Skip to Content
Skip to Content
March 10, 2014

Affordable Multifamily Housing: Weighing the Options

David Brickman
David Brickman, EVP Multifamily Business

Significant debate around multifamily housing policy focuses on the question of whether the federal government should support affordable multifamily rental housing and, if so, how the government should encourage private capital to flow toward it. While important, that question tends to overshadow the equally important, but more nuanced, question of what the need is for affordable housing, how to help meet it most effectively, and what trade-offs come with choosing one policy approach over another.

The definition of what an affordable housing unit is, and what makes any given unit affordable – the rent level, tenant, or income level in the surrounding area – is central to the answer.

The first step toward answering the question is to clarify the policy objective. Is it to help low-income renter households? Increase apartment supply in general to bring down rents? Improve the quality of the rental housing stock?  Or reduce the cost burden on renters who otherwise might be forced to spend an outsized share of their incomes on rent?

There’s no right or wrong answer – it’s a choice for policy makers. But each of these objectives calls for a somewhat different approach to meeting it. Each one has implications and involves trade-offs.

Before we dive in, some key multifamily market facts:

  • Today, 35 percent of U.S. households are renters. The number of renter households has risen for nine straight years. Broad demand for multifamily rental housing is expected to keep rising.
  • The U.S. Department of Housing and Urban Development (HUD) deems a unit affordable if the tenant spends no more than 30 percent of income on rent.
    • Moderate-income rental housing is affordable to households earning 100 percent of area median income (AMI) or less.
    • Low-income rental housing is affordable to those earning 80 percent of AMI or less.
    • Very low-income (VLI) rental housing is affordable to those earning 50 percent of AMI or less.
  • More than half of renters spend more than 30 percent of income on rent, up from about 40 percent of renters in 2000, according to the U.S. Census Bureau.
  • According to Harvard University’s Joint Center for Housing Studies, between 2001 and 2011, housing-cost burdens grew fastest for moderate-income renters.

To boost multifamily housing stock in general, policy might, for example, subsidize building new units by lowering developers’ cost of capital. More availability would ease pressure on rents broadly, but households in most need of relief might see relatively little benefit.

Specifically reaching low-income renters or renters with heavy housing-cost burdens would be more complex – and clearly require more targeted intervention.

Without government subsidies, like federal Low-income Housing Tax Credits (LIHTC), little low-income multifamily housing is built. The business case for building affordable rental units is a tough one to make otherwise; costs almost always outweigh potential revenues. And without rent or income restrictions, housing labeled low-income often offers low rents just because of its inferior location, unit size, quality, and/or condition. If multifamily policy simply aimed to channel funds where rents and incomes are lowest, the results largely would run counter to market needs. In this case, the biggest subsidies would disproportionately flow to smaller markets, including exurban and rural areas – where demand for multifamily rental housing tends to be less than in denser urban areas, land relatively inexpensive, and development costs comparatively low.

What about directing subsidies to areas where households spend a high proportion of their income on rent? This presents challenges, too. Large metropolitan areas tend to have the highest demand for multifamily rental housing. They also tend to be high-cost, with high median rents but not necessarily high incomes. Why? In urban areas, land is scarce and more expensive and construction costs are higher; consequently, rents are higher than in other markets. Not surprisingly, under current AMI-based policy, these areas have few affordable units eligible for subsidy but many cost-burdened renters. On the other hand, where it’s less expensive to build and rents already are more in line with incomes, more units qualify for subsidies.

Let me illustrate with an example from Freddie Mac Multifamily’s research paper, Multifamily Affordability: Market Conditions and Policy Perspectives. AMI is measured across specific geographic areas, which may span rural and urban populations. In the Washington D.C., Metropolitan Statistical Area (MSA), AMI in 2012 was $107,500. At the county level, AMI ranged from $71,290 in rural Warren County, Virginia, about an hour’s drive from D.C. to $137,216 in Arlington County, Virginia, just across the Potomac River from D.C. Under the current rule, the VLI threshold was $53,750 for the whole D.C. MSA. Therefore, a unit occupied by a renter earning far more than half of AMI in Warren County might qualify as VLI housing. A renter with the same income living in densely populated Arlington, where rents tend to be much higher, might not have access to VLI housing and is more likely to be cost burdened.

As our paper describes, breaking metropolitan areas into chunks, such as by ZIP code, would allow a more refined approach to identifying affordable housing needs and to qualifying units as affordable in those localities. HUD already uses this approach for its Section 8 Housing Choice voucher program.

The discussion around rental housing finance reform continues. Determining the underlying policy goals and how best to allocate capital to support them would better frame the debate and promote a more productive, meaningful reform process. The outcome will affect not only the multifamily housing finance industry, but millions of people’s quality of life.

  • Feedback

    Have a comment or question about this post? Email us to let us know what's on your mind.

    Maximum of 250 characters.