Follow up to fare setting post

Posted: January 15, 2014 in Transport
Tags: , , , ,

Correction: A calculation error was made in the initial post. These errors have been corrected (original figures shown struckout). The graphs have also been corrected. While these figures still support bus plus train fare integration, given the similar cost per passenger km for those 2 modes, it does make achieving this appear more difficult given that there is no longer a small gap in the fare charged per passenger km for those same modes. Therefore, doing so remains the most likely outcome, but would now require a large (circa 50%) increase in train fares relative to bus fares.

A post published here last week about fare setting resulted in a fair amount of interesting discussion, enough to warrant a follow up, starting with a recap.


When setting fares, one of two approaches can be taken: a cost based approach and a distance based approach.

The first approach is to require fares to represent the cost of providing the service. The more expensive it is to provide that form of transport, the more expensive the fares should be. This uses price signals to encourage passengers to travel on the mode of transport which costs the least to provide. The average cost of transporting a passengers a single km on each of the different modes works out to: $1.27 $0.96 on a ferry, $0.79 on a train, and $0.59 $0.79 on a bus. This suggests that, given a similar length journey, fares for buses and trains should be equal, while ferries should be about 115% 22% higher than for buses and trains, and fares for trains should be 34% higher than for buses. However, the average fare for a passenger traveling a single km on each of the different modes works out to: $0.41 $0.31 on a ferry, $0.17 $0.23 on a bus, and $0.13 $0.15 on a train. Thus, ferry fares are 141% 35% higher than buses (too high: they should only be 115% 22% higher), while train fares are actually 24% 35% lower than buses (too low: they should be 34% higher the same). Thus, in order to properly represent operating costs, ferry fares would need to be cut by 11% 9% and train fares would need to be raised by 57% 53%, with bus fares remaining steady.

2014-01-15 Operating cost per km

The second approach requires fares to represent the distance traveled by passengers, effectively integrated fares. Two people traveling 1km on public transport should be charged the same, regardless of which or how many modes of transport are used. This ensures that passengers use the most efficient and effective route to reach their final destination, rather than prioritise one that minimises transfers. As previously mentioned, the average fare for a passenger traveling a single km on each of the different modes works out to: $0.41 $0.31 on a ferry, $0.17 $0.23 on a bus, and $0.13 $0.15 on a train. Thus, in order for fares to be the same for traveling the same distance, ferry fares would need to be cut 59% 35% and trains would need to be raised 31% 53%, with bus fares again remaining steady.

2014-01-15 Fares per km

The previous post concluded that if integrated fares was the goal, then it would be easier to achieve fare parity for trains and buses, given the smaller disparity in fares similar operating cost per passenger km than compared to that between ferries and buses/trains.

Update: The following paragraph was added at 3:03PM, 15 January 2014

However, doing so would require a 50% increase in train fares relative to bus fares. This does not necessarily mean a change in the base fare. For example, much of this is possible via the removal of heavily discounted periodical fares for trains, which account for 45% of train users, that appears to be occurring with the rollout of Opal.


This conclusion is based on certain assumptions which do not always hold up well, some of which have been pointed out in comments to the earlier post.

2014-01-14 TandemTrainRider

The post assumes that the fare per km and cost per passenger km are constant within each mode. In reality, these vary wildly based on things like total distance (short trips have higher fares per km than long distance ones), availability of concessions (children/pensioners/students pay a lower fare than working adults), geographic location (highly patronised inner city services cost less per passenger km than sparsely patronised outer suburban services due to costs being divided among a greater number of passengers), etc. As a result, claiming that fares cannot be integrated because one mode costs more than another overlooks the fact that each mode is made up of a number of routes, some of which will have higher costs and some of which will have lower costs.

