Facebook Google Menu Linkedin lock Pinterest Search Twitter


Public transport

Feb 21, 2013

Do public transport fares cost too much?

Governments with the spine to increase transit fares in real terms should be applauded, not pilloried. Public transport is already very heavily subsidised financially but desperately needs more money


Sydney's CityRail performs poorly on % operating costs recovered from fares compared to other metro rail operators (source RailCorp)

In a story titled Public transport costs race past inflation for Melbourne commuters, The Herald Sun reported yesterday that “fares have risen twice as fast as inflation since public transport was privatised”.

The paper says that since 1999, the cost of commuting has increased by up to $1200 per year.

Yearly tickets have doubled in price. Monthly and weekly tickets have also risen by more than double the 41 per cent rise in inflation in this period. Rises for daily and two-hour tickets have been less steep but still outpaced inflation, jumping 52 to 80 per cent

As usual, I’m gob-smacked at how the mainstream media selectively mix up dollar and percentage figures; blithely show increases but not the starting or ending figures; and don’t show inflation over the period until near the end of the story (it was 48.5%).

I don’t believe for a moment that “fares have risen twice as fast as inflation” as the paper implies i.e. that the average fare has increased in real terms by 100% over the last 14 years (that’d be a 7% per year real increase on average!).

In these sorts of cases, the taboid approach is to identify one ticket type that’s increased significantly in price and then slyly infer the increase applies across the board.

Average fares were increased in real terms this year (6.8% nominal) but, according to this December 2011 report, train and tram fares have only previously risen faster than the rate of inflation on two occasions since privatisation back in 1999 – in 2004 and 2012.

The Public Transport Users Association took the numbers at face value. The President, Tony Morton, is reported in the Herald Sun saying:

Public transport users aren’t getting value for money….We’re now paying twice as much for a system that is by no means twice as good as it was in the 90s, and in some ways worse.

Putting aside the question of  the size of the increase, I haven’t seen any objective data that shows by how much the system is better or worse now than it was in the 90s. It would be an interesting exercise to devise a fair and balanced methodology for assessing the proposition.

Due regard would have to be given to the significant increase in patronage over recent years. The Transport Minister reckons patronage grew by 200 million trips to 536 million over the 14 years. He says many more routes now have more frequent services.

What interests me most, though, is the logic that travellers can’t reasonably be expected to pay higher fares unless they’re offered a proportionate increase in service.

This is a pretty common expectation. If you read the comments on this story, you’ll see there are even users who rely on this sort of argument to justify deliberately evading fares.

The key fact, though, is that public transport passengers in Melbourne (and in Australia generally) pay only a small proportion of the real cost of their journeys.

They pay none of the capital costs and, in the case of rail (the major public transport mode), only around one third of operating costs.

As the exhibit shows, Sydney’s CityRail recovers a similarly low proportion of its operating costs. Moreover, its level of cost recovery is very low compared to its international peer group (CoMET and Nova are groups of rail operators who share benchmarking information).

The NSW Auditor General reported last year that the average subsidy for each rail journey in the state is $10.01.

Revenue from fares is important for public transport but it’s modest compared to total capital and operating costs. An increase in fares obviously can’t lead to a proportional increase in service; not even close.

Even a relatively large one-off increase in fares, like the average 8.6% (about 5% in real terms) brought in at the start of last year by the Baillieu Government, will defray total system costs by only a small amount.

Rather than opposing fare increases, those with a commitment to improving public transport should be arguing for a higher level of cost recovery from beneficiaries.

As I’ve noted before, that would provide more revenue to build a better system, lower the disincentive (i.e. of ongoing subsidy) to governments of building more transit, and be less regressive.

And as I also wrote previously, the rail system alone needs billions spent on things like improved signalling, track upgrades and duplications, more train sets, new rail lines, improved security, and much more. Patronage has grown at around 5% p.a. and that increases costs.

Spending more to improve the system is vital. Experience shows system improvements are more likely to lead to increases in patronage than lower fares.

