Where are cities growing?
There’re some interesting takeaways from the latest ABS estimates of population growth released last week. You can get a broad overview from demographer Bernard Salt’s analysis, but I think there’re a few other points pertinent to cities that warrant examination.
For a start, I expect many will be surprised at how small a proportion of each city’s population lives in the inner city. In each of the four largest cities in Australia – Sydney, Melbourne, Brisbane and Perth – less than 10% of metropolitan residents live in the inner city.
Defining comparable areas across cities with big differences in population, geography and housing stock is an inexact science, so I’m going with what the ABS defines as the inner or central city in each case.
Inner Melbourne, which roughly equates to a 5 km radius around the CBD and stretches out to St Kilda, Prahran, Fairfield and Kensington, has a current population equal to 8.0% of the metropolitan area’s population. Inner Sydney and Central Perth have similar shares, but Inner Brisbane, which includes Bowen Hills, Kelvin Grove, Dutton Park and Paddington, has only 4.7%.
Brisbane of course was always a suburban city. Even historic suburbs like Spring Hill which today are on the edge of the CBD were comprised mainly of detached houses. Although founded nine years earlier than Melbourne, terraces were uncommon in the northern capital.
Another interesting observation is the high rates of growth in the central areas. Inner Sydney and Central Perth both grew by around 23% over the ten years from 2001 to 2011; Inner Melbourne by 32%; and Inner Brisbane by a spectacular 43%. That seems to gel with the high levels of residential redevelopment visible in and around the centres of Australia’s capitals over the last ten years.
However – and it is a big ‘however’ – the inner city’s share of metropolitan population barely increased, as illustrated by the first exhibit. For example, the share of metropolitan residents living in Inner Sydney – which includes Botany Bay, Leichardt and Marrickville – increased from 7.1% to 7.8% between 2001 and 2011. If this glacial trend were to continue, by 2031 Inner Sydney would have 10.3% of the metropolitan area’s population!
Inner Melbourne and Inner Brisbane had similar small increases. In fact Central Perth, which includes Subiaco, Claremont, Cottesloe and Cambridge, actually lost share over the period, dipping slightly from 8.8% to 8.6%. The explanation is simple – the population grew even faster in the rest of the metropolitan area, especially in the outer suburbs.
Clearly the vast bulk of population growth in capital cities took place beyond the inner city. In Sydney’s case, much of it was located within established middle ring suburbs like Auburn, Holroyd, Parramatta and Blacktown. However this wasn’t the case in the other cities.
For example, the second exhibit illustrates that while Melbourne’s middle suburbs added 202,000 people over 2001-11, the outer suburbs contributed a much more impressive 384,000. In fact population grew by a third in Inner Melbourne and the outer suburbs (32% and 36% respectively) but by a mere 9% in the middle ring suburbs. The middle ring’s tardy performance is a problem because it had the lion’s share (62%) of metropolitan population in 2001.
Recent research by the Grattan Institute suggests a substantial proportion of settlers in the outer suburbs of Melbourne would prefer a location closer to the centre if it were affordable. They would be prepared to forego space and a detached house in order to avail themselves of a more accessible location.
Yet it is starkly clear from the data that the middle ring suburbs, despite housing most of Melbourne’s population, simply aren’t responding to this demand. They’re not contributing – the necessary apartments and town houses simply aren’t getting built. A key reason is existing residents and councils tend to oppose multi unit housing developments.
The inner city isn’t maximising its contribution either. Most of the growth over 2001-11 took place very close to the CBD. For example, Leichardt, Marrickville and Botany Bay together accounted for just 17% of population growth in Inner Sydney – the rest was located in the City of Sydney (and most of that in the South and West of the municipality).
Not everyone wants to live on the fringe or on the edge of the CBD. Ways have to be found to increase housing supply in middle ring and inner city suburbs. This applies to all capitals but is especially pressing in Melbourne, Brisbane and Perth. The ‘bicycle wheel’ model – where growth is focussed on the hub and rim – is untenable.
I’ve previously discussed various policies proposed to increase housing supply in established suburbs. These include multi-unit developments in activity centres and along transport corridors (here, here and here); infill development (here and here); and redevelopment of disused industrial sites (here).
These are valuable approaches but they’re not delivering supply of new housing on the required scale and on trend aren’t likely to. Other approaches need to be considered. I’ve suggested some, including developing surplus land in large institutions like suburban universities (here); retrofitting existing industrial areas (here); and creating major new suburban activity centres in places like Melbourne’s Clayton (here).
