Is inner city living the solution to obesity?

How much exercise is needed to burn off food?

It’s often pointed out that residents of the inner city, on average, are less obese than residents of the outer suburbs. Since the inner city is denser, more walkable and has much better public transport access than any other part of the metropolitan area, the conclusion seems obvious to many – a key strategy to address obesity should be to encourage higher dwelling densities and better public transport in the suburbs, especially the newer, fringe areas.

The flaw in this thinking is it fails to observe that the inner city – defined roughly as the area within 5 km of the CBD – is a different world. Relative to the suburbs, the inner city has an emphatic over-representation of younger, well educated and affluent residents with fewer dependents. The proportion of the population made up of young singles is three times that of the metropolitan area as a whole and there are twice as many young couples without children.

These are the sorts of people who on average are slimmer because they’re younger, who are of an age where appearance is enormously important, and who are well educated enough to know about nutrition and eschew fast food. They can afford to buy high quality fruit and vegetables and pay for gym memberships. Because they’re more affluent, they have fewer children on average and hence less need for a car.

They live in smaller dwellings so they can be near the CBD and take advantage of its enormous and unparalleled concentration of high-paying professional jobs, its matchless endowment of cultural attractions and its huge and diverse range of social and entertainment opportunities. There’s no other concentration of activity within the metropolitan area that comes even close to the richness of what the inner city offers.

Because they live at higher density, driving is too hard for many trips – roads are congested and parking costs range from expensive to impossible. So residents often walk or use public transport instead. That’s O.K., because they happen to live in that transit-rich, small and unique geographical area where every train line and tram line in the entire metropolitan area – the result of 130 years of construction and at least one spectacular land boom – converges.

So population density and access to public transport are not the underlying forces driving this group’s superior average BMI. Rather, it’s a combination of the small but highly specialised group who can afford to live there, on the one hand, and the special characteristics of the area, particularly the presence of the CBD, on the other.

It’s pie in the sky to imagine the sheer scale and complexity of the highly specialised attributes offered by the inner city could be replicated in the suburbs – much less the outer suburbs – within the foreseeable future. The inner city is focussed on the CBD and in almost every city in the world, the number of jobs in the city centre is an order of magnitude larger than any suburban centre (Atlanta is possibly the sole exception). In Australia, the centre offers the cream of corporate jobs.

The importance of proximity to the CBD in explaining the special character of the inner city is demonstrated by the fact walking’s share of work trips plummets from 13% in the inner city to just 2% immediately one locates in the adjacent inner suburbs. This share is only marginally better than the outer suburbs.

Will building at higher densities and providing better public transport in the outer suburbs significantly lower the incidence of obesity? Not likely. Even if all outer suburban dwellings were townhouses, the incentive to walk is much lower if there’s no CBD, cultural precinct, river, beach, historic buildings, hundreds of cafes, and hundreds of thousands of jobs to walk to. Perhaps most importantly, the outer suburbs don’t have the constraints on driving and parking that often make walking or public transport a superior alternative in the inner city.

And outer suburban residents overwhelmingly work at dispersed locations where car travel is a much superior option to any realistically feasible level of public transport service. There are also many more workers in the suburbs who really do need their car for work and who routinely travel to different work places each week e.g. tradies. There are many more families with young children who won’t forgo the convenience of a car for tasks like shopping. There’re probably more limited household budgets too, making it worthwhile to travel to large discount stores rather than shop within walking distance of home. And then there are differences in socioeconomic status which could mean there are more people who will eat fast food no matter what mode of travel they use.

I don’t doubt that any individual outer suburban resident who chose to take public transport over driving would get more exercise, presumably with consequent health benefits. But that’s a personal choice for individual travellers. From a policy perspective we should be concerned about how many residents are likely to make this choice and what ratio of benefits to costs it will yield relative to other non-transport options. To do that we need to understand the underlying drivers of obesity and not be taken in by the symptoms.

I suspect it’s far more cost-effective to focus on what goes in our mouths – after all, it takes an hour and a half of walking to burn off the calories from one Big Mac. I also wonder what the risks are in making health a headline objective of transport policy. These are issues for another day, although I’ve touched on them before (e.g. here, here, and here).


