Follow us on Twitter

Analysing Auckland’s car dependency

My post the other day about the importance of public transport in reducing car dependency’s financial cost, generated a lot of interesting discussion. There was a bit of debate over whether New Zealand mirrored the USA’s split in household expenditure (it seems we spend a bit less of our incomes on transport), but what there was general agreement on is how car dependency might vary from place to place – particularly across different parts of Auckland.

Using 2006 Census Data (yes I know it’s horribly out of date but for comparative purposes that might not matter too much), we can look at transport modeshare results by Census Area Unit (generally around 4000-5000 people). This tells us some interesting stories. Full excel worksheet here for your information.

Firstly, let’s look at Census Area Units (CAUs) with the highest public transport modeshare: It would seem that key characteristics of parts of Auckland with high PT use include proximity to the city centre and also proximity to high frequency services (bus, train or both). I’m pretty sure every single one of these CAUs is on the isthmus – and most of them fall within parts of Auckland developed prior to the Second World War. Interestingly, they have a pretty wide range of levels of car use and walking/cycling use.

If we turn to walking/cycling next, we see an even more dramatic concentration in and around the city centre – which is no surprise: 
These areas with the highest levels of walking and cycling generally correspond with being those that have the lowest level of car-use, with Whenuapai West being an interesting exception (due to the airbase I imagine).

If we look at Census Area Units with the highest levels of car-use, unsurprisingly the first 20 are pretty much exclusively rural areas: Ormiston (the Flat Bush area) is an interesting inclusion there – though not surprising as back in 2006 there was probably no public transport and almost certainly nowhere to walk to. What the rural nature of these areas highlights is that by living in the countryside you really do become massively dependent on the private car.

The next 20 highest CAUs start to show us a few more urban areas: We definitely start to see a few more urban places here – though generally on the periphery (Weymouth, Dannemora, Wattle Farm, Silverdale etc.) This suggests that Auckland’s urban form in these peripheral areas isn’t really encouraging either the use of public transport or walking and cycling as viable alternatives to driving.

Taking things one step further, I was curious to see whether modeshare had any basic relationship with income – building on my hypothesis that many people in Auckland perceive the PT network to be so bad that their families own multiple cars – regardless of their income. So let’s firstly look at CAUs with the highest proportion of households earning over $100,000 a year: No super-clear pattern seeming to come through there. Many quite wealthy areas (Ponsonby & Grey Lynn) seem very willing to catch public transport, with many (Stanley Bay, Parnell, St Mary’s Bay) also very willing to walk or cycle to work. Perhaps this highlights the connection between richer areas being central and central areas being less car dependent, more than anything else.

Now let’s look at poorer areas, those with the lowest proportion of households earning more than $100,000 a year: The data is complicated a bit by a few very unusual CAUs (Gulf Islands in particular), but generally once again there doesn’t seem to be much of an obvious pattern. Some of these areas have comparatively low car dependency (Lynnmall being perhaps the best example), but many are pretty high – ironically including Mangere Station CAU, which has no public transport use at all even though it’s named after a (now closed) train station. Places like Manurewa and Papakura remain extremely car dependent – even though these centres have become two of Auckland’s busiest rail stations.

I probably need to learn more statistics to properly analyse the information, but the various tables above highlight a few things in my mind:

  • Proximity to the city centre seems the strongest determinant of car dependency
  • Rural areas are extremely car dependent
  • Public transport use is highest in city fringe areas and along well established bus routes like Dominion Road
  • Income, both at the high and low ends, seems not to be related to modeshare much at all, except through other correlations such as central areas being high income areas

What I’d really love is to see some of this information mapped. Guess I should take a GIS course next semester.

Update:

A reader has kindly mapped the car modeshare, which highlights to us in progressively lighter colours the areas with lower car modeshare – pretty much proving that proximity to downtown Auckland seems to be the most important factor in determining whether or not you drive to work:

14 comments to Analysing Auckland’s car dependency

  • This is out of the States too of course but its pretty clear that it is happening here also: Best way to change the value of your neighbourhood upwards? Increase its walkability: http://www.nytimes.com/2012/05/27/opinion/sunday/now-coveted-a-walkable-convenient-place.html?_r=2&src=recg

    How to do that? Improve local services and destinations and improve public transport links and reduce auto priority.

    So it is completely the reverse of the C&R and gov. idea that we can’t afford PT and street and neighbourhood services; it’s that we can’t afford not to.

