The Auckland Council Chief Economist’s most recent economic commentary highlighted a vital challenge for the city: the rising cost of adding road capacity (and the dubious returns from adding capacity).

International studies have shown that expanding the roading network, either by building new roads and/or widening existing roads, typically does not keep up with the increase in the number of private vehicles because of induced demand. For instance, Duranton and Turner have shown a near one-to- one relationship between expansion of highway capacity and the kilometres of car travel in the US.

In summary, the more roads you build, the greater the incentive there is for people to drive on them.

But there are also other challenges. In the case of Auckland, geography makes it increasingly difficult and costly to add motorway capacity. Land availability constraints mean that engineering costs are escalating as most “easy” roading projects have already been completed. And because roads are funded by both the tax payer and the rate payer, we are subject to competing demands for roading in other parts of the country as well.

This is an issue that’s concerned me for a while. The rising cost of adding transport capacity is going to fundamentally affect the way the city grows. I’ll talk about that in a moment, but first, some numbers.

I tracked down some data on the costs to build roads over the last decade and the costs for the next few major road projects that are planned. To enable a standard comparison, I’ve estimated the cost per lane-kilometre – ie a $100 million project to build two kilometres of a two-lane road would have a cost of $25m per lane-km. Here’s a chart:

This chart should strike fear into the hearts of Auckland Transport and the NZ Transport Agency.

It shows that the costs to build roads have steadily increased in recent years, and that the cheapest major roads we’re going to build over the next decade are as costly as the most expensive roads we’ve previously built.

Prior to now, the most expensive road on a per-kilometre basis was the Victoria Park Tunnel, which cost around $60m per lane-kilometre. We’ll be lucky if we can build the Reeves Road flyover, East-West Connections, and Additional Waitemata Harbour Crossing for a similar cost – these projects are likely to blow that record out of the water.

Even a highway on the edge of the city – Mill Road – is going to be more expensive than last decade’s major motorway upgrades.

Why is this happening?

A fundamental cause of this cost escalation is that we’re out of easy-to-build transport corridors. It’s comparatively cheap to lay down highways on paddocks, and expensive to retrofit them into built-up urban areas or thread them through geographic pinch-points.

Going forward, the options for new transport corridors in Auckland are:

  • Dig a tunnel (eg Waterview);
  • Build a bridge or viaduct (eg Reeves Road flyover);
  • Reclaim some land (eg East West Connections); or
  • Buy and bowl some houses.

All of these options are costly.

However, as the case of Mill Road shows, the cost of acquiring a corridor isn’t the only problem. As I wrote in a post a few years back, there has been significant cost inflation in civil construction as a result of recent governments’ decisions to spend more building roads. Inflationary cost pressures have probably gotten worse over the last few years, driving up the cost to build just about anything in Auckland.

We probably have quite a lot of room to bring down construction costs. International data on rail tunnelling costs shows that while the City Rail Link isn’t unusually costly, it’s not that flash either. Spain and South Korea can build underground rail for more like US$100m per kilometre, or around 25% of the cost of the CRL.

What can we do about this?

As the Chief Economist’s report outlines, the rising cost to build roads means that it’s going to be increasingly difficult to keep up with transport pressures from growth.

There are three basic things that we can do about this.

The first is to ensure that we get the best use out of any new transport capacity we add. If it now costs us more to build corridors, we need to make sure we’re maximising their ability to move people. As I wrote last year, that means rapid transit first:

If you start a sentence by saying “we need more land for housing…” the next words out of your mouth should be “… and therefore here are some rapid transit investments we should make to support it.”

The second is congestion pricing to get the best economic use out of our existing road network. This isn’t easy to do, although we’re moving towards it. As the Chief Economist points out, we’re going to have to solve some political problems to get there:

…there are at least two pre-conditions to making any policy that wants to smooth road use work:

  • Good transport alternatives: Better public transport (frequency, speed, and coverage area) would entice more commuters out of their cars, freeing up the roads for those who continue to use them. It would also mean that road prices required to get a certain volume of traffic off the roads would be lower.
  • Impact on equity: Any road pricing scheme must address any differential impacts it might have on different socio-economic groups, in particular those who may not have good alternatives to car travel.

Finally, we need to work out how to build cheaper tunnels and bridges. We’re probably going to need to build more of them. It would behove us to do it more productively and at a lower cost. Learning from people who are building a lot of tunnels relatively cheaply – like the Spanish and South Koreans – would be a good start.

What do you think of the cost of building roads in Auckland?

Share this


  1. “Any road pricing scheme must address any differential impacts it might have on different socio-economic groups, in particular those who may not have good alternatives to car travel.”

    Does this same concern exist for public transport pricing schemes and those that do not have the option of car travel? It seems odd that we only care about equity effects of pricing on driving but not about alternatives (or other true essentials like water, food, etc.) Is that only because it is a new cost / people worry about the effects of change?

