Auckland’s public transport patronage results for August are now available and there are some decent numbers on show. This was partially expected thanks to there being two extra business days in August this year compared to August last year but even accounting for that, numbers are up. August is traditionally a strong month for patronage with its 31-days and no school or public holidays, and the month didn’t disappoint clocking in with the third highest patronage behind March 2015 and 2016. The month was significant as halfway through we finally had integrated fares roll out, something that Auckland has needed for decades. Changes like we had normally don’t have an immediate impact though and so it will be some time for us to see the full extent of the new structure and for many, cheaper fares.


Overall patronage was up 8.7% for the month (normalised to 3.9% when taking account of the extra weekdays) and 7.9 million trips were taking on PT. Drilling down to the PT modes:

  • Trains once again led the charge up 18.4% (normalised to 14.5%) and on a 12m rolling basis, we surpassed 17 million trips for the first time. Looking at the rail numbers we’re still seeing fantastic results but the percentage increases are slowly starting to reduce, guess we can’t grow at 20%+ per annum for ever. The next boost is likely to come from the roll out of the new network.
  • Buses have been struggling lately despite some key routes such as the busway growing impressively. This month we’re still seeing that overall trend with this month the busway looking even more impressive after posting a 34.6% increase in August. On a 12m rolling basis, Busway usage could soon exceed usage on the Eastern Line. In fact, patronage growth has been so strong that AT say Ritchies will increase the number of double deckers on Northern Express services in October from 16 to 29 and there will only be two non-double decker buses used (all off peak services will be double deckers too). Other routes that have had double decker love are also said to be posting some good growth. But with stagnant patronage on buses overall, it means those routes seeing crazy growth are offsetting declines elsewhere and the two areas experiencing this the most are the south and the west. More on this later in the post.
  • Ferries have continued to show relatively good and consistent growth over the last 18 months or so.


As part of some travel planning, AT conducted a survey of employees in a number of large office buildings in the CBD on how they travelled to work. From over 10k responses an impressive 51% said public transport.


In some analysis of bus patronage performance, AT have broken the results down by area and eventually route. As you can see from the last image, many of the routes in the south have been on a bit of downward trajectory. Hopefully the New Network launching at the end of next month will help address this.


Looking at some other results, farebox recovery was expected to take a bit of a hit, and it has, but not by too much. We really need to wait to see a few months with integrated fares to see just what impact it has but a promising start at least. Related to integrated fares, AT say 84% of all PT trips were taken by using a HOP card.


Share this


  1. Great results but pretty useless in isolation against the increasing private transport trips. The data shows more people using PT but is it showing a reduction on private vehicles? Are we ahead of the curve?

    1. Dale yes. And no.

      Yes; it is absurd that we are spending billions and billions incentivising people to drive into the city, especially from distance, simply of course to have them stuck in traffic. Which is to say we are now investing in Transit, but also are investing even more furiously in traffic building.

      No because while gross vehicle kilometres driven is increasing, per capita they are not. There are more cars in AKL, but they are also parked more [and not just stationary on congested roads]. More cars used less is sadly crazy from an economic efficiency point of view. But, despite the hopes of tech evangelists and economists everywhere; we are still showing very little inclination to share the use of our cars with each other….

  2. We don’t have integrated fares – if I catch a bus then ferry then train this morning I will be charged for three separate trips. It is expensive and ridiculous, and definitely not what I would call integrated. Really surprised to see you imply that it has been delivered in a remotely acceptable way when AT are basically stealing my rates.

    1. I appreciate your enthusiasm, but can’t help feel that you’re being a bit emotional about integrated fares. There’s simply too many inequities all over the shop about how rates are collected and used for this to qualify as “stealing”. The reality is that all ferry users have known in advance what the fares would be for their service, and any reduction from this point on should be treated as a bonus – not an entitlement.

