Facebook Google Menu Linkedin lock Pinterest Search Twitter


Airports & aviation

Dec 14, 2011

User login status :


Ground transport mode share for Melbourne Airport passengers (%) - data from Melbourne Airport

Yesterday’s post on the unreliability of predictions fits nicely with the latest round of calls for a rail line to the airport. The stimulus this time is a report in The Age last week on Melbourne Airport’s plans to upgrade freeway access and build a new terminal.

It set off a predictable and familiar landslide of calls for a train line. There were 141 comments on the article, virtually all of them advocating an airport train. I must say that I’ve hardly met a Melburnian who doesn’t think an airport train should be a high priority of any and all governments.

Some doubtless think others would use a train and thus, they imagine, reduce congestion on roads leading to the airport. But I expect most see themselves avoiding gridlock, punitive airport parking fees, or high taxi fares by using the train for most of their airport travel.

And yet if the train were built, there’s no doubt their prediction would prove to be enormously over-optimistic. Brisbane has a train from the CBD to the airport that carries just 5% of all travellers (another 3% come by bus). Sydney has a train too – it only carries 10% of all travellers (and a further 2% access the airport by bus). As Jarrett Walker observes, the political popularity of airport rail “is always several orders of magnitude above its actual ridership”.

Is there any reason to think that a train to Melbourne airport would increase public transport’s existing share of travel by a significantly greater amount than the trains have in these other cities?

Even without a train, Melbourne Airport already has a higher public transport mode share than either Sydney or Brisbane, with 14% of travellers accessing the terminal by bus. The former Government’s specification for a future airport train was a $16 fare, 20 minute trip time and 15 minute frequency. That’s much the same as SkyBus provides at present.

It’s true trains are generally more appealing than buses, but I can’t see that’s likely to lift public transport’s share significantly – certainly it hasn’t been enough in Brisbane and Sydney. It’s more likely it would cannibalise SkyBus and perhaps gain one or two additional percentage points of mode share.

If the latent demand for better public transport service between the airport and the CBD was as strong as readers of The Age think, then SkyBus – which offers the best frequencies and span of hours of any public transport service in Melbourne – should be doing much better than it is now (and it’s doing quite well).

It’s often argued that if an airport train were priced at a Zone 1-2 fare, it would attract higher patronage than SkyBus. That’s likely to be true, but it’s totally unrealistic – no Government is going to spend billions on an airport rail line and then subsidise its operations. And nor should it.

In any event, I doubt the increase in patronage would be anywhere near as dramatic as some assume. There is a host of reasons why the great majority of travellers would still prefer to drive or take a taxi than pay even a Zone 1-2 fare.

For example, most airport trips are to or from homes and workplaces in the suburbs – a taxi or a car is usually going to be more convenient than going to the local station and transferring to the airport service at Southern Cross. For many regular travellers, taxis and parking are cheap because they’re a business cost.

For tourists, it’s easy to justify a taxi for an occasional and important trip. Most tourists also travel with at least one other person, so in many cases that will improve the competitiveness of a taxi, or the long term car park, relative to public transport (I’ve elaborated on these reasons in previous posts – see Airports & aviation category in sidebar).

These are much the same sorts of reasons why Melburnians drive for the great majority of work and non-work trips. They’re also the same reasons why all those readers of The Age who imagine they, or others, would use a train instead of a car if they had the choice, in fact wouldn’t.

It seems human beings aren’t very good at predicting how they’ll behave in hypothetical situations. Perhaps supporters of a rail line implicitly assume the example of public transport’s success in the CBD can be transferred to the airport.

If so, they fail to account for some important differences. For example, the CBD is served not by one but by many radial tram and train lines. It’s also much less friendly to cars than an airport owner who earns substantial revenue from parking.


Book giveaway: follow this link to be in the running for one of two copies of Jarrett’s Walker’s new book, Human Transit

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

50 thoughts on “Would we use an airport train (as much as we say we would)?

