GAO Releases Report on National Housing Trust Fund

The Government Accountability Office (GAO) released Affordable Housing: Improvements Needed in HUD’s Oversight of the Housing Trust Fund Program (GAO-23-105370) on August 8. The GAO report provides information about 70 projects completed with support from the national Housing Trust Fund (HTF) in 12 states, providing details about activity types (e.g., new construction or rehabilitation), number of bedrooms, racial/ethnic composition of occupants, targeted populations, per-unit costs, and additional funding sources used to develop the projects. The report also discusses weaknesses in the oversight and monitoring of HUD that have little practical relevance to the uses or benefits of HTF resources.

The GAO report was issued in response to a request sent on April 14, 2021, by Representatives Patrick McHenry (R-NC) and Steve Stivers (R-OH). At the time of the request, Mr. McHenry was the Ranking Member of the U.S. House of Representatives’ Committee on Financial Services, and Mr. Stivers was the Ranking Member of the House Subcommittee on Housing, Community Development, and Insurance. Currently, Mr. McHenry is the Chair of the House Committee on Financial Services, and Mr. Stivers is no longer a member of Congress. NLIHC sent GAO a reply on November 15, 2021, addressing most of the questions posed by Representatives McHenry and Stivers. 

The HTF allocates funding by statutory formula each year to states to build, preserve, rehabilitate, and operate rental housing for extremely low-income households (ELI) – those with income less than 30% of the area median income (AMI) or less than the federal poverty line, whichever is greater. The authorizing statute requires 90% of HTF dollars to be used for rental housing, and currently 100% of HTF-assisted units must be occupied by ELI renter households. HUD’s Office of Affordable Housing Programs (OAHP) within its Office of Community Planning and Development (CPD) oversees the HTF program.

GAO selected 12 state HTF grantees and reviewed 70 of their completed HTF projects, accounting for 42% of the HTF funds granted nationally in 2021. GAO reviewed HUD data on 2,186 completed HTF-assisted units as of March 1, 2022, making it possible for the report to describe those projects.

GAO found that 57% of the completed units were newly constructed, and 42% were rehabilitated, with 1% simple acquisitions. GAO also looked at 4,275 units in development, finding that 64% would be newly constructed and 35% would be rehabilitated. NLIHC obtained project-specific information from each state and the District of Columbia reflecting projects awarded funds from their 2018, 2017, and 2016, (2016 Supplement) HTF allocations. This information indicated planned (not completed) projects. NLIHC’s review of projects planned for rehabilitation found that all but a few projects should more accurately be portrayed as substantial rehabilitation essential to preserve existing federally assisted housing from leaving the affordable housing stock, along with some adaptive reuse projects. CPD uses an online management information system, IDIS, which only allows a state to indicate whether a project is “rehabilitation,” “acquisition and rehabilitation,” “acquisition only,” or “new construction.”

GAO also reports that of the 2,186 completed HTF-assisted units:

  • 41% had one bedroom, 16% were efficiency units, 26% had two bedrooms, 14% had three bedrooms, and 3% had four or more bedrooms.
  • 61% were rented by White households, 28% were rented by Black households, and 11% were rented by Latino households.
  • Most occupied units received some form of rental assistance, with only 26% not receiving any rental assistance.

GAO found that many states aligned their HTF application process with those of their other multifamily housing programs, such as their Low-Income Housing Tax Credit (LIHTC), HOME Investment Partnerships program, or state program application, eligibility, and ranking processes. (NLIHC has made note of this practice as well.) GAO found that non-HTF funds generally represented a larger proportion of projects’ total funding than HTF funds; NLIHC has also noted that for most projects, HTF is an important source of gap financing. GAO indicated that LIHTC equity supported 47 of the 70 properties in its sample, representing about 40% of total funding. Various private sources of funding constituted about 27% of total funding, while state and local funds provided 19% of total funding. A mere 5% of the funds derived from the HOME program, the Federal Home Loan Bank System’s Affordable Housing Program (AHP), and state or local contributions of their federal Community Development Block Grant program.