      My understanding is 1) AT have recently delivered integrated fares for bus and rail journeys and 2) are planning to integrate ferry fares in the near future, once budgets and operators allow. Budgets are important because, well, they rely on funding from Auckland Council (rates) and NZTA. The latter impose farebox recovery targets, which AT must also work towards. Operators are also relevant, because under integrated fares, some ferry routes will experience a significant reduction in fares. If these routes are short on capacity, as some indeed are, then it doesn’t seem to make much sense to reduce fares before additional capacity is ready to go.

      Unless you want to spend rate-payers money subsidising services to the point where people are super-squashed and/or left behind?!?

      1. I’d point out that fare integration for ferries is not the same as fare parity. Integration so that you don’t have separate charges for multiple services is planned but doesn’t mean it will cost the same as the same journey by bus. Likely to be that you pay the same zonal fare that now exists with a small ferry premium which is based on which route you use. Reason for this is AT expect demand for ferries would be too high if there was fare parity and cost of providing the extra service needed would be huge

          1. Ferries don’t scale all that well. Tend to be very peaky usage, more so than other modes, and currently fleets are able to be used off peak for tourist services so costs shared. Adding services means the full cost of running services (and they have high opex) comes back to AT.

        1. I’m not too sure how much of an issue this is. Ferries aren’t exactly quick and there can’t be too many journeys where it makes sense to PT to a ferry terminal, cross the harbour and then use more PT on the other side. Most places on the Shore you might as well bus to Britomart directly.

          However a lot of ferry users park at the terminal to avoid queuing for the motorway – but should this be rewarded by cheaper fares? Why go car-ferry-bus/train instead of bus-bus/train?

          I’d argue there is a case for integration for cyclists though, using the mixed modes of cycle-ferry-train-cycle to extend their range.

          And on the point of mixed modes, I wouldn’t know how to answer that survey. I cycle to work apart from the wet bit in the middle, where I take the ferry (by distance about 2/3 bike to 1/3 boat) unless I take the bus in which case it’s something like 3/4 bus, 1/4 walk. Isn’t that the whole point of a successful PT and active mode system, mixed modes increase the convenience.

          1. You cycle most of the trip that you do most of the time. So answer that you cycle and ignore ferry and bus. Its the same for drivers (to a much lesser degree) . They usually answer that they drive and ignore the walk from the carpark to the office.

    2. I understand ferry is the odd one out…? Let’s hope that gets fixed soon. For those of us using bus and train, integrated ticketing has been a huge step forward. I’m getting about $34 per week “free” travel compared to the old system, plus the time saving of those additional local transfers. I feel like that’s great value for my rates!

  3. Assuming it is accurate, that CBD survey is great because it has such a large sample of workers. 2/3rds of people don’t arrive by car. Imagine if you included the thousands of students as well. At the very least we need half the available road space dedicated to buses/cyclists only, immediately.

      1. 10k is a pretty big catch. But Ari is absolutely right to point out that the private car figure would be even lower than 34% if the survey covered journey to education as well as work. Given that education makes up a huge part of demand that AT needs to meet it seems illogical to ignore it. Of course MoT and the Ministers, and every anti Transit speaker all love the journey to Work figure because it presents a much lower Transit and Active demand picture. In Australia the census question is; ‘journey to work or education’.

    1. Would be interesting to see the split of drive alone between motorbike and car. I presume taxi is lumped in here.

      Its the 5% that responded “other” that intrigues me, how did they travel?

      1. I do see a few people skateboarding. Funny that car-pooling only makes up 3% from that survey. Always here politicians saying people should car-pool we must car-pool … Doesn’t seem to work!

        1. It’s a great idea in theory, all those empty seats!, but is very very hard to get to happen in practice. It seems few people like sharing the intimate private space of the car with each other. Tech evangelists and those uber-rational economists believe we are always on the verge of overcoming the oh so very human issues around this but I’m not so sure. For example here’s a view on Uber pool in the US:

          Furthermore there are geometric issues which create structural issues: dispersal of employment and habitation hinders ride share more than Transit use for a key sharing demographic: The occupants of the same dwelling. In a more dispersed city there is less chance these people are heading the same way each day; hence households with 3, 4, 5 cars, especially in more auto dependent areas. It is harder to get the secondary riders, eg kids, to rideshare with a stranger when parents suddenly are going in a different direction….

