The patronage results for July are now available and they show another strong month of growth.

Auckland public transport patronage totalled 72,740,387 passengers for the 12 months to Jul-2014, an increase of +0.5% on the 12 months to Jun-2014 and +5.9% on the 12 months to Jul-2013. July monthly patronage was 6,268,752, an increase of 343,651 boardings or +5.8% on Jul-2013, normalised to ~ +5.4% accounting for additional special event patronage only, same number of business and weekend days in Jul-2014 compared to Jul-2013.

Rail patronage totalled 11,552,643 passengers for the 12 months to Jul-2014, an increase of +1.0% on the 12 months to Jun-2014 and +14.4% on the 12 months to Jul-2013. Patronage for Jul-2014 was 1,089,839, an increase of 117,561 boardings or +12.1% on Jul-2013, normalised to ~ +9.9%.

The Northern Express bus service carried 2,460,177 passenger trips for the 12 months to Jul-2014), an increase of +1.4% on the 12 months to Jun-2014 and +7.6% on the 12 months to Jul-2013. Northern Express bus service patronage for Jul-2014 was 233,814, an increase of 33,433 boardings or +16.7% on Jul-2013, normalised to ~ +15.2%.

Other bus services carried 53,653,594 passenger trips for the 12 months to Jul-2014, an increase of +0.4% on the 12 months to Jun-2014 and +4.6% on the 12 months to Jul-2013. Other bus services patronage for Jul-2014 was 4,578,804, an increase of 228,637 boardings or +5.3% on Jul-2013, normalised to ~ +5.2%.

Ferry services carried 5,073,973 passenger trips for the 12 months to Jul-2014, a decrease of -0.7% on the 12 months to Jun-2014 and an increase +1.6% on the 12 months to Jul-2013. Ferry services patronage for Jul-2014 was 366,295, a decrease of -35,980 boardings or -8.9% on Jul-2013, normalised to ~ -8.9% (no special events).

14 - July AK Patronage table

14 - July AK Annual Patronage

Again it’s the rail network showing the most growth up 13% with the 12 month rolling total up 14.4%. One of the interesting aspects about this result is the Western Line managed a 14.2% increase despite there being no additional services other than half hourly services on Weekends in October last year. The Onehunga line continues to show strong growth since it was converted to using electric trains – although part of the month saw the old diesels return as Auckland Transport and others try to address some ongoing power supply issues. The most impressive result was on Manukau services in the first full month that the new MIT campus was completed. Patronage on those services was up over 26% and it will be interesting to see if that level of growth continues.

14 - July AK Rail Patronage

To highlight the growth that’s occurring this graph shows the average patronage on each weekday which since July last year has risen from just under 37,500 to about 42,000 per day. AT say that on average seven services per day are over the planned capacity ratio of four people standing for every 10 sitting while a further six were very near to that level

14 - July weekday average

While the rail network is increasing the fastest the biggest growth by overall number continues to be the bus network which was up 5.8% when you combine the Northern Express with other bus services. The graph below shows the NEX patronage while the one after shows the average weekday patronage for the other bus services

14 - NEX Patronage

14 - July other bus weekday average

As AT said last month, from now they have finally dropped the self-reported bus reliability and punctuality stats and have instead moved to reporting them using the on-board GPS tracking equipment. A separate report on the stats highlights the reasons why the old self-reported stats were so high.

Under existing contracts, bus operators provided AT with a monthly service delivery report. Two main variants of contract exist: ‘North Auckland Spine’ (~5% of services), and the remainder (~70% of services). Commercial services (~25% services) are exempt from performance reporting. The majority of contracts reported contracted performance rather than actual customer experience, i.e. excluded trips where performance was impacted by factors outside of operator control e.g. adverse weather, exceptional passenger loadings and significant traffic congestion, resulting in artificially high performance reports. Various metrics were used including reliability at within 30 minutes of start of trip.

