Growth Areas

Mar 5, 2018

What’s wrong (and right) with “suburban sprawl”?

Our major cities require a more sophisticated understanding of the role of fringe development in accommodating forecast growth than reflexively dismissing it out of hand as "sprawl"

Alan Davies — Editor of The Urbanist

Alan Davies

Editor of The Urbanist

New townhouses in Mernda, 26 km from the CBD in outer Melbourne’s North Growth Corridor

One of the essential tropes of contemporary planning is that expansion of the urban fringe is irredeemably bad. Here’s a report published in Domain last week on a new book,  Urban Choreography, Central Melbourne 1985, co-edited by City of Melbourne design and projects director Rob Adams (Cut back on cars, stem urban sprawl to improve Melbourne and its suburbs, authors say):

Fewer cars and an end to urban sprawl is the way forward for Melbourne and its suburbs, according to the authors of a new book taking stock of the city’s planning past… (The book) pinpoints areas for improvement, and for Mr Adams, the most important is stopping the spread of metropolitan Melbourne. “If I only had one wish, it would be that we don’t subdivide another single piece of farmland on the periphery of the city,” he says.

The go-to prescription to limit sprawl is to use tools like the Urban Growth Boundary to contain all or most growth within the existing urban fabric. It’s not easy though. The Urban Growth Boundary established in 2002 in the Bracks’ Government’s strategic plan, Melbourne 2030, sought to restrict the fringe Growth Areas to circa 30% of population growth; however in the face of falling housing affordability it was subsequently expanded a number of times, including by a whopping 430 sq km in 2010.

The Brumby Government updated Melbourne 2030 in 2008 with a new document, [email protected] Million, which recognised that the Growth Areas were accounting for about half of all growth. More recently, the Andrews Government’s “refreshed” urban strategy, Plan Melbourne, reinstated the 30% target.

What’s wrong with fringe growth? The key objection relates to the relatively low densities of new suburbs compared to the alternative of multi-unit developments within the existing urbanised area.

The main charges levelled against “sprawl” include:

  • High levels of car use (“car dependence”)
  • Inadequate public transport
  • Long trip distances and travel times, particularly for the journey to work
  • High cost of providing infrastructure headworks
  • Lengthy delays until essential community services “catch up” with population growth
  • High incidence of physical and mental health problems
  • Sterilisation of farmland and pollution of bushland
  • Large dwellings with high embodied and operating resource use, especially electricity.

But fringe development also offers advantages. After all, every city in history expanded at the periphery and, absent some insurmountable barrier like an ocean, they still do (see Why do cities still sprawl?). When affordability improves, the fringe tends to capture a higher share of population growth. Targets like 30% are based on periods of poor affordability when low to average income households struggle to buy a home.

The principal advantages claimed for fringe development include:

  • Low land costs per sq m
  • Low building costs per sq m
  • More indoor and outdoor living space, making it better suited to larger households
  • More privacy, both within and between dwellings
  • Greater autonomy e.g. no body corporate
  • Greater accessibility by car to suburban jobs and services
  • Greater scope for greening streets
  • Lower cost of providing infrastructure
  • Better infrastructure e.g. larger school grounds
  • Protects established suburbs from redevelopment

What’s instructive here is most of the claimed positives are private benefits, whereas most of the alleged shortcomings are social costs. Given prevailing prices (including implicit subsidies), households who want affordable housing are voting with their feet (or yes, their cars). Many of them would prefer a location closer to the city centre, but the refusal of governments and existing residents to increase supply of new multi-unit housing in established areas rules that out.

The fringe doesn’t provide all the benefits of the established suburbs, but then it costs a lot less, has other advantages, and for many households still represents their optimal trade-off. Are the social costs of providing affordable housing on the fringe acceptable? I long ago came to the conclusion that the problems associated with it are exaggerated (see Problems with fringe-dwelling are peripheral):