          1. Issues with car pooling are far more practical than you let on.

            The reality of modern life is that a fair number of people don’t simply go to work at the same time every day, then go home after work at the same time every day. Using myself as an example I’m doing other activities straight after work between 2-4 nights a week. My schedule would be totally incompatible with anyone wanting to carpool with me and vice versa.

            To put it different why don’t we see people riding tandem bikes? It would be more efficient.

          2. TRM – it’s a bit of both I think. Within a household or group of friends you are correct, people often have different schedules. However, when it comes to carpooling with random strangers I think Patrick is correct people just don’t like sharing a car with someone they don’t know. The only place I’ve seen this work successfully (there may be others) is San Francisco where a carpool lane on the Bay Bridge is free, and generally more free flowing. This gives an incentive for people to call into the designated carpool pick-up spots in Oakland and Berkeley to pick up a stranger.

        2. Many of those that don’t mind carpooling are doing it already. It’s called catching the bus. So i agree the numbers to gain from technology making private carsharing more feasible could be quite small.

          1. No there is a big difference in being with strangers in a public place and being with strangers in an intimate one. Transit is a vary different place to someone else’s car. Each have their issue, but it is important to understand this difference to understand why one scales well and the other doesn’t.

      2. From my experience on the bus (great south road from Papakura) the Takanini area is terribly slow and (i think) would be perfectly suited to peak time bus lanes, the road is already wide enough.

        The takanini motorway area is a bit of a bottleneck but that would be a lot trickier to deal with.

        Heading south in the PM peak could also benefit from bus lanes (past the manukau hospital) down to the manurewa town centre. And then again once the buses get through the Takanini motorway area, bus lanes along ‘Takanini straight’ most of the way to Papakura would be fairly easy to introduce.

        1. Yeap, much of which will be addressed by the motorway works completing (changes the diamond shape, shifts load, and should reduce people getting of the motorway because of the queues), or the new service arrangements for South Auckland when they start.

        2. Takanini got stuffed up by the unused bike lanes. The smartest thing would have been two marked lanes each way. Even on the weekends its appalling. The road is certainly wide enough but the fools in charge chose to destroy the efficiency of a major thoroughfare, ultimately benefitting no one.

          1. I’d disagree, the on street parking provided would be much more utilised with bus lanes each way during ‘peak times’. Parking can then be used during the middle of the day. The bike lanes aren’t that well utilised right now as they are so intermittent and dis-jointed. I’ve biked Papakura to Manukau a few times and it can be scary when the bike lane ‘just ends’.

      1. From my experience on the bus (great south road from Papakura) the Takanini area is terribly slow and (i think) would be perfectly suited to peak time bus lanes, the road is already wide enough.

        The takanini motorway area is a bit of a bottleneck but that would be a lot trickier to deal with.

        Heading south in the PM peak could also benefit from bus lanes (past the manukau hospital) down to the manurewa town centre. And then again once the buses get through the Takanini motorway area, bus lanes along ‘Takanini straight’ most of the way to Papakura would be fairly easy to introduce.

        1. My experience is that all of the buses that travel from Britomart to Manukau are never on time by the time they are travelling through Mangere or Otara. I am hoping that the new system which starts soon will improve this plus the higher frequencies will lift patronage.

  4. About the farebox ratio, anyone know what SOI stands for? And why it has fallen since May? Also the subsidies for ferry services (I asume this is just the subsidised ones) increased suddenly from around 10c per kilometre to 14c. Was there some big drop in ferry fares? or patronage? or something else? I do think the long term aim should be to keep increasing the farebox recovery ratio (ideally through increased patronage) so as to gain extra funds for capital improvements to the system

    1. SOI = statement of intent – the document that sets out ATs goals and targets for the year.

      Farebox fell on rail due to increasing western line services. On ferries I think it might have been from either new contracts or changes making the costs but transparent. Bus probably from things like more services extra buses and double deckers on some routes

    2. SOI = statement of intent.
      The arbitrary Farebox Recovery required by NZTA is a crude attempt by National to restrict funding and growth whilst not doing any sort of subsidy reduction for roading.

