oversight

FHA Loan Limits: Finance Board Data Are a Reasonable Source of Home Sales Prices

Published by the Government Accountability Office on 1999-03-29.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                 United States General Accounting Office

GAO              Report to the Ranking Minority Member,
                 Committee on Banking and Financial
                 Services, House of Representatives


March 1999
                 FHA LOAN LIMITS
                 Finance Board Data
                 Are a Reasonable
                 Source of Home Sales
                 Prices




GAO/RCED-99-78
      United States
GAO   General Accounting Office
      Washington, D.C. 20548

      Resources, Community, and
      Economic Development Division

      B-281434

      March 29, 1999

      The Honorable John J. LaFalce
      Ranking Minority Member
      Committee on Banking
        and Financial Services
      House of Representatives

      Dear Mr. LaFalce:

      Purchasing a home is one of the most significant financial undertakings for
      most families. For many of those seeking to do so for the first time, the
      Department of Housing and Urban Development’s (HUD) Federal Housing
      Administration (FHA) often helps make home ownership possible. FHA does
      this by insuring mortgages so that lenders will make loans to borrowers
      who might be unable to otherwise get a mortgage. FHA also sets limits on
      the dollar value of individual loans it will insure. The maximum
      FHA-insured mortgage a borrower may obtain varies throughout the
      country depending on local housing prices, resulting in hundreds of
      different loan limits. To measure local house prices and determine these
      loan limits, FHA largely relies on data from a Federal Housing Finance
      Board (Finance Board) survey that collects house purchase prices and
      other related data on non-government-insured mortgages from mortgage
      lenders.

      Because the FHA loan limits for some areas seemed high, you asked us to
      collect and analyze data about house prices from sources other than the
      Finance Board. This report (1) compares data on house prices from the
      Finance Board with data HUD’s Office of Federal Housing Enterprise
      Oversight collects to measure house price changes;1 (2) presents the views
      of officials of the agencies involved on the results of this analysis;
      (3) analyzes the effect on median prices of supplementing the Finance
      Board’s and the Office of Federal Housing Enterprise Oversight’s data with
      information each does not already include on lower-priced homes with
      government-insured mortgages; and (4) provides information on FHA’s

      1
       Neither of these sources collects information on purchase prices for all home loans or all home sales.
      The Federal Housing Finance Board collects home price data for mortgages that are not insured by the
      federal government and that are not for the purpose of refinancing an existing loan. The Office of
      Federal Housing Enterprise Oversight collects data on loans—few of which are government-insured
      and none of which may exceed a statutory limit—purchased by the Federal National Mortgage
      Association (Fannie Mae) and the Federal Home Loan Mortgage Corporation (Freddie Mac). While the
      Finance Board publishes data on house prices as well as price changes, the Office of Federal Housing
      Enterprise Oversight publishes only data on house price changes, not the house price data it provided
      us for the analysis in this report. Both Fannie Mae and Freddie Mac consider these data on house
      prices proprietary and confidential.



      Page 1                                                          GAO/RCED-99-78 FHA Loan Limits
                   B-281434




                   recent efforts to explore alternative sources of data for measuring median
                   home prices. The data we analyzed cover 42 selected metropolitan areas
                   for which the Finance Board publicly reports data annually (see app. I for
                   a list of these areas).


                   The Finance Board and the Office of Federal Housing Enterprise
Results in Brief   Oversight’s estimates of 1997 median home sales prices were similar in
                   about two-thirds of the metropolitan areas we reviewed. That is, in 27 of
                   the 42 areas, the two agencies’ estimates were within 5 percent of each
                   other. Therefore, loan limits based on either set of data would be similar in
                   these areas. For the remaining 15 areas we reviewed, the Finance Board
                   estimated a higher median home sales price in 10 of the areas, while the
                   Office of Federal Housing Enterprise Oversight’s estimate was higher in 5
                   areas.

