Talking about affordable housing

I’ve wanted to write a quick explainer about affordable housing for some time. Mostly, I hope to clear up confusion and misconceptions that fly around whenever the topic is discussed. Let’s see whether I can accomplish that.

  1. What is affordable housing?

The term gets thrown around a lot without a lot of understanding. The problem is that “affordable housing” seems self-evident. It’s right in the name! Except, not really, at least not from a policy standpoint. Technically speaking, “affordable housing” means housing with these three characteristics:

  1. Rents (or prices, though I want to focus on rental housing here) are restricted by deed. Usually, this means that rent is capped at 30% of someone’s pre-tax income. Side note here: in expensive coastal markets, people say “Below Market Rate” (BMR) to mean “affordable”. But in cheaper markets, affordable housing can be at or even above market rent!
  2. Tenant incomes are also restricted, and tenants need to income qualify. This can entail a ceiling and a floor. The income ranges allowed depend on the financing sources being used. For the biggest source, the Low Income Housing Tax Credit (LIHTC) program, incomes are restricted to 60% of a county’s median income. Tenants need to certify their income each year but usually do *not* have to vacate if their incomes go over the cap.
  3. The housing is built with some public or private subsidy, either via some state or federal program, or using cross-subsidy from “market rate” units.

So, affordable housing isn’t housing that happens to be cheap, or designed to be inexpensive, or a particular typology. When we talk about “affordable housing” in a policy context, we are generally talking about housing that falls under the criteria above.

2. Who lives in affordable housing?

As I mentioned above, when we talk about affordable housing, we’re talking about housing that is pitched at a particular income band. The band is determined by funding sources. If we’re talking about LIHTC housing, and we usually are, we’re usually talking about folks making around half the median income of an area. To put that in context, here in San Francisco, for a household of 4, 60% of median income is $69,200 a year (pdf). So here we’re talking about folks making in the $50k — $80k range depending on household size. Some may disagree, but I would frame that as a household with one low-to-moderate earner (SFUSD starting teacher salaries fall in this range, for example), or a household with two lower-wage earners. What your politicians refer to as “workforce”, basically. This is a large percentage of who lives in “affordable housing”, but not the only definition.

But there’s also housing that targets the deeply poor, and housing that targets folks who are more solidly in the middle class. The former type combines tax credits with other subsidy sources to support operating costs, because the residents cannot support property operations through rent. The latter housing is what you hear about when you hear about projects having an “affordable” component in places like San Francisco that have inclusionary requirements. These units range from 80% of median income to 110% ($90k to $120k for a family of 4), so you’re getting into households with one pretty strong earner or two moderate ones.

On top of all that, there’s special needs housing, senior housing, youth aging out of foster care, formerly homeless folks, and so on. Point being: there’s no real tenant profile here. It depends entirely on the particular project. Our economy leaves a lot of folks in need of help, and there’s a massive and complex (but inadequate!) patchwork of programs to try to provide that. So whatever your image is of who affordable housing “helps” is probably both right and wrong.

3. Who builds it?

A wide variety of developers produce “affordable housing”. The world of LIHTC housing has a mix of non-profit and for-profit players (much more the former in California). The for-profit developers who build LIHTC tend to focus on that market, since it is a specialized area. However, there are a lot of “market rate” developers who produce affordable housing as well, either in partnership with LIHTC-focused firms as part of larger projects, or non-LIHTC affordable units that they are required to build under inclusionary housing rules.

4. What is LIHTC?

There are a bunch of subsidy sources for affordable housing, but as I’ve mentioned, the largest source by far is the Low Income Housing Tax Credit. The basic idea of LIHTC is that the cost of constructing an affordable housing project generates tax credits. Under LIHTC, “affordable” means that resident income is capped at 60% of the area median. The amounts of these credits is calculated based on a formula. There are two types of credits: 4% and 9%. The basic idea is that for every $1 of development cost, the developer can issue either $0.04 or $0.09 of tax credits each year for 10 years. In California, the 4% credits are available to any qualifying project, while the 9% credits are allocated competitively. There are lots of details that complicate things, but that’s the general idea. These credits are bought by some entity with tax liability to shield their income. The proceeds from this sale become equity in the project.

LIHTCs can be bought by any entity with tax liability, but are frequently bought by large corporations, particularly banks. The credits sell based on market pricing, so the amount the buyers pay per dollar of tax credit moves with supply and demand. A variety of things can affect this pricing, like the strength of a particular housing market, or the number of projects seeking investors. But, critically, it also depends on the amount of corporate income that needs shielding. This means that in a weak economy, pricing on the credits can fall off a cliff. Similarly, changes to the tax code can have a big impact — that’s why pricing fell after Republican victories last November, which suggested lower corporate rates were a strong possibility. This was before we realized how incompetent this gang was, but I digress!

Here’s another important thing about the LIHTC program — the restrictions on the housing built using the credits are temporary. Currently, the law says that the units absolutely must remain restricted for 15 years, and then there is a 15 year “extended use” period during which the property owner can seek relief from the restrictions. In any case, after 30 years, unless the property is refinanced with a new set of tax credits, the affordability restrictions expire. (except in states with longer periods, and on projects with other sources of financing that carry longer restriction periods. In California, 55 year requirements are common.) In most markets, this doesn’t necessarily make a big difference since 30 year old apartments don’t rent for much more on the open market than the restricted rents. In high-cost markets, however, it’s a different story, which is why local money in these locations tends to carry longer use restrictions.

