AT have kindly sent us the Train Station HOP data for for the last two calendar years. Note that these data are incomplete, not including those travelling on legacy paper tickets, transferring, or on special event services. See here for Matt’s mid year post where on these data were then.


As expected these are great numbers; there’s spectacular growth across the network. Highlights include:

  • Manukau City takes off now MIT is open: 118% growth jumping in rank from 24th to 13th. Strong growth is likely to continue once the Bus Interchange there opens.
  • Panmure is the next big mover, leaping up 52%, from 12th to 5th. I guess we can expect a similar burst at Otahuhu too once the new Interchange is up and running.
  • Britomart adds a million new movements each way. The top 10 stations are now over 400k, last year only 3 were.
  • Next year should see Britomart over 5mil, Newmarket 1 mil, and most of the rest of the top 10 over 500k.
  • Grafton still the most asymmetrical station other than Britomart; 69k more alightings than boardings, showing that downhilling is still strong there. This is people heading to the city via Grafton but returning via another route, many likely using Britomart, which shows more some 169k more boardings than alightings.

AT HOP Station Data 2014-2015

Here’s the top 15 ranked by 2015 boardings. The positive movers are all on the Eastern Line, which has had the new trains the longest, and biggest upgrade in frequency. And the biggest two movers have shiny new stations: Manukau City with the new MIT above, and Panmure with a new bus interchange. The Eastern Line also has very good bones; it has no level crossings, is fast, straight and direct and now some good attractors to unlock those advantages. As well as the two stations mentioned above, the mall at Sylvia Park is clearly drawing customers by train, which adds to the long strong destinations of Papatoetoe, Middlemore, and GI. Even the minor stations on the line improve well over the year: Puhinui the 3rd highest proportionate change at 43.9%, Meadowbank; 5th, 33.0%, and Papatoetoe, by no means minor; 6th, 31.3%. Papatoetoe still the forth busiest station in 2015, but will it be overtaken by Panmure this year? Which would be impressive as Papatoetoe has twice the number of services. It is clearly time that businesses took advantage of all those people at Panmure station; it’s still sitting in a land-use desert.


It’s pretty clear what works; investment in stations and interchanges [Panmure], alignment with land use [Manukau City], and improved service. I think it is likely that the Eastern Line still has more growth in it, as the results of improvements to frequency and capacity on the Western Line planned for this year may not fully come through until next year. If we have learnt nothing else from the changes to places like Sylvia Part and Manukau City is that it can take a little while for these changes to be reflected in pax numbers. Although the lower growth percentages from Western Line stations does suggest they are being held back by capacity and frequency constraint [exception: Avondale; jumping 29.3% up one place to 16th busiest].


What else can we learn from these data?

Share this


  1. Fantastic analysis Patrick. It’s interesting to see how things have changed at Panmure and Manukau – the next test will (hopefully) be Otahuhu after its bus interchange is completed.

  2. On a related note to station usage: What is the latest plan regarding closing Westfield? Removing the station from near such a critical junction (NAL/NIMT) would help regulation quite a bit as well as speeding up overall journey time.

  3. Can see how Pukekohe is lagging without electrification… This route should be booming with a growing population and worsening congestion on the Southern Motorway, but who wants to ride an old diesel train.

  4. Well, we went Boxing Day shopping by train this year. Drove down to Orakei station and rode to Sylvia Park. So glad we weren’t circling the carparks in scorching heat for 30 minutes plus like all the other chumps. I think this will continue to catch on – long live the Eastern Line.

    Glen Innes is also begging for some apartments and further developed retail next to the train station. As you said it’s an amazingly quick trip into town from there.

    1. I also went to Syliva Park by train on boxing day which was so much easier than driving except for the fact there were only rail buses from Pukekohe which was a pain in the ass – so I drove to Papakura even through I live only 4 minutes walk from the Pukekohe train station.
      If the route had been electrified then I’m sure the trains would have been running and made it even easier.

  5. As this if for the 12 months to the end of December and my last post is the 12 months to the end of June it means we now have a six monthly update view in station changes. Having a brief look last night that shows some interesting trends. For example Panmure has moved up a couple of places even within the last six months. Will look to do a separate post on it.