                   Officials familiar with these data, including officials from the Finance
                   Board as well as the Federal National Mortgage Association (Fannie Mae)
                   and the Federal Home Loan Mortgage Corporation (Freddie Mac), cited
                   the low number of substantive differences in our analysis as an indicator
                   of the validity of the Finance Board’s data. Because no substantive
                   difference existed between the two sets of data in about two-thirds of the
                   areas we reviewed, the officials indicated the Finance Board’s data are a
                   reasonable measure of an area’s median sales price for homes without
                   government-insured financing. In those areas for which substantive
                   differences did exist, officials from the different agencies involved agreed
                   the reason was that the Finance Board includes larger loans, and thus
                   higher home purchase prices, in its survey than does the Office of Federal
                   Housing Enterprise Oversight. The Finance Board’s 1997 data include loan
                   amounts up to $500,000 and house purchase prices up to $750,000. The
                   Office of Federal Housing Enterprise Oversight’s data for 1997 include no
                   loans greater than $214,600. These officials also cited normal variations
                   associated with surveys and statistical sampling as a reason for some
                   differences between the two sets of data.

                   Supplementing the data from either the Finance Board or the Office of
                   Federal Housing Enterprise Oversight with data on homes financed with
                   government-insured loans would lower the estimated median home sales
                   price in any given area by 2 to 31 percent. This is true because neither the
                   Finance Board’s nor the Office of Federal Housing Enterprise Oversight’s
                   data include homes financed with government-insured loans. The
                   purchase prices of homes financed with mortgages insured by FHA and the



                   Page 2                                          GAO/RCED-99-78 FHA Loan Limits
             B-281434




             Department of Veterans Affairs2 are, on average, lower than those of
             homes bought with privately insured financing. These lower prices result
             from the limits on the size of individual loans FHA may insure and because
             government-insured financing tends to be focused on first-time
             homebuyers, who typically purchase less expensive homes.

             FHA is engaged in an effort to use additional sources of data, such as the
             Office of Federal Housing Enterprise Oversight, to improve its
             measurement of median home sales prices. In part, this is due to recent
             legislative changes that make it necessary for FHA to know the median
             price for each county within a metropolitan area as well as for the entire
             area when setting loan limits.3 Currently, FHA relies heavily on the Finance
             Board’s survey for the data it needs to set its loan limits, but to a limited
             extent it has also supplemented that survey with data its field offices
             gather on local home sales prices. In addition to exploring the use of the
             Office of Federal Housing Enterprise Oversight’s data, FHA is considering
             using data on loans that neither Fannie Mae nor Freddie Mac has
             purchased.


             One of FHA’s primary goals is to assist those households that are unable to
Background   meet the requirements of the private market for mortgages and mortgage
             insurance or that live in underserved areas where mortgages may be
             harder to obtain. In doing so, FHA applies more flexible underwriting
             standards than the private market generally allows. Borrowers seeking
             FHA-insured loans may make smaller down payments (as a percentage of
             the purchase price) than the private market requires and may also include
             in the amount they borrow most costs associated with closing the loan,
             rather than using cash for those expenses, as private lenders generally
             require. FHA is required by statute to set limits on the dollar amount of
             individual loans it will insure. These limits are based, in part, on local
             median home prices.

             The Finance Board surveys major mortgage lenders each month,
             collecting information on the terms and conditions (including the sales
             prices of homes) of conventional single-family home loans closed during




             2
              Although the Department of Veterans Affairs guarantees mortgages for U.S. veterans and their
             families rather than insuring them, this report refers to these loans as government-insured.
             3
              P.L. 105-276 sec. 228(b) mandates that the highest loan limit of any county within a metropolitan
             statistical area must apply to loans insured in all the counties in that area.



             Page 3                                                           GAO/RCED-99-78 FHA Loan Limits
B-281434




the last 5 business days of the month.4 The Finance Board may not require
that lenders participate in its survey. Those doing so participate
voluntarily. Fannie Mae and Freddie Mac, both government-sponsored
enterprises, are parts of the secondary mortgage market, through which
many single-family home mortgages are ultimately sold.5 Federal law
requires that Fannie Mae and Freddie Mac use information from the
Finance Board’s survey on the year-to-year change in house prices to
annually adjust the conforming loan limit (currently $240,000), which is a
legislative restriction on the size of any individual loan that either may
buy.6

FHA also uses information from the Finance Board to set limits on the
dollar value of loans it will insure, which are based on the conforming loan
limit and median home prices. That is, FHA sets an area’s loan limit at the
greater of 48 percent of the conforming loan limit or 95 percent of the
median home sales price for the area, but no greater than 87 percent of the
conforming loan limit.7 Consequently, FHA loan limits vary depending on
the location of the home and the median home sales price there but are no
lower than $115,200 and no higher than $208,800—48 percent and
87 percent, respectively, of the conforming loan limit.

