Singapore’s weekend Straits Times was full of debate about the recent fare system changes, which finally eliminated fare penalties for connecting from one service to another.
Eliminating these penalties is a crucial step in creating an integrated and versatile transit network, because (a) networks designed around connections are more legible and frequent than those that aren’t and (b) transferring is already enough of a hassle without these penalties. The new system means that your fare from A to B will now be the same regardless of the path you take and the number of times you transfer. This, in turn, will allow the transit agency to design a simpler and more reliable system.
To take just one example, many bus lines duplicate MRT rapid transit corridors, partly because fare penalities for transferring encouraged riders to stay on these buses instead of change to the MRT. This was always silly from the transit agency’s standpoint. Remember: To get a rider to her destination sooner is not just good for the rider. It’s good for the transit agency, too, because the operatings costs of transit are measured mostly in time, so a faster trip uses less of it.
However, like many (but not all) great ideas, eliminating these penalties isn’t free. When you eliminate a fare penalty that is bringing in a certain amount of revenue, you have two choices: increase the base fare — which is a direct fare increase for people who aren’t transferring — or accept a drop in revenue. Singapore’s transit agency did a little of both, and therefore got attacked from both sides. The new fare system has been running for a couple of months now, and inevitable complaints about the fare increase (on passengers who didn’t transfer) are in full swing.
I’m impressed with the Straits Times coverage of this issue — to which I’d link generously if it weren’t behind a paywall. The reporter actually understood the financial reality underlying the choice and explained why other choices might have been worse. The prevailing critique of the transit agency was not that they did it, but that they oversold the change as a fare cut. In fact, it was a cut if you were already transferring and an increase if you weren’t, and they probably could have been clearer about that.
Still, this is a good move. Singapore is in the early stage of rationalising its huge but confusing bus system. This was an essential step toward making that possible. When New York City eliminated bus-subway connection penalities in 1996, the result was a small jump in subway ridership and a huge jump in bus ridership. I’ll be interested in seeing Singapore’s results.
When you need to change between bus lines is a penalty in itself. You should get money back instead of paying more.
The easiest would be a fully integrated system: one ticket for all lines including metro and commuter rail. Most riders from within the city using monthly passes then, so it does not matter what you do with one trip tickets. And one trip tickes so expensive that frequent riders will buy a monthly ticket.
When there are complaints about a fare increase in Singapore after transfer penalties were eliminated it can only mean that a lot of people pay for each single trip. What is not very productive for a transit city.
Singapore doesn’t have any unlimited passes. The closest thing it has to a discount is that fares are much lower with the ezLink smartcard than with cash; since the ezLink costs $5 and has a minimum $10 initial stored value, it acts as a sort of frequent-user pay-per-ride discount.
I think it’s fairly common for Asian cities with robust smartcard systems to have no unlimited discounts. I don’t think Hong Kong has one, and I know that Shanghai only gives you a 10% discount over a certain number of rides per month. Tokyo has unlimited monthly passes, in fact with larger discounts over the base fare than in Germany, but the passes are only between specific stations, rather than within a zone.
I can think of three explanations for this. First, smartcard systems have very low transaction costs, so that the benefits of making everyone pay just once a month are reduced. Second, Asian cities have fares increasing based on distance rather than based on zones; this makes implementing unlimited cards somewhat harder. And third, East Asia does not learn from the transit practices of the German-speaking world even when they’re superior.
I don’t think the benefit of this arrangement is all that clear-cut. Light users have a legitimate complaint against having to subsidize heavy users. It’s a trade-off, and like all trade-offs, there are winners and losers.
The perverse effects of fare penalties for connections are profound: huge amounts of wasted service and a bus system that's too complicated to use for anything but routine trips. There is always a tradeoff about whether to do it via accepting reduced revenue or increasing the base fare.
I don’t think the reason Singapore doesn’t have unlimited fares is a fairness issue. The transit operators are private businesses and care about maximum profit, not about equal fare per passenger-km. As it is, light users already subsidize heavy users if they don’t use ezLink, riders who travel short distances subsidize riders who travel long distances, and subway riders subsidize bus riders.
