Office of the Assistant Secretary for Policy Development and Research, HUD.
Notice of proposed changes for estimation of Fair Market Rents (FMRs).
Section 8(c)(1) of the United States Housing Act of 1937 requires the Secretary to publish FMRs periodically, but not less than annually, adjusted to be effective on October 1 of each year. The primary uses of FMRs are to determine payment standards for the Housing Choice Voucher (HCV) program, to determine initial renewal rents for some expiring project-based Section 8 contracts, to determine initial rents for housing assistance payment contracts in the Moderate Rehabilitation Single Room Occupancy Program, and to serve as rent ceilings for rental units in both the HOME Investment Partnerships Program and the Emergency Solutions Grants Program. HUD also uses FMRs in the calculation of maximum award amounts for Continuum of Care grantees and in the calculation of flat rents for Public Housing units. In furtherance of that effort, HUD proposes two changes in how FMRs are estimated in this notice and seeks public comment on the proposed changes.
Comment Due Date:
July 5, 2019.
HUD invites interested persons to submit comments regarding the proposed changes to the calculation of the FMRs to the Regulations Division, Office of General Counsel, Department of Housing and Urban Development, 451 7th Street SW, Room 10276, Washington, DC 20410-0001. Communications must refer to the above docket number and title and should contain the information specified in the “Request for Comments” section. There are two methods for submitting public comments.
1. Submission of Comments by Mail. Comments may be submitted by mail to the Regulations Division, Office of General Counsel, Department of Housing and Urban Development, 451 7th Street SW, Room 10276, Washington, DC 20410-0500. Due to security measures at all Federal agencies, however, submission of comments by mail often results in delayed delivery. To ensure timely receipt of comments, HUD recommends that comments submitted by mail be submitted at least two weeks in advance of the public comment deadline.
2. Electronic Submission of Comments. Interested persons may submit comments electronically through the Federal eRulemaking Portal at http://www.regulations.gov. HUD strongly encourages commenters to submit comments electronically. Electronic submission of comments allows the commenter maximum time to prepare and submit a comment, ensures timely receipt by HUD, and enables HUD to make them immediately available to the public. Comments submitted electronically through the http://www.regulations.gov website can be viewed by other commenters and interested members of the public. Commenters should follow instructions provided on that site to submit comments electronically.
Note: To receive consideration as public comments, comments must be submitted through one of the two methods specified above. Again, all submissions must refer to the docket number and title of the notice.
No Facsimile Comments. Facsimile (FAX) comments are not acceptable.
Public Inspection of Public Comments. All properly submitted comments and communications regarding this notice submitted to HUD will be available for public inspection and copying between 8 a.m. and 5 p.m. weekdays at the above address. Due to security measures at the HUD Headquarters building, an advance appointment to review the public comments must be scheduled by calling the Regulations Division at 202-708-3055 (this is not a toll-free number). Individuals with speech or hearing impairments may access this number through TTY by calling the Federal Relay Service at 800-877-8339. Copies of all comments submitted are available for inspection and downloading at http://www.regulations.gov.
Start Further Info
FOR FURTHER INFORMATION CONTACT:
Questions on this notice may be addressed to Adam Bibler, Chalita Brandly, or Peter Kahn of the Program Parameters and Research Division, Office of Economic Affairs, Office of Policy Development and Research, HUD Headquarters, 451 7th Street SW, Room 8208, Washington, DC 20410; telephone number 202-402-2409 (this is not a toll-free number), or they may be reached at email@example.com. Persons with hearing or speech impairments may access HUD numbers through TTY by calling the Federal Relay Service at 800-877-8339 (toll-free). For technical information on the methodology used to develop FMRs or a listing of all FMRs, please call the HUD USER information line at 800-245-2691 (toll-free) or access the information on the HUD USER website https://www.huduser.gov/portal/datasets/fmr.html.
Electronic Data Availability. This Federal Register notice will be available electronically from the HUD User page at https://www.huduser.gov/portal/datasets/fmr.html.
Federal Register notices also are available electronically from https://www.federalregister.gov/, Start Printed Page 26142the U.S. Government Printing Office website.
