Feb 15, 2012

Why do subways cost so much more here than elsewhere?

There‘re few issues more likely to get the juices flowing than debates about the cost of transport projects. How is it, public transport advocates ask, subways can be built in Spa

Alan Davies — Editor of The Urbanist

Alan Davies

Editor of The Urbanist

Costs of subway projects in selected countries (data prepared by Alon Levy)

There‘re few issues more likely to get the juices flowing than debates about the cost of transport projects. How is it, public transport advocates ask, subways can be built in Spain for $40 million per km, but in Australia and the US its hundreds of millions of dollars per km?

Or why have new rail projects been built in Perth for a fraction of what governments say they will cost in Sydney, Brisbane and Melbourne? The Perth to Mandurah rail line has more than 70 km of rail line, ten stations and a tunnel into the city centre, and was constructed for a claimed cost of around $1.7 billion. Yet in Melbourne, for example, a 3 km rail extension from Epping to South Morang is costing around $560 million.

Americans are just as concerned as we are about why they’re paying more for new transport infrastructure than other countries. Alon Levy at Pedestrian Observations has put together a table of comparative costs of selected subway projects in the US, Europe and Asia, adjusted for differences in purchasing power (see exhibit). Clearly caution is required in interpreting the table, but some of the differences are extraordinary.

New York costs are the shocker, but look also at the proposed Crossrail project in London. The 22 km of tunnelling required for the project is estimated to cost $1,000 million per km, whereas the 5.8 km Sants to La Sagrera tunnel under central Barcelona completed last year cost only $40 million per km. (Differences in the length of projects only have a minor relationship with per kilometre cost, at least for this sample – and if New York were removed on the argument it’s an outlier, there’d be no relationship).

Alon Levy doesn’t address in any detail why there are such large variations, but he points out that although labour costs are generally lower in developing countries like China, so is productivity. Further, while the high Japanese costs are probably due to strong property rights, Italian costs are much lower than US and UK costs despite weaker institutions.

Some of the reasons the cost of projects differs are shown by an analysis undertaken a few years ago by Toronto transit advocate Steve Munro. He analysed a report by transit agency Metrolinx comparing the cost of tunnelling for Toronto’s new 6.4 km Sheppard Subway with that for the new 40.5 km MetroSur line in Madrid. Madrid is a popular benchmark because it has literally built hundreds of kilometres of new heavy/light rail lines over the last 40 years.

After adjusting for differences in how land acquisition is costed, he says the respective costs of Sheppard and MetroSur were $142.5 million per km and $87.1 million per km. Both lines opened at the same time. The key differences Mr Munro identifies are:

  • No environmental assessment was conducted on MetroSur
  • The standards for fire safety are more stringent on Sheppard
  • Stations are 50% longer on Sheppard
  • Non-tunnel construction was undertaken 5×12 on Sheppard, compared to 7×24 on MetroSur
  • Sheppard was built with dual tunnels, MetroSur with a single tunnel. Mr Munro says “the trains in Madrid are smaller and require a smaller combined tunnel than would be the case in Toronto. Single tunnels eliminate the need for cut-and-cover box structures at crossovers and effectively reduce the scope of excavation at stations where these crossovers are located”.
  • The use of cut-and-cover tunnelling on Sheppard was confined to stations, but it was used for 30% of MetroSur
  • It was more expensive to tunnel through the glacial rocks and streams of Toronto than the compacted sand of Madrid
  • MetroSur provided greater economies of scale as it was one of a number of projects. “Construction activities simply moved from one project to another rather than being reconstituted for each expansion, and more of the design was done during construction.”
  • Sheppard has two large interchange stations over its 6.4 km, whereas MetroSur has five interchanges over its 40.5 km

Physical factors like geology explain a lot of the difference between Sheppard and MetroSur, but so too do standards. Community expectations on a range of variables – for example environmental standards, engineering and operating standards, safety standards, the level of citizen input – appear to be key drivers of higher costs in Toronto. It seems there is a very strong commitment in Madrid ‘to get on and get it done’.