  6. These are good numbers, but I suspect the Panmure numbers are going to plateau fairly soon.

    We’re 18 months from the new eastern bus network, that should deliver a big increase, I’ve no idea why this is taking so long to implement

  7. Wow, these numbers show the power of investment in public transport, and they show such impressive growth in such a short time. I’d make just a few observations:

    – the lack of North Shore destinations looks even more bizarre seeing the list of all 41 Auckland destinations aggregated like this
    – Aucklanders really want to use public transport, and newer trains and better timetables make it a good option compared to driving
    – it looks like we are really dependent on Britomart, with a combined 8 million arrivals and departures there. Is this dependency a risk?

    I would personally draw the conclusion from these numbers that we have an urgent need for the CRL – the Shore gap, the passenger growth and the dependency on Britomart station all point to it becoming an essential part of our network sooner than most people realise.

    1. yes the Northern Busway data could well be represented in the same table. Be good to see how the “RTN” compares.

      Isuspect a couple of NB stations would jump to close to top of list, given that it carries almost the same as a rail line but from only six stations (including Britomart) or seven if you include Silverdale.

    2. Yes we need rail to the Shore.
      Have posted before but the latest inflation figures were released today (-0.5% for the quarter and annual rate of 0.01%) time to do some quantitative easing of between $5-10b and pump that into big PT projects (AWHC – rail) and convert NEX to HR all the way to Albany, fund CRL (incl Beresford Sq entrance), HR to Airport via Otahuhu, electrify NIMT to Frankton (and Palmerston North-Paraparaumu, and Hamilton-Tauranga), LR for Christchurch.
      Creates a little bit of short term inflation (which is needed), long term lowers inflation by improved productivity and reduced costs, allows for lower interest rates and a lower NZ$ which helps exporters and the overall economy, not to mention the construction jobs in the meantime.

      1. +1

        Couldn’t have said it better (or more succinctly).

        [If only Muldoon hadn’t thought small for Auckland while he was thinking big for the rest of the country – we would have been enjoying a by now expanded Robbie’s Rapid Rail system.]

    1. Removing Westfield is still a great gain for all the Southern and Eastern Line passengers, and should mean a couple of minutes shaved from every journey.

    2. I am surprised Te Mahia station is not used more, given it’s good access from the Takanini interchange from the Southern Motorway

      1. Te Mahia is pretty dodgy station and it is hidden, also there is not really a Park and Ride.
        If AT is to keep Te Mahia, they need to upgrade the station and make it feel safe.

      2. I was told by a Manurewa Board representative that when they did an independent survey of this station they found significantly more users than the official figures. this suggests there is a lot of fare evasion at this station. But it is also the one of the worst stations in the southern network and it is very close to a major bus route on Great south road. All of which I suspect limit the use of the station. I’m told in 2017-19 Fletcher living is building a new housing area at the Manukau golf course which would be an ideal time to upgrade the station and hopeful the use of the station will increase. If they could add that small bit of rail between the southern line and the Manukau branch line and then run diesels to Manukau from Puke…..that would help as well.

        1. There are a LOT of works commencing just south of Te Mahia for the motorway widening and I half suspect a bit of something to be done to link Mahia Rd with Spartan Rd through that area in the future, with grade separation of some kind. A bit more wild speculation might suggest Te Mahia and Takanini stations eventually being shifted south to become a Spartan Rd and Glenora station respectively. That would tie in with various commercial/development agendas currently evolving. The obvious lack of development at both the existing stations does hint at either grand neglect or future plans are being considered.

  8. On a different note, does anyone here know if Auckland Transport is planning on integrating AT HOP with mobile payment? Ive heard that Wellingtons Snapper can be used with Semble. Be great to be able to tag on and tag off with my mobile device.

    1. There’s probably a bullet point on a plan somewhere. However at the pace AT move I’d be surprised if you see this before 2020 by which time the technology will have changed anyway

  9. Without me looking & for discussion, will the new network make better use of Orakei & Meadowbank? Catchment areas pretty small with all the water & slows down the Eastern Line somewhat.

      1. Orakei is pretty maxed out due to limited access, no buses go past it and the local streets are already saturated with park and hiders. The Park and Ride only handles 200 cars and fills up really early.

