This is part 2 of a 3-part series on household emissions in NZ. Part 1 looked at whole-of-life emissions from housing, and this part does the same for transport. Part 3 will tie them together.
Last time, I started with a graph showing the carbon footprint for an average New Zealand household, with transport very much in “big toe” position:
Stats NZ have kindly provided more of a breakdown on the transport figures, letting me make the new graph below:
An average NZ household creates 5.2 tonnes of ‘direct’ emissions a year from using petrol and diesel, with another 0.9 tonnes of ‘indirect’ emissions coming from the supply chain. These figures will vary between different households. ‘Household composition’ makes a big difference – is there one adult who drives to work or are there four flatmates who all drive? And there are regional differences: Wellington households have consistently lower emissions than Auckland ones.
But today I’ll focus on the difference that location makes within the Auckland region.
So how can we measure transport emissions for households in different locations? There are several different data sources we can use, because there’s a largely linear relationship between travel, fuel consumption (or fuel spending), and emissions. On average, cars use a litre of petrol for every 10 km they travel (real-world results, often 30% higher than the lab-tested results that get quoted), and that petrol will create exactly 2.3 kg of emissions.
There have been a number of New Zealand studies using these various data sources, and we’ve often featured them on Greater Auckland. These include Peter Nunns’ analysis of census commuting data…
… and my analysis of petrol station spending data. With a few assumptions, I can convert this into an estimate of emissions (including the upstream ones):
There isn’t a straight-line relationship between distance and emissions – it’s logarithmic instead, which means that the biggest increases come early on (e.g. emissions increase more going from 5 km to 10 km than going from 25 km to 30 km). But they do increase throughout. People living a long distance away from the middle of town travel much further distances – not just for work, but to access education, shops, services and so on.
The equation on this graph suggests that a household living 5 km from the city centre will create five tonnes of driving emissions a year, a household living 10 km away creates six tonnes, and a household living 30 km away creates seven tonnes.* 5 km would be Balmoral, 10 km would be Stonefields, and 30 km would be Kumeu, Takanini or most of the other places where Auckland is sprawling.
Looking ahead to the next decade when major decarbonization will be required, it will be much easier for central households to cut their emissions than for remote households.
Even just looking at the existing emissions, the difference between five and seven tonnes might not seem huge, but it is significant. If anything, I think it’s understated for new housing. A lot of the outliers below the trend line are town centres – Papakura, Warkworth etc – and I suspect that people have historically chosen to live there because they work nearby. New subdivisions are more likely to be above the trend line, as these new areas tend to take on a much larger share of housing growth than they do of employment growth. Unless transit is built in from the start, they’ll probably be more car-dependent than the suburbs next to them.
Lastly, a quick reflection. “Food and beverages” slots in between transport and housing in the average carbon footprint. I was honestly quite surprised at how big a contribution it makes. I mean, I love meat, and I don’t drive much, so for me it’s probably the big toe (or a delicious Achilles heel?). My third-ever post for Greater Auckland said that “as consumers, the best thing we can do to reduce our contribution to global warming is to change our transport habits”. But our eating habits are right up there too; I’ll certainly be taking another look at mine.
* I’ve done some regressions against demographic variables – household income, household composition and the average number of people employed per household – which still give very similar results for the distance effect.