Is obesity really caused by suburban sprawl?

Suburban sprawl is often linked with rising obesity – for example, see this submission to last year’s Urban Growth Boundary Review from Kelvin Thompson, Labor Member for the Federal seat of Wills, or this article in the Sydney Morning Herald.

The customary argument is that because the incidence of obesity is lower in the inner city where densities are higher, it follows that low density outer suburban development is the cause, or at least a very significant contributor, to obesity.

At first glance this seems to make some sense. For example, only 1.1% of workers in Melbourne’s outer suburbs walk to work, compared to 12.9% in the inner city.

But for all its faults, is it reasonable to put the blame for obesity on sprawl?  No, it isn’t reasonable. We would we better off focusing our energies on the real issues associated with sprawl rather than being distracted by sideshows.

The key reason is that what goes in our mouths is more important than how much we exercise. You have to walk the dog for an hour and a half, or cycle for an hour, to burn off the calories in just one Big Mac.

The inner city has a lower incidence of obesity primarily because the residents eat better. And they don’t eat better because of higher density but because they have higher incomes than residents of the outer suburbs and, importantly, higher levels of education. They are more likely to know about the importance of good eating and they are more likely to be able to afford to eat better food. They also have smaller households on average so it’s easier to cook healthy food at home rather than go out for fast food.

Possibly most important of all, they are much more likely to be single and young, with a strong incentive to manage their food intake carefully in order to look good.

This misunderstanding of the demographics of the inner city comes up in other contexts, too. For example, the fact that inner city residents have a higher per capita ecological footprint than suburban residents has been used to condemn density when in fact the former’s environmental profligacy is due to their higher incomes – richer people tend to consume more of almost everything. The correct interpretation is that inner city households have an appalling environmental performance despite living in smaller houses – it would be worse, given their smaller household size, if they lived in McMansions.

Calories out is the other key variable in the obesity equation. We know inner city residents walk more than outer suburban residents but much of the reason for that, as I argued here, is because they live close to the enormous concentration of jobs and attractions in the CBD and inner city, rather than because of population density. As soon as you move  from the inner city (about a 5 km radius) to the inner suburbs (about 5-10 km radius), the proportion of workers who walk to work drops from 12.9% to 2.1%.

Replicating the inner city’s density of destinations in the suburbs in a way that significantly affects walking levels seems unlikely (let’s not forget that less than 10% of Melbourne’s population lives within 5 km of the CBD and that they have 28% of the metropolitan area’s jobs).

But it’s going to be hard to offset that Big Mac just by walking to your local pub or restaurant. I’m not aware of any data that suggests outer suburban residents play significantly less sport or undertake less formal exercise than their inner city counterparts. In fact outer suburban workers are more likely to have a job that involves physical effort, like a trade, than inner city professionals. If there are significant differences in per capita exercise levels, I suggest it would have more to do with demographic factors than with density.

Densities are rising in new developments in the outer suburbs, so can we expect a fall in obesity? I doubt it. It’s highly unlikely the new residents will have jobs within walking distance because the great majority of Melbourne’s jobs are dispersed across the suburbs at low density.

There’s a quaint notion they’ll walk to the corner store rather than drive to the district shopping centre. Structure plans accordingly make provision for convenience stores – for example, the draft structure plan for Toolern, Melton, has three plus a large activity centre. I don’t buy that.

Why would households burdened by mortgage payments and childcare costs forgo the cheaper prices available by driving to bigger centres, in favour of the higher cost at a local convenience store? And what does “convenience” really mean anymore? The use-by date on fresh milk at my Safeway is around 10 days and UHT milk is ubiquitous. There’s obviously a market for stores like 7-11, but last time I looked mine didn’t sell fresh vegetables or fruit, the very things that do go off quickly. Now that smoking is out of favour, I don’t see much hope for small walking-based convenience stores in new estates.

I think many more kids could walk or cycle to school both in the outer suburbs and everywhere else. But the key constraints there aren’t housing density (although I do wonder if school density is lower in the outer suburbs?). They’re traffic and over-wrought perceptions of “stranger danger”.

All in all, the current levels of walking (2.1%) and cycling (4.5%) to work that apply to the inner suburbs seem like a reasonable, but ambitious, target for the outer suburbs, given that job density is higher in the inner suburbs.

