Melbourne’s great big tunnel of pork

ScreenHunter_05 Apr. 15 22.08

By Leith van Onselen

Fairfax’s Michael West has thrown a great big hand grenade at Melbourne’s East-West Link – Melbourne’s controversial road tunnel project – which according to West could cost an incredible $1 billion per kilometre, making it one of the more expensive road projects in the world:

Guess which road in Australia is shaping up to cost $1 billion per kilometre to build?..

As the state election nears, the Napthine government is declaring all sorts of unfunded public works, some $27 billion in infrastructure projects crowned by the controversial East West Link. Conceptually, this road is not a bad thing, even though the city’s trains are clogged, but the price is stupefying.

Actuary Ian Bell has kindly put some figures around this for us, deploying what scant information is in the public domain. Bell puts the capital cost of this road at up to $18 billion for 18 kilometres, a grand $1 billion per kilometre. One could be forgiven for thinking that the business case for the East West should be made public so the people who are paying for it could discuss it. Alas, not so…

Whether the $18 billion figure cited by West is accurate remains to be seen. Previous estimates have claimed the road would cost some $8 billion, or around $440 million per kilometre.

Irrespective, it is an expensive tunnel by global standards. As noted in The AFR today, a 10-kilometre road tunnel in France, known as “Duplex A86”, cost €1.56 billion ($2.25 billion), which works out at around $226 million a kilometre. France, of course, is renowned for its 35-hour working week and being a relatively high cost nation for infrastructure investment.

What is more concerning is that the East-West Link appears to have been approved without passing a rigorous cost-benefit analysis, with The Age earlier this year revealing that the project could deliver a net loss to Victoria:

Infrastructure Australia head Michael Deegan told a Senate committee that the government’s unpublished business case also provided an alternative estimate, showing a benefit-cost ratio of 0.8 when ”wider economic benefits are not included”.

Under this scenario the project would return just 80¢ for every $1 spent, suggesting an economic loss if the ”stock-standard” analysis preferred by Infrastructure Australia is used…

In a submission to a federal infrastructure inquiry, Infrastructure Australia targeted Victoria for failing to submit a ”robust” business case for the east-west link, singling the project out as an example of why the public are cynical about ”big ticket” infrastructure announcements.

With the mining capex boom set to unwind, Australia needs to find ways to fill the void to both growth and jobs.

One obvious option is to boost investment in productivity enhancing infrastructure – i.e. projects whose expected benefits far outweight their costs and whose cost of financing can be funded by the increased productivity (and therefore tax revenue) that they create, making the project’s debt “self liquidating”.

Unfortunately, the East-West Link appears to fail this most basic test, with the project instead looking more like a politically motivated exercise in pork that risks undermining Australia’s productive capacity and living standards.

[email protected]

Unconventional Economist


  1. The older $ 8 billion figure was just for the eastern section (tunnel) between City Link and the Eastern Fwy.

    I’m not against motorway projects but the fact is the info leaked so far paints a very sorry picture of the business case for this project. If the case for it is so damn good, why hide it? Why the rush to sign it up before the election? It seems all to be about subsidising the FIRE sector investment banksters, consultants and the trucking magnets and it will do that by forcing motorists to use it, ie, effectively making tolls into taxes on ordinary city commuters.

    Personally, I’d like there to be a good hard look at the so-called Missing Section of the Ring Road in Melbourne’s north-east between Greensborough and Ringwood. Currently, trucks and cars coming down the interstate and state highways in the north (Hume and Calder Fwys) have to push their way down through inner Melbourne on City Link and out along the choked Monash to get tp Melbourne’s sprawling south eastern suburbs. But this option was summarily dismissed by the 2008 enquiry, me thinks because it would go through some rather marginal, high income seats in the Yarra Valley green wedge, whose residents will paint their semi-urban, semi-rural environment as being as precious as the Daintree! But since this option was brought to my attention it seems to make mighty good long term sense given that eventually the Port of Hastings to Melbourne’s south-east will become a massive deep water port.

