I’ve criticised Auckland Transport in the past for having so many items on their closed agenda. For example this was from their meeting on Tuesday:

Items for Approval / Decision

i. Rail Deep Dive
ii. Pitt Street Lease Expiry
iii. Relocation and Disposal of AMETI Property
iv. Investment Framework
v. Draft Election Policy

Items for noting

i. RLTP/LTP Update
ii. Internal Management Audit
iii. Health and Safety – May Report
iv. CRL Update
v. AMETI Communications & Engagement StrategyUpdate
vi. Glen Innes Tamaki Cycleway

I still stand by that criticism however I also have noticed that AT have been pro-actively releasing papers from the closed sessions after they are no longer considered confidential. This is a good thing and I think more council and government agencies should take this approach.

One of the papers that has been released from the April meeting is a fascinating comparison of a wide range of metrics between the Auckland and Wellington rail networks. The authors note that it can be very difficult to do a proper comparison due to issues like

  • the length, layout and topography of the two networks
  • how the services are operated
  • the differences in the age and types of rolling stock
  • the maturity of the Wellington network vs the rapid change being experienced by the Auckland network.

As such it is far from a complete comparison but does provide some useful bits of information about the two networks.

First up is a comparison of some key statistics.

AKL vs WLG Key Metric Comparison

The metrics show that Wellington rail commuters are generally travelling a lot longer than those who use trains in Auckland (23.7km per trips in Wellington vs 13.7km in Auckland). On a cost basis Wellington commuters also pay more however that reverses when you compare the average fares to the average distance travelled. On a per KM basis users in Wellington pay about 15c per km compared to 21c per km in Auckland.

Perhaps the most important difference is the operational costs. On a per km basis the difference will be even far more pronounced however we don’t have the number of service km’s that were run to do that comparison properly. Further on the report does break down the operational costs further though.

AKL vs WLG OPEX Comparison

There are a number of significant differences between the two cities. Some of these are explained as:

  • Fuel costs are obviously a lot higher in Auckland due to running diesel trains. The diesel cost was $3.86 per service km compared with an equivalent energy consumption of $1.22 per service km in Wellington. These should come much closer together once electric trains are rolled out across Auckland.
  • Labour costs are considerably higher in Auckland. The authors aren’t able to give a definitive answer for this but suggest a combination of factors might be at play.
    • The mixed fleet meaning Auckland had two separate driver rosters (Loco drivers and DMU drivers) combined with now former situation where some drivers were hired from Kiwirail at a premium rate (it changed in January this year). This is also thought to have led to an increase in driver training costs.
    • Slower trains which means increased trip times for services and therefore more crew hours are needed.
    • Auckland’s costs include all of those incurred by Transdev whereas it is suspected that in Wellington some support roles and corporate overheads are effectively absorbed by Kiwirail.
    • Different fare collection staffing models. They note that it wouldn’t be possible to replicate Wellington’s fare collection model without potentially a lot more staff and/or fare leakage.
  • Station expenditure is higher in Auckland. Britomart alone costs about $3.5 million per year to run and all stations in Auckland have more extensive use of CCTV and security patrols.
  • Higher rolling stock maintenance costs due to the aging diesel fleet including approximately $5.7 million to Kiwirail for facilities, management overheads and hiring the diesel locomotives. The cost per km to service Auckland’s trains is $7.32 per service km vs 2.71 per service km in Wellington. The Auckland costs are expected drop significantly after electrification.

The one area Auckland does seem to exceed in is with the customer satisfaction scores which are significantly higher than those in Wellington. I suspect there’s a heap of reasons behind this and perhaps one is Aucklander’s are more accepting of crappy infrastructure/services as we don’t have the history of high quality to look back on.

All up a fairly fascinating report and while Auckland doesn’t look good in many of the metrics the good news is that improvement are on the way. The Auckland network should move much closer to that of Wellington from an expenditure point of view in coming years as the electric trains are rolled out and the savings they provide. That is likely to also be influenced by the re-tendering of the rail services which I suspect will attract a number of bidders from international rail operators as well as Kiwirail.

Transdev is hoping to secure a longer contract from 2016 following the transition to electrification, although Scott expects to face competition. Auckland Transport is asking prospective operators to attend a market sounding event on July 2 where it is seeking interest for the city’s passenger rail services for mid-2016

Share this


  1. Having lived and ridden through the years-long upgrade of the Wellington rail system, I suspect there is lingering dissatisfaction with the service from that. Glad to see that operational reliability has improved in Wellington. Hopefully satisfaction will pick up. Wellington lags on taking advantage of density possibilities at stations other than downtown.