    1. Dudley Horscroft

      Those who have found the bus “completely fine” must have been very lucky. When I have used the service the bus was full, so much luggage that the aisle was jammed, people had to climb over the bags to get to a suitable standing spot, it was noisy and uncomfortable.
      There are only two decent options. One is to extend the existing Airport West tram – yes, it will take a long time to get to the City, but more frequently than Skybus, and cheaper as the fare will be usable for transfers to other tram lines (not with Skybus). The other is to run a branch off the Keilor Line direct to the Airport. The killer in that MAGTP Report was “with the current airport passenger and staff numbers”. But there will be greater patronage, and, providing that the track and station is provided at reasonable cost, the line will be economically sound. Given the additional passengers for a very short line, possibly even financially sound.
      Yes, Brisbane’s Airport line only runs every 30 min – based on the QR refusal to recognize that the suburban services are a Metropolitan Railway and should be operating at no more than 15 minute service intervals on all lines. But unless you have a car load of people going there, it is a better bet than taking a hire car. And it is pretty well reliable and on time.
      Railway lines do not need to be gold plated. Start with “What is the absolute minimum we need, and how can it be done at what cost.” And then look to improvements which will bring in greater revenue than their cost.

    1. Harold Winthrop

      With respect, I’m not sure how you think Singapore and Brisbane’s airport lines are ‘premium’ or ‘tourist focused’. Both lines use standard rolling stock and in Brisbane’s case form part of the larger network.

      In Singapore’s case the terminal station at Changi’s great and the cross-platform interchange to East-West line services at Tanah Merah’s a good example of what can be achieved on a dedicated airport shuttle line. However, I can get downtown with my luggage in a taxi faster (albeit at greater cost) than on the MRT and not inconvenience other, non-tourist MRT users.

      As for Brisbane, the 1/2 hourly service frequency’s a killer. While transferring at Brisbane to an international flight in 2010, we were given vouchers by the airline to use the Airtrain between terminals. Having just missed a service by about a minute and not wanting to wait 29 minutes for the next train, the taxi won the day. Yes, it cost my $4.50 instead of being free, but the frequency was ‘right now’ and I had a stress-free connection.

      I don’t think the airport railway is viable (yet) for Melbourne. Public transport is already enough of a loss leader that it doesn’t need to be saddled with the white elephant of an airport line. There’s many, many other projects that are more deserving of the limited pool of PT funding than a line to the airport from the CBD. If you’re biased against Skybus, don’t want to park at the airport and are prepared pay for a premium service, take a taxi. That ticks all the boxes except dealing with the traffic.

  1. Harold Winthrop

    I must say, I quite like the Skybus service as it is. When I have to go to or from the city to the Airport, Skybus is my first choice, as it is fast, frequent and convenient. The price is vastly cheaper than the taxi alternative. The free CBD hotel pickup service is also useful as me and my luggage aren’t taking up space on a train or tram going to and from Southern Cross. When going to the airport from my home in the northern suburbs, taxi is the better mode as it is also fast and direct, takes my luggage and also means I’m not gouged for long-term parking. The main taxi issue is drivers grumbling about the ‘short trip’ (i.e. a $35.00 fare) from the airport to home.

    I think there’s a number of issues behind the call for an airport rail line in Melbourne. Firstly, the aspirational logical fallacy that seems to make up a lot of the argument goes a bit like this:
    “Melbourne is a global city. Global cities have airport rail links. Therefore Melbourne should have an airport rail link.”

    Whether or not you believe in the whole system of ranking and assessment of global cities, Melbourne still seems to rate quite well on a number of indices without a railway to the airport. Sydney and Brisbane fell for the ‘global cities need an airport railway’ trick in the 1990s and the taxpayers of those States are still dealing with the aftermath of those decisions with sub-standard rail services and ticket prices that are still the equivalent of Skybus for an arguably lower level of service.