  5. Boardings is a useful measure, but it treats a one-stop trip the same as a long-distance commute from eg Pukekohe or Waiheke when quantitatively they’re quite different. Passenger km is a better measure for the total volume of passenger travel, and would present a very different picture of the significance of the modes and the contribution being made by PT. The information isn’t quite as simple to collect as boarding, but not that hard with such a high proportion of Hop use. Does AT collate/publish this information?

    1. Yes they have that info and very easy to get with HOP measuring trips. I have seen it annually but not more frequent than that. NZTA has the info for all regions in their website

  6. I wish we could get more detail on the bus patronage data. There are so many bus routes so a single number doesn’t tell us much. We know there is decline on buses, but is there growth anywhere on the bus network (other than the NEX of course)? Would love to see how the new network in Hibiscus Coast and Green Bay/Titirangi is going in terms of patronage. Is it working?

  7. 1. I fully agree with Mike (TLO) comment above regarding the desirability of having passenger km data, certainly at the modal level and desirably in more detail. AT clearly has this data, as it is used in their chart 2.4.3 (Subsidy per passenger kilometre) above. It would be good if AT could add passenger km statistics to their monthly patronage report.
    2. I wonder if part of the reason that bus patronage has been struggling is because some previous bus passengers have been ‘cannibalised’ by the rail service improvements and are now rail passengers? It should be possible to get a good handle on this by examining monthly patronage trends on each of the rail lines and on the parallel (competing) bus routes. I have never seen such analyses (has anyone else?): I would think these would be useful in shedding more light on the overall impacts of the rail service improvements over recent years on the public transport system as a whole?

    1. AT could supply average passengers per train/bus service but it would most likely be too scary for the ratepayers to see how badly their PT subsidies are being spent. Peak travel is the only time the PT services (and not even all) run anywhere near the levels we are told they are capable of.

      1. That’s probably why it’s called the peak…

        So if subsidies are currently being spent “badly”, how would you suggest that they were spent “well”?

        1. We are constantly told when comparing PT to cars that the average people per car is 1.2 but never the average person per PT vehicle just their capacity.

          If we are told that a 6EMU makes 16 trips per day with 468 seated capacity twice it carries 600 passengers and 12 times it carries less than 20, is that well spent money?

          1. Yes, those 20 people would either be driving both ways or be unable to make their trip and running a service is cheap once the track and rolling stock are there.

          2. Ted you have odd obsessions. Or rather a poor understanding of capacity utilisation. Motorways are empty at 4am [which by the way is all that the TomTom metric tells us] but full at the peaks. So have full capacity utilisation at the peaks? Er no, because the passenger vehicles on them consistently have that low 1.2 occupancy. The rail system too is at capacity at the peaks, but the difference is that vehicle utilisation is also at maximum then; full trains. Additionally this is also when every single ride on the rail system is of its highest value, especially to road users, cos otherwise these trips would be on the road, either in a bus or car, making that system even worse.

            This tells us that we should aim to do [at least] three things in order to increase efficiency in these systems [value for money]. 1. Increase car occupancy at peaks, 2. increase train numbers at peaks, 3. shift both riders and drivers to the off peak. The last one is tricky, but we don’t try very hard. We have long pushed for off peak Transit discounts and time variable road pricing, hopefully we are heading there. Rideshare that works would be good too [not often seen in the wild]. But otherwise the best thing we can do is 2. increase peak time rail capacity; bingo: CRL.

            Expecting some kind of Goldilocksian constant capacity utilisation is a waste of energy. Only systems under enormous demand pressure, like the London Underground exhibit that, hence Crossrail and the Overground investments. Demand is constantly high there, and overflowing to all other modes, including cycling, and despite high pricing, because of supply limits and a thriving city. Really a much bigger problem than some empty-ish trains on the Onehunga Branch Line.