Complete documentation of the impact of these methodology changes and calculation of hypothetical FY 2019 FMRs with these changes are available at https://www.huduser.gov/portal/datasets/fmr.html. Small Area FMRs for all metropolitan FMR areas incorporating these material changes in methodology have also been calculated and are also available at: https://www.huduser.gov/portal/datasets/fmr.html.
End Further Info
Start Supplemental Information
Section 8 of the United States Housing Act of 1937 (USHA) (42 U.S.C. 1437f) authorizes housing assistance to aid lower-income families in renting safe and decent housing. Housing assistance payments are limited by Fair Market Rents (FMRs) established by HUD for different geographic areas. In general, the FMR for an area is the amount that would be needed to pay the gross rent (shelter rent plus utility costs) of privately owned, decent, and safe rental housing of a modest (non-luxury) nature with suitable amenities and is set at the 40th percentile of the distribution of gross rents for recent movers. HUD's FMR calculations represent HUD's best effort to estimate the 40th percentile gross rents paid by recent movers into standard quality units in each FMR area.
In recent years, the most prevalent comments concerning FMRs are that FMRs need to incorporate more local and more timely data. HUD has enumerated potential solutions to these concerns in a recent report to Congress entitled “Proposals to Update the Fair Market Rent Formula”.
The proposals outlined in this notice address the concern of using more local data; however, HUD believes that the use of local trend factors will also address some of the concerns regarding the timeliness of the data used to calculate FMRs.
II. Procedures for Developing FMRs
Section 8(c)(1) of the USHA requires the Secretary of HUD to publish FMRs periodically, but not less frequently than annually. Section 8(c)(1)(B) as amended by the Housing Opportunities Through Modernization Act of 2016 (HOTMA) (Pub. L. 114-201, approved July 29, 2016), requires that HUD publish for comment in the Federal Register a notice of proposed material changes in the methodology for estimating FMRs and a notice containing HUD's final decisions regarding such proposed substantial methodological changes and responses to public comments.
The calculation of FMRs may be reduced to three parts: An estimate of gross rents paid by recent movers from the American Community Survey (ACS), an inflation adjustment measured using components of the Consumer Price Index, and a trend factor. In the following section, HUD describes proposed changes to the trend factor calculation, and a change in the way Small Area FMRs are calculated for ZIP Codes with insufficient data.
III. FMR Methodology Changes
1. Trend Factor Changes
Following current methodology, calculation of FMRs for FY 2020 requires HUD to update the ACS-based gross rent “as of” 2017 rent through the end of 2018 using the annual change in Consumer Price Index (CPI) components from 2017 to 2018. Following the application of the appropriate CPI update factor, HUD calculates a trend factor, incorporating economic assumptions used in the formulation of the President's Budget, which brings the estimate forward seven quarters from CY 2018 to FY 2020 using a forecast of the Gross Rent Index. The Gross Rent Index forecast is made up of two independently forecasted components of the Consumer Price Index: Housing, Shelter, Rent of Primary Residence; and Housing, Fuels and Utilities.
The forecasts of these two series are combined using the long-term average expenditure combination factors of approximately 80 percent and 20 percent, respectively.
Commenters on prior year's FMRs have remarked that FMRs are not timely enough or based enough on local information and that this may cause operational difficulties in program operations for the HCV program. In a 2017 Senate Report,
the Committee on Appropriations called for HUD to improve its FMR estimates to better reflect the rent inflation that occurs between the time that American Community Survey data is collected and the fiscal year for which the FMRs are produced. The report further recommended that HUD explore means of accelerating its research on improving its FMR estimates.
As a result of these FMR accuracy concerns, HUD tasked a multi-disciplinary research team to explore ways to refine its current trend factor methodology to address these issues. The final report of this research is available https://www.huduser.gov/portal/sites/default/files/pdf/deriving-local-trends-factors.pdf. Within the report, the research team documents that using local CPI data instead of national CPI data to derive local trend factors, using similar methods to what is currently used to derive the national trend factor, can provide better estimates of the expected change in gross rents for local areas where data is available.
