How to make social housing work in Chicago, part 1

Explaining our new affordable housing tool

Unless you live under a rock or have been blessed by the winds of fortune, you probably are aware we face a housing crisis. This month, Chicago took one of the boldest steps in the nation to address this crisis by passing Mayor Brandon Johnson’s much-touted Green Social Housing Ordinance, or GSH for short.¹

GSH, which follows up on similar programs recently launched in Montgomery County, Maryland, and Atlanta, Georgia, establishes a city-owned non-profit to fund mixed-income affordable housing developments. These programs have been hailed in venues from NPR to Vox to Bloomberg because, unlike with “traditional” affordable housing programs, they promise to create affordable housing through innovative bond-financing tools that remove the need for federal subsidies known to be mind-numbingly complex, notoriously unstable, and short of demonstrated need.

Most pieces you will read on “social” housing are heavy on praise and light on detail. This piece seeks to rectify that issue–just because social housing works in Vienna does not mean it will work in the capital of Vienna Beef. In today’s piece, I’ll explain what social housing is, how it differs from existing affordable housing programs, and the basics of Chicago’s new legislation. Next week, I’ll explain what needs to happen in order for our program to succeed.

Let’s dive in.

What is social housing? Sourdough starter vs. instant yeast

In a standard market rate development, 60-80 percent of a project’s cost is funded by debt (e.g., a construction or permanent loan from a bank), while equity investments (e.g. investments in ownership stakes of a building) fund the rest. After a building is constructed, rents from the newly-constructed units are used to pay down the debt, and (hopefully) generate a profit for the equity investors.

Market rate investors demand a return on that debt and equity, one that must account for the risk inherent in the project and potential alternative uses for their capital. In the current, high-interest rate environment, that rate is often in the range of 15-20%. When construction costs are high as well, , privately-financed developments have to borrow a lot of money at a high rate – and therefore can only build projects with rents that are high enough to cover those costs.

In affordable housing developments, lower rents limit the amount of debt that can be used to finance a project. This means that about 30-70 percent of project financing will need to be subsidized. The Low Income Housing Tax Credit represents the most popular way to provide that subsidy, by issuing tax credits that equity investors buy with the promise to keep the project’s housing affordable for a set term. When that term expires, the project will need another set of credits to refinance for affordability, or it will likely convert to market-rate rents. You can think of the LIHTC subsidy like instant yeast to generate affordable units.

By contrast, social housing is a sourdough starter. The revolving loan fund that underpins the program (and that the city just approved), is designed to be a renewable resource that funds project after project, returning back to the fund once the loan is paid off. Importantly, this can be used either to fund new projects from the ground up, or to jumpstart projects that were in a development pipeline, but have stalled out due to high costs, delays, or other factors.

The power of this revolving loan fund comes from its ability to offer low-cost, short-term financing to housing projects, replacing the double-digit interest rate that would otherwise be required by private financing. At the highest-level, you can imagine it this way: If GSH offered $25 million for 7 year’s worth of financing required to close a deal at 5 percent interest, as opposed to private equity’s demand for 15 percent returns on a similar amount, you’d save nearly $17.5 million on the costs of one project. That is certainly something!

It gets better. In a social housing scheme, the City can either convert its interest on GSH debt to equity or inject equity from another source. Both of these approaches would lower cash outflows, which would enable the project to safely cover operating expenses (like maintenance and property taxes), and generate a reasonable cushion of profit for use to accommodate GSH’s set-aside of affordable rentals.

That lower cost of capital, plus the higher profits from the building’s majority market-rate units cross-subsidize the lower profits (or losses) resulting from affordable units. After the short-term loan is paid off 5-10 years later (“revolving” back into the $135 million revolving loan fund), voila! The city has created affordable housing in which it holds a majority equity stake, and therefore ownership that ensures these units will maintain long-term affordability, without the need for outside subsidization.

This also helps address a key risk with other programs in our current affordable ecosystem: A lack of any return on government investments, and a disconnect between long-term capital needs and the fickle political nature of the budgeting process. Public housing units and vouchers both commit yearly appropriated sums to closing the gap between rent that individual low-income households can pay and location-specific fair market rents (FMR) determined by the U.S. Department of Housing and Urban Development, and both are underfunded. While vouchers do this through direct rental assistance payments that supplement a given household’s income, public housing does so by providing capital and operating grants necessary to maintain government-owned housing assets.

The problem is that the federal government only funds vouchers enough to cover about a quarter of current demonstrated need, and, additionally, vouchers result in no development of government equity or assets.² Public housing has suffered, meanwhile, because buildings actually require long-term and stable capital planning not accounted for in fluctuating year-to-year budgets. The current national backlog of public housing repairs is estimated at around $70 billion and climbing for a reason.³

Two final points about the structure of social housing efforts:

First, social housing is a tool for affordable housing and not a class of affordable housing. It can be combined with other programs (like LIHTC) to ensure development feasibility, to deepen affordability, and to achieve other development goals: for instance, combining revolving loan funds with proposed new powers of Northern Illinois Transit to encourage transit-oriented growth.

Second, social housing is premised on the initial thesis behind public housing in the 1930s: that public housing should be high quality, and serve residents at a range of incomes who the private market failed. Before we paved a road to hell with good intentions by concentrating low-income residents in public housing units, projects were built at such high quality and filled such a glaring need that they successfully competed for tenants whose combined incomes could cover public housing’s operating costs via rents. How much more preferable was public housing to the private market in the 1930s and 1940s? We threw a friggin’ parade to celebrate the opening of the Ida B. Wells Homes! If you’re curious to learn more about why this stopped being the case in most cities by the 1950s and 1960s, I recommend reading Gail Radford’s Modern Housing for America (1996), D. Bradford Hunt’s Blueprint for Disaster (2009), and Nicholas Bloom’s Public Housing That Worked (2008).

What would GSH look like and how would it work in Chicago?

Now, let’s get to the details of Chicago’s approach. We have a rough idea of what GSH would look like in Chicago based on contents of the final ordinance, so here are the key details before we dig in to how it would work for a newly constructed multifamily project:

• Developments have to consist of 30 percent affordable units for households making up to 80 percent of Area Median Income (AMI), with affordable rent guidelines set by DOH. The Board of the corporation, however, can permit some of these set-aside units to be affordable to households with incomes at 80-100 percent AMI if necessary to make some “deeply affordable” units (defined as being affordable to households making 30 percent AMI or less) pencil out. For reference, the 2024 Area Median Income for a household of three people in Chicago is $100,900, or what might equate to a dual-income household in which both partners earn about $50,000 and share custody of a child. The maximum affordable rent (combined with utilities) determined by the city for that household in 2024 stood at $2,811.

• Developments must comply with the sustainability code of the Department of Planning and Development, unless otherwise noted by the Board. DPD’s sustainability code applies to all buildings that receive city funds or zoning approvals. DPD’s code is not the same as the DOH’s ATSM, whose terrible effect on affordable housing costs has already been covered by this blog…so could be a plus!

• All affordable units, and 20 percent of market units, must be accessible as defined by Chapter 14B-11 of the Municipal Code. This is a strangely broad requirement to place on 44 percent of building units, given that only 10 percent of the population in Chicagoland identify as having a disability (the same is true of about half of all voucher holders), but I digress.

These standards are different, and generally more challenging than the ones that apply to Montgomery County’s Housing Production Fund:

• First, incomes in Montgomery County are much higher than incomes in Chicago. The median income of a family of four in Montgomery County was $42,000 dollars higher than the median income of a family of four in Chicago in 2024, at $154,700 versus $112,400. (Montgomery County, in fact, is one of the wealthiest counties in the United States.) Though Montgomery County has tighter income requirements for its affordability set-aside, the max affordable rent for a 3-bedroom unit for a family of four at 65 percent AMI ($2,614 in rent) is still nearly $400 per month higher than the max affordable rent for a 3-bedroom unit for a family of four at 80 percent AMI in Chicago ($2,228 in rent).

• Second – construction costs in Montgomery County seem to be lower than comparable construction costs in Chicago. Take the $388K per unit cost of the Montgomery County social housing development The Laureate,⁴ compared to the $400-550K project cost per GSH unit quoted in the graphic above.

• Third, the Housing Production Fund does not appear to have defined energy efficiency and accessibility requirements like those the GSH ordinance carries. This gives projects extra flexibility to ensure they can pencil out.

Put those factors together, and this means projects in Montgomery County will require significantly less financing support and have more flexibility from local government to close the gap between affordable rents and building costs. As the chart shows below, there is a healthy gap between rent collected for a household making 80 percent AMI in Montgomery County and the cost to finance a single $388K unit at 5 percent interest over 40 years. This provides a cushion to ease covering operating costs for units, whose affordability would be additionally ensured through loans that further lower the cost of capital. For a family of four making 65 percent of AMI in Montgomery County, rent payments also cover debt service but operating costs would need to be more heavily cross-subsidized by other units at market rents.

By comparison, the higher cost of constructing affordable units and lower household incomes in Chicago removes that cushion for cross-subsidy and requires drastically reduced interest payments to pencil out. Generously assuming that GSH units only cost the City of Chicago $550,000 each to build (a midway point picked between 2023’s average LIHTC cost and what the city estimates that GSH units should cost), a four-person household with 80 percent AMI in Chicago could not afford to cover the debt service for a unit at affordable rents financed at 5 percent interest over a 40-year period. Rents for a four-person household at 65 percent AMI in Chicago would only cover two-thirds of that debt service.

An exciting opportunity with minimal room for error

This is an exciting program, but Chicago’s version is going to have to be run really well in order to succeed. In part two, I’ll do a deeper dive into what that looks like in a pro forma, and talk about what the city needs to do in order to make this model work.

1. Unclear if this was meant to reference the George Stanley Halas initials on Chicago Bears jerseys, but a cool branding opportunity nonetheless.

2. Rental Assistance Demonstration projects excepted.

3. I realize that some of you might ask, “Well wait a second, what about inclusionary zoning?” Chicago of course has its own such program, which requires developers to set aside a certain amount of affordable units if they receive city land, entitlements, financial assistance, or other benefits like zoning variances to make a project pencil out. Theoretically, ARO units require no or minimal subsidies, which many cast as a boon for the program. But whether or not inclusionary zoning policies are truly effective is hard to determine because inclusionary zoning policies vary by locality and by the frequency with which they are automatically paired with subsidy programs; that therefore makes their ultimate effectiveness difficult to determine, too. My short answer to avoid a much longer discussion about inclusionary zoning in an article dedicated to social housing: Inclusionary zoning only works because status quo exclusionary zoning limits affordability from the outset. When the proverbial “carrot” merely represents a city deciding not to beat you with a stick, one wonders why the city would not just forgo that awful stick first.

4. You can find the Laureate’s value by checking its county assessment on this website. Searching for Map GS23, Parcel N601. The building was assessed at $104M, with 268 units. Since this was only one year after completing construction, I’m assuming the assessment and construction costs map nearly 1:1.