Some thoughts on transport network pricing

On Saturday I participated in an Infrastructure Victoria session on transport network planning. It was described as a kind of speed-dating session – seven speakers doing quick 12 minute conversations with groups of 4-5 people.

They suggested kicking off with an introduction to your view on the topic, then seeing where the conversation takes you. After a couple of goes you pretty quickly get into a rhythm.

The thoughts below started as an approximation of my opening comments, but I’ve added a bit along the way:

Fare pricing can be an important tool to help ease congestion – both on the roads and on public transport.

Crowded train and platform at Caulfield during service delays

The fare system we have now is a hangover from the 1980s.

Before then, trains trams and buses had complicated, separate fares.

In the early 80s they had a shake-up – and introduced the three zones in Melbourne, one inside the other.

There have been tweaks and adjustments, but that’s basically what we still have today.

  • In 2004 they removed the Short Trip ticket from zone 1
  • In 2007 they got rid of zone 3; it merged with zone 2
  • In 2010 they removed the City Saver zone from Myki
  • Then in 2015 they made it so if you pay for zone 1, you’ve also paid for zone 2.

So we now basically have a flat fare system in Melbourne. If you want to travel two stops on the tram, it’s $4.40. If you want to travel from Werribee right across Melbourne to Pakenham, that’s also $4.40.

Given governments have an eye on cost recovery, this may mean upward pressure on that flat fare. It’s already risen at CPI plus 2.5% for four years running.

And for short trips this can discourage people from using public transport – they might see it as cheaper to drive.

There’s no peak/off-peak difference. If I travel on the train at 8am when it’s packed, it’s the same cost as at 11am when there’s plenty of space. (Of course, there has to also be a good frequent time-competitive service available at 11am. Off-peak service is very patchy around the network.)

“Peak” crowding is not confined to commuting hours. CBD trams were already crowded at lunchtime before the Free Tram Zone was introduced. This has considerably added to crowding.

Crowding on tram route 96 in Bourke Street

There is the Earlybird discount – free rides before 7:15am. This is bit of a blunt instrument – it reflects what Metcard was capable of when it was introduced last decade. A free ride in the morning, but the usual price going home later. So there’s no incentive to make your trip home in the afternoon at a time when it’s quiet. It also only applies to Metro trains. Catch a bus to the station? Then you pay. This can encourage people to drive to the station instead.

There is a discount of sorts after 6pm – you only pay one fare for unlimited travel until 3am. That’s good, but it’s also a reflection on the old ticket system – when paper tickets were around, there wasn’t enough space for a notch for every hour of the day!

2-hour Met ticket from 1991-92

So having spent a billion dollars on a complicated smartcard system, we have it charging a $4.40 flat fare for almost everybody.

But the Melbourne flat fare doesn’t apply outside zone 2. It jumps dramatically if you travel beyond Melbourne.

People coming in from Geelong to Melbourne pay $13.40 one way – if instead they drive through suburban Geelong to Lara and get on the train there, it’s $4.40. The fare system is providing a huge incentive for people to drive through Geelong.

Free station car parks are also an issue. Some of them get misused by non-public transport users. But apart from that, they cost tens of thousands of dollars per space – then they’re given away to whoever is lucky enough to show up early enough in the morning to get them – even if they have options to use a connecting bus, or walk or ride to the station. Those options need to improve, of course. (And maybe they could if all the money wasn’t spent on parking.)

So there’s plenty of scope to reform fares – Myki is capable of more zones, and off-peak discounts, and concessions for those who need them. This could be the way forward – but whatever the system, it’s important that the fares are affordable, logical, and equitable.

They could start with easy stuff like re-introducing the once-proposed weekly cap, which would encourage Monday to Friday users to also use the system at weekends (as well as removing some confusion and doubt for people trying to decide between Myki Money and Myki Pass).

King Street, lunchtime

Road pricing isn’t really my area, but it’s not hard to see how it’s similarly flawed. The only road pricing in Victoria is the toll roads. Again, it’s a hangover from past decades – tolls are to pay private toll companies that built Citylink and Eastlink.

This means it’s free to drive through the CBD, but it costs money to bypass the City and drive over the Bolte Bridge instead, which is a crazy outcome.

So the pricing on public transport and on roads is problematic.