    1. Why don’t we have a communty services card discount for PT? We do it for super gold cars and student ids do why not for those who are most likely to need cheap transport.

      1. We can do this already, and we don’t need a new card [or the overhead to manage it] either.
        Its called “capped weekly” (or capped monthly) fares.
        HOP can do it, AT just need the balls to sit up and “make it so”.

        With that in place, low income earners who use PT for nearly everything will find their costs are slashed once they hit the weekly or monthly cap.
        And its not just low income folks we need to target with this either.
        Everyone should be targeted to use more and more PT, especially out of normal “peak” hours.

        Agreed that not everyone can use PT, but lets keep the focus on getting folks who don’t need to be on the roads, off them, and using PT.

        1. Already have that in Christchurch with the MetroCard. Maximum cost per day on the buses in Chch city is $5.10, and $25.50 max per week ($7.50 and $37.50 respectively for travel to the outlying towns).

        2. Great to hear GlenK. This means no one can use the usual (and very thin) excuse of “it can’t work in NZ”)

          Let’s get fare caps in place.

          BTW sorry for getting the “Glen” handle first LOL 🙂

  2. Congestion charging – IT can now manage this in a cheap sophisticated way. Bring it in now just to get the concept accepted – say $1 on the harbour bridge but only when more than ‘x’ vehicles per hour are crossing it. Simple App to tell drivers when they can cross for free – surprising how many people can and will adjust their travel times to save money. Attach charges to car rego fees. If a few old bangers are scrapped rather than pay their congestion charges who cares.
    Then with a working system new motorways can have their congestion fees planned before they are built – no surprises for drivers.

    1. No thanks! All you are suggesting is costs, costs, costs to be imposed on people who AT haven’t provided adequate public transport for.

        1. I agree. Peak period travel in NZ is unpriced. The government and Auckland should stop stuffing about and get on with introducing time dependent peak period tolls like in Singapore which are reviewed every 6 months.

          We build roads for peak capacity & without pricing demand is unconstrained. Some of the benefits include:
          – removing or delaying the need for expensive road upgrades that this article discusses, i.e. long term cost savings
          – more people ride PT meaning higher revenue and lower subsidies
          – Some drivers move to driving outside of peak and thus use some of the wasteful offpeak space capacity
          – Some congestion toll revenue can be used to cross subsidize PT meaning lower rates for property owners.
          – Some congestion toll revenue can be used to add additional peak PT services to cover the increased PT demand.
          – It provides an economically more efficient allocation of resources
          – It helps drive more dense development around PT nodes as more people want access to PT to avoid the tolls.

          Until such time a the congestion tolls are put in place high demand commuter parking locations should be hit with a congestion levy so that the cost of commuter parking (public and private) increases and is used as a proxy to manage congestion.

  3. “”Learning from people who are building a lot of tunnels relatively cheaply – like the Spanish and South Koreans – would be a good start.”” NZ has many clever and experienced engineers – I’d be surprised if there is some magic bullet that solves our problem but worth checking.

    1. Without fact checking, my guess is that Auckland’s geotechnical and seismic conditions will contribute to the cost. All (most?) tunneling equipment is imported as well. Though having a Spanish or South Korean construction company on a tender list would make it more competitive!

      1. You might know that Naples also has quite a big volcano on its footsteps, millennia of history buried underground that needs to be carefully unearthed and a peculiar business environment (mafia) and it still manages to build a much cheaper tunnel than AKL.

      2. Auckland doesnt have geotechnical or siesmic issues. Most of the city is Waitemata sandstone as seen in the cliffs, and under that is a harder sedimentary rock seen in the Hunuas- but most tunnels wouldnt get down to that. Some areas are volcanic basalt but thats mainly on surfce. Very easy to tunnel using the current boring machines.
        Seismic issues dont effect tunnels broadly speaking- unless you are cutting through a fault line. The big Christchurch quake with very high ground g was centered near the Lyttleton tunnels which were unscathed.
        Each boring machine costs around $80-90 mill for the 3 lane size used in Waterview tunnels, smaller would be cheaper. The other costs are for the concrete segments which line the tunnel, the disposal of the spoil and then the final fitout.

        Waterview was more than just two tunnels, there was the bridges and road works at each end , with a major flayovers at Pt Chev. Additional costs were the widening of the SH16 from te Atatu to Western Springs and with that went the raising of the causeway across the harbour bed.
        The actual tunneling cost per km would be easy to find out, but doesnt seem to be revealed. But as for everything to do with construction in NZ the materials, eg concrete, steel and so on is normally much higher than elsewhere

      3. Maybe we should get in touch with Elon Musk and his Boring Company. He’s talking about reducing the cost of TBM tunneling by a factor of 10 and seems to have a good record of putting his money where is mouth is to get results in his endeavours to date.

