Fair Market Rents for the Housing Choice Voucher Program and Moderate Rehabilitation Single Room Occupancy Program Fiscal Year 2005

AGENCY:

Office of the Secretary, HUD.

ACTION:

Notice of final fiscal year (FY) 2005 fair market rents (FMRs).

SUMMARY:

Section 8(c)(1) of the United States Housing Act of 1937 (USHA) requires the Secretary to publish FMRs periodically, but not less than annually, adjusted to be effective on October 1 of each year. FMRs are used to determine payment standard amounts for the Housing Choice Voucher program, to determine initial renewal rents for some expiring project-based Section 8 contracts, and to determine initial rents for housing assistance payment (HAP) contracts in the Moderate Rehabilitation Single Room Occupancy program. Other programs may require use of FMRs for other purposes. Today's notice provides for all areas final FY2005 FMRs that reflect the estimated 40th and 50th percentile rent levels trended to April 1, 2005.

Proposed FY2005 FMRs were published in the Federal Register on August 6, 2004. The proposed FMRs were calculated for the first time using 2000 Census data and new Office of Management and Budget (OMB) metropolitan area definitions. Both changes in how FMRs were calculated had significant impacts. A number of public comments from public housing agencies (PHAs) and major interest groups raised concerns about the magnitude of FMR changes experienced by many areas. HUD is required by law to utilize the most recent available data in calculating FMRs, and all federal agencies are instructed to use current OMB metropolitan area definitions unless there are strong program reasons to use alternative definitions. As a result of public comments and further consideration of the proposed FMRs, HUD determined that there was sufficient reason to not use the new OMB metropolitan area definitions in calculating the final FY2005 FMRs. The final FY2005 FMRs provided in this publication are therefore based on the most recent available data but use the same FMR area definitions used in the FY2004 FMR publication, which were based on old OMB metropolitan area definitions.

EFFECTIVE DATE:

The FMRs published in this notice are effective on October 1, 2004.

FOR FURTHER INFORMATION CONTACT:

For technical information on the methodology used to develop fair market rents or a listing of all fair market rents, please call the HUD USER information line at 800-245-2691 or access the information on HUD's Web site, http://www.huduser.org/datasets/fmr.html . Any questions related to use of FMRs or voucher payment standards should be directed to HUD's local program staff for the area in question. Questions on how to conduct FMR surveys or further methodological inquiries may be addressed to Marie L. Lihn or Lynn A. Rodgers, Economic and Market Analysis Division, Office of Economic Affairs, Office of Policy Development and Research, telephone 202-708-0590. Persons with hearing or speech impairments may access this number through TTY by calling the toll-free Federal Information Relay Service at 800-877-8339. (Other than the HUD USER information line and TTY numbers, telephone numbers are not toll free.)

SUPPLEMENTARY INFORMATION:

I. Background

Section 8 of the 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 to FMRs established by HUD for different areas. In the Housing Choice Voucher program, the FMR is the basis for determining the “payment standard amount” used to calculate the maximum monthly subsidy for an assisted family (see 24 CFR 982.503). In general, the FMR for an area is the amount that would be needed to pay the gross rent (shelter rent plus utilities) of privately owned, decent, and safe rental housing of a modest (non-luxury) nature with suitable amenities. In addition, all rents subsidized under the Housing Choice Voucher program must meet reasonable rent standards. The interim rule published on October 2, 2000 (65 FR 58870), established 50th percentile FMRs for certain areas.

Electronic Data Availability: This Federal Register notice is available electronically from the HUD news page: http://www.hudclips.org . Federal Register notices also are available electronically from the U.S. Government Printing Office website, http://www.gpoaccess.gov/fr/index.html

II. Procedures for the Development of FMRs

Section 8(c) of the USHA requires the Secretary of HUD to publish FMRs periodically, but not less frequently than annually. Section 8(c) states in part as follows:

Proposed fair market rentals for an area shall be published in the Federal Register with reasonable time for public comment and shall become effective upon the date of publication in final form in the Federal Register. Each fair market rental in effect under this subsection shall be adjusted to be effective on October 1 of each year to reflect changes, based on the most recent available data trended so the rentals will be current for the year to which they apply, of rents for existing or newly constructed rental dwelling units, as the case may be, of various sizes and types in this section.

