Auckland’s journey to work data from the census was released yesterday by the council on their site Journey to work is a useful metric but it does have some serious flaws in that as the name implies it’s only recording how people got to work whereas there are generally a lot of other trips at peak times, like to school. In Auckland for example tens of thousands of students enter the CBD each day to go to the Universities or other education providers and those students all have a big impact on transport networks. This can be quite important when looking at PT trends as students tend to be much stronger users of PT than other parts of the population.

I’ll go through the data and how it’s changed over time in the next few days but here are some images from the maps showing the results which in themselves are quite telling.

First up travel to work by car, truck, van or company bus. Unsurprisingly the lowest car use is in the areas surrounding the central city as well as the lower North Shore. Whenuapai West will stick out on many of the graphs which I assume is due to airforce staff having very localised trips. The area around Pakuranga/Howick/Botany really stands out as being quite car dependant which is unsurprising seeing as the PT network in that part of the city have been so poor.

2013 - Auckland Journey to Work - Car

Next we have trips by public bus. What I find most interesting – and completely unsurprising – is that the areas with the strongest bus usage also happen to be the same areas where the most bus priority and frequency exists. Of the dark blue areas, those that surround Dominion Rd happen to have the highest bus usage.

2013 - Auckland Journey to Work - Bus

On to train and that is obviously focused primarily on the areas next to the rail network.

2013 - Auckland Journey to Work - Train

For cycling the highest use is once again focused on the inner suburbs and on those along Tamaki Dr

2013 - Auckland Journey to Work - Cycle

Like many of the other measurements walking to work is something primarily seen in the inner suburbs although there are some stronger patches in some of the suburban centres.

2013 - Auckland Journey to Work - Walk

Lastly Other under which ferries sit and because of that it’s unsurprising to see the areas with ferry service stand out strongly.

2013 - Auckland Journey to Work - Other

As mentioned earlier I’ll be looking into the results in more detail in coming days however what is quite clear just from looking at these maps is that the areas with the higher quality PT, walking and cycling links also happen to be the ones with the lowest car usage. In other words giving people high quality alternatives will see more people choosing not to drive.

Share this


  1. Wow just look how car dependent the southeast is. Why the hell was that area allowed to grow so much without any PT infrastructure?

    1. There is hope for the future: Ongoing investment in the RTN (e.g. electrification), as well as the investment in the AMETI corridor (providing bus/rail connectivity at Panmure), should see a surge in PT use in that area by the time the next census rolls around.

      So we’re a little bit late on the uptake, but heading in the right direction.

    1. LOL, that was why I switched from using “public transport” to “passenger transport” because there’s only one way that you will ever mis-spell public

  2. Hmmm, maybe I’ve done something wrong but maybe there is an issue with the data. I exported journey to work modeshare for all area units in Auckland, and it gives me 5,508 for ‘public bus’. That’s less than train, or bicycle.

    Is it really true that less people get the bus to work than the train across the whole of Auckland?

  3. How long before we can get the census data on journeys to work for the “work address” (i.e. journeys into each census area unit not from each area)?

    That will be the one thats truly interesting as it will hopefully show the mode share changes e.g. into the CBD and other locations over time and allow the impact from Bus priority etc enhancements e.g. along Rem Rd and other locations since 2001 (and 2006) to be evaluated and also show the NEX effect too I think.

    The other thing is that the travel to work data doesn’t allow comparison with earlier census data on the Council website – only can see 2013 numbers – any idea why?.

    Also on the “Other classification” it would be nice to know how people who caught ferries and who do not live adjacent to the ferry terminal got to the terminal (e.g. those on the Devonport peninsula Census Areas Units for example), But of course the Census only assumes you used one mode per journey and only if you are working.

    So for the next census when those in Auckland using PT for eaxmple living on the Devonport Peninsula again), could in theory have to use a Bus, then a Ferry, then a Train (from Britomart) then another Bus and all to get from home to work (under integrated fares and the new PTOM model) – and so what mode do they answer as to what mode they used?

    Or is it that the census needs to consider these multi-modal journeys which will become more common in Auckland in future.

    And if you can still only pick one (e.g. the “main” mode) is this “main” mode the one you spent the most time “on” or doing e.g. walking or the one you went the furthest distance on (e.g. the train)? I think that the Stats Dept need to think about this question for next time.

    I actually think they need to ask people about how many journeys using each mode they made on Census day and for what purpose(s) e.g. work, education leisure etc. As that will give a much fuller breakdown on how is using what modes and why.

    1. The “travel to work” question is based on the mode you travelled the greatest distance on – the census form explains this, so it’s pretty straightforward when you’re ticking the boxes.

