There’s a view that, since much of American inequality can be explained by the dynamics of housing wealth (as famously argued by Matthew Rognlie), then we could remedy inequality, or at least prevent its increase, if we took a battering ram to the gated city by eliminating height limits and exclusionary zoning and other restrictions that make it difficult for developers to add housing supply as prices increase in desirable cities and neighborhoods. I think this view is mistaken. It gets causality backwards.

It is not hard to find explanations for the increase in inequality in the US. There has been an evisceration of labor unions; selective globalization that puts the working class but not the professional class in competition with labor in developing countries; skill-biased technical change and automation that substitutes labor for capital or threatens to. All of these reduce the power of labor to bargain for their share of the economic pie. Market power increasingly concentrates income and wealth, via Facebook or Google with their unassailable platforms, via Pharma and Hollywood thanks to the expansion of narrowly conceived patents and copyrights into amorphous “intellectual property”. Deregulation of finance and various forms of “financial innovation” made it possible for skilled financiers to lay claim to gains while offloading risks, creating a class of people who won big until they didn’t but never lost very much. The professional class, thanks to the general bailout of creditors when things came apart, hitched a ride on the coattails of the gamesters of finance and came through the great financial crisis largely unscathed, while the middle and working class lost their homes and marriages and self-esteem to a margin call on housing, which has since recovered in other people’s hands.

You can certainly add having bought the right properties in the right cities in the 1970s and 1980s to the list of drivers of inequality, but I don’t think it is a big piece of the puzzle. Instead, I think it is more accurate to point out that one of the first and most valuable amenities people purchase when they become wealthier is wealthier neighbors. Wealthy people self-segregate, and the places to which they self-segregate become valuable, because the way you get a place limited to wealthy people is by bidding up the price of being in that place. The community, or the city, is gated for a reason.

Now this sounds like a story of dastardly rich people. It is not. It is a story of humans, and how humans naturally and understandably behave in the society that we have built. To be successful in our society, to be a good person, is to be a successful capitalist. One should accumulate educational and financial resources, steward them responsibly, and invest them in labor and capital markets or business entrepreneurship to yield decent returns. Failing to do this is waste. Consideration of the welfare of others is mostly delegated to the state, or else to arms-length charities to which one budgets as one sees fit. The misfortune of your neighbor, or of your cousin, is not your misfortune directly. One could never be a successful microcapitalist, and therefore a good person, if one took it upon oneself to indemnify the mishaps and misfortunes of ones neighborhood and extended family. To do too much of that is a kind of squandering, a kind of waste. If it puts the welfare of your own family at risk, especially if it reduces your child’s quality of life or education, it segues from failure to sin.

This ethos is very difficult to maintain, for human beings most of whom do strive to be good, do try to be virtuous, if we live directly among misfortune. We strive to find ways of reconciling being good and doing well, and to consider ourselves good we want to feel and be viewed as generous within our own, directly experienced communities. But we cannot simultaneously be successful microcapitalists and generous people according to the norms of less fortunate communities. Because while in theory, under mixed-economy capitalism, the state provides the less fortunate with the insurance they require to live decent lives in a topsy-turvy economy, in practice, in the United States, it does a pathetically inadequate job of it. In poorer communities, people manage their risks by pooling them directly, helping a neighbor with a rent check to prevent an eviction, or else letting her crash at their place for a while. Poorer people insure one another by forming lasting human relationships under which they make directly available, or directly draw upon, one another’s real and financial resources. Wealthier people “self-insure”, but of course that is an oxymoron. What insurance means is to create claims upon others’ resources that we can draw upon if we suffer misfortune. The “self-insurance” of the wealthy replaces the interpersonal claims of informal insurance with financial claims exercised via arms-length markets. Wealthier people save money they can draw upon in times of trouble. They purchase formal insurance contracts. Most of us try very hard not to trouble our neighbors with our misfortunes, but wealthier people are much more likely to succeed. Which makes wealthier people desirable neighbors, especially for wealthier people upon whose disproportionate resources the misfortunes of poorer people might make strong emotional and moral claims.