Note that Singapore govt's intention is to shift to net cost
contracting, which eliminates the operator's dependence on fare
There is some more coverage of the issue in the Today newspaper this morning. Fortunately not paywalled: see http://www.todayonline.com/Singapore/EDC100823-0000029/Fairer-fares-message-lost-in-transition
I think the main reason Singapore has only very limited ‘season pass’ options (more than they used to but still very limited) is the franchise arrangement with the operators retaining fares and hence taking revenue risk.
It would be much easier to offer attractive season passes if the system does indeed shift to a system of service contracts with gross cost arrangements (public agency takes revenue risk and operators get paid for seat km or such like).
However, the current plan appears to be to introduce competitive tendering with net cost contracts. So it seems likely that the operators would still face some revenue risk. [These details are not finalised publicly but this is the sense I have from earlier announcements.]
I would have thought that net cost contracts would still make it complicated to offer generous season passes, right? [Jarrett. Did you mean net cost or gross cost in your comment above?]
I think better technology would eventually make it easy to charge fares based on distance – which seems most “fair” to me and which should eliminate the perverse effects you describe.
As for discounts, I approve if they bring in more revenue (which is the whole point in the private sector) but in transit discounts just seem to lead to more deficits and budget battles.
Rhywun. What you suggest is exactly what Singapore is now doing. The
political problem is the transition to that system, which is what
they're going through now.
@Jarrett: “What you suggest [charge fares based on distance] is exactly what Singapore is now doing.”
How has Singapore implemented distance-based fares on buses? Does everyone slide or tap a card both when boarding the bus and when getting off? Do you have to do it for every vehicle, or only at the beginning and end of the trip?
I would like Los Angeles to move to an integrated, distance-based fare system based on the “TAP” RFID card now being used by Los Angeles Metro. It would be ideal to be able to tap the card once at the beginning fo the trip, and then at the very end, with the whole system (including separate municipal buses) being proof-of-payment.
Ah, my apologies. I didn’t RTFA, but instead assumed you were describing a “flat fare” scheme. It’s not clear from the article but can I assume that riders who transfer are paying some sort of “combined” fare that represents the total distance they travel? That sounds fair to me. I still think that folks who travel just a few stops (i.e. some significant amount less than the length of the “average” trip) should get a price cut – assuming the system’s finances are otherwise sound (which rules out systems like the one in my NYC which are subject to the endless whims of politics for their very survival….).
If I remember correctly, when I transferred between buses, I’d tap at entry on both buses and at exit on the second bus. I wouldn’t get charged the maximum fare on the first bus for not tapping out.
People who travel short distances do get a price cut, but they don’t get a very large cut. As of 2006, the minimum fare was 63 cents and the maximum fare was about $2. This is a factor of about 3 in fares, in a system where the longest trips are far longer than 3 times the shortest trips.
The franchise agreements are widespread on buses, but the vast majority of buses are operated by SBS and SMRT, as is the entire subway network.
I think I have a link for the Straits Times coverage of this issue on Saturday (the article Jarrett would have linked to if allowed).
I hope this works for everyone.
Thanks for the link, Paul.
But honestly, the article reminds me just how much of a government lackey the Straits Times is. An American newspaper would probably have probed the 68% claim more, or done its own analysis to check what percentage of commuters would see a fare reduction. It would probably have also asked whether the average taken over all trips is weighted for ridership or not. It would not have used language like “The government can’t win.”
Alon. Certain Australian newspapers, on the other hand, are interested only in news of the form: “How did the government screw up today?” That extreme isn’t helpful either.
Alon Levy: An American newspaper would probably have probed the 68% claim more, or done its own analysis to check what percentage of commuters would see a fare reduction.
I admire your faith in the media, but I think most newspapers would have either have reported such a figure without analysis, or 100% attacked/suported the move. I think only a few would have gone off and done the calculations.
The only reason you’d probe the number is if you’d have reason to suspect it was (very) wrong.
