This is the accessible text file for GAO report number GAO-07-879 
entitled 'Federal Housing Administration: Agency Should Asses the 
Effects of Proposed Changes to Manufactured Home Loan Program' which 
was released on August 29, 2007. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as part 
of a longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the printed 
version. The portable document format (PDF) file is an exact electronic 
replica of the printed version. We welcome your feedback. Please E-mail 
your comments regarding the contents or accessibility features of this 
document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

Report to Congressional Requesters: 

United States Government Accountability Office: 

GAO: 

August 2007: 

Federal Housing Administration: 

Agency Should Assess the Effects of Proposed Changes to the 
Manufactured Home Loan Program: 

FHA Title I Loan Program: 

GAO-07-879: 

GAO Highlights: 

Highlights of GAO-07-879, a report to congressional requesters 

Why GAO Did This Study: 

Pending legislation to the Federal Housing Administration’s (FHA) Title 
I Manufactured Home Loan program would increase loan limits, insure 
each loan, incorporate stricter underwriting requirements, and set up-
front premiums. GAO was asked to review (1) selected characteristics of 
manufactured housing and the demographics of the owners; (2) federal 
and state consumer protections for owners of manufactured homes; and 
(3) the potential benefits and costs of the proposed changes for 
borrowers and the federal government. In addressing these objectives, 
GAO analyzed select Census data; researched federal laws and laws in 
eight states; interviewed local, state, and federal officials; and 
analyzed various scenarios that might affect Title I program costs. 

What GAO Found: 

According to 2005 American Housing Survey data, most manufactured homes 
(factory-built housing designed to meet the national building code) 
were located in rural areas in southern states, and most were occupied 
by lower-income owners rather than renters. Although the market for new 
manufactured homes declined substantially from 1996 to 2005, buyers 
increasingly bought larger homes and placed them on private property 
rather than in manufactured home parks. In addition, some states are 
experiencing park closures, with the properties being converted to 
other uses. Overall, manufactured homes can be an affordable housing 
option, with monthly housing costs lower than for other housing types. 

Owners of manufactured homes generally have more consumer protections 
if their homes are considered real rather than personal property, but 
protections provided by laws in the states GAO examined vary. Consumer 
protections extending to lending and settlement processes for personal 
property loans are not as broad as those for real property loans 
(mortgages). Also, delinquent Title I borrowers can be subject to 
repossession or foreclosure, but the consumer protections for 
repossession are often less extensive than those for foreclosure. State 
laws give owners of manufactured homes on leased land varying levels of 
notice, protection, and compensation related to length of leases, rent 
increases, evictions, and park closures. 

According to some FHA and lending officials, potential benefits of the 
proposed changes for borrowers include loans big enough to buy larger 
homes and more financing as more lenders participate in the program. 
The program insured about 24,000 loans in 1990 but only about 1,400 
loans representing $54 million in mortgage insurance in 2006. While the 
changes could benefit borrowers, according to FHA and the Congressional 
Budget Office, the potential costs could expand the government’s 
liability. To gain an understanding of the effects of the proposed 
changes, GAO presented various scenarios. Although risk factors unique 
to manufactured home lending (such as placement on leased land) as well 
as commonly used predictors of loan performance (such as credit scores) 
are associated with default risk, these data were not available. 
Instead, GAO modeled different variations of borrower default risk and 
other factors (such as premiums and lender recovery) that were based on 
the experience of FHA loans to illustrate how variations in these key 
factors could affect potential gains and losses to FHA’s General 
Insurance Fund. The analysis suggests that in all instances where 
borrowers had medium or high default risk, the fund would experience a 
loss. However FHA has not articulated which borrowers would be served, 
how the loans would be underwritten and priced under a risk-based 
structure, or collected data on credit scores and land ownership type. 
FHA explained that among other reasons, it had not done so because the 
Title I program was currently a low-volume program. As a result, the 
effects of the proposed changes are unclear. 

What GAO Recommends: 

GAO recommends that the Secretary of Housing and Urban Development 
(HUD) direct FHA to assess the effects of the proposed changes to the 
Title I program and develop an approach for collecting the information 
needed to effectively manage the program. HUD agreed with these 
recommendations. 

[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-07-879]. 

To view the full product, including the scope and methodology, click on 
the link above. For more information, contact William B. Shear at (202) 
512-8678 or shearw@gao.gov 

[End of section] 

Contents: 

Letter: 

Results in Brief: 

Background: 

Manufactured Homes Were More Likely to Be Located in Rural Areas in the 
South, Owned by Lower-Income Individuals, and Cost Less Than Other 
Types of Housing: 

Owners of Manufactured Homes Have More Consumer Protections If Homes 
Are Considered Real Rather Than Personal Property, and Protections 
Provided by States Vary: 

More Information Is Needed to Determine the Impact of Proposed Changes 
to the Title I Program: 

Conclusions: 

Recommendations for Executive Action: 

Agency Comments and Our Evaluation: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: Scenario Analysis Methodology: 

Appendix III: Comments from the Department of Housing and Urban 
Development: 

Appendix IV: GAO Contact and Staff Acknowledgments: 

Tables: 

Table 1: Differences between Manufactured, Modular, and Site-Built 
Homes: 

Table 2: Characteristics of Eight States Selected for Semistructured 
Interviews: 

Figures: 

Figure 1: Examples of Manufactured Homes, Single-Wide and Double-Wide: 

Figure 2: FHA Title I Loans Received by Top 20 States, 1990-2005: 

Figure 3: Concentration of Manufactured Housing, by Census Tract, 2000: 

Figure 4: Manufactured Housing as a Percentage of All Occupied Housing, 
by Region, 2005: 

Figure 5: Percentages of Manufactured Home Occupants Who Were Owners 
and Renters, 2005: 

Figure 6: Income Characteristics of Owners of Manufactured and Site- 
Built Homes and Apartment Renters, 2005: 

Figure 7: Percentage of Owners of Manufactured and Site-Built Homes and 
Apartment Renters in Selected Income Categories, 2005: 

Figure 8: Number of Manufactured Homes Purchased and Placed, by Size, 
1996 and 2005: 

Figure 9: New Manufactured Home Placements, Owned versus Leased Land, 
1996-2005: 

Figure 10: Examples of Affordable Manufactured Homes in New 
Subdivisions That Look Similar to Site-Built Homes: 

Figure 11: Definition of How Many Homes Constitute a Manufactured Home 
Park and Presence of Licensing Requirements, for the States We 
Reviewed, 2007: 

Figure 12: Monthly Housing Costs of Owners of Manufactured and Site- 
Built Homes, 2005: 

Figure 13: Requirements for Written Leases and Notices of Rent 
Increases in the States We Reviewed, 2007: 

Figure 14: State Provisions for Displaced Occupants of Manufactured 
Homes in the States We Reviewed, 2007: 

Figure 15: Comparison of Installation Inspection Programs in the States 
We Reviewed, 2007: 

Figure 16: Number of FHA Title I Loans and Percentage of Loans in 
Default, 1990-2006: 

Figure 17: Results of Scenarios Based on Variations in Default Risk, 
Lender Recovery Costs, and Premiums Paid by Borrowers: 

Figure 18: Number of FHA Loans and Percentage of Loans in Default, 1990-
2006: 

Figure 19: Assumptions of Default Risk Used in Our Analysis: 

Abbreviations: 

AHS: American Housing Survey: 
CBO: Congressional Budget Office: 
FHA: Federal Housing Administration: 
HUD: Department of Housing and Urban Development: 
MHS: Manufactured Homes Survey: 
RESPA: Real Estate Settlement Procedures Act: 
TILA: Truth in Lending Act: 

United States Government Accountability Office: 

Washington, DC 20548: 

August 24, 2007: 

The Honorable Christopher J. Dodd: 
Chairman: 
Committee on Banking, Housing, and Urban Affairs: 
United State Senate: 

The Honorable Charles E. Schumer: 
Chairman: 
Subcommittee on Housing, Transportation, and Community Development: 
Committee on Banking, Housing, and Urban Affairs: 
United State Senate: 

The Honorable Jack Reed: 
United States Senate: 

Manufactured housing (factory-built housing designed to meet the 
Department of Housing and Urban Development's (HUD) national building 
code) provides affordable housing for approximately 17 million 
Americans. Relative to other forms of housing (generally referred to as 
site-built housing), a manufactured home can be more affordable, 
particularly when a home is purchased without the cost of the land. In 
such a purchase, the home generally is not titled as real property, but 
is considered personal property or chattel, which denotes property that 
is movable and personal, such as an automobile or furniture. 
Consequently, lending for manufactured homes differs from other home 
lending because prospective buyers can receive either a real estate or 
a personal property loan. 

Currently, the Federal Housing Administration (FHA) of HUD offers the 
only active federal loan guarantee program that includes an option for 
a "home-only" product; that is, a personal property loan for the 
purchase of a manufactured home without the land on which the home will 
be located. However, FHA officials explained that the purpose of the 
Title I Manufactured Home Loan Program (Title I)--to protect mortgage 
lenders against the risk of default through insurance or a guarantee-- 
is not currently being met because the current design of the program 
passes the majority of the insurance risk to the lenders who in turn 
charge borrowers higher interest rates. In addition, the lending market 
associated with manufactured homes has undergone significant changes 
over the last 15 years. Market growth in the 1990s was followed by a 
large number of repossessions from 2000 to 2002. Because of the amount 
of origination fees manufactured home lenders received, some lenders 
focused on increasing sales volume to the detriment of assessing 
borrowers' creditworthiness. As a result, officials and literature 
suggest that the quality of the manufactured home loan pool began to 
deteriorate and less creditworthy borrowers began to default on their 
loans, causing a high number of repossessions (personal property is 
repossessed rather than foreclosed). Subsequently, many lenders exited 
this market, resulting in a decrease in the availability of private 
financing for manufactured homes. Furthermore, loan volume generated 
through Title I declined by 94 percent from 1990 to 2006, with 1,438 
loans insured in 2006 representing $54 million in mortgage insurance. 

In addition to the relative scarcity of financing for manufactured 
homes, the owners of manufactured homes are in a unique position 
relative to other homeowners in terms of the federal and state consumer 
protections applicable when they buy, finance, and occupy the housing. 
That is, whether the manufactured home is considered real or personal 
property affects what consumer protections apply and what recourse is 
available. For example, consumer and tenant protections, particularly 
at the state level, are especially pertinent for owners of manufactured 
homes who lose leases to their underlying ground because they lived on 
land sold for commercial or other residential development. Some owners 
of manufactured homes found that their homes lost most or all of their 
value in such situations. And, according to state and local officials, 
the loss of manufactured housing parks, particularly in growing 
metropolitan areas, has exacerbated the shortage of affordable housing. 

Legislation has been introduced to make changes to the Title I program 
that may increase the demand for and availability of loans for 
manufactured homes.[Footnote 1] With the potential expansion of the 
program, you asked us to review the proposed legislation and consumer 
protections available for owners of manufactured homes. Specifically, 
this report (1) describes selected characteristics of manufactured 
housing and the demographics of the owners, (2) compares federal and 
state consumer and tenant protections for owners of manufactured homes, 
and (3) describes the proposed changes to FHA's Title I Manufactured 
Home Loan program and assesses potential benefits and costs to 
borrowers and the federal government. 

To determine selected characteristics of manufactured housing and the 
demographics of the owners, we analyzed Census data from the 
Manufactured Housing and American Housing Surveys. Specifically, we 
used Manufactured Housing Survey data from 1996 through 2005 to examine 
changes in the manufactured housing industry, such as the number of 
homes placed and the size of these homes. We relied on 2005 American 
Housing Survey data to provide information on the demographics of the 
manufactured homeowner. We did not use earlier years of American 
Housing Survey data because the sample of manufactured homes in the 
survey changed in 2005.[Footnote 2] However, in both the Manufactured 
Housing and American Housing Surveys, data about the ownership of the 
land (that is, owned or leased) on which the home is placed are 
limited. We assessed the reliability of the Manufactured Housing and 
American Housing Surveys by reviewing information about the data, 
performing electronic data testing to detect errors in completeness and 
reasonableness, and interviewing knowledgeable officials regarding the 
quality of the data. We determined that the data were sufficiently 
reliable for the purposes of this report. To compare federal and state 
consumer and tenant protections available for owners of manufactured 
homes, we researched relevant federal laws and laws in the eight states 
(Arizona, Florida, Georgia, Missouri, New Hampshire, North Carolina, 
Oregon, and Texas) that we selected for our review. The eight states 
were selected based on a combination of factors including the volume of 
FHA Title I loans in the state from 1990 through the first quarter of 
2007; the concentration of manufactured housing as a percentage of 
housing units in the state; information from our interviews of industry 
and consumer officials; and previous studies conducted on manufactured 
housing. In each state, we conducted semistructured phone interviews 
with the state regulator, representatives of the manufactured housing 
industry, and a consumer group, such as the state manufactured 
homeowners association.[Footnote 3] To assess the potential costs and 
benefits of the proposed changes to the Title I program, we conducted a 
literature review; interviewed FHA officials, FHA lenders, Ginnie Mae 
officials, and officials from federal and other lending programs, such 
as Fannie Mae, Freddie Mac, U.S. Department of Agriculture Rural 
Housing Service, and the Department of Veterans Affairs; and reviewed 
policies and procedures from programs that provide financing for 
manufactured homes to determine what additional factors, such as 
default risk and the location of the manufactured home, could be 
considered to mitigate risk for lending for manufactured housing. To 
illustrate potential costs and other effects of the proposed 
legislation, we conducted an analysis of different scenarios of 
potential loan performance that incorporated these additional factors. 
We used this approach because we did not have sufficient data on the 
credit scores of FHA borrowers or the location of the homes to perform 
a more in-depth analysis. Appendixes I and II contain additional 
information about our methodology. We conducted our work in Washington, 
D.C., Atlanta, and Chicago, from October 2006 through June 2007 in 
accordance with generally accepted government auditing standards. 

Results in Brief: 

Available data on selected characteristics of manufactured homes and 
their owners in 2005 indicate that most manufactured homes were located 
in rural areas, more were located in southern states than in other 
regions, and most were occupied by lower-income owners rather than 
renters. For example, according to data from the American Housing 
Survey, almost 50 percent of all owners of manufactured homes earned 
less than $30,000 compared with 23 percent of owners of site-built 
homes in 2005. Although the market for new manufactured homes declined 
significantly, from about 332,000 manufactured homes sold in 1996 to 
about 118,000 homes in 2005, buyers increasingly bought larger homes 
and placed them on private property rather than in manufactured home 
parks. According to local officials we interviewed, few new 
manufactured home parks have been built since the early 1980s, largely 
as a result of local zoning issues, and some states are experiencing 
park closures with the properties being converted to other uses. But 
overall, manufactured homes can be an affordable housing option, with 
monthly housing costs considerably lower than other housing types. For 
example, according to 2005 American Housing Survey data, more than 50 
percent of the owners of manufactured homes had monthly housing costs 
of $100 to $499.[Footnote 4] In comparison, a little more than 25 
percent of owners of site-built homes paid $100 to $499 in monthly 
housing costs in 2005. However, while manufactured homes can be more 
affordable than other housing types and are often thought to be mobile, 
few placement opportunities and the high cost of moving the homes 
limits their mobility. 

