Financing Low Income Housing in Africa by emc12393


More Info
									                             Current Practices for
                             Financing Affordable
                             Housing in the United

Cambridge, MA
Lexington, MA
Hadley, MA
Bethesda, MD
Washington, DC               1996
Chicago, IL
Cairo, Egypt
Johannesburg, South Africa   Prepared for
                             Fannie Mae Foundation Office of
                             Housing Research
                             1996 Tri-Country Conference on
                             Housing and Urban Issues

                             Prepared by

                             Michael J. Lea
                             Cardiff Consulting Services

                             James E. Wallace
                             Abt Associates Inc.
Current Practices for Financing Affordable
    Housing in the United States*

Michael J. Lea
Cardiff Consulting Services

James E. Wallace
Abt Associates Inc.


This article first sets a context for reviewing affordable housing finance in the U.S. by addressing the need for
subsidy and the risks involved to private suppliers of funds for affordable housing. U.S. tools for financing
affordable housing include capture of equity for rental housing in the low-income housing tax credit (LIHTC), a
housing block grant (the HOME program), effective interest subsidies provided through tax-exempt bonds and
below-market-interest funds provided to members of the Federal Home Loan Bank (FHLB) system, mortgage
insurance and guarantees, regulatory influences on mortgage capital through the Community Reinvestment Act
and through housing goals of the major secondary mortgage market entities (Fannie Mae and Freddie Mac), and
state and local devices such as housing trust funds.

The current options for financing affordable housing in the United States provide housing affordable to moderate-
income home purchasers and to renters in the range of 50 to 60 percent of local median income. Assistance
beyond that provided in the housing finance system is necessary to reach households at lower income levels. The
diffusion of incentives to state, local, and private sources of financing and the use of implicit or explicit tax devices
make it very difficult to assess their overall impact or cost effectiveness.


Affordable finance of housing has been a concern of government policy since the creation of the
FHA in 1932. Over the years there has been considerable experimentation with different tech-
niques including mortgage instrument design, interest rate subsidies, down payment grants,
insurance and guarantee programs and tax incentives to lenders, builders and buyers. Today the
U.S. does not rely on a particular approach or program. Affordable housing finance programs
exist on the federal, state, and local level, and incorporate a wide variety of techniques to reduce
financing costs.

 This research was supported by the Fannie Mae Foundation Office of Housing Research. The Foundation has
agreed to permit publication of the paper.

                                                          1996 Tri-Country Conference on Housing and Urban Issues
2   Michael J. Lea and James E. Wallace

This situation has arisen in part because of government budgetary constraints and in part from a
desire to de-centralize the provision of housing programs. In recognition of declining government
support for affordable housing, there has been an increased emphasis on stimulating private sector
finance for affordable housing in the form of goals and requirements and through linking of
regulatory approvals to lender performance.

This approach has both strengths and weaknesses. On one hand, the use of multiple techniques by
multiple different organizations has led to innovative programs and flexibility in design. On the
other hand, this approach has created a bewildering array of programs and program features that
increase the cost of provision and monitoring. Furthermore, many of the programs offer relatively
shallow subsidies (reflecting the limited resources provided by different levels of government),
which frequently necessitates combining several different programs to finance one project.

This article provides an overview of the main approaches to the provision of affordable housing
finance in the U.S. today. In the next section we provide a definition of affordable housing and
identify the major barriers that exist to the private sector provision of finance for both owner-
occupied and rental housing. We then review the major approaches that can be used to assist in
the finance of housing including subsidization and risk-reallocation. Finally, we review the major
government programs, and in the concluding section we provide an assessment of the current
approaches to the finance of affordable housing.

What is Affordable Housing?

Need for Subsidy

We define affordable housing as housing for low-income households (incomes at or below 80
percent of local median income, adjusted for family size), and very low income households
(incomes at or below 50 percent of local median income, adjusted for family size).1 In 1993 there
were 26.5 million households with incomes under 50 percent of U.S. Department of Housing and
Urban Development (HUD) defined adjusted area median income (14.7 million renters and 11.8
million owners), and an additional 16 million households with incomes between 50 and 80 percent
of HUD adjusted area median income (6.4 million renters and 9.7 million owners) (HUD 1996a
A-2). Affordable housing can either be owner-occupied or rental tenure. Current federal
programs and private sector incentives do not directly address those households in what HUD and
the U.S. Congress have defined as “worst-case” housing.

