Advice, stories, and expertise about work life today.



We know your inbox is protected space, so we promise to send only the good stuff, twice a month.

Fun Maps Science of Maps Maps

The surprising science of cities

Cities are like Jekyll and Hyde. They’re the origin of the world’s biggest problems, such as water scarcity, pollution, and wobbly financial markets. But they’re also the solution, according to Prof Geoffrey West of the Sante Fe Institute, because they act as magnets for smart people, innovation and ideas.


gwestAnd they don’t come much smarter than West, who believes the solution to such problems is the appliance of science. To that end, and with help from fellow lead investigator Prof Luis Bettencourt and others, he’s been developing a universal theory of cities. “The idea is not to replace traditional methods but to complement them. If we just use historical, intuitive kinds of arguments, we are doomed to all kinds of unintended consequences,” he says.

The theory is being developed by gathering a huge amount of data on cities, from wildly different sources. Maps are an integral part of presenting some of this data - augmented maps have displayed the median income in areas of Cape Town and New York, levels of violent crime in US cities and clusters of social groups in a Chinese city.

West was originally a particle physicist. He left the field when funding for the Superconducting Super Collider - America’s proposed equivalent of Europe’s LHC - fell through in 1993. And it wasn’t long before he discovered a new scientific passion: biology.

He set about applying the principles of physics to biology’s scaling laws - the discovery that, despite being complex, sometimes life follows simple rules. For instance, if you plot the metabolic rate of various animals against their weight on a graph, you get a straight line. “A whale, a giraffe, a human being and a mouse are scaled versions of one another even though they look completely different,” he says.

Networks are everywhere

The reason for this, West discovered, is that networks are found in all living things - from human circulatory and respiratory systems to the roots of a tree. All can be described using the same mathematics and physics. And what’s more, networks are everywhere, he says. “It became clear that one of the extensions of this might be social organisations.”

Social organisations don’t come any bigger than cities, where more than half of us now live, and nearly 80 per cent of us will reside by 2050. “The first thing I wanted to address was: do cities scale? Is New York a scaled-up Los Angeles, which is scaled up Chicago?”

It’s a question West and his collaborators set out to answer by developing a theory of cities based on networks. They turned up a surprising relationship. If you double the size of a city, any socioeconomic activity goes up by more than double. Everything from wages to patents and the number of AIDS cases gets a 15 per cent extra boost.

“The amazing thing was that it appeared to be true across completely different metrics. But at the same time there was an extraordinary economy of scale. When you double the size of a city, infrastructure only increases by roughly 85 per cent. You don’t need twice as many petrol stations or roads,” he says.

Our cities reflect us

Even more surprising, the same rules applied to cities in every country for which West and his colleagues could find data, from Colombia and Chile to France, the Netherlands and the USA.

“It strongly suggested that the structure, growth and evolution of cities isn’t just arbitrary. Underlying principles that transcended history, geography and culture were operable in all cities,” says West.

So what was going on? “Cities are places where human beings come together. Human beings are pretty much the same all over the world and somehow these remarkable scaling laws are a manifestation of that.”


The challenge was then to turn the data into mathematics that would, in turn, form a theory that could be used to make useful predictions about a city’s future. “There can never be a theory of cities like Newton’s laws. But in a ‘coarse-grained’ way, the average behaviour of the average city forms a baseline for understanding what’s going on,” he says.

Looking at this baseline, or rather deviations from it, formed the next stage of West’s research - one that would prove interesting reading for any mayor or urban planner.

“If you give me the size of a city, I can tell you how many police it should have, how much violent crime, how many AIDS cases, what the length of all the roads will be, and so on. Within 10 to 20 per cent, those numbers will be correct,” says West.

“But that’s the baseline. How does an individual city perform relative to what it “should” be doing? On any given metric, some will overperform and some will underperform. We might predict a particular city ‘should’ have 227 police, but it might actually have 219,” he says.

The resulting published paper ranked the performance of 365 US cities and examined how their performance had changed over 50 years. The answer was: not much. With a few exceptions, cities that were doing well in 1950 are still doing well today - and vice versa. “There’s great resilience - you need decades in order to make changes,” says West.

Looking deeper

The research is now going beyond averages to deconstruct cities into their constituent neighbourhoods, suburbs and downtown districts. They include slums in Africa and India. Backed by the Bill & Melinda Gates foundation, researchers from the Santa Fe Institute analysed maps and other data from Shack/Slum Dwellers International (SDI). The aim was to standardise the way informal communities collect data about themselves, which in turn helps to improve slum conditions.


A similar research strand is looking to classify subdivisions of cities. “What are the suburbs? What is the downtown? The literature of urban geography and urban economics are poor at giving operational definitions, and that’s huge problem,” he says.

One way around this problem is to use mobile phone data - lots of it. West and his colleagues secured data sets consisting of billions of calls, all anonymised, for the Greater Boston area, Singapore and Senegal. “From these we can trace out the movements of people in terms of how often they visit places, and from how far away,” says West.

This research, which has yet to be published, has once again turned up a surprise. “What was extraordinary was that the patterns we saw in Boston are the same as the patterns we see in Singapore and in Senegal. That was amazing and we are still struggling with trying to integrate that into the general theory.”

So how close are we to having a complete, universal theory of cities capable of making accurate predictions based on mathematical laws? “There’s huge amounts to do – in one sense we’re still scratching the surface. But I think we’ve made tremendous progress, given that we started from nowhere and no one had really thought about cities and urbanisation from this viewpoint.”

Graham Southorn is the consultant editor of BBC Focus magazine


Urban Mobility

Optimize on-demand and scheduled mobility operations to enable seamless intermodal journeys.

Find out more

Sign up for our newsletter