[tweet 421381662245543936 align=’center’]

The figures used also only consider operating costs, and not any capital costs. This is most significant for trains, which require a large up front investment in the form of railways, often underground, whereas for buses and ferries these costs are often small or nil. It could be argued that these are sunk costs: they have already been made and cannot be reversed, so should not be considered in decision making. It is also the case that rail operating costs (2013: $4.0bn) are many times the size of its capital costs (2013: $1.6bn) according to Railcorp (p. 8) But given the billions being spent on expanding and maintaining the rail network, it remains difficult to eliminate capital costs entirely from consideration.

David Caldwell made a strong case in favour of including ferries in any multi-modal fare integration in one of the comments to a post he wrote about Opal back in 2012. It’s too long to replicate in its entirety here, and the post itself is even longer, but both are definitely worth a read.

2014-01-14 Alex

Finally, there is also the possibility that different modes of transport may retain their differing fares, but with only a single flag fall per journey. The dual standard currently applied by Opal is worth noting here as currently two trips made one after another are considered a single journey for the purposes of reaching the 8 journey per week level after which all travel is free, yet each trip within that journey has a separate fare. Each of those fares has a flag fall component (akin to the $2.50 flag fall paid to a taxi driver for merely boarding the taxi) and a distance component (which increases roughly in proportion to the distance traveled). It would be quite achievable to remove the flag fall, but retain separate fares for different modes.

The argument here that Treasury would be opposed due to the loss of fare revenue is valid. But Treasury has already appeared to have lost that fight on single mode fare integration, given that the fare for two bus trips is now calculated as though only one was used. However, this was likely achieved because the distance component of bus fares is the same for all buses, and so it would be difficult to extend this to other modes until two or more modes have similar fare calculation methods.

That is why the previous post recommended that buses and trains adopt similar fare bands. This is easiest for buses and trains because the disparity in fares between them (24%) is much lower than that for ferries and buses (59%).

  1. Greg Cameron says:

    Greg Cameron 62598145

    Freight rail bypass better idea than WestConnex

    PUBLISHED: 14 JAN 2014 06:44:27 | UPDATED: 14 JAN 2014 06:44:27

    Port Botany container trucks will require 50 per cent of Sydney’s M5 East westbound tunnel’s capacity and 25 per cent of the eastbound tunnel’s capacity when the terminal reaches “natural capacity” of 13 million by 2040. This defeats the purpose of building WestConnex.

    Increasing the port’s selling price (“ The inside story of the astonishing $5.1 billion NSW Ports sale”,AFR, January 3) was achieved by the NSW government abolishing the cap on container throughput, which existed because of inadequate transport infrastructure. Maximum rail capacity is a meagre 2 million containers, while 5.5 million truck movements would be needed in 2040 to transport another 11 million containers.

    NSW government policy, however, is to build a freight rail to bypass Sydney, between Glenfield and Newcastle. Railing all containers between a container terminal at Newcastle and an intermodal terminal at Eastern Creek would pay for the new line.

    Freight can then be removed from Sydney’s rail network, increasing rail’s share of urban travel, and $4.4 billion will be saved by cancelling stages two and three of the Northern Sydney Freight Corridor.

    Interstate goods entering Sydney can be railed, not trucked.

    The freight rail corridor between Epping, Strathfield and White Bay can be used for the new North West Rail passenger service. Sydney Airport’s environmental performance and passenger capacity can be improved by extending the parallel runway and building a second cross runway, using the container terminal site.

    Creg Cameron Florey, ACT

    The Australian Financial Review

  2. TandemTrainRider says:

    Another great post @Bambul.

    Before I get into the nitty gritty I think it’s worth reflecting on what we are, and what we should be, trying to achieve with pricing.

    Is it any or all of these thing:

    – Lower the to-taxpayer cost of provision of PT serivces through better cost recovery
    – Maximise the “externalalities” return on PT investment by pricing it to encourage it’s use
    – Send appropriate price signals to users to moderate their consumption
    – Send appropriate price signals to operators to drive more appropriate service provision

    IMHO the last point is something completely absent in NSW :-).