 See Wallace 1995 for a more comprehensive discussion of the definition of affordable housing. Both the LIHTC
and the GSE affordable housing goals (which will be discussed extensively in this article) use a definition of 60
percent of local median income.

1996 Tri-Country Conference on Housing and Urban Issues
                                     Current Practices for Financing Affordable Housing in the United States       3

As a measure of housing need, HUD annually produces a congressionally-mandated report on
“worst case” housing needs. Worst case renters are those not receiving federal housing assist-
ance; with incomes below 50 percent of median family income in their area, as adjusted by HUD
for family size; who pay more than half of their income for rent and utilities; or who live in
severely substandard housing. The number of worst case renters had reached an all time high of
5.3 million in 1993. Almost 2 million of these have a working member of the household. Despite
large and growing needs for housing affordable to worst case renters with extremely low incomes
(under 30 percent of median area income), housing markets are not supplying units affordable for

Affordability of Owner-Occupied Housing

From the borrower’s perspective there are two dimensions of affordability: the ability to service
the loan (the payment burden) and the ability to provide the down payment. As reported by
Harvard’s Joint Center for Housing Studies (1995), the average first-time buyer (household
income of $24,265) buying the average-priced ($67,959) house in 1994 had an after-tax cash
burden of approximately 30.4 percent of income, the lowest level recorded since 1986 (figure 1).2
This buyer, however, had a ratio of up-front costs (down payment plus closing costs) to income
of 58 percent. Lowering the down payment requirement to 10 percent reduces the up-front cash
payment from $14,200 to $7,600, or 31.3 percent of annual income.

  After-tax cash cost is the ratio of after-tax housing costs (mortgage payment, property tax, insurance, fuel,
utilities, and maintenance) assuming a 30-year fixed-rate conforming mortgage with a 20 percent down payment
for the American Housing Survey median value house purchased by a first-time home buyer aged 25 to 29 divided
by the average income of young renters (married couple renters aged 25 to 29 in 1989 dollars). Down Payment/
Closing Cost is the ratio of down payment (20 percent of the median value house) plus average closing costs
divided by the average income of young renters. The average income of all renters ($15,814) is considerably less
than the average income of young renters ($24,265). See Table A-1 in the Joint Center study for more detail.

                                                      1996 Tri-Country Conference on Housing and Urban Issues
4   Michael J. Lea and James E. Wallace

                               Figure 1. Home Ownership Cost Burdens

                          Source: Joint Center for Housing Studies

These data suggest that the down payment constraint is the major obstacle facing first-time home
buyers today.3 One solution is the provision of higher LTV loans. A household can get a 95
percent loan with private mortgage insurance. (A 5 percent down payment requirement would
lower the household’s up-front savings requirement to $3,398, or 14 percent of annual income.
The annual premium for mortgage insurance would increase the after-tax cash burden to 29.9
percent of income.) The median net wealth of young renters, however, is less than $2,500,
suggesting that a majority of these households would be rationed out of the market even with 95
percent loans.

 There is obviously a great deal of regional variability in these numbers. In high cost areas (e.g., New England,
coastal California) these ratios are considerably higher.

1996 Tri-Country Conference on Housing and Urban Issues
                                        Current Practices for Financing Affordable Housing in the United States   5

Of course, the average first-time buyer does not have to purchase the average house. If the
household purchased a house with a value of only 75 percent of the average with a 95 percent
down payment, the required down payment would fall to $2,548.4 Moving down the income
scale, however, the situation becomes more problematic. For a low- to moderate-income first-
time home buyer with income at 80 percent of average, the down payment burden for a house that
is 80 percent of median is 15.6 percent of annual income. The ratios for this class of home buyer
are likely to be a bit higher than those of the average-income household, reflecting the fact that
some of the home ownership costs are fixed and not a function of house value (e.g., maintenance,

