It’s always fun having debates about city densities and public transport. A post by Jarrett Walker at makes an excellent contribution to the debate – as we seek to answer the age old question of “does density matter when it comes to the viability of public transport?” As I noted in a blog post a while back, there’s an ugly myth that perpetuates thinking in Auckland, the myth being that we are one of the lowest density cities in the world, and therefore public transport won’t work.

In that particular post, I had a good dig through a very detailed set of statistics on city sizes, both in terms of area and population (and therefore obviously in terms of density). Some of the results were probably quite surprising for some. Here’s the list of selected cities by density: While obviously Auckland’s density is well below many of the large developing world cities like Mumbia and Dhaka, somewhat surprisingly our density is higher than Sydney, Vancouver, Melbourne, Perth, Brisbane and even that of greater New York. Yet all of these cities have public transport systems that function far better than Auckland’s and are far more popular – even Brisbane with its population density of less than half of Auckland’s.

Melbourne based transport academic Paul Mees used this data in his excellent book Transport for Suburbia to argue that “density is not destiny”. I wrote a blog post on that matter earlier this year outlining Mees’s argument that, while population density might have some effect on public transport use at the extreme ends of the scale (in that it’s utterly essential for somewhere like Hong Kong to function, and damn near impossible to operate effectively somewhere like Atlanta), in the middle where most cities are other factors – such as the simple quality of the public transport system – are likely to have far greater impact on whether or not people use the system than density will.

While one could be negative about such a finding, wondering whether it removes one of the big arguments for trying to contain sprawl, there’s also the big potentially positive point-of-view that “we can still have a great public transport system, even when our city isn’t enormously dense“. This is basically why Mees has written Transport for Suburbia, to show cities like Auckland that we shouldn’t wait around for another 50 years, hoping that our city will end up looking like a dense European city, before finally getting around to properly improving our public transport system. I fully agree with this and it annoys the heck out of me when people like John Banks continue to go on about Auckland being the “second most spread out city in the world” (presumably Los Angeles in the first, even though the statistics actually show LA is the densest city in the USA) to help justify why we need to complete the road network before getting around to focusing on the public transport system.

However, there’s also something in the “density is not destiny” argument that doesn’t quite feel right. Visiting one of the densest places on earth, Manhattan, and seeing the tremendous number of public transport users such an urban environment generates (due to lack of space for parking, building roads and so forth) does make it difficult to believe that the relationship between population density and public transport use isn’t significant. But if that’s the case, why do we get such weird numbers like this: Looking at those numbers, one would just about have to conclude that there seems to be absolutely zilch relationship between density and transit mode share for work trips. Yet that seems illogical – so what’s going on here?

This is where Jarrett’s post comes in to make some excellent points – including asking some very necessary questions.

But emotions often hide inside things that look like facts, and density “facts” are a great example. In transit arguments, people say things like “the net density of Toronto is ###/hectare,” because this sounds like a fact and therefore conveys some authority to their argument. In reality, though, there are several possible meanings of “density” in that sentence, as well as several possible definitions of “net” and ”Toronto.” Briefly:

  • “Density” in an urban planning context is always some kind of quantified human presence divided by some kind of land area, but it can be residential density (residents or homes per acre) or it can be total development density — including homes, businesses, schools etc – or it can be an economic density such as the number of jobs in an area.
  • “Net,” as opposed to “gross,” means that the human presence is being divided by a smaller unit of area instead of a larger one. In calculating the area, for example, you might take out undevelopable land, and bodies of water, and the land taken up by streets and highways — or you might not. There are arguments for or against excluding each of these things from “net density,” and an opinion about each of them is hiding inside the word “net.”
  • “Toronto,” of course, can be the City of Toronto, or the Toronto Transit Commission area, or the whole urban mass of greater Toronto. These obviously have utterly different average densities.

So the statement “the net density of Toronto is ###/hectare” is really as subjective as the statement “I think that cities should have more open space, narrower streets, and should have single governments covering the entire urban area.” Because each of those opinions can affect how you choose to define “net,” “density,” and “Toronto,” which in turn determines the number that you declare, with cold factual authority, to be the net density of Toronto.

So really how do we know whether we’re comparing apples with apples? This is a point that Mees makes quite strongly in his book though, that many of the past analysis of population density have been flawed through inconsistent measurements. He has certainly tried very hard to make sure his statistics don’t fall into the same trap, but it just goes to show something seemingly as simple as calculating the population density of the a city is actually very complex indeed. A corollary to this of course is the question of “how much does the density of outer outer New York really matter when assessing the city as a whole’s suitability for public transport? Isn’t the fact that it has hugely dense inner suburbs vastly more important?