Politicians are terrified of changes, because inevitably someone ends up being disadvantaged, but it’d be good to see them have the courage to introduce reforms to fix some of these problems.

Some people will always object to reform, but if the benefits can be quantified and explained, the broader community will take it on board.

By Daniel Bowen

Transport blogger / campaigner and spokesperson for the Public Transport Users Association / professional geek.
Bunurong land, Melbourne, Australia.
Opinions on this blog are all mine.

29 replies on “Some thoughts on transport network pricing”

Daniel – thanks for all your good work, this is another great thought starter. You’re 100% right about living in the past. With the current system we have much more data and capability. Maybe if the govt would make available the dataset of all trip data, it could be analysed by various Unis / transport think tanks etc. With some decent modelling, we could predict impacts of changes to pricing models….I know that lack of tap off on trams reduces the accuracy of the data but lets use what we have….Gary in St Kilda Rd

Very interesting post, thanks.
To most people, “reform” means that they are personally better off as a result.
e.g. for someone living in Sandringham and commuting to the city, “reform” occurred when zone 1 fares included zone 2.
Charging someone a fee for driving into the CBD is a “reform” but only for those who never drive into the CBD.
So I wish Infrastructure Victoria luck with network pricing.

Thanks for the post Daniel, you raise quite a lot of interesting questions. With a system like Myki it should be possible to have a highly nuanced charging system whereby people could be charged per kilometre or per stop passed, cutting out when it reached the maximum cap for the day, and easily having different charges for different times of the day, for example outside the peak times and weekends. This would encourage people to use public transport for much shorter trips as the charge would be very much reduced and also encouraging people to take their trips outside the main peak times if this was practical for them..

Pricing also factors into government investment decisions.

The main reason politicians love park and ride instead of bus improvements is that there is essentially no ongoing operating cost. Sure, the capital cost of a carkpark is very high, but it’s a one off. The bus operating expenses need to be paid each and every year, indefinitely. The government will ask themselves, do I want to commit the state to still annually funding this bus in 20 or 30 years time?

This ongoing budget drain is a huge pill for governments to swallow. One possible answer to this conundrum is to improve cost recovery, through fare reform so that the ongoing burden of ongoing operating costs is lower, or maybe use congestion pricing to pay for operating expense of PT, so that there is an ongoing income stream to pay for ongoing services.

I think having poor bus services is more expensive as it may be cheaper to run the buses, the revenue is much lower as it’s unattractive to passengers. Although many will connect to the train being revenue neutral many will use for local trips and if combined with cheaper off peak pricing then it will help fill up the bus and the train.

Two hour fares for zone 2 only are still only $3 (with longer fares and passes being similarly cheaper than their zone 1 counterparts), not the $4.40 of zone 1 fares, so the fares system is not quite flat.

*Weekend caps are however not commensurately cheaper for zone 2 and similarly do not apply to single 2-hour fares in either zone price.

On most of the rail network, the lines are busiest in one direction that the other and so people making cross town trips are usually using up spare capacity or capacity that is much cheaper to increase than in the other direction and thus are not imposing the same cost across their whole trip, reducing the argument for them being charged by distance. However the inner section of the Burnley group is very busy in the counter-peak and there may be one or two other exceptions.

Off-peak fares, including off-peak passes, are a good idea for spreading the patronage load and attracting people away from cars outside peak (when cars are more competitive). They could also solve the weekend fare issues outlined above as well.

As you have not doubt realised, fares for Public Transport is a very complicated can of worms. I’ll just mention one issue with putting a cap on weekly fares which then encourages people to use public transport on weekend. Sydney tried this and had to put an end to it because it caused a big problem: Excessive demand on weekends. This meant that extra services, particularly bus and ferry services, had to be operated to carry the people. But there was no cost recovery for running the extra services because most people were not paying to use them (having reached the cap for travelling all week). This was also not helping with traffic congestion on weekends because it was found that a lot of the induced weekend public transport travel was discretionary. While there was a drop in weekend traffic congestion it did not balance against all the people travelling on the free public transport. So the conclusion was that many people, particularly families, who were travelling for free on weekends would otherwise have staid a home if it had not been for the free public transport. And since there was no cost recovery on this extra induced travel, it was discontinued.