    1. RE the denisity possibilities next to stations. Absolutely right. Areas around stations, especially in the Hutt, are crying out for medium density mixed use development. Maybe the council is reluctant to support this because of the already struggling Lower Hutt CBD, but that’s the wrong approach.

      1. Maybe the inherent problem is that there is already little correlation between the denser developments near train stations, and people living in them actually ever catching the trains. So there is little point in doing more of it than there already is.

        Park-and-ride is the only thing that has kept commuter rail alive (on massive life support, at that) in the modern era in most cities. The systems in the outlier cases of Manhattan and central London is a convenient gouge for wealth transfers to the global finance sector, in the form of mobility for their workforces in their unique local economies. It is ironic that the highest income locations in the world capture the most public subsidy.

        Hong Kong did it right, and integrated the land ownership (and hence rent capture) with the subway enterprise.

  2. I would expect to see Auckland beating Wellington in almost all those metrics within 5 years or so

    1. If AT staff refuse to use trains, why would metrics improve? Heaps of unhappiness to come with the roll out of electric trains, line disruption from CRL, cheap signaling software means frequency will be limited, and Kiwirail will continue to give freight a priority. In Wellington Kiwirail staff get free trains to work each day, so any problems are instantly identified.

      1. “In Wellington Kiwirail staff get free trains to work each day. . .”
        Not free. Actually 75% discount. But your point is valid that many Kiwirail staff use the service (though not sure how many of the managers do, or Regional Council staff who now control the service).

        1. And if something goes wrong, the colleague who is responsible will soon know about in a manner much stronger than end of month statistics of inter department memo.

      1. How about on public subsidy cost per rider km, including capital costs? When do you expect to see Auckland “beat” Wellington on this, and by “beat”, I mean a lower cost, not a higher one.

        I don’t see that in the above data.

        I would also like to know when the public is ever going to be told by the media, just how high that cost is, and why the NZ Herald won’t even print a letter raising the subject.

  3. The track length and distance travelled per passenger wont increase for Auckland until Pukekohe is electrified AND new track is built to extend the network. Wellington has those long lines up the Kapiti Coast and Hutt Valley which move large numbers of people very long distances.

    1. not necessarily a good thing. Those long distance services heavily commuter oriented, and few off-peak or anti-peak trips. Auckland has much more potential to be an all day metro type service, Wellington still very much stuck as commuter rail and little sign of change.
      In Auckland distance travelled may drop with high frequency making it attractive to make short hops. especially with CRL.

      1. > not necessarily a good thing

        It’s not at all a good thing. In a regional sense Wellington really sprawls, even compared to Auckland, with long lines of disconnected dormitory and retirement suburbs up the coast and the Hutt Valley. (“Satellite Towns”, is I believe the polite term). Mostly it’s inevitable, because of the geography, but partly just because the development is a bit leapfrog. Greater Wellington’s ended up with an urban area about as large across as Auckland, despite being home to far fewer people.

        1. Ironically, as Alain Bertaud’s research has pointed out, the more systemically unaffordable that housing is, the longer commutes people are forced into because they cannot afford the housing nearer to the city centre. This effect is pronounced all the more in Wellington by the fact that it is an international outlier for centralisation of employment.

          This is why it is also an international outlier for commuter rail mode share for a city of as small a population – a combination of geography, corridor form, centralisation of employment, and deliberate neglect of road capacity, making it also an international outlier for traffic congestion delay. Even Los Angeles, the worst in the USA, is not as bad as Wellington, with only 4% of the population that LA has. US cities of half a million generally have a congestion delay 80% or more lower than LA – it is an incredible achievement for Wellington to actually beat LA and indeed cities like New York, Paris and London on this metric.

          I raise the question whether the cost burdens from this congestion delay affecting so many people, justify the gains alleged from a few percent of people commuting to the CBD by train instead of by car, especially when the cost of the trains is so high per person km of travel on them. The public subsidy cost is many times higher than the public subsidy and externalities to motoring in per person km terms, and even the total cost per person km is quite close to automobility.