This leads on to the question of whether the reason we’re getting such weird results is actually because we’re using the completely wrong measuring stick – in the form of average density. This is what Jarrett argues:

To me, Mees’s table proves that average density over a whole urban area is the wrong kind of density for understanding transit. The impression you probably have of the densities of these cities is actually closer to the kind of density that matters.

Transit reacts mainly with the density right around its stations. It is in the nature of transit to serve an area very unevenly, providing a concentrated value around its stops and stations and less value elsewhere. So what matters for transit is the density right where the transit is, not the aggregate density of the whole urban area.

Of course, what matters even more precisely is how much stuff is within walking distance of a station, so it’s not just density (the amount of stuff in a fixed radius) but the completeness of the pedestrian network. A poor connected pedestrian network can ensure that much of the stuff that’s within a 400m radius is not in a 400m walk. Consider Las Vegas, which Mees finds to be denser than highrise Vancouver. The Las Vegas economy is based on hospitality and entertainment, labor-intensive industries with low average wages. So the city needs a lot of housing suited for lower incomes. The Las Vegas way is to build utterly car-dependent apartment buildings on a vast scale, achieving all of the disadvantages of density with none of its benefits.

I completely agree. What matters is not what the density of a city is, but it’s how that density is structured. Do we have higher areas of density clumped around train stations – like you see in Canadian cities such as Toronto and Vancouver – or do we see the uniform densities of Auckland and Los Angeles? Furthermore, how conducive to public transport are our street patterns? What is the environment like for someone wanting to walk to public transport from their house? It is highly arguable that all of these issues matter far more to the success or failure of public transport in a city, than average density does.

So, returning to Mees’s question – is density destiny? I would certainly argue that average density has nothing to do with our public transport destiny: just compare New York and Los Angeles for your answer there. However, urban densities around public transport – particularly around rail – will determine the success or failure, the destiny, of that public transport system. Park and rides, feeder buses and so forth can all help in making public transport work better in low density areas, but ultimately if you want a situation where people don’t feel as though they have to own a car to live a meaningful life – I really do think you’re going to need some density.

Share this


  1. Would it be more accurate to say that workplace concentration is a better predictor of public transport use than residential densities? Wellington’s CBD provides a third of the region’s jobs, and half the people who work there use either bus or rail. In Auckland, half the CBD’s workers use public transport – but the CBD only accounts for ten-twelve percent (?) of the region’s jobs. And use of public transport to get to work is far higher in a CBD than outside it. So: the concentrated downtowns of London and New York gets lots of public transport use; the dispersed working environment of LA is another matter.

    Profoundly, there is also a cultural issue, which I’ve become increasingly aware of in my time in the UK. It was pointed out to me years ago that Canadian cities, geographically on a par with their American counterparts, got *twice* the public transport (=bus) use rates that the American cities did. The Canadian middle class, Europeans, residents of Edinburgh and Wellington, will use the bus; the American middle class, residents of Auckland and Glasgow, will not. This is why in the States there is so much enthusiasm for light rail over BRT; because that is the only way that the middle class will use mass transit – buses are there for the poor and minorities.

    1. The New York v Los Angeles example certainly points towards the importance of employment densities. However, Vancouver has a lower proportion of its jobs in the CBD than Auckland yet achieves 16% modeshare v Auckland’s 7%.

  2. It’s fine to discuss the overall, mean density of an entire city, but what about looking at the residential/commercial/employment density of the areas actually served by the existing rail infrastructure? About a third of southern and western line stations (Britomart, Newmarket, Sylvia Park, Henderson, etc.) seem to qualify as stocked with patronage potential, but the majority of these stations “serve” catchments with fewer than 500 working age adults in a 500m radius: Fruitvale, Ranui, Baldwin Ave, Southdown, Middlemore, etc. Integrated ticketing will increase the catchment slightly, but in reality, until the Council grows up and permits greater density in the proximity of the existing train lines, the ridership will always be limited by rail lines that predominantly serve sparsely-populated residential suburbs.

    1. I think even better examples exist around many busway stations. Sunnynook is the only one with a decent walk-up catchment.