Owners of manufactured homes generally have more consumer protections 
if their homes are considered real rather than personal property, but 
the laws in the states we visited provide varying protections. Consumer 
protections that extend to the lending and settlement process for Title 
I personal property loans are not as broad as those for real property 
loans (mortgages). For example, Title I borrowers who obtain a home- 
only loan (that is, a personal property loan) are not entitled to the 
settlement cost disclosures of the Real Estate Settlement Procedures 
Act. Further, delinquent borrowers can be subject to repossession (if 
the loan was for personal property) or foreclosure (if for real 
property), but the consumer protections for repossession are often less 
extensive than those for foreclosure. In addition, state laws give 
owners of manufactured homes on leased land varying levels of notice, 
protection, and compensation related to length of leases, rent 
increases, evictions, and park closures. For example, in the states we 
reviewed, notice requirements for rent increases range from 60 to 90 
days, but not all the states have provisions on rent increases. 
Further, states vary in what programs or tools are available to help or 
compensate tenants displaced because of park closures. For instance, 
the states of Arizona, Florida and Oregon offer financial help through 
relocation funds or tax credits, but the remaining five states we 
reviewed do not offer such aid. As a result, purchasers of manufactured 
homes who do not own the land underneath their home and experience 
adverse conditions over which they have no control, such as rent 
increases or park closures, have differing degrees of recourse 
depending on the state in which they live. 

Legislative proposals to change FHA's Title I program would increase 
loan limits, insure each loan made, incorporate stricter underwriting 
requirements, establish up-front insurance premiums, and adjust the 
annual premium; however, the effects of the proposed changes remain 
unclear. For instance, limits for a home-only loan would rise from 
$48,600 to $69,678, loan guarantees would apply to individual loans 
rather than be capped at 10 percent of the value of a lender's 
portfolio, and underwriting requirements would be revised with the 
stated intent of strengthening the financial soundness of the program. 
According to some FHA and industry officials, the potential benefits of 
proposed changes for borrowers include obtaining loan amounts 
sufficient to buy larger homes, additional financing as more lenders 
likely would participate in a program where a greater portion of their 
portfolios could be insured, and an expansion of the secondary market 
that could provide more liquidity for lenders to make more loans. 
However, the ability of the owner of a manufactured home to build 
equity may be limited when the land is leased, which also often 
increases the risks associated with the loan. For instance, if a 
borrower with a home on leased land were to default, lenders could face 
higher costs and lower recoveries (relative to site-built homes) in 
trying to repossess, move, and resell the personal property. To gain an 
understanding of the effects of the proposed changes, we developed 
various scenarios. Although risk factors unique to manufactured home 
lending (such as placement on leased land) as well as commonly used 
predictors of loan performance (such as credit scores) are associated 
with default risk, these data were not available. Instead, we modeled 
different variations of borrower default risk and other factors (such 
as premiums and lender recovery) that were based on the experience of 
FHA loans to illustrate how variations in these key factors affect 
potential gains and losses to FHA's General Insurance Fund.[Footnote 5] 
The results of our analysis show that in all cases when borrowers had 
medium or high default risk, the fund experienced a loss. But FHA has 
not yet articulated which borrowers would be served if the program were 
expanded, specified changes in its underwriting requirements, developed 
a risk-based pricing structure for the proposed legislation, estimated 
costs to the General Insurance Fund, or collected data on credit scores 
and land ownership type. Our internal control standards for the federal 
government require that an agency identify risks that may be posed by 
new legislation.[Footnote 6] FHA officials have stated that they have 
not made those risk assessments because the current volume of the Title 
I program is low and they did not know if the legislation would pass. 
They said they devoted their resources to making changes to the much 
larger Title II program, which guarantees loans for single-family home 
mortgages. As a result, the effects the proposed legislation may have 
on the volume of lending and claims and the overall financial soundness 
of the program are unclear. 

GAO recommends that the Secretary of Housing and Urban Development 
direct FHA to assess the effects of the proposed changes to the Title I 
program and develop an approach for collecting the information needed 
to effectively manage the program. 

We provided HUD with a draft of this report for review and comment. HUD 
provided comments in a letter from the Assistant Secretary for Housing-
-Federal Housing Commissioner (see app. III). HUD agreed with the 
recommendations in our report and described plans for implementing the 
recommendations. 

Background: 

Manufactured homes differ from site-built homes based on how they are 
constructed, classified, financed, and appraised, with many differences 
resulting from the home's status as either real or personal property. 
Manufactured home parks have a variety of ownership models, ranging 
from sole to corporate ownership and including cooperative and 
nonprofit ownership as well. FHA's Title I program dates to 1969, where 
it has served primarily low-income individuals and the majority of the 
lending has been geographically concentrated.[Footnote 7] 

Differences between Manufactured, Modular, and Site-Built Homes: 

The National Manufactured Housing Construction and Safety Standards Act 
of 1974 set a national building code for the construction of 
manufactured homes, known as the HUD Code, which became effective on 
June 15, 1976.[Footnote 8] For the purposes of this report, we define 
manufactured homes as factory-built housing units designed to meet the 
HUD Code. Manufactured homes can be single-wide, double-wide, or multi- 
wide (see fig. 1). The federal standards regulate manufactured housing 
design and construction, strength and durability, transportability, 
fire safety, and energy efficiency.[Footnote 9] Units constructed and 
completed prior to June 15, 1976, are not considered HUD-approved and 
generally are considered mobile homes. 

Figure 1: Examples of Manufactured Homes, Single-Wide and Double-Wide: 

[See PDF for image] 

Source: GAO.

[End of figure] 

Every home built to the HUD Code is identified with a red metal tag, 
known as the HUD certification label.[Footnote 10] This distinguishes 
manufactured homes from modular homes. Both types of homes are factory- 
built, but modular home "modules" are then assembled on a site. And, 
unlike manufactured homes that are federally regulated under a national 
building code, modular homes must meet the state, local, or regional 
building codes where the home is to be sited. Finally, site-built 
housing is constructed on a lot and must meet local building codes (see 
table 1). 

Table 1: Differences between Manufactured, Modular, and Site-Built 
Homes: 

Manufactured home; 
Type of building code: HUD Code (1976); 
Production method: Factory built. Single/multiple sections transported 
to site for installation; 
Qualifies as: Personal property: [Check]; 
Qualifies as: Real property: [Check]. 

Mobile home; 
Type of building code: Pre-HUD Code (built prior to June 15, 1976); 
Production method: Factory built, to voluntary industry standards later 
enforced by most states; 
Qualifies as: Personal property: [Check]; 
Qualifies as: Real property: [Check]. 

Modular home; 
Type of building code: State, local or regional codes; 
Production method: Factory built. Modules are transported to site for 
assembly; 
Qualifies as: Personal property: [Empty]; 
Qualifies as: Real property: [Check]. 

Site-built home; 
Type of building code: Local codes; 
Production method: Built on-site; 
Qualifies as: Personal property: [Empty]; 
Qualifies as: Real property: [Check]. 

Source: GAO. 

[End of table] 

Manufactured Homes Can Be Considered Personal or Real Property, with 
Corresponding Differences in Financing and Appraisals: 

Unlike site-built homes, which are titled as real property and usually 
financed through a mortgage, a manufactured home may be financed as 
personal property or as real property. When a homebuyer purchases a 
manufactured home without tying the purchase to land and does not title 
the home as real property, the home is generally considered personal 
property, or chattel, which denotes property that is movable and 
personal, such as an automobile or furniture. Private sources--such as 
national consumer-finance companies and manufactured home lending 
specialists who work directly with manufactured home dealers and also 
through FHA Title I approved lenders--provide home-only or personal 
property financing, which is more akin to a consumer loan such as an 
automobile loan than a mortgage. Typically, these loans have higher 
interest rates than mortgages due to factors such as quick credit 
approval and their availability to those with marginal credit 
histories. To begin the process, a customer submits a credit 
application to the manufactured home lending specialist, who may or may 
not be affiliated with the dealership. The credit application also may 
be sent to a local bank. The lender reviews the applicant's credit and 
makes a decision on whether to approve a loan. Manufactured homes not 
considered real property do not undergo market-based appraisals. 
Instead, they undergo a loan-to-invoice appraisal, where the 
manufacturer's certified invoice, in effect, substitutes for an 
appraisal. 

In contrast, when a manufactured home is attached to the underlying 
land by a permanent foundation and the home and the land are treated as 
a single real estate package under state law, the home is generally 
considered real property and borrowers can obtain conventional real 
estate mortgages, which include conventional and government-assisted 
mortgage financing obtained through traditional mortgage lenders. Home 
and land financing for manufactured homes is similar to conventional 
mortgage lending for site-built housing. Manufactured homes that are 
financed using a conventional real estate mortgage undergo an appraisal 
that factors the location into the appraised value and also includes 
comparable prices of manufactured homes. 

Ownership Types of Manufactured Home Parks: 

Manufactured homes can be placed on either private property where the 
homeowner typically owns the land or in a manufactured home park. In a 
manufactured home park, also known as a mobile home park or a land- 
lease community, owners of manufactured homes pay rent for the land 
underneath the homes in addition to the loan payments they make for the 
units (the homes).[Footnote 11] The park owner typically provides 
sewer, water, electrical systems, landscaping, and maintains the roads 
and other common areas. 

Manufactured home parks have a variety of ownership models. Investors, 
ranging in size from small family operations to large conglomerates 
that own several properties across the country, own most of the 
manufactured home parks. Tenants of these parks may or may not have a 
lease and have no control over rent increases. According to officials 
we interviewed, in states such as Florida, California, and New 
Hampshire, resident-owned communities are more prevalent; that is, park 
tenants collectively purchased their community by forming either a for- 
profit or nonprofit cooperative corporation.[Footnote 12] Cooperative 
ownership allows residents to control the land by buying memberships or 
shares in the corporation and have more control over membership dues 
increases. Another ownership model involves a land trust, typically run 
through a nonprofit organization, in which the nonprofit owns the land 
and ensures against the possibility of sale or foreclosure of the land. 

Title I Manufactured Home Loan Program: 

FHA first insured loans for manufactured housing in 1969, under a 
program that came to be known as the Title I Manufactured Home Loan 
Program. The program was created to reduce the risk to lenders through 
insurance or a guarantee and encourage lenders to finance manufactured 
homes, which had traditionally been financed as personal property 
through comparatively high-interest, short-term consumer installment 
loans. Under Title I, FHA can guarantee loans for manufactured homes, 
for manufactured homes and the property on which they are located, or 
for the purchase of a manufactured home lot. FHA insures Title I 
manufactured home loans under the General Insurance Fund, which is 
supported by lenders' insurance premiums (currently an annual premium 
of 1 percent, based on the initial loan amount). Since 1998, three 
lenders have originated the majority of Title I loans. 

Almost all of Title I loans are for the manufactured home-only loans 
rather than for home-and-land or land-only loans. In 2005, FHA Title I 
Manufactured Home lending accounted for only 2.8 percent of the 
personal property loan market; conventional lending accounted for the 
remainder. According to data from FHA, from 2004 to mid-2007, 66 
percent of FHA Title I borrowers were 34 years or younger compared with 
2.7 percent who were 65 years or older.[Footnote 13] From 2004 to mid- 
2007, the majority (73 percent) of the borrowers had a monthly income 
from $1,000 to $3,000 (or approximately $12,000-$36,000 annually). From 
1990 to 2005, the majority of FHA Title I lending has been in southern 
states. Twenty states, primarily in the South, Southwest, and the 
Midwest, received more than 85 percent of the FHA Title I loans (see 
fig. 2). 

Figure 2: FHA Title I Loans Received by Top 20 States, 1990-2005: 

[See PDF for image] 

Source: GAO analysis of FHA data; Art Explosion (map). 

[End of figure] 

FHA's Insurance Operations Division administers the Title I program, as 
well as a property improvement program.[Footnote 14] The majority of 
the staff and budget allocations are for the property improvement 
program. In fiscal year 2006, the division had a staff of nine and a 
total budget of $1.1 million, approximately $350,000 of which supported 
the manufactured home loan program. 

Manufactured Homes Were More Likely to Be Located in Rural Areas in the 
South, Owned by Lower-Income Individuals, and Cost Less Than Other 
Types of Housing: 

Available data on geographic and demographic characteristics of 
manufactured homes and their owners indicate that most manufactured 
homes were located in rural areas of the South and were occupied by 
lower-income earners who owned, rather than rented, the homes. The 
market for new manufactured homes declined significantly from 1996 to 
2005, but homes that were purchased were larger in size and more often 
placed on private property. Although limited data were available on the 
number of manufactured home parks, regulatory, industry, and consumer 
officials from seven of the eight states in which we conducted 
interviews told us that manufactured home parks were closing because 
rising land values were driving redevelopment. Housing costs for 
manufactured homes were lower than costs for other housing types; 
however, the costs of moving manufactured homes were relatively high 
and options for placing homes in new locations were few, which affected 
owners' mobility. 

The Majority of Manufactured Homes Were Found in Rural Areas, Mainly in 
the South, and Owned by Low-Income Earners: 

Manufactured homes were located in every state, but were most often 
located in rural areas. According to 2000 Census data, manufactured 
homes were more concentrated in rural areas, particularly in the South 
and desert Southwest, as a share of total housing units (see fig. 3). 
In 2005, according to data from the American Housing Survey, 
approximately 6 percent of occupied homes in the U.S. are manufactured 
homes. The majority of the occupied manufactured homes (68.5 percent) 
were located in rural areas, while 31.5 percent were found in suburban 
areas and central cities. State, industry, and consumer officials in 
more than half of the states we reviewed also told us that manufactured 
homes were more likely to be located in either rural or suburban parts 
of their states. 

Figure 3: Concentration of Manufactured Housing, by Census Tract, 2000: 

[See PDF for image] 

Source: 2000 Census (data); MapInfo (map). 

[End of figure] 

Compared regionally, manufactured homes represented a larger share of 
occupied homes in the South than in other areas of the nation. For 
instance, 10 percent of occupied housing in the South consisted of 
manufactured homes, compared with 6 percent in the West, 5 percent in 
the Midwest, and 2 percent in the Northeast (see fig. 4). Overall, in 
2005, 57 percent of occupied manufactured homes were located in the 
South, 19 percent in the West, 17 percent in the Midwest, and 7 percent 
in the Northeast.[Footnote 15] 

Figure 4: Manufactured Housing as a Percentage of All Occupied Housing, 
by Region, 2005: 

[See PDF for image] 

Source: GAO analysis of AHS 2005 data; Art Explosion (map). 

[End of figure] 

Our analysis of 2005 American Housing Survey data showed that more 
occupants of manufactured homes were owners than renters (see fig. 5). 
A majority (79.5 percent) of those living in manufactured homes owned 
their homes, compared with 17.4 percent who rented their manufactured 
homes.[Footnote 16] 

Figure 5: Percentages of Manufactured Home Occupants Who Were Owners 
and Renters, 2005: 

[See PDF for image] 

Source: GAO analysis of AHS 2005 data. 