Comparing projects on a like-for-like basis is also a key challenge. The scope of what’s counted can often explain seemingly irreconcilable differences. For example, the Epping-Sth Morang price cited earlier includes some necessary related ancillary works such as 8 km of track duplication, stabling facilities and ‘downstream’ signalling improvements (although this doesn’t explain all the differences vis a vis projects like the Perth-Mandurah line).

One of the most important insights I’ve gained from looking at comparative data on cities is simply the enormous variation on a range of variables, whether it’s construction costs, residential density, employment density, public transport mode share, cycling mode share, and so on. The high variability applies both between countries and within countries. It’s going to be amplified when drilling down to the project level.

It pays to be very careful about assuming what applies in other places can be applied directly to one’s own city.

I’ve looked at differences between projects. I’m also interested in the complaint that projects are simply getting more expensive – see my post on this closely related issue, Why is infrastructure so bloody expensive? See also If WA can build rail lines cheaply, why can’t the other states?


Leave a comment

Your email address will not be published. Required fields are marked *

20 thoughts on “Why do subways cost so much more here than elsewhere?

  1. McCaskill Malcolm

    The Los Angeles MTA have addressed this problem by owning their own construction authority to build rail lines. They have just finished one line and are moving on to another. While this goes against the prevailing ideology in Australia that the private sector can do things cheaper, it addresses many of the issues that inflate costs
    – high project start-up costs (~20% of costs)
    – high project closure costs (~10%)
    – insufficient tunnelling experience by bidders
    – high target profit margins by bidders
    – high capital costs of specialist equipment (they own 2 tunnel boring machines)
    – costs of bank finance for working capital
    – retention of specialist skills
    – high salaries/wages because of the short-term nature of the work

    Because the Los Angeles MTA has a dedicated construction budget from sales tax revenue, and these flow to the same entity (without a rebidding process), it avoids many of the costs of short-term project work. If only we can get this message through to politicians that we’re paying perhaps 30% more on our rail infrastructure because there’s not a pipeline of projects. The issue is less in road construction, because the funding streams are more reliable.

    Subway construction is an area of market failure, because there are few purchasers (usually a combination of the Federal and State governments) and few providers (a few large construction companies). Competition is restricted because of immigration restrictions, and compliance costs that a difficult for overseas bidders to estimate. Are we better off setting up construction authorities that contract out components of the project, but have the flexibility to directly purchase their own tunnel boring machines and operate them if local companies don’t have sufficient experience ?

  2. Bellistner

    michael r james @ February 16, 2012: at least part of the difference in cost is the length of the tunnels. I understand the ‘set up’ costs for a TBM are not insignificant, so the longer a tunnel is, the more further you can distribute the set up cost. But the difference between NY and euro tunneling still seems manifestly excessive.

  3. michael r james

    @daft descender Posted February 19, 2012 at 11:43 am

    Dear Daft,
    That is one of the more ridiculous ideological bits of nonsense posted here. Take the NSW Waratah project that has blown up.
    From SMH: “But consider also what role the bankers who put together Reliance Rail, and the main company behind it, Downer EDI, played. They aggressively pitched for a lucrative government contract, under-bidding other interested manufacturers. They won the contract, and the financiers paid themselves upfront success and other fees of $50 million.”

    So this company rejected Australian manufacturers to accept a grotesquely under-bid contract from China, from a company that had no record of building such trains and which in the end could not complete the contract. The Chinese company had to give it to another Chinese company to finish them (hence the huge time over-runs and cost-over-runs). The NSW government has bailed the scheme out to the tune of $175 million though of course the total funds that have flowed offshore are closer to $400M. As Jacob Saulwick explained, despite all this the “clever” financiers who did this stupidity, keep their $50 million.