The wealthy huff a lot about efficiency, but what fundamentally distinguishes the insurance behavior of the wealthy and the poor is that the poor insure one another with much greater capital efficiency than the rich. The wealthy prefund their insurance individually, each household accumulating cushions of financial savings and contingent assets, the majority of which are rarely drawn upon. The poor never hold much in the way of assets they do not require, but draw upon the resources of their community and family on an as-needed basis. Wealthy communities hold financial assets multiples in value of what members of the community will ever actually use, but each household within a wealthy community may genuinely have no resources to spare, in the sense of having endowed themselves sufficiently to buffer their family’s customary lifestyle against potential shocks. At a social level, the capital inefficiency of financial “self-insurance” need not be a problem. Financial resources aren’t inherently scarce like real resources, and one can imagine a policy regime in which some sort of financial claim was made so broadly available that all households could “self-insure” in this way without increasing any burden on real resources. But in the world as it is, wisely or not, only the wealthy can afford the luxury of dormant, underutilized financial claims. The not-so-wealthy must find ways of managing their risks without the intermediation of money and markets. They call upon on another for help, and when disputes arise, as they often do with respect to need-based claims, rather than hire a lawyer to fight the insurer as a wealthy household might, neighbors and families must argue it out, in discussions that may become painful and personal and destructive of valuable relationships. Like most efficiencies, capital-efficient mutual insurance has costs invisible to the outputs-over-inputs computation. As we become wealthier, we trade less intensive use of the financial claims at our disposal for the peace of not having to field or make claims upon our neighbors.

This would all be true even if it were not the case — but of course it is the case — that various sorts of crime and discomfiting behavior correlate geographically with poverty. That correlation is itself, I think, a function of reconciling inequality with a liberal society, an effect much more than a cause of that inequality. You can agree with that or not, but the correlation still stands, and people want to move to nice neighborhoods, where “nice” is defined by the behavior of your neighbors, and wealthier people are more likely to be “nice” in the sense of troubling you less than poorer people.

Segregation is a normal good, for individuals and families. As people become wealthier, they want to insulate themselves from the chaos of other people’s lives. In a relatively equal society, there is no community of people more or less capable of substituting formal, maket-intermediated forms of mutual insurance for interpersonal and relationship-based mutual insurance. As a society that celebrates personal accumulation and self-reliance becomes unequal, then unless the state itself provides sufficient formal insurance, those capable of “self-insurance” will migrate away from those who survive by sometimes tugging on their neighbors heart strings and purse strings. Some will move explicitly into gated communities. For the more liberal and cosmopolitan among the wealthy, the quiet gate of high market prices — which they themselves must work and sacrifice to pay! — is more spiritually congenial. Individually we take prices as just a fact of nature, so the control prices exercise seems natural and legitimate, even if, from some idealistic perspective, lamentable.

In the US, segregation is overdetermined. If by the path-dependence of horrible history, we find that wealth and race become correlated, this tendency of the wealthy to segregate themselves from the poor would be sufficient to engender racial segregation. Add, gently, a modest preference for same-race neighbors or, less gently, outright racial animus, and it’s unsurprising that in the US we see the sharpest segregation across racial lines that are also economic lines. And of course, while segregation may be an outcome of yesterday’s economic tournaments, it also shapes the outcomes of tomorrow’s tournaments. The children of parents who could afford to withdraw themselves to gated communities with great schools are much more likely to be able to do so themselves, as they inherit both social capital and much of the financial capital their parents held as insurance. Race and racism very obviously shape and harden segregation in America. But we would and will have it in some shape or form regardless, as long as we are so unequal in our capacities to insure ourselves by impersonal means, and liberal enough to accommodate the preference of the wealthy to withdraw from the very personal claims of the less well insured. And just as race shapes segregation, segregation shapes race. Segregated communities don’t remain just groups of different individuals for very long. Across the lines, the different communities give one another names, turn anecdote and experience into stereotypes, attribute differences in circumstance to differences of character, behave in ways that presume and reinforce group differences. Over a generational time horizon, the phrase “racial segregation” is a tautology.