The cost of short trips on Myki is definitely frustrating for me. I live in Southbank, and it’s great being able to catch the 96 tram in to the CBD or out to Albert Park, the 109 out to Port Melbourne, or the 58 to the Queen Vic Market (though that one’s usually a bit tricky with a pram), but it’s not cheap — if my wife and I take our son out to the (excellent) playground in Albert Park, and don’t make it back within 2 hours, that’s nearly $18 to get us there and back. On a per-trip basis, driving and parking almost always works out cheaper.

That’s ignoring the cost of the car itself, of course; we don’t use it often, and ideally we wouldn’t own one, but our friends and family are mostly in the suburbs, or out in the country, so we often have no practical alternative. We used Flexicar before we had our son, but young kids and car sharing don’t really mix — you *can* do it, but lugging around a heavy car seat and having to repeatedly install and remove it was a huge source of frustration.

One thing that neither Sydney nor Melbourne has done would be, as our First and Best Tommy said, counter-peak travel. I’d like to see that happen in both cities. Anyone travelling away from Melbourne CBD should get a discount in the AM, and vice versa in the PM. Say for example:
Anyone catching a tram caught in the inner suburb (say St Kilda Road in South Melbourne) and disembarked in the outer suburbs (say Malvern or Toorak) in the morning will get a 10-20% discount.
Similarly, anyone travelling outwards on the train in the morning (say, Melbourne Central to Ringwood) will also get a discount.
The question would be to consider how to apply the discount when there is cross-CBD/City Loop travel involved.

Sydney would be somewhat more complicated, as there is quite a bit of travel to Parramatta and Chatswood, even in directions that would otherwise be considered counter-peak, but they could apply the off-peak discount for commuters travelling from, say Parramatta to Penrith for work.

Station car parks aren’t necessarily a bad thing – as long as you don’t make it a multistory building, you would essentially be setting aside adjacent room for the station to grow in the future (eg space extra tracks, extra platforms, or extra bus bays, take your pick). Just don’t place them right in front of the station entrance, and reserve that space for connecting buses instead. And, when the carpark has to go, just only provide replacement spaces nearby for the disabled. (Speaking of which, what really needs to happen is to have the bus timetables be linked to the trains. This means buses being held back when there’s a train delay.)

Road pricing is a very bad idea and any Victorian government that does it deserves to be thrown out. People hate it. In a democracy those preferences matter.

What surprises me is that so little is invested in reducing demand for road space through incentivising school cycling, or that safe routes to and from train stations and parking are not supplied. In this sense, Victoria doesn’t have a transport system so much as a set of kingdoms and orthodoxies.

albert, what do you claim was “discontinued” in Sydney ?

The congestion which occurs on weekends on Sydney occurs in a very few places – the Manly Ferry in particular.

You falsely attribute this to a weekly fare cap.

Yet the swarms of people from the western surburbs which you see on the Manly ferry at the weekend, are not there because of any weekly fare cap. Most of them are probably not even regular weekday commuters. Unless they work in the CBD, very few people are regular weekday commuters – they drive everywhere.

The reason the ferries are so popular on weekends are/were, the $2.50 Sunday fare. ( which is 2.70 or something now, and has not been discontinued ), combined with what used to be called “family funday sunday”, in which any number of children could travel with two adults ( but not single, widowed or divorced parents ), for FREE. Which not only saved money, but avoided having to get opal cards for them, when they almost never used public transport. Avoids bureaucracy and hassle of getting the card, not losing them, and not having to check they have credit on them.

It was certainly an advantage for people who hit weekly caps to be able to travel for no incremental cost at weekends. But the number of people taking advantage of that would be outweighed by the other facts I mentioned. And people using weekly or monthly travellpasses could also do that. And the $60 weekly cap is pretty steep.

@Malcolm: The only problem with per-kilometre ticket charges is that public transport is rarely in a straight line. You can pick virtually any single bus route in Melbourne and it will look like a square wave weaving through the side streets to pick up and drop off the passengers in suburbia while almost completely avoiding the main roads.