FHA is not required by statute to use a particular source of information on
home prices to determine the median price of homes in an area and,
consequently, the loan limit for the area. However, FHA has chosen to use
the Finance Board survey for this purpose. FHA relies heavily on the survey
to measure median home sales prices because it is the most
comprehensive source of published house price data readily available to
the agency.

The Office of Federal Housing Enterprise Oversight (OFHEO) also collects
information on home sales. Specifically, both Fannie Mae and Freddie Mac


4
 Conventional loans are all home loans not insured or guaranteed by the government. The survey also
excludes refinancings.
5
 The Congress created the enterprises to enhance the availability of credit to qualified borrowers by
purchasing mortgages from lenders, who can then use the proceeds from those sales to make
additional mortgage loans.
6
 12 U.S.C. 1717(b)(2) and 12 U.S.C. 1454(a)(2). Loans greater than the conforming loan limit are called
jumbo loans.
7
 Alaska, Hawaii, Guam, and the Virgin Islands may have even higher loan limits because the Congress
has designated these states and territories as special high-cost areas, allowing FHA to set its loan limits
there up to 50 percent higher than the limits applicable elsewhere. Also, FHA permits its field offices to
use non-Finance Board data to adjust loan limits. Any interested party who does not believe that the
FHA loan limit for a given area accurately reflects the median house prices there may appeal for an
increase. Anyone doing so must submit to FHA sufficient house sales price data to justify granting an
increase.
Page 4                                                             GAO/RCED-99-78 FHA Loan Limits
                        B-281434




                        provide data to OFHEO on all of the mortgages they purchase in order for
                        OFHEO to construct a house price index.8 OFHEO uses the house price index
                        to account for changes in the values of the homes securing the mortgages
                        that the enterprises have purchased and their potential impact on credit
                        risk. By definition, the index includes only conforming loans—those with
                        values less than the conforming loan limit—because neither enterprise
                        may purchase loans that exceed the conforming loan limit. In addition, the
                        index excludes all government-insured loans.9 In 1997, Fannie Mae and
                        Freddie Mac purchased 37 percent of all conventional loans originated
                        that year for single-family homes.


                        In about two-thirds of the 42 metropolitan areas we reviewed, no
The Finance Board       substantive difference existed in the 1997 median house prices calculated
and the Office of       with Finance Board and OFHEO data. As figure 1 shows, in 27 of these
Federal Housing         areas, the difference between the higher and lower estimates of median
                        prices according to the two sets of data was 5 percent or less. In an
Enterprise Oversight    additional 10 areas, the two agencies’ estimates were within 10 percent of
Usually Estimate        each other. For the 15 areas where the two estimates differed by more
                        than 5 percent, the Finance Board estimated a higher median home sales
Similar Median Prices   price than OFHEO in 10 areas, while OFHEO estimated a higher median in 5
                        areas.




                        8
                         The house price index measures changes in home values over time, using the home’s appraised value
                        from a refinancing and either the home’s appraised value or sales price from purchases. OFHEO
                        estimates a value for each home with a repeat transaction and compares that to the home’s previous
                        sales price or appraised value, with any difference between the two becoming the basis for the index.
                        Although OFHEO publishes the index, it does not publish the data that Fannie Mae and Freddie Mac
                        provide it that form the basis for the index. Fannie Mae and Freddie Mac consider those data to be
                        proprietary and confidential.
                        9
                         Fannie Mae and Freddie Mac can and do purchase government-insured loans, but they buy very few,
                        and OFHEO screens these out of its house price index as well as the data it provided us. The
                        Government National Mortgage Association (Ginnie Mae) securitizes nearly all government-insured
                        mortgages, guaranteeing the timely payment of principal and interest on privately issued securities
                        that are backed by pools of government-insured mortgages.


                        Page 5                                                          GAO/RCED-99-78 FHA Loan Limits
                                      B-281434




Figure 1: Degree of Difference
Between 1997 Median House Prices in                                                     Greater than 5% and less than or
42 Metropolitan Areas Using the                                                         equal to 10% (10 areas)
Finance Board’s and OFHEO’s Data
                                                                                        Greater than 10 percent (5 areas)




                                                  • 12%


                                          • 24%

                                                                   64% •                Less than or equal to 5% (27
                                                                                        areas)




                                      Sources: GAO’s analysis based on median house price data from the Federal Housing Finance
                                      Board and the Office of Federal Housing Enterprise Oversight.