      While I am a great fan of feeder buses to improve catchment size and drive patronage growth, I get the feeling that people within a walk-up catchment of rapid transit are much more likely to use it for non-work trips than those within feeder bus catchment.

      Why does this matter? Well, I think that one of the biggest economic benefits of a top-notch PT system is a reduced reliance on the need to purchase and maintain so many cars. Being able to use PT easily for non-work trips is critical in increasing the ability to live a decent life without owning a car.

  3. Dan, your example shows how easy it is to simply discount important factors: MIDDLEMORE has more than 500 working age adults in a 500m radius. It is on the front porch of one of the largest hospitals of the country! So that’s staff AND patients, and I can assure you that it is well used even during the off-peak, despite being, in some statistical terms, in a relatively “empty” area.

    You know, 78% of all researchers recommend you to be wary of statistics you haven’t made up yourself 😉

  4. Dan, Paul Mees notes in his work that it is only (from memory) ten percent of Vancouver skytrain passengers who walk to the station. The vast majority come by bus due to the integrated system of lines and timetables. So rather than integrated ticketing increasing patronage slightly, if Auckland does some planning Vancouver style it could increase the catchment of each station tenfold. Increasing density around rail nodes is probably a good way to manage growth, but it’s not a very good way to increase patronage.

    1. Nick, I don’t disagree and also sincerely hope that integrated ticketing will enhance the real catchment potential for train patronage (in addition to better coordination of the buses – but I have a sense that ARTA will not pick up on that little relationship for some time). However, my original statement was with a general claim that Auckland density is sufficient for trains. My point is that that may be true in some areas, but not necessarily where it counts.

  5. ingolfson, you’re correct about the made-up stats. The ‘500 adults within a 500m radius’ is indicative at best. A simple GIS analysis could quantify it. However, the main point is valid: the majority of the stations currently are not situated near areas with any reasonable density. I’m not saying this as an excuse to reduce priority or focus on rail infrastructure. The local transport infrastructure would simply be much better served if areas within some radius of the stations were encouraged to intensify – whether local residents wanted it or not.

  6. I would be a lot easier to extend the radius of each station to about 5km through a proper integrated bus system.
    That way you’d have almost all of Auckland within the radius of a train station, rather than forgetting about the 80% of people that are beyond walking distance. Auckland is never going to progress very far if it only ever considers walking access to stations.

    A single bus and driver could get several hundred people an hour from their existing homes to a station. That’s a lot more efficient and effective that building several hundred new homes within walking distance.

    Again I have no problem with intensification around transport nodes, but fogretting about anyone who doesn’t live within 500m of a station can’t be basis of transport policy.

  7. As with most such arguments, this is a case of rediscovering knowledge known in the 1940s, i.e. before we went down the asphalt path so completely. See Ernest Fooks, “X-Ray the City!” (1946), Fook’s point being that it is critical to identify the denser transport spines and not blend them into the low density suburbs between. They have a copy in the Auckland University Architecture library and Paul Mees also knows it well.

  8. Auckland is a great example of the pointlessness of average density, and the bad soundbites that arise from it. From a transit standpoint, the average density of permanent-low-density places like Howick is a matter of complete irrelevance. What matters is the structure of major dense centres around the PT network.

    When it comes to strong dense centres beyond the CBD, Takapuna and Manukau are already of a scale that Melbourne, for example, can only dream of. In fact, among Aus cities only Sydney has an effective network of high-density centres on its rail network. Auckland is actually way out in front, second only to Sydney in Australasia, on that score.

  9. Jarrett, the irony is in Auckland those myths are always trotted out, but in comparison to the rest of Australasia it’s closer to the opposite. Auckland is too low density, Auckland spreads too far, Auckland’s shape doesn’t suit transit etc. In reality Auckland is relatively compact, has it’s development focussed along a series of corridors and is constrained by the harbours and mountains such that it can’t ‘do a Los Angeles’.

    I believe Auckland’s geographic constraints and historic patterns of growth and form mean that it is actually very well suited to a transit based system over a motorway based one, and that the drive to build an extensive motorway network is especially wasteful as a result. At least this means once Auckland gets it’s act together (which finally seems to be happening) the city should take to transit like a duck to water.

    Melbourne does have a couple of nascent district centres, i.e. Box Hill and Dandenong are comparable to Takapuna or Manukau. But of course these are in a city four times the size of Auckland, so I guess relatively Auckland is still out in front.

Leave a Reply

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