    • Dear Sainted,

      i understand your preference for the NE ring. i live near the Bulleen Road/Eastern Freeway interchange and my leafy green is sorely tried by the endless queueing of dirty trucks and uncouth tradies in SS utes. Removing this multi-layered offence would do wonders for local land prices, and for all properties East of here.

      If the Victorian government had any brains, they would build the outer Western ring road Werribee to the Hume Highway. A great deal of land would abruptly become attractive and prompt the investment and economic activity Victoria sorely needs.

      Instead, the billions spent on the East-West tunnel will simply become embedded in mid and outer East land prices, as would the NE ring. Great politics; poor economics.

      • Exactly; there were debates in the USA in the 1950’s over the Interstate Highway system, which meant the Federal Government was paying for highways that went right to the centre of every city; often several highways that radiated out from the centre.

        Many planners and economists actually wanted the Interstates to bypass the cities, which would be better served by a grid network of arterials, than monstrous radial highways. One of the primary grounds for this, is the effect on land values – “radial” concentrates them in the centre, benefiting big property rentiers there; “grids” disperse and flatten land rents, democratising both home ownership and “amenity”, and better enabling new business start-ups and new types of agglomeration.

        Sadly, the “radial” interests won the debate. Mind you, it wasn’t a “debate”, it was basically a battle for politicians souls.

      • Both you and Phil seem to make excellent arguments and I did ask for a “good hard look” at the NE missing section of the Ring Road, meaning a very open, public inquiry by experienced experts using tried and true benefit-cost analysis, not all this smoke, mirrors and rabbits out of hats nonsense we’re getting with the E-W or Sydney’s West-Con and god knows what other “nation building” infrastructure. Spain has great infrastructure. Hmmm

  2. Good call. Without sufficient accountability, these projects cost far too much, negating the benefit that might have been there otherwise. Plus, the way they are paid for need to be targeted, especially at the owners of land at the locations where the most trip destinations will be – otherwise they are the beneficiaries of a wealth transfer from the people who are paying, viz the general taxpayer.

    Tolling won’t work if there are alternative routes that remain “unpriced”. I have tried to explain this before.

    I wish someone else would tell me they “get it”.

  3. Mark HeydonMEMBER

    How much of the NBN could be built in Victoria for $18bn? Given the cost estimates for building out the whole of Australia, much of which is vastly less populated than Victoria, I suspect $18bn could pay for pretty much the entire state to have FTTP. Compared to one road tunnel, the NBN would be a bargain.

    • Great call. Invest in the way of the future in the 21st century, not the past in the 20th century.

      • Folly.
        We should continue building roads to accommodate (mainly) single occupant commuters at the expense of a cohesive mix of PT and bicycles.

        We can then encourage more cars, more congestion, more centralisation into cities, more urban sprawl.

        I only see upside.

      • More urban sprawl lessens congestion. Intensification concentrates it and worsens it. Everyone needs to be clear about this.

        There are two elements involved – the concentration of vehicles in a given space; and the centralisation of primary trip destinations. The greater the dispersion of jobs, the less congestion there will be, all other things being equal. Density hardly matters if jobs are centralised, the same “crush” of congestion will be created in the centre regardless of whether people are coming from nearby or further away. But jobs are never centralised any more, and this decentralisation leads to the next “rule”: the lower the density, the less the traffic congestion delay.

        Alain Bertaud and colleagues are trying right now, to work out a formula by which traffic congestion delay can be predicted. Road lane miles per capita is in it as an obvious element, with “more” predicting less congestion delay, all else being equal. Anti-road activists claims that the reverse is true, i.e. building more road space creates so much induced traffic that congestion is worse, is utter propaganda worthy of a Goebbells or a P T Barnum. It shows how easily people can be fooled, that anyone believes this.

    • Yep. NBN as it was originally envisaged should be picked up by one state in Australia, it may as well be Victoria.

      When I think about fibre optic cabling I always think back to when I first heard about it. I was a wee lad and Clive James was drinking beer in a German environ espousing how Australia was ahead of the Germans on this technological front. It was an advert for Telecom but still I don’t think that we can make a claim like that today.