    I’ll give an example …

    CityRail used to – and I think these services still operate but I’m not sure which agency NSWTrains of SydneyTrains wears the cost or revenue – operate commuter 850 seat trains from Wyong (~100km from Sydney) via the Nth Shore, stopping at Woolstonecraft (~5km from Sydney) on it’s way to Wynyard. How much should each user pay?

    The cost of running the 3000ton steel box to deliver the seat to the city is the same regardless of where the PAX boards. A fare structure properly reflecting a cost recovery percentage would rightly charge each PAX the same. Indeed this is the real concept behind zonal fare structures: the cost of providing the service is related to the route length, not the journey length. Indeed the cost of the relevent MyMulti periodical ticket reflects this: $64/week vs $46/week. If you look at the rail only fare – which is more distance based (sort of), the headline casual use fare is $8.60 vs $3.80.

    The real question here is who’s fault is it that a 5km journey is delivered by running a service of over 100km? Should the pricing system deter a short haul user from “wasting” a long distance PT resource? Should the operator bare responisibility for an (aledgedly) inefficient operating patern? I don’t think there is any obvious answer, but whatever that answer is, it is down to the objectives and priorities of the pricing structure.

    I think it’s worth reflecting on the forces that drive PT pricing: the narrow desire to recover direct costs as much as possible, and the broad desire to appeal to a broad audience at the statewide political level. MyZone was as much about lower PT fares (in a cost effective way) to users in what were marginal seats in far western Sydney and the central coast than any sort of serious price reform. This is a very very long way from a market mechanism to appropriately allocate resources.

  3. TandemTrainRider says:

    @Greg Cameron

    > NSW government policy, however, is to build a freight rail to bypass Sydney, between Glenfield
    > and Newcastle. Railing all containers between a container terminal at Newcastle and an
    > intermodal terminal at Eastern Creek would pay for the new line.
    Do you have a link for this?

    > Freight can then be removed from Sydney’s rail network, increasing rail’s share of urban travel,
    > and $4.4 billion will be saved by cancelling stages two and three of the Northern Sydney Freight
    > Corridor.
    ATM no freight operates on the Sydney passenger rail network during peak times. There is no serious prospect of increasing rail’s share of urban travel by re-directing freight. The NSFC is about allowing freight and passenger rail to interoperate a bit less painfully, thus allowing a reduction in the peak hour restrictions on freight rail.

  4. Simon says:

    TTR, I see where you are coming from, but I’d argue by stopping at Wollstonecraft they can let on standing passengers which would otherwise be less likely. It’s a bit of a moot point anyway because they need to stop there.

    Putting that aside, I don’t think such a thing justifies making complicated fare rules to more fairly redistribute the costs. Jarrett Walker has a pretty good spiel about “do we want fares to be fair”. It’s in his book, but probably on his website too.

  5. Anon says:

    I think a distance methodology could be easily rolled out with some additional “flag falls”/ surcharges.

    Same bands for distance
    – $1/2 extra for ferry journey (reflecting higher cost)
    – $5/6 for Fast Ferry services (if the existing private operators join the system
    – $8 (or so) Gate Fee for Airport station (not that I am in favour of this but it is current policy)

    And then potentially a
    – $1 mode change fee (to capture some of the fixed flag fall cost for each mode), but still to provide some passenger benefit.

  6. TandemTrainRider says:

    Like a $2 cross harbour surcharge

  7. michblogs says:

    Well other economic theorists would tell you, the fare should be based on the incremental or marginal cost.

    Fares based on an average cost are nonsense. Is the average cost of carrying ten people 5 km on the carlingford line ( driver’s wages, electricity, wear and tear on the train, variable overheads ), the same as carrying 500 people for 5 km on the western line ?

    If they were going to be covering the average cost, you’d have to be charging the carlingford passengers $170 each.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s