                                      Because one basis for measuring the median price—the Finance Board’s
                                      data—does not result in a substantively different price than
                                      another—OFHEO’s data—for about two-thirds of the areas we reviewed, FHA
                                      loan limits in those areas would be similar using either source of data. The
                                      median price determines whether FHA sets its loan limit at a percentage of
                                      the conforming limit or at 95 percent of the median home price.


                                      The Finance Board’s data were a reasonable measure of an area’s median
Officials See Data                    home sales price (for homes with conventional financing), according to
Sources as Mutually                   officials at the Finance Board, Fannie Mae, and Freddie Mac. As support
Supportive, With                      for this position, they cited the similarity in the median prices that the
                                      Finance Board and OFHEO calculated in about two-thirds of the areas we
Differences Largely                   reviewed. Given the differences in the nature of the data the Finance
Due to Jumbo Loans                    Board collects through a survey of lenders versus OFHEO’s data, which




                                      Page 6                                                     GAO/RCED-99-78 FHA Loan Limits
                      B-281434




                      represent all the conforming loans Fannie Mae and Freddie Mac
                      purchased, these officials stated that they viewed each data set as
                      supportive of the other. Regardless of the similarities and differences in
                      the median home prices derived from these two sources, the different
                      methods by which the Finance Board and OFHEO collect home price data
                      could explain some variation between them. For example, while the
                      Finance Board survey is intended to estimate house prices at the national
                      level, estimates at the local level are likely to be less representative of all
                      non-government-insured home sales for that area. These officials added
                      that there would be even fewer differences between the two if we analyzed
                      trends in median price data over a number of years.

                      The Finance Board’s survey included some higher-priced homes that
                      would not be reflected in OFHEO’s data, sometimes resulting in a higher
                      calculated median price than OFHEO’s data reflected. In fact, in 26 of the
                      areas we reviewed, the Finance Board data showed higher median home
                      sales prices than did the OFHEO data. According to officials from Fannie
                      Mae and Freddie Mac, including these jumbo loans is the primary reason
                      for that difference. Specifically, the Finance Board’s data included
                      purchase prices up to $750,000 and loan amounts up to $500,000.10
                      Conversely, OFHEO’s data excluded all jumbo loans because they exceed
                      the conforming loan limit ($214,600 in 1997), meaning neither Fannie Mae
                      nor Freddie Mac could have purchased them.


                      Supplementing the Finance Board’s or OFHEO’s data with information on
Median House Prices   prices of homes financed with government-insured mortgages reduces the
Are Lower When They   estimates of median prices across the board and within all of the
Are Based on          metropolitan areas we reviewed. Homes financed with
                      government-insured mortgages typically cost less than homes financed
Conventional and      with conventional mortgages, but the Finance Board and OFHEO (with few
Government-Insured    exceptions) collect data only on homes financed with conventional
                      mortgages. Hence, including data on government-insured mortgages in the
Loans                 calculation for a given area results in a lower median price of homes.

                      As table 1 shows, the effect of adding data on homes with
                      government-insured mortgages to the Finance Board’s and OFHEO’s median
                      price estimates is not uniform across all of the metropolitan areas; that is,
                      it does not reduce the median in each area by the same amount. When we
                      added data on homes with government-insured loans to the Finance

                      10
                       The Finance Board excludes purchase prices and loan amounts above these thresholds to control for
                      possible data entry errors on the part of the lenders participating in its survey. In 1998, the Finance
                      Board raised these thresholds to $856,000 for purchase prices and $571,000 for loan amounts.



                      Page 7                                                          GAO/RCED-99-78 FHA Loan Limits
                                      B-281434




                                      Board’s data, median prices in individual metropolitan areas were 2 to
                                      30 percent lower. When we added these data to OFHEO’s data, median
                                      prices were 6 to 31 percent lower.