“(b) transferring is already enough of a hassle without these penalties.” (from Jarrett)
“When you need to change between bus lines is a penalty in itself. You should get money back instead of paying more.” (first commenter)
I could have sworn I’ve seen you argue against this position here in the past. I know Alon has.
At least two systems that have won the APTA Outstanding Public Transportation System of the Year, HART (Tampa, FL) (http://www.hartline.com) and the L.A. Metro (http://www.metro.net), have no bus transfers whatsoever.
What they have instead are relatively inexpensive day passes, purchasable on the bus. HART’s base fare is $1.75, but it’s day pas is only $3.75.
Cheap all-day passes to be a great compromise to increase ridership while not having to waste time handling transfers for everyone who needs to change buses.
What is a fair system?
Obviously in the eyes of some commentators here a system where everyone pays the fair price for his trip that he caused.
Can there be a fair price?
1. You want that distance is an important factor, okay.
2. Then it would be important to make a difference if you use bus, tram or subway, all systems generating different costs.
3. Then you would have to make differences if someone uses transit during peak hours (very high demand, high price), normal time or evening (again higher price because of lower vehicle occupancy)
4. Then you make differences if it is a weekday or weekend, if it is a schoolday or not
5. Then some people will claim it would be fair to give a discount to school children or the elderly. But is it fair that then other riders will have to pay for it? Or the government – that means also people who do not ride transit?
6. Then you can argue it is important to include infrastructure costs, so that the car users will not pay for public transit. Or you can say the car drivers also need to pay for public transit cause they can be happy that their streets are less congestet because of some brave public transit riders.
7. Then you should ask for the income when you buy a ticket, to make it really fair.
8. Should frequent riders get a discount? Some people here say no, it would be on the costs of infrequent riders. But in the normal economy companies and shops do give a discount for people buying big packages, a dozen cans etc. – and they do that cause it works.
Sorry, in that way you ruin the public transit. If a private, profit oriented company would think in that way it would not sustain.
First is the basic question what do I want to achive with public transport. If it is more than bringing welfare recipients and school children across the city, then I should take into consideration:
– If a main goal is to change the modal split and get people out of the cars and onto public transport because you think that improves quality of life in a city, saves the environment or reduces congestion then you must make it attractive for people.
– I am in competition with the car. Every mile travelled by a car more makes the mile relatively cheaper, cause I have the car already and just need the gas.
– At least here in Europe, different businesses like to go to some kind of flat rate – telephone/cell phone/internet etc. and even some clubs and discos for their drinks. They do that cause if pays off to make their products so easy to use as possible. I don’t want to think about tariffs, fare structures, if I pay more when I do another trip, if I should rather disembark one stop earlier to save money etc. If I want to promote frequent use of public transit I should get people to use monthly passes.
– Once you have a monthly pass you are more likely to use public transit on a daily basis. A transit company can calculate with you, moneywise and from the amount of vehicles needed. A transit company that has a lot of infrequent riders has sooner or later a problem. A lot of people who use public transit only when it rains cause a problem. People who use public transit only for their commute to work cause a problem cause during most time of the day the vehicles will be empty. So the more stable you can make a system the better.
That means, you should punish infreqent use by relatively high fares and encourage daily use by monthly passes. What also makes it easier for the passenger and for the company. Cause when someone buys a monthly (or in Europe even a yearly pass), then the transit company has the money even in advance and can operate with it.
This is maybe not fair. It is just logic.
Please get people’s positions right. Jarrett’s position is that every transfer incurs a penalty, but it can be made very small, small enough to be outweighed by the increased service simplicity and frequency. My position is the same, but also that there exist zero-penalty transfers, but only timed cross-platform transfers between trains.
In Singapore, you can’t time the bus-train connections. It’s not suburban Zurich, where everything is on a predictable takt. Everything runs frequently in the city. In addition, the MRT stations are not designed for maximally convenient bus-rail transfers; some Swiss and German stations have learned to put bus stops as close to the tracks as possible to reduce the bus-rail transfer penalty, but in a dense urban environment such as Singapore, it’s not possible.