The Department's regulations at 24 CFR part 888 provide that HUD will develop proposed FMRs, publish them for public comment, provide a public comment period of at least 30 days, analyze the comments, and publish final FMRs. (See 24 CFR 888.115.) Final FY2005 FMRs are published on or before October 1, 2004, as required by section 8(c)(1) of the USHA.

III. Proposed FY2005 FMRs

On August 6, 2004 (69 FR 48040), HUD published proposed FY2005 FMRs. As noted in the preamble to the proposed FMRs, the FMRs for FY2005 were based on two significant changes to the statistical methodology used to compute FMRs.

The first change was the introduction of 2000 Census data as a benchmark for FMRs. The 2004 FMRs were based on updated 1990 Census data except in areas where Random Digit Dialing (RDD) surveys or American Housing Surveys (AHS) had been conducted. Census 2000 data only recently became available in the level of detail (recent mover, standard-quality unit rents by number of bedrooms) necessary to calculate FMRs. The Department refers to the use of new decennial census data to revise FMRs as “rebenchmarking.” This process involves replacing the base year FMR estimates with those developed from new Census data and then updating the Census-based estimates from the date of the Census to the midpoint of the program year during which the FMRs will be in effect.

The second change was the use of new metropolitan area definitions issued by OMB to define FMR areas. As part of the 2000 Census process, OMB released new metropolitan area definitions on June 7, 2003, and updated them on February 18, 2004. These new metropolitan area definitions contain substantial changes from the old metropolitan area definitions because they incorporate the 2000 Census data and a substantially revised standard for defining metropolitan areas.

In response to the August 6, 2004, proposed FMRs, HUD received 370 public comments. The majority of the commenters were opposed to the proposed FMRs and cited various reasons. The primary reason given was that the proposed FY2005 FMRs were significantly different from the FY2004 FMRs and additional time was needed to examine the proposed FMRs. Many commenters asked HUD to delay issuing FY2005 FMRs.

As noted in Section II of this preamble, HUD is required to issue FMRs to be effective October 1, and the FMRs to be issued by HUD must be based on the most recent available data trended to the mid-point of the year in which they will be used. While HUD cannot delay issuance of the FY2005 FMRs, HUD has made changes to the proposed FY2005 FMRs announced in this notice in response to the public comments. (The public comments are discussed in more detail in Section V of this preamble.)

IV. Final FY2005 FMRs and FY2005 FMR Procedures

In setting the final FY2005 FMRs, HUD took into consideration a large number of comments objecting to the magnitude of changes caused by use of new OMB metropolitan area definitions and the inadequate time given to evaluate and respond to the proposed changes. While HUD is required by statute to use the most recent available data in setting FMRs, and by regulation to use current OMB metropolitan area definitions, HUD's regulations allow HUD to make exceptions to the use of the most current OMB metropolitan area definitions. Therefore, HUD is not obligated to use the new OMB metropolitan area definitions, and has determined to use the old OMB metropolitan area definitions, that is, the 2004 FMR area definitions, in calculating the final FY2005 FMRs. Use of the 2004 FMR area definitions generally produce fewer and smaller differences between the FY2004 FMRs and the FY2005 FMRs set forth in this notice for two reasons. The first is that the geographic area over which the 40th (or 50th) percentile rent is determined is unchanged, eliminating FMR differences resulting from changes in geography. The second reason is that some areas retained post-2000 Census RDD survey-based increases by reverting to the old definitions, whereas these increases could not be applied to the proposed FY2005 FMRs because the new areas differed too much from the old areas. Therefore, the FY2005 FMR schedules contained in this notice are based on 2000 Census and, when available, more current data, but were calculated for the same geographic areas used in preparing the FY2004 FMRs. Schedule B(1) lists Fair Market Rents for each area by state. FMRs that are at the 50th percentile, or median rent, are denoted by an asterisk. For informational purposes, Schedule B(2) shows what the 40th percentile FMRs would have been for the 39 areas where the FMR is set at the 50th percentile.

A. 2000 Census-Based FMRs

For areas where the base-year estimates were developed from the 2000 Census, the 40th and, where appropriate, 50th percentile gross rents for standard-quality units occupied by recent movers were calculated for differing numbers of bedrooms. The rent distributions were modified to eliminate public housing and other units with similarly low rents, so that only market-rent units would be considered. FMRs are calculated for all metropolitan areas and non-metropolitan counties.