    So, even if the Australian-manaufacturing option was $200 million “more expensive” than the Chinese option, it would have been cheaper, as well as employing Australians and building on our existing heavy rail manufacturing capacity.

  4. deft descender

    Productivity costs are a big part of the issue IMO. Huge wages bills miles above award rates and on costs that more than double the final charge out rate. All of these excessive rates and imposts are driven by the participating unions that see any big infrustructor project as a cash cow to be milked until dry.

    Current IR “reforms” by Labor and the abolition of the building watch dog ABCC has made and will make futher projects even more expensive.

    It’s not just large infrustructure projects where costs are rising, your local builder servicing the domestic market is feeling the effects of changed regulations. In July even more reporting requirements will be imposed on contractors under new rules wished upon by unions. As a result the public should expect their next house extension to become more expensive.

  5. michael

    It’s those public-private “partnership” arrangements that pad out the costs enormously. What a scam. Just look at the dismal failure with the latest rail stock for Cityrail in Sydney. Over budget, long delayed and the few trains delivered are of poor quality. Private sector efficiency, indeed.

  6. Alan Davies

    Stephen Luntz:

    All explained by Daniel Bowen

  7. Stephen Luntz

    My recollection was that the South Morang cost estimates included a bunch of things that had nothing to do with the South Morang line – changes to the Hurstbridge Line for example, as well as building an underpass for a tiny suburban street that gets perhaps 1 car an hour.

    If this is true it suggests that the costings were being deliberately inflated to discredit the idea, and if that is the case it would hardly be surprising if they were being inflated at other points.

    Cost variations between cities of a factor of 2 per km are credible, even for similar type works. Variations of factors of ten make me very suspicious.

  8. michael r james

    @Stephen Posted February 15, 2012 at 8:52 pm

    Quite. And don’t forget cushy board positions when the politicians retire. Where did Bob Carr go on retirement? Macquarie Bank whose entities own several toll roads and the juicy monopoly of Sydney Airport. Could you wish for any more stark demonstration?

  9. feelgoodcause

    What about insurance? Thats going to be higher in some countries more than others.

    CHina would have none but replace damaged infrastructure at cost, some of the time.

  10. michael r james

    Here, I will take the liberty of reposting part of a comment on the Elon Levy article. Bottom line is that in NYC the phenomenally high costs cannot be accounted for by engineering type issues. (Square brackets are my additions for clarification.)

    [MobilMan says: 
2011/05/25 at 22:03 
I’m told that in these parts of Europe it usually costs €300/m³ to excavate for a basement of a normal house.
Let’s look at SAS [Seventh Avenue Subway, NYC] Phase 1: “$4.9-5.7 billion in 2007-17 for about 3 km”. Wikipedia tells me the tunnels are 6.7 m in diameter with 3,200 meters length: 2 * (6.7/2)² * pi * 3,200 = ~225,641 m³. The TBM launch box is 248m *23m * 15m = 85,560 m³. Let’s allow 50,000 m³ for additional shafts etc. The theoretical worst case (%100 for tunneling) is $13,565 to $15,780 per m³. Assuming 75%: $10174 to $11,835 /m³. 50%: $6,782 to $7,890 /m³. 30%: $4,069 to $4,734 /m³.
I have some numbers of a few German tunnels (2006 Euros) the cheapest being the Katzenberg tunnel part of Karlsruhe-Basel at slightly above €200/m³ (TBM).
U3 in Munich to Moosach (cut&cover, TBM), Adler tunnel Basel (TBM) ~€450/m³.
There are some tunnels built with the New Austrian Tunneling method (Hasenberg, average tunnels of Nuremberg-Ingolstadt) that start above €500/m³ and go up to €700/m³. The tunnel part of the Berlin Hauptbahnhof excluding the North-South tunnel cost €800/m³. Some of the geologically most difficult tunnels that will ever be built (though let’s hope they’ll stop the insanity) are part of the new high-speed line Stuttgart-Ulm. Even under the most desperate circumstances such as high-pressure karst with cavities under water and the New Austrian Tunneling method, costs would not exceed €1,000/m³ in 2010 Euros.