    There’s also some technological bias against buses and in favour of rail going on, with many of the anti-bus/pro-rail arguments stated in the comments above. The fact is that Skybus can get higher frequencies still (I’ve also heard the 2-3 minute headways being bandied about) and that the fleet are not all 3-axle artics means it is not yet at saturation capacity. The main issues inhibiting Skybus are road space allocation (the need for a permanent Bus Lane is important), off-motorway priority for Skybus in West/North Melbourne and the CBD edge, dedicated access to Southern Cross Station and terminal capacity in the Airport and at Southern Cross. These treatments are relatively cheap (transponders to change traffic signals, pre-paid ticketing for all-door boarding, fare gates at Southern Cross bus terminal) There’s still a lot of slack to be pulled up before people can realistically say that a rail line is justified on capacity grounds alone.

    In an ideal world, the airport rail link from Southern Cross to Tullamarine would be the ‘down payment’ on a Sydney – Melbourne high speed rail corridor, but operated with dedicated medium-speed (160-200km/h) rolling stock optimised for airport travel. It should be constructed on a wholly new alignment separated from the suburban and ARTC freight networks and engineered to HSR standards enabling quick retrofitting for HSR operations. It would cost a lot of money, but as the Federal Government report into HSR outlined, the access to the cores of Sydney and Melbourne (and Brisbane) would be the largest cost element.

  2. Dudley Horscroft

    ” no Government is going to spend billions on an airport rail line and then subsidise its operations. And nor should it.” There is a massive assumption here – that a rail line to Tullamarine Airport (TA) would cost billions. There is a suitable route – [Flinders St], Southern Cross, North Melbourne, Footscray, Sunshine, Keilor East, Airport West, TA. Rather roundabout, but fast. And except for the last half km or so into the Airport, and possibly the curve from the freight line into the Tullamarine Freeway median, all at ground level. The cost should be more like a hundred million, not billions. The track to Sunshine is already electrified, and thence to Airport West requires electrification. The only new track is the aforementioned curve and then the line in the median to TA and the station there. (The old tracks may need upgrading with concrete sleepers and long-welded rail, but that is normal maintenance.)

    The State Government already subsidizes operations on the rest of the Melbourne rail network, why not subsidize this line? It also subsidizes roads in Melbourne, provided free of cost to the user (courtesy taxpayers), except of course the very few toll roads.

    And what would it cost to create a congestion free route for Skybus? Either you take away existing traffic lanes, or you build new ones. The first won’t happen, the second will indeed cost billions.

    Sydney is not a good example – the line was tunnelled and cost far too much, there is a massive additional charge for airport passengers and patronage is very poor. Brisbane is better, the line is single track, elevated instead of tunnelled, and after a shaky start is no doing fairly well. London, Heathrow and Gatwick, relies on rail for the workers and passengers, Frankfurt and Zurich have excellent rail services to their airports. Even small towns like Cleveland in the USA built rail lines to their airports.

    A rail line costs, but it is an INVESTMENT which will return benefits for 100 years. Widening a Freeway is a no-hoper.

    1. Alan Davies

      If it only cost $100 million or therabouts then previous governments would’ve committed to it long ago. There’s no doubt it has immense political appeal.

      Instead, you have statements like this from the Melbourne Airport Ground Transport Plan: “The Victorian Department of Transport has investigated the feasibility of introducing a dedicated rail service to Melbourne Airport and found that the market imperative is: $16 fare, 20 minute travel time and no more than 15 minutes between train departures, which make the construction of a branch line to the airport and operation of the train service both unfinancial and uneconomic with the current airport passenger and staff numbers”. Maybe in the future (although SkyBus says it can get down to 2-3 minute frequencies).

      I reckon the airport station alone is probably $500 million (note the HSR feasiblity study estimated the cost of a new station at SXS at $1 billion).

      Brisbane’s Airtrain is not a good example either. It runs on half hour frequencies (when they’re not cancelled) and up until this month only ran to 8 pm (now 10 pm). Skybus is a 24/7 operation and runs at 10 minute frequencies for the great bulk of the day.

      1. Julian Wearne

        The Victorian DOT shouldn’t be let anywhere near station costing. I’ve brought it up a few times on this blog Alan and I’m yet to have had any reply on the matter.