FMR estimates are calculated for two-bedroom units, which are the most common rental units. Rent relationships between two-bedroom and other unit sizes are then calculated using local unit size rent relationships to the extent statistically feasible. For the past several years, bedroom ratios have been based on 1990 Census data. The FY2005 FMRs are the first to make use of 2000 Census data to more closely reflect market rent differentials between units with differing numbers of bedrooms. The rents for three-bedroom and larger units continue to reflect HUD's policy to set higher rents for these units than would result from using normal market rents. This adjustment is intended to increase the likelihood that the largest families, who have the most difficulty in leasing units, will be successful in finding eligible program units. The adjustment adds 8.7 percent to the unadjusted three-bedroom FMR estimates and adds 7.7 percent to the unadjusted four-bedroom FMR estimates. The FMRs for unit sizes larger than four bedrooms are calculated by adding 15 percent to the four-bedroom FMR for each extra bedroom. For example, the FMR for a five-bedroom unit is 1.15 times the four-bedroom FMR, and the FMR for a six-bedroom unit is 1.30 times the four-bedroom FMR. The FMRs for single-room occupancy units are 0.75 times the zero-bedroom (efficiency) FMR.

A further adjustment is made for areas with local bedroom-size intervals above or below what are considered to be reasonable ranges or where sample sizes are inadequate to accurately measure bedroom rent differentials. Experience has shown that highly unusual bedroom ratios typically reflect inadequate sample sizes or peculiar local circumstances that HUD would not want to utilize in setting FMRs (e.g., luxury efficiency apartments in New York City that rent for more than typical one-bedroom units). Bedroom interval ranges were established based on an analysis of the range of such intervals for all areas with large enough samples to permit accurate bedroom ratio determinations. The final ranges used were as follows: efficiency units are constrained to fall between 0.65 and 0.83 of the two-bedroom FMR, one-bedroom units must be between 0.76 and 0.90 of the two-bedroom unit, three-bedroom units must be between 1.10 and 1.34 of the two-bedroom unit and four-bedroom units must be between 1.14 and 1.63 of the two-bedroom unit. Bedroom rents for a given FMR area were then adjusted if the differentials between bedroom-size FMRs were inconsistent with normally observed patterns (e.g., efficiency rents were not allowed to be higher than one-bedroom rents and four bedroom rents were set at a minimum of three percent higher than three-bedroom rents).

For low-population, non-metropolitan counties with small Census recent-mover rent samples, Census-defined county group data were used in determining rents for each bedroom size. This adjustment was made to protect against unrealistically high or low FMRs resulting from insufficient sample sizes. The areas covered by this new estimation method have fewer than 33 two-bedroom Census sample observations.

After base 2000 Census estimates were established for each FMR area and bedroom size, they were updated from the estimated Census date of April 1, 2000, to April 1, 2005 (the midpoint of FY2005). Update factors for the 2000 through end of 2003 period were based either on the area-specific Consumer Price Index (CPI) survey data that were available for the largest metropolitan areas or on HUD regional RDD survey data.

For areas with local CPI surveys, CPI annual data on rents and utilities were used to update the Census rent estimates. Three-quarters of the 2000 CPI change factor was used to bring the FMR estimates forward from April to December of 2000. Annual CPI survey data could then be used for calendar years 2001, 2002, and 2003. Trending to cover the period from January 1, 2004, to April 1, 2005, was then needed. An annual trending factor of three percent, based on the average annual increase in the median Census gross rent between 1990 and 2000, was used to update estimates from the end of 2003 (i.e., the last date for which CPI data were available) until the midpoint of the fiscal year in which the estimates were used. The 15-month trending factor was 3.75 percent (3 percent times 15/12).

For areas without local CPI surveys, the same process was used except that regional RDD survey data were substituted for CPI data. Regional RDD surveys were done for 20 areas—the metropolitan and nonmetropolitan part of each of the 10 HUD regions. Areas covered by CPI metropolitan surveys were excluded from the RDD metropolitan regional surveys.