Table 1: Effect of Adding Data on
Government-Insured Loans to Finance   Percentage drop in
                                      estimated median                                          Number of areas
Board and OFHEO Median House Price
Data in 42 Metropolitan Areas         house price when                           Median based on         Median based on
                                      government-insured                   Finance Board data and         OFHEO data and
                                      loans are added                   government-insured loansa government-insured loans
                                      5 percent or less                                                 5                          0
                                      6 to 10 percent                                                 12                           1
                                      11 to 15 percent                                                12                          16
                                      16 to 20 percent                                                  7                         16
                                      More than 20 percent                                              3                          9
                                      a
                                      Finance Board data were not available for this analysis in three metropolitan areas.

                                      Sources: GAO analysis based on data from the Federal Housing Finance Board, the Office of
                                      Federal Housing Enterprise Oversight, FHA, and the Department of Veterans Affairs.



                                      The effect that including government-insured loans has on the estimated
                                      median price of homes in any given area depends on how much
                                      government-insured lending (relative to all other types of lending) was
                                      taking place in that area. Where government-insured lending was a
                                      relatively higher percentage of home loans, median prices decreased by a
                                      greater degree than the decrease for the 42 areas taken as a whole.
                                      Conversely, where there was relatively less government-insured lending in
                                      any given area, median prices also decreased—but to a lesser degree than
                                      in metropolitan areas with more government-insured lending.


                                      FHA is exploring additional data sources to supplement the Finance
FHA Plans to Use                      Board’s data and to improve its own measurement of median house prices.
Additional Data                       In part, this is in recognition of the importance and value of timely and
Sources, Including                    comprehensive data on house prices at the local level for its own purposes
                                      as well as larger, research-oriented uses. Also, recent legislative changes
OFHEO, to Improve                     have made it more important for FHA to have accurate local-area measures
Its Measurement of                    of house prices on which to base loan limits. However, FHA has found no
                                      source that systematically collects house price data on an ongoing basis in
Median Home Sales                     all of the areas—metropolitan areas and counties—for which FHA must set
Prices                                loan limits. As a result, FHA has stepped up its efforts to determine the
                                      availability of, and any limitations associated with, additional data sources
                                      on home prices.



                                      Page 8                                                         GAO/RCED-99-78 FHA Loan Limits
               B-281434




               FHA’s most pressing reason for developing additional data sources is a
               provision in recent legislation mandating that the highest loan limit of any
               county within a metropolitan area must apply to loans insured in all the
               counties in that area. To implement this provision as part of a recent
               comprehensive update of all FHA loan limits, FHA supplemented its primary
               data source, the Finance Board survey, with data from the National
               Association of Realtors and a private marketing firm that collects and sells
               data from real estate transaction records. To a limited extent, FHA also had
               its field staff work with local interested parties, such as realtors’
               associations, to gather sufficient recent data on which to base an estimate
               of an area’s median house price. Nonetheless, FHA officials told us that for
               over half of those areas whose loan limits were not automatically indexed
               to the conforming loan limit, the Finance Board’s survey was their primary
               source of median house price data.11

               FHA’s goal is to comprehensively update all of its loan limits annually and,
               in doing so, to make use of additional data sources to broaden the extent
               to which its estimates of median house prices cover more of the housing
               market. To do so, FHA recently initiated preliminary discussions with OFHEO
               about making use of its data (similar to the data OFHEO provided to us) in
               its next comprehensive update. In addition, FHA is considering obtaining
               data on jumbo loans as well as other loans that Fannie Mae and Freddie
               Mac have not purchased. FHA has no specific timetable for including such
               data, in large part because the sources of the data on some of these loans
               do not include information on house sales prices, which makes using the
               data much more methodologically complex and time-consuming than
               using a database such as OFHEO’s.


               FHA has substantial discretion in choosing the source of median house
Observations   price data it will use to set loan limits because, unlike the conforming loan
               limit, there is no statutory requirement for it to use a specific data source.
               Lacking a nationwide source of data that systematically collects
               comprehensive house price information in each and every area where FHA
               must set loan limits, the agency is left with the challenge of assembling the
               best data available to it. At present, its use of the Finance Board’s survey
               appears reasonable given that the only more comprehensive source of



               11
                 For the recent comprehensive update, FHA used the Finance Board as its data source for 30 of the 42
               areas discussed in this report; for 6 areas, FHA used data its field office staff gathered on local real
               estate sales; and, in the remaining 6 areas, field office staff determined that one county in the area had
               a higher loan limit than the others and invoked the recently enacted provisions applying that county’s
               higher limit to the entire area.