          3. Or to apply Ted’s logic to roads, because motorways only ever look busy at peak, and even then they’re not full because of all those empty car seats, we should stop investing in motorways and reduce them all to one-lane highways.

          4. For some reason I posted too quickly and needed to slow down so I will try again.

            Sailor Boy I agree in part but maintenance is also part of those costs so wasteful trips (or running 6EMUs 16 times in a day when it is only required twice, 3EMU are more than sufficient for the other 14) increase costs.

            Mike its not that there is necessarily a better way to spend, it would be good to know the average passengers per unit before we increase the units off peak as many are calling for.

            Patrick yes motorways are empty (well almost) at 4am but so are all the trains and buses (with the exception of possibly an airbus). I’m not saying we need to build motorways 10 lanes wide to cope with peak but we also need to spend money wisely rather than running more frequent off peak services just because someone thinks it is a good idea, without giving us an average people per current services.

            Andy or to turn your comment around we could only supply off peak capacity during peak as that’s all that should be required. The comment was nothing to do with roads or motorways it was in addition to Ian mentioning getting “passenger km statistics” to have passengers per service.

            Onehunga has an average of 10-12 passengers per service, up to around 50 in peak but we still run 238 services there every half hour 7 days per week (55692 seats).

          5. “it would be good to know the average passengers per unit before we increase the units off peak as many are calling for.”

            No, it wouldn’t as that isn’t the point of increased daytime frequency.

            We are not increasing the day time services because we need more capacity. We are increasing the day time services because people are much more likely to catch a train if they know that they won’t have to wait. Increasing the frequency also means that transferring between services becomes more reliable meaning that we can remove the bus routes that run parallel to the rail lines and save money that way(especially as those buses are all very empty, but adding 10 buses with 5 people each is 1/8 of a full 3 car train). Making public transport more useful for off peak journeys allows more people to make the choice to travel by PT off peak which may mean that the other part of their journey (the return) is at a peak time when we need the road capacity.

            Increased off peak services isn’t about accommodating existing capacity demand, it is about increasing capacity demand.

            Yes, we absolutely should investigate using 3 car sets on all off peak routes.

          6. Right, Bigted, so you’ve got no idea what you mean by the subsidy being spent badly, nor any idea of how, if it is being spent badly, to spend it well. You want AT to produce an average figure for current services, but you’ve no idea what use that information could be put to. (If you do have any idea about any of the above, you’re not telling!)

            Thanks for your contribution.

          7. Sailor Boy and Mike the point is IF AT produced such a stat maybe they would then look themselves at the how wasteful running 6 car units when 3 cars are sufficient.

            It is all very well pat themselves on the back because the patronage has increased over 80 million trips but they don’t tell us how many KMs each trip is (currently Swanson, Manukau or Pukekohe to Britomart or even Swanson to Pukekohe is one trip but so is Takanini to Te Mahia) and how many services are run to. It is all very well saying we moved 1000 people from here to here but if you then if they say we ran 700 services to do it, is that money well spent?

          8. What is wasteful about running six car sets? They still need the same number of operators and i doubt driving to the depot and back to park up half the train would save any significant runnings costs compared to the disruption it would cause. It would probably cost more as youd need a second driver if you wanted to keep the other half running in service.

          9. Dan
            “What is wasteful about running six car sets?”
            -Extra ks means extra maintenance and of course extra costs.

            “They still need the same number of operators and i doubt driving to the depot and back to park up half the train would save any significant runnings costs compared to the disruption it would cause.”
            -Splitting is by the push of a button, the crew that bought it in then could put it away and the crew that was taking it out take half the unit, where is the extra cost and disruption?

            “It would probably cost more as youd need a second driver if you wanted to keep the other half running in service.”
            -It would be run instead of not in addition to the next service. This would all be happening at the end of peak when there is still sufficient crews and there is already units being put away, all that changes is more 6 car units get put away.