Although it’s difficult politically, increasing fares significantly in real terms on an ongoing and systematic basis must be a key part of any strategy for improving public transport.

Advocacy groups like the Public Transport Users Association should understand the nexus between revenue and better services. They should avoid playing politics and leave the populist stuff to the Opposition!

To genuinely serve the interests of public transport users, advocacy groups should promote a sustained strategy of real increases in transit revenue, including via fares, tied to long-term operating and capital improvement plans.

Alan Davies — Editor of The Urbanist

Alan Davies

Editor of The Urbanist

The Urbanist is edited by Dr Alan Davies, a principal of Melbourne-based economic and planning consultancy, Pollard Davies Consulting.

Get a free trial to post comments
More from Alan Davies


We recommend

From around the web

Powered by Taboola


Leave a comment

36 thoughts on “Do public transport fares cost too much?

  1. Smith John

    Dudley #34, elasticity of demand:

    ‘general agreement’ on typical elasticities: I’d be interested to know any overview references for this.

    ‘large fare rises are rare’: note for researchers: Sydney train fares were increased by about 50% in, from memory, about 1975, in what was presumably a kneejerk reaction to declining cost-recovery. A less certain memory is that an early act of the Wran Government elected 1976 was to undo part of that increase.

  2. Alan Davies

    Dudley Horscroft #34:

    To add to that, The Independent Inquiry into a long-term public transport plan for Sydney assumed a 6% fall in patronage in response to a 38% real increase in fares (raising farebox recovery for rail and bus to circa 70%). That fall though would be more than offset by the increase in patronage from investing the additional revenue (circa $2 billion p.a. by 2040) in system improvements. PT’s share of all travel is estimated to increase to 25% as a result of the suite improvements (higher real fares is only one source of the additional funding required).

  3. Dudley Horscroft

    To follow up on John Smith’s point, elasticity figures are difficult to come by as few systems bother to calculate them, and often various other changes are made at the same time. That said, there seems to be a general agreement that the elasticity of demand against fare changes is about -0.3 in general, about -0.5 to -1 for off-peak and about -0.1 for peak travel. This means that for a 50% rise in peak fares, Melbourne and Sydney would lose about 5% of the peak patronage. Much of this would shift to shoulder periods, where,say,a 20% rise could be justified. Caveat: such large rises are rare and the elasticity is probably not the same for a 50% rise as for a 10% rise. But even if Sydney lost 20% of peak patronage, most of this would have had to shift into the shoulder or off-peak periods, or people would go for a longer work day (hence work 4 days per week) and, given traffic congestion in Sydney, it is probably that most of the lost demand would return. In adddition, such a demand reduction in the peak would solve problems at Wynyard and Town Hall, leading to sufficient reductions in dwell time that CityRail could operate 24-30 trains on each track in place of the existing 17-18, leading to better travel conditions and (shudder to say) more patronage! And Cityrail could afford to build the new tunnels on the NWRL full size, adding 400mm to their proposed diameter.

  4. Alan Davies

    IkaInk #32:

    Well that’s a perennial problem for writers – how much supporting detail and evidence can you include and still keep a reader engaged? I find around 800- 1,000 words is about as much as you can get away with, but if someone has a deep-seated or ideological disagreement then no amount of words will ever be enough.

  5. IkaInk

    The arguments surrounding fare rises are far more complex than the above article covers. Sure you’ve linked to some previous posts you’ve done on the matter, that offer some more depth, but that depth should have at least been mentioned if your aim was to convince anyone.

    Apologies to any other commentators that have mentioned any of these points.

    Firstly and most obviously public transport doesn’t exist in isolation. It must compete with other modes, most predominately the car. Farebox rises that drive people away from public transport don’t make any sense if externalities are considered.