[End of figure] 

Although those who lived in manufactured housing were more likely to 
own their homes, they tended to have lower annual incomes (see fig. 6). 
More owners of single-wide and double-wide homes earned less than 
$49,999 compared with owners of site-built homes, who were more likely 
to earn $50,000 or more. For example, in 2005, of all owners of single- 
wide homes, 15.1 percent earned $10,000 or less annually and 23.6 
percent earned from $10,000 to $19,999. In comparison, 6 percent of 
owners of site-built homes earned $10,000 or less and 8.3 percent 
earned from $10,000 to $19,999. 

Figure 6: Income Characteristics of Owners of Manufactured and Site- 
Built Homes and Apartment Renters, 2005: 

[See PDF for image] 

Source: GAO analysis of AHS 2005 data. 

[End of figure] 

Almost half of all owners of manufactured homes earned less than 
$30,000 in 2005 (see fig. 7). More specifically, 49.4 percent of owners 
of manufactured homes earned this amount compared with 23.4 percent of 
owners of site-built homes. Officials we interviewed from six states 
told us that owners of manufactured homes were more likely to be low- 
income individuals. Apartment renters also were proportionally lower- 
income than owners of site-built homes, with 56.8 percent earning less 
than $30,000. 

Figure 7: Percentage of Owners of Manufactured and Site-Built Homes and 
Apartment Renters in Selected Income Categories, 2005: 

[See PDF for image] 

Source: GAO analysis of AHS 2005 data.  

[End of figure] 

Fewer, but Larger, Homes Were Sold and Placed on Private Property, but 
the Number of Manufactured Home Parks Is Unknown: 

The total number of new manufactured homes sold decreased from 1996 to 
2005. According to Census data from the Manufactured Housing Survey, 
332,000 new manufactured homes were sold in 1996 compared with 118,000 
sold in 2005, a net decrease of 64.5 percent. California and Florida 
had the highest number of new manufactured home units sold in 2005, a 
change from 1996 when North Carolina and Texas reported the highest 
number sold. According to officials that we interviewed, several 
factors may have contributed to the decrease in manufactured home 
sales, such as lower interest rates available for site-built homes, the 
decrease in available financing for manufactured homes due to 
consolidation experienced in the industry, and a large number of 
repossessions that flooded the market with units and increased the 
supply of manufactured homes. For example, as a result of the decrease 
in financing options for manufactured homes, industry officials 
explained that manufacturers lowered production of manufactured homes 
and instead built more modular homes, because more financing options 
were available. Modular homes can often be built in the same factory as 
manufactured housing but are not required to meet the HUD Code. 

Although consumers purchased fewer new manufactured homes in 2005 than 
in 1996, according to the Census data from the Manufactured Housing 
Survey, they bought more double-wide or multisection homes. In 2005, 76 
percent of the manufactured homes purchased were double-wides or 
larger, compared with 51 percent in 1996 (see fig. 8). However, FHA 
data shows 82 percent of the loans originated through FHA's Title I 
Manufactured Home Loan program for fiscal years 2005 and 2006 were for 
the purchase of single-wide homes. Officials we interviewed attributed 
this trend to FHA loan limits that were too low to enable borrowers to 
purchase larger, multisection homes using guaranteed loans. 

Figure 8: Number of Manufactured Homes Purchased and Placed, by Size, 
1996 and 2005: 

[See PDF for image] 

Source: GAO analysis of Census MHS data. 

[End of figure] 

Manufactured homes were more likely to be placed on private property. 
From 1996 to 2005, more new manufactured homes were placed on private 
property than in manufactured home parks, even though placements 
overall (in both parks and private property) decreased since 1996. 
According to data from the Manufactured Housing Survey, in 1996, 
229,790 new manufactured homes were placed on private property compared 
with 88,420 homes placed inside a manufactured home park. In 2005, 
80,757 manufactured homes were placed on private property and 28,850 
were placed inside a manufactured home park (see fig. 9). Because FHA 
does not collect placement data, it is unclear where manufactured homes 
purchased with FHA Title I loans were located--on owned or leased land. 
However, FHA officials told us that, based on their review of lender 
insurance claims, most of the Title I loans are for manufactured homes 
on leased land. 

Figure 9: New Manufactured Home Placements, Owned versus Leased Land, 
1996-2005: 

[See PDF for image] 

Source: GAO analysis of Census MHS data. 

[End of figure] 

Similarly, officials we interviewed from five states reported that more 
placements were occurring on private property than in manufactured home 
parks. The officials cited a variety of reasons why new manufactured 
homes were more likely to be placed on private property. First, the 
lack of financing available for manufactured homes to be placed on 
leased land decreased the likelihood of units being placed in a 
manufactured home park. For example, one official stated that the lack 
of manufactured home financing resulted in more manufactured homes 
being placed on private land because of the increased availability of 
financing for homes that are considered real property. Second, the 
increase in the size of manufactured homes to double-wides or 
multisection could prevent the homes from fitting into park spaces 
designed for smaller units. Third, both industry and consumer officials 
suggested that the quality and style of new manufactured homes had 
improved, allowing them to blend in with other site-built homes on 
private property. Developers have created affordable housing 
opportunities by using manufactured homes in infill lots located in 
urban areas or subdivisions. For example, in Seattle, a community 
development corporation used manufactured homes to create affordable 
single-family and town homes in a development called Noji Gardens. In 
Kentucky, Frontier Housing, an affordable non-profit housing developer, 
built affordable housing communities using a combination of 
manufactured, modular, and site-built homes (see fig. 10). 

Figure 10: Examples of Affordable Manufactured Homes in New 
Subdivisions That Look Similar to Site-Built Homes: 

[See PDF for image] 

Source: Frontier Housing and Homesite. 

[End of figure] 

Variability of State Requirements Makes It Difficult to Use State Data 
to Determine the Number of Manufactured Home Parks: 

Data were not available on the number of manufactured home parks 
because states define and license them differently (see fig. 
11).[Footnote 17] For example, the definition of a manufactured home 
park in New Hampshire is a parcel of land that accommodates two or more 
homes; however, in Florida, certain provisions apply to manufactured 
home parks with 10 or more homes. Moreover, most states do not require 
manufactured home parks to be licensed; this is typically done at the 
local level. As a result, data on the number of manufactured home parks 
in each state and at the national level are limited. Anecdotally, 
several officials we interviewed suggested that the creation of new 
manufactured parks was uncommon, with few parks being developed since 
the early 1980s. The officials suggested that local zoning restrictions 
prevented manufactured home parks from being built and that localities 
often preferred to promote other land use options to attract 
development with greater potential to raise the tax base. Officials 
from most of the states that we reviewed told us that most manufactured 
home park closings were caused by rising land prices and subsequent 
pressure to redevelop the site. 

Figure 11: Definition of How Many Homes Constitute a Manufactured Home 
Park and Presence of Licensing Requirements, for the States We 
Reviewed, 2007: 

[See PDF for image] 

Source: GAO analysis of select state statues or landlord/tenant law. 

[A] In Florida, in a park with 26 or more lots, park owners are 
required to file a prospectus, which includes the park bylaws and other 
information, with the state for its approval. 

[B] Georgia and Missouri do not have provisions in place that define 
what number of homes or spaces constitute a manufactured home park. 

[End of figure] 

Although anecdotal data indicate a number of manufactured home parks 
have closed, the extent to which closures have occurred is unknown. 
Through a database search of national and local newspapers, we found 
closures had occurred in 18 states between May 2005 and May 2007. In 
some cases, other types of housing (such as condominiums, town homes, 
and single-family homes) were built on the former park sites, while in 
other cases the parks were converted to commercial use. A few parks 
also were converted from investor-owned parks to resident-owned parks. 
In some instances, local municipalities tried to curb the number of 
closures by placing a moratorium on park owners selling to developers. 

Costs of Living in a Manufactured Home Were Lower Than for Other Home 
Types, but High Costs of Moving and Few Placement Options Limit 
Mobility of Homes: 

Manufactured homes can be more affordable than other housing types. 
According to 2005 American Housing Survey data, monthly housing costs 
for manufactured homes generally were lower than for site-built homes 
(see fig. 12).[Footnote 18] More than half of the owners of 
manufactured homes (54.7 percent) had monthly housing costs from $100 
to $499. In comparison, a little more than a quarter (27.4 percent) of 
owners of site-built homes had monthly housing costs from $100 to $499. 

Figure 12: Monthly Housing Costs of Owners of Manufactured and Site- 
Built Homes, 2005: 

[See PDF for image] 

Source: GAO analysis of AHS 2005 data. 

[End of figure] 

The costs of moving manufactured homes can be high, and, according to 
state, industry, and consumer officials we interviewed, the cost- 
prohibitive nature of moving manufactured homes was one reason why 
owners moved them infrequently. Officials explained the price could 
range from $3,000 to $25,000. According to officials, a variety of 
factors influence moving costs, including distance of the move and the 
size of the home. In addition, moves involve set-up and dismantling 
costs, such as utility and other work to prepare the land. Several 
officials suggested that homeowners, particularly those on fixed 
incomes, could not afford the cost of moving their manufactured homes 
because they did not have the financial means to do so. As discussed 
later, in cases of park closures, some states have a relocation fund 
and, sometimes, property owners or developers might provide some funds 
for displaced residents to move their manufactured homes, assuming the 
displaced residents can find a place to move. 

Owners of Manufactured Homes Have More Consumer Protections If Homes 
Are Considered Real Rather Than Personal Property, and Protections 
Provided by States Vary: 

Borrowers with loans for real property are generally entitled to a 
broader set of protections under a federal law governing the loan 
settlement process than borrowers with personal property loans. For 
instance, borrowers taking out loans for real property receive uniform 
settlement statements, as well as escrow statements. Additionally, 
although state law for situations of foreclosure (real property) and 
repossession (personal property) varies, consumer protections for 
foreclosure are generally broader than for repossession. Finally, 
tenant protections--involving issues such as the length of leases for 
land, requirements for notice and frequency of rent increases, notice 
of eviction, and park closures--vary across the eight states we 
reviewed, as did state aid for displaced residents of parks that 
closed. 

Fewer Federal Protections Apply to Loan and Settlement of Personal 
Property Than Real Property: 

Generally, borrowers with personal property loans are entitled to fewer 
consumer protections under federal laws than borrowers with real 
property loans. Under the Truth in Lending Act (TILA), borrowers 
(including Title I borrowers) who purchase homes using personal 
property loans receive certain disclosures. For instance, creditors 
generally are required to provide the amount financed; the finance 
charge and the finance charge expressed as an annual percentage rate; 
the number, amount, and due dates or periods of payments; and the 
provisions for new payment, late payment, or prepayment. The 
disclosures are intended to make borrowers aware of the cost of the 
loan and policies for paying the loan, so that lenders cannot charge 
arbitrary rates or implement policies that are not disclosed to the 
borrower. 

Borrowers who take out loans for the purchase of real property are 
entitled to additional protections under the Real Estate Settlement 
Procedures Act (RESPA), which is intended to ensure that consumers 
receive information on the nature and costs of the real estate 
settlement process and are protected from unnecessarily high settlement 
charges caused by certain abusive practices. RESPA also protects Title 
I borrowers or other buyers of manufactured homes if their federally 
related mortgage loans are secured by land on which a manufactured home 
sits or on which a manufactured home will be placed within 2 years. 
Borrowers are entitled to receive a good faith estimate of settlement 
costs within 3 days of submitting a loan application. At settlement, 
RESPA requires a uniform settlement statement that shows all charges in 
connection with the settlement both before and at the time of the 
settlement. RESPA also requires an initial escrow statement that 
itemizes the estimated taxes, insurance premiums, and other charges 
expected to be paid from the escrow account in the first year.[Footnote 
19] RESPA generally prohibits kickbacks and unearned fees for 
settlement services and charges for the preparation of certain 
documents. Additional disclosure requirements--an annual escrow 
statement that summarizes deposits and payments and a servicing 
transfer statement if the loan is transferred to a different lender-- 
apply after the loan is settled. 

States Generally Provide More Protections for Borrowers in Foreclosure 
Than Repossession; However, Federally Insured Borrowers Are Entitled to 
Additional Protections: 

State law generally provides more consumer protections in connection 
with foreclosures of real property than in connection with 
repossessions of personal property; however, borrowers in certain 
federally insured loan programs receive additional protections. 
Depending on state law and the mortgage contract, the two most common 
methods of foreclosure are judicial foreclosure and nonjudicial 
foreclosure by power of sale.[Footnote 20] The level of protections to 
the homeowner in case of a foreclosure varies by state. 

All states let the homeowner redeem the mortgage by paying off the 
total outstanding debt before the sale. However, only some states let a 
homeowner cure a default by paying the installments due and the costs 
to reinstate the loan prior to the resale of the home. Some states may 
allow the homeowner time to redeem the property from the purchaser, 
which often is the lender, after the foreclosure sale by paying the 
purchase price for the home, plus related costs and interest. For 
example, in North Carolina, a homeowner has 10 days to redeem the 
property after the foreclosure sale. 

Personal property loans typically are subject to repossession rather 
than foreclosure. As with real property, the procedures can be judicial 
or nonjudicial. Generally, creditors use judicial action procedures to 
repossess manufactured homes. The Uniform Commercial Code, a model code 
adopted by states in various forms, also authorizes a secured party, 
upon default, to take possession of the collateral without judicial 
process--self-help repossession--if that can be done without breach of 
the peace. Because it may be difficult to avoid breaching the peace 
when repossessing manufactured homes, this process is not likely to be 
used often with manufactured homes.[Footnote 21] Time frame and notice 
requirements for repossession can be less stringent than the 
corresponding requirements for foreclosures. For example, the Uniform 
Commercial Code does not prevent a creditor from immediately 
accelerating the note and repossessing the collateral; however, some 
states do impose restrictions on acceleration of repossession. Of the 
eight states we reviewed, five have provisions in place that permit 
acceleration of repossession in certain transactions only when the 
borrower is in default or in breach of the agreement or when contract 
terms permit it under certain conditions. In addition, some state 
statutes provide a right to cure a default prior to the acceleration or 
repossession of a manufactured home and for other consumer transactions 
in certain cases. 

However, Title I Manufactured Home Loan borrowers are entitled to 
additional protections under FHA regulations. For instance, lenders may 
not begin the process of repossession or foreclose on a property 
securing a Title I loan in default unless the property has been 
serviced in a timely manner and with diligence and reasonable and 
prudent measures have been taken to get the borrower to bring the loan 
account current.[Footnote 22] Title I borrowers, like borrowers in 
certain other federally-insured loan programs, are entitled to receive 
written notice of their default. For Title I borrowers, this notice 
includes a description of the lender's security interest, a statement 
of the nature of the default and the amount due, a demand upon the 
borrower to either cure the default or agree to a modification 
agreement or a repayment plan, and a statement that if the borrower 
fails to either cure the default or agree to a modification or a 
repayment plan within 30 days of the notice, the maturity of the loan 
is accelerated and full payment is required. Further, for federal home 
loans that HUD, the Department of Veterans Affairs, and Rural Housing 
Service guarantee, a lender cannot start foreclosure proceedings for a 
default in payment until at least three full monthly installments are 
past due. 