          1. Has anyone heard anything more on the ambitious project to build a large structure over the Orakei station/bridge/carpark etc? (nevermind, comments below have gossip)

    1. Grant,
      Under the new network as proposed (but not yet finalised) Orakei will benefit, Meadowbank not.

      The current bus service (655) that services Meadowbank station area (note, it comes about 300m close to but doesn’t actually go past the Meadowbank station as thats a dead end road) will be axed according the proposed new network services. There will be a much less frequent “wind everywhere slowly through local back streets” “connector” bus between there and Orakei instead. So Patronage will probably plummet as a result at Meadowbank.
      [and/or the hide and ride will get a lot worse]

      Orakei will be on a Cross town bus route – the one that is called the 007 will go back to using this route and renumbered – the new route is Crosstown 6, and goes right past the station in both directions, so should dramatically increase the potential catchment. From both Eastern Suburbs and Orakei/Remuera as well.

      But, depending on how the integrated ticketing handles Orakei station itself – currently its 1 stage from Britomart, under the proposed integrated fares it becomes 2 stops, as per the rest of the isthmus.
      So Orakei’s advantage/attraction as the cheapest and closest park and ride in Auckland will be reduced a lot by this change.

      I am really surprised by the numbers how well Meadowbank bears up compared with Orakei. But I know a lot of people walk to Orakei even though they live in Meadowbank as that saves them a 1 stage each way fare for the costds of a 800m or so walk. Whether that will continue under the integrated fares is too hard to know as both station may have the same price.

      1. Yes apart from fare/stage changes, Orakei should do a lot better. Plus there were those plans to build that development over it. Seems Meadowbank is too close & dead end area. That combined with community calls for a station out the back of Selwyn College/ASB Stadium, would seem like a good place to move it to.

        1. It would be a better location but it won’t happen anytime soon – the road/pedestrian access to it there is likely just too damn difficult (expensive) to do.

          Orakei development seems to have stalled big time.
          Early purchasers are demanding their deposits back after 2-3 years of nothing doing.

      2. To offer a counter-view, Meadowbank is rapidly overtaking Orakei.. the walk-up catchment far exceeds Orakei, and is likely to increase when the GI-Tamaki Shared Path brings parts of Kohi within a short, flat walk / ride to the station.. and a 10 min congestion free trip to Britomart (compared to bus 45 mins in traffic with no priority lanes). The SHAs and other developments in Meadowbank, some very close to the station will give another boost. And with Orakei becoming two-stages rather than one to the CBD, fewer people will park in Meadowbank and walk across the boardwalk to Orakei.

        Feedback on the bus route changes was widespread and robust on the need to retain and for that matter improve bus connections to Meadowbank Station. If that happens, and I expect it will (Meadowbank is the only station on the entire network not proposed to be served by a nearby bus service.. an aberration and not aligned with AT policy) then patronage will increase further.

        On the other hand, with the Orakei Point development stalled, Orakei Station remains surrounded by water so I can’t see its patronage growing much.

          1. Sadly, yes. Cancelled even it seems! That development looked excellent from many viewpoints, not least transit oriented development.
            The company website says this: “This World Class seaside Transport Orientated Development was planned by Auckland Council and Equinox .
            As a result of Auckland Council changed requirements the project was terminated in late 2015.”

  10. When will we get an increase in services on the Western Line? And I assume that with the start of works for the CRL and imminent announcement of Government support for this project that extra EMUs are being arranged, which will be necessary to continue with the growth we’re seeing.

  11. Interesting data. Imagine if the top 15 ( or top 10) were gated stations!

    I was at Henderson station last week. The lone security officer was trying to stop a group of 8 teenagers free riding, on that one train alone.

    Yesterday on an Eastern Line train ticket inspectors stopped the train at Panmure and forced off 12 people roughly 16 – early 20’s off the train. Just that one train alone.

    1. That’s such a pity. I assume that the case for not gating any except the very busiest stations was based on previous passenger numbers. Where those boardings have increased by 30-50% in the last two years the case would presumably be now be much stronger.

    2. I think AT plan to progressively gate stations, which is good.

      Worth bearing in mind, however, that the benefits of gating stations quickly become marginal at smaller stations.

      if you look at Grafton, for example, then the costs of gating (if it is even possible) would probably be around $5 million. If we assume that fare evasion through that station is currently 10% and it would reduce to close to 0% with gates, then you’re talking about saving 10% of 250,000 passengers p.a., or 25,000 passengers p.a. If 50% of those who are currently fare evading chose to continue to use the train after gates were installled (remembering that many will not), then you’re talking about spending $5 million in order to generate an additional 13,000 passengers p.a.

