On this post, Patri Friedman commented,
Think about the famous “double size, increase infrastructure cost by only 85%” rule for cities. So roughly every 16x size increase halves cost. You are trying to compete with established, funded incumbents who are easily 256x bigger than you and thus with 1/4 the infrastructure costs. And with high transaction costs for their customers to leave.
I think that the last sentence, concerning transaction costs of leaving, is interesting. I am not sure what the equilibrium would like if you brought those costs to zero.
Suppose that a big challenge with creating a new city is that the value is in the people there. This creates a Catch-22. You cannot convince me to move to a city until I know there are people there with whom I want to interact. And there won’t be people with whom to interact until you convince people to move to the city.
If there were zero transaction costs in changing cities, then you might get me to try a city before I am sure that it has enough interesting people for me. However, if I know that there are zero transaction costs to changing cities, then when I see interesting people in a city I may not be confident that they will stay there. So what do I do in that case?
I think that in that scenario, cities would behave like dating bars. Such bars tend to surge in popularity until they suddenly lose clientele.