linked trip basis, ridership numbers would fall. Support for transit among the political class could diminish.
@Steve Lax:
About point 2:
In a POP system it does not matter. No matter if you override a zone or if you board in the wrong zone, if a ticket inspector catches you with an invalid ticket for that zone, you pay 40 Euros fine (here we call it translated “increased delivery charge”). Everyone knows that price. So someone with a monthly pass can buy an extra ticket for the additional zone at the ticket machine.
About point 3:
One ticket, one price for all modes of transit is the best to handle. And works in most European systems without a problem and without anyone feeling discriminated or unfairly treated. Moreover in a big city system usually a bus rider is more expensive for the agency than a rail rider. And for the passenger it is easier if you have one price, one ticket, cause you do not have to think in categories like subway, tram, bus but simply “transit”. Equity issues can be risen by a lot of things in our western societies. Being able to use a tram, a subway or a bus with the same ticket and for the same price is not among them, from my perspective.
@Steve Lax
“2. Proof of payment is an alternative to smart cards; but it is extremely difficult to administer in a zone based system. For example, a person has a one zone monthly pass or one zone fare receipt and is riding in the second zone of a two zone bus route when a POP inspector appears. Did the passenger board in the second zone or did the passenger override and board in the first zone? (The fare receipt could be stamped with location of purchase; but there are some issues with this, also.)”
I’m not sure I see the problem.
In Metro. Vancouver the system is broken up into 2 zones. Now if you purchase a ticket on the bus or at a station. It stamps the time it was purchased and the zone. It also gives an expiry time. Which is 90 minutes at point of purchase. If someone has ticket that is stamped zone 1. But they are found to be in zone 2. And they have no other ticket to prove that they added fare (bought a ticket to cover the difference in cost in crossing into another zone). Then they get a $173 fine.
Same with monthly passes. My monthly pass is crossed off for zone 1. If I happen be in zone 2 when I should have payed the extra fare. And I haven’t they I would get a fine.
Metro Vancouver
3 Zones.
Same fare to travel anywhere in 1 zone.
Going into a 2nd or 3rd zone costs more.
Every ticket gives you 90 minutes to go anywhere. You can get on and off as many times as you like. So long as the ticket is still valid for the zone you are in.
There is not extra charge to get on the rapid transit system. The only train that costs extra is the West Coast commuter train. But that only runs M-F during peak hours.
After 6:30pm and on weekends there is not extra cost to cross into another zone.
Also those with a U-Pass get unlimited all zone access.
There is no difference in price nor any extra time on the expiry. Whether you purchase a ticket at 8AM in the middle of rush hour or at 3AM when only a few select night owl routes are running.
I think it’s pretty common to offer POP discount in the German-speaking world. Berlin offers a reduced-fare pass valid only after 10 am, and I think in Switzerland you can buy reduced-fare passes valid only after 9.
In the German-speaking world, POP is actually bundled with zone pricing fairly well. The zones aren’t complex – even Berlin only has three – so you know exactly what zones you’re allowed to be in with your ticket. An inspector who boards a train in zone A and catches someone with a pass for zones BC will immediately know the rider is dodging the fare. It’s much easier to remember than the station-to-station distance-based pricing common in Asia.
But that doesn’t mean smartcards and POP can’t work together. On Singaporean buses, the inspectors board with handheld card readers, which confirm that everyone on the bus has tapped in but not out. Smartcards aren’t strictly necessary, but they simplify inspections on systems where the fare calculation isn’t simple.
@TransitPlannerMunich and @Paul C –
Jarrett’s blog and the comments in response to his posts are useful both as a discussion of best practices in the transit industry and in expanding each of our individual knowledge bases. At the same time, we all must remember that each of our comments is based in large part on the transit systems with which we have had the most experience. And each transit system is unique in some ways, from historic development to its politics to its topography to the size of its geographic territory and so on.
Accordingly, my comments to the issue of whether POP could work in a multi-zone system were with respect to a much larger zone system than Vancouver or Munich.
For sake of argument, envision a system five zones wide by five zones long with each zone being five miles by five miles and identified as zones 1 through 5 east-west and zones A through E north-south (as a map grid) so zones went from A1 in the northwest to E5 in the southeast of the transit territory. Because of the many permutations of zone travel within the territory, the transit property sells monthly passes good for one through five zones. Under the current system, a passenger boards a bus in a given zone (say Zone B2) and shows the bus the pass. The operator rings up a fare receipt based on the number of zones (let’s say 2) on the pass and the zone in which the passenger boarded and the zone to which the passenger can travel. If the bus is traveling from Zone B1 to Zone B4, the fare receipt would read B2-B3. The operator is responsible for checking the fare receipt when the passenger disembarks and needs to collect an override if the passenger disembarks beyond Zone B3 (unless the passenger paid the override on boarding.) Obviously, this is a time consuming process that smart cards could improve. (For connections, however, since the pass is a generic two-zone pass, the passenger could make up to a two-zone connection at no additional charge.)
Now assume a POP system without smart cards. Let’s say the inspector boards in Zone B3. There is no way for the inspector to know whether a one zone or two zone monthly pass holder boarded in zone B1, zone B2, or zone B3. Thus, my comment that POP does not work well in a multi-zone system.
And now, my reality. The system with which I am most familiar is 43 zones long by 21 zones wide with typical zone about four miles by four miles and low first boarding zone fares to assist low income transit dependent people. The longest route is 160 miles and 39 zones long; the shortest route about one mile long. The longest commuter route (15 minute or better peak frequency and hourly service for 16 hours or more, including weekends) is 76 miles and 23 zones long. Of course, most routes are five zones long or less. The system handles about 550,000 bus boardings daily.
Steve, the POP systems that work best do not have this concept of zone. First, their zones are concentric and not gridded, but that’s not too important. Second and more relevantly, the commuter passes don’t just specify a number of zones, but also the specific zones you’re allowed to be in. It may be that traveling 3 zones costs the same no matter which 3 they are, but the tickets still say Zone ABC, BCD, etc.
Which system is 43 zones by 21?
That’s how it works with TriMet–tickets, including passes, are available in All-Zone or 1-2 Zone forms. (TriMet has three concentric fare zones). If you buy a multi-use 1-2 zone pass, you have to specify whether the pass is valid for Zones 1 and 2, or Zones 2 and 3. (Holders of a 1-2 zone pass can use their pass on a all-zone journey by paying an extra 30 cents).
@Alon of 11/29 –
To answer the question – “Which system” –
NJ TRANSIT. (The 43 zone number is a rough estimate as no route goes 43 zones. I took the 39 zones of the NYC-Cape May route and added four zones for the distance north of the Lincoln Tunnel covered by routes serving Bergen and Passaic Counties. If you include the route into Orange County, NY the zone count could be higher. The 21 zone number represents Philadelphia to Atlantic City. Actually, there are five separate layers of fare zones. Northern NJ intrastate, northern NJ interstate, northern NJ long distance intrastate, southern NJ intrastate, and southern NJ interstate.)
As I noted in my earlier post, the large number of permutations of possible zone travel in this large gridded zone network would make it extremely costly to issue unique zonal passes; so generic zone count passes are issued.