      To put that in context, that’s an annualised cost (assuming 20% cost of capital) of $1 million p.a. for 13k passengers p.a. So about $80 per passenger generated. Of course, that passenger will pay some money in fares, say $5, which brings the net cost down. Still fairly expensive compared to other ways to spend that money (e.g. additional rail and bus services).

      So beyond a certain point fare evasion is better managed with on-board inspectors. The benefit of inspectors is that they can not only check that people have a ticket, like gates do, but also that people are carrying the right ticket. There are some types of fare evasion that gates won’t stop, such as concessionary fare evasion (people using other people’s supergold cards) and/or zone over-riding. So in some circumstances on-board inspectors are more effective than gates.

      1. $10 million seems awfully high to simply install some gates. They’re not free, obviously, but what accounts for so much cost?

        1. In most situations, gating requires fundamentally changing the way the station operates. There’s also a lot of ancillary costs that people forget.

          For example, aside from the technology and installation costs, you would also need to install a new ticket machine outside the gates for people to top up. And if you wanted people to be able to use paper tickets then you need to pay someone to check paper tickets at the station, i.e. they need to be manned. You also need to secure the platform – no point installing gates if people can access the platforms easily via another manner. In fact installing gates without securing the stations could be a real safety risk, from people crossing the tracks to avoid paying tickets.

          So if it’s a choice between paying someone to check tickets at a station (and paying for the station to be gated) versus paying someone to check tickets on the train (and avoiding the costs of gating), then often the latter is more cost-effective.

          1. It’s a shame that you need to have someone manually checking tickets. I’d imagine that this could be phased out by having barcodes on printed tickets which the gates could read, but it’s probably too late to put all that in place now.

          2. They just need to take a note of how National Rail in the UK works (and similar Transport for London services which even though there is a big push for the Oyster Card and electronic payments) still use cash fares on machine readable tickets

          3. Even if you got rid of paper tickets, you would still need to have someone manning the gates. In Melbourne we don’t have paper tickets, but all gates have to be manned in case (for whatever reason) people can’t get in or out.
            Also you mention the need to buy more ticket machines, surely you’d just move the ones from the platform to the gate?

      2. Seriously, $10M to gate one station? That sounds ridiculously over-priced, where is that figure from?

        Remember that once you have gated a station, it is forever gated, so the “50,000 passengers p.a.” that you mention will eventually pay it off. Meanwhile, an inspector will always be collecting a salary and not catching everyone who is fare evading, and when they do find someone on board there is the hostile situation of evicting them, not to mention the unnecessary delays to hundreds of other passengers.

        Also, 25% of all passengers aren’t paying their fare?! Wow, we really need to gate the stations!

        1. yep gating stations is expensive. The $5 million figure is derived from the costs to gate and secure (fairly comparable) rail stations in Australia. Costs include design, installation, and maintenance (oh maintenance, so easy to forget!). As I pointed out above, you need ti pay someone to at least check paper tickets at each gated station. You also need to secure the station itself, e.g. by installing fences, so people don’t cross the tracks so as to avoid the gates.

          The general rule is that when you start looking at old rail stations, such as those that exist on Auckland’s network, then the costs of gating increase fairly quickly.

          Maybe AT can do it cheaper, but I don’t have any more information to go on. Do you?

          Why does gating somewhere like Grafton cost so much? Well:
          – Number of entry/exit points: Grafton has 4 entry exit points. At each entry point you’d want at least 2 gates = 8 gates in total. I’d guess-stimate $500k per gate = $4 million just for device readers equipment.
          – Modifications to entry/exit points: The stairs at Grafton station have not been designed to facilitate gating, i.e. they are too small/narrow at the top, so would need to be modified. You may also need a wider gate so people with wheelchairs/prams can access the lift.
          – All weather protection: You’re talking about putting machines with expensive/sophisticated electronics outside in Auckland’s climate. Rain, humidity etc etc. Not cheap.
          – Associated systems/services: You’d need to locate a ticket machine outside one of the gates (so people could top up if their balance was low) and a person checking tickets. Plus CCTV, power supply, etc.