Although 95 percent LTV loans are available in the U.S., they are not available from all lenders or
to a large proportion of borrowers. The reason is that such loans are considerably more risky
than lower LTV loans.5 Affordable owner-occupied housing default risk is likely to be higher
than other types of single-family mortgage loans due to lower down payments as well as less
stable incomes. Research by Freddie Mac indicates that default rates on conventional 95 percent
loan-to-value (LTV) loans are more than 15 times greater than on loans with LTVs of 80 percent
or less (Van Order 1995). Default losses from the higher LTV loans are more than seven times
those with the lower LTVs.

Van Order and Schnare (1994) analyzed the relative default experience of affordable owner-
occupied housing loans with loans to higher income households in Freddie Mac’s portfolio. As
shown in table 1, default rates rise substantially with LTV. Default rates for both low-income (51
to 80 percent of metropolitan statistical area [MSA] median) and high-income (120 percent or
more of MSA median) borrowers are about 20 percent higher than default rates for moderate-
income borrowers. Low-income borrowers with LTVs between 91 and 95 percent experience the
highest default rates of these groups. Loan losses (as a percent of the outstanding balance before
mortgage insurance reimbursement) for the different income groups are compared in table 2.
Loans to low-income households have the highest severity of default loss, which also rises with
LTV. As noted by the authors, these results are similar to the Federal Housing Administration
(FHA) experience.

  As discussed by Bradley and Zorn (1996) there are many non-financial factors that also deter young renters from
becoming homeowners. These include the fear of rejection and complexity in the loan application process, the fear
of purchasing the “wrong house,” and concerns over employment stability.
    Affordable housing loans are also smaller and more costly to service.

                                                         1996 Tri-Country Conference on Housing and Urban Issues
6     Michael J. Lea and James E. Wallace

                                Table 1. Probability of Ever Defaulting

    Household Income as                                                LTV
    Percent of MSA Median                     0-60%        61-80%      81-90%        91-95%         All
    51-80%                                      0.1%         1.2%        3.3%          8.8%         2.8%
    81-120%                                     0.1          0.9         2.6           7.2          2.3
    120%+                                       0.1          1.4         4.1           8.1          2.8
    All                                         0.1          1.2         3.4           7.9          2.6

                                            Table 2. Loss Severity

    Household Income as                                                LTV
    Percent of MSA Median                     0-60%        61-80%       81-90%       91-95%          All
    51-80%                                      59%          47%          57%          67%           60%
    81-120%                                     51           35           44           54            48
    120%+                                       44           31           36           47            39
    All                                         51           36           43           55            47

Source: Van Order and Schnare 1994.

There are several approaches to improving the affordability of owner-occupied housing. These
include programs to lower the interest cost of mortgages (e.g., interest rate subsidies or
buydowns), lowering the risk of granting mortgages to low- and moderate-income home buyers
(and thus indirectly lowering the interest cost) through guarantees or insurance, or lowering the
down payment requirement through granting of higher LTV loans (typically combined with
insurance). Interestingly, the provision of lump-sum grants to home buyers, conditioned by
savings behavior, has not been extensively used in the U.S.6

The difficulties in providing mortgage loans on affordable terms to lower-income households
suggests that programs to lower housing cost and encourage savings could be significant in
improving the affordability of owner-occupied housing for first-time home buyer households. The
President’s Commission on Affordable Housing (1983) detailed a number of ways to reduce the
cost of new housing, primarily through relaxation of land use regulations, building codes, and
local taxes. Encouraging savings for housing has not been tried in the U.S. A number of
countries have well-established contract savings programs for housing. These programs have
been shown to be effective in stimulating aggregate savings as well as creating a pool of long-term

  Economists have long extolled the efficiency of such grants, and they are extensively used in developing
countries. For a review of the Latin American experience, see Ferguson et al. 1996.

1996 Tri-Country Conference on Housing and Urban Issues
                                      Current Practices for Financing Affordable Housing in the United States      7

funds for housing.7 They typically combine a tax exemption on interest earned, a cash bonus paid
by the government, and in some cases the availability of a low-interest second mortgage when the
savings target is reached.