    Secondly, there is are two sides to getting better cost recovery. Incoming revenue, and outgoing expenditure. I know its been argued on this blog that most low-hanging fruit has been picked in terms of public transport improvements, but frankly that is rubbish. There are plenty of improvements that will improve services and reduce costs. These are win-win improvements that must be pursued. In Melbourne I would argue the most important improvement is fixing the tangled, slow, meandering, duplicated bus network. In Sydney it would be introducing a cheap and effective multimodal ticketing system.

    Thirdly, an unbalanced loading will always make a transport system cost dramatically more, as peaks will cost a lot to be able to provide for, whilst huge amounts of the system are left barely used outside of the peak. Is it really the best use of funds to spend tens of billions on a new rail tunnel to add extra peak capacity (which is only necessary during peak times) when a fraction of that expenditure will produce a much bigger mode shift outside of peak hours?

    Following improvements that are either “low hanging fruit” or “off-peak” improvements (obviously there will be a lot of crossover here) can of course increase farebox recovery ratios. This then makes big-ticket items, built for peak hours be easier to afford and easier to justify? These arguments also make a lot of sense for differentiating peak, counter-peak and off-peak fares.

    Fourthly, if you are going to compare farebox recovery from different systems around the world; it is equally fair to compare service quality, and the actual fares being paid. Are Australian’s getting good value? I doubt it. Is that because our fares are expensive? Only partially, it’s because our fares are pretty expensive compared to the service quality.

  6. Tom the first and best

    Around the world the more car based cities do worse then cities, in similar economic situations with better non-car options. It is also true when comparing parts of cities with the other parts.


  7. mook schanker

    I don’t understand the cost recovery model argument which is much like hypothecation. Public transport for example provides economic benefits to business. What better way should we ‘tax’ business directly for benefiting?

    Farebox is an arbitrary and artificial function with no real linkage to transport costs, IMO the way it should be. Just look across the way to Auckland and you will see how ‘hypothecation’ has strangled public transport to death over decades (at least now they’re actually doing something about it and getting into the eletric train age, 80 years later than Wellington btw!).

    There is some smokes and mirrors in solely comparing Comet/Nova figures globally, especially comparing costs from Asian transport systems. Typically Asian systems are a lot newer with the asset base not as life expired as Australia where assets are hitting +100 years old! (cue loads of bridges). Hence the Capital/O&M workload is miles more compared to typical Asian systems. I am unsure exactly what normalising the Comet/Nova figures include but I would think they wouldn’t normalise for asset age and ‘capital’ works Australia would treat as regular O&M and Asia would never do as they haven’t hit serious capital works as yet. It is also this reason why cost recovery ratios don’t seem to make sense in the one country where factors such as pricing power parity, labour costs, systems and farebox are not wildly different.

  8. Burke John

    Thanks David#28 for addressing such a far fetched notion. However there is no reason for it to be a “glaring logical fallacy” at all just because most people would find it a surprising or novel statement.

    Of course you could always google up a few statistics and have a look but I’ll give you a tip. Expect countries such as Greece and Spain to have the lowest cycling rates and so on without any great anomolies…oh except for Ireland from memory. Ireland provided tax breaks for cyclists after the financial meltdown so watch that one.

  9. David Sanderson

    “Look at Europe since 2008. The countries that suffered the most economically and the rates of cycling and public transport usage make arresting comparisons.”

    There is a glaring logical fallacy here. Linking bike riding and public transport use to economic success is pretty far-fetched.

  10. Burke John

    The subsidization of private vehicles is way beyond that of public transport in Australia. The whole design of dwellings and infrastructure of our cities is around the suitability of such to be negotiated by private vehicles. The cost of roads could be a mere fraction of current levels and savings for dwellings without carports as 2 examples. There are 10s of billions to be saved by winding up this out of date unsustainable car dream.
    Alternately, continued reliance on private motor cars (electric or petrol for that matter) will send us broke.

    Look at Europe since 2008. The countries that suffered the most economically and the rates of cycling and public transport usage make arresting comparisons.