          Main message is that it’s cheap to do 1) at large stations and/or 2) when designing a new station, such as Newmarket, Panmure, and Otahuhu.

          1. That sounds outrageously expensive for something that uses about $200k of metal and another $100k to install (for a large station). $500k for a gate? How many multiple middle men are clipping that ticket??!! Is it gold plated with embedded diamonds in it or something?

          2. feel free to ask Thales or Cubic (or any of the other technology providers) to supply you with a cost-estimate to gate your favourite station and see what they come back with.

            Then get back to us on whether the cost reflects 1) ticket clipping or 2) complexity :).

          3. $500,000 per machine that can’t deal with the rain? The ones the skifields use can’t cost that much and they deal with a lot worse than auckland weather. I’m thinking something like this.

            They don’t have to be impassable barriers, if someone wants to get on for free they will. What they do need to do is stop those people that might try to sneak on but wouldn’t jump a barrier, and if they do jump the barrier then it is clear to the cameras that they are fare evading

        2. Jonty:

          – No, 25% of passengers aren’t evading their fare. I made that number up to make the calculations easy. The true fare evasion % is probably closer to 10%, depending on station/time-of-day etc. I’ve changed the number to be closer to what I expect it to be.
          – Even with many gated stations, you would still need some on-board collectors for the reasons I note, i.e. people travelling without a ticket is only one type of fare evasion. Many people may have the wrong ticket, e.g. travel on a concession card that does not belong to them. Gating does not stop that type of fare evasion.
          – Gating stations requires that you man stations to sell paper tickets. So it’s really a choice between paying people to work on trains or paying people to work on stations. At some places the equation will favour onboard staff, and at other places it will favour station staff.
          – Don’t think that gates are free once they are installed. They need to be maintained, and probably even replaced every 10-15 years. So you’re talking about an expensive asset that depreciates fairly rapidly. Hence you need to account for their capital cost as a annual charge. In the example above I used a figure of 20% p.a.

          Sure if you were designing a new rail line/network from scratch then you might be able to gate every station cost-effectively. But many of Auckland’s stations pre-date electronic ticketing and are therefore not particularly well-suited to gating. Basically, the problem is not as simple as saying “AT should gate every station”. Don’t get me wrong: I support gating stations *in the right places*, but often it’s more expensive and/or more complicated than people think.

          $5 million for somewhere like Grafton is definitely at the high end, and not indicative of most stations. But even the simplest stations could well hit a couple of million $$$.

          1. I remember seeing it quoted that gating stations adds about $400k each in operational costs primarily due to the need for staff to man the gates (to stop people jumping them, help people etc.). Given most stations now seem to have security guards posted at them during the day I wonder how the cost changes of they include that.

            As for whether gates are worth it, consider that the average fare is about $2.7 so that’s a lot of trips you’d need to gain to make it worthwhile.

            As I understand it faster evasion is at about 7% across the network but some stations are worse than others, e.g. Henderson is bad but Grafton is actually ok.

          2. If connecting to a bus at Grafton, a passenger would want to tag off so the bus then lets them tag on again. Most of the destinations within walking distance, such as the new university campus, would be extremely peaky, and could be picked up by roving ticket inspectors.

        1. From what I have heard, Manurewa is the next station to be gated, which I think is a good choice as a lot of South Auckland fare evaders use this station so it will catch a fair number of people getting off at other stations – plus I don’t think it would be to hard to gate.

          1. “South Auckland fare evaders” wow you sure some evaders didn’t come from elsewhere or are we just gonna stick with pinning it on South Auckland?

            Fare evasion is a widespread issue mate. Let’s leave the stereotypes out.

          2. Someone is feeling sensitive I can see.
            It’s a simple fact that a fair number of south Auckland fare evaders enter at Manurewa and for some strange reason the Western & Eastern line fare evaders don’t seem to use Manurewa – strange that.

          3. Yes they obviously use the southern line. But are they purely from South Auckland? Anyone from anywhere in Auckland can catch a train to & from Manurewa. Point, just say fare evaders instead of trying to bring others down in an attempt to make certain other areas & its people sound more honest.

          4. I was not making any claims about any areas being more dis-honest or honest than others.
            Just chill and relax dude.

    3. I took the train to Sylvia Park during the holidays – first time Quaxing and it was great – the train station a real asset to the mall.