States Give Manufactured Homeowners Varying Levels of Notice, 
Protection and Compensation Related to Length of Leases on Land, Rent 
Increases, Evictions, and Park Closures: 

Tenant protection issues affecting owners of manufactured homes include 
the length of the leases for land, rent increases, requirements for 
eviction, and park closures. We analyzed state laws in eight states and 
found varying written lease requirements (see fig. 13). For instance, 
five of eight states have provisions for written lease requirements. 
The terms range from any amount of time agreed upon by the landlord and 
tenant to a minimum of 2 years. However, officials in some states with 
whom we spoke suggested that enforcing this requirement was difficult. 
Notice of rent increases range from 60 to 90 days; however, some states 
do not have notice requirements on rent increases, such as Georgia, 
Missouri, North Carolina, and Texas.[Footnote 23] States also qualify 
the rent increase provisions in varied ways. Arizona provides that 
rents generally can only increase upon renewal or expiration of the 
lease and the owner has to give 90 days notice. New Hampshire requires 
60 days notice to raise rents but is silent on the number of times the 
rent can increase in a given year. Industry and consumer officials 
suggested the lack of ability to control monthly payments created 
additional risk for both lenders and borrowers. 

Figure 13: Requirements for Written Leases and Notices of Rent 
Increases in the States We Reviewed, 2007: 

[See PDF for image] 

Source: GAO analysis of select state statutes on a landlord/tenant law. 

[A] Arizona law provides the landlord is entitled to a rent increase 
effective at the expiration or renewal of a lease. However, it also 
states that the landlord can increase the rent immediately to account 
for the actual costs of certain expenses if the written agreement so 
provides. 

[B] Missouri and North Carolina do not have requirements for minimum 
lease terms or rent increases. 

[C] New Hampshire law does not use the term "written rental agreement," 
but requires that landlords disclose in writing all terms and 
conditions of the tenancy including rental, utility, and service 
charges, prior to entering into a rental agreement with a prospective 
tenant. 

[End of figure] 

Unlike owners of site-built homes, owners of manufactured homes living 
on leased land can be subject to eviction for nonpayment of rent or 
noncompliance with terms in lease agreements. Additionally, nonpayment 
of rent can be a signal that the homeowner is behind on loan payments 
as well. All states that we reviewed require good cause for eviction; 
however, the amount of time that the affected party has to cure the 
cause for the eviction (that is, to bring the late rent payments 
current) ranges from 7 to 30 days from receipt of notice.[Footnote 24] 
Failure to cure an eviction for an owner of manufactured home on leased 
land could require the homeowner to move from the manufactured home 
park. However, as mentioned earlier, such a move may be cost- 
prohibitive. 

Homeowners also can be forced to move because parks close. Notice 
requirements for those residents that had to move for this reason vary 
from 120 to 545 days in the states we reviewed (see fig. 14). But the 
states we reviewed also have a range of tools to aid the displaced 
owner of a manufactured home, such as offering the park residents the 
right of first refusal (the first opportunity to bid on the purchase of 
the park) and also offering relocation funds or tax credits for 
displaced residents. For example, one of the eight states we reviewed 
offers residents the right of first refusal. However, although Arizona, 
New Hampshire, and Oregon do not have a right of first refusal law, the 
states do have laws that provide notice of the park sale and time in 
which to prepare a bid.[Footnote 25] In New Hampshire, state law 
requires both the park tenants and state financing agency receive 
notice when a manufactured home park is sold. The New Hampshire 
Community Loan Fund then works with the park tenants to form a 
nonprofit cooperative in which the tenants would own both the land and 
their homes.[Footnote 26] Three states we reviewed have a relocation 
fund or tax credit for displaced residents (Arizona, Florida, and 
Oregon). Some interviewees suggested that in some park closures, 
especially those with a lot of publicity, the developer or buyer of the 
land would partially compensate the displaced residents. 

Figure 14: State Provisions for Displaced Occupants of Manufactured 
Homes in the States We Reviewed, 2007: 

[See PDF for image] 

Source: GAO analysis of select state statutes on landlord/tenant law. 

[A] Georgia has no provision for such a notice. 

[B] In Oregon, the notice requirement can be 180 days if the landlord 
finds acceptable space for the tenant to move and pays moving expenses, 
up to $3,500. 

[End of figure] 

Although a few states offer relocation funds for displaced manufactured 
home residents, officials from all states we reviewed cited potential 
barriers in finding a place to relocate the homes, such as a lack of 
vacancies in nearby parks, age requirements that park owners or 
municipalities place on units, and costs associated with moving and 
relocating homes. For instance, many parks will not allow homes built 
before 1976, and localities in some states may have laws prohibiting 
placement of homes that are more than 5 or 10 years old. Further, in 
states such as Florida, wind zone requirements for certain areas may 
prevent the relocation of a home not rated (certified) to withstand 
winds of certain speeds.[Footnote 27] In addition to costs, officials 
also cited potential damage to the home as a barrier to movement. 

More Information Is Needed to Determine the Impact of Proposed Changes 
to the Title I Program: 

Legislative proposals to change the Title I program would increase loan 
limits; insure each loan made; incorporate stricter underwriting 
requirements; and establish up-front premiums and adjust annual 
premiums; but the potential effects of the changes on the program and 
the insurance fund are unclear. According to some FHA and industry 
officials, the potential benefits for borrowers include larger loans 
with lower interest rates to buy larger homes. Also, increased access 
to financing for borrowers could occur since more lenders would be more 
likely to participate in the program because individual loans could be 
insured. Industry officials also identified several factors unique to 
manufactured home lending, such as the decreased ability of borrowers 
to build equity, the location of the home (owned or leased land), and 
the cost of recovery to the lender after defaults that can increase the 
risks of manufactured home lending. To illustrate the effects of the 
proposed changes, we developed an approach that used variations of 
unique risk factors associated with manufactured home lending, as well 
as commonly used predictors of loan performance, such as credit scores, 
to illustrate default scenarios. Our analysis suggests that loans for 
homes on leased land and to borrowers with poor credit have greater 
risk of default. And, in all instances where borrowers had medium or 
high default risk, we show the fund experiencing a loss. However, FHA 
has not yet assessed risks associated with the proposals or detailed 
changes to its underwriting requirements. The agency also has not yet 
collected data needed to help assess risks such as credit scores and 
land type. FHA officials explained that it had not done so because the 
Title I program was low-volume and because they were unsure if the 
legislation would pass. FHA officials said that they chose to devote 
their resources to changing the much larger Title II program. As a 
result, the effects of the proposed changes to the Title I program are 
unclear. 

Proposals to Change the Title I Program Would Increase Loan Limits, 
Insure Each Loan Made, Incorporate Stricter Underwriting Requirements, 
and Set Premiums: 

Several bills introduced in Congress from 2005 to 2007 detailed 
proposed changes to the Title I Manufactured Home Loan program, but the 
majority of the bills contained similar provisions.[Footnote 28] For 
example, all would increase the loan limits of the program and index 
them annually. In the latest bill that passed the House in May 2007, 
the loan amount for a home-only loan would increase from $48,600 to 
$69,678. For the land-only loan, the loan limit would increase from 
$16,200 to $23,226 and for combined home and land loans, the loan limit 
would increase from $64,800 to $92,904.[Footnote 29] 

All but one of the bills would require a change to the mechanism that 
FHA uses to insure against its insurance risk.[Footnote 30] Currently, 
FHA accounts for its insurance risk by insuring only a portion (10 
percent) of a lender's Title I Manufactured Loan portfolio. For 
example, if a lender's portfolio in a given year totaled $1,000,000, 
FHA's guarantee to the lender would not exceed $100,000. The proposed 
legislation removes the portfolio cap and insures each loan on an 
individual basis. However, the current risk-sharing mechanism on 
individual loans between FHA and lenders (where FHA covers 90 percent 
of the loss if there is a claim on a defaulted loan and the lender 
absorbs the remaining 10 percent) would not change. 

Moreover, FHA would be required to establish specific underwriting 
criteria to ensure the financial soundness of the program within 6 
months of the passage of the legislation. Currently, Title I 
regulations require a lender to exercise prudence and diligence in 
underwriting a loan to determine whether the borrower is an acceptable 
credit risk, such as requiring lenders to conduct a credit 
investigation and obtain a credit report. But the Title I regulations 
do not contain provisions that would address other factors specific to 
manufactured homes, such as whether the home is placed on owned or 
leased land. For Title I, FHA reviews the lender's underwriting only 
when a default occurs within the first 2 years of the loan and a lender 
submits a claim for insurance. FHA then has 2 years to deny a claim for 
insurance even after FHA has certified the claim for payment. The 
proposed legislation also would require FHA to provide incontestable 
insurance endorsements, meaning that no claim could be denied because 
of underwriting issues--absent fraud or misrepresentation.[Footnote 31] 

All but one of the bills would establish up-front mortgage insurance 
premiums, not to exceed 2.25 percent of the loan amount and adjust the 
annual insurance premiums up to 1 percent of the remaining unpaid 
principal balance, rather than the original loan amount as stipulated 
in current law.[Footnote 32] The remaining bill would give the agency 
flexibility to establish premiums through risk-based pricing. If such a 
provision were made law, FHA officials told us that they would provide 
a range of premiums based on historical analysis of FHA loan data. 

Furthermore, all but one of the bills would require operations at 
negative subsidies--that is, without cost to the government. Currently, 
the Title I program is operated at a positive subsidy-, meaning that 
the present value of estimated cash outflows (such as claims) to FHA's 
General Insurance Fund exceed the present value of the estimated cash 
inflows (such as borrower premiums). According to the Federal Credit 
Supplement, FHA's Title I Manufactured Home Loan program is expected to 
require a $487,000 subsidy in fiscal year 2007 and a $76,000 subsidy in 
2008. FHA officials state it is unlikely the program could generate 
negative subsidies because of the proposed premium structure and 
potential for depreciation of the assets underlying the loans (the 
manufactured homes). 

Some of the bills also would require that the claim and disposition of 
property for the Title I program be similar to the Title II program, 
where FHA disposes of the used homes once the lender receives insurance 
benefits. FHA opposes this change to the bill and proposes to continue 
having the lenders dispose of the property. As discussed later, 
recovery cost for manufactured housing are higher than for other types 
of housing and lenders require strong recovery practices, such as a 
network for selling homes in place, to recoup more than half the loan 
balance after a default. 

According to FHA and Industry Officials, Potential Benefits of Proposed 
Changes Include Increasing the Number of Borrowers, More Lender 
Participation, and Expansion of the Secondary Market: 

FHA and lending industry officials with whom we spoke cited benefits 
that could accrue to borrowers, the industry, and the Title I program 
if the proposed legislation were enacted. These officials suggested 
that increasing the loan limits would allow more borrowers to buy 
manufactured homes at lower interest rates and also larger homes. As 
noted earlier, in recent years buyers have expressed a heavy preference 
for purchasing double-wide or multisection units. 

FHA, Ginnie Mae, and lending industry officials also suggested that 
increasing the limits and eliminating the portfolio cap would increase 
lender participation and demand for Title I loans, which in turn could 
increase competition and decrease borrower interest rates. In 
particular, Ginnie Mae officials stressed that eliminating the 
portfolio cap would be central to their decision to expand their 
participation in the secondary market for manufactured home 
loans.[Footnote 33] This, in turn, could provide more liquidity to 
lenders and greater access to credit for borrowers. Ginnie Mae was the 
main guarantor of securities backed by FHA Title I loans on the 
secondary market up until 1989 when Ginnie Mae placed a moratorium on 
new manufactured housing issuers because of the high risks associated 
with the product. Currently, Ginnie Mae has four lenders in its 
manufactured home program with just one active. According to Ginnie Mae 
officials, it imposed the moratorium because structural features of the 
Title I program, such as the portfolio cap and the nonspecific 
underwriting requirements, exposed Ginnie Mae to greater risk and 
losses. According to Ginnie Mae, once claim amounts were reached on 
troubled portfolios, lenders had little incentive to continue servicing 
the portfolios and make payments to security holders. Ginnie Mae then 
sustained substantial losses when it assumed the portfolios of lenders 
that reached FHA coverage limits. In addition, one lending official 
suggested more stringent underwriting requirements would be beneficial 
to the industry, which still is recovering from the defaults and 
repossessions of the early 2000s.[Footnote 34] Industry officials 
suggested that federal agencies, such as FHA, and the government- 
sponsored enterprises could help facilitate changes in the industry, 
such as improving underwriting requirements. 

However, according to FHA and the Congressional Budget Office (CBO), 
the elimination of the portfolio cap could increase significantly the 
amount of claims paid and expand the government's liability under the 
program since each loan would be insured on an individual basis. FHA 
officials also said that they believed risk-based pricing would help 
compensate FHA's insurance risk. The extent to which risk reduction 
would occur and what borrowers would be excluded would depend on 
underwriting requirements, such as the ranges of credit scores allowed. 

Lending Officials Identified Factors Unique to Manufactured Home 
Lending That Can Affect Loan Performance: 

Industry officials identified several risk factors unique to 
manufactured home lending, such as the decreased ability of borrowers 
to build equity, the lack of consistency and transparency in appraising 
and pricing homes, the location of the home (owned or leased land), the 
cost of recovery to the lender after defaults, and issues related to 
the installation of the home. Based on our review of literature and 
interviews with lending industry officials, owners of manufactured 
homes generally have less ability to build equity than the owners of 
site-built homes. As assets, manufactured home can depreciate in value 
after purchase, similar to automobiles. For example, officials 
explained that manufactured homes bought with personal property loans 
generally depreciated in value if not attached to land. The officials 
emphasized that, even after years of making payments, a borrower could 
choose to default on a loan if the home was worth less than the loan 
balance. 

In general, manufactured homes are appraised differently when 
considered real property compared to personal property. When a home is 
placed on real property, the value of the home is determined based on 
comparable homes in the vicinity. When a home is considered personal 
property, the value is based on the price that the manufactured home 
dealer had determined for the unit. However, lending officials with 
whom we spoke suggested prices varied by dealers and that pricing of 
manufactured homes was not transparent because dealers are not mandated 
to display a manufacturer's suggested retail price. In addition, states 
conduct little or no recording of sales data. Further, the officials 
suggested that the lack of transparency resulted in some consumers 
overpaying for the manufactured home, particularly in instances where 
the dealer would present the price in monthly payment terms. One lender 
identified California as a model state, because it requires all 
manufactured home purchases to go through escrow, whether real or 
personal property, which helps to monitor sales prices.[Footnote 35] 
This lender's loans performed significantly better in California than 
in other states, and the lender suggested that California's transparent 
pricing was one of the main reasons. 