Nine of the 12 states reviewed by GAO awarded HTF funds to projects that targeted special needs populations, such as people experiencing homelessness, formerly incarcerated people, people with disabilities, older adults, and veterans. NLIHC’s reports for 2016, 2017, and 2018 reflecting all states and the District of Columbia show similar results. GAO reported that nine of its 12 states that targeted special needs populations also prioritized the development of permanent supportive housing. NLIHC has also identified significant numbers of HTF-assisted projects devoted to permanent supportive housing, likely explaining GAO’s reporting that 57% of the completed units in its survey had one bedroom or were efficiency units.

GAO reports that North Dakota and California set aside some of their HTF funding for projects developed by members of Indian Tribes or that are located on reservations or in other tribal areas. NLIHC’s reports indicated that North Dakota and South Dakota had set-asides. For those three early years, California did not have a Native American set-aside. GAO also indicates California and Utah set aside 20% of their HTF funds for rural areas, while Tennessee and Utah award extra points for rural applications. When NLIHC reviewed the HTF allocation plans or related program materials for these states in 2016, 2017, and 2018, they did not have rural set-asides or significant extra points for rural areas.

GAO found that the average overall per-unit (non-HTF-assisted and HTF-assisted unit) development cost of the 70 projects was $232,000. In addition, nine of 11 HTF-assisted projects that were similar to LIHTC projects in a prior GAO report had per-unit development costs within the range of the LIHTC comparison projects. The average per-unit development cost for new construction projects was $74,000 higher than for rehabilitation projects. The average per-unit development cost for projects developed by nonprofit entities was $40,000 higher than for projects developed by for-profit entities. To explain the difference, GAO refers to studies observing that nonprofits focus more on populations that are more costly to serve, such as special needs tenants who may require additional or enhanced facilities. In addition, nonprofits tend to focus on smaller projects that in general have higher per-unit costs due to the inability to benefit from economies of scale.

GAO also addressed several CPD-related technical issues that have little practical bearing on how HTF is used or whom it benefits. One issue, however, is of particular interest. The McHenry/Stivers letter to GAO criticized the HTF program, claiming it cost $1 million per completed unit of housing. As NLIHC explained to GAO, the “Leveraging” section of CPD’s National Housing Production Reports used misleading terminology. It has a term “Total Dollars for Completed HTF Units” and a corresponding numerical figure; that numerical figure confusingly reflects both the amount of HTF dollars plus other (non-HTF) dollars in projects. A different table on page 3 provides the total number of completed HTF units for a given month. Simply dividing the reported number of completed HTF-units into the misleading “Total Dollars for Completed HTF Units” results in greatly inflating the perceived HTF cost per-unit because most HTF-assisted projects contain relatively few HTF-assisted units. The actual average HTF cost per unit for completed projects for the month cited by the McHenry/Stivers letter was $113,552, with average HTF costs per unit in subsequent months declining to an average of $97,226.

CPD explained to GAO that the template used for the National Production Reports is used for multiple programs and not easily customizable. Therefore, CPD will add explanatory language to the National Production Report webpage. NLIHC notes that this fix depends on the public reading that webpage closely and perhaps simply reviewing each month’s report – overlooking the explanation.

The other technical CPD monitoring weaknesses identified by GAO are as follows:

  • Because CPD does not review grantees’ final drawdown and completion dates, it has been unaware of grantee noncompliance with and confusion about the requirement to enter project completion information into IDIS within 120 days of the final drawdown of funds.
  • CPD’s data on the total number of units (non-HTF units plus HTF units) in completed HTF-assisted projects are inaccurate. CPD could take additional steps to identify likely errors in IDIS total units data and to instruct grantees on requirements for reporting these data.
  • CPD has not effectively communicated requirements for grantees to obtain cost certifications for completed HTF projects.
  • CPD has not scheduled or conducted a comprehensive assessment of fraud risks using GAO’s Fraud Risk Framework or HUD’s own fraud risk management policy.

Read Affordable Housing: Improvements Needed in HUD’s Oversight of the Housing Trust Fund Program (GAO-23-105370) at: https://tinyurl.com/ycx3adpm

Read NLIHC’s HTF reports:

More information about HTF is on NLIHC’s HTF landing page and state-specific page, as well as on page 3-1 of NLIHC’s 2023 Advocates’ Guide.

Read the McHenry/Stivers letter requesting a GAO analysis at: https://tinyurl.com/5wnnuadr

Read NLIHC’s response to GAO at: https://tinyurl.com/ymk5j7um