Affordability of Rental Housing

The economics of affordable rental housing finance is based on the cash flows produced by a
project, the discount rates used by lenders and investors in assessing project feasibility, and the
cost of debt and equity. The example shown in table 3 illustrates the parameters the developer
and lender would look at to determine the feasibility of investment in affordable multifamily rental
housing. The numbers are for illustration only, but are meant to be broadly representative of this
type of investment.8 In this example, the developer is assumed to reserve 20 percent of units for
households earning 60 percent of U.S. median income, reserve 20 percent of units for households
earning 80 percent of median income, and rent the rest of the units to households earning 120
percent of median income. Total project development cost is $85,000 per unit for a 100-unit
project. Households are assumed to pay 25 percent of their monthly income in rent after allow-
ance for utilities of $120 per month. As shown in the second section of table 3, the project
generates an annual gross rental income of $846,000 and a net operating income of $543,700.

The calculations at the bottom of table 3 show the maximum loan this project could support,
assuming all of the net operating income is applied to debt service, or $5.4 million. This would be
63.7 percent of development cost in this example. This leaves the developer with a financing gap
that must be covered by equity, subsidies, or non-interest-bearing loans. In reality, the developer
is unlikely to get a first mortgage for this amount. The debt-service coverage ratio (DSCR) in this
example is only 1.0 (all of the net operating income is dedicated to debt service), well below the
usual minimum acceptable ratio of 1.2. A DSCR of 1.2 implies a loan of $4.5 million at the 8
percent interest rate, which leaves a financing gap of $4 million, or 47 percent of project cost.
Another reason why the developer is unlikely to get a loan of $5.4 million is that the lender will
not appraise the project value at the development cost. Rather, the lender will use a discounted
cash flow valuation or apply a capitalization rate to the first year (or stabilized) net income
stream. Assuming a capitalization rate of 9 percent, the project appraised value would only be $6
million (in part reflecting the effects of the affordable units on the project’s net income stream). A

  International experience with targeted contract savings for housing programs suggests that they do increase
aggregate savings. See Lea and Renaud (1994) for a discussion of European experience, and Engelhardt (1994) for
a discussion of the Canadian experience.
   From Miles, Haney, and Berens (1996, F307-8), Figure 15-5. The typical terms for apartment loans made by
life insurance companies in 1995 were for 10 to 15 years at yields with fees around 8 percent at LTVs of 70
percent, debt-service coverage ratio of 1.5, and capitalization rates around 9 percent (American Council of Life
Insurance, 1996).

                                                       1996 Tri-Country Conference on Housing and Urban Issues
8   Michael J. Lea and James E. Wallace

lender using an 80 percent LTV ratio would lend $4.8 million, except that the DSCR approach
would limit the loan to $4.5 million in any case.

                Table 3. Economics of an Affordable Multifamily Rental Project

Assumptions                                 Inputs
Loan Term                                            240 months
Loan Rate                                        8.00% annual
Total Units                                          100
Project Development Cost                    $8,500,000 Total
Vacancy %                                        5.00%
Per Unit Annual Op Exp                         ($2,600) per unit
Median Income                                  $39,600

% of Total Units (Input)                             60%       0%          20%       20%      0%
Number of Units                                       60           0        20        20       0
% of Median Income (Input)                        120%      100%           80%       60%    25%
Monthly Income                                  $3,960      $3,300       $2,640    $1,980   $825
25% of Monthly                                       $990    $825         $660      $495    $206
Utility Allowance                                    $120    $120         $120      $120    $120
Rent Income Available                                $870    $705         $540      $375     $86
Total Monthly Rent                             $52,200          $0      $10,800    $7,500     $0

Total Annual Rent                             $626,400          $0     $129,600   $90,000     $0

Gross Annual Rent                             $846,000
Vacancy                                       ($42,300)
Total Annual Operating Expense              ($260,000)
Net Annual Rent Available                     $543,700

