I don't have time to comment or peruse the study in depth, but this looks interesting. From Randal O'Toole:
Harvard economists have proven one of the major theses of American Nightmare, which is that land-use regulation is a major cause of growing income inequality in the United States. By restricting labor mobility, the economists say, such regulation has played a “central role” in income disparities.
When measured on a state-by-state basis, American income inequality declined at a steady rate of 1.8 percent per year from 1880 to 1980. The slowing and reversal of this long-term trend after 1980 is startling. Not by coincidence, the states with the strongest land-use regulations–those on the Pacific Coast and in New England–began such regulation in the 1970s and 1980s.
Forty to 75 percent of the decline in inequality before 1880, the Harvard economists say, was due to migration of workers from low-income states to high-income states. The freedom to easily move faded after 1980 as many of the highest-income states used land-use regulation to make housing unaffordable to low-income workers. Average incomes in those states grew, leading them to congratulate themselves for attracting high-paid workers when what they were really doing is driving out low- and (in California, at least) middle-income workers.
As Virginia Postrel puts it, “the best-educated, most-affluent, most politically influential Americans like th[e] result” of economic segregation, because it “keeps out fat people with bad taste.” Postrel refers to these well-educated people as “elites,” but I simply call them “middle class.”
I have not read the study, but I think the word "proven" in the first sentence likely goes to far. Economic systems are way too complex to absolutely show one variable among millions causes another. I am convinced that the way we have regulated the housing market and promoted home ownership has reduced labor mobility.