Reliability and punctuality has been predominantly monitored through manual self-reporting systems. AT has been working with operators to transition to an automated system based on actual performance data generated from new GPS-tracking equipment. Reporting reliability and punctuality using GPS-tracked performance data will commence from 1 July.

New and consistent, PTOM KPIs will be reported – reliability (trips started within 10 minutes of schedule and completed) and punctuality (trips started within 5 minutes of schedule). In future punctuality at points through the trip and at the final destination will also be measured.

This new methodology reports on customer experience with no exclusions or exemptions such as congestion or adverse weather. An expected punctuality is 100% at start of first each duty timetable trip (operator reaching the trip start) and for all other trips, allowing for an element of average statistical non-performance from outlying high congestion, poor weather, accidents, etc, and compounded where successive trips are linked, 95% at trip start for non-right-of-way (mixed with traffic) and 98% for right-of-way (busway) services.

As a result of no exceptions, the GPS-tracked reliability and punctuality will be lower than previously reported, however performance data collected will permit improvements in service delivery through an ongoing iterative programme of six to twelve monthly timetable reviews.

It’s almost unbelievable that operators were allowed to ignore commercial services, services with lots of passengers, services caught in congestion or results when the weather was bad and it’s no wonder they always managed 99% of services on time. Like the self-reported stats this only reports buses based on when they start their run however AT say they are also looking at performance based on certain timing points too. Below is the punctuality results for July compared to December last year which I assume was when AT finalised their tracking methodology.

14 - Bus Punctuality

So Airbus, the service you probably most want to be on time has the worst performance. This graph shows how the performance has changed since December.

14 - Bus Punctuality graph

The PTOM target is the target that will apply once the new contracts are rolled out as part of the new network. They say that performance above or below the target will be subject to financial bonus or deductions so based on the info above I bet all the operators are happy those contracts aren’t in force yet. The same also applies to bus reliability (whether the bus even starts it’s run) for which there will be financial deductions for results less than 98% with again no operators yet meeting that level.

14 - Bus Reliability graph

It’s great to finally have some proper visibility around this and something we’ve only been calling for for about 3 years.

The biggest downside to Julys results was with cycling numbers which were down 21% on last July for some reason (despite what the text in the report says). If anyone has any reasons they might have fallen so much please let us know in the comments but a quick check shows that July was drier than normal with rainfall in Auckland at only 50-79% of an average July.

14 - July cycling

Share this


  1. A 21% drop in cycling number for July is very surprising. I haven’t seen any less cyclists on my daily commute if anything an increase in cyclists, despite the colder weather!

    One explanation could be cycle counters are not capturing the majority of cyclists. For example northwestern cycle way has been moved in some places to facilitate the causeway upgrade.

    It would be interesting to see data from Auckland Transport that shows location of counters and count by location.

    1. Damian might be correct. From AT’s website ” Automatic monitoring data from nine sites: Upper Harbour Drive, Great South Road, Highbrook, Lake Road, North-Western cycleway Kingsland and Te Atatu, Orewa Cycleway, Tamaki Drive (eastbound) and Twin Streams path.”
      I would be surprised if the counter at Te Atatu is working with all the current works going on there. I couldn’t find any data about the individual counts for each site. Hopefully some one with AT contacts can provide this info.

    2. Auckland has typically about 1.5 million cycle trips a month (from household travel survey data). So the counters are only capturing about 5% of all cycling.

    3. Hi Damian – I have access to the background data for CAA (I wish AT would finally give public data access!!!) , and have just done an analysis of July 2013 vs July 2014. Some very weird things happening – Orewa Cycleway is down *extremely* (does anyone know whether that cycleway has been closed during recent months?). Te Atatu counter is also down strongly, but still at a level where construction and weather might allow for it.

      Overall, the decline is across all counters but with funny stuff happening like the Orewa Path, I have asked AT for comment before I blog further on it.

      Max Robitzsch
      Cycle Action Auckland

      1. May also be that all of the rain that did fall fell on weekends for the Orewa one? Or that there was an event there last year?