Furthermore, industry officials suggested that the location of the home 
on owned or leased land is a predictor of loan performance. According 
to our review of literature and interviews with industry officials, 
loans for manufactured homes placed on owned land (titled as real 
property) tend to perform better than loans for homes on leased land 
(titled as personal property), as they tend to appreciate more. Some 
officials suggested appreciation could occur on leased land, but that 
appreciation would be dependant on location and amenities available 
(such as pools, club houses, or golf courses). 

In contrast to other housing types, many lending industry officials 
suggested that the cost of recovery for lenders when a loan defaulted 
was greater with manufactured homes. For instance, for manufactured 
homes the costs to the lender in a foreclosure or repossession (which 
may involve the movement of the home) would be proportionately higher 
relative to the loan amount than for more expensive site-built housing. 
Some states have lien holder statutes in place that may help the lender 
protect its collateral in cases of borrower default by requiring 
notification of lenders in case of abandonment or eviction. Of the 
states we reviewed, Arizona, New Hampshire, Oregon, and Texas have such 
a statute.[Footnote 36] Some lending industry officials suggested that 
losses and high recovery costs could be mitigated by selling the home 
in place. They suggested that lease agreements between lenders and 
community owners should ensure that manufactured homes located on 
leased land could be sold in place if borrowers defaulted. 

According to some of lending officials we interviewed, the size of the 
home also was a predictor of performance. Loans for larger manufactured 
homes (double-wides or multisection units) tend to perform better than 
loans for single-wides. The officials with whom we spoke suggested that 
these loans performed better because the income level of those 
borrowers tended to be higher. However, the majority of Title I loans 
have been for single-wides, which according to FHA and industry 
officials was because of the current loan limits. 

In addition, many industry officials suggested the type and quality of 
installation of the home affects the value of the home and that, in 
theory, states with stronger inspection programs help maintain the 
value of the home for the consumer. The Manufactured Housing 
Improvement Act of 2000 set standards for installation inspections 
across the country, but states continue to differ in how they monitor 
installation of homes.[Footnote 37] Until recently, many states did not 
have a program to inspect the installation of manufactured homes. In 
our review of eight state installation programs, we found the level of 
inspections varied by state (see fig. 15). For example, five of eight 
states require 100 percent inspection (Arizona, Florida, New Hampshire, 
North Carolina, and Oregon). All of those that require 100 percent 
inspection had installation programs in place prior to the 
implementation of the Manufactured Housing Improvement Act, except for 
New Hampshire, whose requirement went into effect in July 
2006.[Footnote 38] The remaining states relied on state officials 
inspecting from at least one manufactured home installer in Georgia or 
from 10 to 35 percent of manufactured homes in Missouri and Texas, but 
Georgia and Missouri made changes to their installation programs after 
the passage of the act. Prior to these changes, the two states 
inspected installations on a consumer complaint basis. Further, state 
programs differ in how they conduct installation inspections. For 
instance, Florida, New Hampshire, and North Carolina rely on local 
jurisdictions to conduct the inspections; Arizona and Oregon use a 
combination of both state and local officials; while Georgia, Missouri, 
and Texas use only state officials. 

Figure 15: Comparison of Installation Inspection Programs in the States 
We Reviewed, 2007: 

[See PDF for image] 

Source: GAO analysis of select state MH installation programs. 

Note: Georgia does not specify a percentage of manufactured homes to be 
inspected. Rather, each installer is inspected once annually. If a 
problem is found with a particular manufactured home installer, the 
state office conducts multiple inspections of the installer. 

[End of figure] 

Our Analysis Suggests That Poor Credit and Homes Placed on Leased Land 
Are Key Factors Associated with Greater Risk of Defaults of Loans for 
Manufactured Homes: 

In the absence of available data on the credit of FHA borrowers and the 
location of the homes (owned or leased land), we developed scenarios 
using assumptions based on various risk factors, such as the default 
risk of borrowers and the ability of lenders to recover losses. In 
addition, we considered the experience of FHA's Title I program since 
1990 and of non-FHA personal property manufactured housing loans. For 
example, from 1990 to 2002, FHA's cumulative defaults expressed as a 
percentage of originated loans, did not drop below 10 percent and have 
exceeded 25 percent in 8 of the 13 years (see fig. 16). However, loans 
from 2003 to 2006 may not be reflective of the default experience 
because they are recent loans and lending industry officials explained 
that the peak default period for these types of loans generally occurs 
from the third to the fifth year. Non-FHA manufactured housing loans 
also had high cumulative losses, typically above 15 percent for loans 
originated between 1997 and 2001, but lower than FHA's cumulative 
losses. 

Figure 16: Number of FHA Title I Loans and Percentage of Loans in 
Default, 1990-2006: 

[See PDF for image] 

Source: GAO analysis of FHA data. 

[End of figure] 

Our scenarios incorporate assumptions based on factors such as annual 
default rates for different yearly intervals, loan interest rates, and 
loan terms.[Footnote 39] Once we established these parameters, we 
included additional factors, such as variations on the lenders' ability 
to recover their losses in cases of default and the borrowers' 
insurance premium schedule (based on the premiums suggested in the 
proposed legislation). Our assumptions about default rates reflect an 
important characteristic of home-only manufactured housing loans. Even 
after years of loan payments, a borrower may not have enough equity in 
the home to avoid a default in the face of adverse financial conditions 
or may choose not to pay off a loan if the home is worth less than the 
loan balance. 

Based on discussions with lending industry officials and our review of 
available manufactured home lending data, we assumed three variations 
of default: a low default experience, a moderate default experience, 
and a high default experience. In general, the low default experience 
would reflect conditions in which borrowers possessed good credit 
quality (credit score), lenders used high-quality underwriting 
requirements, and lenders' security interests in the collateral were 
well protected in terms of those factors that are associated with the 
preservation of value, such as placement of the home (owned versus 
leased land) and installation. The high default experience would 
reflect conditions in which borrowers had poorer credit quality, and 
collateral values and lenders' security interests also were lower. 

The following assumptions in our analysis were based on discussions 
with lending industry officials on possible recovery outcomes and 
possible legislative changes regarding FHA's upfront and annual 
premiums: 

* If lenders had a strong recovery program (which may include a good 
network of dealers who resell manufactured homes) they would achieve a 
net recovery of 50 percent per claim. Alternatively, we assumed a 
moderate recovery rate would be 33 percent of the claim, and a low 
recovery rate would be 25 percent. 

* Two different potential up-front premium amounts--a high up-front 
premium of 2.25 percent of the original loan amount and a low up-front 
premium of 1 percent of the original loan amount. 

* Two different annual premiums--a high annual premium of 1 percent of 
the declining loan balance and a low annual premium of 0.5 percent of 
the declining loan balance. 

To determine the potential impact on FHA's General Insurance Fund, we 
used the above assumptions to calculate the relationship between the 
amount and timing of both expected claims and premiums to FHA. Similar 
to a subsidy calculation, we estimated the present value of estimated 
cash outflows (such as claims) net of the present value of the 
estimated cash inflows (such as premiums) to FHA's General Insurance 
Fund. 

The results of our analysis show that in all instances where borrowers 
had moderate or high default risk, the fund experienced a loss--that 
is, the present value of estimated cash outflows exceeded the present 
value of cash inflows (see fig. 17). The range of the loss was 
determined by the lender's ability to recover its losses and the 
premiums the borrower paid. For instance, in cases where the borrower 
paid high up-front and annual premiums (2.25 percent and 1 percent, 
respectively) and was a moderate default risk, and the lender had a 
high net recovery rate (50 percent), the loss to the fund was less than 
1 percent. However, if we varied the scenario to lower the recovery 
rate for lenders (25 percent), the potential loss to the fund was 4.4 
percent. Similarly, when the borrower paid low up-front and annual 
premiums (1 percent and 0.5 percent) and had moderate default risk, the 
losses ranged from 4.4 percent if the lender had a high net recovery 
rate to 8.5 percent if the lender had a low recovery rate. The fund had 
the potential to experience gains in instances where the borrower had 
low default risk, premiums were higher, and lenders had a higher 
probability of high recovery of losses. 

Figure 17: Results of Scenarios Based on Variations in Default Risk, 
Lender Recovery Costs, and Premiums Paid by Borrowers: 

[See PDF for image] 

Source: GAO. 

[End of figure] 

Our analysis also showed that there is potential for FHA's General 
Insurance Fund to experience a wide variation in the level of losses 
but little potential for gains. The results suggested that there is 
greater risk of loss from borrowers who have either moderate or high 
default risk. Typically, these are loans where the borrower may not 
have a high credit score, and the property is located on leased land-- 
in which case the lender's security interest may be uncertain because 
of the variability associated with rent increases, lease terms, and the 
potential for the manufactured home park to be sold. In addition, the 
amount of the loss was influenced by the amount of premiums paid. For 
instance, where borrowers paid the highest up-front and annual premium, 
the loss was 11 percent in cases where the borrower also had high 
default risk and the lender had low recovery, compared with a 15 
percent loss in instances where the borrower paid a low up-front and 
annual premium. However, since FHA does not currently collect data on 
credit score or where the property is located--owned or leased land--it 
is unclear how these scenarios may actually affect the General 
Insurance Fund. See appendix II for a more detailed discussion of our 
scenario analysis methodology. 

Effects of the Proposed Changes Are Unclear Because FHA Has Not Yet 
Articulated Which Borrowers Would Be Served, Assessed Its Insurance 
Risks, or Developed Underwriting Requirements: 

FHA has not yet assessed the effect of the proposed changes to the 
Title I program. More specifically, it has not developed criteria or 
models to assess the potential effects of the proposed premiums or risk-
based pricing or developed specific underwriting requirements. Such 
assessments and requirements are central to effective operation and 
oversight of a revised Title I Manufactured Home Loan program. Our 
internal control standards for federal agencies state that effective 
management involves comprehensively identifying risk as part of short- 
and long-term planning.[Footnote 40] Such planning would encompass the 
identification of risks posed by new legislation or regulations. The 
results of our scenario analysis also suggest that FHA could use 
modeling to illustrate, in a general way, potential gains and losses to 
FHA's General Insurance Fund and that premium structures play a key 
role in determining these outcomes. 

Although the purpose of the Title I program is to serve low-to moderate-
income families, it is unclear which borrowers a revised program would 
serve because FHA has not yet shared the specifics as to how it plans 
to compensate for risk, including how the premiums would be set. 
According to FHA officials, they do not plan to develop criteria for 
assessments of proposed premiums or risk-based pricing until the 
program is approved by Congress. FHA officials told us that they have 
begun to analyze a range of up-front premiums and a maximum premium 
amount based on a historical analysis of receipts and claims, but that 
they had not yet reached any conclusions. In addition, FHA has not 
conducted an analysis to determine under what conditions the program 
could operate at a negative subsidy if the proposed changes were 
enacted. As mentioned earlier, FHA's Title I Manufactured Home Loan 
program is expected to require a $487,000 subsidy in fiscal year 2007 
and a $76,000 subsidy in 2008.[Footnote 41] According to HUD officials, 
they expect to calculate the new subsidy rate based on projected 
defaults, interest and fees, and loan characteristics (such as loan 
maturity, default and recovery rates, and up-front and annual fees) for 
the 2009 budget. CBO estimated that, if the legislation were enacted, 
FHA could achieve a near zero subsidy for the Title I program assuming 
default rates would be at 9.5 percent or lower. CBO also acknowledged 
that because of the uniqueness of FHA's program and lack of comparative 
programs in the private market, the potential costs of the program are 
uncertain. The results of our analysis suggest that, in almost all 
situations, there is potential for loss except when borrowers have 
lower default risk (based on credit scores and other information). 
While credit score is one of the key factors used to determine default 
risk, FHA does not collect this information (discussed further below). 

FHA officials also stated they have not yet developed specific 
underwriting requirements for a revised program. Although industry and 
FHA officials with whom we spoke discussed the unique risks for 
manufactured home loans, the information FHA provided us about any 
changes to its underwriting criteria have not addressed the specific 
characteristics of manufactured housing. FHA officials did explain they 
would like to establish review procedures when a loan is submitted for 
insurance, similar to the procedures in FHA's Title II loan program. In 
Title II, FHA conducts post-endorsement reviews of 10 percent of its 
loans, with FHA staff going over the lender's underwriting decisions 
and calculations. 

In explaining the agency's limited assessments, FHA officials noted 
that the agency is focusing its resources on assessing the impact of 
proposed changes to the much larger Title II Mortgage Insurance 
program. As of May 2007, FHA's risk-based pricing proposal for the 
Title II program established six different risk categories, each with a 
different premium rate, for purchase and refinance loans. FHA used data 
from its most recent actuarial review to establish six risk categories 
and corresponding premiums based on the relative performance of loans 
with various combinations of loan-to-value ratio (the ratio of the 
amount of the mortgage loan to the value of the home) and credit 
score.[Footnote 42] Further, since the current volume of the Title I 
program is low, FHA officials did not anticipate large losses for the 
fund. However, the programmatic changes in the proposed legislation are 
designed to increase the demand for Title I manufactured home loans. 
FHA officials told us that once the legislation is passed, it would 
take up to a year to implement changes to the program and to work on 
developing the risk-based pricing strategy; however, they were unsure 
if they would implement the program in stages or all at once. 

As a result of FHA not conducting risk assessments or determining 
underwriting requirements, potential effects of changes to the Title I 
program remain unclear. Without such risk identification, FHA's 
planning may be adversely affected. In particular, the agency may lack 
timely indications of whether the program could generate positive or 
negative subsidies, which in turn would affect decisions about pricing 
premiums. 

FHA Also Has Not Yet Developed an Approach for Collecting the 
Information Needed to Manage the Program: 

Currently, FHA does not collect information on the credit scores of 
borrowers or the type of land on which manufactured homes are placed. 
Our internal control standards for federal agencies state that an 
agency must have relevant, timely, and reliable information to run and 
control its operations. Of the factors identified as risks affecting 
manufactured home lending, FHA maintains data only on the size and 
condition (that is, new or existing) of the manufactured home. In 2004, 
FHA started to collect information on borrower demographics, such as 
gender, address, birth date, and monthly income. And, because FHA 
currently monitors a lender only when a claim is filed for insurance 
and not before the loan is originated, the information collected is not 
as thorough as would be generated if the program required review prior 
to the endorsement of the loan. 

FHA officials told us it would like lenders to electronically capture 
more information about borrowers during the underwriting process, but 
that the current information system for the Title I program would need 
to be updated to accommodate expanded data fields. FHA officials also 
told us that they plan to collect more detailed borrower, property, and 
loan-level data to improve tracking and performance measurement, but 
did not have specific details as of July 2007. However, our interviews 
with lending officials and the results of our scenario analysis both 
suggest that credit score and the location of the home (on owned or 
leased land) are important predictors of loan performance. Without more 
comprehensive data on its borrowers and lenders, FHA may not be able to 
successfully estimate default risks in its portfolio, mitigate risks to 
the insurance fund, and, thus, effectively manage the program. 