    So wow, we [NYC] have an order of magnitude difference. The complications of tunneling in a high-density city alone can’t explain the difference. What is it? Noo Yawk construction? Have I missed something?

  11. feelgoodcause

    Perth tunnel hit an unforseen which requires a “variation order”. Come in taxpayer. Private operator gets open cheque book.

    Go back to government run departments so your wallets don’t catch fire.

    I suppose the political party gets a nice donation as well.

  12. michael r james

    It is nice to have some engineers post some comments, however, the other comments are closer to the truth. This is mostly socio-political, and as we can also see, overwhelmingly an Anglo thing, in fact a neoliberal economic rationalist view that has reigned in the Anglo world for the past 3 decades.
    None of those engineering reasons account for the huge disparity between the London, Sydney, NYC and Brisbane versus Madrid and Barcelona costs. Or Tokyo (earthquake hotspot), San Francisco (loose sand terrain) or Paris (hyper-crowded with existing subterranean structures).
    In Australia it is worse because we tend to look to London as our guide –I cannot see a worse possible approach in this world. As I wrote on this blog last week (and previously in Crikey articles) the absurd cost estimate of any HSV for the Australian east coast is in fantasy land and up to 4 times the current estimates of building such things in France and Spain (and incidentally the UK has not actually built a single HSV of its own yet).

    The proposed Brisbane Metro, a single line of about 9 km, is costed at $16 billion! (The current version being touted at $8B is not the full plan.) It is a tunnel so no expensive property resumptions (or very little; I do not understand Beachcomber’s comment because the line will have a station under Albert St not impinging on any buildings.) It surfaces at both ends on existing government-owned railway land.

    So the real reason is that in places like Brisbane the current set of politicians are half a century behind the times and in thrall to the road lobby; and to be fair to them a majority of voters are likewise. If Newman becomes Premier, forget a Metro for another generation, especially as all the car toll tunnels collapse financially in the next 5 years. There is also the developer and construction lobby. In other places in the world, construction companies embrace building public transport but not in Australia, possibly for the simple reason that none of them have much if any experience in building them!

  13. beachcomber

    Most subways in Europe have entrances that are small gaps in the footpath, surrounded by an iron fence, with a sign. The new underground proposed for Brisbane, to go under the river, requires the removal of an entire high rise hotel, to allow construction of a station. The engineers have delusions of grandeur, and think erections above the ground glorifying their magnificance are more important than the important bits underground that people will use.

  14. Stephen

    Admittedly, Perth is built on a flat sandplain, which makes it easy to do Lego there. But, by and large, rail is deliberately designed to succeed in Perth, where there are no private tollways. By and large, rail is deliberately designed to fail in Sydney and Melbourne, where there are private tollways. Routinely, corrupt state treasuries staffed by ‘professional economists’ multiply the true costs of rail and divide the benefits, whereas they divide the true costs of tollways and multiply the benefits. It keeps on happening because both parties are committed to the corruption.

  15. IkaInk

    I’m a firm believer that a good portion of the added expense in Victoria happens precisely because successive Governments have demonstrated that they are prepared to pay more and more for similar projects. Railway stations are the prime example. Each time new stations are being built they are millions more than previous stations, sometimes over a matter of months between projects.

    Why would any contractor offer to do it for less if they know they can get paid more?

  16. Michael Wilbur-Ham (MWH)

    On Monday night I missed the 10:07 pm train home by 2 minutes. I had to wait 28 minutes for the next train.

    It would have been quicker to cycle the 15 km home (waiting time plus travel time), and of course if I took a car / taxi home I would have been home before the train even arrived at Parliament station.

    Public transport infrastructure is so expensive in Victoria because neither Labor nor Liberal care about PT. Talking about a big expensive tunnel means they can ignore all the fairly simple practical things that need to be done NOW to improve the system – like more trains running at night, and trains running well after midnight.

