Balranald in outback NSW with a population of 1,600 is as urban as Sydney according to the ABS
Balranald in outback NSW with a population of 1,600 is as urban as Sydney according to the ABS

I rarely watch ABC-TV’s Q&A (I see enough narcissists during the day, thank you) but my attention was drawn to a claim made on Monday’s program by Liberal Senator Eric Abetz. Responding to a rambling question about the implications of the Trump victory for the future of the Turnbull government, he noted that Trump’s success in regional areas has lessons for Australian politics. According to the Senator:

We are, as a country, very much a rural, regional country. We are decentralised. I remind the ABC of that from time to time – that Australia is not Ultimo and ABC in Sydney, but the rural regional areas, and we forget them as a mainstream centre-right party at our peril.

Some viewers reacted by pointing out on Twitter that the Senator is wrong; Australia is in fact highly urbanised. For example, one tweeted: “If Eric Abetz thinks Australia is “rural/regional”, maybe he should meet the 90% of us who live in metropolitan cities”.

So what is regional/rural and what is urbanised? The ABS has two key measures that show around:

  • 66% of the Australian population lives in a capital city.
  • 89% of the population are “urbanised” i.e. live in a township or city with a population of 1,000 or more.

I’m not sure what a “metropolitan city” is exactly, but if it means our sprawling capital cities then “90%” is too high. Nevertheless, the Senator’s claim that we’re a regional/rural country is wrong. If he’d restricted his assertion to his home state of Tasmania he’d have been on stronger ground – only 42% of Tasmanians live in Hobart. He’d also have had a better case if he’d referred to Qld where a majority of the population lives outside Brisbane.

Then again, add the population living in large regional centres like the Gold Coast and Newcastle to the capital cities count and circa 80% of Australia’s population live in urbanised areas. Maybe not Tasmania, but it’d certainly rule out Qld as an exception. Senator Abetz could claw back some dignity though by pointing out he specifically referred to “rural/regional”; pretty much any city that isn’t a capital comes under the rubric of “regional” in Australian political discourse.

We should be careful however about attributing too much significance to the ABS’s measure of urbanisation since it starts at a population of just 1,000; the last 10 percentage points or so of the 89% figure are questionable. A township with a population as small as 1,000 persons – something like Balranald in outback NSW with 1,600 residents – is a pretty small place. Does it explain much to describe the residents of Cobram and Kyabram in Victoria, which have populations around 5,000, as “urbanites”?

The ABS is a statistical body, not a planning or population agency. It’s definition of urbanisation is mechanistic, essentially based on the extent of built-up area. There’s limited explanatory value in a geographically based definition that includes towns like Emerald and Dalby in Qld (with around 10,000 residents) in the same category as urbanised Sydney with 4.5 million residents and urbanised Melbourne with 4.3 million.

Country towns don’t deliver the agglomeration benefits of diversity and specialisation in employment, culture and social life that draw people to Australia’s big metropolitan areas. That’s especially true for very small centres but it’s also true, albeit to a lesser extent, for the larger ones like Ballarat (100,000) and Bendigo (93,000). It’s why young people leave the regions, including the larger provincial centres, for the major metropolises.

When looking at the attitudes and needs of populations in different geographies, we need another metric for urbanisation-rural that provides a more sophisticated understanding of the differences between places; and that might even account for differences in political views. With the strong push for decentralisation in Australian politics, can we take comfort in assuming there’re dozens of university researchers currently looking closely at policy-relevant questions like this? I doubt it, unfortunately.