Conclusions: 

Manufactured homes are an affordable housing option, but they differ 
from site-built homes in the way they are financed, sold, and the 
consumer protections available. These differences create additional 
risks for both the borrowers and lenders of manufactured homes. For 
example, the ability for the homeowner to build equity is constrained 
if the property is located on leased land and the land ownership also 
affects the ability of the lender to recover its losses relative to 
other types of lending. These risks are reflected in the performance of 
the Title I program, which has a history of high default rates, as does 
the manufactured home lending industry. However, the Title I program 
also provides a unique product as the only active federal program 
offering insurance for home-only (personal property) loans. According 
to recent FHA data, the majority of its borrowers are younger and lower-
income, suggesting that Title I helps them achieve homeownership. 

But changing and expanding a lending program can introduce new risks 
and increase existing risks. FHA only insured slightly more than 1,400 
loans in 2006. Changes to the Title I program are expected to increase 
loan volume, which could generate the desirable outcome of providing 
more lower-priced loans to lower-income individuals desiring to 
purchase a home. Yet, both FHA and CBO suggest proposed changes can 
increase FHA's insurance risk and expand the government's liability. 
The extent of gains or losses to FHA's General Insurance Fund will 
depend on a variety of factors, such as the borrower's default risk, 
the lender's ability to recover losses, and the amount of premiums 
paid. However, FHA has not articulated which borrowers would be served, 
how the loans would be priced under a risk-based structure and the 
expected increase in risk to the General Insurance Fund, how the loans 
would be underwritten, and the additional data it plans to collect to 
manage the program. Thus, the agency lacks vital information for 
implementing any changes to the program. If FHA were to conduct such 
risk identification, it could plan to anticipate changes to the 
program, target new borrower populations, and more effectively manage 
existing loan portfolios. In particular, with indications of whether 
the program could generate positive or negative subsidies, the agency 
could make appropriate and well-informed decisions about pricing 
premiums. For example, an analysis similar to the one we performed 
would provide at least an indication of what scenarios would produce 
the highest risks for losses to the fund. Finally, more comprehensive 
data on its borrowers and lenders could allow FHA to mitigate the risks 
inherent with the manufactured home product. 

Recommendations for Executive Action: 

In light of the growth that a revised Title I program could spur and 
previous experience in the manufactured home loan industry that 
included a high number of defaults and repossessions, prior to the 
implementation of a revised program, we recommend that the Secretary of 
Housing and Urban Development direct the Assistant Secretary for 
Housing and Urban Development--Federal Housing Commissioner to assess 
the effects of the proposed changes. At a minimum, this action should: 

* articulate which borrowers would be served if the program were 
expanded, including the financial conditions and creditworthiness of 
the served borrowers; 

* develop criteria or economic models to assess the potential effect of 
the proposed changes including risk-based pricing; that is, determine 
what circumstances or pricing structures would most likely result in a 
positive or negative subsidy if the proposed changes were enacted; and: 

* develop detailed proposed changes to its underwriting requirements 
that account for unique attributes of manufactured housing and the 
characteristics of FHA's targeted borrower population. 

We also recommend that the Secretary of Housing and Urban Development 
direct the Assistant Secretary for Housing and Urban Development-- 
Federal Housing Commissioner to develop an approach for collecting the 
information needed to manage the program, including the credit scores 
of borrowers and whether the manufactured homes are on owned or leased 
land. 

Agency Comments and Our Evaluation: 

We provided HUD with a draft of this report for review and comment. HUD 
provided comments in a letter from the Assistant Secretary for Housing-
-Federal Housing Commissioner (see app. III). HUD agreed with the 
recommendations in our report and described plans for implementing 
these recommendations. More specifically, HUD agreed with our 
recommendation to assess the effects of the proposed changes prior to 
the implementation of a revised program. FHA noted that it recently 
initiated a review of the credit subsidy calculation for the Title I 
Manufactured Home Loan program and that the results of the study will 
be used to develop models to test underwriting and premium pricing 
options. As we noted in our report, this type of analysis or an 
analysis similar to the one we performed could provide an indication of 
the risks for losses to FHA's General Insurance Fund. 

HUD also agreed with our recommendation to develop an approach for 
collecting the information needed to manage the program. As we 
mentioned in our report, HUD stated it began collecting additional 
data, such as borrower information on age and income in 2004. HUD 
stated that it did not collect information on the location of the homes 
(owned or leased land) because the program requirements for both types 
of homes were essentially the same; however, HUD plans to collect these 
data under a revised program to track loan characteristics. HUD also 
agreed to collect appropriate credit and application variables such as 
credit scores. Finally, the agency noted that it intended procedures 
for originating and underwriting Title I loans to mimic those of FHA's 
real estate financing programs. 

As agreed with your offices, unless you publicly announce the contents 
of this report earlier, we plan no further distribution until 30 days 
from the report date. At that time, we will send copies of this report 
to the Ranking Member, Senate Committee on Banking, Housing, and Urban 
Affairs; Ranking Member, Subcommittee on Housing, Transportation, and 
Community Development, Senate Committee on Banking, Housing, and Urban 
Affairs; Chairman and Ranking Member, House Committee on Financial 
Services; and Chairman and Ranking Member, Subcommittee on Housing and 
Community Opportunity, House Committee on Financial Services. We will 
also send copies to the Secretary of Housing and Urban Development and 
will make copies available to other interested parties upon request. In 
addition, the report will be made available at no charge on the GAO Web 
site at [hyperlink,http://www.gao.gov]. 

If you or your staff have any questions about this report, please 
contact me at (202) 512-8678 or shearw@gao.gov. Contact points for our 
Offices of Congressional Relations and Public Affairs may be found on 
the last page of this report. Key contributors to this report are 
listed in appendix IV. 

Signed by:

William B. Shear: 
Director, Financial Markets and: 
Community Investment: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

The Chairmen of the Senate Committee on Banking, Housing, and Urban 
Affairs and its Subcommittee on Housing, Transportation, and Community 
Development and Senator Jack Reed requested that we evaluate the 
Federal Housing Administration's (FHA) Title I Manufactured Home Loan 
program. Specifically, the objectives of this report were to (1) 
describe selected characteristics of manufactured housing and the 
demographics of the owners, (2) compare federal and state consumer and 
tenant protections for owners of manufactured homes, and (3) describe 
the proposed changes to FHA's Title I Manufactured Home Loan program 
and assess potential benefits and costs to borrowers and the federal 
government. 

In summary, to address our first objective we analyzed Census data from 
the Manufactured Housing Survey from 1996 to 2005 and the 2005 American 
Housing Survey. To address our second objective, we researched relevant 
federal laws and laws in eight states (Arizona, Florida, Georgia, 
Missouri, New Hampshire, North Carolina, Oregon, and Texas) and 
conducted semistructured phone interviews with state, industry, and 
consumer group officials in those eight states.[Footnote 43] We also 
used the information gathered in the interviews to inform our 
discussion in the first and third objectives. For our third objective, 
we interviewed FHA officials and lending officials from programs that 
provide financing for manufactured homes. To learn about risk- 
mitigation practices, we also reviewed policies and procedures from the 
programs mentioned above. Finally, we conducted an analysis using 
different scenarios that incorporated assumptions of risk for 
manufactured housing lending to illustrate potential costs of the 
proposed legislation. We conducted our work in Washington, D.C., 
Atlanta, and Chicago, from October 2006 through June 2007 in accordance 
with generally accepted government auditing standards. 

Selected Characteristics of Manufactured Housing and the Demographics 
of the Owners: 

To determine selected characteristics of manufactured housing, we 
analyzed Census data from the Manufactured Housing Survey. Census 
conducts the Manufactured Housing Survey on a monthly basis and samples 
approximately 350 manufactured home dealers or 1 in 40 of the 
manufacturers that ship manufactured homes each month. The sample of 
manufactured home dealers surveyed fluctuates based on the total number 
of manufactured homes shipped. Specifically, we used Manufactured 
Housing Survey data from 1996 through 2005 to examine trends in the 
manufactured housing industry, such as the number of homes sold, 
average sales price, where the homes were placed (owned or leased 
land), and the size of these homes (single-wide versus double-wide 
units). To determine demographic characteristics of manufactured home 
owners, we relied on the 2005 American Housing Survey. Census conducts 
the American Housing Survey every 2 years, sampling approximately 
55,000 housing units to gather data on apartments; single-family homes; 
manufactured or mobile homes; vacant housing units; age, sex, race and 
income of householders; housing and neighborhood quality; housing 
costs; equipment and fuels; and the size of the housing units. We 
choose to use 2005 American Housing Survey data since they were the 
latest available. We did not provide information on trends in earlier 
years because the sample of manufactured housing used in previous 
surveys (through 2003) changed, making it difficult to compare 2005 
data with previous data. Data on land ownership for manufactured homes 
(that is, owned or leased land) was limited in the Manufactured Housing 
and American Housing Surveys; as a result, we could not report 
differences in the data for where the manufactured home was 
placed.[Footnote 44] We assessed the reliability of the Manufactured 
Housing and American Housing Surveys by reviewing information about the 
data, performing electronic data testing to detect errors in 
completeness and reasonableness, and interviewing knowledgeable 
officials regarding the quality of the data. We determined that the 
data were sufficiently reliable for the purposes of this report. 

Because Census data used in our American Housing Survey analyses are 
estimated based on a probability sample, each estimate is based on just 
one of a large number of samples that could have been drawn. Since each 
sample could have produced different estimates, we express our 
confidence in the precision of our particular sample's results as a 
confidence interval. For example, the estimated percentage of occupied 
manufactured homes located in the South was 56.7 percent, and the 
confidence interval for this estimate ranges from 56.6 percent to 56.8 
percent, with a percentage point error of 0.1 percent. This is the 
interval that would contain the actual population value for 95 percent 
of the samples that could have been drawn. As a result, we are 95 
percent (or more) confident that each of the confidence intervals in 
this report will include the true values in the study population. All 
variables from American Housing Survey that are included in this report 
have 95 percent confidence intervals of plus or minus 5 percentage 
points or less. 

We conducted a literature review and examined relevant studies on 
manufactured housing. We also conducted a review of newspaper articles 
from May 2005 to May 2007 to identify where manufactured home park 
closures occurred in the United States. Because states collect 
different types of information on manufactured home parks and even 
define them differently, the consequent variability of the state data 
makes determining the number of manufactured home parks extremely 
difficult. Thus, we relied on a database search of national and local 
newspapers to provide anecdotal information on park closures. We used 
several different search parameters and keyword searches and identified 
park closures in 18 states; however, it is possible that other closures 
occurred in other states during the period we reviewed, but were not 
identified in our searches.[Footnote 45] 

Federal and State Consumer and Tenant Protections for Owners of 
Manufactured Homes: 

To compare federal and state consumer and tenant protections for owners 
of manufactured homes, we reviewed federal laws relevant to 
manufactured housing, such as the Real Estate Settlement Procedures Act 
and the Truth in Lending Act. We reviewed prior work on state laws for 
manufactured housing conducted by the National Consumer Law Center and 
the American Association of Retired People and also interviewed 
officials from these organizations. We then selected eight states and 
reviewed statutes related to the consumer protections provided for 
foreclosure and repossession and the tenant protections applicable to 
contracts or acts, such as written lease requirements, rent increases, 
evictions, and park closures. 

The eight states were selected based on a combination of factors 
including the volume of FHA Title I loans in the state from 1990 
through the first quarter of 2007; concentration of manufactured 
housing as a percentage of housing units in the state; information from 
our interviews of industry and consumer officials; and previous studies 
conducted on manufactured housing. The table below indicated the 
characteristics of the states we reviewed. 

Table 2: Characteristics of Eight States Selected for Semistructured 
Interviews: 

State: Arizona; 
FHA total loan volume from 1990-2007 (1[ST] quarter): 1.7%; 
Large concentration of manufactured homes in state: 9% to 13%; 
State identified as familiar with manufactured home park issues: Yes; 
State identified as active in manufactured housing policy development: 
No; 
Ownership model of manufactured home park (condo or co-op): [Empty]; 
Officials interviewed cited policies in place for manufactured home 
parks and/or park closures occurring: No; 
Geographic representation: West. 

State: Florida; 
FHA total loan volume from 1990-2007 (1[ST] quarter): 3.3%; 
Large concentration of manufactured homes in state: 9% to 13%; 
State identified as familiar with manufactured home park issues: Yes; 
State identified as active in manufactured housing policy development: 
Yes; 
Ownership model of manufactured home park (condo or co-op): Condo; 
Officials interviewed cited policies in place for manufactured home 
parks and/or park closures occurring: Yes; 
Geographic representation: South. 

State: Georgia; 
FHA total loan volume from 1990-2007 (1[ST] quarter): 4.7%; 
Large concentration of manufactured homes in state: 9% to 13%; 
State identified as familiar with manufactured home park issues: No; 
State identified as active in manufactured housing policy development: 
No; 
Ownership model of manufactured home park (condo or co-op): [Empty]; 
Officials interviewed cited policies in place for manufactured home 
parks and/or park closures occurring: No; 
Geographic representation: South. 

State: Missouri; 
FHA total loan volume from 1990-2007 (1[ST] quarter): 3.6%; 
Large concentration of manufactured homes in state: 6% to 9%; 
State identified as familiar with manufactured home park issues: No; 
State identified as active in manufactured housing policy development: 
No; 
Ownership model of manufactured home park (condo or co-op): [Empty]; 
Officials interviewed cited policies in place for manufactured home 
parks and/or park closures occurring: No; 
Geographic representation: Midwest. 

State: New Hampshire; 
FHA total loan volume from 1990-2007 (1[ST] quarter): < 1%; 
Large concentration of manufactured homes in state: 6% to 9%; 
State identified as familiar with manufactured home park issues: 
No; 
State identified as active in manufactured housing policy development: 
No; 
Ownership model of manufactured home park (condo or co-op): Co-op; 
Officials interviewed cited policies in place for manufactured home 
parks and/or park closures occurring: Yes; 
Geographic representation: Northeast. 

State: North Carolina; 
FHA total loan volume from 1990-2007 (1[ST] quarter): 11.2%; 
Large concentration of manufactured homes in state: 13% - 20%; 
State identified as familiar with manufactured home park issues: No; 
State identified as active in manufactured housing policy development: 
No; 
Ownership model of manufactured home park (condo or co-op): [Empty]; 
Officials interviewed cited policies in place for manufactured home 
parks and/or park closures occurring: Yes; 
Geographic representation: South. 

State: Oregon; 
FHA total loan volume from 1990-2007 (1[ST] quarter): < 1%; 
Large concentration of manufactured homes in state: 9% to 13%; 
State identified as familiar with manufactured home park issues: Yes; 
State identified as active in manufactured housing policy development: 
No; 
Ownership model of manufactured home park (condo or co-op): [Empty] ; 
Officials interviewed cited policies in place for manufactured home 
parks and/or park closures occurring: No; 
Geographic representation: West. 