B. FMRs Based on Post-2000 Census Surveys

There are a number of areas where AHS and RDD telephone surveys of rents have been conducted since the 2000 Census. Both the AHS and RDD surveys have been proven to provide statistically reliable results within the limits of their stated confidence intervals.

The RDD technique involves use of large, randomly selected samples to obtain data on current rents paid for one- and two-bedroom rental units occupied by recent movers. RDD surveys exclude public housing units, newly built units and non-cash rental units. They do not exclude substandard units because there is no practical way to determine housing quality from telephone interviews. These surveys, however, also exclude units without a telephone, and past analysis has shown that the slightly downward rent estimate bias caused by including some substandard units is almost exactly offset by the slightly upward bias that results from only surveying units with telephones. This relationship held true across a variety of areas.

RDD surveys that meet HUD criteria have a high degree of statistical accuracy. There is a 95 percent likelihood that the 40th or 50th percentile recent mover contract rent estimates developed using this approach are within three to four percent of the actual 40th or 50th percentile. Virtually all survey estimates of contract rent will be within five percent of the actual 40th or 50th percentile value.

A number of RDD surveys were conducted after the 2000 Census. The results of RDD surveys conducted in 2001 and 2002 were used in the FY2004 FMRs and were evaluated for use in the final FY2005 FMRs. RDD surveys are used to provide a rebenchmarked FMR in lieu of updating the previous year's FMR when there is a statistically significant difference. RDD estimates are updated using the same types of data used to update Census estimates.

RDDs covering 24 areas were conducted in August 2004 and completed in time for use in this publication. The first column of the following table identifies the RDD survey area. The second column shows the final FY2005 FMRs that would have been published based on updated Census and 2001-2002 AHS and RDD surveys. The third column shows the August 2004 RDD results, trended to the middle of FY2005. The fourth column shows whether or not the RDD results were statistically different enough to justify replacing the Census or other survey estimates with the RDD results. The survey results were as follows:

HUD is directed by statute to use the most recent available data in its FMR publications. The RDD survey results are being implemented in the final FY2005 FMR publication consistent with that requirement.

HUD uses AHS data to calculate rents from the distributions of two-bedroom units occupied by recent movers. Public housing units, newly constructed units, and units that fail a housing quality test are excluded from the rental housing distributions before the FMRs are calculated. Thirteen areas were covered by AHS surveys conducted in 2002. Two surveys did not have enough recent mover cases to provide reliable estimates. More current AHS results were used to replace FMR estimates based on Census or RDD survey data if the Census- or RDD-based estimate was outside the 95 percent confidence interval of the AHS estimate. The AHS results produced statistically different FMR estimates and were used to rebenchmark FMRs for the following areas in Schedule B(1):

Orange County, CA, Portland-Vancouver, OR-WA, and Riverside-San Bernardino, CA. As noted in the proposed FY2005 publication, the AHS reduced the Portland FMR. The subsequent 2004 RDD survey confirmed this result. Orange County and Riverside-San Bernardino had increases as a result of the AHS. All three of these areas are 50th percentile FMR areas.

C. Impacts of New Data on Final FY2005 FMRs

The use of the 2000 Census rent data corrects for estimation errors that have accumulated during the past decade, and results in a larger than usual number of FMR revisions this year. The availability of more detailed local information on public housing, which is excluded from FMR estimates, also improved these estimates. Post-2000 AHS and RDD surveys provide more current estimates of market rents than those available from the Census, and serve to document the need for FMR changes in areas where recent mover rents do not follow regional or CPI rent trends. New AHS and RDD survey results were incorporated into this publication. The following table shows the distribution of impacts resulting from use of the 2000 Census and the AHS and RDD surveys used in this publication:

There are an additional 22 areas in the country where HUD has begun RDD surveys that could not be completed in time for this publication. Additionally, seven areas will be surveyed beginning in October 2004. Because the FY2005 FMRs for these areas will not have the benefit of a completed RDD survey by the date of submission of this document for publication in the Federal Register, and thus not have the benefit of the most recent rental data, HUD is allowing PHAs in those areas to wait, if they so choose, for completion of the RDD surveys and issuance of a notice that contains revised final FY2005 FMRs that reflect the completed RDD surveys. The notice of revised final FY2005 FMRs for these areas will reflect the RDD survey data and, at that point, housing authorities must use these published FMRs. Areas where HUD is currently conducting RDD surveys are:

D. Regulatory Procedures for Exceptions to Established FMRs

For housing authorities in areas that are not undergoing RDD surveys but continue to have concerns with the FY2005 FMRs announced in this notice, HUD's regulations in 24 CFR 888.113 provide the procedures by which HUD may make exceptions to established FMRs.

