Housing is considered affordable if it costs less than 30% of a household budget. Transportation is the second largest expense for families, but few consider these costs when choosing a place to live.  Center for Neighborhood Technology

We’ve run a few posts lately discussing the topic of housing affordability and transportation costs. Of course you can’t separate the concept of housing affordability from transportation costs, since most people have to travel to work to pay for their housing.

The posts were based on recently published research from the University of Otago, by Kerry Mattingly and John Morrissey titled, “Housing and transport expenditure: Socio-spatial indicators of affordability in Auckland”. Their work included a series of maps that showed a significant transportation premium associated with living in the distant suburbs. While the maps they published did not assume everyone would be working in the cbd, the results of the research showed that households furthest away from the centre generally travel greater distances for work and therefore spend more of their income on housing AND transportation.

There was a robust discussion in the comments about how the data was not so useful since everyone’s situation is so different. This is where Alex Raichev and Saeid Adli come in. Using the methodology of Mattingly and Morrissey as a starting point, they developed a dynamic website called Affordable New Zealand that shows housing (rent) costs + transportation costs and allows users to identify housing and travel costs based on specific location of work.

They have included a number of factors which  contribute to the overall cost of living such as parking and car ownership. The car ownership cost reflects the sunk cost of the car, even if the mode of travel chosen is public transport (or walking, cycling). Further travel costs for the car are determined by distance  traveled to work. The public transport costs are estimated using the formula from Mattingly and Morrissey.

Here is a look at the results using Albany as the job location.

Where to live? Job in Albany.
Where to live? Job in Albany.

Another way to interrogate the data is to consider the pin drop as a housing choice and then the data reveals the relative affordability of accessing employment areas.  Here is a comparative fictitious example of a young family with one parent working trying to decide between living in Titirangi or Freemans Bay. The Titirangi family has 2 cars and use 1 car to go to work paying $15 for parking. The Freemans Bay family has 1 car, but uses PT to get to work.

Affordable Lifestyle Choice #1: Titirangi
Affordable Lifestyle Choice #1: 2 Cars in Titirangi (source: affordability.org.nz)
Affordable Lifestyle Choice #2: Freemans Bay
Affordable Lifestyle Choice #2: 1 car + PT  in Freemans Bay (source: affordability.org.nz)

The site is still under development. One idea is to expand the search criteria to suit a couple/household scenario with two unique job locations.  Another improvement will be refining the data to provide a more accurate cost of public transport.  Wellington and Canterbury versions are coming soon. The project is open source (Github), and if you have any comments about the project leave them below or contact Alex from the notes page.

Share this


  1. Very neat. It would be great to be able to apply a cost to commuting time too. Although I could live in Clevedon and cycle through to the CBD every day, there would be an opportunity cost for the large amount of time I would be sacrificing to do so.

    1. I apply the same amount as I’m getting paid. Get a lot of people saying that I can’t do that but it’s my time that I’m losing. It’s why I think that travel time needs to be a cost on the business that’s doing the employing. As it is the cost of all that lost time is born by the employee and it’s not being accounted for and thus making sprawling cities seem cheaper.

      1. But it’s no different from the time I spend doing the breakfast dishes or feeding the cat… And indeed I’m on the bus right now, just cleared my emails for the morning and now have a little blogging time.

        It needs to be a bit more sophisticated than applying your hourly wage. I don’t actually value my commute time that highly because, well no one is going to pay me to not do it but people sure do pay me to go to work. Same with tidying up and keeping Miss Nibbles fed, it’s never a monetary thing.

        A related question, why do we always have to monetise non-monetary costs? Can’t we just say you’ll lose seven hours of your life each week if you do that commute?

        Perhaps that’s another way to represent the data, hours per week of commuting time.

        1. But it’s no different from the time I spend doing the breakfast dishes or feeding the cat…

          But it is. It’s time taken away from you by someone else.

