Most of the comments not related to specific areas requesting a reevaluation discussed inaccuracies of the FMRs and a need for more current and local data. These comments and their responses are discussed in greater detail below. Before addressing specific issues, HUD would like to remind commenters that HUD receives no geographic information from regulations.gov for the location of the commenter. Therefore, HUD cannot respond to or take action on comments regarding the level of the FMR when there is no corresponding information about the FMR area in question.
Comment: FMRs in the town of Windham, CT are going down and this will impact Housing Choice Voucher (HCV) holders' ability to find units.
HUD Response: While HUD wants to be responsive to comments regarding the level of FMRs, HUD cannot reevaluate the FMRs in an area without data more current than what was originally used to calculate FMRs.
Additionally, valid reevaluation requests must be submitted by or supported by public housing agencies servicing at least 50 percent of the voucher holders in the FMR area. Finally, Windham Town, CT is part of the Windham County, CT HUD Metro FMR Area, which consists of the following towns: Ashford town, CT;
Brooklyn town, CT; Canterbury town,
CT; Chaplin town, CT; Eastford town,
CT; Hampton town, CT; Killingly town,
CT; Plainfield town, CT; Pomfret town,
CT; Putnam town, CT; Scotland town,
CT; Sterling town, CT; Thompson town,
CT; Windham town, CT; and Woodstock town, CT. Any data collected in support of a future reevaluation request would need to be representative of the rental units across the entire FMR area.
Comment: Baltimore City, MD requested a reevaluation to maintain their FMRs at the 50th percentile.
HUD Response: 50th percentile FMRs were abolished as part of the Small Area FMR Rulemaking (81 FR 80567) and Baltimore's three-year use of 50th percentile FMRs ended with the FY 2018 FMRs. Public housing agencies operating in metropolitan areas that revert to 40th percentile FMRs from
50th percentile FMRs and where Small Area FMR are not required have several avenues to maintain current payment standards or maintain the use of 50th percentile rents. These options include:
1. The Housing Opportunities
Through Modernization Act allows PHAs to maintain their current payment standard amounts (regardless of their percentage relative to the new FMRs) for in-place tenants. The Small Area FMR rule further provides that PHAs may select a payment standard between the current in-place payment standard down to 110 percent of the new FMR. This is characterized as a gradual reduction in the family's payment standard. For more information, please see PIH Notice 2018-01 (
https://www.hud.gov/sites/dfiles/PIH/documents/PIH-2018-01.pdf), item (4)(e).
2. HUD's regulations at 24 CFR 982.503(f) provides a mechanism for PHAs in this position to request approval to maintain payment standards based on 50th percentile rents.
3. HUD's regulations at 24 CFR 982.503(e), Success Rate Payment Standards, allows PHAs to set payment standards based on 50th percentile rents if they qualify based on HCV holders having limited success in finding suitable units.
4. The PHA may adopt exception payment standards of up to 110 percent of the applicable Small Area FMRs as described in 24 CFR 982.503(b)(iii).
Comment: One of the comments requesting a reevaluation of the FY 2020 FMRs in the Oakland-Freemont, CA HMFA also included several comments regarding the use of their December 2015 rent survey, an objection to the use of 5-year ACS data, an objection to the use of a national trend factor, and a
comment that the "public housing cut off" rent is too low for one county within the metropolitan area.
HUD Response: Due to the nature of the questions posed regarding the need to conduct a new survey within the comment, HUD has already provided indepth responses to the commenter; however, HUD is providing additional commentary here for the benefit of all FMR commenters.
HUD cannot continue to use the survey conducted in the area in December 2015 because data collected through the 2016 ACS are more recent than the survey. This is true even though the Department received the local survey in early 2016 and used the information to revise the FY 2016 FMRs,
and used the local survey as the basis of the FY 2017 and FY 2018 FMR calculations. Due to the rolling nature of the ACS, HUD has adopted the following policy concerning the longevity of local area surveys. Surveys with an "as-of" date of January through June ARE NOT considered to be more
current than the same year ACS survey since at least 50 percent of the ACS responses are newer than the local area survey. Local area surveys with "as-of" dates in July through December ARE considered to be more current than the same year ACS since less than 50 percent of the ACS responses are more current than the locally conducted survey.
