Wednesday, September 07, 2016

Cities can grow exponentially for very long, so what can organizations learn from it?

Recently, I saw this old but great TED talk by Geoffrey West on the mathematical nature of growth of cities and I can't resist sharing it.  

 

The main insight we can get here is that in living beings and companies, growth is a sub-linear phenomenon and hence it plateaus with increase in size of multi-cellular organisms, animals and companies. This is because their growth is determined by the economies of scale. 

Cities, on the other hand have shown to have super-linear growth patterns because they grow through innovation and wealth creation. The speaker says that the constraints faced by cities are Malthusian because managing the needs of such large populations with limited resources is a big problem. But cities are still able to grow because they are able to innovate their way out again and again from this problem of limited resources. We can assume (at least I assuming) that cities do so by getting either more efficient (cheaper) access to far-away resources or increasing the efficiency of resource utilization. This kind of implies that smart cities is the natural evolutionary outcome for all mega cities that survive.

As we are talking about cities, another study worth noting here is mentioned in the following article:
http://www.citylab.com/housing/2016/06/the-price-of-happiness-in-cities/487823/

It suggests that people living in suburban or rural areas tend to be a bit more happy, probably because of the self-selection effect i.e. non-urban areas tend to attract people who are generally happy with their lives and more agreeable. Maybe they have more aspects in their lives for balancing (e.g. family, preference for leisure, etc) than fast paced young professionals flocking to the cities for greater opportunities. 

The article says it's basically a matter of choice, "At the end the day, the positive characteristics of cities—their fast-paced life, diversity, greater opportunities, and heightened exchange of skills and knowledge—come at a price. It’s up to individuals to decide which ones are worth the price of happiness."

On the whole, my personal take-away beyond these new insights mentioned above is more like an idea for further research or some new hypotheses to test with data. 

To set the background for the research questions I will be proposing later: my assumption is that the sub-linear growth is a result of the extra control required by biological beings and companies for their survival, or thermodynamically speaking, they prefer low entropy/disorder, which means more effort is required to keep the order. This need for extra control over everything increases the costs of coordination to maintain this order/control (or low entropy) - entropy is supposed to keep increasing in a closed system, but we could decrease entropy locally through external efforts (think of AC throwing out the heat from your room by using much more electrical energy to do that task). 

In case of companies, at a high level we will borrow the idea of transaction/coordination costs from the theory of firm by Ronald Coase who got the Nobel prize for his theory which also places limits on the growth or organizations based on the coordination costs. 

Cities on the other hand aren't (or can't be) controlled so tightly like internal body functions of an organism or internal transaction/co-ordinations within a firm. Therefore the costs of coordination are much less (because of higher and always increasing entropy cities enjoy) and hence we can observe the super-linear/exponential growth. Such growth is also a function of the network size (population) and thinking minds we have i.e. emergent combinatorial phenomenon of mixing of different ideas leading to innovation (and wealth creation). 

At the end it is about the evolution of a new type of organism (i.e. cities) which is not tightly controlled (like living beings or companies). Hence its evolution is limited not so much by the inertia (or pull) of the internal coordination costs (or costs of maintaining low-entropy) as much as by the Malthusian resource constraints. So cities keeps growing forward with the evolutionary force of new ideas (i.e. innovation) by pushing this constraints ceiling/limit further and further through efficiency gains in either resource utilization or access.

Firms everywhere also try to keep out-innovating their competition for the sake of survival. So what would happen if we come up with a new type of firm (or legal structure or business models)? Something that doesn't require running a tightly controlled ship or where the coordination costs are decreased significantly (through the use of technology). Could such firms outlive their peers in the market? If yes, then what should be such a structure/entity/model? Or what are the elements that make such enterprises and markets?

Researchers could work on determining whether sharing economy, platform models or on-demand service provider companies have grown fast mostly because of reduced coordination costs (through the use of technology). Or if there are any other type of innovations (except tech innovation) which would allow for the enterprises to grow exponentially in a sustainable manner. 

Also, is technology based reduction in coordination costs an advantage that is sustainable in the long term for a firm to keep out-surviving its competition? What kind of ownership structure should such firms have and how that ownership structure evolves? Would such firms requires different type of markets rules to exist? How to allow the increases in entropy for companies while improving (or at least without diluting) the customer experience of their offerings in the market? What does it mean in terms of the happiness of people who would be working in such organizations and markets i.e. is their a happiness trade-off like it has been observed in urban v/s suburban/rural living?

Another thing I feel is that while some of these questions could be tested on real firm data, probably for internet economy firms, but in most cases simulating such markets using agent based modeling could be more fruitful. Though in that case also I really have no idea right now on how to gather the data for starting set of assumptions or scenarios to be simulated. Hope this could lead to some real research project someday.

Update 8th Sep, 2016: Just now found another interesting (but longish) talk by Geoffrey West that could be interest for this topic.



Will write more about this talk when I get some time.