    • in 10 or so years very few people will commute to work anyway. Work from home will become so normal that we will be asking ourself about sanity of commuting to work to perform work entirely using computers and phones

      • (my response on deleted post about my resistance to urban sprawl)
        I’m against deregulation of suburban land use because there is no need to compromise and destroy our cities and surroundings just when a need for people living in suburbs is vanishing. Telecommuting will allow people who prefer quite and secluded life to live in rural areas without making big compromises by living in poor life quality urban sprawls. There is no need for anyone to live is such a depressing environment. People who love space, tranquillity and privacy will be able to enjoy life on big blocks or mini-farms in rural towns. The others who love city life will be able to enjoy city life without being overcrowded by suburban commuters who are forced into city.

        Urban sprawl was post WWII compromise: people who preferred rural lifestyle were forced to live close to their factories. Factories are gone, telecommuting is our future so need for the compromise is gone.

      • No offence intended but you are absolutely dreaming.

        Look at the NBN. They’re supposed to support this type of work, but my close friends inside the business tell me it’s barely support and not encouraged.

        I worked in Telco for over 10 years and there’s almost zero drive from C suite execs to decentralise. I quit my job because they wouldn’t support me doing it (I moved north from Melb for health/lifestyle reasons).

      • drsmithyMEMBER

        in 10 or so years very few people will commute to work anyway.

        I’d love to believe this, but the technical capabilities to allow huge numbers of office workers to work from home have been in place for 10+ years, yet few are allowed to do so.

        Without some sort of financial incentive to mitigate the perceived (and real) risks of employees working from home en masse, it won’t happen.

      • Your pessimism might be well founded DrSmithy, but it is worth remembering that social change tends to massively lag technological change – the technology of the flush toilet was available for more than fifty years before its widespread adoption by the wealthy in England, for example, in part for the paradoxical reason that it lessened the requirement for servants.

        Personally I suspect that newer, smaller companies will adopt telecommuting as they are founded, giving them a lower cost base, and slowly supplanting the ‘bums on seats’ model of corporate organisation. Larger, older companies only adopt anything new when it is vital for survival – if their competitors are also old, large and equally hidebound, they might hold off the inevitable for as long as the Australian property market holds off its inevitable correction.

      • major issue with telecommuting is not technology but management. The way people are managed today (mainly stopwatch) will change as generations of older managers die and young “multitasking” generations enter labour force. Huge financial incentive is already there, office space is extremely expensive. It takes time for a social change to happen, but once it starts happening it develops quickly (remember cubicle culture in late 80s).

        Some big companies already started by introduction of hot desks. That is first step toward telecommuting. It will start as one day a week, but quickly turn into one day a week in the office.

        As petrol prices rise, telecommuting will became favourable by employees so companies will be forced to move on this issue to attract better employees. It will also enable companies to spend some of the money saved on space to attract good employees.

        eventually even IBM will start telecommuting 🙂

      • If you can put computing power and storage into the cloud, the only business purpose for floorspace in the context of a service company is for meeting customers.

        Moreover,it’s got to be tempting to force employees to purchase their own desk and floorspace.

        EDIT: IBM seems to be outsourcing at least part of its sales force into ‘IBM Business Partners’. My (not completely uneducated) guess is that a fair swag of IBM Business Partners are run out of a kitchen with a laptop, so in a sense IBM are already there.


        enable companies to spend some of the money saved…[on good] employees

        Yeah, now you’ve gone back to fantasy land. It might ‘enable’ companies, but the money will go straight to CEO’s bonuses.

      • Why isn’t the “things are changing rapidly” argument, not one in favour of NOT trying to “plan” too much?

        The biggest problem we should avoid is creating massive wealth transfers that pay for precisely nothing. Technological progress will occur QUICKER if there is more discretionary income and discretionary capital investment around rather than being hoovered up by a land-rent monster. Houston will be the first western city into the high-tech future, NOT the last. Precisely BECAUSE it is not “planned” to death. Even if “automobile dependent sprawl” is going to prove to be the wrong direction within the next 30 years, the cost of this “mistake” is MUCH lower than the cost of driving house price median multiples up to 7 – 10. There are no gains in the latter paradigm over the former one anyway, to compensate for the wealth transfer losses. Those being gouged do not end up living more efficiently, because their choices are all bad ones when they can afford FEWER of the good ones than they would have been able to before.