For the purposes of uniformity and transparency, HUD's FMR documentation systems shows the 5- year ACS data for each FMR area along with the calculation of a recent-mover factor for each area based on information available from the 1-year ACS data. In practice, for large areas with statistically reliable 2-bedroom recent-mover estimates from the 1-year ACS tabulations, the recent mover "factor" is an explanation artifact for conformity in HUD's online documentation systems . The 1-year recent mover data replaces the 5-year ACS data in the FMR calculation. That is, the FMR for the area is based solely on the 40th percentile 2-bedroom recent mover rent calculated from the 1-year ACS data and this is true for the Oakland area. HUD has addressed the issue of local trend factors in prior FMR Federal Register notices. Beyond HUD's own research into this topic, HUD has employed third-party researchers to assist with this important component of the FMR calculation. Following the successful completion of this research, should local trend factors be viable, HUD is compelled to announce the new trend factor calculation methodology in a notice of FMR material changes and solicit public comments on such changes prior to implementing them.
Due to the lapse in Federal appropriations at the end of 2018, HUD is not able to commit to this change being ready for a material change notice prior to the calculation of the FY 2020 FMRs.
Finally, HUD reiterates that the "public housing cut-off rent" used to eliminate the bottom end of the distribution of rental units from the ACS before the 40th percentile rent is calculated is a proxy because the ACS does not collect whether or not a respondent is receiving some form of housing assistance.
HUD uses a consistent method to calculate this distribution cut off for each HUD region and sees no reason to make a change to this method at this time.
Comment: Several commenters suggested that HUD should provide additional funding to PHAs who undertake local area surveys.
HUD Response: HUD reminds PHAs that paying for local area rent surveys is an eligible expense to be paid from ongoing administrative fees or their administrative fee reserve account.
Comment: HUD received several comments from the Lenawee, MI community regarding the level of their FMRs. The commenters noted that the FMRs were too low and one commenter requested information on how information is gathered to calculate FMRs.
HUD Response: The requests for reevaluation of the FY 2020 FMRs in Lenawee County, MI were not accompanied by supporting statements from PHAs representing at least 50 percent of the voucher holders in the area and HUD was unable to grant the request. As for information on how HUD
calculates the FMRs, HUD provides comprehensive documentation on how FMRs are calculated for each area. For FY 2020, please visit
https://www.huduser.gov/portal/datasets/fmr.html#2019_query and follow the prompts after selecting the "Click Here For FY 2020 FMRs" button.
Comment: HUD should use more timely data when calculating FMRs. HUD should request more money to reimburse surveys. HUD should continue to refine the FMR calculation methodology.
HUD Response: There are no other data on gross rents paid that are consistently collected on a nationwide basis, available to HUD, that are more current than the data we receive from the ACS dataset. Proprietary rental data cannot be used as the basis for the FMR calculations because they are not
consistently available for all areas and frequently are not statistically representative of the rental markets they cover. Some of these sources focus on rents for major apartment projects only. Other sources include single family homes, which are at least 30 percent of the rental market in major metropolitan areas and a greater portion in rural areas, are typically compiled from internet-based ads, or the small subset of professionally managed single-family rental units. Online listings of rents are similar to newspaper ads which have been excluded as a source of rent data for FMRs since the mid-1980s due to a directive issued by HUD's Inspector General.
As HUD has stated before, the Federal Government invests a substantial amount of resources in collecting socioeconomic data through the ACS.
Furthermore, the Census Bureau has statutory advantages in compelling responses to the ACS and receives significantly higher response rates than HUD could achieve if it was to undertake its own survey program.
Comment: Annual "erratic fluctuations" in FMRs within the same area do not follow established housing market dynamics and represent serious flaws in the FMR calculation methodology.
HUD Response: HUD has implemented steps to attenuate the fluctuations found in the annually updated survey data; however, with the statutory directive to use the most recent data available, HUD is compelled to update the data behind each area's FMR calculation when new data is released.
Consequently, FMRs will change from year to year in accordance with changes in the underlying survey data. Furthermore, the commenter provides no references to academic literature that seeks to establish that it is impossible for rents to move both up and down over time. HUD emphasizes
that the primary data source for FMRs is a survey (ACS) and while surveyors do their best to select unbiased random samples of the population they wish to study, sampling error still persists within survey statistics.