  • The impact of Melbourne’s expansion on productive agriculture is extremely small. The area of land used for agriculture in Melbourne’s vaunted Green Wedges comprises just 1.7% of all agricultural land in Victoria. The wedges are mostly made up of extensive areas of protected natural bushland, semi-rural uses like hobby and lifestyle “farms”; and urban uses like airports, sewage works, prisons, sporting facilities, quarries and more (see Is sprawl a serious threat to food security?)
  • The average journey to work trip time is much the same in Melbourne’s outer ring suburbs (38 minutes) as it is in the middle ring suburbs (37 minutes) and inner ring suburbs (37 minutes). The average trip time for all purposes is also much the same i.e. 22-23 minutes (see How big is the “transport divide” between inner and outer suburbs?). The claim of “forced car ownership” is overstated (see Transport disadvantage in the suburbs).
  • Residents of fringe Growth Areas commute by car because they mostly work in the suburbs and because most jobs are in the suburbs. While most recent jobs growth was in the centre, the effect on the spatial distribution of jobs across the metropolitan area was tiny. Half of all Melbourne’s jobs are more than 13 km – and 70% more than 5 km – from the CBD (see The jobs are already in the suburbs). Four-fifths of suburban jobs aren’t in major suburban centres; they’re in dispersed locations which are much easier to access by car than by public transport (note that even for residents of Melbourne’s relatively accessible inner suburbs, the average commute takes 40% longer by public transport than by car).
  • There’s no reliable evidence that it costs significantly more on average to provide infrastructure on the fringe than it does elsewhere (see Does infrastructure cost a lot more on the fringe?). It’s likely though that the idea established suburbs have money-saving “spare capacity” doesn’t apply to all types of infrastructure (e.g. see Is unused infrastructure capacity in the inner suburbs all used up?).
  • Community services have always lagged development of new areas and always will. It happens in redevelopment areas too, as well as with services provided by the private sector. It’s largely a question of scale. I’m not aware of any research on how it’s changed over time, but my sense is new estates are vastly better serviced in the early years nowadays than they were in the past (see Are the politicians trying to con us on this one? and Is the planning of new fringe suburbs getting better?).
  • Detached dwellings in Growth Areas are larger than multi-unit housing in established suburbs, but their thermal efficiency has increased significantly over the last few decades. Moreover, the average Growth Area dwelling is much smaller than the term “McMansion” implies (see Why are new suburban houses so bloody big? and Are huge homes irresponsible?).
  • The claim that fringe development leads to poorer health outcomes – especially obesity – than redevelopment is perhaps the weakest criticism. It overlooks demographic and class-based explanations, self-selection, and the astounding increase in calorie density of affordable foods over the last 40 years (see Is it healthy to assume correlation means causation?, Does urban sprawl really make us fat?, Does living on the fringe make people sick?, and Is better health a key rationale for urban policy?.
  • Growth area structure planning today is a far cry from the stereotype of outer suburban wastelands. For example, the structure plan for Rockbank in Melbourne’s west envisages the great bulk of lots will be relatively evenly distributed between 325 and 525 sq m; average density will be the same as some of Melbourne’s old inner suburbs; and the share of lots expected to be used for detached housing is similar to the proportion in inner suburban Coburg (see Is the planning of new fringe suburbs getting better? and How dense are Melbourne’s outer suburbs?).

Of course, there are downsides to peripheral development just as there are to redevelopment in established areas, or to channelling growth to dormitory suburbs in regional centres (see Suburban sprawl or regional sprawl?). For example, there are high levels of car travel (kms) on the fringe; but there are other ways of tackling the associated impacts – like mandating cleaner vehicles – rather than demonising fringe growth. The key issue with “sprawl” is to evaluate the private and social costs in the context of the corresponding benefits.

Australia’s big cities need to accommodate immigration-driven growth both within the built-up area and, given the failure to significantly increase supply in existing suburbs and to address taxation policy, on the periphery. The consequences of implementing the wrong policies and getting the balance wrong are severe e.g. exacerbating chronically poor housing affordability.


Leave a comment

Your email address will not be published. Required fields are marked *

13 thoughts on “What’s wrong (and right) with “suburban sprawl”?

  1. Adam Ford

    This is excellent. When you are DATA-driven, you invariably offer the right perspective.
    Airport rail is the wrong priority. Fantasies about accomodating 5 million people EQUITABLY within the current city footprint NEED to be gainsaid, because of what we are doing to the prospect of home ownership for people on below average incomes.

    I would really prefer we were having a debate on how to build the new growth estates in a way that creates much better communities than we did when planning Pakenham/Cranbourne.

    The Regional Rail link was actually about URBAN policy, it was about offering greenfields estates that could be properly planned as walkable cities based around heavy rail.

    UNFORTUNATELY, go out to Tarneit and have a look at how little planning towards that ideal is taking place, and at THIS point I throw up my hands in despair. But what is needed now is voices crying out for the proper design and planning to be happening NOW, all the people complaining about suburban fringe development (and Rob Adams is guilty as all get out – his version of future melbourne is a complete fantasy that all Melbourne should be like the inner city that he loves so much, and without any thought for the people who are shut out) are dialling themselves out of the real advocacy that urbanists should be urgently taking the most interest in.

    Great article.

    1. Tom the first and best

      The urban growth boundary restricts growth in house availability (either owner-occupier or rental) not home ownership. Flats can be and in many cases are owner-occupier. Fringe estate houses can be bought by investors just as easily as infill flats.

      The declining rates of home ownership, among independent households, are more a result of tax breaks for landlords (especially negative gearing) and the decline in stable mortgage-friendly employment available to young people and lower income groups of all ages.

      Scrapping negative gearing, especially on existing properties, is the best way to increase home ownership, followed by heavier labour market regulation in favour of permanent employment.