      A couple with a kid got on a Glen Innes, approached by the ticket inspector and put on this act:
      “Did you buy a ticket?”
      “No, I thought you did?”

      Ticket inspector let them go too. Probably because the guy was big, rather than believing his marvellous acting skills. GI is a top 10 station and would benefit from gating.

      1. There’s a certain demographic that likes to take the piss in this regard, which you’ll find is a regular feature at GI.

        But really, the free-ridership problem isn’t as catastrophic as it seems. Even if people are catching the train for free, there’s generally zero marginal cost to AT for them to do so (outside peak times). And that’s still another vehicle that’s not on the road.

        My only concern, as per the Broken Windows Theory, is that fare evasion is the first step on the slippery slope of antisocial behaviour and vandalism on trains. It’s why i believe that maintenance budget should be a priority to keep the new electric carriages looking ship-shape.

          1. Agree, was about to make the same point.

            If he is going to make a racist slur he should come straight out with it

          2. Well with these statements it can go two ways.

            If that is just an assumption, then it would indeed be racism. The same for personal observations, which are skewed by confirmation bias.

            But on the other hand, if it is indeed the case, then saying so is not racism. And if people are afraid to say so because they are afraid to be branded as a racist, then we are in for way worse than just fare evasion.

  12. Still need to fix the dwell time issue, which should lead to further patronage increases. Compared with Melbourne, Auckland’s trains seem SLOW. I timed dwell times Saturday night on the Western Line, and they were 45 to 50 seconds. Doors were only open 20 seconds.

    1. Ridiculous. In Melbourne (and, I’m sure, in hundreds of other urban rail systems) a typical offpeak minor station dwell is 20-25 seconds (5 seconds opening and closing**, 15-20 seconds handling passengers).

      I’m genuinely mystified by how a blank slate electrification project managed to end up with such poor outcomes on this very important detail. Did no-one think to look for world’s best practice and order that?

      ** That includes the time between train stop and door start opening, and between door finish closing and train start.

    2. Yes, Melbourne has dwell times down pact and also has passenger operated doors. I don’t buy the argument that passenger operated doors are the reason for Auckland’s being so incredibly slow. Melbourne shows that it can be done.

      I caught the train last Saturday, and the majority of stations had dwell times UNDER 20seconds.

    3. The point you are missing about Melbourne is how much they are delayed when somebody with a wheelchair needs to get on. This adds minutes as they don’t have level boarding so staff have to go and help. Apparently there therefore build slack into their timetables to cope with this, which isn’t required in Auckland.

      That said, DOO would help in Auckland a lot I think.

      1. That is absolute rubbish. Minutes?? There’s disabled people that get on my station all the time. At most, it adds about a minute. A lot of the time, it adds nothing as disabled people in Melbourne know to wait by the drivers door at the end of the platform and driver comes out with board, it’s a very quick process most of time. They’ve started having raised platforms at the end of each platform, so now they can board without the driver even getting out as it’s now level boarding in most cases.
        Even with the slack in, Melbourne’s trains run a lot quicker and more efficiently than Auckland’s, even with Auckland’s newer fleet and newer signalling system.

        1. Quicker and more efficiently until you get to the City Loop approach and sit there for six minutes waiting for a signal to cross four sets of tracks.

      2. The Auckland schedules have significant slack built in to cope with network limitations, namely junction management. These days many of the dwell time gripes that people want to insist are the Train Manager’s fault, are actually just the requirement to not leave before the scheduled time. It’s not unusual to see a train stopped at Penrose or Otahuhu for two or three minutes as it has arrived there that far ahead of schedule. By the same token, though, it’s not unusual for that slack to be soaked up on the other side of the junction waiting for another train to cross, a signal, a platform, speed restriction, whatever. Only the CRL can fix that.

  13. I see that Penrose appears to indicate that its data is just for the Southern Line (Platforms 1 & 2 only ??), or does this also include the Penrose 3 platform which services the Onehunga Line?

  14. There seems to be some confusion over the name of the Manukau station. In this article it is called both Manukau and Manukau City. It has only ever been called Manukau as Manukau City ceased to exist when we all became part of Auckland Council.

  15. Sadly, I think Len will be remembered for his affair and not the great work that he put into making Auckland an international city.

Leave a Reply