    • Geoff Houtman

      Walkability- exactly.
      Our neighbourhood is rich and walkable, and we have pretty damn good public transport (unless you want to go to Queen St). I’m proud that people are walking round here, there are some very similar neighbourhoods to the East who are coloured a very dark red.

      A difference in attitude by Community Boards? This is all pre-Superauckland right?

  • Great analysis, but I think there are two important things to note when viewing this data:
    Firstly it is journey to work modeshare, so the modeshare across all trips would be different. Presumably car modeshare for all trips would be higher in most places, except the CBD and fringe where people might walk/cycle even more often outside of their commute.
    Secondly, this is the main means of journey to work for the people who live in the areas, not for the people who work or pass through each area. So for people visiting the CBD you have about 50% non-car modeshare, spread across all the residential areas of Auckland.

  • Peter M

    You’re right Nick. Dealing with the limitations of the data.

  • David

    Anyone else get the feeling that the government is quite comfortable with not doing another census?

    One of the ironies of car-dependency like Auckland’s is that inner-urban neighbourhoods (generally higher income, although not exclusively) get to live in expensive neighbourhoods, and own expensive cars, but don’t need to use them. They choose to spend lots of money on flash cars because they can.

    Meanwhile, the less well off live miles from the city, struggle to pay for two or more crappy cars (if there are working teens in the house) and have to drive everywhere. The old arrangement of the industrial cities where poor neighbourhoods were crowded but close in and well-served by transport, post-gentrification, is a whole lot less friendly to the urban poor than it used to be.

  • Luke

    I’ve looked at this data before too. Few other points. The ‘Other’ Category includes Ferries. In most areas the other caterogry is around 2%, but in places like Devonport, Bayswater and Half Moon Bay this is up much as high as 10%.
    So these suburbs us PT much more than the data suggests. Maybe be able to subtract the standard value for ‘other’ to give a better figure for public transport.
    Another really interesting thing to do once we get the next set of data would be to find out which suburbs have the highest increase of PT (especially rail) use.
    There are also a few free GIS programmes around like MapWindow and qGIS that can do a reasonable amount of stuff, and are able to visualise this data.

  • Logan

    If you can provide a drawing of the area boundaries, I’d be happy to work up some graphs graphically.

    • For those just go to the Statistics New Zealand website, select interactive boundary maps, then zoom in until it pops down to area unit level. That will show the boundaries.

      There is also a GIS data file somewhere on the site with the same, if you want to get really stuck in.

  • Chris Randal

    I work night shift in Victoria/Queen St. I live in Manurewa. My employer pays for parking.

    I take the train Sun-Fri nights and drive on Saturday night because when I finish at 0500 I am not going to wait until 0740 for PT.

    Occasionally I will catch the bus to AKL airport because I get a discount and my wife picks me up from there.

    I fail to see why we can’t have PT 0500-2400 – on Jodi’s basis of “provide it and they will come”

  • LucyJH

    Yeah, I found this data surprising because stuff I have seen elsewhere actually suggests that cycling rates are often higher in smaller towns than in auckland (weird I know, but there you go, I think it’s about the difference between perceived safety and real safety – i.e., people tend to see a quiet country road as being safer than Ponsonby Road at rush hour, even though you’re actually much more likely to die if a car hits you while going at 100 km on a rural road, than if somebody pulls out into you on ponsonby going at 25 km/hour).

    But it makes sense to me that people living on the periphery of Auckland would drive elsewhere for work, while they might walk/cycle to the local dairy/shops etc along their (comparatively) quiet suburban roads. So I do wonder if maybe trips to work are showing you a bit of an unbalanced picture compared to total trips?

    Would anybody like to do a map of walking/cycling across region. I would love to see it.

  • morecityplease

    Awesome map. Of course this correlates with where property prices have been increasing.

  • Matt L

    It will be really interesting to see how this changes in the next census, especially in the areas around the rail network

  • morecityplease

    Good idea Matt L. Of course remembering the geometric characteristics of the streetcar suburb facilitate car-free living by a big factor. My suggestion is that we start seriously infilling the pink areas and make them white as opposed to spending too much trying to make the red areas pink. In very general color-coded terms of course…;)

  • James Pole

    I remember being surprised when I was doing the Census in 2006 that I wasn’t able to complete the transport questions. The questions were structured so that you would answer the transport questions (if I recall) only if you worked full time. As I was a student at the time I was forced to skip the transport questions. I wonder how much impact including students into the figures would have on suburbs where there are higher than normal number of students. Most likely one would find higher PT usage.

Leave a Reply