      Rail access for the soon to be suburbs through which the line passed was one of the major reasons, but not the only one, for the RRL. If it have been entirely about serving suburbs it would have been a suburban only electrified service. It was built a dual purpose line and is suffering today because of it.

      All of Melbourne being like the inner-city would make the inner-city less expensive, not more, because it would make the things that people value about the inner-city more available. The inner-city is expensive because of scarcity, not its its urban structure.

  2. Tom the first and best

    Not all existing urban area densification requires body corporates/owners-coperations. Any development with individually access dwellings (independently accessing the street) and no fully shared driveways does not need an owners-corperation. Party wall agreements and easements reduce the need for owners-corperations.

    Fringe development being without body-corporates/owners coperations is not universal. Some places have fringe development with owners corporations. Fringe development owners-corperations are common in the USA due to various factors. Gated communities are also likely owners-corperation heavy (unless they are single landlord owner).

  3. Horst (Oz) Kayak

    There is evidence that Nature contributes in improving the health of a population. Which urban morphology allows residents to experience more potential exposure time to Nature? Advocacy groups such as the Town and Country Planning Association (TCPA) would argue that more green space per resident is required within walking distance of the increasingly denser inner urban metropolis.
    The health impact of urban sprawl or otherwise is more than just about the personal inactive time spent in sedentary travel.

  4. Joe Flood

    So-called sprawl has always been misinterpreted and frequently misunderstood, based on a number of weak arguments as the author explains. On the one hand -as long as jobs decentralise at the same rate as population then the extra travel time associated with urban expansion can be minimised. At another level, inner city households are heavier energy users on average than suburban households, through their heavier use of power for a range of amenities which suburban households avoid. The arable land argument is also weak – as well as being barely significant; it conceals the reality that suburban households find productive uses for their land through gardens and growing food; so that suburban land actually supports more biodiversity than monocultures..

    The real problem however is that the restriction of urban boundaries through green belts substantially increases the price of housing – typically doubling the price of housing within the belt. Increasing density pushes up prices for everyone, invariably resulting in greater inequality.

    1. Alan Davies

      Joe Flood

      Good points. See Will Sydney and Melbourne implode as they get bigger and bigger? on how increasing the population/employment size of a city increases accessibility to jobs by much more than the increase in commuting time. See Are inner city residents bad for the environment? on how resource consumption is strongly related to income.

    2. Tony Morton

      It’s tempting to think that travel distances and energy consumption must be lower if jobs and population decentralise together, but it’s never worked out that way in practice. What happens instead in cities like Houston or LA is that you replace a predominantly radial pattern of scale-efficient transport corridors with a random network where homes and workplaces bear no relation to each other. Travel distances and times turn out about the same in one as the other, but in the random network you lose all the returns to scale, so energy consumption is vastly greater. The lack of scale efficiency also counts against the effectiveness of public transport, so another result is much greater car use and road congestion.
      I think you may also be mistaken to attribute higher energy consumption by inner-city households to location-specific amenities. It was the ACF Consumption Atlas that came up with this conclusion some years ago, and it turns out they got there using input-output analysis, which almost by definition attributes higher consumption to wealthier consumers. Inner-city households consume more because they’re wealthier, not because of where they’re located.
      Care is needed with the effect of ‘green belts’ too. When the first Melbourne Urban Growth Boundary was introduced in 2002, house prices had been rising strongly for all of the previous five years. In the four years from 2002 to 2006, Melbourne house prices barely increased above inflation. In practice, growth boundaries are careful to include generous provision for fringe growth so as not to constrain the market – they’re more about where the ‘sprawl’ goes than about how much there is. London’s green belt is the most celebrated, but that went hand in hand with development of the ‘New Towns’ further out. I tend to favour green wedges rather than belts, because they allow for contiguous development along corridors complemented by green space nearby.

  5. Tony Morton

    Alan – if there’s one thing that has changed in the transport policy debate in the last five years, it’s the way the whole transport planning narrative has been reframed around population growth and migration. Completely dominating the public mindset about urban transport and land use now is the idea that roads are so much more congested and trains so much more overcrowded than a decade or two ago, and that this reveals the folly of the record rate at which we’re adding to our national population.

    There’s some truth in this – our population growth rate does appear to be unprecedented in absolute terms, though I’m unsure how it compares in relative terms to the postwar era. But the fact this has come to so dominate thinking about transport planning indicates that we can’t be so complacent about maintaining the status quo in our cities.

    People are not the least bit relaxed and comfortable about leading car-dependent lifestyles in the suburbs of capital cities; it’s just become easier to blame immigration for the consequences of car dependence than to question governments’ failure to rebalance investment and planning effort away from private car transport.