  11. Alan Davies

    Dylan Nicholson #25:

    That might well be right, but when it comes to big ticket items, like new rail lines, real money is needed

  12. Dylan Nicholson

    Except I’m sure there are at least a few things that could be done that would simultaneously reduce costs (in the medium term) and improve services. More often than not it costs far more when things go wrong than when they go right, so there’s ample scope for reducing the former sorts of costs. Even if you really did need to spend more to improve services, much of it could be recouped by the eventual increase in patronage.

  13. Alan Davies

    David Sanderson #23:

    It’s generally acknowledged, including by PT advocacy organisations, that improving services will win more patronage than reducing fares. Even the PTUA is opposed to free PT.

    Improving services requires money and, given the very low farebox recovery ratios in Australia relative to peer operators, there’s considerable room to improve.

  14. David Sanderson

    How is the impact of cars measured? Cars have a disastrous impact on the amenity of city life and we also put up with an appalling toll of death, injury and more insidious health detriments. While public transport users are subsidised they are entitled to those subsidies because they are not imposing all the detriments car users bring upon their fellow citizens.

  15. Alan Davies

    Farebox recovery ratios for US metros.

    Among heavy rail systems, the average farebox recovery ratio is 46.4%, and systems range from 20.2% (Baltimore Metro) to 76.6% (New York City Subway). The range for light rail systems is between 12.0% (Dallas) and 57.4% (San Diego), with an average of 29.6%. Commuter rail averages 33.3% and ranges from 6.2% (Portland WES) to 62.3% (Metro-North).

  16. Dylan Nicholson

    Also, 81% of services might only equate to 60% of trips, seeing as the most delays occur in the busiest times. Though I was quite surprised when out at ~6:30 this morning to see a tram running past the zoo that was completely full.

  17. Dylan Nicholson

    I wonder if this change: http://www.theage.com.au/victoria/not-good-enough-new-yarra-trams-boss-tells-staff-20130221-2eudu.html will actually make a difference. Only 81% of services running less than 5 minutes late…that’s pretty shocking. I should hope they’re aiming for something like 81% of services running less than 1 minute late.
    But at least the trams can blame that partly on other road traffic. The train network has no such excuse.

  18. Tom the first and best

    For increasing tram and bus routes fare recovery there is traffic light priority and segregated lanes to reduce staff and passenger time wasted.

    For Melbourne buses there is also the restructuring of the routes to be straighter, director and not winding down side streets so that they have time competitive routes.

    The buses and trams can also be run at a higher frequency.

  19. Smith John

    Agree in principle, but you have to look out for the elasticity of demand. Not much point increasing fares if you’re on a part of the demand curve where that kills enough demand to decrease revenue.

    That extreme is unlikely at present, but you still have to think about how much lost patronage you want to wear for may be a small percentage improvement in fare box cost recovery ratio.

    Start with increasing peak period fares where elasticity of demand is less (less lost patronage) and the marginal cost of providing the service is higher (so the gap between peak and off peak fares is arguably fair). Or to be more politically palatable, increase the off peak discount (ie hold off peak fares steady) at a time of general fare increase.

  20. Tom the first and best

    There are plenty of ways to increase fare revenue without across the board fare increases.

    Increase the relevance of the rail system to non-peak and/or not into the CBD trips. One obvious way to do this is to built Southland station so the Frankston line (which is currently has a lower proportion of non-CBD-bound peak trips than other lines) has more counter-peak, outer part only peak direction (using seats otherwise left empty until further in), inter-peak, and weekend daytime patronage with little increase in running costs due to most of the extra patronage going on services already run.

    Run train at higher frequencies more of the time so they are more competitive with the car.

    Increase fares only in peak hour in the peak direction with lower off-peak fares.

    Cut seniors card concession as the only means testing is that their holders must work less than 35 hours a week or not at all. This means that wealthy self funded-retirees and non-fulltime workers (of whatever income) can get very generous concessions while far more needy people have to pay full fare. Not forgetting that pensioners would still be able to get concession with their pensioner cards. Failing that a higher fare for seniors in peak hour (about 8% of travel in the peak is seniors apparently).