Currently, HUD uses a national Gross Rent Index forecast to trend rents to the current fiscal year. HUD's national gross rent index forecast model is a composite of forecasts for national rent of primary residence and national fuels and utilities. The national rent of primary residence relies on forecasts of residential fixed investment from the Bureau of Economic Analysis National Income and Product Accounts. These forecasts come from the economic assumptions that correspond with the President's budget submissions. The national fuels and utilities forecast are based on forecasts of the price per barrel of West Texas Intermediate Crude Oil, the price per short ton of bituminous coal, and the seasonally adjusted Consumer Price Index, All Urban Consumers (CPI-U). The CPI program currently calculates Rent of Primary Residence and Fuels and Utilities for 22 metropolitan areas and for four regions at three different size classes (while data for 10 U.S. divisions and the Riverside-San Bernardino-Ontario, CA metropolitan area are available, they do not currently contain enough data observations to construct reliable forecast estimates). Approximately 42 percent of Housing Choice Voucher families live in an area covered by one of the 22 CPI metropolitan areas. FMR areas without a corresponding CPI metropolitan area will use a regionally based local trend factor.
The multi-disciplinary team HUD tasked examined multiple models and model structures for forecasting shelter rent and utility components of gross rent at the local level. The performance Start Printed Page 26143of local forecast models was tested by comparing actual data to an in-sample forecast (or validation period). Models were estimated using approximately 20 years of quarterly observations up to 2016 (Q1) and forecasted out through 2018 (Q1). These comparisons revealed how close rent and utility predictions of the validation period were as measured by the Root Mean Square Error Statistic (RMSE). Models yielding the lowest RMSE were determined to provide the most accurate estimates.
Based on these results, the team recommended that forecast of local vs. national rent of primary residence data from BLS be informed by the forecast of national residential fixed investment from the Bureau of Economic Analysis National Income and Product Accounts, as in the calculation of the national Gross Rent Index. This method is referred to as a “National Input Model” (NIM) approach for rent. In contrast, the team's research did not find that using a utility NIM model was the best approach for forecasting local fuels and utility data from BLS. Instead, a “Pure Time Series” (PTS) approach produced the best model results. In a PTS approach, the local forecasts are based upon previous values of the variable of interest; in this case prior values of the local fuel and utilities index. Additionally, the team also analyzed a Local Input Model (LIM) approach, where forecasts are developed based off of local exogenous variables such as local building permit data and employment data for rent, and electricity prices for utilities. While the LIM specification produced a lower RMSE in some areas, the research team did not recommend the LIM approach for use in the trend factor as a one-size model for calculating rent or utilities. The above recommendations were based on the study team's finding that the NIM model had a lower RMSE in 10 of the 22 geographic areas for rent, while the PTS model had a lower RMSE in 9 of the 22 geographic areas for utilities.
As a result of the recommendations provided by the research team, HUD is proposing the following to address concerns of FMR accuracy. Overall, HUD proposes using metropolitan and regional Gross Rent Index forecasts to calculate and apply more locally-based trend factors to address concerns of FMR accuracy. While the research provides recommendations to use the NIM forecast for the calculation of the rent of primary residence, and the PTS forecast for the calculation of fuels and utilities, the research shows that one model does not fit the rent and utility data better in all geographic areas. HUD proposes to build on the research team's approach for calculating local trend factors for each CPI area by selecting model forms unique to each area that minimize the RMSE for each rent and utility forecast for each CPI area as opposed to one cross-cutting model form calibrated with the data for each CPI area. This will ensure the best performing models and optimal functional forms are used. As a result, there is a possibility that a forecast model for a CPI area may change over time as additional data become available each year and forecast models are re-estimated. For instances when HUD changes the functional form of the model (NIM, PTS, LIM) for a geographic area that is different from the previous year, HUD will ensure the change is not due to overfitting the model or outliers in the data.
To ensure transparency in this process, HUD will include the model specification used to calculate local trend factors for each area in the on-line Fair Market Rent Documentation System.