As I’ve argued elsewhere, sprawl appears to have only a small independent effect on obesity. A recent study by Zhenxiang Zhao of the University of Illinois, Effects of Urban Sprawl on Obesity, examined data on 53 large metropolitan areas using a methodology designed to minimise selection effects. The author concludes that “overall, my results suggest that urban sprawl did cause an increase in obesity, but its effect was relatively modest….a 1% decrease in the proportion of the population living in dense areas increased the prevalence of obesity by 0.1% to 0.2%”.

Sprawl has it’s downsides but over-egging the pudding doesn’t help rational debate.


11 Comments on “Is obesity really caused by suburban sprawl?”

  1. Zuko says:

    Thompson does not make the obesity/sprawl nexus in that document.

    • Alan Davies says:

      See pages 14, 16 and 19. On page 19 he says “Obesity, congestion and air pollution are all anchors that are weighing our economy down. Expanding the Urban Growth Boundary will add further shackles to achieving economic growth in these uncertain times by fuelling each of these issues”.

      Yet I’ve been a bit hasty. A closer reading of the article in The Age, Urban sprawl – Melbourne growing obese: Labor MP, indicates Thomson was more focused in that specific article in likening Melbourne to an obese man “who thinks he can solve the problem by loosening his belt”. I’ve therefore removed the reference to The Age. Thanks for the feedback.

  2. Michael says:

    I feel like there is a connection between the lack of time people have to prepare food and the over reliance on cars with obesity and urban sprawl, although blaming obesity on sprawl is way too simplistic. If people do spend longer in passive modes of transport, and you can’t get much more passive than door to door driving then that’s going to take time away from exercise and food preparation. If suburbs where designed to encourage cycling then that would at least give people an option, but I hold out zero hope that this will occur without a serious increase in the cost of petrol. Until then you will only get dabbling.
    The US academic Elizabeth Warren from Harvard identified the relationship between increased private debt, access to high value infrastructure such as good schools and car ownership in her lecture “The coming collapse of the middle class” . She compares a typical family from the 70’s with a contemporary family and she finds that apart from increased costs of housing (driven up by limited supply of good schools etc.), the only other area where costs for families had risen was the move from single car families to two car families.

  3. Russ says:

    Alan, do you know of any good statistical studies looking at the potential for walking in Melbourne? I say this because the state government “Walking and Cycling: Census analysis” did an adequate job discussing the demographics of existing walkers, but failed to address in any way the question of “who might be able to walk, but doesn’t? (and therefore why not?)”

    That is the question that actually matters. A comparison of walking rates between the inner and outer suburbs is of very limited use, because as you state, demographic factors affect are the primary reason for the difference. But it is also true that if (making figures up here) 80% of the people in inner suburbs who live within 1-2km of their workplace walk, and 20% of people in the outer suburbs who live within 1-2km of their suburbs walk, then there is probably a lot of improvement that could be made to the urban environment in the outer-suburbs. If the rate is the same, then it may be a waste of time focusing on environmental factors. Similarly for the distance walked (or cycled), or for any gender differences in the rates amongst the “potential” group.

    It just seems a lot of debates over these modes bounce between advocates stating that rates are very low compared to other [country, city, suburb] and detractors responding, “yes, but…”, without ever getting anywhere.

    • Alan Davies says:

      No, I don’t really know of any studies on walking in Melbourne, but I plan to dig up some numbers soonish and post something about walking in the outer suburbs.

      As I said here there are a number of factors that determine why inner city residents walk more than others. Demographics is part of that although I think they are more important in explaining obesity in particular rather than walking in general.

  4. […] I’ve pointed out before, the main culprit isn’t exercise but what goes in our mouths. You have to walk the dog for an […]

  5. […] angle (an issue I’ve discussed specifically in the context of obesity before – here, here and here). And as always, correlation doesn’t mean […]

  6. […] planners is too limited to have a major effect, as I’ve discussed in various contexts before (see here, here, here, here, here, here, here and here). These are not fundamentally issues of land use […]

  7. […] planners is too limited to have a major effect, as I’ve discussed in various contexts before (see here, here, here, here, here, here, here and here). These are not fundamentally issues of land use […]

  8. […] and health outcomes like obesity is fraught, as I’ve explained before (e.g. see here, here, here and here). The Victorian Legislative Council’s Environment and Planning References […]

  9. […] policy. These are issues for another day, although I’ve touched on them before (e.g. here, here, and […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s