If this were necessarily true - that all industries would naturally gravitate to the regions of highest density - then why is so much of the banking industry now in North Carolina?
> all industries would naturally gravitate to the regions of highest density
I used a bit of shorthand, above, in saying "the relative size of cities in almost every geographic region is remarkably stable." Let me expand on "geographic region," because it reveals the tension that keeps the world from collapsing into Chongqing [1].
The size of a "geographic region" for this purpose is best defined as the distance the average economically-productive person can commute in a day. Back when people walked, towns tended to be about 10 miles apart. As transport technology improved, Manhattan ate its neighboring boroughs. It is now consuming Northern New Jersey and Connecticut. Barring a quantum leap in transportation tech, Charlotte is safe.
(On the seeds for dispersion: Certain economic activities, like mining and agriculture, occur somewhat independently of population centers. You go where the mineral veins and fertile land are. The fruits of those activities navigate the transportation geography, navigable waterways, flat areas across which rail lines can network and natural harbors, their wealth concentrating at the hubs. Thus, a fundamental industrial tension between distributed economic activity and the concentrating effects of transportation emerges.
In the modern world, those industrial distribution factors remain. Added to them are activities that are cheaper away from population centers, e.g. large-scale manufacturing and nuclear-power generation. Nevertheless, technology so far appears to be increasing the concentrating factor faster than the dispersion term by making (a) the distributed economic activities less labor-intensive and (b) increasing profits to those organizing our social structures-whether in finance, logistics or government–who benefit from the chance encounters proximity promotes.)