               Page 9                                                            GAO/RCED-99-78 FHA Loan Limits
                  B-281434




                  data that we found—HUD’s Office of Federal Housing Enterprise
                  Oversight—usually yielded a similar median price.

                  Nonetheless, while both the Finance Board and OFHEO offer measures of
                  median prices that capture one particular segment of the housing
                  market—homes with conventional financing—neither covers all of the
                  housing market. As a result, FHA’s efforts to broaden its coverage of the
                  housing market will be guided by a need to identify what its data sources
                  are not capturing and a need to consider the implications for its loan limits
                  and potential FHA borrowers of using any new data sources.


                  We provided a draft of this report to the Department of Housing and Urban
Agency Comments   Development (HUD), the Federal Housing Finance Board (the Finance
                  Board), the Office of Federal Housing Enterprise Oversight, the Federal
                  National Mortgage Association, and the Federal Home Loan Mortgage
                  Corporation for their review and comment.

                  HUD  agreed that the Finance Board’s survey is a reasonable source of home
                  sales price data even though neither the survey nor the Office of Federal
                  Housing Enterprise Oversight’s information on home sales covers the
                  entire housing market. HUD commented that the report effectively
                  describes the practices and resources it used to set FHA loan limits and
                  identifies the data collection obstacles associated with this activity. HUD
                  also provided technical corrections to the report, which we have
                  incorporated. HUD’s comments are included as appendix II of this report.

                  The Finance Board agreed that our analysis indicates its survey is a
                  reasonable measure of 1997 home sales prices in the areas we reviewed.
                  However, the Finance Board also commented that because its data come
                  from a voluntary sample of mortgage lenders, it cannot ensure that its
                  sample size in individual metropolitan areas or counties is large enough to
                  provide statistically reliable results. The Finance Board stated that if the
                  Congress wants HUD to use the survey, it should provide the Finance Board
                  with the authority to require lenders to participate in the survey. We have
                  reported in the past that users of the Finance Board’s data suggested the
                  sample size would need to be expanded to make the data more reliable for
                  measuring local housing prices.12 We have revised our description of the
                  Finance Board’s survey to clarify that lenders participate in it voluntarily.
                  The Finance Board’s comments are included as appendix III of this report.

                  12
                   Housing Finance: Implications of Alternative Methods of Adjusting the Conforming Loan Limit
                  (GAO/RCED-95-6, Oct. 5, 1994).



                  Page 10                                                       GAO/RCED-99-78 FHA Loan Limits
              B-281434




              The Office of Federal Housing Enterprise Oversight provided technical
              corrections and clarifications to the report, which we have incorporated as
              appropriate.

              The Federal National Mortgage Association and the Federal Home Loan
              Mortgage Corporation also commented on the draft report. Both stated
              that they consider the data they provide the Office of Federal Housing
              Enterprise Oversight to be proprietary and confidential. We agreed to add
              this information to the report. The Federal Home Loan Mortgage
              Corporation also provided technical corrections to the report, which we
              have incorporated as appropriate.


              Our review covered homes sold in selected metropolitan areas in calendar
Scope and     year 1997 (1) for which FHA insured or the Department of Veterans Affairs
Methodology   guaranteed the mortgages on the homes; (2) about which the lenders
              issuing the mortgages for the homes reported data on the loans in the
              Federal Housing Finance Board’s monthly interest rate survey; or (3) that
              had mortgages that Fannie Mae or Freddie Mac subsequently purchased,
              reporting data on those loans to OFHEO.13 For this analysis, we focused on
              the 42 metropolitan statistical areas (MSA) for which the Finance Board
              publicly reports data annually.14 By definition, MSAs have at least one city
              with 50,000 inhabitants or are urbanized areas with a total metropolitan
              population of at least 100,000. Most MSAs consist of more than one county.

