Social housing is enjoying a moment in the United States. And, as we saw in Part 1 of this series, the proposals on the table today look, deliberately, very unlike the public housing that has been deeply troubled, stigmatized, and controversial for almost all of its 80-year history in the United States. They take their cues, instead, from models found overseas, perhaps most notably that of Vienna, Austria, where over half of residents live in publicly subsidized and rent-regulated housing, much of it of very high quality.
But we can’t just copy-paste these models into a vastly different political and land use context. We need to understand how the mechanism that social-housing advocates are proposing in the U.S. would work here.
The New Social Housing Model: Mixed Incomes and Cross Subsidy
Almost all American social housing advocates today explicitly advocate mixed-income housing. In addition to promoting integration, this brings social housing closer to being financially self-sufficient, not dependent on perpetual support from a political establishment with effectively no skin in the game. This is an understandable reaction to the past failures of public housing, which ended up concentrating only the poorest of the poor in often terrible conditions.
Basically every mixed-income proposal relies on what is called cross-subsidization, in which higher-income tenants subsidize the rents of lower-income tenants and the expansion of the program. This is not an untested idea but a feature of successful models in places like Vienna.
The logic of cross-subsidization is the same as that of inclusionary zoning, a widespread policy in which cities require private developers to include a certain percentage (typically 10–20%) of low-income units within a market-rate apartment building. The low-income subsidy comes out of the market-rate rents.
This in itself is no silver bullet for costs, and we know that because inclusionary zoning already isn’t. Think about the premise. Cross-subsidy works because market-rate rents are significantly more than high enough to cover the operating expenses and retire the construction debt: in other words, because market-rate development is highly profitable. If market rents weren’t so inflated, cross-subsidy wouldn’t look so attractive. It’s a textbook case of where the policy remedy to a very expensive housing market depends on those very high prices for its viability. The treatment perpetuates the disease.
Cross-subsidy is generally able to produce very limited numbers of affordable units relative to market-rate units, and if that ratio gets too high, development becomes unprofitable. We recently witnessed exactly this in Portland, where an unworkable inclusionary zoning requirement caused market-rate development applications to fall off a cliff with all the grace of Wile E. Coyote.
If it would cause private developers to lose money, it is also going to cause a supposedly self-sustaining government to lose money, and so we’re back at the subsidy problem: who’s paying for social housing.
Mythbusting: “With This One Neat Trick, Social Housing Is Basically Free!”
The pie-in-the-sky version of social housing advocacy involves a highly suspect claim that it’s a free lunch. Some advocates assert that a social housing agency can simply pay for itself at scale, without requiring any ongoing subsidy out of taxpayer dollars. Development is profitable, after all, and this is a developer like any other, only it’s part of the public rather than the private sector, and instead of retaining a profit margin at the end of the day, its mission is to keep rents low for poor residents.
The challenge is that even a public sector developer is subject to most of the same forces that have driven market-rate housing costs sky-high.
A foundational insight of Strong Towns is that the government does have to be in the black. Not in everything it does, of course: the public sector does some things at a loss (provide parks, for example) and makes up for them with revenue from elsewhere. But in aggregate, the public sector has to be solvent. And specific public programs need a stable revenue stream to keep operating, or they get into trouble.
Building buildings is tremendously expensive. Social housing construction will not be exempt from this. Laypeople sometimes assume that non-profit housing will be vastly cheaper to build and operate than for-profit housing, but they overestimate the difference between non-profit and for-profit. It’s more like the difference between x ≥ 0 and x > 0.