Note: The LGAs I’ve classified as ‘outer suburbs’ of Melbourne are: Mornington Peninsula, Frankston, Casey, Cardinia, Yarra Ranges, Nillumbik, Whittlesea, Hume, Melton, Wyndham.













Please login below to comment, OR simply register here :
Thank you for registering, we have just sent you a confirmation email, which includes your new password to be entered below.
Perhaps the first thing to do is to remove the existing barriers to redevelopment. These are things like height limits (a maximum of two storeys does not help), requirements for less than 50% floor space ratio (if one story, a max ground floor area of 50%, if 2 story, a max GFA of 25%, etc), minimum requirements for car parking (if the building owner decides that only 200 car spaces need be provided, why require him to provide 250 (this would mean either buying additional buildings to convert into car parking, or to add another parking floor), unreasonable frontage set backs (why should a building have to be set back so that a ‘stretch limo’ could be parked on the ‘drive’), unreasonable frontage requirements (why require a double fronted garage facing the street when a single garage front but extending back far enough to accommodate two cars would be as effective).
Getting rid of these requirements would be a good start, no doubt your readers can think of others.
And remove the ‘right to object’ to all but the immediately adjacent neighbours (and perhaps not even them). If someone buys the land, and wishes to develop it for housing, it should be up to him to decide what is the best way to do so, and “put his money where his mouth is”.
Melbourne has something like 1.6m households and Melbourne @ 5 million (2008) proposed a roughly 50% split between fringe growth and consolidation for the next 600,000 households over 20 years.
That the 50/50 split hasn’t happened (yet) is, as you rightly point out, a failure to achieve development in middle suburbs. The ‘bicycle wheel’ analogy is apt.
While Rob Adams’ suggestions for high density development along transport spines had a number of practical holes when it came to implementation, he did demonstrate that we don’t need major density increases everywhere to accommodate lots more households. The trick will be to identify the locations where differing approaches will actually be encouraged, and what they are. There has been lots of vague talk about this in the past (e.g. GoGo, SlowGo, NoGo zones) but no concrete action.
As for the Grattan Institute research, one of the impediments to the community actually buying non-detached housing is the inertia and transaction costs inherent to our system. We have high levels of home ownership. With stamp duty, agents’ fees and removalists we tend to buy a house for more than just our current needs – we look to the future, or often continue to live with the past, because it frequently doesn’t make economic sense to do anything else. This one might be tougher nut to crack!
I’m not sure you can ever fully remove a “right to object” – local governments are representatives of their citizens and if councillors consistently fail to take into account of resident’s objections they’ll be replaced by others that are happy to.
I would say though I’d much rather we had a system of requiring the payment of financial recompensation to neighbours if a proposed development can clearly be shown to be detrimental. I suspect more than a few residents would be rather more happy to accept a redevelopment blocking some of their afternoon light or view from the balcony if there was some monetary bonus involved (plus of course, it would discourage developers from putting forward proposals that would be likely to cause such issues).
What I’d actually prefer to see is laws strongly discouraging or even prohibiting developers or speculators from sitting on undeveloped blocks for long periods of time – not only are they eye-sores and an inexcusable waste of space, but they represent significant foregone revenue for governments. There’s a very large block of land not far from me in Hawthorn that has been a wasteland for years – apparently it’s even known as a drug-dealing haven, yet obviously there’s insufficient pressure or motivation for a developer to do something about it. I doubt a single nearby resident would prefer what’s there to even a 20 storey downscale apartment building.
If we assume say that Melbourne is circular and 100km in diameter, then the ‘inner city’ of 5km radius is 78.5km2 in area, which represents 1% of the total area. So from the report data there is 8% of Melbourne’s pop’ living in 1% of it’s area.
If distribution of population growth were to be pro rata based on area, then the inner city need only take on 1% of growth in order to be doing ‘it’s share’.
@Wiz Aus April 10. I’m pretty sure Mary Drost convenor of Planning Backlash and local resident and would be more than happy to see a block sit vacant than have even a 4 story building on it to be honest… I’ve been in VCAT with almost that exact situation (and suburb) and her group arguing against the development.
The middle ring Council’s are allowed to hide behind excuses like ‘protecting the back streets’ while they aren’t forced to open up there activity centres to increased development providing no ‘give and take’ in the development debate.