        1. he says reducing the cost by factor of 10 is a target, not something that has been accomplished. Essentially its taking short cuts

          “In an onstage talk at TED in Vancouver last month, Musk explained that to make this project work the team would need to cut the costs of tunnelling by a factor of more than 10, while also speeding things up. Shrinking the size of the tunnel is one way to do this, so rather than a typical 28-foot wide tunnel (8.5 m) for a one-lane road, the team will burrow 14-foot (4.25-m) tunnels and shuttle cars along electric sleds. This should apparently reduce tunnelling costs by three or four times.”

          Electric sleds ?…..hmmmmm, i wonder who could make those ?

  4. What do I think? Take the roads out of ATAP. They are only going to cause growth in traffic, dispersal land and impoverishment. Just think how much money there would be to spend on HR, LR, RT and active modes!!

  5. Part of the problem of increasing costs is capacity restraints in the industry. For all the major players there is no shortage of work available at the moment.
    I work for a minor road marking company and we are flat out a lot of the time. When we receive a tender opportunity for a new project where we will need to increase capacity to take on i.e. hire more staff buy more plant etc. We put up the price accordingly.
    If miss out there are plenty of other opportunities, if we still get it, then it makes it worth our while.

    If everyone is doing this then there is no surprise that there is a significant rise in costs.

      1. Potentially it could, if there were fewer major road building projects it might encourage the major players to put in a more competitive tender to get it. Especially if it came on the back of a period od expansion as they will suddenly have spare capacity they will want to use.

        There have also been increases in material costs. For example the cost of Road Marking Paint is going up approx. 10% this year and we are looking at a similar increase next year.

        1. Just paid a road marking crew a couple of boxes of beer to put yellow line on the road outside my house. If that cementy paint is getting expensive maybe I owe them another box 🙂

        2. That example certainly looked like it needed the yellow lines! The pot-hole filling anarchy movement has a similar benefit to society.

          I wonder if a pot of David’s paint could be used to re-route some disgustingly car-centric flow paths mapped out on our city by yellow dashed lines… the exit from Motat onto Meola Rd being one example, where the cars are literally directed to cut across the walking space that connects the kerbed footpath down the driveway and the footpath along Meola Rd. Only school kids and people visiting Motat and the theatre use it. Perhaps they don’t matter as much as the cars???!!!

      2. surely it comes down to basic supply and demand, the ronps have got to be a large increase in demand for road building in a small economy like NZ.

    1. A lollipop young man I know gets plenty of work. 6 or 7 days a week. He left school with no qualifications. His take home pay some weeks is a very good $2200. Many women in care work etc. do just as important work.

  6. What’s happened to all the doom and gloom that was trotted out on this forum about the Waterview Tunnel?
    That hasn’t worked out too well has it?

    1. I was never against the project just thought doing the NW widening without doing busway at same time was stupid.

      1. Fair enough.
        That’s the NZ way though.
        Completing the third main should have also been done when the electrification project was underway.

        1. Doom and gloom? Rather, it was celebrated as “completing the motorway network”, which would mean getting on with building the missing moods.

    2. I’ve used the Waterview tunnel four times, all just off-peak. Each time the traffic has been moderate, but have never managed to travel at more than 70km/hr due to dawdlers or trucks crawling up the hill at either end. Each time the bottleneck has been Onehunga to Puhinui and joining SH1.

    3. 1 Month is way too soon to tell how well its working.

      When the next major motorway infarction happens on the bridge or the Newmarket viaduct, or some clown takes out the Penrose over-bridge and that side of the motorway with their over-height-limit load causing yet another bridge strike, then lets see how well it lets the network cope.

      After all this link was touted as being all about providing alternatives to SH1 north of Manukau and south of Albany including the harbour bridge. Waterview is all about providing network resilience they said.

      So until we actually have a problem on that part of the network between Manukau and Albany that tests that resilience, then of course the new link will work fine.

      I’m pretty sure we’ll have a good idea of how its working by the end of the year though.

      And 5 years from now, all the touted benefits like resilience will have been lost thanks to induced traffic. Unless we invest in demand side management before then that is.

      1. Yeah right.
        It was predicted on this forum that chaos would ensue once the schools came back from holidays.
        It’s working just fine and even has plenty of safety (speed) cameras to help pay for it.
        A brilliant piece of infrastructure.

  7. Our road construction are not very productive. When ever I passed our road works, I often see 4 people watching one person do to the work.

    One person in suit holding a file to watch
    One person is along with the suit guy
    One person supervising
    One person smoking in the car
    And finally one poor guy doing all the work

    Thats why our roads are so expensive

    I believe the root cause is NZTA just accepted whatever quote they are given. The quote are generally rip off price compare to overseas.