  2. I cycle every day and my memory of July was that it was a grim month to cycle. We had some big storms come through which meant we had a week of strong winds, even if we didn’t have a lot of rain overall. Also foggy mornings.

    1. Agree I have cycled to work every day for about 4 years- I think last month I had the most ” you didn’t cycle today did you” questions from colleagues ever!

    2. Agreed, without looking at the weather charts, I recall about a two week period where the weather was wet and windy, and not at all pleasant to be out on a bike.

    1. I’d be willing to bet money that the city link has more boardings than any specific route except the NEX. Combing all the Dom Rd, and Mt Eden Road routes may be higher but actual single routes will be lower than City Link.

  3. Not surprised that howick and eastern has such poor punctuality and reliability compared to the other services. H&E are probably the worst operators in Auckland.

  4. I still find the bus stats pretty meaningless, considering pretty well all buses I catch from Albert Street that have started their run from Britomart are late, I fail to fathom how that could only supposedly be occurring ~3% of the time for NZ Bus. Better stats are needed and it’s great AT are moving in that direction, but the blame for delays due to congestion does still fall on AT as they’re the only ones that can install bus lanes. Ultimately, I hope they use this data to target upgrades for buses where the most delays are seen and this should be very easy to do with global HOP GPS data across all routes. However, the cynic in me thinks AT will simply claim any changes need to wait until the new network and then once that’s rolled out that changes need to wait until they’ve collected enough data to know where changes are needed, and ultimately nothing will happen to due the reduction in LOS for cars any such change would result in.

  5. Are we officially past Wellington on annual boardings now?

    Also, the weekday boardings for buses have gone up 3% since the HOP rollout was completed just 4 months ago. that’s a big jump.

      1. Do you think it’s safe to say that the gap in annual patronage still exists because of the annual shutdowns Auckland has? Using Metlink’s figures ( scroll down to the bottom for the excel worksheet) as well as AT’s, I see that Auckland’s monthly patronage has been higher than Wellington’s every month this year except for January when Auckland’s is lower by over 200,000.

        1. Not quite, as of end of July they were 400,000 ahead themselves and they also changed the counting mechanism which reduced rail boardings as well, so while our summer shutdown are part of the explanation its not the whole story.

          Regardless I expect Auckland will be well ahead of Wellington by December this year on a like for like basis, the big question is how much of a summer shutdown are we going to have this year.
          It seems there will be one, but it may only be 2 weeks or so, and so should mean any lead we get we should keep into January.

          1. Wellington 11.78m trips, Auckland 11.55m trips. Difference about 230k not 400k. Also of note that based on individual months Auckland has been ahead every month since Feb this year. Prior to that the only times Auckland had higher patronage in an individual month was during RWC, April 2012 and June 2013 (when part of the Wellington network was shutdown due to a storm washing out the tracks along the edge of the harbour

          2. Matt 400K was not the difference between Wellington and Auckland annual totals as that is about 200K currently.

            It is how much higher Wellingtons “last” 12 months figures (ending July 2014) were than the previous 12 months (ending July 2013).

            Given that their rail growth has been fairly flat/falling for some time now, that is actually quite a leap of patronage for them (thats why I said they are 400K “ahead of themselves” i.e. 400K higher than if they had the same pattern of flat/falling numbers as the recent years have shown), and that spurt in the last 12 months is the underlying reason why we haven’t passed them yet when we should have.

            Our Xmas shutdowns have not helped in that regard, but are not the full story here. They have actually have an uptick too, but we seem to have got more momentum and so should pull ahead soon.

          3. I actually think the uptick has been hidden slightly by the events of June last year which took out trains to the Hutt Valley for about a week. Had it not been for that the growth would have shown earlier. In some ways it’s similar to the impact that the rollout of HOP had on rail patronage in Auckland as parts of November and December 2012 were hammered as people used up their old paper tickets which had been counted in the patronage results of earlier months and that impacted on the results for a year which is a part of the large increases we’re seeing now.

