Here are some graphs with a little more detail on incomes between cities. Here I am using Closed Access cities (LA, SF/SJ, San Diego, NYC, Boston) and Other Cities, which in my labels refers to the top 20 cities that aren't either Closed, Open (Dallas, Houston, Atlanta) or Contagion (Phoenix, Riverside, Miami, Tampa). The Other Cities includes Chicago, Philadelphia, Detroit, Washington, St. Louis, Minneapolis, Baltimore, and Seattle. A cross section of struggling cities, mature cities and "closed access lite" cities.
Generally, the Open and Contagion cities look like the other cities, but with incomes across the distribution that are 20-30% lower. And the US outside these MSAs has income levels similar to the Open and Contagion cities, but with a lot less income variance by this measure. Some of this could be because rural income variance is more likely to happen within a zip code while urban incomes are geographically segregated on a larger scale.
In any event, in 1998, the earliest year of IRS data, it looks like a chunk of median quintile incomes have been taken out of the Closed Access cities and replaced by a range of above-average incomes. By 2013, the same shift is apparent, but moreso. But, using Zillow median rents (which aren't available for 1998) to estimate incomes after rent, in 2013, we see that income after rent looks about the same as the income distributions in 1998, except now there is a hump at the bottom of the distribution too.
I think this is related to the migration pattern I have been describing, where high income workers can move into the Closed Access cities to tap into lucrative labor markets. Since those labor markets are more lucrative because of the limits imposed by the closed access housing policies, this migration creates a bidding war on that limited asset - housing. At the top of the income distribution are households voluntarily making a move, and natural economic forces push incomes and home prices to an equilibrium that pulls the outcomes for those households back down to a net income level, after rent, that is similar to opportunities in the other cities.
The secondary effect of that voluntary migration is for the housing stock to be bid up across the existing range of units, which raises costs for households at the bottom of the income distribution. (Why don't we call that trickle down economics? At least the economic program normally derided as "trickle down" promises growth. In the Closed Access version, it is deprivation that trickles down.) In this part of the income distribution, migration comes involuntarily, after costs impose enough distress to induce one more household to move out, making room for more of this migration pattern. The Closed Access cities are the pattern of distress that seems to be informing American politics these days - a hollowed out middle, high gross incomes at the top, and households at the bottom running to stand still, at best.
The solution to this problem isn't so much affordable housing as it is housing in general. The nearly 4 million domestic households that moved out of the Closed Access cities from 2000 to 2007 were moving because their homes were becoming too expensive. Their homes were becoming too expensive because they were the best units available for high income workers who were moving into the city. I don't think, on net, this problem is ameliorated by imposing local pressures for affordable housing that, in the end, reduce the total number of units.
Steve Randy Waldman has a typically challenging and interesting post up today about this issue. He makes some good points about the legitimate concerns that are sometimes referred to as Nimbyism. And, he is probably right that the residents most likely displaced by infill development will be the economically least advantaged. But, I think there is a bit of a forest & trees thing going on here. That is happening now. The most distressed households are the ones that are displaced now. Now they are displaced by remorseless market forces, haphazardly, a family at a time. If they were displaced by infill development, it would be in identifiable groups that would be part of local political negotiations that account for some of those dislocations.
Those dislocations are now in the service of dysfunctionality and more dislocation coming down the road while the dislocations of infill development would be in the service of functional, accessible housing in the future. But, the targets of frustration of the dysfunctional dislocation are vague while the targets of infill dislocation are specific. This seems to be the bias at the heart of the problem. And, the problem of Nimbyism seems to be that the power to negotiate against those specific forces of development has become overdeveloped.
The Oceanwide project - a skyscraper in the heart of San Francisco - seems like a great solution to infill expansion that minimizes dislocations. Yet, if my math is right, taxes, fees, etc. are being imposed on the project that amount to more than $200 per square foot of commercial and residential space.
It seems clear that the overwhelming impediment to these sorts of projects is not the lack of potential for manageable infill development. It is the political power that is targeted at stopping it.
The equilibrium of the local housing market, then, does not settle based on a supply and demand balance between cost and value. It settles based on the balance between the amount of kickbacks that can be captured by local political forces today and the certainty that the developer has about how dysfunctional local housing will be in the future. If there was any hope that San Francisco would ever allow enough development even to simply stop rent inflation, let alone pull it back toward unencumbered costs, a development that had taken on $200 in fees per square foot would not have been feasible to begin with. It's a sort of strange equilibrium. The political obstruction is what gives the development its value. The higher the fees, kickbacks, and obstacles, the more the building itself is worth, as long as the developer can be sure those dysfunctions will remain in place to prevent future developments.
The higher the fees and kickbacks, the fewer new developments, now and in the future, and the higher the rent inflation will be, and we're back to the migration pattern where low income households are displaced anyway, but their anger is just vented at an amorphous market instead of a single developer.