  8. If only we had some sort of government owned entity that could use size and monopoly position to drive massive cost efficiencies. Some sort of “Ministry” that would undertake “Works” or something

  9. Good post Peter! Australia is following a similar depressing trend. I think that also has an effect in NZ. Staff shortages do bid up prices. Sydney is currently a wastravaganza of freeway tunnel projects, all at prices in the billions. They are sucking up engineering resources from elsewhere in Australia and probably NZ too.

    The cost of land acquisition and service relocations is going up on all construction projects regardless of staff costs. So this problem is not going away. It highlights one more of the negative impacts of spending money on unneccesary or damaging projects like EW Link – they not only reduce funds left to build what we do need, but increase their cost as well. The best way to build a new rail or PT project quickly and cheaply in Auckland would be to delay/defer EW Link 2 or 3(00) years. A steady funding program with predictable cash flow over time is much more efficient than an industry with short term spikes in work and costs. Better quality work too.

    Finally, as to why does this happen, industry lobby groups with noble sounding names are forever badgering ministers to plough another billion into the construction industry. See

  10. With Waterview the urban motorway network in Auckland is all but complete, some rats and mice to go, but it is now functionally connected. Time to add the complementary Rapid Transit one so that it will still function. Oh yes and price it. Done.

  11. I’ve just seen a prime reason why reading costs in Auckland have risen so much. Case in point – 2 men unscrewing a metal panel on the K Road overbridge. About 4-5 m from the road. Also in attendance: an entire row (or two) of orange road cones, and three large road closure trucks, each with two people in. Totally bloody unnecessary. How much did that shenanigans cost? In Spain they probably just send one man out on a scooter, with a screwdriver. Total savings: about 99% of our bullshit “Health and Safety” bogus cost.

    1. I think it is fair enough that we provide fool proof protection to people working on a road while 1.5 tonne machines pass at 50+kmh. While there is some pretty big box ticking in health and safety, the concept that people can go to work and come back in one piece is very reasonable.

      1. Yes, but they were NOT working on the road, but on a fence, behind the bus shelter, 4-5m away from the kerb. Nowhere near any moving lumps of metal. Someone in Council should send the invoice back. It’s a total rort of the system.

        1. The Council and its contractors work under Health and Safety regulations set by central government.

          If you don’t like them, tell your MP.
          Better still, vote for a change in September.

        2. Agreed. “Safety” is turning into a rort.

          Meanwhile, the real source of danger which is our whole, excessive reliance on a poorly-regulated road-transport system continues unquestioned.

    2. Sorry Patrick, there are endless projects that are probably being dreamed up even as I type. (whoops -forgetting that they likely only work 10 to 4). Here’s one reported from the Devonport Flagstaff of 28 July. “a flyover to give access from a high occupancy lane on the south side of Esmonde Road to the northbound side of the motorway..”
      I had never realised there was a problem in this way. Endless expenditure of public resources for very little return!
      And dishonest. Consultation on Esmonde Road has just finished and this proposal was nowhere in sight.

  12. For comparison on the chart above, the Melbourne Metro Tunnel will cost $11b AUD including stations, rolling stock, and signalling, and be 9km in length.

    There are a number of reasons why Australia is expensive, but the high wages of construction workers are a large part.

  13. One of the reasons for the escalating costs could be that the roads are being built to a higher standard than previously, and methods used in the past are not acceptable any more. When the North Western causeway was originally built, the base was bassalt quarried from nearby Mt Albert. Try doing that today. Nowadays, most new motorway works are sitting on bases of rock quarried from the Hunuas or Pokeno. However, expanding public transport through more bus services also has its problems, as modern buses run intensive services on streets that were never designed for them, many of them little more than bitumen and chip laid on a minimal base. Fine for the residents’ cars when the subdivisions were built, but no good for a three axle bus today. By the way, today’s safety regulations weren’t dreamt up by someone looking for something to occupy their minds, they usually came in response to an incident, and the desire not to let it happen again.

  14. “In summary, the more roads you build, the greater the incentive there is for people to drive on them.”. Unbelievable. Just unbelievable. The age of the Luddite continues, unimpeded by logic or sanity.

    1. It is incredible.
      Despite experience:
      Despite research:
      Despite the dependency people have on cars in our own city, and the lack of opportunity this means for people who are too young / old / unable / disgusted to drive, some people still don’t get the simple statement you quoted.

  15. Between 1977 and 1995, the productivity within the non-agricultural industries just rose
    about 1. There will not bbe a worry in transferring for waste simply because your machinee make use
    of the complete trash providing as energy due to gasification process.
    It can save much time for entrepreneurs to sivn a partnership overr the Internet.

Leave a Reply

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