Max Loan Given Net Rent                     $5,416,806
Mortgage Payment                            ($543,700)
Loan as % of Development Cost                   63.73%
Financing Gap                               $3,083,194
Financing Gap/Unit                             $30,832
Financing Gap % of Devlp. Cost                  36.27%

1996 Tri-Country Conference on Housing and Urban Issues
                                      Current Practices for Financing Affordable Housing in the United States   9

This example shows why there is little unsubsidized private market financing of affordable newly-
constructed rental housing. Quite simply, at the income levels for these households, costs of
development and market mortgage interest rates, the “numbers do not work.” The developer is
unlikely to find $3.9 million of private equity capital for this project, which generates an internal
rate of return of only 8.5 percent on investor capital (assuming the standard depreciation allow-

Although the unassisted mortgage insurance programs have no income targeting requirements,
they do typically have per-unit dollar caps that target them toward the lower cost end of the
market. From data provided in the 1991 national sample of the American Housing Survey, one
can estimate the percentage of rental units that would be affordable (at 30 percent of income) to
households of various incomes relative to the area median. The percentage of households at
median income that could afford rents can be estimated by type of structure and whether
existing or new, as shown in table 4.

                         Table4. Percentage of Affordable Units by Property
                                      Type and Income Level

                                        Affordable by                           Affordable by Tenants with
                                        Tenants with Incomes Under 50           Incomes Under Area Median
 Rental Unit Type                       Percent of Area Median Income           Income

 Existing 1–4 Unit(s)                                   NA                                  80%
 New 1–4 Unit(s)                                        20%                                 66%
 Existing 5+ Units                                      41%                                 78%
 New 5+ Units                                            9%                                 87%

Source: Tabulations of the 1991 American Housing Survey
Note:   “New” is defined as units constructed within the three years prior to the survey.

The economics of affordable rental housing are further complicated by the fact that investors view
multifamily housing as more difficult and risky to manage than other forms of commercial real
estate. As noted by DiPasquale and Cummings (1992), the management of rental housing is less
of a business relationship than commercial real estate. There are complex legal relationships
governing leases and tenant eviction protections, as well as the possibility of government reg-
ulatory restrictions on the rents that can be charged and the ability to convert the property into
condominiums. The upside of rental housing investment is often limited by the location of the
properties and the ability to raise rents in line with expectations and market developments.

                                                       1996 Tri-Country Conference on Housing and Urban Issues
10   Michael J. Lea and James E. Wallace

Multifamily loans are viewed as more risky than single-family loans, reflecting the fact that they
are investment properties in which both the probability of default and the loss per default is
higher.9 American Council of Life Insurance data show much higher delinquency and foreclosure
rates on multifamily than on single-family loans. Nothaft (1994) notes that the spread between
commercial (including multifamily) and residential single-family loans increased from negative
spreads of 20 to 60 basis points in the 1983-90 time period to positive spreads of 50 to 60 basis
points in the 1991-93 period, an increase he attributes to increased default risk. The negative
spread in the 1980s reflects the value of the call protection common in multifamily mortgages.
This measure is imprecise because the prepayment and liquidity aspects of commercial and
residential loans also differ.

It is less clear whether affordable rental housing projects have more or less risk than purely
market rate projects. DiPasquale and Cummings provide a number of reasons why such loans
may be less risky. These include lower vacancy rates reflecting the tightness in the affordable
housing market, the presence of income subsidies that create more rent payment stability, and
more intense and specialized management of such projects by specialized managers. There are,
however, several potential offsetting risk factors. If the project is located in a poor neighborhood,
the maintenance and security expenses may be higher and more variable and the project may have
less upside potential.

The affordability of rental housing finance can be improved through programs to lower the inter-
est cost of debt finance. As with owner-occupied housing, this can be accomplished through
interest rate subsidies, buydowns, and insurance. Because equity is an important component of
rental housing finance, programs that lower the cost of equity (e.g., through tax incentives) are
also a source of government assistance. Affordability can also be improved through reductions in
the cost of inputs (e.g., land use regulations) and enhancements in rental income (e.g., through
direct tenant assistance).

