Here Wendell Cox again makes the case that most U.S. urban growth is outward. Because of the usual nostalgia for a "return to the cities" many analysts like to argue that some long standing trends are reversing. I recently added some comments to the Russ Roberts-Charles Marohn discussion. Each time, Marohn responded with his concerns over suburban growth. He cited subsidies to suburban development. One cannot repeat too often that suburbanization is international; trying to pin it to peculiar U.S. policies must address that fact. Policies are actually a mish-mash and do not point in a single direction. Decentralizing forces are too powerful to yield to the various "smart growth" policies.
Bu there are two related points that are more important. First, "suburbs" vs "center city" is an artificial and misleading distinction. The big cities are much too complex. There are sub-centers everywhere. I say this without even knowing how to define them. A three-dimensional (population or employment) density surface representing any major metropolitan area would show some peaks as well as many low-rise hills. Which ones are the "sub-centers"?
The more important point is that spatial organization is not a zero-sum game. Complex metropolitan areas are an economic unit. They persist and grow (those that do) because their spatial organization is functional and passes an on-going market test. Labor and capital are mobile; the latter is especially picky. The spatial organization (including the physical infrastructure and governance) must provide a competitive package of benefits vs costs. This is much too complex to fold into the old and cliched cities vs suburbs discussion.