              For the 42 areas, we obtained (1) from FHA and the Department of Veterans
              Affairs, data on the median price of all of the homes sold for which the
              federal government insured or guaranteed the mortgages15 and (2) from
              the Finance Board, the median purchase price of all the homes sold whose
              mortgages were reflected in the Board’s monthly interest rate survey.
              Using data on loan amounts and loan-to-value ratios, OFHEO calculated and
              provided to us an estimate of the median price of all homes sold in these
              areas that had mortgages that were subsequently purchased by Fannie

              13
                See appendix I for a complete list of these 42 areas.
              14
                The Finance Board also reports data on 18 consolidated metropolitan statistical areas (CMSA), which
              are MSAs that have a population of 1 million or more, separate component areas that are identifiable
              and meet certain standards for designation as a metropolitan area, and local support for the
              component areas. Because the Finance Board and OFHEO were more readily able to provide data
              arrayed by MSA (rather than also disaggregating each CMSA’s data into the component MSAs), we
              focused our analysis on just the 42 MSAs.
              15
                In 1997, the Department of Agriculture’s Rural Housing Service guaranteed $2.4 billion in rural
              housing loans made by private lenders. However, because these loans are a relatively small part of the
              mortgage market and are concentrated in rural—as opposed to metropolitan areas—we have excluded
              them from our analysis.



              Page 11                                                         GAO/RCED-99-78 FHA Loan Limits
B-281434




Mae or Freddie Mac. For this calculation, OFHEO used the data Fannie Mae
and Freddie Mac provide it for the calculation of its house price index
(unlike OFHEO’s house price index, the data it provided us for this review
are not publicly released). We then compared the median purchase prices
according to these sources of data. Because homes financed with
government-insured loans are typically lower priced and neither the
Finance Board nor OFHEO includes data on government-insured mortgages,
we also calculated median purchase prices that included data from FHA
and the Department of Veterans Affairs with the data from OFHEO and the
Finance Board.

Throughout our review, we discussed issues related to data sources for
measuring house price changes with officials from FHA, OFHEO, HUD’s Office
of Policy Development and Research, the Finance Board, Fannie Mae, and
Freddie Mac. We also supplemented this information by discussing these
issues with officials of private organizations having an interest or expertise
in this area, including the National Association of Homebuilders and the
Mortgage Insurance Companies of America. We also discussed the results
of our analysis comparing median prices from the various sources with
officials from the agencies that provided these data.

We did not directly assess the reliability of the data we obtained from FHA,
the Department of Veterans Affairs, OFHEO, or the Finance Board. To
assure ourselves that each data set was sufficiently reliable for our
purposes, we reviewed the procedures each agency had in place to ensure
its data are reliable and accurate.

We conducted our review from July 1998 through March 1999 in
accordance with generally accepted government auditing standards.


We are sending copies of this report to the appropriate congressional
committees; the Honorable Andrew Cuomo, Secretary of Housing and
Urban Development; the Honorable Bruce A. Morrison, Chairman of the
Federal Housing Finance Board; the Honorable Mark Kinsey, Acting
Director of the Office of Federal Housing Enterprise Oversight; the
Honorable Franklin D. Raines, Chairman and Chief Executive Officer of
Fannie Mae; the Honorable Leland C. Brendsel, Chairman and Chief
Executive Officer of Freddie Mac; and the Honorable Jacob J. Lew,
Director of the Office of Management and Budget. We will make copies
available to others upon request.




Page 12                                         GAO/RCED-99-78 FHA Loan Limits
B-281434




Please call me at (202) 512-7631 if you or your staff have any questions
about the material in this report. Major contributors to this report are
listed in appendix IV.

Sincerely yours,




Stanley J. Czerwinski
Associate Director, Housing and
  Community Development Issues




Page 13                                         GAO/RCED-99-78 FHA Loan Limits
Contents



Letter                                                                                           1


Appendix I                                                                                      16
Metropolitan Areas in
Which GAO Compared
Median Home Sales
Price Data
Appendix II                                                                                     17
Comments From the
Department of
Housing and Urban
Development
Appendix III                                                                                    18
Comments From the
Federal Housing
Finance Board
Appendix IV                                                                                     20
Major Contributors to
This Report
Table                   Table 1: Effect of Adding Data on Government-Insured Loans to            8
                          Finance Board and OFHEO Median House Price Data in 42
                          Metropolitan Areas


Figure                  Figure 1: Degree of Difference Between 1997 Median House                 6
                          Prices in 42 Metropolitan Areas Using the Finance Board’s and
                          OFHEO’s Data




                        Page 14                                      GAO/RCED-99-78 FHA Loan Limits
Contents