          I think this is probably the number one reason Wellington’s economy is under-performing even though its per capita government spend is the highest in NZ. Capital cities generally have the highest per-capita spend and generally benefit from this weightless income. London, Paris, Washington……. But Wellington is an odd one out because its small population means its commuter rail system is cost-inefficient even with the outlier advantages it possesses, and of course running “transport” policy on the basis that congestion is a tool to increase commuter rail mode share by a few percent, would be a massive net cost-burden-increaser.

          Ironically, if housing was more systemically affordable and employment was more decentralised, there would be more scope for shorter commute distances by whatever mode. You could have no commuter rail at all, and major efficiency gains would be possible in shorter commute distances, less congestion delays, and more walking and cycling to work. If you want to suggest that centralisation of employment, systemically unaffordable housing and ribbon sprawl still enhances walking to work opportunities (presumably because some rich enough or rugged enough people live in Wellington CBD) I would love to know what the response would be from the Wellington CBD workers priced out to housing in the Wairarapa.

      2. This is what I do already. I live just 3 stops down the line from work and frequently get on/off at the intermediate stations between home and work to go shopping and do other activities. I mix train with the bus (since the timing of both services aren’t too different for my journey patterns) but I’m sure with better frequency I would take the train for many more trips.

      3. Luke C – Wellington’s longer-distance routes actually have quite a lot of off-peak and contra-peak trips. They’re pretty well patronised all day in both directions.

        1. Wellington has longer distance trips because it has a big swathe of empty space between its city and suburbs. That doesn’t mean much except that it is a long distance between the hutt and Wellington, arguably that is dis benefit that costs more in operations and time to get people somewhere useful.

          Auckland luckily doesn’t have ten km of empty hills between Newmarket and Downtown, so we are better off.

        2. Hmmm, I don’t see particularly full carriages any time I see a Wellington Metro train anywhere at off-peak time……. the cost per rider km at these times would have to be several times as high as that for private cars, or even more so, for shuttle vans which the Akl Council understands is the best option for its own staff.

          1. Perhaps a car would be cheaper, a single car, but we both know more than one person is catching the train interpeak. Besides, the greatest cost for transit is covering the peak, interpeak is basically a free lunch.

  4. But burning money on diesel and clapped out old trains will cease over the next two years. And ridership will continue to grow, well if frequency, service hours, off peak, etc does. Roll out of interchange stations, bike ways to stations, and integrated fares….

  5. The longer distance travelled by Wellington commuters is the result of two factors:

    1) Upper Hutt and Waikanae/Paraparaumu passengers are incentivised to use rail by offering express services, which are more competitive than all-stoppers with car travel. Auckland should do this as well.

    2) Wellington has outer urban trains as well, which increase the overall average. Auckland doesn’t, and in fact is planning to shrink its network back significantly from that envisioned by the Rail Development Plan. Auckland needs to reverse this thinking.

    1. how is it planning to “shrink” its network?

      AC/AT have proposed to open more stations (Drury, Wesley, and Paerata) than they’ve proposed to close.

      It seems to me like they’re proposing to close underperforming stations, while opening new stations in areas that are likely to perform better.

    2. P.s. And giving huge subsidies to long distance travel is not necessarily a good transport/land use strategy in the long run, especially when you consider long term impacts on locational choices and land use patterns etc … not very “sustainable” – in either an environmental or a fiscal sense.

      I like this quote by Anthony Trollope (rather unfortunate name) made in 1857: “It is very difficult to say where the suburbs of London come to an end and where the country begins. The railways … have turned the countryside into a city”. You can see it at the TfL museum in London.

      Railway induced/enabled greenfields development is still sprawl, just slightly more concentrated. Unless you provide huge P&R – in which case it is sprawl, pure and simple.

      1. Stu, satellite towns and cities are not sprawl. Pretty much every major city on the planet with a rail service, has urban and outer urban trains. I thought you were in favour of Waikato trains, linking all the towns between Auckland and Cambridge? Have you changed your mind?

        You seem to be in line with AT and National Government thinking, that rail should be urban only, and everything else buses and cars on motorways. I don’t think many people here will side with you on that.

        1. Where rail investment stimulates new towns to develop that wouldn’t otherwise, then yes that is a form of sprawl.

          My views on trains to Waikato trains are a little more nuanced than you suggest. I’m in favour of investigating whether we could run an Auckland – Hamilton service (along the lines of what Nicolas proposed) and as part of that stopping in between where it makes sense.

          But it’s the anchors/endpoints that are important to the success of that line. That’s quite different to what you have out west beyond Swanson, i.e. a lot of track (capital/operating expenses) and not much demand.