      • Um, Phil in this case it is in favour of “not planning too much”! The argument is “car journeys could be about to fall due to telecommuting, so move very slowly and carefully before you spend that $18bn on EastWestLink.”

        It is argument for doing nothing, not doing something else.

        The NBN stuff isn’t all that relevant to the overall question – you can do it now, or wait until telecommuting is popular, and the next election depends on it.

      • in 10 or so years very few people will commute to work anyway.

        There is currently a shortage of roads, but this is about to change due to some factor that far-sighted super-intelligent beings can foresee. Therefore, factoring this in, there is no shortage and no need to build more. Just wait and the problem will fix itself.

        Now where have I heard that argument before?

      • There is currently a shortage of roads

        A shortage! A shortage! Is there anything we’re not short of Claw?

  4. with projected daily traffic at 60k/weekday, 30 years pay-off period with 4% credit and 2% operational cost per annum it will cost almost $100 per car per trip to pay investment off in 30 years.

    on the other hand $18 billion can be used to build more than 100km of metro lines with capacity of almost 1 million passengers a day (at least it can be build even in very expensive cities like Zurich or Copenhagen)

    • And at some point in the future, the total public costs – capital plus operating – of that metro system, will top the total public costs of the same investment in road space, and it will be all downhill in sunk costs from that point onwards.

      Imagine if the Romans had spent their “transport” budgets on moving people around in horse-drawn PT vehicles instead of building roads. Who cares if the PT might have been cheaper for the first 100 years?

  5. The preliminary report puts costs at between $6-8b – would need to understand Bell’s methodology before sober comparison can be made. It seems to me that most road infrastructure in urban areas is quickly utilised and my understanding was that EWL specifically targeted freight transport etc.

    Btw the Duplex 86 appears to have cost more than the AFR suggests and heavy freight vehicles are banned from using this purpose designed low clearance tunnel.

    From a distance EWL has degenerated into a political divide which may obscure benefits or otherwise of the project. Build it, and they will come…

    • Seeding confusion?

      The $6 billion to $8 billion of the East-West refers only to the first section to be built, namely the EASTERN section of the EWL, joining City Link with the Eastern Fwy. This will be built first.

      The total cost of the EWL of some $15 to $18 billion also includes its WESTERN section of the project, joining City Stink with the western (section) ring road. These are actually separate roads, linked by a section of City Link.

      Future “stranded assets”, who’s losses will be made up by taxpayers under some sort of income guarantee. PPP (Public Pay Pounds of flesh)

      • Having now looked at the route maps, despite often finding myself stationary in a car wishing for a viable alternative to Elliot avenue to get from Footscray to Alexandra parade and the Eastern Freeway, which the EWL nominally is, I can’t shake the feeling that the EWL’s only agenda is to bring new custom to CityLink, as it seems to provide a way for people who have no need of CityLink to suddenly want to use it.

  6. Précisément Stat Sailor.

    Hence my moniker:

    City STINK.

    It seems Victoria revolves around this great axis of evil. Might be something to do with the contract with the state government.

    Thanks Kennett, the gift to the FIRE sector that just keeps on giving!

    Now what’s going on with Sydney’s own version of the E-W, what’s it called, Westcon tunnel or something? I heard it is now misaligned with the planned new airport. Do tell, I thought the Suppository of Wisdom was an EXPERT on infrastructure. Self taught from the back of corn flakes packet. No need for benefit-cost analysis for his genius! Now in five minutes we’ll learn how to cure the world of all its diseases and build box girder bridges, but first….

    • The airport itself is not worth the money either cost benefit wise. The market in western Sydney flies the least (and I mean on average less than one flight a year).

      It will be small and owned by private interests; anything big would not have the demand to match. In the end Mascot will remain the big airport for the forseeable future. Yet 10 billion in road spending is goi g there for roads that don’t have traffic in peak hour – that’s what I call pork