Abbreviations

CMSA       consolidated metropolitan statistical area
FHA        Federal Housing Administration
HUD        Department of Housing and Urban Development
MSA        metropolitan statistical area
OFHEO      Office of Federal Housing Enterprise Oversight


Page 15                                     GAO/RCED-99-78 FHA Loan Limits
Appendix I

Metropolitan Areas in Which GAO Compared
Median Home Sales Price Data

              Albany-Schenectady-Troy, New York
              Atlanta, Georgia
              Austin-San Marcos, Texas
              Birmingham, Alabama
              Buffalo-Niagara Falls, New York
              Charlotte-Gastonia-Rock Hill, North Carolina-South Carolina
              Columbus, Ohio
              Dayton-Springfield, Ohio
              Fresno, California
              Grand Rapids-Muskegon-Holland, Michigan
              Greensboro-Winston-Salem-High Point, North Carolina
              Greenville-Spartanburg-Anderson, South Carolina
              Hartford, Connecticut
              Honolulu, Hawaii
              Indianapolis, Indiana
              Jacksonville, Florida
              Kansas City, Missouri-Kansas
              Las Vegas, Nevada-Arizona
              Louisville, Kentucky-Indiana
              Memphis, Tennessee-Arkansas-Mississippi
              Minneapolis-St. Paul, Minnesota-Wisconsin
              Nashville, Tennessee
              New Orleans, Louisiana
              Norfolk-Virginia Beach-Newport News, Virginia-North Carolina
              Oklahoma City, Oklahoma
              Omaha, Nebraska-Iowa
              Orlando, Florida
              Phoenix-Mesa, Arizona
              Pittsburgh, Pennsylvania
              Providence-Fall River-Warwick, Rhode Island-Massachusetts
              Raleigh-Durham-Chapel Hill, North Carolina
              Richmond-Petersburg, Virginia
              Rochester, New York
              Salt Lake City-Ogden, Utah
              San Antonio, Texas
              San Diego, California
              St. Louis, Missouri-Illinois
              Syracuse, New York
              Tampa-St. Petersburg-Clearwater, Florida
              Tucson, Arizona
              Tulsa, Oklahoma
              West Palm Beach-Boca Raton, Florida



              Page 16                                     GAO/RCED-99-78 FHA Loan Limits
Appendix II

Comments From the Department of Housing
and Urban Development




              Page 17         GAO/RCED-99-78 FHA Loan Limits
Appendix III

Comments From the Federal Housing
Finance Board




               Page 18        GAO/RCED-99-78 FHA Loan Limits
Appendix III
Comments From the Federal Housing
Finance Board




Page 19                             GAO/RCED-99-78 FHA Loan Limits
Appendix IV

Major Contributors to This Report


                        Judy A. England-Joseph
Resources,              DuEwa A. Kamara
Community, and          Bill MacBlane
Economic                Mathew Scire
Development
Division, Washington,
D.C.




(385768)                Page 20                  GAO/RCED-99-78 FHA Loan Limits
Ordering Information

The first copy of each GAO report and testimony is free.
Additional copies are $2 each. Orders should be sent to the
following address, accompanied by a check or money order
made out to the Superintendent of Documents, when
necessary. VISA and MasterCard credit cards are accepted, also.
Orders for 100 or more copies to be mailed to a single address
are discounted 25 percent.

Orders by mail:

U.S. General Accounting Office
P.O. Box 37050
Washington, DC 20013

or visit:

Room 1100
700 4th St. NW (corner of 4th and G Sts. NW)
U.S. General Accounting Office
Washington, DC

Orders may also be placed by calling (202) 512-6000
or by using fax number (202) 512-6061, or TDD (202) 512-2537.

Each day, GAO issues a list of newly available reports and
testimony. To receive facsimile copies of the daily list or any
list from the past 30 days, please call (202) 512-6000 using a
touchtone phone. A recorded menu will provide information on
how to obtain these lists.

For information on how to access GAO reports on the INTERNET,
send an e-mail message with "info" in the body to:

info@www.gao.gov

or visit GAO’s World Wide Web Home Page at:

http://www.gao.gov




PRINTED ON    RECYCLED PAPER
United States                       Bulk Rate
General Accounting Office      Postage & Fees Paid
Washington, D.C. 20548-0001           GAO
                                 Permit No. G100
Official Business
Penalty for Private Use $300

Address Correction Requested