          Ooo I love the old “you’re a tory and we don’t like your types around here” argument. For the record I’m a bleeding heart cheap riesling socialist who thinks rail works well in situations of high linear demands.

          I don’t think cars or buses on motorways are very good. In fact they’re more of a bad. Which is why I’ve argued against many of the Government’s (current and last) investment in state highways.

          Buses running on moderately dense, diverse urban streets makes sense.

          1. “Not much demand” – and that is precisely why AT have adopted a policy of making it all about Waitakere, and ignoring Kumeu. The “not much demand” argument disintegrates when you consider the massive traffic flows through Kumeu (more commuter traffic than there is from Pukekohe), and even more so when you consider current and future growth of that area, which is already beyond what MRCagney suggested a mere 14 months ago.

            Lester Levy in reply to my numerous questions about Kumeu, redefined every single one to make it about Waitakere as well. He bluntly refused to discuss Kumeu rail or acknowledge the history of plans for it, which extended even into his reign.

            “who thinks rail works well in situations of high linear demands”

            That’s precisely the view of the current government. Have you been to Melbourne, Victoria, Stu? All those regional passenger train routes radiating out from the city, to such places as Bendigo, Shepparton etc, could be handled by buses. Under your thinking, Victoria would have the extenseive rail network that it does. Trains would stop at the edge of the city. Don’t kid yourself, you’re not pro-rail.

          2. Stu, those insights are onto it. I have often raised the question, why are long train rides desirable? You sum it up beautifully:

            “……..giving huge subsidies to long distance travel is not necessarily a good transport/land use strategy in the long run, especially when you consider long term impacts on locational choices and land use patterns etc … not very “sustainable” – in either an environmental or a fiscal sense……”

            Systemically affordable housing and dispersion of employment would actually maximise the co-location options that lead to shortening of trips. Shortening of trips is by far the most important thing for transport efficiency, not the mode. And a shortening of trips increases the opportunities for walking and cycling.

            I argue that an integrated transport and land use policy that pursues dispersion and systemic urban land affordability, would achieve (along with good urban design) as great an increase in mode share for walking, as the status quo policy achieves in commuter rail mode share. The latter unfortunately involves long average trip lengths and a massive public subsidy cost. Walking would not need to involve any public subsidy cost, and systemically affordable urban land would reduce overall cost burdens on households by a considerable amount. Of course the big property and big finance sectors won’t like the sound of this.

            Also as an incentive to efficient co-location, mileage pricing and congestion pricing should be charged to motorists. The USA’s average urban density is many times lower than the UK’s and yet commute to work times are lower in the USA. This is because of co-location opportunities from systemically affordable housing; and less traffic congestion affecting most trips. If you socked a petrol tax onto the US driver as high as the Poms pay, you would see the already quite admirable co-location efficiencies in the US cities increase and average commute times fall still further to absolutely slaughter the outcomes in the UK (and all other back-to-front train-first-planning cities).

          3. HI Phil. yes I suspect our views are somewhat aligned. I tend to see rail (and public transport in particular) as potentially commercially viable in a future of accurate transport and land use price signals (e.g. deregulated parking supply, congestion pricing, removal of limits on intensification).

            I fully agree first best transport/land use policies are preferable to our current approach, which seems to be “he let’s screw up the price signals” and then “oh crap we need to regulate the resulting mess”. Our approach has lead to the reliance on blunt regulatory instruments, such as the MUL, as a means for keeping Auckland from dispersing “too much”. I use quotes on “too much” because I appreciate this is a subjective comment. However I do think the complexity of these issues and the paucity of local research mean that at some point you do have to make personal judgement calls on what’s optimal.

            And I think it’s simply at that point that we diverge, i.e. our personal intuition on the likely optimum that would be achieved once accurate transport/land use price signals are applied differ somewhat. I can respect your view, but I wouldn’t agree that shutting down rail is the right way to start …

            Instead, my focus in the short run is to ensure the Government investment in public transport that does happen seeks t maximise benefits and deliver systemic efficiencies/economies of scale (i.e. lower operating costs) thereby increasing revenues per user and decreasing costs (i.e. reducing subsidy on a per passenger basis). This will lead to higher cost recovery over time, and possibly even lower total subsidies.

