Nov 20, 2012

Will compact cities deliver on the environment?

The orthodox view says making cities more compact is essential to improve sustainabity significantly. However new research suggests the environmental pay-off is modest

Alan Davies — Editor of The Urbanist

Alan Davies

Editor of The Urbanist

Alternative spatial designs for urban regions (source: Echinique et al)

A paper published a few months ago in the prestigious Journal of the American Planning Association (JAPA) casts doubt on the strength of the environmental claims made on behalf of compact development.

The key finding – what JAPA calls the key “takeaway for practice” – is:

Urban form policies can have important impacts on local environmental quality, economy, crowding, and social equity, but their influence on energy consumption and land use is very modest; compact development should not automatically be associated with the preferred spatial growth strategy.

The paper, Growing cities sustainably: does urban form really matter?, was written by four academics from Cambridge, Leeds and Newcastle universities (Marcial H. Echenique, Anthony J. Hargreaves, Gordon Mitchell, and Anil Namdeo). As I discuss below, it’s provoked a debate among US planners about the normative role of research.

Using statistical models, the researchers simulated three possible 30-year spatial futures – compaction, expansion, dispersal – for three distinctly different English city-regions. These are Tyne and Wear in NE England, the South-West centred on London, and Cambridge sub-region.

They used 26 environmental, social and economic indicators to compare explicitly the benefits and costs of each of the three options. They conclude that land use and transport policies promoting compact development:

have virtually no impact on the major long-term increases in resource and energy consumption. They generally tend to increase costs and reduce economic competitiveness. The relatively small differences between options are overwhelmed by the impacts of socioeconomic change and population growth.

That’s not a novel finding for studies at the broader metropolitan level. For example, in a study of 31 of the largest US cities, Gary Barnes from the Centre for Transportation Studies at the University of Minnesota found:

Even very large changes in land use have very little impact on travel behaviour, in good ways or in bad. Apparently the larger effects sometimes observed in neighborhood-scale studies are just that: neighbourhood-scale effects that do not extend their benefits to the larger urbanized area.

As I noted when reviewing this study last year, his analysis implies that increasing residential density by 100% would increase transit share by only 5-6%.

To get a 1% increase in walking and cycling’s combined mode share would require an increase in residential density of 5,000 persons/sq mile (1,931/sq km). Similarly, a 14% increase in density would only yield a 0.5% decrease in in-car travel time per person.

Closer to home, Paris Brunton and Ray Brindle examined the effect of density and land use mix on travel behaviour in Melbourne and found only “superficial support for a relationship between density and trip-making.”

They found accessibility to activities has a stronger bearing on travel choice than density. However they concluded that after income is taken into account “urban form characteristics as a whole played a relatively insignificant role in determining car travel in Melbourne.”

I pointed out last month that Sydney is twice as dense as Brisbane (using the superior population-weighted density metric). Yet public transport’s share of all work trips is only five and a half percentage points higher – Sydney’s public transport mode share is 14.8% compared to Brisbane’s 9.3%.

There are of course other reasons why as a society we might want to promote density and limit urban expansion. But when it comes to the specific objectives of reducing emissions, pollution and energy consumption, it appears plausible land use policy is likely in most cases to make only a modest contribution.

That moderate gain needs to be compared against the costs. Echinique et al argue that compact development reduces housing choice and exacerbates congestion. They could’ve added that it’s also a very slow mechanism and politically thorny.

Indeed, while they’re all difficult, there are arguably more politically feasible, more technically effective, and certainly quicker ways of promoting environmental objectives than changing land use.

A key one is converting power generation to non-carbon sources. In the urban sphere, another approach is to implement regulatory and/or taxing policies that lower the demand for driving and provide incentives for the use of more efficient vehicles.

There needs to be more consideration of evidence-based research by those interested in cities. One reason why there isn’t is illustrated by the reaction to the Echinique et al paper by some members of the US Association of Collegiate Schools of Planning (ACSA).

Lisa Schweitzer, an academic at USC’s Price School of Public Policy, reports it was widely discussed on-line by planning academics and was the subject of a seminar earlier this month in LA to consider how ethics, politics and normative values factor into what gets published.

According to Dr Schweitzer’s account, the Chair opened the seminar with a critique of the paper that appeared to boil down to this proposition:

(Planning) practitioners have a tough time convincing people to pursue Smart Growth, as a result, JAPA has no business publishing things that do not reflect practitioner’s goals and values……

The nub of Dr Schweitzer’s response, with which I agree, is that the goal should be to support the achievement of desirable outcomes, not specific solutions. The problem with the sort of critique offered by the Chair:

is it suggest that researchers “owe” it to practitioners to only inquire within the framework that compact development is unambiguously meritorious and sprawl is unambiguously not……

All solutions need to be constantly subjected to testing and analysis, not protected from review. If evidence suggests they’re not achieving desirable outcomes at an appropriate scale, they need to be reappraised.