5. Where does the other money come from?

LIHTC equity is a big piece of the puzzle, but other sources of funding are needed as well, especially in high cost areas. These take a few forms,:

  • Other federal programs that “loan” money to projects on a competitive basis, or that support operating expenses. It’s not really a “loan” in the conventional sense, since full repayment is not usually required. There’s just a claim on some portion of whatever cash flow the project generates.
  • State and local money. Particularly in a high cost area like San Francisco, where I live, this is a big piece of the pie. This can be a grant, but more often it is a “soft loan” that doesn’t require full repayment, but has a claim on some of the project’s cash flow. This funding also tends to have a lot of extra requirements attached, depending on state and local priorities. For example, California directs some of the proceeds of its cap-and-trade program to funding affordable housing. This funding is distributed competitively, and the decisions are driven largely by how well a project does at reducing carbon emissions. As another example, local funding in the Bay Area tends to require projects to provide special needs housing of some sort (e.g., housing for formerly homeless individuals), and often also has local hiring and prevailing wage requirements.
  • Good old bank debt. Most projects will support some amount of regular old bank debt. The amount of debt that a project can get depends on how much operating profit it can generate. Typically in affordable projects, this is a fairly small amount, as the restricted rents don’t necessarily exceed operating expenses by very much. Affordable projects that are targeted to families with very, very low incomes (like under 30% of median income) may not generate much operating profit at all. These types of projects will often have operating subsidy support to make the numbers work, like Section 8 vouchers that fill in the gap between what a tenant can pay, and what the government believes a fair rent to be for an area, or subsidies that cover some chunk of a property’s expenses. These subsidy sources are then used to support the borrowing needed to build the project.

6. Putting it all together

I learn best through examples, so now that we’ve talked about the basics, I’ll work through a very simplified example of a project and how the money gets put together.

Let’s stick with LITHC housing for the moment. Here’s the deal:

  • We’re going to build 100 units of affordable housing, targeted at 50% of median income.
  • Let’s say we are building 30 1 bedrooms, 50 2 bedrooms, and 20 3 bedrooms.
  • The total development cost is $500k per unit, so a total of $50 million. We are getting the land for free from our generous city government!
  • Our operating costs are a skinny $6,500 per unit.

OK, so how do we fund the project? Well, let’s assume that all of our construction costs are basis eligible, which means that we can get credits for all of our costs. Let’s go after 4% tax credits, which are non-competitive. We could think about 9% credits, which are more lucrative, but they are allocated through a competitive process, a key part of which is minimizing the costs you get credits against. So for simplicity and for the project’s feasibility, we can stick with 4% credits.

As stated before, the 4% credit is supposed to provide $0.04 in tax credit for every dollar of cost, in theory. The reality is trickier. Rather than a straight 4%, the program provides credits according to a “credit rate” determined by the IRS, which is based on the government’s cost of borrowing. So, in recent times as interest rates have been extremely low, the credit rate has been low as well. Currently, rather than 4%, it sits at 3.23% (as of Oct 2017, it changes monthly). For our project, we will assume a “credit rate” of 3.25%. You’ll recall that the buyer gets 10 years of credits. Let’s assume that the credits are priced at $1. The amount of equity we get is:

$50,000,000 * 3.25% * 10 years * $1.00 = $16,250,000 of equity.

The next big source of funds is conventional debt. To get to this, we need to work out how much profit the building can generate. Here is the rent table for San Francisco:

With our unit mix above, that gives us annual potential revenue of $1.54MM. Applying a 5% vacancy rate, that goes to $1.46MM.

Then we said that our expenses were $6,500 annually per unit. This is a total of $650,000 for the 100 units. That means that our operating profit is $816,000 annually. That’s what we can use to support debt.

Banks calculate debt based on a ratio of income to the size of the required mortgage payment. For affordable projects, that ratio can be as low as 1.15, i.e., the income needs to be 1.15 times the mortgage payment. That gives us a possible payment of $707,000 annually.

How much debt can that payment support? Depends on the rate and term of the loan.Without getting into too much detail, let’s say that the loan has a rate of 5% and it’s 30 years. In that case, an annual payment of $707,000 can support a loan of $10.9MM.

So we have about $16.25MM of equity, and $10.9MM of debt for our $50MM project.

As you can see, this leaves a pretty big gap, amount to over $200,000 a unit. In order for this project to happen, that money needs to be cobbled together from a bunch of different public pools, usually competitive.

7. So is that why we don’t get that much affordable housing?

Pretty much. At least in areas with high construction costs (like the Bay Area), affordable development is far from self-supporting. The numbers above were a gross approximation, but in the ballpark of what deals tend to look like. This means that affordable developers need to chase money from a bunch of different sources, at the federal, state, and, in well-off places, local level. Each source comes with its own application process and timeline, and its own scoring criteria and conditions of funding. This is time-consuming for the developer and delays projects getting done, especially because a number of funding sources require that you show they are the last bit of money you need, which leads to a weird dance as projects juggle many processes simultaneously. It also means that projects are loaded up with features demanded by each source — unit mixes, lower income brackets, “green building” and the like. None of these are bad, but loading several of them onto individual projects is questionable, and again, makes these projects harder to build.

In a perfect world, we’d streamline the funding and make it more direct — why go through the banking sector? — and we’d also work to bring construction costs down so that public money generates as much housing as possible. We aren’t close to that world yet. But I think the first step is to understand what we talk about when we talk about affordable housing. Hopefully this is a start.

BGI -> POS -> NJ -> NYC -> SFO