Hopefully developments like ‘Ikon’ in Glen Waverley will open developer’s eyes to the latent demand in these areas (particularly in the eastern suburbs as the ‘middle’ north (Coburg/Preston) have many developments of this size already). Council’s eye to the potential for an increased rate base. and Resident’s eyes to the IMPROVED amenity that higher density development can bring to an area.
Improved amenity comes from the simple fact that more people = more demand = more provision of services. If your average suburban shopping strip with a dozen offices/shops closed by 5pm and a fish and chip shop, video store and milk bar open to 9pm with benefit hugely from conversion to 4-8 storey apartment buildings. The extra people would allow more food and services on both the main street and around the corners serving the apartmetn dwellers and benefiting the back street detached house residents as well by providing them with a local (walkable) thai/indian/chinese/pizza etc. restaurant to go with their fish and chip shop.
Krammer – you are right, the trick is to find a way to get Council’s to actually identify and zone these areas for development. I actually think this should be taken out of Council’s hands. A blanket methodology similar to Rob Adam’s could be adopted with Council’s having the right to object to a site specific height limit increase on limited grounds (eg. unrecognised Heritage or sensitive uses, or perhaps ‘consistency’ of streetscape). This might lead to a rolling height limit along a particular road, with one block of largely intact heritage buildings being height limited to the highest parapet of any of those heritage buildings, then the next block of 70′s warehouses being stepped up to a maximum of 6 levels, and then stepped down to 3 stories for a largely detached residential precinct etc.
A cautionary note on apples and statistical oranges. The inner Brisbane area in the ABS data is only 28 sq km – about 3km radius. Quite different to the other three. And that makes the effect of the river, railyards at Bowen Hills, Kelvin Grove Golf Course, and various other large chunks of educational and parkland usage more significant.
Popn density is about the same as the inner Melbourne district. And that 3km radius doesn’t look suburban any more. Dramatically changed landscape over the last decade, with a lot of high and medium rise apartment living.
Richard: Thanks, that’ll explain Inner Brisbane’s 4.7% in 2011. The City of Sydney is also about 25 sq km and it had 4% of metro popln in 2011, so Brisbane is at least on par. Do you know why ABS uses a smaller area for Brisbane than other cities? Does beg the question of whether the concept of the “inner city” can be the same absolute area for cities of very different sizes e.g. Sydney 4.6 million vs Brisbane 2.1 million.
BTW Melbourne has very substantial non-residential uses within 3 km too e.g. industrial and port uses within City of Melbourne. I can’t agree however that Inner Brisbane “doesn’t look suburban anymore”. Yes, there’s been enormous redevelopment (e.g Newstead) but on a relatively small footprint – areas like Hill End, West End, Highgate Hill, Milton, Rosalie and even New Farm (where I used to live and visited a year ago) are still predominantly detached housing. Would be useful to check the Census for what % of popln in Inner Brisbane live in multi unit housing. AD
I would also be cautious with the ABS population distributions generally. They now use “mesh blocks” as the lowest spatial unit in which they define population groups. These don’t match actual blocks or follow any particular logic other than past historical boundaries within SLAs or CDs. I note the ABS publication linked to refers to “inner city LGAs”, not any specific distance. I don’t really understand how they could come up with an accurate figure for all those within a (say) 5km radius. So there is really no way to compare Australian cities against this issue. Aneccdotally however, I think ther ehas beena dramatic change in inner residential density in the past ten years, except here in Adelaide (because of our great planning!).
Statistics aside, I personally think the perception that many suburban residents are against higher densities around them is true, but is a product of the way the issue has been (badly) handled in the past. Many cities with higher overall population density and high quality of life feature medium density (townshouses 2 to 5 stories) with plenty of parks, not high density. Examples like Geneva, Vienna, Vancouver and Munich come to mind, with hardly a building in sight above 5 stories and many beautiful gardens. I think 20 story residential towers are inappropriate except near PT nodes (TODs). Yet developers and government planners often imagine them as the “solution” for economic reasons. We really don’t seem to be very good at understanding what causes high amenity in cities. The proposals that destroy amenity only make people more reluctant to consider a level of change that might be practical.
Hands up everyone who wants higher densities and reduction of amenity thrust upon them so developers can line their pockets?
How many people is enough? What is the end game – 35M, 50M, 100M?
Population densities – how can you plan your route when you don’t know where you are going?
Please login below to comment, OR simply register here :
Thank you for registering, we have just sent you a confirmation email, which includes your new password to be entered below.