  21. Dylan Nicholson

    Yep, my response would have “Yeah I know, thanks for reminding me why I ride my bike!”. But there are of course legitimate reasons why that’s not a practical option for more than a certain fraction of the population, and there should be good alternatives.

  22. SBH

    Dylan, I got in the lift yesterday morning. One colleague moaned about how bad the traffic was even on his motor cycle, another commiserated with how bad the trains also were and a third said it had taken them 40 minutes to go 15 tram stops.

    Seeing as everyone was contributing I noted that the ride was joyous on that beautiful morning and took 25 minutes.

  23. Dylan Nicholson

    Well yes, it may be better than driving/parking, because that alternative is far worse. But we should be able to do better in a first-world international city like Melbourne to have to choose between awful and abominable (there’s also “awesome” – riding your bike. Eventually people might catch on…)

  24. Tom the first and best


    The reason PT is competitive in peak hour to the CBD is the traffic and the cost of parking compared to the higher frequencies and close proximity of stations to workplaces. Outside peak hour the trains are less frequent (and there are fewer expresses) and the traffic is less so it is faster.

  25. SBH

    Not even two thirds Alan, more like half

    For Victoria the farebox figure of $577,532K published in the DOT 2011 Annual Report and the operating costs of $2,139M published in the 2011/12 Victorian Budget papers.

    This calculates out as the farebox only covering 27% of the cost of delivering public transport.

  26. Dylan Nicholson

    I’d also say the only reason I personally would be ok with paying double is because I actually get pretty good value out the system, given I travel quite a long distance all within Zone 1 (from one side of the city to the other). When I used to travel only a short distance between Zone 1 and Zone 2 I didn’t think it was good value at all.

  27. Dylan Nicholson

    AD, bringing cost recovery percentages in-line with other countries surely has more to do with running the system efficiently and getting decent patronage levels than increasing ticket prices. To getting decent patronage levels you need to provide a good service.
    If the PT system needs to go further in debt in order to prove it can provide high levels of service, then so be it. I’m not convinced it needs to be spend as much money as is often claimed though – for example I’m not aware of a single case of a train being delayed because of a lack of rolling stock. OTOH, ensuring every signal box has a redundant back-up couldn’t cost all that much, but given the number of times Metro have blamed delays and cancellation on signal faults, would surely be money worth spending. And I’m willing to go out on a limb and suggest that a few relatively simple timetable adjustments would mean that both a) at busy times nobody would ever have to wait more than 5 minutes for a train and b) the system would probably cost less to run. For a start, nobody’s ever explained to me the point of running express and stopping-all-station trains along the same track. You might save a fraction on electricity usage, but you can’t possibly move passengers at a faster rate than having every train stop all stations.

  28. Jim Wright

    I have been involved inthe design and construction of a number of railway projects over the last 50 years ago and the organisational model was pretty much the same everywhere. The government or council would have an engineering department, which would engage a consultant to design the railway and the two together would select a contractor from submitted bids. The funding model allowed time to be spent on optimising the project and the financing model (bonds at interest) allowed for flexibility in amortising the debt over time. Don’t forget that the return on the railwasy was not only fares and haulage, but also returns (e.g. land tax) from the new business generated by the railway. Contrast this with the modern model, where financiers shop around for the cheapest design and construction bids (often without regard for technical merit). Their only returns from the project are from fares and haulage, plus money from any sob story they can pitch to the client, plus multiple layers of fees form the esoteric financial structures they set up. From what i can see, the only merit in this model is the political one that it gets the project off the books. I did a back-of-an-envelope calculation once, costing some of the old projects at todays rates and reckon that under the new model, railways cost 100%+ more than they might have done.

  29. Alan Davies

    Dylan Nicholson #5:

    Your preparedness to pay double current fares would only bring cost recovery up to two thirds, still well below best practice Asian systems (and behind even some Nth American systems).

    Isn’t saying you have to prove you can run it well before you can have more money a Catch 22? Sure, it’s not just money (work practices are relevant too) but it’s really important e.g. to improve signalling; buy rolling stock.

