Labor and capital in the U.S. have been moving from "frostbelt" places to "sunbelt" destinations for many years. Climate matters.
A recent paper by Gyourko-Glaeser (I still cannot link from my European hotel computer but the paper is easy to find) made the point that much of urban captal stock, including housing, is durable and takes many years to whither. Simply cover variable costs and this stuff (including many old U.S. cities) hangs on for many years. And variable costs can be covered by low-income and/or welfare populations. Old and run-down places will be around for years to come.
What about the exceptions to the rule? Ed Glaeser's recent NBER paper cites New York City and the productivity gains from high densities of human capital. Tom Wolfe famously asked whether the CEOs like Manhattan because the great chefs are there or whether it is vice-versa.
Densities, however, are a funny thing. They vary drastically as geographic definitions change. New York has the nation's highest city density while Los Angeles has the highest urbanized area density. Agglomeration economies surely matter for Silicon Valley but the place, according to some, straddles San Francisco Bay.
Apparently, it takes both insights -- the long life of the NYC building stock that houses immigrants and other low-income families and the propinquity of CEOs, chefs, sex-and-the-city types, etc. -- to explain the Frostbelt success stories. Besides, even the sunbelt-to-frostbelt migration cannot go to the point where it empties half the country. There will always be some outposts.
Also, Fred Siegel's writings re New York City politics point to much that is dysfunctional -- and a continuing challenge to the area's peculiar advantages.
When the housing "froth and fizz" dissipate, differential rates of slowdown will be worth studying.