What Is the Quality of America’s Assisted Housing?

What Is the Quality of America’s Assisted Housing?
Sandra Newman, Scott Holupka
Publication Date:
Find Full Text

As part of its role in designing, implementing, and monitoring most of America’s assisted housing programs, the US Department of Housing and Urban Development (HUD) must ensure that its residents live in safe, physically adequate dwellings. What is the state of the quality of America’s assisted housing? This study uses data from the 2011 American Housing Survey to answer this question. The authors acknowledge that “housing quality” has no definitive criteria. They base their measures on the dwelling’s physical integrity (e.g., holes in the floor) and housing systems (e.g., heating breakdowns) rather than factors reflecting the resident’s housekeeping or behavior (e.g., leaving unsafe cleaning products within a child’s reach).

First, the authors developed three composite indexes of housing quality—a market value index, a consumer rating index, and a normative standards index—and tested their validity. Each of these measures includes 33 housing conditions. Next, the authors applied these indexes, along with HUD’s long-standing housing quality index, to the assisted and unassisted stock to identify disparities in quality. They also identified the locations, household types, and assisted housing programs most likely to experience housing problems. As a final step, they examined how well assisted housing in good condition aligns with HUD’s Fair Market Rents (FMRs), or the 40th percentile of rents in each housing market.

Key findings

  • Fifty-five percent of US rental housing stock had none of the 33 quality problems in the analysis, and fewer than 5 percent of rental units accounted for more than 75 percent of problems.
  • Despite its intuitive appeal, the market value index performs poorly. In contrast, both the consumer rating index, based on modeling the respondent’s house rating on a 1-to-10 scale, and the normative standards index, based on a factor analysis, perform well and achieve convergent and predictive validity.
  • The quality of assisted housing is comparable with the quality of unassisted housing based on these two indexes.
  • There is little variation in housing quality among the different assisted housing programs.
  • The Northeast region and households with a disabled member have lower housing quality than average.
  • The analysis supports the FMR because the 40th percentile of rents is roughly the inflection point for maximizing assisted housing quality on both housing quality indexes tested.