      1. Okay…. so thats clear as mud and works fine when you only get a bus or a train or whatever to work..

        But, next census, when you have travelled to your work by a bus, a train, & another bus (or worse, bus, ferry, train, bus) – how will you correctly answer that question?
        You can’t – not unless you dig out Google maps and measure each leg or if you travel one huge leg on on one mode making it obvious.

        And more to the point, how will users of the Census question responses know if your response is accurate or not?

        You can (and do) have billion dollar (thats right a B) or more of spending hanging off the interpretation of what the travel question means in a Auckland, or NZ (as Auckland is 1/3rd of the countries population) context.

        I also read last week in the Wellington paper that Wellington Council is going to do something similar with their trains and buses as to what we’re doing here, and probably have it in place by next census too – which for most people there it will be obvious which mode they used the to cover the most distance.

        Even so I think this question needs expansion – yes the main question stays for consistency with earlier census, but further questions are needed to tease out the real data underneath.

  4. One thing I think that needs to be explored in more detail is the “outliers”.

    In the case of walking, for example, Auckland Central likely achieved a 30-40% walking mode share. In those maps, the highest band is 6.7%, so this tends to underplay the real contribution of walking to managing travel demands in the area with the most congestion.

    In this context, investment in walking/cycling/public transport makes much more sense than major road capacity expansions: Because the former provide a more cost-effective way to meet travel demands in the parts of the city that experience congestion.

    Out in the burbs cars generally work OK for most medium-distance trips. But in the city centre and town centres we need alternatives so that those who are travelling short/long distances can walk/cycle or use PT respectively.

    That’s a long-winded way of saying we need a “horses for courses” approach to transport investment, rather than the “one size fits all” (i.e. roading) based approach that has characterised the last 50 years of Auckland’s development. Horses for course will ultimately be much more effective/efficient …

    1. If you go visit the interactive map on you can mess around with the legend in more detail: bump up the number of bands (max is 9, one more), and configure the range of values used for each band.

      Quick example of walking modeshare:

    2. Stu, if you hit the wrench icon you can uncheck “estimate best distribution” and check “show maximum and minimum values”. Much better, put on walking and the CBD and inner suburb pop out bigtime.

  5. The only blue square of car dependency on the isthmus is stonefields. Would be good to see tamaki station promoted a bit more, as stonefields is just getting bigger and bigger.

    1. would it not be better to connect Stonefields to existing train stations (and town centres) at Glen Innes, Panmure, Sylvia Park and Ellerslie with high frequency bus routes? I mean, Stonefields is literally surrounded by train stations – they are just difficult to access by foot. And I don’t see Tamaki changing that: It’s still quite a long way to walk, so connecting bus services will still be required.

      1. Agreed,
        Most of Stonefields is as far from GI station as the (yet to be reopened) Tamaki station would be by foot. (and its over 800 metres away as the crow flies, which not the same as wlaking distance, which will be higher). So Tamaki station doesn’t solve that problem.

        While some sort of bus based PT is available now, the layout of Stonefields and access only from the north end makes its practicality near zero.

        To my mind what Stonefields is missing is a (PT) link into the quarry from the South West corner (by Lunn Ave/Nahue corner) as this would then allow buses to travel around and through Stonefields as part of longer journeys rather than have some shuttle bus service that only serves Stonefields (from the north) that trundles nowhere else much.

        This link was supposedly investigated by the Quarry Developers as required by the ACC, but was ruled out as not possible by the developers. So was then not required.

        Yet if you walk to the Quarry on the corner of Lunn Ave and Ngahue Drive you can see the ease with which a PT path (busway) could be put into place there without impacting either residents or businesses too much. The main issue would then be where this linked to the Lunn Ave/Abbots Way/Ngahue intersection.
        If ever there was a good reason/place for building a bus ramp – this would be it.

          1. I’ve cycled along that several times, and lots of people use it. So its not quite the dungeon you imagine it is.

            In fact, the trees provide shade during summer, and some respite from wind and rain in winter.

            Yes lighting and the path need work (and it could be wider), but thats not unique to this part of town.

            To be honest I’d rather a decent wide shared path link be put in place from the GI Station platform to the new supermarket/retail area on the corner of FM and Merton roads – right across the park for this link would be fine as this park is not used for anything except the BMX’ers down there who use the northern end of the park as their car parking lot.