  17. Venise Alstergren

    Put away all those graphs and learned inter-state comparisons. Of course Victoria would have the highest costs. How else could our developers pay our politicians, if not from over-estimating their accounts-aka soaking the public?

  18. Jim Wright

    I came to Australia to work on the Rail Standardisation project in WA. I recently did some comparisons (admittedly off-the-cuff) between the cost/km in 1966 and in more recent projects, allowing for the decline in the value of money. Through some other sources with matching interests, I was also able to compare desalination plants in South Australia and Victoria (the Wonthaggi project – ugh!). In general, I found that the costs of comparable railway projects (e.g. no tunnels and other expensive adjuncts) were about twice what one would expect. The delivered cost of water from Wonthaggi would be between 3 and 4 times the cost in South Australia.
    I believe that one explanation lies in the funding models used. Contracting out the work and operation to PPPs or to finance companies operating on the hierarchical Babcock and Brown model is intrinsically more expensive. I can identify at least three possible reasons for this. The first is that allowance must be made in the price for profit. The second is the cost of risk. The third is the limited time in which costs must be recouped and profits made. In “the olden days”, the design would be carried out by government engineers and architects (or outside consultants such as the firms I worked for, who received fees according to a formula). Tenders would be called for construction and the winning tenderer would carry out the work under the supervision of the designers. Funding would either be provided from general income or through the issue of bonds. Another important point is that the retrieval of costs over a specific time was only ever an estimate for management purposes. It was never a driver, because the lifetime of an infrastructure project is such that many new sources of revenue and other indirect revenue from the community benefit could arise, even though not anticipated at the time of construction.

  19. hk

    To assist in meaningful project cost comparisons, maybe more insight would be gained by including quantifiable civil engineering unit cost rates for concrete casting etc as well volumes of concrete per km for the various projects. The same applies for steelwork, track installation, signaling solutions and similar. Land acquisition costs and service relocation, always a major cost item could be separated out. This level of detail in making local and international comparisons has served the bridge designer well over many years, and assisted Vicroads in choosing cost competitive structures for Victoria. Politically selected solutions for improved links in the PT network required similar rigour.

  20. Austin M

    The question also needs to be asked how much of the Perth to Mandurah Railway was discounted by being co-located in the freeway reserve in the areas near to the Perth CBD (which are typically the most expensive bits of land in any city).
    The co-location would have made all the overpasses of the freeway reservation longer and thus more expensive (most of the beam spans were cast onsite/near site rather than pre-cast and trucked as is typical). The freeway grade would have been limited by the rail grade thus increasing the cut/fill and costs. The increase in cut and fill would require more extensive batters/reservation and more land acquisition costs.
    The point being that the original freeway would have effectively subsidised a lot of the land acquisition, formation construction, and structural costs that would have otherwise occurred if the railway was to be built in an independent reservation. (i.e. how much would it have been if the rail was funded to be built with alowance to upgrade to a freeway as opposed to the logical alternate that was delivered, and how did the freeeway costs compare in terms of other similar freeways being built at the time)

    As you have pointed out tunnel costs are much more dependent on geology, environment and methodology (with risk and overheads thrown in for good measure).

    Recent projects rise in cost is another good topic (look at peninsula link compared to pakenham bypass in terms of costs. Both have a similar length, and number of bridges, etc.) Granted penlink has more REwall but the cost difference between them even with the deliverey model considered and slight increases in industry costs doesnt seem to add up. The question is are the most appropriate deliverey models and arms of government delivering the coresponding services (i.e. for planned metro tunnel – Melbourne water or similar who do lots of tunneling for sewers etc. may be better capable of planning and delivering the tunnel component of the project etc.)

Share this article with a friend

Just fill out the fields below and we'll send your friend a link to this article along with a message from you.

Your details

Your friend's details