Because neither the subway nor the long-distance buses serve the university, I’d always have to transfer. It would take me about 1-2 minutes to make the transfer, and then I’d have to wait a few minutes for the train or bus to arrive. Depending on which line I took it’d mean a transfer time of 5-15 minutes including wait. The German-speaking world has gotten this transfer time to less than 4 minutes, but operates in a different environment.
In Singapore, parts of what you say do not apply. The reason is that there are no choice riders on transit. High car taxes and congestion pricing keep the roads clear enough that transit is almost never faster. As a result, nearly anyone who can drive does: about 30% of households own cars, and about 30% of commuters drive to work. There are still benefits from unlimited cards, since they induce people to travel more and act as an off-peak discount, but there’s no point in changing people’s mode choice.
At this point, Singapore can’t even emulate New York, Copenhagen, or Paris and take lanes from cars and give them to pedestrians and cyclists to make driving more difficult. The arterials typically do not have any pedestrian activity on them, and have traffic speeds too high to make cycling safe. Unlike American cities, Singapore has not placed commercial activities on strip malls by the sides of arterial roads, but in enclosed malls in shopping districts.
I think the term “penalty” (in transfer penalty) is being overloaded to mean two different things, and these meanings are being confused:
1) Transfer penalty is a term used by transit planners to model/approximate how thoroughly riders dislike transfers (in general, or in a particular case). Often times expressed as a time penalty, or as a function of the wait time involved (time spent waiting for a transfer often counts more than time spent on a vehicle, depending on the circumstances of the transfer).
2) In the context of this article, Singapore has (until recently), required each boarding of a vehicle to incur a separate payment of fare, rather than allowing a single fare to cover an entire journey (as is commonly done, and commonly expected, in much of the west)–essentially penalizing riders financially for multi-hop trips. (While number of hops is sometimes cited as a proxy for trip distance, it functions rather poorly in that role.)
These two things are not the same.
Maybe most riders don’t have a choice in Singapore.
But that should not stop a public transit agency to make life easier for both sides: the rider and the transit operator.
The easier it is, the better to handle for the transit operator.
And when you speed boarding of the buses by not worrying about extra money to pay for transfers or even ticket sales by the driver, you might even save enough time to cut on vehicles needed for a line and save money. What also makes it desirable to have monthly passes.
It is some practical and neat kind of efficiency, and that should be a key factor in any modern society. I thought people in Asia would love efficiency.
@Scotty: in my comments, I use the term exclusively to mean the time penalty. The second definition you give I just call “paying for transfers.”
Regardless, my recollection is that bus-bus transfers in Singapore were always free. I think mine were, and I remember a professor use bus transfers as an analogy in class. What was not free is bus-subway transfers.
@TransitPlannerMunich: you’re right that it’s inefficient. As I said, I suspect there’s a not-invented-here thing going on here: Europeans and Asians generally refuse to learn from each other’s innovations.
However, in the case of Singapore, SBS (as well as the other operators) has done well for not having unlimited tickets. The buses have a proof of payment system, so you can tap your ezLink at the bus stop or at any door; you only need to pay the driver if you pay cash. In addition, on a crowded bus it’s standard procedure for the driver to close the doors and start moving before everyone has tapped yet.
I’m not familiar with Singapore, but IIRC in Hong Kong, you can’t transfer from bus to MTR subway, either–boarding the subway requires a separate fare. Hong Kong’s subway is fairly limited in scope, compared to the city as a whole–it focuses on downtown and Kowloon, and with one branch to the airport and two more to the New Territories–but it is generally viewed as a premium service, and for any MTR line you can think of, there are quite a few parallel bus routes.
The MTR (and the NT light rail) both use distance-based fares (you swipe your Octopus card on entry and exit; for the light rail a zone system is used for cash fares). But it’s highly common in Hong Kong for riders to stick to the bus system (which is frequent but a bit on the slow side) in order to pay a lower fare.
The MRT is quite limited as well. It misses many key destinations (NUS, NTU, the industrial zones, the science parks in Buona Vista), and serves many others poorly (Orchard is only on one line). However, Singapore has a policy of shifting riders from buses to the subway, which includes building a lot of new subway lines.