E. Manufactured Home Space Rents

Manufactured home space rents are set at 40 percent of the two-bedroom rent. Exceptions to this rent are granted when justified by survey data. All approved exceptions to these rents that were in effect in FY2004 were updated to 2005 using the relevant update factor. If the result of this computation was higher than 40 percent of the rebenchmarked two-bedroom rent, the exception remains and is listed in Schedule D.

F. FMRs for Federal Disaster Areas

Under the authority granted in 24 CFR part 888, the Secretary of HUD finds good cause to waive and hereby waives the regulatory requirements that govern requests for geographic area exception FMRs for areas that are declared disaster areas by the Federal Emergency Management Agency (FEMA). HUD is prepared (1) to grant disaster-related FMR exceptions up to 10 percent above the applicable FMRs for those areas. HUD field offices are authorized to approve such exceptions for single-county FMR areas and for individual county parts of multi-county FMR areas that qualify as disaster areas under the Robert T. Stafford Disaster Relief and Emergency Assistance Act, if (2) the PHA certifies that damage to the rental housing stock as a result of the disaster is so substantial that it has increased the prevailing rent levels in the affected area. Such exception FMRs must be requested in writing by the responsible PHAs. Exception FMRs approved by HUD during FY2005 will remain in effect until superseded by the publication of the final FY2007 FMRs, and replace lower, published FMR values.

V. Public Comments

In addition to the changes made in this notice for final FY2005 FMRs, HUD is continuing to accept public comments on FY2005 FMRs through November 5, 2004. HUD will consider these comments in determining revisions that may be needed to the FY2005 FMRs. Any such revisions will be announced in a subsequent FY2005 Federal Register notice.

In response to the August 6, 2004 proposed FMRs, HUD received 370 public comments covering 75 FMR areas during the initial comment period ending September 7, 2004. The majority of these comments concerned changes in the metropolitan area definitions that resulted in higher or lower FMRs. The New England region, where there was a proposed change away from the city-town designations previously used to define metropolitan areas to new OMB county-based metropolitan definitions, provided 148 comments. Many of the New England comments noted that the change in the geographic definitions was significant and had the impact of reducing FMRs in most metropolitan areas in New England. Some county parts shifted from one metropolitan area to another (e.g., part of Brockton to Providence, part of Boston to Providence), and some non-metropolitan areas were added to metropolitan areas with substantial increases for the non-metropolitan area (Chelmsford to Cambridge) and significant decreases for the metro area. Some metropolitan areas were combined (Bridgeport and Stamford-Norwalk) with increases for one area and decreases for the other area. The change in geography prevented the use of some 2001 and 2002 HUD RDD surveys in the proposed FY2005 FMR publication, because the metropolitan area coverage for these areas had changed so much under the new OMB area definitions that it was not considered valid to apply survey results based on different area definitions. Returning to use of the old metropolitan area definitions has the effect of re-instating RDD-based increases for the following areas (that are either a Metropolitan Statistical Area (MSAs) or a Primary Metropolitan Statistical Areas (PMSA):

Boston, MA-NH MSA

Portland, ME MSA

Brockton, MA PMSA

Lawrence, MA-NH PMSA

Lowell, MA-NH PMSA

Worcester, MA-CT MSA

Though not all comments from the New England region discussed the new geography, most stated that the proposed reductions were inconsistent with recent rental market history. In general, however, Census and more current surveys correct for what can be several years of accumulated estimation errors, and should not be thought of as solely relating to the change in rents that occurred between FY 2004 and FY2005. Many national housing and legal aid organizations also provided comments opposing the use of the new county-based metropolitan designations for New England and other areas and cited fair housing concerns.