  30. Dylan Nicholson

    (and yes, that’s assuming you’re fortunate enough to be able to live near a train line. But in fact most people ARE able to live with in easy driving or bicycling distance of a train line. That many choose to buy houses in areas where that’s not feasible isn’t really the fault of planners).

  31. Dylan Nicholson

    Strewth – “planners have made little serious effort to make the system an attractive alternative to car travel, other than for peak-hour CBD commuting”…seriously? The *least* attractive time to use the current system is commuting into the CBD at peak hour. The rest of the time it’s actually pretty good.

  32. Dylan Nicholson

    “those with a commitment to improving public transport should be arguing for a higher level of cost recovery from beneficiaries.”

    Sorry, it works the other way around for me. First prove you can run the system well, then you can justify asking for more money. I would have no problem paying twice the current fares if the trains were even half as reliable as they are in most other OECD countries. I’d also pay more to guarantee getting a seat.

  33. Alan Davies

    Moving Brisbane #3:

    I have looked at those very issues before e.g. Should cars be subsidised?

    Cars and PT are subsidised at around the same level. The big difference is most of the PT subsidy is cash (i.e. from the budget) and most of the car subsidy is economic costs (i.e. negative externalities like congestion).

  34. Moving Brisbane

    I think that Strewth raises the central tenet of the issue – one of “user pays”; with the point that roads are subsidised to a greater extent than PT (I’d love to see a side-by-side comparison of this incidentally).

    But leaving that argument to one side. Alan, what about the other costs in the equation; for example, congestion?

  35. Alan Davies

    Strewth #1:

    Indeed I do think the price of driving has to be increased in parallel, as set out in my earlier article I linked to above (Should public transport be subsidised?) and in others. Also see What did abolition of petrol excise indexation cost?

    The CPI period used in the Hun article starts from 1999 (not 1990) as they say that was when privatisation took effect. The 48.5% CPI figure is in the article (sourced from the Government).

    It would be very useful to see an objective figure of the real increase in average PT fares between 1999 and 2013. I’d also like to see how that compares to the change in median wages and median benefits.

  36. Strewth

    Alan, I assume you likewise support the government hiking petrol tax to ensure the price of fuel increases in real terms, to better recover the cost of building, maintaining and policing the roads and treating car crash victims. At present of course we don’t even allow fuel excise to increase with CPI.

    And that’s the larger issue of course. The public transport system doesn’t exist in isolation: it competes directly with private car use. If road user charges do not recover the full cost to the public purse of providing a road transport system, it makes no sense to demand the public transport system recover its own costs. Moreover, if fares increase without a similar increase in fuel prices, this will literally drive passengers away from the system – and full-fare paying passengers in particular. The old Met discovered this when they hiked fares in the 1980s and 1990s in an effort to reduce their operating deficit, and found it had the exact opposite result.

    As Asian and European cities know only too well, the easiest way to boost cost recovery is to ensure high patronage throughout the day, and this can’t be done unless fares are set on a competitive basis with incremental car travel costs. On a purchasing-power basis, fares in those high-cost-recovery cities are mostly well below ours (London and Tokyo possibly excepted).

    Our cost recovery is low in part because our system is inefficiently managed, and also because planners have made little serious effort to make the system an attractive alternative to car travel, other than for peak-hour CBD commuting. It isn’t necessarily fair that passengers should be made to pay more because of planning failures.

    Finally, going back to those inflation claims: according to the ABS, the CPI for all goods has increased by about 70% since 1990. The CPI for private motoring costs increased by a similar amount. But the CPI for public transport user costs in Melbourne rose by 150%. In that sense there is something to the claim that ‘fares have risen twice as fast as inflation’. Though yes, the actual relative cost increase in real terms is 2.5 divided by 1.7 or around 50%. Obviously, it would have cleared things up a lot if the paper had published the actual CPI figures and let them speak for themselves.

Leave a comment