      2. GI station is the closest to stonefields, but there is a steep climb up stonefields ave which must be repelling people. I think this is called Taylor’s hill or something and is actually a volcano in its own right. Stonefields to panmure is also inconvenient as you have to climb out of the quarry. Stonefields markets to tainui road is flat and easy to walk. Tamaki station also serve the suburb of tamaki in the other side of the tracks. A little too far to walk to panmure or GI. Tamaki is a much more valid station than meadow bank, and it’s surrounded by a massive brownfields site that is on the radar of redevelop company. So much is invested in the infrastructure that runs through the east so why not allow people to access the infrastructure when Westfield is closed the eastern line will be one station down anyway.

        1. I would think what is repelling people from walking or biking to GI is the horrid arterials in their way. 4 lane, difficult to cross, roads everywhere.

          1. College rd is pretty bad, especially the roundabout next to bluegrey ave. the vegetation makes it hard to see.
            Anyway, when the AMETI rd is finished all those cars going on to the city will probably go via college rd. but peds could still make it to tai nui rd fairly easily by walking along the south side of college and cutting through to homestead.

          2. Looking at the Auckland Council GIS tonight, i think a nice route for both Stonefields and the old St Johns area could be easily established. I’ll draw it up and post it tomorrow.

          3. There is no route through Homestead to Stonefields currently (it has 4ft high fences between it an Stonefields) and that land in Stonefields bordering Homestead is all small lot development so won’t have any kind of walkway ther -, so most Stonefields folks will have to continue to walk out to College Road somehow then along College towards either Morrin Rd (if going to Tamaki/Panmure) or up the hill from Stonefields Ave towards Morrin then along to GI.

            Note: once the Ameti Rd is finished there will be a (flat) walk along Morrin to Panmure Stn beside AMETI Road.
            And its actually no further than any walk to Tamaki Stn would be as it has to go via Morrin Road to Jellicoe and then double back – so for now it will be easy to walk that way to Panmure. (its still 1.4Km though).

        2. Tamaki
          The Volcano you refer to is known as “Purchas Hill” which is a remnant stump of a scoria cone volcano that preceded Mt Wellingtons cone by 500 or so years (and which was surrounded by ash and lava flows from Mt Wellington). Purchas Hill’s,crater was still standing intact, centered more or less where current College Road and Morrin Roads meet, until about 1940, then the southern half was quarried away completely leaving the land flat, and the northern half of the cone was gone the by 70’s.

          Purchas Hill was thought to have been totally lost, until during the land shuffle/swap for the current Netball courts, its remnant stump was “rediscovered” in the undergrowth and Auckland Council paid Winstones a lot of money for the land it stands on – for what is basically a scoria stump not really worth saving and they have left it ever since with the promise to make it a park someday.

          Meanwhile they have since shoved a 4 lane road up its arse, further insulting the area.

    2. More people used the train on census day who lived in Stonefields than caught the bus, so the Train is already used by folks (albeit not many of the total residents do so).
      In fact Stonefields train use was higher than the Sylvia park station “adjacent” Hamilin and Mt Wellington Sth Census Areas. So folks are using the train already.

      But it is one of the few places in the isthmus areas with a low bus patronage – its actually the lowest %age of Bus users of the entire isthmus area at 3.1% – trains was 4.5%), and thats probably a bigger indictment on the poor connections to PT down there than not wanting to walk to the train station.

      And opening up Tamaki won’t make a train even 100m closer for most Stonefields residents in this census.
      Future census may show more people using trains, but thats because the north eastern corner which is the closest to ALL/ANY train station is being built up now, so in 5 years time it will have a small uptick for sure in train usage from this corner of the development and for little other reason.

      1. It would help somewhat if the GI station could be integrated into the suburbs to the western side of the station rather than continue to be cut off.

        1. College Rd is now a 4 lane arterial. That alone should never have been allowed. It is a crime against walkability.

          1. Well that was sold on the basis that the outer lane in each direction will become dedicated bus lanes.
            Fat chance.

        2. It does – there is an under pass from the Station, and from GI itself that links to the park and then to Felton Matthew Avenue.

          There is also access to the station platform from Merton Road on the western side.
          Remember there has been a huge swathe of reserve land for a future motorway sitting there for 50+ years which has always severed the train station from the community there more than anythng else. Currently the western edge is nothing but a timber yard. And until they lift the motorway designation nothing more will be done.

          1. I’d say the commercial businesses have as much responsibility for the severance. It’s not very inviting to me much less my wife and child so I’d imagine local residents mostly feel the same way.

          2. Tamaki, get on to your local board and get all the trees cut down along the walkway from Felton Matthew to the train station. Add some good lighting and seats for people to dwell. That’d be a cheap start. Likewise here Get the path out from under the trees. Make it light and safe feeling. Build a nice bridge from the walkway, across and joining the station, over Apirana Ave to join onto Delwyn Ave. Narrow, gloomy underpasses are horrid things that do not encourage use. Gosh we do mange to destroy neighbourhoods in the name of progress don’t we?