        How is Western Australia (where Labor costs are higher than Victoria) able to continuously build stations and rail projects at a fraction of the cost of Victoria?

        The ‘New MetroRail’ project is a shining example here. Look at what Perth was able to get for $1.66billion and imagine how much it might cost in Victoria considering individual, above ground stations are now being built for $110mllion. I can’t find article at the moment but I remember reading an article on a new railway station for Western Australia that seemed to have similar functionality (ramps, an overpass, etc) off the top of my head I believe it cost $17million.

  3. Julian Wearne

    In this argument I think you’ve made some good points and bad points Alan.

    Firstly, I agree that setting up a train that operates at the same frequencies as the Skybus, at the same cost will not likely attract more passengers. Nor do I think the airport rail line is the most pressing nor urgent public transport upgrade our city needs (that priority definitely lies in relatively cheap job of fixing the bus network!).

    On the other hand, I agree with some of the commentators above, I see no reason why we should subsidise the the construction of the freeway, and the airport, then encourage both the toll operators and the (already subsidised) airport to make a mint on the proceeds of encouraging everyone to drive to the airport. Public transport should be subsidised because of the positive externalities it provides. If the freeway is being widened, it shouldn’t be to allow encourage yet another lane of cars to drive to the airport, it should be to allow the Skybus and other bus services to have a dedicated lane.

    You’ve stated that a lot of regular airport travellers are travelling for business purposes. It is also fair to assume that many of these people are employees in the CBD. If the airport link was faster and cheaper than taxis I’d imagine plenty of businesses would encourage employees to take that option. Again the best way to test if this is true would be to dedicate a lane of traffic along the freeway as a bus lane, and additionally, drop the fare to a Zone 1+2 fare. I imagine the result would be that Skybus’ numbers would increase dramatically, whether its enough to warrant a rail link I’m not sure. The more I’ve thought about this issue, the less I’ve been convinced a rail link is a good idea at the moment (your blog has heavily influenced my re-think on this). What I am certain of is that widening the freeway for more private vehicles is a bad idea. Every time the freeway is widened for this purpose it only makes public transport the less attractive option and creates more negative externalities.

    1. Alan Davies

      Julian, you’ve pointed to some “bads” but not all of them are points I’m making. I’m not arguing for widening the airport freeway for cars or for prioritising cars over public transport as a means of accessing the airport. That’s a misunderstanding.

      But yes, I don’t think SkyBus fares should be subsidised (other than for airport workers). I’ve made my case here (perhaps I’ll revisit that issue shortly – there’re some new points). I think cheaper fares would increase SkyBus’s mode share, but I don’t think it would be as dramatic as you do. And that’s because I think the car/taxi option is decidedly more attractive than SkyBus for the great majority of travellers (but bang up airport parking fees dramatically, or start tolling the Tulla, and that could change things dramatically)

  4. Ash

    Why not extend the Upfield line to the airport? Upfield station could be closed/used as stabling.
    The rail link could head west just north of Camp road, along the old goods rail alignment, and then a tunnel underneath Pascoe Vale road, the shopping centre and some houses, and then travel above ground or on a viaduct through the Attwood green wedge. It would mean that the Upfield line could have an improved service and improved patronage.
    A few stations along the line would have to be closed to provide higher average speeds and several grade separations would need to be undertaken, but it would be the start to utilising the Upfield corridor to its full capacity. (As we know, a spur off the Craigieburn line is also possible but it is pretty much full already in peak hours)

    The idea of an airport express train assumes that at one certain point in time there are 300 people going to the airport from Southern Cross, and that idea is why many airport trains are so underutilised (it is just a point to point service)

    Furthermore, the Hume bus reviews suggested several more bus routes serving Melbourne Airport, including:
    Sunshine/St Albans – Airport,
    Frequent Sunbury – Airport service
    Footscray – Airport

    The local bus routes needs to be fixed as the most immediate priority, the rail link options should be studied further, as I think the government is doing now. We will see in a year or two years about the results of the government’s Tullamarine rail link study.

Leave a comment