There’s lots of good things about density, smart growth and compact development, but we should always understand their limits. We need a more informed debate about what favoured policies can really achieve.

(Visited 4 times, 1 visits today)


Leave a comment

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

17 thoughts on “Will compact cities deliver on the environment?

  1. Dylan Nicholson

    Don’t remember sorry – it was an article in the Age from memory though, but I can’t find it from a quick Google.

  2. Alan Davies

    Dylan Nicholson #15:

    I linked to it because it relates to the topic, not as an endorsement. Far as I can see though, you have to look hard to find people who’re aren’t either advocates or aligned with some ideological position. Personally, I find value in all sorts of viewpoints. Wendell Cox maintains a data base of cities which is held in high regard by a range of observers.

    BTW who’s made that argument about radial vs circumferential rail and sprawl? Sounds interesting.

  3. Dylan Nicholson

    I just noticed the author of that article is Wendell Cox, who has long had a thing against what he calls ‘anti-sprawl’ policy. Maybe it’s different in the U.S., but I don’t think I’ve seen a single piece of legislation enacted here that is obviously ‘anti-sprawl’, and plenty that almost certainly encourages it. He’s also a member of the Heartland Institute and various other conversative and firmly right-wing organisations, but apparently seems to believe that sprawl is the natural result of letting people do what they want with minimal legislation, despite plenty of historical evidence to the contrary. E.g. it’s been persuasively argued Melbourne is so sprawled in large part due to the design of our rail network, which stretches out long distances from a central hub, but with almost no inter-connectivity.

  4. Dylan Nicholson

    …oh and on that last point, I disagree that the way to encourage smart growth is to “artificially” restrict growth boundaries. The reality is that whenever a city expands at the edges governments need to be involved to ensure basic infrastructure is supplied to such areas – and my suspicion is that currently in many cases unnecessary sprawl occurs because of over-regulation (or at least, poorly thought out regulation) rather than excessive restrictions. Further, the main barrier to increasing density in existing developed areas is often existing regulations (whether they be heights controls or excessive car parking requirements etc. etc.)

  5. Dylan Nicholson

    Re increased traffic congestion and localized pollution – I’d argue that the problem with low density development is that there’s the same amount (per head of population) of pollution and negative side-effects caused by the use of automobiles and industry, but because they’re far less immediately perceptible, there’s far less incentive to do anything about them.
    Lack of housing affordability is a more interesting problem – but surely primarily high housing prices in densely built-up areas is a direct result of more people *wanting* to live in such areas than housing is available, so therefore the obvious solution is to ensure we continue to develop more such areas.

  6. Alan Davies

    More on the reaction in the US to this paper – Separation of church and urban planning.

  7. Dudley Horscroft

    Re me at 10 – Corrections: Should read “237 buses that use City”, and “at much less cost”.

  8. Dudley Horscroft

    Re Tom at 9, the other major thing holding Sydney back is the excessive use of buses to do a job which should be done by much higher capacity vehicles. Where there is very high demand, eg, Parramatta Road, there are a large number of buses and they get in each other’s way, and produce congestion for themselves and others. There are approximately 237 buses use City Road Junction to Railway Square and then about 322 buses from Town Hall to Bridge Street, in the 2 hours 0700 to 0859. Using large trams (60 m) the 18,000 passengers the buses carry could be carried far faster and at must less cost, and George Street would not be congested. At low demand, buses are best, but at Sydney levels, trams are needed. And if Sydney densifies (horrible word) Waste Connex would not be needed, or even proposed.

  9. Tom the first and best


    As I point out in my comment at #3 Sydney`s public transport usage percentage is actually bad for a city of its density. Non-density factors are holding it back.

    The major things holding it back are; an insufficiently multi-modial ticketing system, overly radil bus network and lack of rail capacity from Central to Epping.

  10. Russ

    Krammer, improving supply-chain efficiency for a single business is going to reduce transport distances (or at least improve transport efficiency) within that part of the supply-chain. Whether it improves transport efficiency across the total length of the supply chain, including the store to customer, is something planners should be researching and acting on – it is a classic market inefficiency problem. I don’t know of any research that looks at the problem though.

    Alan, the issue around academic and practitioners brought to mind an old paper by McLoughlin discussing planning education. He made the point that early planners distinguished the profession from other fields by focusing on “procedure rather than substance, form rather than content”, which diverted attention from research into substantive issues and tended to demand planning schools produce uncritical (if idealistic) bureaucrats. To quote liberally from the conclusion on bringing practice and theory together:

    In the first place, we can have practice defined in a very narrow sense of physical, land-use, statutory planning agencies demanding ‘relevance’ from academia, which in turn meekly and uncritically deliver the goods in the form of uncontroversial bureaucrats. This kind of linkage between practice and academia may well result in very little advancement of knowledge – – certainly critical appraisals of the world of practice are unlikely – – and in such an incestuous, circular and self-justifying world no research is done because no research is needed.