          3. Sorry, the reply buttons have dropped off some posts.
            Bryce P, yes the Felton Matthew underpass is horrid. It will close soon to cut the trees, and I’ve heard that they will be doing up the underpass with new lights etc. They are currently building a new countdown supermarket and retail centre on the corner of Felton Matthew and Merton, so it’s a good time to do up the underpass.
            Whoever said there is a fence at the end of homestead, I vaguely remember there was also a wooden turnstile. Regardless, fences were made to tear down, and this wouldn’t stop a ped walkway if thats what made sense.
            the unitary plan has two special precincts in tamaki, one of them indicates another 9,000 people around the old tamaki station site. Yes, they could all walk 1.5km to panmure station, or they could catch a ferry type bus to GI station (although this has never really worked in akl), but if there is no station then it’s likely many more of them will start to depend on the internal combustion engine. This would be a criminal shame given the billions invested in the rail corridor passing underneath them. Add in the future 10,000 in stonefields and the 3 to 4 thousand in between panmure and GI stations and you have a very compelling case. Not, note only for these people to commute to the city, but for others in the city to commute to tamaki where there is forecast to be high jobs growth. This ticks so many boxes it’s a no brainier.

          4. Correction Tamaki, Stonefields population was predicted at time of Environment Court approval to be 7,500 not 10,000 and that was based on the proposal to build 2,700 dwellings on every meter possible in the Quarry that they could.

            And a large chunk of those were to be high rise apartments. furthermore some of the land adjacent to Mt Wellington was blocked off by the Environment from ever being built on (apartments were planned for there), which reduced the number of dwellings down by about 10% early on, and the Developers backed away form that 2,700 number after the GFC as they publicly stated that they could not sell as many apartments in Stonefields like they assumed – so built more terrace housing instead, so the actual final density of Stonefields will be between 5,000 and 7,000. We should know next census.

            As for fencing – well its 6ft high fencing at Homestead put up by the Stonefields developers obviously to keep the riff raff out – so its not exactly Ped friendly.
            Once that part is developed they may put up some nicer fences and maybe a walkway, but that is not for some time.

            I don’t disagree with reopening Tamaki and never have but there are other considerations needed, like how close it is to both Panmure and GI stations which does introduce operational issues with a station there. And it won’t help Stonefields one jot.

            And don’t forget this is right now little more than a raised island platform in a sea of rail tracks right now.

            Without access and facilities its useless to anyone and those roads and access methods from it are not coming until some time after AMETI road opens when the industrial land is redveloped. I don’t see 3000 or whatever dwellings suddenly appearing around Tamaki Station either – thats a 10 year project to do it properly.
            And what form that industrial land will end up with who knows. Although a fair chunk of it is owned by Auckland Council (from the old ACC).

            So there is time and a need to take time, to get it right.

          5. Sounds like the chicken and egg problem from manukau city circa 1995. Let’s get lots of residents first then worry about public transport later. To encourage high density living you need the transport infrastructure first. This also encourages investment.

          6. Exactly right Tamaki. Transit needs to planned at the design stage and service delivered at the beginning. This is simply a ordinary level of competence we should expect from our institutions. No one would expect to move into a new place like Stonefields without running water or a functioning waste water system. So it is with bike lanes and an adequate and scaleable Transit service.

          7. Not quite Manukau, we have PT (trains and buses) there already, and its being enhanced as we write.

            Since Tamaki is a SHA is up to the Tamaki Redevelopment Corporation or whatever they call themselves to manage this co-ordination of PT with everything else.
            As they are the government (local and central) appointed agency to do all this – so they have a clear job to do and a mandate to do it.

            As far as I can tell any initial development of the SHA is being done in the northern/eastern portion of GI nearest the Tamaki river which is pretty far away from Tamaki station and Panmure and closer to GI than anything else.
            But it won’t be “next door” and so will be at the outer edge of walkability to ANY station. So buses will have a big part to play.

            The PTOM model changes to bus route have not yet begun consultation around GI, and thats what will ultimately govern what PT is put where and how accessible it is.
            [and needs integrated fares so you can hop the bus then the train quickly and easily with only one fare being charged to make it all work].

            Don’t forget while the draft UP is operative in the Tamaki SHA now, it is not operative anywhere else around there yet and won’t be until 2016. So untilthen Tamaki SHA is an island to itself.

            That’s why I say there is both time and a need to take some time, to get this redevelopment totally right, or we will end up with something worse than we have now longer term.

