Why cities grow and corporations die
and what can be done about it

With commentary by +Kevin Kelly +Dave Gray and +Tom Bennett

This is a thread that started in email but now I think it's interesting enough to post here for a wider group to chew on. It begins with a comment +Kevin Kelly made on this Long Now blog post by +Stewart Brand about a talk by Geoffrey West of the Santa Fe Institute.



+Kevin Kelly said: All companies die. All cities are nearly immortal.

Both are type of networks. But there are two basic network forms: organisms or ecosystems. Companies are like organisms, while cities are like ecosystems.

All organisms (and companies) have share many universal laws of growth. Creatures age in the same way, whether they are small animals, large mammals, starfish, bacteria, and even cells. All ecosystems (and cities) also share universal laws. They evolve and scale in a similar fashion among themselves — whether they are forests, meadows, coral reefs, or grasslands, or villages.

Geoff West from the Santa Fe Institute has piles of data to prove these universal and predictive laws of life. Organisms scale in a 3/4 law. For every doubling in size, they increase by less than one, or .75. The bigger the organism, the slower it goes. Both elephants and mice have the same number of heartbeats per lifespan, but he elephant beats slower.

Ecosystems, on the other hand, scale by greater than one, or 1.15. Every year they increase in wealth, crime, traffic, patents, pollution, disease, infrastructure, and per capita by 15%. The bigger the city, the faster it goes.

A less than one rate of exponential growth inevitably leads to an s-curve of stagnation. All organisms and companies eventually stagnate and die. A greater than one rate of exponential leads to a hockey stick upshot of seemingly unlimited growth. All cities keep growing. As West remarked: We can drop an atom bomb on a city and 30 years later it will be thriving.

The question Geoff West could not answer at tonight's Long Now talk was:

Is the internet more like a company or more like a city?

UPDATE: I've add my own answer to my question.

Is the internet more like a company or more like a city?

I'd bet it is more like a city.

I think the difference between the development of an organism and a ecosystem, or a company and a city, is that the later in each case evolves rather than grows. Growth is always self-limiting, while evolution is unlimited. Evolution is the infinite game; it remakes itself again and again from within so that its growth cannot catch up or stagnate.

The other question, also unanswerable right now, is: how does one make a company more like a city?

West's answer to that one was: bring in the crazy people! A city is full of crazies. You can't get rid of them, but they are the ones that keep the evolution going. Companies get rid of crazies, which in the long term will prevent them from evolving. I don't buy the crazies. Rather I think it is the defined boundaries of a company which prevent it from evolving. It is too closed, too bounded. Cities on the other hand or ill-defined, loosely coupled, permeable, center-less beings, and therefore capable of constant transfiguration without changing their essence.

That's why I think the ill-defined center-less internet will evolve and live forever. Or at least as long as a city.


+Dave Gray said: This topic fascinates me.

In short I don't think it's the crazies or the fuzzy boundaries although both contribute. I think it's the underlying structure.

Both cities and companies have hierarchical and network characteristics. But cities are managed as platforms and companies tend to be managed as hierarchical command structures. Because they are platforms that lots of people can build on, cities are magnets for entrepreneurial activity. The density of entrepreneurial activity in a city fosters all kinds of opportunities and connections. Because companies are hiererchies, they tend to repel wealth-creating entrepreneurs, who don't like to be ordered around. Also the company structure which is organized for efficiency (division of labor) has a side effect: it tends to reduce density and therefore reduces the opportunities and connections entrepreneurs thrive on.

City platforms have built-in feedback loops between the managers and the population, like elections, taxes and so on. It's clear that the city managers' job is to collect money and use that money wisely to support the platform. If they don't do that they get replaced.

Company feedback loops have long lag times, partly because they must move all the way from the customer interface up to the control center before decisions can be made and passed back down. That delay introduces all kinds of distortion and noise.

There are cities which have died and companies which have lived a long long time. A lot of the cities that died (like Uruk for example) were organized hierarchically with godlike kings at the top. A lot of the long-lived companies were decentralized and operated as platforms that supported a wide array of business units, like Unilever and Stora in Sweden, and Nokia before they divested all their non-mobile businesses, which may have been their death blow.

Most companies today are organized like the Soviet Union, with five-year plans and control centered in the hands of a few. Hierarchies are inherently unstable structures and don't adapt easily. I think companies can become more citylike and increase their productivity and lifespans by changing their structure to be more citylike. This is the subject of the book I'm working on.


+Tom Bennett said: Yes, fascinating. …and pardon my possibly dense next question, as I am only coming to this recently…

But relating to companies, what about the difference in goal or stated focus? Cities come into existence organically as a way to serve the needs of all (or at least the privileged) within their walls. Companies are willed into existence to serve the needs of 'customers', but that conception might be from one head or mind (Steve Jobs anyone?). No matter how successful, that mission or concept has its own lifespan and set of issues to resolve over time… while a city, by its very diffuse nature, can go on serving the needs of each individual citizen, on their lifespans…

Like your Nokia example… if we create an ecosystem to develop consumer products (tires or phones… whatevs) and run it like a city.. Aren't we in better shape since we've got a set of mission-diversity going?

I am essentially asking about mission, versus the 'biological' theme line.
Shared publiclyView activity