It has been really refreshing to see transport discussed so much in parliament this week – with the results of the exchanges spilling into the media, as evidenced by the interviews on Breakfast TV a couple of days back.

What seems to have really kicked this off are numbers coming out of the Ministry of Transport, and in a series of answers to written questions, highlighting the ever-increasing dominance of our transport budget by projects that have very low cost-benefit ratios. This was first highlighted in the Ministry of Transport’s briefing to the incoming minister – which included this graph: A series of written questions from Phil Twyford to Gerry Brownlee has dug up some further detail on the numbers that sit behind the graph above (at least for the last couple of years) and also updated it with 2010/2011 data. I’ve put together the answers to a series of written questions into the table below – first by dollar amount and then by percentage: Finally, a couple of questions asked by Mr Twyford look at the proportion of the state highway spend on projects with low cost-benefit ratios that are related to Roads of National Significance projects. The answers highlight that in 2009/2010, $527 million of the $587 million spent on projects with low cost-benefit ratios related to RoNS project (just under 90%). In 2010/2011, $468 million of the $583 million spent on projects with low BCRs related to RoNS projects (just over 80%).

Now let’s put them all together into a graph showing what’s happened since 2005/2006 – effectively adding the 2010/2011 data to the earlier graph in this post: 

Geez what happened from 2008/2009 onwards that triggered such a dramatic lowering in the cost-effectiveness of our state highway spending? Oh that’s right, the current government came to power and introduced the RoNS projects.

Share this


  1. That last graph is pretty brutal in its impact- that’s not agood look for the government.

    Though I guess they could spin it along the lines of “crucial RoNS for regions have been neglected/ignored in previous years because of a blind slavery to the so called wisdom of the “BCR’s”….”

    1. Shit. Brownlee is probably reading this.

      I’m not going to read that excuse at Question Time in parliament on Monday, am I?

  2. Absolutely shocking. What an ideologically-led, evidence-averse government we have.

    Just to be clear, what is meant by low, medium, and high in these graphs? I assume that low means between 0 and 1.0 (ie. benefits accrued are less than the cost of the project), medium between 1.0 and 2.0, and high over 2.0? I’ve looked through the BIM and they don’t describe these labels.

    1. “Just to be clear, what is meant by low, medium, and high in these graphs?”

      I wonder the same thing.

      I also wonder what any of this has to do with the RONS. As far as I’m aware, none of the RONS projects are approved yet. (Except Vic Park Tunnel?… but everyone seems to think that has a great BCR.) But both the graphs are clearly labeled “Percentage of $ approved”.

      1. As you say VPT has already started and is nearly finished but as for the others:
        The WRR has started now but I imagine there were some decent costs for the EPA process
        There has been work on the Waikato Expressway
        Tauranga Eastern Link has been in full construction for a while
        No physical work yet on Wellington Northern corridor but I imagine quite a bit of design work
        There has been quite a bit of work going on with the Christchurch motorways.

        1. Of course you’re right. I forgot about Waikato Expressway since it has been going on since the early 2000s. And Waterview was a Helen Clark project, tunneling under her electorate. I wasn’t aware that Tauranga had started. NZTA lists Waikato and Tauranga as both having BCRs of 1.8 (inc WEBs with discount rate of 8%), and Waterview has a good BCR as well. So it isn’t these projects making the graphs look bad.

          I had to drive through Hamilton over the Christmas holidays. At one stage I crossed a motorway over bridge, but there was no motorway underneath. I always assumed you’d build the road surface first and then add the junctions, but it looks like they were doing it the other way around.

        2. Depends on how you measure the project I guess, the extent to which you “break it up”. The cost-effective sections of the Waikato Expressway are the sections which have been built. They were pretty cheap and had massive benefits.

          It’s the sections left which are much more expensive and have less significant benefits which drags down the graphs above (as well as projects like Puhoi-Wellsford). I think the Hamilton & Huntly bypasses are the really expensive parts of the Waikato Expressway left, and they are the ones with really low BCRs.

