Here (pdf) is a great paper from Peter Ganong and Daniel Shoag with several interesting empirical findings regarding migration patterns.
They find empirical confirmation of many of the patterns I have been covering here. And they find that the reversal of historical migration patterns toward income opportunities is responsible for creating income inequality. High income workers still retain the historical migration pattern because their higher incomes provide a higher return on the payment they have to make for housing - the gatekeeper to opportunity in a closed access city. For low income workers, the return to tapping into productive labor markets isn't high enough to cover the cost of limited access housing, so they are induced to migrate away from income opportunities.
Here is Figure 6 from their paper. In past periods, all of these regressions would have positive slopes. In other words, workers migrate to places with higher income potential. But, in the recent period, the regression with migration and gross incomes for low skilled workers has a negative slope. Yet the regression for incomes net of housing expenses remains similar to the historical norm.
Housing expense is the dominant issue of our time, and it is the dominant issue directing the migration of low income workers.
What we called a housing bubble was actually a refugee crisis. Migration patterns explain what happened much better than banking activities. Blaming the housing bubble on bankers is like blaming illegal immigration on coyotes. We impose desperation and exclusion on people, then when the context those people face is invariably peppered with incivilities and difficulties, we demonize the very agents who are trying to facilitate access to opportunity despite all of our efforts to deny it.