    Our 1920s ancestors could afford to provide public transport service to all new suburbs. Now, we just make excuses not to. The “four-fifths of suburban jobs are outside major suburban centres” claim simply says that most of our suburban employment isn’t concentrated the way it is in the CBD. But that says nothing about how easy access is, or could be, by public transport. Our largest suburban job concentrations are around Clayton / Monash Uni and Melbourne Airport, which are notoriously difficult to access by public transport (though Monash has become easier of late). Meanwhile, the neighbourhood shops adjacent to almost all suburban railway stations are for the most part classified as “dispersed suburban” locations and their potential function as convenient public transport hubs ignored.

    I think the major drawback with the continuing outward expansion of urban development is the way it restricts accessible employment opportunities relative to infill development. There may indeed be plenty of jobs available to someone in an edge-of-city location without contemplating a long commute, but the range is going to be substantially restricted compared with a more central location.

    1. Alan Davies

      Tony Morton

      There may indeed be plenty of jobs available to someone in an edge-of-city location without contemplating a long commute, but the range is going to be substantially restricted compared with a more central location.

      It’s true the range will be restricted compared to a more central location because in every city in the world the average job is closer to the CBD than the domicile of the average worker. Two things though: (a) those who choose to settle in the outer suburbs are more likely to work in the suburbs than in a more central location i.e. it’s a selection effect, (b) most of these settlers also see other advantages in an outer suburban location that makes any compromise in job accessibility worth it i.e. it’s a trade-off.

    2. Alan Davies

      Tony Morton

      Our 1920s ancestors could afford to provide public transport service to all new suburbs.

      The key point in the 1920s was that cars were very expensive so most travellers used public transport; cars didn’t really become a mass market phenomenon in Australia until after WW2. That made it much more likely governments would spend on public transport. The situation is very different today, with circa 90% of motorised travel in Melbourne by car.

      1. Tony Morton

        Quite so Alan – car ownership was much lower in the 1920s so in a sense governments had no choice but to fund adequate public transport.
        Governments today do of course have that choice, which makes it more of a value judgment. If you think public transport is inherently bad, then of course you’ll fund it to the absolute minimum possible. But if you think public transport can provide more public benefit than car travel in a large city, that should incline you to fund public transport generously even if a lot of people drive cars. Cities like Vancouver and Zurich have made that choice quite consciously.
        Anyway, my point was that value judgments aside, the Victorian government in the early 20th century found that well-provisioned public transport was quite affordable. That’s despite the fact that governments’ capacity to raise revenue is much greater now than 100 years ago.

        1. Alan Davies

          Tony Morton

          I think there was more to it than governments having no choice: (a) by the 1920s most of the rail and tram lines were already built (electrification was a major investment but would’ve paid its way) (b) load factors were high and so therefore was revenue relative to costs.

          Nowadays, urban public transport gets a large share of government funding for both capital and operating costs relative to its modest mode share. Under the former Labor federal government, funding for both modes was running neck and neck. Note roads are also used for urban freight distribution and buses/trams. There’s a good case for more government funding for public transport and for increasing taxation of car travel, although of course that’s also true of lots of other areas that’re competing for funding e.g. education, health.

          1. Tony Morton

            That’s quite true: until the postwar era, public transport systems generally expected to recover their costs from user fares. In this sense, the idea that public transport networks should be funded similarly to roads as public goods is relatively recent. As of course is the idea that all motorised modes of transport should be subject to transparent user charges to recover their costs.

            I think there’s a real challenge these days to the idea that recurrent costs for roads are minimal. We’ve previously been able to push the long-term maintenance issue to one side because our road network is for the most part less than half a century old. For the rail system in Victoria, we have notional Capital Asset Charges that are supposed to recognise the holding costs of rail assets and the need for ongoing maintenance. (Inter alia, it’s curious that John Stanley doesn’t seem to know about this item that weighs down ‘Output Costs’ in Budget papers.) But all our roads and bridges are subject to long term wear and tear, pretty much all of which happens off-budget.

            In the USA the road maintenance backlog is now a serious issue affecting Federal budgets, thanks to a lot of their highway construction having taken place in the 1950s and 60s. It’s only a matter of time before we face this in Australia – it’s already an issue for most of our rural councils but this still goes largely under the radar.

            Imagine for a minute that every time a State government or local council gazetted a new road or railway, it was made subject to an annual charge to provide for ongoing maintenance. On the rail side, this could replace of the much more nebulous Capital Asset Charge. Such charges would accrue in the budget and be offset against actual maintenance expenditure.

            Such a measure would doubtless transform the budgeting approach to maintenance and operating expenditure, for roads in particular. It would also no doubt have consequences for project assessment. The biggest advantage though is that operating authorities and councils would no longer always be on the back foot financially when it comes to sustaining the public’s asset base.

Share this article with a friend

Just fill out the fields below and we'll send your friend a link to this article along with a message from you.

Your details

Your friend's details