Comments for other parts of the country also expressed concerns about geographic definitional changes that resulted in significant increases and decreases in FMRs that could adversely affect local programs. The new OMB definitions combined the two large metropolitan areas of New York City and Bergen-Passaic to create the New York—Wayne-White Plains, NY-NJ Division. As a result, the rents for Bergen-Passaic declined significantly, and comments were received about this decline. Dutchess and Orange counties in New York, formerly both separate metropolitan areas, were combined under the new geographic definitions. The proposed FY2005 FMRs for this combined metropolitan area significantly lowered rents for Dutchess County. Counties that were formerly non-metropolitan were concerned that HUD's current voucher renewal policy meant that they would be unable to continue to assist all current voucher families if FMRs increased, because voucher assistance funding would be set at the previous year's expenditure level plus a modest inflation adjustment. This was the primary concern from counties added to the metropolitan areas of Clarksville, TN, Lafayette, IN, Jonesboro, AR, and Anchorage, AK. In other comments, objections were raised by counties that had previously been included in metropolitan areas that were removed and designated as micropolitan areas with their own, much lower FMRs. This category includes comments received from Lincoln County, NC (formerly in Charlotte), Richmond County, KY (formerly in Lexington), Genessee County, NY (formerly in Rochester, NY), and Webster Parish, LA (formerly in Shreveport). Warrenton County, NJ, had a large proposed FMR reduction because it was removed from the Newark metropolitan area and placed in the Allentown, PA, metropolitan area. Anderson County, SC, had a large FMR reduction because it was taken out of the Greenville metropolitan area and made a separate metropolitan area with its own, much lower FMR.

In Puerto Rico, the new OMB definitions, especially for San Juan, were considered a cause of the proposed FMR reduction because many lower rent metropolitan and non-metropolitan municipios were added to this metropolitan area. The main source of these reductions, however, was the 2000 Census. RDD surveys of all Puerto Rico FMR areas will be conducted starting in October 2004.

A form letter-writing campaign by landlords in the Lake County-Kenosha County, IL-WI Division, resulted in 134 comments. Taking Lake County out of the Chicago metropolitan area and merging it with Kenosha County resulted in a significant decline in the FMR for Lake County and a significant increase for Kenosha County.

Many commenters with no changes in geography expressed concern over reductions in their FMRs. This included Cheyenne, WY, Fargo, ND, and Knox County, IN, and all areas in Hawaii. In Johnstown, PA, the removal of Somerset County was noted as a possible cause of the reduction in the proposed FMR, although the primary cause of the reductions proposed for both parts of the old metropolitan area was related to the use of 2000 Census data. The change from the use of state minimums to Census-defined county groups produced FMR declines in areas such as Sumter County, FL, Evangeline Parish, LA, and Harlan and Knox counties in Kentucky. Concerns about the large changes in county rents in Texas, Georgia, and Oklahoma were noted in general comments by the Texas Tenants' Union, the Georgia Department of Community Affairs, and the Oklahoma City Housing Authority. Other commenters stated that the proposed FMRs resulted in too many significant changes and would hinder the application of the program. This included comments from the States of Georgia, New York, New Jersey, Massachusetts, Vermont, Rhode Island, and Puerto Rico.

The Louisville Metro Housing Authority commented that significantly higher utility costs in the past year warranted higher FMRs. Louisville is currently undergoing an RDD survey that was started in September 2004. The utilities used for the FMR come from the utility schedule of the PHA, so the higher utility costs will be included in determining the RDD estimate.

Data of some form were provided to support comments made for 30 FMR areas. For the most part, the data consisted of rent reasonableness studies, data on local housing market conditions, newspaper ads for rental units, voucher rent data, and apartment rent data for projects. Only the data submitted with the comments of the Okanogan County Housing Authority met the minimum statistical requirement for acceptance. Accordingly, the FMRs for the County of Okanagon, WA, will be increased from the proposed FMR.

A group of major industry organizations (e.g., CLPHA, NAHMA, NAHRO, NLIHC, and others) jointly submitted a set of comments which argued for more time to evaluate the impacts of the new OMB definitions, more time to permit RDD surveys (including HUD's) to be completed, and postponing implementation of decreases until all RDD results were available and sufficient time had been allowed to consider all public comments. They also recommended an analysis of the new OMB definitions with the objective of minimizing the impact of their implementation (e.g., by allowing for submarket areas patterned after old area definitions when appropriate). HUD-conducted surveys of all areas with significant decreases in large-unit FMRs were proposed by commenters, as was the continued use of state minimum FMRs and the same minimum bedroom ratios used in the FY2004 FMRs. Concerns about large-unit FMR calculations were also expressed. The final FMRs address some but not all of these concerns. Use of old OMB definitions is the simplest way of addressing the concerns raised in many comments about use of the new definitions, and permits the impacts of use of new definitions to be distinguished from the impacts of new data. Many of the other requests were at odds with the requirement that HUD use the most recent available data in setting FMRs. PHAs continue to have the discretion to fund their own RDD surveys, but HUD's budget for this purpose has been and will continue to be limited.