        3. The Tauranga Eastern Link is something to behold. It’s huge, and connects Tauranga with the other side of Te Puke. To be clear, there are high volumes on an unseparated dual carriageway, and no doubt people die on this road too regularly. Improvements are needed. Whether they could have been done now at cost significantly of less than $100m (cf Project Lifesaver / Puford RONS), we’ll never find out. This road was rushed into with unseemly haste.

          While this is a heavily used road, that the road was chosen without any evaluation suggests that it’s more of a gift to the voters of the BOP than any necessary force for economic and social improvement.

      2. As Matt says, most of the RoNS have had their funding approved (except Puhoi-Wellsford and the Wellington one I think). They just haven’t been built yet.

      1. So measured using the same NZTA criteria that they apply to highway projects, the CBD Rail Tunnel BCR would be less than low. That’s insane, and maybe puts the graphs in to perspective.

        But… why is it to hard to find BCRs on rail projects? It took me about 20 seconds to Google the BCRs for Waikato and Tauranga RONS and find the figure in the linked page. But I wanted to compare these to the BCR for the Manukau rail branch, and couldn’t find anything. Why so secret?

        1. Obi our argument is, again, that the MoT and the government are producing unrealistically high BCRs for highways by using inaccurate vehicle volume and fuel price projections, and, as we have shown, produced one strangely low BCR for the CRL also based on crazy assumptions like infinite space for more cars and buses in the CBD, as well as a projection at the low end of use. But even their junk math produces BCRs that don’t support what they are currently uncritically planning to build.

          What would a rational policy look like? They should generate BCRs that reflect at least the possibility that the trends of the last seven years may continue, or even accelerate. A little humility in the projections would be good; I would like to see a range of possible outcomes. If they did this you would see that their current BCRs for State Highways are at the extreme high end of any sensible analysis. Too much to ask?

          Because they won’t do this, or even consider analysis of their assumptions one can only conclude that they are using the BCR process to justify doing what they want to regardless of the facts. Also evidenced by Joyce just ordering a new BCR with fiddled assumptions for Puford after the first one came in at 0.4. Let’s have one now based on the real current numbers, one that takes the difficult geography of the route and the impact of tolling on vehicle numbers into consideration.

          Would it get close to 0.2? The last one requires 4% pa rise vehicle numbers; yet we’ve had seven years of 0% growth. So why is it easy for you to find their BCRs?; because they are nothing but made up numbers for PR; to help you feel good about your taxes being pissed into a big hole in the Auckland countryside, and into Stan Semenoff’s back pocket.

        2. As it stands with Manukau – it was done as a total fist job by the former Manukau City Council with that station being short 110 metres, running over budget and missing its Southern-Papakura link where I bet your bottom dollar and a pint the bulk of your revenue customers would come from until MIT is built on top and can attract passengers from the what I call Inner Circuit (between Otahuhu and New Lynn incl. Onehunga)

          I have called Dick Quax out about the $6.12m saved from moving the station from the South Entrance of the Manukau Mall to its currently hopeless position. I reminded Quax that the extra $6.12m getting that station 110 metres further up than when it is now would of paid itself off very quickly with increased potential to bring in more passengers and get a situation like Sylvia Park where people catch the train to head to the mall.

          However we have to live with what we got with Manukau.

          The BCR for Manukau obi? Go lodge an OIA request with KR and Council.

      2. Obi, I’m pretty sure the AC/AT review of the CRL put the BCR at around 1.4. The Ministry of Transport review ignored the giant ‘elephant in the room’ by measuring bus capacity to and through the CBD as unconstrained. For a project whose benefits significantly arise from “adding transport capacity to an area that cannot easily add transport capacity in other ways” ignoring the capacity constraints is an utterly massive, completely unforgivable, oversight.

