December 23, 2024
TO: NCRC members and allies
FROM: Josh Silver, Senior Fellow
RE: Final FHFA Affordable Housing Goals Rule for 2025-2027
Dear Colleagues:
On December 19, the Federal Housing Finance Agency (FHFA) issued its final affordable housing goals for Fannie Mae and Freddie Mac, known as Government Sponsored Enterprises (GSEs) for the years 2025 – 2027.
The goals are a series of performance measures for GSE purchases of single-family and multi-family lending. For single-family lending, the goals cover the percentage of GSE purchases of loans that finance housing for low-income homeowners and homeowners that reside in low-income census tracts or predominantly minority census tracts.
The FHFA assesses GSE performance using two performance measures for each goal. The first one, the benchmark level, is a goal expressed as an annual percentage of GSE purchased loans for each of the three years from 2025 through 2027 that the FHFA establishes based on economic forecasts. The second performance measure is a market level that reflects the percentage of loans that all lenders, as a group, made to a borrower category in a particular year. The GSE must meet or exceed the lower of the benchmark or the market level in each year. For example, if the FHFA established a benchmark of 25% for low-income home purchase loans and the market level is 23% for 2025, a GSE must meet or exceed at least 23% for 2025 to pass on the low-income home purchase goal.
If a GSE does not meet or exceed the lower of the two performance measures for a goal, it generally must submit a housing plan indicating to the FHFA how it will improve performance on the goal (there are exceptions to the housing plan requirement described below).
Below are the major takeaways on the final rule
FHFA Proposed Rule
FHFA Final Rule
The good news is that the FHFA did not lower any goal targets. The bad news is that the agency did not raise any targets despite our efforts to convince them to do so.
We applauded the FHFA’s minority tract goal of 12%, which was an increase from 10% in the previous goals for 2022-2024. On the other hand, NCRC believed the low-income tract goal was too low since about 9% of the GSEs purchases in recent years were loans from low-income tracts. The FHFA reasoned that it did not want to facilitate gentrification by pressuring the GSEs to purchase higher percentages of loans in low-income tracts. NCRC countered that for gentrifying tracts identified by a methodology similar to the one used by NCRC’s research department, the FHFA could disqualify loans from counting if they were issued to middle- or upper-income households. The FHFA did not take up our suggestion; we will keep pushing. There were other issues related to the single family goals; we will refer you to NCRC’s comment here.
For multifamily goals, NCRC had likewise advocated increases in the goals. For example, the FHFA proposed and finalized that 61% of the units in multifamily rental housing be for households that were low-income (up to 80% of area median income). Based on historical GSE performance, we believed the goal could have been about 7 percentage points higher. Again, there were more issues related to the multi-family goals which you can see in NCRC’s comment letter.
Measurement buffers could promote poor GSE performance
The FHFA added measurement buffers to provide some allowance for adverse economic or housing conditions impeding GSE performance on the goals. In other words, during years in which the benchmark level based on FHFA forecasting is higher than the market level, the FHFA will allow the GSEs to perform below the market level by specified amounts and avoid the requirement to produce a hosing plan. The benchmark level could be higher than the market level during years of recession in which the market is not performing as well as a forecast conducted two or three years ago estimated.
While NCRC understands the need to accommodate unexpected economic downturns, we do not think avoidance of housing plans is the appropriate way to take sudden economic changes into account. The measurement buffers run the risk of excusing poor GSE performance since there will now be circumstances in which they can perform below both the market and benchmark levels. NCRC sees nothing inherently punitive in a housing plan since it is a document submitted to the FHFA that describes how the GSEs will improve their performance. Furthermore, the FHFA has discretionary authority regarding whether to require a housing plan. Thus, in particularly adverse economic conditions, the FHFA can choose not to require a plan. In contrast, the measurement buffers, when they apply, provide automatic relief from the housing plan requirements.
The one concession the FHFA made to NCRC’s concerns is that measurement buffers cannot be applied to one or more of the goals if a GSE fails to meet goal levels for three consecutive years. In other words, measurement buffers can be used to avoid the housing plan requirement for the first two years but not during the last year of the three-year Affordable Housing Goal term for one or more of the goals.[1]
FHFA considers but declines to adopt NCRC proposal for two-year goal setting
NCRC has consistently advocated for two-year goal setting instead of the FHFA’s three-year goal setting. NCRC reasons that forecasting which is used to establish the benchmark level is generally more accurate over shorter time periods. In addition, a two-year goal setting would provide more opportunities for public comment. The FHFA considered this suggestion but rejected it, stating that a three-year period provides better opportunities for the GSEs to develop strategies for meeting the goals.[2] NCRC remains unconvinced and will see if we can further develop our arguments for two-year goal setting the next time the FHFA proposes goals.
FHFA declines to provide more opportunities for public comment
NCRC had also suggested that the public be provided with an opportunity to review and comment on draft housing plans when the FHFA requires a GSE to develop them. This is akin to the Community Reinvestment Act (CRA) strategic plan procedure. The FHFA declined to do so, stating that a housing plan may contain proprietary information.[3] NCRC is not convinced; housing plans can be developed just like CRA strategic plans that do not contain proprietary information. A public portion of a housing plan can be developed without proprietary information.
The FHFA also states that it produces an Annual Report reviewing GSE performance on the goals.[4] It would seem reasonable to provide formal public comment opportunities regarding GSE performance as described in the Annual Reports.
FHFA publishes state data but will not calculate local GSE data
NCRC commended aspects of the current FHFA data disclosures on its website. On a state level, the FHFA publishes excel tables regarding the performance of the GSEs on the affordable housing goals.[5] NCRC suggested that the FHFA publish this data on a local level (county and/or metropolitan level) so that stakeholders can see how the GSEs perform in their locality compared to state and national performance. Stakeholders can then encourage GSEs to perform better when GSE performance in their locality lags compared to other geographical areas. The FHFA stated that comparisons of local data can be misleading without context regarding loan volume fluctuations and other economic conditions. This is not convincing, considering that home loan data for banks is available on a local and census tract level. Careful analysis can consider economic conditions on a national, regional, and local level. The FHFA did commit to further researching this issue.[6]
[1] Final rule, p. 29, https://www.fhfa.gov/sites/default/files/2024-12/2024-12-19_2025-2027_ehgs_final_rule_web_5.pdf
[2] Final rule, p. 21.
[3] Final Rule, p. 22.
[4] Final Rule, p. 22.
[5] Final Rule, p. 23.
[6] Final rule, pgs. 23-24.