HUD is permitting areas where HUD RDD surveys are being conducted to wait for issuance of updated final FY2005 FMRs for these areas that reflect completed RDD surveys. HUD will do as many RDD surveys in the future as available funding permits, and will continue to concentrate on areas with believe that use of local bedroom ratio data is appropriate in instances where there are large samples. The Census provides the best available measure of bedroom size rent relationships for most larger areas, and shows that small differentials, between bedroom sizes occur and are valid. In response to comments, however, HUD is using standard national ratios in instances where the statistical reliability of local ratios is questionable and has adjusted the calculation for the four-bedroom FMR to ensure that it is higher than the three-bedroom FMR by the minimum typical percentage differential even when Census data show no difference.

Both the increases and decreases in the bedroom ratios based on 2000 Census data reflect actual rent relationships, and HUD continues to add significant rent bonuses for units with more than two bedrooms. The decrease in the differentials between two-bedroom and larger rental units that occurs in some FMR areas is due to the availability of more current and reliable Census data. The same FMR bonuses for larger bedroom sizes used in the past were also applied in calculating the FY2005 FMRs.

VI. Manufactured Home Space Surveys

The FMR used to establish payment standard amounts for the rental of manufactured home spaces in the Housing Choice Voucher program is 40 percent of the FMR for a two-bedroom unit. HUD will consider modification of the manufactured home space FMRs where public comments present statistically valid survey data showing the 40th percentile manufactured home space rent (including the cost of utilities) for the entire FMR area.

Manufactured home space FMR revisions are published as final FMRs in Schedule D. Once approved, the revised manufactured home space FMRs establish new base year estimates that are updated annually using the same data used to estimate the Housing Choice Voucher program FMRs. The FMR area definitions used for the rental of manufactured home spaces are the same as the area definitions used for the other FMRs.

VII. HUD Rental Housing Survey Guides

HUD recommends the use of professionally-conducted RDD telephone surveys to test the accuracy of FMRs for areas where there is a sufficient number of Section 8 units to justify the survey cost of $20,000-$30,000. Areas with 500 or more program units usually meet this criterion, and areas with fewer units may meet it if local rents are thought to be significantly different from the FMR proposed by HUD. In addition, HUD has developed a simplified version of the RDD survey methodology for smaller, nonmetropolitan PHAs. This methodology is designed to be simple enough to be done by the PHA itself, rather than by professional survey organizations.

PHAs in nonmetropolitan areas may, in certain circumstances, do surveys of groups of counties. All county-group surveys must be approved in advance by HUD. PHAs are cautioned that the resulting FMRs will not be identical for the counties surveyed; each individual FMR area will have a separate FMR based on its relationship to the combined rent of the group of FMR areas.

PHAs that plan to use the RDD survey technique may obtain a copy of the appropriate survey guide by calling HUD USER on 800-245-2691. Larger PHAs should request “Random Digit Dialing Surveys; A Guide to Assist Larger Housing Agencies in Preparing Fair Market Rent Comments.” Smaller PHAs should obtain “Rental Housing Surveys; A Guide to Assist Smaller Housing Agencies in Preparing Fair Market Rent Comments.” These guides are also available on the Internet at http://www.huduser.org/datasets/fmr.html .

HUD prefers, but does not mandate, the use of RDD telephone surveys, or the more traditional method described in the small PHA survey guide. Other survey methodologies are acceptable if they provide statistically reliable, unbiased estimates of the 40th percentile gross rent. Survey samples preferably should be randomly drawn from a complete list of rental units for the FMR area. If this is not feasible, the selected sample must be drawn so as to be statistically representative of the entire rental housing stock of the FMR area. In particular, surveys must include units of all rent levels and be representative by structure type (including single-family, duplex and other small rental properties), age of housing unit, and geographic location. The decennial Census should be used as a starting point and means of verification for determining whether the sample is representative of the FMR area's rental housing stock. All survey results must be fully documented.