11 Comments on “Is inner city living the solution to obesity?”

  1. Michael says:

    It seems to me that once you plan suburbs for cars then getting around anywhere on foot is going to become an inconvenience. It would be interesting to know if there was any suburban developments created after the eighties that have high pedestrian numbers. Cars and low density mean that you don’t need to do any business in your local area and the consequence for much of Melbourne is that there is very little to walk to so you might as well drive, further reducing the amenities of the suburbs for pedestrians. Maybe we should follow the US and get rid of footpaths in areas where no one walks. Even inner city areas of Melbourne are unpleasant to walk around because of high car congestion and pollution.

    • Precisely.

      There are urban forms that will promote more walking, cycling, etc without trying to turn the outer suburbs into the inner city. These urban forms are not going to solve obesity in a single stroke, but they may promote healthier lifestyles, as well as other benefits.

    • Simon says:

      For post-80s walkable and bikable suburban developments, it might be interesting to look at the towns and cities of the netherlands.

    • wizofaus says:

      Much as I’m very much an advocate for trying to increase the supply of higher-density mixed-used zones in cities, I’d tend to agree that hoping it’s likely to measurably reduce obesity on its own is probably wishful thinking. However I also think that to suggest that immediate proximity to the CBD is the many reason people walk/cycle more is not entirely fair – I live nearly 10km from the CBD but you only need to spend 5 minutes in the neighbourhood and 5 minutes, in say, a newer development in, say, Point Cook (~20km from the CBD) to see a world of difference in the modes of transport typically employed. And I’ve been fortunate enough to sample various cities throughout the world and seen much the same pattern – higher density mixed-used zones enable a higher variety of options for getting around, and encourage people to rely less on powered means of transportation and the necessary externalities they impose.

      • Alan Davies says:

        Surely we know by now that it’s a fallacy to think that simply increasing density in suburban areas is going to produce significant increases in transit, cycling and walking’s share of travel and correspondingly large decreases in car travel.

        • wizofaus says:

          Sure, increasing density may not be a sufficient condition to notably decrease dependency car travel. But I suspect it’s a fairly necessary one, and in many parts of Melbourne at least (I can’t speak enough for other cities to say for sure), there’s very good reasons to believe it will work. It may not significantly reduce km driven, if the time/money saved from less car dependency and a healthier lifestyle etc. results in people choosing to use their cars for longer “getting away” type trips, but sensible resource pricing will sort that out eventually.

          • Alan Davies says:

            I think the line of causation works in the other direction. If we were to increase the price of car travel significantly wrt other modes, I expect we’d get some increase in density of houses and jobs, some shift to other modes and a big shift to more efficient cars, more efficient car trip planning (e.g. chaining) and some suppression of low value trips.

            Generally, I think the effects of higher residential density are confused with deeper underlying factors (e.g. like proximity to the CBD or a valuable natural feature, peculiar demographics, etc) that are hard to reproduce in other parts of the city. The one instance where density in and of itself really does seem to make a big difference is concentration of jobs. Of course planners need to understand why some firms choose to cluster (and why most prefer not to).

          • wizofaus says:

            Except that density is to a significant degree dependent on government policy. Indeed, a lot of my concerns about suburban development come back to perverse policies that actively prevent development from happening the way it did 100 years ago. Even the fact that the widespread availability of cars makes a difference is only true because of the amount governments invest in attempting to facilitate car travel. Which is not to say I think if governments sat back and did virtually nothing at all things would suddenly get better, but I do sometimes wonder if it might be a good start!

        • wizofaus says:

          Old thread I know, but how do your claims that higher density areas don’t show an obvious reduction in km driven square with diagrams like http://en.wikipedia.org/wiki/File:Revised_petrol_use_urban_density.JPG?

          • Alan Davies says:

            I’m not saying there isn’t a correlation or even that they aren’t linked to some degree, but rather that residential density isn’t usually the major underlying cause of kms driven. This is important because simplistic solutions like increasing density on the fringe of a city like Melbourne, say, won’t produce the sort of mode shift or change in kms driven that advocates of this policy imagine it will. It’s necessary to understand the underlying causes – there are other variables at play.

            The question you raise is addressed more directly in this earlier post. Some more on that graph in the Comments section of this post.


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