State: Texas; 
FHA total loan volume from 1990-2007 (1[ST] quarter): 15.8%; 
Large concentration of manufactured homes in state: 6% to 9%; 
State identified as familiar with manufactured home park issues: No; 
State identified as active in manufactured housing policy development: 
No; 
Ownership model of manufactured home park (condo or co-op): [Empty]; 
Officials interviewed cited policies in place for manufactured home 
parks and/or park closures occurring: Yes; 
Geographic representation: South. 

Source: FHA data and GAO. 

Note: Concentration of manufactured homes illustrates the states with a 
high percentage of manufactured homes as part of their housing stock. 

[End of table] 

We also conducted semistructured interviews with regulatory, industry, 
and consumer officials in each state. We pretested our interview 
questions on-site in Georgia and conducted the remaining interviews by 
telephone. We used interview responses on state statutes to check our 
interpretation of the state statutes containing consumer and tenant 
protections applicable to manufactured home owners. In each of the 
states, we interviewed officials who represented (1) the state 
regulator for manufactured housing, (2) the state industry group who 
are affiliates of the National Manufactured Housing Industry group, 
known as the Manufactured Housing Institute, and (3) a consumer 
advocacy group, such as the state manufactured homeowner's 
association.[Footnote 46] In total, we conducted 25 interviews across 
the eight states. 

To synthesize interview data, we compiled the responses by interview 
question into a document for each state (state summary), which we 
reviewed for accuracy and completeness. Next, we identified themes 
among the interviews and created categories within a response, noting 
the state and type of official interviewed. For example, in our 
question on placement options for owners of manufactured homes when the 
park in which they live closes, we identified categories, such as a (1) 
a neighboring park, (2) private land, and (3) lack of space to move. We 
then identified the states that provided a response fitting with each 
category and totaled the number of states in each category. We also 
used this method to compare installation programs across eight states 
based on our interviews with state regulators. 

Proposed Changes to Title I Program and Potential Benefits and Costs to 
Borrowers and the Federal Government: 

To describe the proposed changes to the Title I Manufactured Home Loan 
program, we reviewed current and proposed FHA regulations and 
legislation. Our review of proposed legislation included Senate Bills 
2123 (109th Congress, 2005) and 3535 (109th Congress, 2006) and House 
of Representative Bills 2803 (109th Congress, 2005) and 4804 (109th 
Congress, 2006), House of Representative Bill 2139 and Senate Bill 1741 
from the 110th Congress in 2007. 

To assess the potential costs and benefits of the proposed changes to 
the Title I program, we interviewed FHA officials, FHA lenders, Ginnie 
Mae officials, and officials from federal and other lending programs, 
such as Fannie Mae and Freddie Mac, the U.S. Department of Agriculture 
Rural Housing Service and the Department of Veterans Affairs, community 
banks, industry and consumer groups, and a rating service. In addition, 
we interviewed officials from HUD's Inspector General Office. To learn 
about risk-mitigation practices, we also reviewed policies and 
procedures from programs that provide financing for manufactured homes 
at the above agencies and reviewed relevant literature. A few industry 
officials also provided information on loan performance for their 
manufactured home loan portfolio. 

We also conducted an analysis using different scenarios that 
incorporated assumptions of risk for manufactured housing lending to 
illustrate the potential benefits and costs of the proposed 
legislation. We incorporated various risk factors unique to 
manufactured home lending (such as site location and loss mitigation 
practices of lenders), as well as other commonly used predictors of 
loan performance such as credit scores, into a model to illustrate ways 
in which these key factors might affect the performance of manufactured 
housing loans and, thus, how variation in these key factors might 
affect potential gains and losses to FHA's General Insurance 
Fund.[Footnote 47] Our estimates relied on assumptions concerning a few 
key inputs such as level of default risk, net recovery rate of lenders, 
and insurance premiums. See appendix II for a more detailed description 
of our scenario analysis methodology. 

We also analyzed FHA data, housed in the F-72 database, on the 
manufactured home loan program. We used these data to review loan 
performance from 1990 to 2005, the size of the units purchased, and the 
states in which the loans were originated. We also used the data to 
generate demographic information on FHA Title I borrowers. However, FHA 
only began collecting demographic data in 2004, so our analysis was 
limited to the period from June 2004 through April 2007. In addition, 
we could not assess where the manufactured homes were placed and the 
credit scores of the borrowers because FHA did not collect these data. 
We assessed the reliability of the F-72 database by reviewing 
information about the data, performing electronic data testing to 
detect errors in completeness and reasonableness, and interviewing 
knowledgeable officials regarding the quality of the data. We 
determined that the data were sufficiently reliable for the purposes of 
this report. 

Finally, we reported information provided by HUD using 2005 Home 
Mortgage Disclosure Act data on manufactured housing and the number of 
personal property loans originated by the FHA Title I program compared 
with the rest of the market.[Footnote 48] We also assessed the data 
reliability of this output and the computer program used to extract the 
information and determined the data were sufficiently reliable for our 
purposes. 

[End of section] 

Appendix II: Scenario Analysis Methodology: 

To gain an understanding of the effects of the proposed changes to the 
Federal Housing Administration's (FHA) Title I Manufactured Home Loan 
program, we developed an approach that could illustrate potential 
effects of the changes on the program. Our model of different scenarios 
used assumptions to illustrate the importance of various risk factors 
unique to manufactured home lending (such as site location and loss 
mitigation practices of lenders), as well as other commonly used 
predictors of loan performance, such as credit scores. For instance, 
the ability of the owner of a manufactured home to build equity may be 
limited when the land is leased, which also often increases the risks 
associated with the loan. If a borrower with a home on leased land were 
to default, lenders could face higher costs and lower recoveries 
(relative to site-built homes) in trying to repossess, move, and resell 
the personal property. 

We developed a model to illustrate some of the ways in which these key 
factors may affect the performance of home-only manufactured housing 
loans, and, thus, how variation in these key factors may affect 
potential gains and losses to the FHA's General Insurance Fund, which 
is supported by insurance premiums and used for several FHA insurance 
programs, including the Title I program.[Footnote 49] Based on 
examining some loan performance data from manufactured home lenders and 
discussions with officials with substantial manufactured housing 
lending experience, we identified some important characteristics of the 
performance of home-only manufactured housing loans. 

Our estimates rely on assumptions concerning a few key inputs: annual 
prepayment rates, annual default rates (which vary over different time 
intervals), and the net recovery rate (which measures the portion of 
the loan balance recovered by the lender in cases of default). Further, 
because FHA has not yet developed its risk-based pricing criteria for 
the proposed legislative changes, we made different assumptions about 
the level of up-front mortgage insurance premiums and periodic 
insurance premium payments based on the amounts discussed in the 
proposed legislation. By varying the default rate, loss recovery, and 
premium rate assumptions, we were able to generate a variety of loan 
performance and recovery scenarios, and illustrate in a very general 
way the potential for gains and losses to FHA General Insurance Fund 
that characterize each scenario. 

In the absence of available data on the credit of FHA borrowers and the 
location of the homes (owned or leased land), we attempted to benchmark 
these scenarios based on the experience of FHA's Title I program since 
1990 and of non-FHA personal property manufactured housing loans. In 
terms of FHA Title I experience since 1990, while the number of loans 
originated dropped significantly from the early to mid-1990s, 
cumulative defaults expressed as a percentage of originated loans did 
not fall below 10 percent from 1990 to 2002 and have exceeded 25 
percent in 8 of the 13 years (see fig. 18). However, loans from 2003 to 
2006 may not be reflective of the default experience because they are 
recent loans and lending industry officials explained that the peak 
default period for these types of loans generally occurs from the third 
to the fifth year. In terms of non-FHA loan performance, cumulative 
losses typically have been above 15 percent for loans originated 
between 1997 and 2001. 

Figure 18: Number of FHA Loans and Percentage of Loans in Default, 1990-
2006: 

[See PDF for image] 

Source: GAO analysis of FHA data.

[End of figure] 

The scenarios incorporate assumptions based on factors such as annual 
default rates for different yearly intervals, loan interest rates, and 
loan terms.[Footnote 50] Once we established these parameters, we 
factored in additional assumptions and variations for the net recovery 
rate of the lender and an insurance premium schedule for the borrower 
based on discussions with lending industry officials on possible 
default scenarios, recovery outcomes, and possible legislative changes 
regarding FHA's upfront and annual premiums. The discussion below 
provides more detailed information on our assumptions. 

* Assumptions on Annual Default Rates. We characterized the peak period 
of default as years 3 through 5, and we described the default 
experience in years after this peak period in terms of a percentage of 
the default rate assumed to hold during the peak period. In general, 
and based on our discussions with lenders and others, we assumed that 
default rates in years after the peak period would be 75 percent of 
what they were during the peak period. In the high loss scenario, we 
assumed that the peak period default rate also held in years 6 through 
9 before dropping to 75 percent of the peak value. Our assumptions 
about default rates reflect an important characteristic of home-only 
manufactured housing loans: Even after years of loan amortization, a 
borrower may not have enough equity in the home to avoid a default in 
the face of adverse financial conditions. 

We present three variations of default: a low default experience, a 
moderate default experience, and a high default experience. In general, 
the low default experience would reflect conditions in which borrowers 
possessed good credit quality, lenders used high quality underwriting 
requirements, and lenders' security interests were well protected in 
terms of those factors that are associated with the preservation of 
value, such the placement of the home (owned land versus leased land) 
and installation. The high default experience would reflect conditions 
in which borrowers are of poorer credit quality, and collateral values 
and lenders' security interests are also poorer (see fig. 19). 

Figure 19: Assumptions of Default Risk Used in Our Analysis: 

[See PDF for image] 

Source: GAO.

[End of figure] 

* Assumptions Based on Annual Prepayment Rates. We assumed that 
prepayments were constant at 4 percent per year. Modest changes in this 
level did not lead to much difference in our results. Based on our 
discussions with lenders and others, we believe manufactured home-only 
loan borrowers were not as likely as other homeowners to prepay in the 
face of favorable refinancing opportunities. As a result, some of these 
loans default in later years, but they also continue to generate annual 
insurance premiums. 

* Additional Scenario Assumptions. Using the prepayment rates and 
default rates that we selected, we calculated the value of claims in a 
given year as the (unpaid) principal balance due in that year based on 
an amortization schedule relating the selected interest rate and loan 
term. Based on assumed prepayment and default patterns, we calculated 
cumulative defaults and losses, expressed as a percentage of the 
original loan balance, losses, and insurance premiums paid by year. We 
also calculate the present value of FHA's share of losses and the 
present value of annual insurance premiums.[Footnote 51] 

* Assumptions on the Net Recovery Rate of Lenders. To provide 
variations in our analysis, we make different assumptions on the 
lenders' ability to recover losses when a loan defaults. Based on 
discussions with industry officials, we assume lenders that have a 
strong recovery program (which may include a good network of dealers 
who resell manufactured homes) may have a net recovery of 50 percent 
per claim. Those lenders who have moderate net recovery are assumed to 
receive 33 percent of the claim, and those lenders with a low net 
recovery may receive 25 percent of the claim. 

* Assumptions on the Insurance Premiums. Insurance premiums may include 
an up-front payment and annual payment. FHA has not yet developed its 
proposed risk-based pricing for potential FHA Title I Manufactured Home 
Loan borrowers. However, several bills introduced in Congress suggests 
the up-front annual insurance premiums would not exceed 2.25 percent 
and the annual insurance premium would be 1 percent of the annual 
unpaid principal balance of the loan. For our analysis, we assumed two 
different potential up-front premium amounts: the highest up-front 
premium was 2.25 percent of the original loan amount and the lowest up- 
front premium was 1 percent of the original loan amount. We also assume 
two different annual premiums; 
the highest annual premium was defined as 1 percent of the declining 
loan balance and lowest annual premium was defined as 0.5 percent of 
the declining loan balance. 

[End of section] 

Appendix III: Comments from the Department of Housing and Urban 
Development: 

Us. Department Of Housing And Urban Development: 
Washington D.C. 20410-8000: 

Assistant Secretary: 
Federal. Housing Commissioner:

August 10, 2007:

Mr. William B. Shear:
Director:
Financial Markets an Community Investment: 
Government Accountability Office: 
441. G Street NW: 
Washington, DC 20548:  

Dear Mr. Shear: 

Thank you for the opportunity to comment on the GAO Draft Report 
entitled Federal Housing Administration: Agency Should Assess the 
Effects of Proposed Changes to the Manufactured Home Loan Program. 

While the Department agrees that manufactured housing is one of the 
most affordable housing options low- and moderate-income families, 
HUD's overarching concern is that the financing options available to 
buyers of these homes are limited and costly. Because the Title 1 
Manufactured Home Loan program is the only active federal loan 
insurance/loan guarantee program for manufactured housing, HUD wants to 
not only preserve this source of affordable housing but to also improve 
the program, We believe that the changes included in the pending 
legislation for this program are critical to this effort. 

The Department also agrees that it is important to assess the effects 
of the proposed changes and will do so prior to implementing a reformed 
program. FHA's Office of Evaluation has recently initiated a study to 
determine if the loss recovery rates currently used to calculate credit 
subsidy should be increased. The result of this study will provide a 
credit subsidy baseline that FHA can use to develop models to test 
various underwriting and premium pricing options.

Ginnie Mae is willing to consider securitizing Title I Manufacture Home 
loans depending on the elimination of the portfolio cap, tightening 
underwriting standards and implementing individual loan insurance. 
Ginnie Mac's willingness to consider expanding the secondary market to 
include Title I manufactured home loans is key to adding liquidity to 
an otherwise tight manufactured home financing market. 

As previously recommended by GAO, HUD has been collecting additional 
data regarding Title manufactured home loans since 2004 including 
borrower information such as age and income. HUD intends to expand data 
collection concurrent with programmatic changes so so that the 
procedures for originating and underwriting Title I loans will mimic 
those of FHA's real estate financing programs. HUD will collect 
appropriate credit and application variables including credit scores as 
part of its implementation of the proposed revised loan insurance 
process.

HUD has not collected data on the placement of the home (i.e, owned vs, 
leased) in connection with home-only Title I manufactured home loans 
because the current program requirements are essentially the same for 
either scenario. HUD will begin collecting this data when it implements 
the revised program in order to better track loan characteristics. 

Again, we appreciate the opportunity to comment on the Draft Report and 
will continue, as recommended by the GAO, to assess the effects of the 
proposed changes to the Title I program, and will collect additional 
information to effectively manage the program when we implement these 
changes. 

If you have any questions regarding this letter, please contact 
Margaret Burns at 2O2-7O8 2121, extension 3989. 

Sincerely:

Signed by:

Brian D. Montgomery: 
Assistant Secretary for the Housing: 
Federal Housing Commissioner: 
Assistant Secretary for Housing - Federal Housing Commissioner: 

[End of section] 

Appendix IV: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

William B Shear, (202) 512-8678 or shearw@gao.gov: 

Staff Acknowledgments: 

In addition to the contact named above, Andy Finkel (Assistant 
Director), Steve Brown, Tania Calhoun, Nadine Garrick, Phil Herr, 
Alison Martin, John Mingus Jr., Marc Molino, Tina Paek, and Barbara 
Roesmann made key contributions to this report. 