A PHA or contractor that cannot obtain the recommended number of sample responses after reasonable efforts should consult with HUD before abandoning its survey; in such situations HUD is prepared to relax normal sample size requirements.

Accordingly, the Fair Market Rent Schedules, which will not be codified in 24 CFR part 888, are amended as follows:

Fair Market Rents for the Housing Choice Voucher Program

Schedules B and D—General Explanatory Notes

1. Geographic Coverage

a. Metropolitan Areas—FMRs are market-wide rent estimates that are intended to provide housing opportunities throughout the geographic area in which rental-housing units are in direct competition.

HUD uses the OMB Metropolitan Statistical Area (MSA) and Primary Metropolitan Statistical Area (PMSA) definitions, but the current definitions from the June 6, 2003 publication have not yet been incorporated. Use of these new geographic definitions will be considered for use in future FMR publications. Schedule B FMRs are issued for the same metropolitan area definitions used by HUD in FY 2004 with the exceptions discussed in paragraph (b). The OMB-defined metropolitan areas closely correspond to housing market area definitions.

b. Exceptions to OMB Definitions—The exceptions are counties deleted from several large metropolitan areas whose old OMB metropolitan area definitions were determined by HUD to be larger than the housing market areas. The FMRs for the following counties (shown by the metropolitan area) are calculated separately and are shown in Schedule B within their respective states under the “Metropolitan FMR Areas” listing:

Metropolitan Area Counties Assigned County-Based FMRs

Chicago, IL

DeKalb, Grundy and Kendall Counties

Cincinnati-Hamilton, OH-KY-IN

Brown County, Ohio; Gallatin, Grant and Pendleton Counties in Kentucky; and Ohio County, Indiana

Dallas, TX

Henderson County

Flagstaff, AZ-UT

Kane County, UT

New Orleans, LA

St. James Parish

Washington, DC-MD-VA-WV

Berkeley and Jefferson Counties in West Virginia; and Clarke, Culpeper, King George and Warren Counties in Virginia

c. Nonmetropolitan Area FMRs—FMRs also are established for nonmetropolitan counties and for county equivalents in the United States, for nonmetropolitan parts of counties in the New England states and for FMR areas in Puerto Rico, the Virgin Islands and the Pacific Islands.

d. Virginia Independent Cities—FMRs for the areas in Virginia shown in the table below were established by combining the Census data for the nonmetropolitan counties with the data for the independent cities that are located within the county borders. Because of space limitations, the FMR visiting in Schedule B includes only the name of the nonmetropolitan County. The full definitions of these areas, including the independent cities, are as follows:

Virginia Nonmetropolitan County FMR Area and Independent Cities Included With County

2. Bedroom Size Adjustments

Schedules B(1) and B(2) shows the FMRs for 0-bedroom through 4-bedroom units. The FMRs for unit sizes larger than 4 bedrooms are calculated by adding 15 percent to the 4-bedroom FMR for each extra bedroom. For example, the FMR for a 5-bedroom unit is 1.15 times the 4-bedroom FMR, and the FMR for a 6-bedroom unit is 1.30 times the 4-bedroom FMR. FMRs for single-room-occupancy (SRO) units are 0.75 times th 0-bedroom FMR.

3. Arrangement of FMR Areas and Identification of Constituent Parts

a. The FMR areas in Schedule B(1) are listed alphabetically by metropolitan FMR area and by nonmetropolitan county within each state. The metropolitan areas in Schedule B(2) are listed alphabetically by the state and metropolitan area for only those 39 areas currently at the 50th percentile for their FMR. The exception FMRs for manufactured home spaces in Schedule D are listed alphabetically by State.

b. The constituent counties (and New England towns and cities) included in each metropolitan FMR area are listed immediately following the listings of the FMR dollar amounts. All constituent parts of a metropolitan FMR area that are in more than one state can be identified by consulting the listings for each applicable state.

c. Two nonmetropolitan counties are listed alphabetically on each line of the nonmetropolitan county listings.

d. The New England towns and cities included in a nonmetropolitan part of a county are listed immediately following the county name.