The subtitle of this article from the Washington Monthly (HT: EV) was so enticing: "Regional inequality is out of control. Here’s how to reverse it." I first went to see if they would have any data on housing's role. Soon, I realized I would have to read the whole thing only to confirm that they never mentioned housing. They didn't.
They have properly identified the problem. The author even recognizes the reversal of migration flows that now has households moving away from high income cities. But, I will just say that the repeal of the Civil Aeronautics Board, the ICC's deregulation of interstate trucking routes, and the toleration of branch banking loom large in his list of causes. I'm not kidding. Click the link if you want more.
But, I mention the article because of this graph. They truncated this graph to before 1980, when the convergence stopped. This is out of BEA table SA1. The regional inequality doesn't diverge again after the 1970's, so in all of the author's comparisons about recent rising regional income inequality, he compares cities like Houston and Atlanta to Washington, New York City, and San Francisco.
So, I downloaded the data to see how it looks after 1980. On the regional level, I think we can see how the technological and commercial innovations that made the world a figuratively smaller place created a more equitable world. (Or maybe it was the ICC, or 70% tax brackets.) We are basically seeing the same phenomenon now, globally.
The regional inequality we have seen since then has been specific to cities. But, on the scale of the metropolitan area, the added gross income for the highest income cities frequently goes to housing expenses. I think much of the topic of the regime shift in income inequality before and after the 1970's relates to these factors. Free economies naturally tend to produce convergence, and this was the case regionally until that convergence had generally played out. Possibly, some divergence on the individual or household scale also tends to grow with economic expansion, and this was mitigated by those regional convergences until the 1970's. A new phenomenon that seems to selectively infect cities across the Anglosphere is sharp limits on housing in high income cities. So a regional phenomenon that reduced aggregate income inequality is largely played out, but a new localized phenomenon has caused aggregate income inequality to rise.
Convergence comes from free flow of capital and labor. Everyone understands this, at least implicitly - this realization is a basis for nativism and protectionism. When deregulation (unfettered immigration and trade) pulls the top 1% (the average American) down toward the mean, everyone understands how deregulated markets create income convergence. (edit: This reflects the protectionist argument. In most cases, I would argue that most or all of the convergence that comes from the flow of capital and labor is due to the mean rising.)
Maybe Bryan Caplan should be applying his open borders project to America's cosmopolitan cities.