2. Using Neighboring ZIP Codes in Place of County-Based Small Area FMRs
In calculating Small Area Fair Market Rents (SAFMRs), HUD attempts to use ZIP Code level estimates where possible. In cases where ZIP Code level estimates are not available or are not sufficiently reliable, HUD's current practice is to assign a SAFMR based on the estimate of gross rent for the county of the ZIP Code. However, because metropolitan counties are often much larger than ZIP Codes,
this approach has the potential to produce discontinuous SAFMR values where the county based SAFMR is not an accurate proxy for neighborhood-level rents. Moreover, in many cases, HUD-defined metropolitan areas consist of only a single county. This means that a ZIP Code without useable local data will use a SAFMR that is exactly equal to the metropolitan FMR, running counter to the purpose of Small Area FMRs, which is to differentiate rents within a metropolitan area.
To address this, HUD is proposing an additional step. If a ZIP Code Tabulation Area (ZCTA) does not have reliable rent data, HUD will then check to see if the ZCTA is bordered by ZCTAs that themselves have reliable rent data. If at least half of a ZCTA's “neighbors” 
have such data, the weighted average of those estimates will be used as the basis for the SAFMR rather than a county proxy, where the weight is length of the shared boundary between the ZCTA and its neighbor. To test the effects of this methodology change, HUD has recalculated FY19 SAFMRs. Adopting this methodology affects the Small Area FMR for 2,677 ZIP Codes, about 11 percent of all published ZIP Codes. Of the 1.87 million voucher holders in metropolitan areas, 4,100, or 0.2 percent are in ZIP Codes affected by this change. The average change relative to the prior methodology is a $49, or 4.1 percent, increase in the two-bedroom Small Area FMR. Of these ZIP Codes, 1,714 experience an increase in the two-bedroom FMR and 963 show a decrease relative to what the SAFMR would be without the use of neighboring ZIP Code rent data.
As an illustrative example, a map depicting a portion of the Pittsburgh, PA metropolitan area prior to the calculation of neighboring ZCTA averages is available at the link specified in the footnotes.
ZCTAs 15086 and 15015 show sharp divergences in rent from their surrounding ZCTAs. Calculating the SAFMR based on the average of the neighboring ZCTAs raises the SAFMR from the county-based value of $950 to $1,340 for both 15086 and 15015. In 2016, there were 105 rental housing units in these ZCTAs. Under this approach, the small ZCTA of 15075 would have its SAFMR lowered from the county-based value of $950 to $890, which is the SAFMR of ZCTA 15024, which surrounds 15075.
To assist in evaluating this proposed change, HUD is publishing a file with actual FY 2019 SAFMRs and hypothetical FY 2019 SAFMRs for ZIP Codes affected by this methodology at https://www.huduser.gov/portal/datasets/fmr/fmr2020/FY19Hypo-SAFMRs-Zipcodes-Material-Change-Notice.xlsx. Note that the hypothetical SAFMRs do not include the proposed revisions to the trend factor discussed previously in this notice.Start Printed Page 26144
IV. Request for Public Comments on Changes
HUD continually strives to calculate FMRs that can serve as an effective program parameter while meeting the statutory requirement to use “the most recent available data.”
These methodology changes are not monodirectional; for example, the use of local forecast trend factors will cause FMRs to be higher in some areas and lower in others compared to using a national forecast trend factor. HUD is particularly interested in receiving comments on its intended approach for evaluating the accuracy of local trend factors using the RMSE and is interested in potential alternative methods for assessing the best local forecast model to select.
Hypothetical FY 2019 FMRs and Small Area FMRs, using these new methodology changes, are published at https://www.huduser.gov/portal/datasets/fmr.html#2020_documents.
V. Environmental Impact
This notice proposes changes in the way FMRs are calculated and does not constitute a development decision affecting the physical condition of specific project areas or building sites. Accordingly, under 24 CFR 50.19(c)(6), this notice is categorically excluded from environmental review under the National Environmental Policy Act of 1969 (42 U.S.C. 4321).
End Supplemental Information
Dated: May 29, 2019.
Todd M. Richardson,
General Deputy Assistant Secretary for Policy Development and Research.
[FR Doc. 2019-11763 Filed 6-4-19; 8:45 am]
BILLING CODE 4210-67-P