FOOTNOTES 

[1] S.1741, 110th Cong. (2007) and H.R. 2139, 110th Cong. (2007). 

[2] The American Housing Survey uses building permit data to draw its 
sample and the sample was changed in 2005 since it appeared 
manufactured homes were undercounted in previous years. 

[3] At one of the states, Georgia, we conducted interviews on-site. 

[4] The American Housing Survey definition of monthly housing costs 
encompasses electricity, gas, fuel oil, other fuels, garbage and trash, 
water and sewer, real estate taxes, property insurance, condominium 
fees, homeowner's association fees, mobile home park fees, land or site 
rent, other required mobile home fees, rent, mortgage payments, home 
equity loan payments, other charges included in mortgage payments, and 
routine maintenance. 

[5] The General Insurance Fund is used to support several FHA insurance 
programs including the Title I Manufactured Home Loan, Property 
Improvement, Home Equity Conversion Mortgages, Mortgage Insurance for 
Condominium Units, and Rehabilitation Home Mortgage Insurance. 

[6] GAO, Standards for Internal Control in the Federal Government, GAO/ 
AIMD-00-21.3.1 (Washington, D.C.: November 1999); and Internal Control 
and Management and Evaluation Tool, GAO-01-1008G (Washington, D.C.: 
August 2001). 

[7] Housing and Urban Development Act of 1969, Pub. L. No. 91-152, 
Title I, Section 103, 83 Stat. 379, 380 (December 24, 1969). 

[8] 42 U.S.C. §§ 5401-5426 with implementing regulations at 24 C.F.R. 
Parts 3280 and 3282. 

[9] In the HUD Code, a manufactured home is defined as a transportable 
structure built on a permanent chassis and designed to be used as a 
dwelling on a building site. The HUD Code also defines a manufactured 
home as being a minimum of 320 square feet; however, the dimensions of 
a manufactured home will vary depending on the number of sections that 
make up the home and the state laws for transporting the sections. 

[10] The label indicates that the manufacturer certifies that the home 
meets the HUD Code and has an identification number stamped on it. 

[11] Homes placed in rented parks are typically financed as personal 
property because conventional single-family mortgage programs usually 
require that the land and property be bundled to qualify. 

[12] For example, while state laws vary, in a for-profit cooperative, 
the manufactured home park returns any profits made, in full, to the 
members of the association; or, in a nonprofit cooperative, the 
manufactured home park returns any profits to the cooperative and not 
to individual members. 

[13] FHA began collecting demographic data (such as age, race, and 
monthly income) for the Title I Manufactured Home Loan program in 2004. 
No demographic data are available prior to this time. 

[14] The FHA Title I property improvement program insures loans to 
finance the light or moderate rehabilitation of properties, as well as 
the construction of nonresidential buildings on the property. This 
program may be used to insure loans on either single-or multifamily 
properties for up to 20 years. The maximum loan amount is $25,000 for 
improving a single-family home or improving or building a 
nonresidential structure. For improving a multifamily structure, the 
maximum loan amount is $12,000 per family unit, not to exceed a total 
of $60,000 for the structure. FHA insures private lenders against the 
risk of default for up to 90 percent of any single loan. 

[15] Census regions are groupings of states that subdivide the United 
States for the presentation of data. There are four Census regions-- 
Northeast, Midwest, South, and West. 

[16] At the time of the survey, the respondents either owned a home or 
were in the process of obtaining homeownership. 

[17] The American Housing Survey only collects data on structure and 
tenure (whether the home itself is owned or rented), but does not ask 
survey respondents whether the land on which the home is sited is owned 
or leased. Moreover, the American Housing Survey does not ask 
respondents whether the manufactured home is located in a manufactured 
home park. The Manufactured Housing Survey does collect data on the 
location of the home (park, court, subdivision, or private property) 
but the data do not indicate whether the land is owned or not. Thus, 
data are limited on land ownership. 

[18] The American Housing Survey definition of monthly housing costs 
encompasses electricity, gas, fuel oil, other fuels, garbage and trash, 
water and sewer, real estate taxes, property insurance, condominium 
fees, homeowner's association fees, mobile home park fees, land or site 
rent, other required mobile home fees, rent, mortgage payments, home 
equity loan payments, other charges included in mortgage payments, and 
routine maintenance. 

[19] Escrow is the holding of funds, documents, securities, or other 
property by an impartial third party for the other two participants in 
a business transaction. When the transaction is completed, the escrow 
agent releases the entrusted property. 

[20] Judicial foreclosures are processed through court actions, but 
nonjudicial power of sale foreclosures are processed without court 
involvement, based on the lender's exercise of the power of sale 
contained in the mortgage or deed of trust. 

[21] A breach of peace is a generic term that includes violations or 
disturbances of public peace or order. Although the use of self-help 
repossession is not common, when it is employed it can place the 
homeowner at a significant disadvantage because the home can be removed 
from the land without much notice and still containing possessions of 
the owner. 

[22] For FHA Title I loans, before acting to accelerate the maturity of 
the loan, the lender generally must contact the borrower in person or 
by telephone to discuss the reasons for the default and seek its cure. 
Unless the borrower cures the default or agrees to a modification of 
the loan terms or a repayment plan, the lender can proceed to take 
action by providing written notice to the borrower. During this time, 
the borrower is given a second chance to cure the default. 

[23] In the states where the law permitted a rent increase during the 
lease term, it did not appear that this would override the terms of a 
written lease agreement. 

[24] Good cause eviction means that eviction must be based upon a 
legally sufficient ground, which might include actions such as 
noncompliance with a provision of the rental agreement, nonpayment of 
rent, or change in use of land. 

[25] At the time of our review, North Carolina and Oregon had 
introduced legislation to pass similar laws. 

[26] Since 1984, the New Hampshire Loan Fund has converted 86 
manufactured home parks to nonprofit cooperative ownership. 

[27] The HUD Code requires that homes be designed and constructed to 
conform to one of three wind load zones. Wind Zone I equates to a 70- 
mph fastest-mile wind speed (i.e., the fastest speed in miles per hour 
within a specified period, usually 24 hours), Wind Zone II equates to a 
100-mph fastest-mile speed, and Wind Zone III to a 110-mph fastest-mile 
speed. The appropriate wind zone used in design is dependent on where 
the home initially will be installed. Homes designed and constructed 
for a higher wind zone can be installed in a lower zone (a zone III 
home can be installed in a zone I or II location). However, a zone I 
home cannot be installed in a zone II or III location. 

[28] Proposed legislation included S. 2123, 109th Cong. (2005), H.R. 
2803, 109th Cong. (2005), S. 3535 109th Cong. (2006), H.R. 4804, 109th 
Cong. (2006), S.1741, 110th Cong. (2007), and H.R. 2139, 110th Cong. 
(2007). 

[29] In 2005, the average price of a single-wide was $33,800 and a 
double-wide was $68,600. 

[30] S.3535 would not change the mechanism that FHA uses to insure 
against risk. 

[31] In contrast, in the Title II program, FHA endorses a loan if it 
meets all applicable regulations and instructions. FHA then issues a 
certificate of insurance, which creates a contract of mortgage 
insurance subject to the regulations in effect at the time. In Title I, 
FHA does not endorse manufactured home loans. That is, it does not 
contractually guarantee payment of eligible claims; rather, it 
acknowledges the insurance through the issuance of premium billing 
statements. 

[32] S. 3535 does not specify the premium amount; rather, it states 
that the Secretary of HUD can establish a mortgage insurance premium 
structure. 

[33] Ginnie Mae defines its mission as expanding affordable housing in 
America by linking capital markets to the nation's housing markets, 
largely by serving as the dominant secondary market vehicle for 
government-backed loan programs. Ginnie Mae does not buy or sell loans 
or issue mortgage-backed securities. Rather, it provides guarantees 
backed by the full faith and credit of the U.S. government that 
investors will receive timely payments of principal and interest on 
securities supported by pools of government-backed loans, regardless of 
whether the borrower makes the underlying mortgage payment or the 
issuer makes timely payments on the securities. All mortgages in the 
Ginnie Mae pool must be insured or guaranteed by a government agency 
and have eligible interest rates and maturities. 

[34] As a result of poor performance in the manufactured home 
portfolios in the early 2000s, government-sponsored enterprises (GSEs), 
such as Fannie Mae and Freddie Mac, changed their underwriting 
requirements to only purchase loans located on owned land. In addition, 
each of the GSEs put size requirements on the type of homes for which 
they would provide guarantees. For example, Fannie Mae underwriting 
guidelines require a manufactured home to be 660 square feet or more, 
while Freddie Mac requires the home to be a double-wide. 

[35] Under California law, when a contract, purchase order, or security 
agreement for the retail sale of a new or used manufactured home (one 
installed on a nonpermanent foundation) is signed, the dealer must put 
the entire down payment or deposit in an escrow account within 5 days 
of receipt. The conditions of sale are written into mandatory escrow 
instructions and require the buyer's signature. The instructions 
specify the conditions of delivery for a manufactured home. For 
instance, if the home is located in a manufactured home park, escrow 
cannot close until the manufactured home owner has signed the park 
rental agreement. Further, the escrow account should not close until 
the home has been delivered and passes inspection. These rules only 
apply to the retail sale of manufactured homes. Manufactured homes sold 
for installation on permanent foundations are subject to escrow 
requirements similar to those of site-built homes. 

[36] The current Title I formula for calculating FHA's payment of 
claims limits the amount of the claim to 90 percent of the loan and is 
based on the best price attainable, which is either the net sale price 
or appraisal value, whichever is higher, and then includes the 
subtraction of the unpaid balance of the loan. 

[37] According to HUD officials, a proposed rule is currently under 
review to require a consistent installation program across the states. 

[38] The Manufactured Housing Improvement Act of 2000 changed methods 
for establishing construction, safety, and installation standards for 
manufactured housing and created a dispute resolution program 
administered by the states. HUD's Manufactured Housing program, which 
is administered through state or third-party agencies, such as the 
Department of Housing and Community Affairs in Texas, the State Fire 
Marshall's office in Georgia, and the Department of Motor Vehicles in 
Florida, monitors these standards. Depending on the state, the state or 
third-party agencies act as (1) design approval and primary inspection 
agencies to check and approve the designs and calculations used in the 
construction of manufactured homes; (2) in-plant inspection agencies to 
certify and inspect manufactured homes during the manufacturing process 
to ensure that the manufacturer is in compliance with the standards and 
with approved designs; and (3) state administrative agencies to handle 
consumer complaints, conduct inspections, make enforcement 
determinations, and conduct hearings. 

[39] The annual default rates consisted of different yearly intervals: 
years 1 and 2, years 3 through 5, years 6 through 9, and years 10 and 
after. 

[40] GAO, Standards for Internal Control in the Federal Government, 
GAO/AIMD-00-21.3.1 (Washington, D.C.: November 1999); and Internal 
Control and Management and Evaluation Tool, GAO-01-1008G (Washington, 
D.C.: August 2001). 

[41] As discussed in previous work, historically, FHA has been known to 
underestimate the subsidy rate. GAO, Federal Housing Administration: 
Modernization Proposals Would Have Program and Budget Implications and 
Require Continued Improvements in Risk Management, GAO-07-708, 
(Washington, D.C.: June 29, 2007). 

[42] GAO-07-708. 

[43] At one of the states, Georgia, we conducted interviews on-site. 

[44] The American Housing Survey only collects data on structure and 
tenure (whether the home itself is owned or rented), but does not ask 
survey respondents whether the land on which the home is sited is owned 
or leased. Moreover, the American Housing Survey does not ask 
respondents whether the manufactured home is located in a community or 
manufactured home park. In contrast, the Manufactured Housing Survey 
collects data on the location of the home (park, court, subdivision, or 
private property) but the data do not indicate whether the land is 
owned or not. Thus, data are limited on land ownership. 

[45] To identify the newspaper articles, we used the newspaper 
databases in Nexis, which included a search of national newspapers and 
wire services, as well as a regional search of newspaper in the 
Southeast, West, Northeast, and Midwest. Our search parameters included 
terms such as manufactured home, mobile home, mobile home park, park 
closure, and conversion. 

[46] Arizona and Oregon each had two separate industry groups, one that 
represented many industry constituents, such as lenders and insurers, 
and another that represented the owners of manufactured home parks. 

[47] The General Insurance Fund, which is supported by insurance 
premiums, is used for several FHA insurance programs such as, Title I 
Manufactured Home Loan, Property Improvement, Home Equity Conversion 
Mortgages, Mortgage Insurance for Condominium Units, and Rehabilitation 
Home Mortgage Insurance. 

[48] The Home Mortgage Disclosure Act requires lending institutions to 
collect and publicly disclose information about housing loans and 
applications for such loans, including the loan type and amount, 
property type, and borrower characteristics (such as ethnicity, race, 
gender, and income). These data are the most comprehensive source of 
information on mortgage lending. To determine the loans considered 
personal property loans, we determined any loan with a rate spread of 4 
percent or higher was considered a personal property loan. We discussed 
this determination with HUD, which agreed with this threshold. 

[49] The General Insurance Fund, which is supported by insurance 
premiums, is used for several FHA insurance programs, such as Title I 
Manufactured Home Loan, Property Improvement, Home Equity Conversion 
Mortgages, Mortgage Insurance for Condominium Units, and Rehabilitation 
Home Mortgage Insurance. 

[50] The annual default rates consisted of different yearly intervals: 
years 1 and 2, years 3 through 5, years 6 through 9, and years 10 and 
after. 

[51] The present value of a future stream of payments or expenses takes 
into account the time value of money. Describing cash flows in present 
value terms leads to better evaluations, particularly if they differ 
greatly in their timing. 

GAO's Mission: 

The Government Accountability Office, the audit, evaluation and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the performance 
and accountability of the federal government for the American people. 
GAO examines the use of public funds; evaluates federal programs and 
policies; and provides analyses, recommendations, and other assistance 
to help Congress make informed oversight, policy, and funding 
decisions. GAO's commitment to good government is reflected in its core 
values of accountability, integrity, and reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, go to [hyperlink, http://www.gao.gov] 
and select "Subscribe to Updates." 

Order by Mail or Phone: 

The first copy of each printed report is free. Additional copies are $2 
each. A check or money order should be made out to the Superintendent 
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 
more copies mailed to a single address are discounted 25 percent. 
Orders should be sent to: 

U.S. Government Accountability Office: 
441 G Street NW, Room LM: 
Washington, D.C. 20548: 

To order by Phone: Voice: (202) 512-6000: 
TDD: (202) 512-2537: 
Fax: (202) 512-6061: 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

Congressional Relations: 

Gloria Jarmon, Managing Director, JarmonG@gao.gov (202) 512-4400: 

U.S. Government Accountability Office, 441 G Street NW, Room 7125: 
Washington, D.C. 20548: 

Public Affairs: 

Susan Becker, Acting Manager, Beckers@gao.gov (202) 512-4800: 

U.S. Government Accountability Office, 441 G Street NW, Room 7149: 
Washington, D.C. 20548: