Construction costs for affordable housing are skyrocketing

Tradeoffs are in fact real

It won’t be news to regular readers around here, but Chicago has a housing crisis. The DePaul Institute of Housing Studies estimates that the city is short 120,000 units of low-cost housing. Estimates of the number of homeless Chicagoans more than tripled last year, according to the 2024 HUD point-in-time count.

As I write, the temperature in Chicago is forecast to hit a low of -7 degrees.

At the same time, it’s getting much harder to produce new units. Costs for city-funded affordable housing projects have skyrocketed. In 2023, city supported affordable projects cost an average of $584,000 per unit.¹ And that number includes rehabs of existing affordable units. Rehabs are great, and generally cheaper than new construction, but they also don’t add to the city’s overall housing stock.² Costs for new construction in 2023 came in at $747,000 per unit. Those prices are growing fast; costs for both rehabs and new construction have almost doubled since 2020.

It’d be bad enough if higher construction costs meant higher taxes. But the problem is even worse than that. Unlike Medicaid or Food Stamps (SNAP), federal funding for housing isn’t an entitlement that automatically grows with demand. Instead, states and local governments receive set outlays subject to the annual appropriations process. I wouldn’t bet on this Congress or President increasing that funding any time soon.

For new construction, Chicago’s primary federal funding source is the Low-Income Housing Tax Credit, which provides a fixed annual subsidy to Chicago. That means that in the long run, when costs go up, we don’t spend more money to build the same number of units. We spend the same amount of money to build fewer affordable units.³

The real losers here aren’t taxpayers (who don’t pay more). Developers also do ok – they just build fewer projects at higher price tags. And because rents are fixed, the residents who move into projects that do get built are also unaffected.

So who’s really paying for higher construction costs?

The Chicago residents who miss out on projects that don’t get built. If costs had held steady over this period Chicago could have built almost twice as many affordable units last year. That’s 360 families that won’t get housing in 2023 alone.

If someone proposed cutting spending on affordable housing in half they’d be vilified, and rightly so. But cost increases with the same result have barely registered. Let’s figure out what’s going on.

What’s making affordable housing so expensive to build?

There are a few reasons new affordable units are so expensive. Judith Crown has an excellent run-down of some of the big issues in a Crain’s article from last year. But you can really boil it down to three big challenges:

I. The structure of the LIHTC process

The Low-Income Housing Tax Credit is an extremely roundabout way to pay for new housing. The federal government allocates tax credits to states (plus New York and Chicago) based on population.⁴ The city then makes awards to developers as part of a Qualified Allocation Plan (more on the QAP in a minute).

Developers generally turn around and sell those tax credits to investors to finance the project, often with the help of a syndicator who matches projects to a group of potential investors. Then investors enjoy the benefits of the tax credits (which they can deduct dollar for dollar against their federal taxes) for ten years.

Source: The Tax Foundation

This process is expensive – each step adds transaction costs and markups. In a 2009 paper Michael Eriksen finds that developers have to sell the credits at an average of 73 cents on the dollar, to avoid the risk of being caught wasting tax credits they can’t use. And because the credits subsidize a percentage of construction cost, instead of a flat-per unit amount, they also tend to incentivize more expensive units. Eriksen finds that in California, LIHTC subsidized properties are 20% more expensive than standard industry estimates. Note: An informed reader points out that recent tax credit pricing is more favorable - generally in the 85-90 cent range. There are still discounts, and the process is still costly, but it may not be quite as costly as it was in California 15 years ago.

While LIHTC is the city’s biggest funding source for new affordable units, it doesn’t cover the full cost of a building. Developers have to cobble together a range of different financing sources in order to get the project to pencil – including TIF dollars, loans from the Chicago Department of Housing and Chicago Housing Authority (if CHA tenants are included), other private loans, grants and more. Each of those sources comes with their own costs and requirements.⁵

You can chalk much of this up to the vagaries of the tax code (and the politics of roping in a bunch of constituencies and framing the thing as a tax cut). But let’s not let the city off the hook completely. When Chicago announced its 2024 LIHTC awards, the press release declared that "proposals will now undergo project underwriting, design review, zoning approvals, and City Council approvals for any public subsidy. In general terms, closing on these projects is anticipated within the next 18 to 36 months."

That’s a lot of weasel words to admit that these projects are years away from obtaining the necessary approvals to proceed. Long delays mean that construction costs will continue to rise (more on that in a minute), and the cost estimates DOH approved are virtually useless. Chicago has taken a convoluted process, and added more transaction costs – lawyers to file zoning applications, time spent in public meetings, and wrangling over other city funds. This timeline is also not normal – Oregon reserves the right to pull tax credits if developers can’t close within 300 days.

Instead of making these announcements in a vacuum, LIHTC awards should come with the full set of available local financing to allow projects to close quickly. And affordable housing projects should have to go to city council once – rather than requiring separate signoffs for tax credits, zoning changes, and other funding streams.

II. Rising construction costs

While developers wade through through years of paperwork, construction costs continue to rise. That began with Covid-era supply shocks which rippled through the economy, driving up prices for labor and building materials. It’s continued thanks to a strong economy and a building boom, which has driven higher construction costs nationally. In total, the producer price index for construction is up 36% nationally since 2020. For context, consumer prices (CPI) are up 23% over this period.

Source: Bureau of Labor Statistics, via St. Louis Fed

City Hall doesn’t have much control over the price of lumber. But I think it’s also a mistake to accept these rising costs as a fixed variable to be passed on to low-income renters in the form of fewer units. As I’ve written previously, we could do more to flex parts of the building code (especially around items like single stair and elevators) and embrace standardized designs to minimize cost increases.

III. Everything else

But even accounting for rising construction costs and LIHTC overhead, the numbers still don’t add up. The city’s costs have risen more than twice as fast as that 36% figure. Not every city is suffering from the same issue we are. Costs for publicly-funded affordable projects in Houston, for example, are growing much more slowly than they are in Chicago:

There are a lot of reasons why Houston’s projects are cheaper. Looser land use restrictions make it easier and cheaper to build. Houston also has much lower construction costs than Chicago overall; according to RS Means data from 2021, construction costs here are 38% higher than they are in Houston. But those aren’t new factors, so they don’t explain why costs have nearly doubled here, while costs in Houston have increased more slowly than inflation.

One thing that did change in 2021 and again in 2023 is Chicago’s Qualified Allocation Plan (QAP), which sets out the process for awarding the city’s tax credits. It’s a detailed document that reflects a great deal of knowledge and care about affordable housing in Chicago. In particular, it sets different goals for different parts of the city – emphasizing a greater percentage of highly affordable units in high-income parts of town, and a greater emphasis on retail and moderate-income housing in lower income areas. That seems like a really good idea in a city highly segregated by race and income.

But the QAP also includes a wide range of preferences and requirements for new projects:

• Permanent Supportive Housing set-aside: In ‘priority’ tracts, projects must set aside 5% of units for permanent supportive housing (PSH). These units must be reserved for individuals who need long-term support, generally for acute conditions (such as prior chronic homelessness, or disabilities). Those services “must be appropriate to the needs and preferences of residents and be available either on-site or closely integrated with the housing.”

• Transit subsidy preference: “DOH encourages developers to provide tenants with a choice of low- or no-cost: 1.) transit (CTA) or 2.) bikeshare (Divvy) passes to new tenants of ETOD developments in the 2023 QAP funding round.”

• Sustainability: In addition to meeting all federal, state and local standards, which now include all-electric construction, the QAP announces a preference for: “innovative building practices that substantially reduce greenhouse gas emissions,” “Above-code insulation and energy efficiency measures that take meaningful steps towards net-zero energy use,” on-site renewable energy and storage, and onsite resiliency elements (“such as flood and extreme heat mitigation and backup power sources”).

Of course, these are just the requirements that are *new* in 2023. Applicants already must meet all standard city contracting rules, as well as the rules outlined in the prior 2021 QAP. Those include meeting targets for including Minority and Women-Owned businesses, a demonstrated commitment to including Black, Indigenous and People of Color (BIPOC) on the development team, the provision of additional supportive services (such as financial counseling, arts, wraparound services, etc.), and high-quality designs that “exceed Federal and local requirements for accessibility.”

Costs are eventually mentioned at the end of the selection criteria. DOH notes that it will look for “reasonable final construction and projected operating costs compared to similar projects.” But there is no indication of what costs are ‘reasonable,’ and no indication that a project that comes in below last year’s average will be more likely to win tax credits.

We’re not done yet. Projects also have to comply with DOH’s Architectural Technical Standards Manual, which function as a second building code for affordable projects. It’s full of extremely prescriptive requirements that leave little to no flexibility if costs increase. For example, costs for kitchen cabinets have spiked since 2020. But good luck trying to design around that: city standards specify the number of required lineal feet (both base and upper) by apartment size, the materials used on cabinet doors and drawer fronts (wood or high-pressure laminate), construction methods for those drawers (“dovetail or reinforced joint construction”), and the cabinet box construction (solid plywood - “no MDF or laminate boxes, cabinet drawers or door fronts allowed”).

This process raises costs in two ways. The first is mechanical – requiring above-code energy efficiency measures means paying for triple-paned windows to be installed by a qualified contractor that meets all the city’s other requirements. But the allocation process also raises costs above and beyond the prescribed standards. When city is clear about its preference for design or environmental efficiency, but is much less clear about what projects should cost, developers push the envelope on what they think the city wants. In her reporting, Crown describes an “arms race” among developers to impress the Housing Department:

Teams often go substantially above and beyond the city code to win tax credits and other funding, says Block of Evergreen Real Estate. Such add-ons can help save utility and operating expenses for tenants and managers, but the difference between a silver and gold LEED-certified building is negligible, he says.

Developer Peter Holsten, president of Holsten Real Estate Development, adds, "There is a lot of architectural ego that goes into these developments. They're pretty high-profile."

Every one of these requirements comes with good intentions. They’re also informed by prior misdeeds. The history of public housing in Chicago is littered with projects that prioritized cost (and political expediency) over tenants. DOH requires high quality design, security features, and accessibility for a reason.

We also want policy objectives to be coordinated across departments. In some cases, that may include sacrificing the efficiency of one goal for another. The city has long deployed Streets and Sanitation crews in high-violence neighborhoods during the summer months to add eyes on the street and keep streets safer. It’s probably not the most efficient way to plan roadwork. But it’s almost certainly worth it if makes an impact on violence.

But how should we balance those tradeoffs? Which priorities come first? How much more is it worth spending to achieve a 10% reduction in building emissions? Is it more efficient to add a slice of permanent supportive housing units in every building, or to prioritize separate PSH projects (which might make it easier to provide on-site services)?⁶

And when the number of requirements add up, they start to operate at cross purposes. It’s important to foster developers who have historically been excluded from city projects. City contracting preferences are designed to help new entrants get a start in the industry. But if contracts for city projects are wildly different from what gets built in private markets, new developers will remain at a disadvantage when they compete for market rate projects.

This tradeoff between priorities is even more perverse when it comes to green building standards. Humboldt Park Passive Living, a project approved by DOH in 2023, is built to new passive house energy standards that builders claim will yield energy savings of 33-50%. But the project also cost $743,554 per unit. Meanwhile, the lowest-cost new construction project in 2023, 43 Green Phase II, came in at “just” $551,324 per unit. The total price tag for the two properties was about the same (roughly $44 million). The difference is that 43 Green will create 80 units of new housing, while the passive house project will only deliver 60.

That means 20 fewer families will move into a new unit built to Chicago’s current exacting energy code. Instead, if they do have housing today, they’re likely living in a much older, less energy-efficient building (the median age of Chicago’s housing stock is 72 years). Based on some simple math from the Department of Energy’s Building Performance Database, the missed savings from not upgrading their housing wipes out more than three quarters of the energy savings from the new passive house construction.⁷

Ezra Klein has referred to this problem as the ‘Everything Bagel’ theory of public policy:

You might assume that when faced with a problem of overriding public importance, government would use its awesome might to sweep away the obstacles that stand in its way. But too often, it does the opposite. It adds goals — many of them laudable — and in doing so, adds obstacles, expenses and delays. If it can get it all done, then it has done much more. But sometimes it tries to accomplish so much within a single project or policy that it ends up failing to accomplish anything at all.

I’ve come to think of this as the problem of everything-bagel liberalism. Everything bagels are, of course, the best bagels. But that is because they add just enough to the bagel and no more. Add too much — as memorably imagined in the Oscar-winning “Everything Everywhere All at Once” — and it becomes a black hole from which nothing, least of all government’s ability to solve hard problems, can escape.

Let’s put tradeoffs on the table

I’m not going to pretend that I know which preferences to tweak to hit the right balance between other policy objectives and housing units. But I do think that there’s a better framework to make these decisions – for housing, and for other public policy challenges.

Instead of listing a long set of preferences, we should quantify the tradeoffs we’re willing to make. How many units are we willing to sacrifice for a project that comes with a transit subsidy? Or one with net zero carbon emissions? If we’re willing to build 10% fewer units to achieve another goal, then we should put that in writing.

Texas, which allocates tax credits for projects in Houston, has an approach that looks like this. Projects must meet all other Federal, State and Local requirements (such as code and Americans with Disabilities Act compliance), plus clear a baseline set of standards.⁸ From there, proposals are scored based on their characteristics: projects get extra points for being located in higher-income areas, adjacent to transit, and for maintaining affordability covenants beyond 30 years, for example. Cost is part of the scoring criteria: projects can receive up to 12 additional points for falling below different cost-per-square foot thresholds.

This makes tradeoffs explicit. It quantifies what the State is willing to pay more for. It generates real information about what add-ons are costly, and which ones are easy to add.⁹ And when costs do spike, it provides flexibility to trade off one feature or amenity for another, minimizing the impact on housing production.

The Ella Grand, a 128-unit affordable senior housing development in Houston approved in 2023 at a cost of $292,372 per unit.

Of course, Chicago’s version would have a different set of requirements. We’d still want to have different preferences by neighborhood, and we’d almost certainly include a broader set of preferences overall. But we could also be clear that developers will be rewarded for proposing projects with lower costs per unit. Given our higher construction costs, we might want to add a heavier weight to cost containment than Texas does.¹⁰

Time to act

The good news is that the opportunity to do so is right around the corner. The city is due to release a new Qualified Allocation Plan in early 2025. It wouldn’t be difficult to follow the lead of Texas, and other states that make costs and tradeoffs explicit. But in the 2023 plan, the DOH signaled its intent to do the opposite – and make the rules even more onerous going forward. That’s a recipe for higher costs – and more families out in the cold.

We can still deliver high quality projects that reflect the progressive ambitions of the city. But we need to set priorities and acknowledge tradeoffs. Chicago is facing a housing crisis. It’s time to act like it.

1. Here, we’re talking about city-funded units that must be rented at below market rates to individuals with low or moderate incomes.

2. Unless the units would otherwise be demolished or unlivable.

3. There are of course other funding sources, including city-funded bonds and tax increment financing (TIF) dollars. But the city is facing a budget crunch as it is – Conor wrote about the S&P credit downgrade earlier this week. The city’s new economic development bond is great, but it also won’t grow in response to higher construction costs. In the long run, a 10% increase in construction costs mean we build 10% fewer income-restricted units.

4. Chicago and New York are the only cities that receive their own LIHTC allocations, distinct from their state Housing Development Agencies. That’s due to the efforts of Congressman Dan Rostenskowski, a clout-heavy appropriator and architect of the program. Rostenkowski stayed on as committeeman of the 32nd Ward even while serving in Congress, and is another Illinois politican whose career ended via federal indictment.

5. This also makes affordable developments more vulnerable to uncertain project approval processes or development delays – with a far more intricate capital stack, there are far more opportunities to miss a deadline or lose a key piece of financing.

6. Notably – there is a separate funding line in the QAP for projects that include a higher percentage of PSH units.

7. That math: An energy efficient multifamily/mixed use building constructed in Chicago in the last 5 years consumes 40 kBtus (a unit of energy) per square foot per year. Assuming that a passive house cuts that number in half, we save 20 kBtus for the 60 households, or a total of 1200 kBtus/psf/year. But the median building constructed between 1950 and 1955 consumes more than twice as much energy: 87 kBtus/psf/year. Since we forgo 20 units of housing, we miss out on the opportunity to upgrade 28 households from 87 kBtus to 40 kBtus. That’s a foregone savings of 940 kBtus/psf/year. If you assume that the folks who miss out on low-income housing are living in units with below-average insulation, or that they could be housed for less than $550,000, the climate benefits of the passive house look even worse.

8. Some of these are straightforward (like square footage by unit type, or kitchen appliances). Some of them are Texas-specific (like prohibiting units within two miles of oil refineries). But even for items like playgrounds and common space, projects aren’t required to offer specific amenities but instead earn enough qualifying points across a range of options.

9. This is another argument for pricing – the proposals developers submit will reflect their real cost tradeoffs, as opposed to cheap talk. If net zero buildings don’t actually cost much more to build, developers should be happy to meet them for a few points on the scorecard.

10. My understanding is that Texas’s approach also has its own issues – in particular, a rigid formula can encourage lawsuits from developers who just miss the cut for tax credit allocation. You may need to require developers to waive their rights to sue and include a discretionary block of points (say 20%) to give DOH discretion to support projects with other benefits or features.