Likewise, the train system, which has been extended umpteen times over the past 120+ years as well as adding way too many superfluous stations in the first half of the 20th century that are barely 1/4 of a mile apart (tram and bus stops are even worse with as little as 100m or 1/16th of a mile apart in some cases). With distance-based fares, lines which are mostly straight such as Frankston, Upfield (beyond Royal Park), Glen Waverley or Pakenham would also end up significantly cheaper than long winding lines such as Belgrave, Lilydale and Hurstbridge which would not be fair to travellers who have no choice. The Hurstbridge line in particular takes a scenic tour around Greensborough rather than going straight through (e.g. under) the other side of the town. Those forced to travel through the City Loop without needing those stations would also be unfairly charged, and a distance-based fare system probably wouldn’t be able to tell whether you’ve switched trains e.g. at Richmond where you don’t need to touch off as myki readers aren’t on trains like they are on every bus and tram (the only “suburban” trains to have ticket machines were the old loco-hauled MTH carriages on the Stony Point line back in the Metcard days, and that was only because they were too lazy to install the machines at every station).

And of course, Melbourne’s trams seem to be considered as buses on rails by the PT companies, and come with the same design of spaghetti routes for the most part. Only very few routes are somewhat direct, such as the 59 and 78, while others zigzag all over the place such as the St Kilda end of the 12, and the incredibly short 82. Others even have a large U or L shaped route, such as the 16 and 72. Most of these zigzag routes all go back to when the tramways, railways and bus companies were all rivals (even from bus company to bus company), you couldn’t have trams and buses duplicating trains and vice versa otherwise there would have been war, while the some of the L- and U-shaped routes were actually separate routes which were combined over the years (e.g. the former 69 along Glenferrie Rd was merged with the 16 and lost its infamous number in the process; a more recent example is the 58, where the 8 and 55 were merged and rerouted, with route 6 replacing the northern end of the former route).

The winding nature of some routes and kilometre pricing was fixed in Sydney for it’s buses and ferries (not trains, yet) by calculating the fare from first touch on to last touch off – and not by route km, but by the straight line distance between start and end points. Even if you have to change buses to complete your journey. So no matter how winding the route is, or which routes you take from point A to point B the fare is the same. This system could be used for Melbourne’s trams, buses and trains to avoid that problem.

@ George
Road pricing is a good idea, you just hate it because you want the tax payer, the cyclists, the pedestrians and the road user to foot the bill for something that should not be free.

The taxpayer pays more because he has to fund the infrastructure without getting any revenue for it he also has to pay for wider roads and larger capacity systems because they are free.

The cyclists and Pedestrians pay for it financially and because the conditions for them are much more difficult on the roads because of the extra cars.

The road user does save money in the short run but also pays in the long run because of congestion and also contributes financially via rego, fuel tax and other fees.

Here is an example of the strangeness of the current fare scheme:
Ormond to Frankston – 18 stations, 43 minutes – zone 2 – $3.00 full fare.
Ormond to Glenhuntly – 1 station, 2 minutes – zone 1 – $4.40 full fare.
It’s a great bargain going to Frankston, but one feels a bit ripped-off going in the other direction. Clearly, any attempt to making the system fairer will create winners and losers.
It just happens that Ormond is the first station in the zone 1/zone 2 overlap.
Much the same thing on the Dandenong line:
Hughesdale to Pakenham – 16 stations, 52 minutes – $3.00
Hughesdale to Murrumbeena – 1 station 2 minutes – $4.40
Ormond and Hughesdale are my 2 local stations, so outwards journeys (not matter how long) are cheaper than any inward journey.

If you had more zones you would have a much fairer system. If you look at value for money as Daniel stated in the article, it’s even more unfair as 60km trips are the same price as 60 meter trips.

More zones is not an improvement. It is a totally CBD focused point of view. With modern technology, start to destination pricing based on distance is the way to go. Small “flag fall” so short journeys are cheap. Sydney bus fares start at 2.20 for trips up to 3 km, compared to Melbourne’s minimum far which is double that. No wonder people don’t pay !

And zero or minimum mode interchange penalty. This used to be a huge disincentive in Sydney, but was mostly fixed a few years ago.

There are other bizarre things in Melbourne. If you make a 100 minute journey on the train, and then change to a bus, you pay a cheaper fare if the bus comes in 5 minutes, than if the bus comes in 25 minutes. That’s right, the customer has a financial penalty if the bus is late, or doesn’t turn up, or because the time-tabling is poor.

The same thing happens if you enter the first station 20 minutes before the train arrives, instead of 2 minutes before the train arrives.