          4. Matt – you may well be right for about the Hutt Valley lines outage impacting 50% of the value of 400K uptick (i.e. it should be 600K not 400K higher).

            A simple back of envelope calculation goes as follows:

            June last year Wellington total patronage = 150K, mode share for rail is at best 50% of that, and lets say the Hutt has 50% of the rail mode share (which looks right as Porirua+J’ville lines is about the same patronage as Hutt+Wairarapa lines)

            On that basis, the daily “Hutt Rail outage” impact would be ((150,000 * 0.5) * 0.5) = 37,500 a day, times 5 working days of the outage = 187,500 “lost” boardings in their figures.

            Now this assumes all trips were total losses i.e. the trips were not made or not made using buses being counted as Rail trips, so it is an extreme, estimate, even so, that is only 50% their present uptick value could have been “lost” by this outage.

            If you compare June last year with June the year before, then there is a 70K difference (lower than expecteed in June 2013), so that suggests the above impact is half to a third of that number in reality.

            What is interesting though is that the June Rail patronage for Wellington for both June 2013 and 2012 were quite a lot lower than Patronage in June 2011. So somethign has been going on with Wellington Rail in June the last 2 years, and maybe the weather is to blame for some of it.

            But yes allowing for all that does show how much Rail is roaring ahead in Wellington too – looks like new trains is the best way to encourage uptake there too.

  6. According to this page: – there is a link to an Excel spreadsheet of patronage figures at the bottom on that page.

    The total annual rail “boardings” for Wellington were: 11.75 million as at July 2014 and with the July AT Patronage figures, we’re about 11.55 million “boardings” on the last 12 months basis.

    So, no, we have not yet passed Wellington. and we’re about 200K behind on a 12 month rolling total basis.
    Wellington is showing a 400K growth since the 12 months that ended in July 2013.

    Which is less than the growth that we are experiencing, so in the next few months I am sure we will exceed their annual figure.

    It is also interesting to note that Wellingtons rail figures are way less peaky than ours, our monthly peaks are similar in time, but higher in numbers, but so are our troughs (caused by that long shutdown over Xmas and Easter each year for the last 6+ years).

    Hopefully, now thats all finished with (I hope) we should leap ahead and stay well ahead even before the integrated fares/PTOM model of buses feeding trains and v.v. is rolled out.

  7. Matt, thanks for your analysis of the stats as usual. Do you know, now that the AT HOP rollout is complete, if AT publish statistics of boardings at the busway stations on the North Shore in the same way that they do for the rail stations? I would be very interested in how the tiny Sunnynook bus station stacks up. It always seems busy especially during the uni terms, and of course there is no P&R or feeder buses serving the station.

      1. Indeed, be also good to see stats for major bus interchanges (New Lynn, Panmure, Onehunga, Otahuhu, Manukau etc) as well.

  8. How much of the Western Line’s 14% growth rate is because of roadworks congestion on the north-west motorway pushing some motorists or bus passengers onto the rail network ?. Once these roadworks and their associated congestion are finished, some of those who have switched to rail may go back to the motorway. It could be temporary growth such as with the Rugby World Cup, set against a longer term growth rate that is less than 14%.

    1. I catch the western at glen Eden, Waterview works has some effect on CBD traffic from here but only limited (still same number of lanes as usual), Causeway works do not really affect CBD traffic from here. I’ve still noticed a huge change in passenger boarding here, so I think that cannot be a major driver for our part of the line.

      I do think housing costs are leading to many more professionals moving out in this general direction. Demographic change in the neighbourhood and on the platform is noticeable. Western line is a sound choice in the context of housing and transport costs, and in terms of expected benefits from CRL.

    1. If you mean how many people travel to the CBD of each city on a daily basis for work? The census 2013 “travel to work” figures will provide that information. There are previous posts about that subject here.
      And the Stats NZ Website will have the figures you can use. That will also break down the travel by mode e.g. Private Car as driver, passenger, PT, walking, cycling etc.

      For the use of PT in both centres on a daily basis, the link ( has this daily information. as do the AT reports.
      [second graph down on shows Wellington Daily PT boardings].