        I may be wrong, but I’m pretty sure I remember it being reported that the Onehunga Line had a BCR of more than 3. And that was based on ridership projections which were vastly lower than what’s turned out to be what’s happened.

  3. @ obi. BCRs calculated using NZ formulas typically under-estimate the benefits of public transport projects and over-estimate the benefits of motorways. There’s a range of reasons for that – but mainly it’s because of the parameters they use in the model.

    One major issue is the discount rates. We set them very high at 8% (like the benefit of the project diminishes by 8% each year after it is completed) which assumes that most transport projects won’t deliver benefits after 10-15 years. This is often, although not always true, for motorways.

    But it’s not true in the case of most rail projects because the infrastructure lasts a really, really long time (look at our shitty rural rail system, in some cases literally built a 100 years ago, had almost no cash put into it for 20 years and yet somehow staggers on).

    There are other issues with the model parameters like the projected cost of petrol they use (typically much lower than it actually is). But the other thing they did with the RoN figures to manipulate the BCRs and make them appear higher than they should be is that they included WEBs in the models. WEBs are typically used when calculating the benefits of public transport projects because they measure the value of agglomeration around a transport route, concentration of businesses etc. So it’s kind of crazy to apply them to a motorway because motorways typically don’t lead to agglomeration – they lead to low intensity, sprawling development.

    This is why there was such a difference between the BCR that Saha calculated for the RoNS and the BCRs that NZTA eventually calculated and published on their website.

    In terms of where the money is going – Remember it’s not just Waterview – the whole Western Ring Route is a RoN. I’m pretty sure that the Hobsonville Deviation was also massively sped up by an injection of cash from the RoN budget. There have been some other projects that weren’t RoNS but were sped up by government investment over last 3 years (e.g., Kopu Bridge, although that actually has a very high BCR but I know there were others in Taranaki and Rotorua).

    Finally, just cos roads haven’t started doesn’t mean they don’t cost money – I remember that two years ago almost the govt put $50 million (approx) into Puhoi to Wellsford, just to do the initial planning, consultation and route investigation.

    1. An 8% discount rate is pretty low actually. Remember we have to compare these to other investments as well as the cost of funds.We also need to incorporate the seemingly high uncertainty around the assumptions in the discount rate.

  4. These are the completed and open bits of the WWR – many of which have opened since 2008.

    SH18 Greenhithe Deviation
    SH18 Upper Harbour Bridge
    SH18 Hobsonville Deviation
    SH20 Mt Roskill Extension
    SH20 Manukau Harbour Crossing
    SH20 Walmsley Road Interchange upgrade
    SH20-1 Manukau Extension

    Some of these were probably built before National introduced the RoNS. But I wouldn’t be surprised if some got injections of funds as RoNS. But I could be wrong…

  5. @Amy: Every single one of the motorway projects you listed were projects of the previous (Labour) government, along with the three CMJ projects. All were built over their planned timeframe, and were not sped up. You mentioned Hobsonville was sped up, but in fact it was always planned to be built over four years, 2008 to 2011 inclusive, and it opened on schedule. I’m not aware of any RONS money being spent on it.

    The only Auckland motorway projects National has given us are the Vic Park tunnel, and Waterview (which themselves were planned by Labour).

    There are no original motorway projects in recent times built or planned by National.

  6. The most depression piece of news that I discovered today is the NZTA annoucing the so called “Holiday Highway” route and have indicated that a construction date could start, even though they haven’t annouced it yet. They haven’t even offically approved the motorway yet and it will cost a whopping $1.7b!!! I just hope in the next couple of years National are kicked out of government and this stupid projects gets dropped.

      1. It’s here

        Love the ‘Notes to Editors’. I’m terrible at that sort of maths but it strikes me that the BCR for the Warkworth to Wellsford section must be either around 0.5 or 0.25. I imagine most of the BCR benefits are in the Perry Road to Warkworth section and that the southern section of this would also be showing similarly negative returns.

Leave a Reply

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