            We don’t have to house 9,000 people tomorrow in the Tamaki SHA – there is no ability to do so (couldn’t build 3,000 houses quickly right now) – even if the PT was there, nothing else is.

            And I’m sure you’d rather a proper job is done than some half-arsed “put up the houses now and do the rest later” solution like what got GI in the present mess 60 odd years ago now right?

          8. Well Patrick,
            Stonefields was promised “an adequate and scaleable Transit service” by both the developer and Auckland City Council.

            Neither has delivered that, some 7+ years after the first residents moved in.
            This is evident from the census bus patronage figures [and how few people actually use the bus service on a daily basis].

            So agreed Stonefields is a PT disaster – doesn’t mean that Tamaki SHA will be or needs to be – for a start its not in a hole in the ground with one way in and one way out.

            But it does have severe severance due to the railway line and AMETI “motorway” being shoved through to Merton Roads currently..

            there are only currently 2 crossings of the railway between Panmure station and GI – Morrin Road over bridge and Merton Road underpass.
            And right now, nothing planned will improve that situation.

            And until its done, you have a problem, so to my mind the first order of business is sorting out the severance issues, then everything else can flow from that.
            Form follows function.

  6. It would be fascinating to correlate this data with obesity statistics. While I don’t recall the subject being addressed in this census (aside from anything else, self assessment of obesity would be too subjective to be accurate), surely it would be possible to extract it from DHB data?

  7. One problem with Journey to Work category is that many such journeys are made up of several components using different modes – eg. a commute from Devonport to Newmarket made up of walking, ferry, and train.

    1. And in the future with Auckland multi-mode journeys being routine this question will become a true lottery as you interpret what the census answers mean.
      I think they need to rethink this section completely.

  8. In Australia I recall the census question asks the mode covered by the majority of the distance.

    While this post is about travel to work, this is only about 20% of trips. In Australasian cities with a more developed PT system, a large proportion of recreational and social trips are by PT, particularly under-30’s and those of non-Caucasian descent. I was surprised on a Monday night recently how many there were in Melbourne CBD for socialising, and how many used PT. Trams were full with standees after 9pm. That was a relatively quiet night. Last weekend Melbourne hosted “White Night”, an all-night arts exhibition in the CBD with interesting projections on the city’s heritage buildings. It attracted about 500,000, and these numbers would only have been possible with a good PT network. Trains and trams ran at a half hour frequency all night, but even with 6-car trains this frequency was totally inadequate. When waiting for an 11:10pm train at Flinders Street I talked with a lady who had been unable to get on the last 2 trains and had been waiting for over an hour. The 11:10 was no better, and we secured a little space on the following train at 11:40. Comments on social media from the few that drove were that it was easy to get parking, so it seems that for a major event such as this in the Melbourne CBD, most people only think of PT. Nevertheless, I did see one of the major arterial roads south of the CBD gridlocked. When there is a good rail PT system, it allows cities to accommodate much larger crowds to such events than could ever be expected from a car-based or bus-based system. Cities with these systems don’t even seem to think about such events. Neither of these systems could work well with most city streets closed for pedestrian use only, as occurred for White Night.

    For the first time I experienced pedestrian gridlock. Normally we think of pedestrian routes having the highest capacity per square metre of space. But the crowd was so great and dense at the corner of Swanston and Flinders Streets that it just seemed to go no-where.

    1. Why would a bus based system be different from Melbourne’s tram based system? I think it’s the management of the event and road closures that is the key thing there.

  9. In Melbourne the “heavy lifting” of passengers is done by the train system, and the tram system operates mainly as a “last mile” service within the CBD, and for travel to the inner suburbs. On White Night the trams stopped a block or two short of the event and turned around. Because trams have a cab at both ends, there was no need to turn around a block, and could get closer to the event. Trams can hold more than a bus, which affects the variable costs of providing the service. The latest E class tram has a capacity for 210 passengers, which is equivalent to 4 buses, so fewer drivers need to be paid for the same level of capacity.

    In a city with a strong bus system to its CBD, a lot more drivers would need to be rostered to provide the same capacity as a tram or train system, and there would need to be careful planning of bus stop locations to avoid areas of pedestrian overcrowding. Bus stops are unlikely to be in their “normal” place which could reduce the confidence users have in choosing to take that mode. It is all perfectly do-able, but I expect would take more planning than for a rail-based system.

    The train overcrowding in Melbourne could probably have been solved by employing as few as an extra 8 drivers, so the frequency on the busiest 2 lines (to Dandenong and Ringwood) could be changed from half-hourly to every 15 minutes.

Leave a Reply