Scaling laws, innovation and the science of cities
Since the late 2000s I have been hearing about exciting new developments around cities. A group of researchers, some of which I even know personally and collaborated with, had discovered impressive empirical regularities in how cities scale. The gist of it is: every human city, across time, space and culture, is a scaled-up (or -down) version of every other city. But such scaling is non-linear: it follows power laws.
- Infrastructure requirements scales sublinearly to city size, as measured by the number of inhabitants. The exponent of the scaling function is 0.85. As you increase the number of a city's inhabitants by 100%, it only needs 85% more kilometers of cables, cellphone masts, etc.
- Social and economic indicators scale superlinearly to city size. The exponent of the scaling function is 1.15. As you increase the number of a city's inhabitants by 100%, its GDP, number of patents, crime rate etc. increase by 115%.
These two phenomena drive the increasing dominance of cities. The size and cost of the support systems needed to support a human live is smallest in the biggest cities. This means cities are a powerful green technology, as argued by Stewart Brand. Also, large cities are conducive to more, faster innovation, and the "buzz" and excitement that city dwellers love so well.
We even know why this happens. Economies of scale in life support are driven by something called hierarchical branching networks. I won't venture in an explanation about this. If you are curious, read Geoffrey Wests's Scale. It's great, superclear and a lot of fun. I want, instead, to take a closer look as why social and economic indicators scale the way they do.
It has to do with social networks. Social networks scale "in reverse". Think of a sublinearly-scaling network like an aqueduct. It has a very big "master pipe" leaving the reservoir of drinkable water. This forks several times into smaller pipes, until it gets to the terminal size good for serving your apartment. Your social network, is not shaped that way. You have strong ties to people closest to you, also in the physical sense. But you also have weaker ties, and those travel farther. For example, they might reach a former classmate who now lives in a different county.
These ties are cheaper to maintain, because you do not talk to these people as much as to your closest friends. But they are also liable to create the most opportunities to you: a new idea, a job opening you did not know about, an interesting book or film. People are more likely to find a new job through distant acquaintances than through close friends. In cities, it is easiest to build low-maintenance, high-potential, long-distance weak ties. And the longer-distance the tie, the higher its potential for novelty and cross-pollination. The flow of ideas feeds on diversity. Unlike with water systems, here thinner, weaker "pipes" (social ties) transport more innovation and opportunities.
This is fascinating stuff, and I plan to go much deeper into it. It also has implications for our own plans with building Edgeryders settlements.
The whole settlement operation grew out of Edgeryders itself. The edgeryders.eu platform is the main connecting layer for our community. We tend to use it in a focused way: no cute cats, celebrities or ranting about politics here. Almost 100% you'll read is people comparing notes, as they try to reinvent their own life and work as part of some kind of societal change. And almost 100% of these conversations are long-distance. This is because reinvention and change can be lonely. If you are trying to reinvent the home, or pushing for social innovation in war-torn zones, chances are you are the only person in your local context doing those things – unless you live in a major city. You are going to need someone to talk to, learn from, check in with. On Edgeryders you can find them, and interact with them in a distraction-free environment.
So, online spaces like Edgeryders provide weak, low-maintenance, long-distance social ties. Exactly the kind of social ties that make cities so innovative. If Geoff West, Luís Bettencourt and the other complex systems urban scholars are right, those ties are what a city is. Except, in our case, they span many contexts, urban and non-urban. If you visualize cities as the branches of a Christmas tree, these social ties are like festoons. They connect people on different branches, whereas city life connects people on the same branch. We are a kind of meta-city.
Two strategies for settlements
There might be a way to test for structural similarity of Edgeryders-style weak ties to city-style ones. I have not looked into it yet. But let's assume, for the sake of the argument, that the meta-city intuition holds true. Then what? What does that have to say about the way we set up spaces to live and work in?
There is one straightforward implication. This: invest in "city-like" weak, long-distance, low maintenance social ties, especially if you don't live in a major city.
In practice, people in Edgeryders are building two very different kinds of settlements. Let's call them "rural" and "urban". Both provide affordable living and working space for project- and small business development. But they do so with different strategies. So, investment in weak ties means different things in different kinds of spaces.
"Rural" settlements, like the one in Sidi Kaouki, offer cheap space in a context of natural beauty. They strive for relative self-sufficiency and a low carbon footprint. This sounds great, and it is, but there is a price to pay: relative isolation. Their world will be small, unless they make constant efforts to enlarge it.
"Urban" Edgeryders settlements, like The Reef in Brussels, offer affordable space and an edgy culture applied to work and business. They strive for sustainability by making the most of what the city offers (public transport, farmers markets etc.). They use their central position to connect to opportunities, and share them across the network.
For rural settlements, investing in weak ties means two things.
First: make your place attractive, and be generous with your hospitality. It's the analogue of the vow of stability Benedictine monks made. Someone must love the place enough to stay with it, and welcome others who might help.
And second: pay a lot of attention to your online presence. You are going to need the long-distance ties that it affords. Documenting your own efforts goes a long way to put you on the map. @matthias's optical coffee sorter, for example, attracted the right kind of attention. Not "go viral", which would be a distraction, but about ten skilled people willing to contribute. This would be a sort of "vow of connectivity": being open to the world and building new social ties from a remote, rural place.
Urban settlements already come with weak ties. Despite this, paying attention to their online presence is a good idea. They are already in one city, but they can and should partake of the meta-city too. Participating in Edgeryders (or other online communities like it) gives them that kind of access. But there is another move they can make. They can position themselves where they can bridge what Ronald Burt calls structural holes in the city's social network. The best candidates for this are at the border of poorer neighbourhoods. There, our settlements can connect people at the edge of urban life to the centers of power. This will make it easier for the settlement to stay affordable. At the same time, it will generate positive externalities and opportunities for our neighbours.
I may be off the mark here. I realize this is all sketchy and conditional. The general point is not that I have found a perfect model for Edgeryders settlements. The general point is that we are now in a position to attempt to make some kind of model, as opposed to steering by intuition. There is a science of cities. We should study it, and use it to make better decision.
Do you know anything about the science of cities? Do you have any information, or ideas we should know about, to help us make location decisions for the next wave of spaces?