That is basically what you are paying for petrol in your car. Mostly, proportional to distance. Otherwise you cannot avoid ridiculous anomalies like at Ormond.

If you want to factor in costs you need zoning, it should be somewhere between flat and kilometres based and zoning is the most efficient at fares based on costs. Whether you go on a train for 1 km or 20km the costs are still similar as most railway infrastructure is a sunk cost. It should be flat but definitely not per kilometre.
Maybe to get rid of the 2 hour quirk push it up to 3hours. The car also has a large sunk cost of rego and insurance so travelling short distance is not as cheap as it looks.

Llib, you are mistaken about the fixed costs of cars.

Unless you are in a position where you can get rid of your car completely [ which I have done, twice, for a number of years ], then you have already paid those costs.

So the incremental cost of using your car for an extra short trip, is very low. It’s the cost of petrol, basically, plus a tiny amount for additional tyre wear. None of your other costs vary depending on whether you drove 10000 km or 10100 km last year. Even the depreciation in the re-sale value of your car depends on age much more than accumulated distance.

I agree with you 100%, that is my point about sunk costs they are fixed and don’t change according to use.

What I am arguing is that of course having a car is much more expensive than driving it as you said.
The comparison I’m making is that if someone chooses to go car free then short trips with zoning would be much cheaper than needing a car.

Per km or multiple zones are a lot closer than you think and much better than flat fares. I just think per km fares are just too low for cost recovery from a government or operator point of view as they themselves have high fixed costs so short trips and long trips are not that much different in terms of operating costs.

Lower prices for short trips encourage lots of short trips, however that means that more passengers trips are made per vehicle trip (by short passenger trips on different sections of the vehicle trip using the same passengers space in the vehicle) and thus brining in more fares per fixed cost vehicle trip.

@ Tom
That’s a simplistic way of looking at it. Like any business your cost is not only the seat provided but all the infrastructure, the stations etc. IE when the train is starting and stopping it loses time and must provide extra stations to let people get on and off as well providing extra infrastructure. Of course parking and facilities for getting to and from the stations must also be provided. If it’s a bus there’s also a lot of wear and tear getting people on and off the bus and the driver checking tickets etc. it’s cheaper but not by a large margin. A person going 1 km is not 50 times cheaper than a person going 50 km because you get economies of scale.

It’s also not always easy separating the cost per passenger because a lot of these costs are bundled and the key is to get people within the capacity for the service
An example is that if a train is full by the time it gets halfway to its journey then it is losing those short fares by the time it gets to the station. And flat fares encourage this which is bad both for cost recovery and obviously quality of service. If short fares are too low the train may never get full with people jumping on and off. This still has advantages as it gets cars off the streets and the service is slow but more more comfortable.
The train may never be full with cheap short fares as long distance fares will encourage passengers to get off the train earlier and operaters may be encouraged to get rid of expresses but under govt contract they probably cannot do this anyway. Also it won’t be fair from an equity point of view because low income people live in the outer suburbs and will find it unaffordable and take more time for passengers to get to distant jobs.

With 5 fare zoning you can encourage both short and long distance fares while adjusting the price to get the right mix for both capacity, equity and cost recovery.

Geelong people could catch the train to Lara for $2.40, then catch the next train Melbourne for $4.40, rather than the full $13.40.

I finally got frustrated with trips to and from Geelong that included a third trip in Zone 1, and the charging for that, that I put a query in to PTV and got a refund of the third Zone 1 charge. I’ll keep doing that in future! Although I think I am correctly charged according to the Fares Manual. The trouble is that the system is so complicated with the interface with regional fares that you need a degree in Myki to understand it.

Dont forget, MyKi ditched zone 3, because of the stress it placed on the MyKi server.

It had to do a calculation on each touch off, and, having that one less zone, meant that one fewer complication in the zone based calculation.

I have always felt that, there needs to be a short/single trip fare, two or so of these increments, like the old rail plus 2 and rail plus 4 tickets prioir to the early 1980s. The Rail plus 2 stayed on until more recently.

Longer multi-zone travel on the same mode, mostly on V/Line, should be a single or return trip fare, perhaps have a daily cap of say three of these, then be charged a 2 hour zone based fare, if they use connecting services at either end of that journey.

Comments are closed.