            My focus in the long run would be on removing the wider transport/land use distortions on which subsidies are premised. And what is the end goal? Well, I like to think somewhere like Edinburgh’s bus system is not far off. That system is commercially viable with one notable exception: The cost of carrying concession passengers is subsidised by the Council. And this is effectively a “social subsidy” rather than a “transport subsidy”, which I think is more appropriate/justifiable. The remainder of the bus network operates at more than 100% cost recovery and is sufficient to cover capital charge on vehicles/depots etc.

  6. The planned network is being cut back from Kumeu to Swanson. They should be expanding the network with outer urban services like most cities that have rail services. Trains to Helensville, Waiuku and Tuakau as a minimum.

    Yes, I know your reasoning against this Stu, we already know we disagree.

    1. So … Kumeu to Swanson = underperforming parts of the rail network. AT ran a trial service out there and demand didn’t respond.

      And when you’re facing fiscal constraints (i.e. “budgets”) then it makes sense to focus investment where there is demand (i.e. “benefits”).

      And demand appears to exist south of Papakura, e.g. Pukekohe has been one of the faster growing stations on the network for about a decade. So trains to Tuakau may be a good idea, just not before:
      1. Electrification to Pukekohe; and
      2. New Stations at Drury, Wesley, and Paerata; and
      3. 30 minute all day trains to Pukekohe.

      Now we don’t seem to have money for #1, let alone #2, let alone #3 … does that help put your expectations of rail expansion out west into context?

      Let’s put it another way: Once we’ve funded all the worthy rail developments in the south (and elsewhere on the network), then it might be reasonable to run our attention to further west.

      P.s. Geoff, I’d be happy if we just “agreed to disagree” – but to be fair you do keep coming back to the blog and trotting out the same old arguments. In which case figure it’s worth me putting forward another point of view just to ensure our readers have a, shall we say, “fair and balanced” perspective?

      1. There has been no trial of a realistic Kumeu or Helensville rail service, only a poorly timed once a day train. The fact that you have to resort to such red herring arguments speaks volumes. As does the total and complete lack of public consult on the dropping of Waitakere. AT knew very well that any such process would have resulted in a revolt, so it skipped it.

        To kill off New Zealand’s longest running passenger rail service (133 years) at a time of rail growth takes staggering incompetence. Those responsible for it will have their names recorded alongside such names as Gerry Brownlee in the history books of New Zealand railway cutbacks.

        1. with all due respect I think you’re off-topic. This post is about how we can invest in Auckland’s rail to gain the greatest benefit, i.e. growing patronage.

          I think you’re talking about expanding geographic coverage, in between talking about herrings, public consultation, revolutions, killing off, staggering incompetence, Gerry Brownlee, and railway cutbacks.

          Back on topic Geoff! Back!

          1. I am surprised that with your insights on the way transport and land use interact, and trip distances, that you are a supporter of rail and UGB’s for NZ cities at all. The long trips associated with rail are a logical outcome of an elevated and spiky urban land rent curve, which is an inevitable consequence of the planning that endeavours to restrict automobile use.

            I am watching with interest what happens to public transport oriented development in cities like Atlanta, because this actually ends up affordable. “Build it and they will come” does not apply to situations where the end price of the housing means that wealthy yuppies are the only prospective buyers.

          2. not sure I agree that long trips by rail are:
            1. Purely an outcome of an elevated and spiky urban land rent curve; nor your implication that
            2. Planning efforts to restrict vehicle use are the only way to achieve such a curve.

            As mentioned to you earlier, my understanding (as a spatial economist) is that historical distortions in transport/land use pricing and policies have created cities that are of much lower density than what would occur in a situation of sustained accurate pricing. My main reasons for holding this view is 1) massive historical subsidies for highway construction in the early 1960s which were not financially viable and caused many of Auckland’s lower density far flung suburbs to develop and 2) ongoing policy settings that favour ownership/use of cars (e.g. minimum parking requirements) and low density development (building height limits).

            In contrast, you believe that restrictions on sprawl (MUL) and vehicle mobility (can’t think of one? Perhaps bus lanes…) are creating a city that is higher density than what would naturally result otherwise. I respect your view, but don’t think there’s much evidence to support it.

            I do agree things like the MUL and massive subsidies for public transport are a very very poor “second best” policy direction.

        2. Very off-topic, but 133 years is nothing for a rail passenger service in NZ. The oldest service has been running for nearly 148 years, between Christchurch and Rolleston – the oldest passenger station in the country.

          1. Can you still book a ticket between Christchurch and Rolleston? I was under the impression that ability was withdrawn some years back.

Leave a Reply