    The other kind of linkage between practice and
    academia is where practice looks to academia at least for specific skills of one kind or another but also for creative constructive ideas. These ideas should arise out of research which is solidly grounded in an empirical but critical assessment of ‘how the city works’, and how planning and other instruments of the state are intervening, with what degrees of success, and with what effects.

    I’d be interested in your perspective on how the planning profession has changed in relation to these issues in the years since the article was written?

  11. Krammer56

    Alan – good to hear some sense being introduced into the denisty debate based on evidence, not dogma.
    The only transport benefit of density is that possible destinations are close by. It doesn’t matter much what the fuel consumption of your car is if you only drive it a few km and non-car alternatives are also more feasible for short trips.
    However, a global density increase isn’t the only way to do this. Reducing the separation of land uses and the scale of developments can achieve similar benefits.
    The common practice that keeps land uses apart increases travel distances. The business mantra about “efficiency” dictates economy of scale arguments for massive shopping centres (another $500m expansion for Chadstone, anyone), massive distribution centres, massive CBD – all requring long travel distances.
    Planning needs to consider challenging this conventional approach, based on considering “whole of society” costs. A more distributed apporach might mean that business costs of a more distributed approach are be a bit higher, but transport and living costs could be much lower. Yet this approach may actually deliver similar transport changes and overall economic consequences will probably balance out.

  12. Alan Davies

    Dylan Nicholson #4:

    I agree density is very important. We just need to be clear in our understanding about what sort of density, where and why it’s important. It’s a much-abused concept.

    Paul Liddell #5:

    I don’t agree. The relative increase is much less important for policy in this context. When the base is small, what counts is the absolute percentage. That’s why I made it perfectly clear I’m talking about “percentage points”.

  13. Paul Liddell

    Interesting that you chose to say that in Sydney public transport share of trips is only 5.5% higher – using the absolute percentage. Really, based on your numbers, Sydney is twice as dense as Brisbane, and has 1.6x the public transport usage. That’s a pretty good result I would think.

  14. Dylan Nicholson

    But Alan surely you’d accept density is a crucial part of what makes a city a city! If we all spread out over the globe at equidistant points (i.e., lived at the lowest possible density), it’s clear there’d be no cities, no trade, no civilisation etc. etc. Local density is arguably the single most important factor in allowing any of those things to occur. Whether there’s any means of determining some sort of ‘optimal’ density I don’t know, or frankly, don’t really care, because it’s something that should by and large sort it itself out based on the choices of individuals, providing there’s no unreasonable impediments to making those choices. Cities in 3rd world countries with virtually no planning laws/regulation at all become very dense indeed, just because it makes sense for the bulk of the people living there – and of course act as a handy warning of what to expect if we were ever to swap over-regulation for under-regulation (though of course there’s a lot more to it than that).

  15. Tom the first and best

    Again, I have to point out that factors other than density that effect the level of PT usage.

    Sydney has a ticketing system that reduces PT patronage by penalising transfers and this in turn causes a more CBD(s) centric bus system which penalises other trips further reducing PT patronage. If Sydney had a ticketing system like Melbourne or Perth with a consequently less radial bus system then it would have higher PT usage. Melbourne, Brisbane or Perth at Sydney`s density would have a PT usage higher than Sydney does now.

    Sydney`s PT usage could be increased by addinmg an extra pair of tracks from Central to Chatswood and the Epping-Parramatta link allowing more trains to run.

  16. Alan Davies

    Dylan Nicholson #1:

    For some industries, the benefits of density are outweighed by increases in factors like accommodation costs and congestion. The link I provided takes you to the paper.

    There’s lots of bunk talked about density. For the great bulk of firms, it’s the fact of being in a city that matters, not the local density. Very high localised densities only provide a net benefit for some firms in some sectors. That’s why, for example, Melbourne’s CBD incl Docklands and Southbank only has circa 15% of all metro jobs and why the sectoral composition of CBD employment differs so much from that of the metro area.

  17. Dylan Nicholson

    I’m intrigued how on earth increased density could “reduce economic competitiveness”.

    I’d think anybody who’s ever given more than passing thought to and done minimal reading on the topic would accept that increasing density has costs as well as benefits – my main issue is that I believe it is something that does and should be allowed to occur naturally (albeit in a well managed fashion), and much of our cities are “unnaturally” spread out due to deliberate polices to keep them that way.

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