      For a quick comparison, Wellington has 140,000 total PT “working day” boardings a day (over all modes) as at July 2014, Auckland for the same month had about 42,000 a working/week day on Rail, and 185,000 a day on Buses (excluding the NEX), you can dig the AT patronage report for the Ferry and NEX daily figures, but you can see that with most Bus and Rail figures on a daily “total PT” usage, Auckland is streets ahead – as it should be – as it has much higher population than Wellington. But on a per-capita basis it will be lower than Wellington’s usage for sure.

  9. A few remarks on the cycling stats:

    1) That July-to-July drop is probably worse than 21% after adjusting for population growth over a year. (For those who’ve heard this addendum before, I apologise for repeating it, but so long as the reports don’t adjust for it, it needs to be remembered.)

    2) I can easily imagine that one or more of the counters mechanically failed, or there was some unreported change in calibration or statistical aggregation/adjustment. E.g. my understanding is that NZTA’s evaluations of counter technology included options that ran on consumer D battery cells, and required scheduled maintenance. Also, the reliability of the counters isn’t all that impressive, often requiring variable signal processing that’s prone to inconsistency. If this kind of thing is being used, it’s easy to see how slip ups can occur. Perhaps Max or Glen K can enlighten us on the technology?

    3) Considering that Orewa is an isolated loop (that users often drive to) and also the strange location of many other counters, it doesn’t surprise me that the stats should seem to fall while there is simultaneously a perception of increased cycling in public. That’s because most cycling — especially publicly observable cycling — takes place in crowded urban streets, at eye level, between the front doors of buildings. Several of the counters are not in these locations where city cycling takes place — instead, they are focused on motorway shoulders and recreational loops that I assert bear only a weak correlation to wider reality.

  10. The roll-out of electric trains provides an opportunity to significantly improve passenger services from the west. Split the western line into two virtual lines: A BLUE line runs from Waitakere-Swanson-Ranui-SturgesRoad-HENDERSON-Sunnyvale-FruitvaleRoad-NEWLYNN-MtAlbert-Morningside-MT EDEN-Grafton-Newmarket-(and on South), and a RED line runs HENDERSON-GlenEden-NEW LYNN-Avondale-BaldwinAvenue-Kingsland-MT EDEN-BRITOMART. The two virtual lines connect at New Lynn and Mt Eden allowing passenger transfer. With each service running six times/hour on peak, the wait time for passengers changing trains would be 4-6 minutes,

    Imagine a BLUE train passing through Henderson on the hour. Six minutes later, a RED train leaves Henderson. The BLUE train stops at Sunnyvale, and the RED train gains approximately two minutes, reducing the headway to four minutes. Then the RED train stops at Glen Eden, whereas the BLUE train passes through. The headway thus oscillates between 4 and six minutes. Likewise he headway between the RED train and the next BLUE train (passing thru Henderson at 10 minutes past the hour) also varies between four and six minutes.

    With this system, the time between Henderson and Britomart would reduce by around 17 minutes, and New Lynn-Britomart by about 13 minutes. Some short trips, such as Avondale-Morningside become really inconvenient (you have to go back to New Lynn) but for the majority of trips, even when passengers have to change trains, the overall time is quicker, and the savings on trips to Britomart may be significant enough to attract a new cohort of passengers.

    The cost turns out to be less than might at first be imagined. Running four trains/hour on the present (3/4 hour Henderson/Britomart) schedule requires 3 train hours per hour. Running 6 trains/hour on each of the RED & BLUE lines (total 12 trains/hour, but only 1/2 hour runs) would still only commit 6 train hours per hour, for a spectacularly better service. With the lower OPEX of the EMUs it would be worth trying to get the boarding numbers up. Six services/hour is getting close to “forget the timetable, turn up and go” level of service. You could do the same thing at a 4 trains/hour level and achieve the same running time improvements, but with longer wait times for passengers who need to change trains, but why aim low.

Leave a Reply