          10. I would hope and expect that Transdev/AT are doing such capacity monitoring as a matter of course, based on the normal maximum load factor of each diagram – matching supply to demand should be standard operational practice, provided the costs/risks of such a reduction don’t exceed the benefits. (The average load factor for any particular trip is meaningless, since trains will tend to be much fuller at Britomart and emptier at the outer terminus, but the Britomart people still have to be carried.)

          11. ‘Splitting is by the push of a button, the crew that bought it in then could put it away and the crew that was taking it out take half the unit, where is the extra cost and disruption?’

            Bigted – do the crews change at the end of a peak period? I would have thought the crews running at say 9am would still have a number of hours still to run before they finished, but I’m no expert.

          12. jezza the crews don’t run continuously all their shift, they have breaks and don’t in most cases (Onehunga and some Swanson service that don’t change crew at Henderson) run straight back out in the same unit they brought in. Off peak they run empties into depots and have meal breaks etc.

          13. Mike there are manual counts done at two points plus the Onehunga and Manukau branches but AT are only interested in the highest count per trip (which makes sense to a point) and the Onehunga/Manukau count. These counts only gives information (and only the higher of the two) at their two prescribed points, the train could have been standing room only or empty only two stops earlier and not get recorded by this method. Manual counts count everyone not just those that pay like all ATs other methods of counting, this information just doesn’t seem to go anywhere.

          14. Bigted – if a 6-car train arrives at Papakura at say 9am and it and it’s crew are going to continue to be in service for the off-peak period, who are the crew that are going to stable or run the not needed 3 cars back to the depot? I imagine if a crew were on an allotted meal break they wouldn’t want to be driving trains.

          15. Jezza a six car does arrive at Papakura at 9am and it then uncouples and half goes out at 9:18 and the other half at 9:36. If the crew was required to put it away their meal break would be allotted to them AFTER they had done it.

          16. You’ve made it clear that you think there are 6-car trains running off peak unnecessarily. I’m still struggling to see where the extra staff are to run the surplus units back to the stabling facilities at Wiri as they can’t all just bank up on the platforms or stabling spots at Papakura and Swanson.

          17. There are no stabling slots at Swanson, they go to Henderson and there are not extra staff required to run units to Wiri as the crews are already in them when they arrive (they are not self driving yet).

          18. Right so no stabling at Swanson and trains don’t drive themselves (I was already aware of that one!), I think you have just answered your own question on why there is ‘wasteful’ running of 6-car units off peak.

          19. I would imagine that would be because they effectively take every second service out for off peak. The problem is the 6-car trains don’t alternate with 3-car trains at peak, they tend to come in blocks when it is busiest.

      2. Bigted – I can’t see where what Ian W has said would require average passengers per bus/train. All it requires is the trends in total volumes for the relevant buses and trains in the area.

      3. A few obvious errors. I have fixed them below:

        “AT could supply average cars per hour per street but it would most likely be too scary for the ratepayers to see how badly their roading subsidies are being spent. Peak travel is the only time the roads (and not even all) run anywhere near the levels we are told they are capable of.”

        No need to thank me.

          1. Funny, I just drove the full length of SH 41 twice and that seems to need a shitload of maintenance even though it has less than 1500 vehicles a day.

          2. That’s embarrassing for you. All state highways in new zealand are maintained continuously under network operating contracts.

    2. Ian as poor bus performance is south and west means it is likley people are switching to train now for more predictable and comfortable journeys. However, the massive spending on motorways is also likely having an effect, tempting bus users into driving. People are, on average, rational, and will switch between modes as they shift in effectiveness and appeal.

      So, if this is largely correct, we can expect continued movement between modes: the Nash equilibrium. As the roads re-congest after widening, which they will, unless the economy crashes, people will be incentivised back to alternatives, depending on their quality. So if AT sort out bus capacity and especially priority we are likely to see them grow again.

      The New Network is key here. Change usually brings a hit, but improvement will win out in the end with ridership, especially in a growing city. But bus lanes, eh.

  8. Transdev do what AT tell them to do.

    They don’t decide what size trains run.

    That is why AT should run the rail service and save the management fee paid to Transdev

Leave a Reply