          I don’t actually value my commute time that highly because, well no one is going to pay me to not do it but people sure do pay me to go to work.

          I do because there’s other things that I can be doing rather than getting stuck in traffic. As such it encourages me to look for work closer to where I live. If it was billed to the company doing the hiring it would also encourage them to look for people closer to the work place and to use PT for the actual accounting. Either way it should encourage a denser city which we need.

          A related question, why do we always have to monetise non-monetary costs?

          Thousands of years of habit 😈

          Seriously though, in the present economic paradigm it’s a way to measure which is the best option so that we can decide the most efficient thing to do. Please note: I’m not exactly in favour of the present economic paradigm as it drives us to destruction through excess resource use.

    2. True. Although if cycling you don’t need to join a gym. So you’ve got no further cost or time commitments. Car commuter has to also pay for gym membership and time to go to gym in addition to commute. It would also be good to see a red flag for schools as well, with option local school or elsewhere. And if elsewhere add in the private school fees.. Obviously quite a few in Auckland will commute significant distance for preferred school and it has a significant impact on household time and money. As well as Auckland traffic.

  2. Excellent, Alex and Saeid, and making this open source was a great move. So much can be done with the platform you’ve built here!

  3. I think instead of adding travel time cost value which would vary for everyone, a better way would be to add limits to the willingness to travel say 30, 45, and 60 mins. Perhaps the meshblocks that exceed the threshold could turn white.

    1. In terms of affordability those things are already reflected in house prices. A house catching the sea breeze on the clifftop of Mission Bay will be substantially more expensive than one next to the motorway in Otara,

    2. C’mon- it’s a great tool to demonstrate a very important point. If you include every possible parameter you make an unusable mess. This is great at showing the two biggest expenses, one of which everyone is obsessed with, and the other that is usually ignored.

      1. I like your comment. Anyone who claims that you can save money by paying a location premium for a like-for-like home and travelling less, does not understand urban economics 101, or basic real estate markets.

        Even RE agents don’t have to be urban economists to know that the savings on transport cost and time capitalise into the value of property at more efficient locations. Everyone knows this property is more expensive because of the old “location, location, location” rule but included in that, is always the reality that you can’t “save money” by moving there instead of accepting a less efficient location.

        I did a long critique of the Mattingley and Morrissey paper that is posted HERE:


        1. Are we assuming like for like?

          I don’t think there is a single property in all of Torbay like a traditional isthmus villa

  4. Great idea! Can the number of days commuted and the aggregator be broken out e.g. 9 days per fortnight, 22 days per month etc.

  5. Great tool. Great assumptions too. Owning 1 car is nearly a “given” in Auckland, so is not part of the cost model. Schools may need to be factored in. Primary schools don’t exist or are full in much of inner Auckland.

    1. And for many, owning multiple cars is a given. It’s not uncommon for households in the poorest parts of the city to own 4 or 5 cars, because of the dispersal of housing and employment (thanks, Manukau City Council) and seriously crap public transport provision. I imagine a huge percentage of those cars are paid for on hire purchase: a huge recipe for poverty among those who can afford it least.

  6. It is a great website. But travel time is based on free flow times rather than congestion. Greenhithe to CBD is given as 2 hours per week when at the moment it is an hour each way or 10 hours per week. It would be good to link it to a congestion model.

    1. Good point. This is where we could use some crowd sourcing. For example, we could assign a time penalty for every trip across the harbour bridge.

  7. There is some work required before this site is of great use. The most glaring example is Waiheke Island where it seems that an assumption may have been made that residents do not work in the city. Also ferries have not been factored in at all as a commuting option!

  8. Cool site … one thing I noticed is that for my specific situation, the calculated bus fare cost is about $17 too low per week on standard hop fares … I get the feeling bus fares have just been calculated on a per km basis or something, rather than the zones? (understandable if so, as I’m sure adding zones would be tricky).

Leave a Reply