Urban density matters – but what does it actually mean?

Shove up: London's Oxford Street, looking crowded. Image: Getty.

In debates about urban density we often find comments about buildings being too tall or not tall enough, about too many people in a neighbourhood or too few, about streets and buildings being overcrowded or empty.

We are told that Melbourne is building at four times the density of Hong Kong, or that density is good and will make us happy.

But as these debates over density in cities continue, what is most often missing is any clear understanding of what people mean when they use the word “density”.

Is it the volume or height of buildings? Or is it the numbers of people? One person’s high density may be another’s sprawl; the same tall building may be experienced as oppressive or exhilarating; a “good crowd” for one can be “overcrowded” for another. One is reminded of Humpty Dumpty’s logic in Through the Looking Glass:

When I use a word it means just what I choose it to mean – neither more nor less.

What such debate most needs is greater density literacy.

Density is a concept borrowed from physics where the meaning is clear – mass divided by volume. Yet when transferred to the city, nothing is so simple.

Are we talking about buildings or people?

First, we need to clarify whether we are talking about concentrations of people or of buildings. If we take population densities first, these are generally measured as residents per hectare based on census data.

But we also need to distinguish between “internal” and “external” densities – the numbers of people in a room or apartment versus those in an urban precinct. If you look at new high-rise housing in the evening, you can find many apartments unoccupied. At the other extreme, internal crowding largely defines a slum. Building density does not mean population density.

Building density does not necessarily mean population density, an example being low-occupancy housing in Melbourne Docklands. Image: Kim Dovey/author provided.

Population densities cannot be based on residents alone, since the numbers of people in a given neighbourhood at a given time include those who work there or are visiting. In a mixed-use neighbourhood, residents may be a small proportion of the population density.

There are also population density rhythms, as people move from place to place throughout the day and week. The same urban precinct can be densely populated during work hours and empty on weekends. Population density is not only the number of residents but fluctuates over time and with functional mix.

And how do you measure building density?

When we shift attention to building densities, we soon encounter some unavoidable jargon. The most common measure of building density is the “floor area ratio” (known variously as FAR, FSI, FSR and plot ratio) – the ratio of floor area to land area. This is the most widely used measure for limiting the bulk of development on any given plot.

However, it does not control the building height, “footprint” (the area occupied by the building) or “coverage” (the proportion of land covered by buildings). Thus it is quite possible to build high-rise low-density (with very low coverage or small footprints) or low-rise high-density (with high coverage or large footprints).

Most high-rise public housing from the 1960s and 70s is roughly the same FAR as the low-rise housing that was demolished to build it.

Building height is not a measure of density, although sometimes the two align. Confusions here abound; press reports (such as here and here) regularly equate FAR with height control.

Areas of different heights but similar building density in Fitzroy, Melbourne. Image: Kim Dovey/Google Earth.

Another common measure of density is dwellings per hectare. This is often used as a means of assessing population and building densities at the same time. But it does neither, unless we know the size of dwellings and of households. Thus dwellings per hectare is a very blunt measure of density, although a useful proxy for comparing housing projects.

Then there is the distinction between gross and net densities. Urban planning controls are focused on the net density on a particular site – yet such measures are of little use in understanding how cities work, because they do not include the public space of streets and parks.

The gross density is always lower than net density and it is the one that matters in debates over urban density. While we might be packed in on a particular site, the street network of a car-based city tends to keep us apart. Net density is not an effective measure of urban density.

As we measure densities of people or buildings at larger scales, we also incorporate water bodies, freeways and unbuildable sites, so the average density diminishes. Where one draws the boundary is a crucial decision in measuring urban density, or in getting the answer one wants.

Even urbanists such as Paul Mees can be guilty of this when using broad metropolitan density measures to advocate for public transport at suburban densities. Elizabeth Farrelly, a vocal proponent of higher density, compares it to the thread-count of luxury sheets (her minimum is 1,000), yet her only measure for urban density is “dwellings per hectare” – whatever the scale, net or gross.

Here we encounter the politics of density literacy and return to the logic of Humpty Dumpty:

The question is which is to be master – that’s all.

There is no single scale at which to measure urban density, but the larger the scale, the lower the density. The best approach is to understand density as multi-scalar: for any location there is an internal density, a net density, a walkable density and a metropolitan density.

High density is no guarantee of urban buzz

Finally, there is the question of streetlife density – of people in public space, of crowds and crowding. Here the complexities multiply. While we can measure the outcomes in pedestrians per minute or per square metre, we are far from understanding the ways in which streetlife is geared to building and population densities.

These connections depend at least on numbers of jobs and visitors, functional mix, car dependency, access networks and walkability. Yet this is where density delivers its greatest benefits in social and economic encounters – what we often call the urban “buzz” or “intensity” – along with disbenefits such as congestion.

Density without intensity in the car-dependent city – in this case, Tampa, Florida. Image: Author provided.

In Australian cities we have become quite good at generating high density without intensity. Think of car-dominated high-rise districts where few people use the street. Yet we also have good inner-city examples of intensity without high density.

For many people, density has become a negative word. Those who want more of it often use other words and phrases: the “compact city”, “urban intensification”, “transit-oriented development” and the “30-minute city”.

