Wendell Cox tipped me to Population: The Growth of Metrpolitan Districts in the United States, 1900-1940. In those days, the discussion cited "metropolitan districts," including and their "satellite areas" which seemingly refers to the suburbs.
On page 8, the report summarizes 1930-1940 changes. "Thus for the first time the major increase in numbers in our metropolitan districts was found in the satellite areas and the slower growth of metropolitan districts as a whole, which might have been expected to retard decentralization, had just the opposite effect."
In 2008, Wendell Cox and Chris Redfearn and I responded to a QJE paper by Nate Baum-Snow in this issue of EconJournalWatch. Baum-Snow had sought to demonstrate that U.S. suburbanization was very much the product of the post-1956 Interstate Highway System. That is the view of many who see suburbanization as being an artifact of policy. We argued that policies followed preferences, rather than the other way around.