In other words, why don’t they put everything into Atlanta or Los Angeles? Paul Krugman has a good blog post on that topic, here is one of his points:
…once upon a time dispersed agriculture ensured that small cities serving rural hinterlands would survive. But for generations we have lived in an economy in which smaller cities have nothing going for them except historical luck, which eventually tends to run out.
Krugman suggests that eventually many smaller cities will indeed fade away, although the process of equilibration may be a long and slow one. All of his points are well-founded, nonetheless I can see a few factors favoring the continuing existence of small cities on a greater scale than many might be expecting:
1. As Alex points out on Twitter, rents are permanently lower, and many people don’t value big city amenities very much.
2. Congestion is likely to be lower. Why should the larger city have worse traffic if it has proportionately more roads? That may require a blog post of its own, but part of the problem is geographic specialization within the larger city, which is not simply some number of smaller cities placed side by side. In other words, sometimes you have to drive all the way across town. Many people don’t like geographic specialization, but wish to find most everything in a small downtown or Walmart (or on Amazon). From this point you can see that Amazon may favor larger cities more than small towns. If it bugs you that in a large city all the shopping of a particular kind is on the other side of town, just order those goods on-line and stay within your cozy neighborhood.
3. Governance may become worse in a very large city. Furthermore, separate and specialized lobbies, as would correspond to geographically specialized parts of a large city, may be a bad influence. Here is a paper on the public choice of mega-city governance.
4. Very large, rich, and famous cities tend to become financial centers, or perhaps movie-making centers, and that is not in the interest of all city residents. Some of this is a matter of rents, in other regards a matter of culture and ethos. Anonymity also increases with size, as does (I think) sexual promiscuity. Smaller locales will have more faux conformism and more real conformism too, which some people prefer. People not wanting to live amongst all the specialization of major cities really is a significant and enduring factor in these comparisons.
5. If you are building a firm for eventual export success, you will prefer to put that firm in a larger city to begin with (“built to scale”). That in turn tends to price out companies and people with less interest in exporting. The larger city will become all that much more globally oriented, which not everyone will wish to pay for or even wish to have at zero price.
6. If I were offered an extra 50% of total salary (nominal, to make this comparison in real terms across all goods and services eliminates the very difference in locales) to move from Fairfax to Washington, D.C. (15-20 miles away), I would decline the offer.
7. The very fact that smaller cities are used to consume non-pecuniary amenities suggests their inhabitants are more diversified than it may appear at first. The shift of gdp into services further enhances this diversification, and the new crop of semi-small cities may be more resilient than the older lot dependent on manufacturing.
8. A significant and enduring trend is the move into warmer and sunnier climates. So while Rochester and Flint decline, Chattanooga and Birmingham are on the rise. I predict the more time you spend in the South, the more optimistic you will be about small and mid-size cities.
9. Here is a good Duranton and Puga piece on the costs and benefits of city size. Here is a short McKinsey piece on complexity as a limit on size. Here is a discussion of city size in Civilization VI.