This can be useful language but the question of what it means remains. The challenge is to raise the standard of urban density literacy – not to make density mean one thing but, as Alice might put it, to understand the ways it is made to mean so many different things.The Conversation

Kim Dovey is professor of architecture and urban design, and Elek Pafka a research fellow, at the University of Melbourne.

This article was originally published on The Conversation. Read the original article.

 
 
 
 

What Citymapper’s business plan tells us about the future of Smart Cities

Some buses. Image: David Howard/Wikimedia Commons.

In late September, transport planning app Citymapper announced that it had accumulated £22m in losses, nearly doubling its total loss since the start of 2019. 

Like Uber and Lyft, Citymapper survives on investment funding rounds, hoping to stay around long enough to secure a monopoly. Since the start of 2019, the firm’s main tool for establishing that monopoly has been the “Citymapper Pass”, an attempt to undercut Transport for London’s Oyster Card. 

The Pass was teased early in the year and then rolled out in the spring, promising unlimited travel in zones 1-2 for £31 a week – cheaper than the TfL rate of £35.10. In effect, that means Citymapper itself is paying the difference for users to ride in zones 1-2. The firm is basically subsidising its customers’ travel on TfL in the hopes of getting people hooked on its app. 

So what's the company’s gameplan? After a painful, two-year long attempt at a joint minibus and taxi service – known variously as Smartbus, SmartRide, and Ride – Citymapper killed off its plans at a bus fleet in July. Instead of brick and mortar, it’s taken a gamble on their mobile mapping service with Pass. It operates as a subscription-based prepaid mobile wallet, which is used in the app (or as a contactless card) and operates as a financial service through MasterCard. Crucially, the service offers fully integrated, unlimited travel, which gives the company vital information about how people are actually moving and travelling in the city.

“What Citymapper is doing is offering a door-to-door view of commuter journeys,” says King’s College London lecturer Jonathan Reades, who researches smart cities and the Oyster card. 

TfL can only glean so much data from your taps in and out, a fact which has been frustrating for smart city researchers studying transit data, as well as companies trying to make use of that data. “Neither Uber nor TfL know what you do once you leave their system. But Citymapper does, because it’s not tied to any one system and – because of geolocation and your search – it knows your real origin and destination.” 

In other words, linking ticketing directly with a mapping service means the company can get data not only about where riders hop on and off the tube, but also how they're planning their route, whether they follow that plan, and what their final destination is. The app is paying to discount users’ fares in order to gain more data.

Door-to-door destinations gives a lot more detailed information about a rider’s profile as well: “Citymapper can see that you’re also looking at high-profile restaurant as destinations, live in an address on a swanky street in Hammersmith, and regularly travel to the City.” Citymapper can gain insights into what kind of people are travelling, where they hang out, and how they cluster in transit systems. 

And on top of finding out data about how users move in a city, Citymapper is also gaining financial data about users through ticketing, which reflects a wider trend of tech companies entering into the financial services market – like Apple’s recent foray into the credit card business with Apple Card. Citymapper is willing to take a massive hit because the data related to how people actually travel, and how they spend their money, can do a lot more for them than help the company run a minibus service: by financialising its mapping service, it’s getting actual ticketing data that Google Maps doesn’t have, while simultaneously helping to build a routing platform that users never really have to leave


The integrated transit app, complete with ticket data, lets Citymapper get a sense of flows and transit corridors. As the Guardian points out, this gives Citymapper a lot of leverage to negotiate with smaller transit providers – scooter services, for example – who want to partner with it down the line. 

“You can start to look at ‘up-sell’ and ‘cross-sell’ opportunities,” explain Reades. “If they see that a particular journey or modal mix is attractive then they are in a position to act on that with their various mobility offerings or to sell that knowledge to others. 

“They might sell locational insights to retailers or network operators,” he goes on. “If you put a scooter bay here then we think that will be well-used since our data indicates X; or if you put a store here then you’ll be capturing more of that desirable scooter demographic.” With the rise of electric rideables, Citymapper can position itself as a platform operator that holds the key to user data – acting a lot like TfL, but for startup scooter companies and car-sharing companies.

The app’s origins tell us a lot about the direction of its monetisation strategy. Originally conceived as “Busmapper”, the app used publicly available transit data as the base for its own datasets, privileging transit data over Google Maps’ focus on walking and driving.  From there it was able to hone in on user data and extract that information to build a more efficient picture of the transit system. By collecting more data, it has better grounds for selling that for urban planning purposes, whether to government or elsewhere.

This kind of data-centred planning is what makes smart cities possible. It’s only become appealing to civic governments, Reades explains, since civic government has become more constrained by funding. “The reason its gaining traction with policy-makers is because the constraints of austerity mean that they’re trying to do more with less. They use data to measure more efficient services.”  

The question now is whether Citymapper’s plan to lure riders away from the Oyster card will be successful in the long term. Consolidated routing and ticketing data is likely only the first step. It may be too early to tell how it will affect public agencies like TfL – but right now Citymapper is establishing itself as a ticketing service - gaining valuable urban data, financialising its app, and running up those losses in the process.

When approached for comment, Citymapper claimed that Pass is not losing money but that it is a “growth startup which is developing its revenue streams”. The company stated that they have never sold data, but “regularly engage with transport authorities around the world to help improve open data and their systems”

Josh Gabert-Doyon tweets as @JoshGD.