Government Role in Affordable Housing

The federal role in affordable housing encompasses programs of tenant-based rental assistance, a
tax credit program to capture equity capital for low-income rental housing, interest rate subsidies,
insurance and guarantee programs, and providing incentives to private lenders to provide mort-
gage capital in all markets. Before reviewing categories of current federal involvement, it is
useful to set the context of existing affordable housing programs. As summarized in “Rental
Housing Assistance at a Crossroads,” a recent document by the U.S. Department of Housing and

   Research by Vandell et al. (1993) has shown that construction and development loans are significantly more
risky than standing project loans. Among property types, multifamily loans are less risky than hotel and office
projects, and more risky than retail and warehouse lending.

1996 Tri-Country Conference on Housing and Urban Issues
                                Current Practices for Financing Affordable Housing in the United States   11

Urban Development (HUD 1996a, 4), affordable housing supported by the federal government

•   Tenant-based assisted housing. Direct rental assistance to 1.4 million renter households to
    enable them to find their own housing on the open market. Begun in 1974, this type of assist-
    ance has accounted for by far the greatest portion of the incremental units, or additions to
    assisted housing, since the mid-1980s.

•   Public housing, 1.25 million units. Owned by local public agencies. Begun in 1937, heavily
    used to produce assisted housing units until the mid-1980s. New production is now limited to
    partial replacement of units lost to demolition.

•   Project-based assisted private housing. Construction and rehabilitation of 1.8 million rental
    units for low-income households developed from 1968 until the early 1980s. Deep rent
    subsidies attached to projects owned by for-profit and nonprofit sponsors, who must rent units
    to eligible households.

Other federal programs produce affordable housing, but households pay the established rent
rather than a percentage of their income. So without an additional subsidy, the poorest house-
holds often cannot afford this housing. The primary programs are:

•   Equity Capture: the Low-Income Housing Tax Credit. Subsidizes the capital costs of units
    with rents affordable to household with incomes at or below 60 percent of area median
    income. This program has produced more than 400,000 units since its enactment in 1986.

•   Block Grants: the HOME Investment Partnership (HOME) Program and the Community
    Development Block Grant (CDBG) Program. HOME is a formula grant to states and local
    governments that can be used to assist existing homeowners, first-time homebuyers, or
    renters. Between 1990 and 1995, HOME produced 63,000 affordable rental units. Qualifying
    rents must be affordable to households with incomes at or below 65 percent of area median
    income, or below local fair market rents.

•   Interest Subsidies: tax-exempt bond financing and the Affordable Housing Program (AHP) of
    the Federal Home Loan Bank (FHLB). Federal tax law makes provision for states to issue
    tax-exempt bonds for financing home purchase and multifamily development. The AHP of the
    FHLB system provides low-cost on-lending to member banks, which are required to provide a
    certain number of these loans.

C   Insurance and Guarantees: FHA and Veterans’ Administration (VA). The FHA (HUD)
    provides full insurance for single-family and multifamily loans (within limits on loan size and

                                                  1996 Tri-Country Conference on Housing and Urban Issues
12    Michael J. Lea and James E. Wallace

     loan-to-value ratio) and the VA guarantees up to 100 percent of the share of homes purchase
     by veterans.

C    Regulatory Influences on Supply of Mortgage Funds: housing goals for the secondary
     mortgage market and the Community Reinvestment Act.

Each of these is discussed before turning to a brief discussion of state housing trust funds.

Equity Capture: The Low-Income Housing Tax Credit

Description. The notable current federal affordable housing production program, the LIHTC, is
designed to capture equity for affordable housing. As the earlier project example illustrated,
substantial equity and funding sources other than amortized loans are necessary to create afford-
able rental conditions in multifamily housing. The LIHTC provides a stream of ten years of tax
credits available to investor/owners. To obtain a claim to the value of these credits, which reduce
the investors’ federal income tax liability, the investors provide up-front cash payments to the
sponsor/developer of a qualifying project. To qualify, a project must have at least 40 percent of
the units occupied by households with incomes at or below 60 percent of median area income
(adjusted for family size) or, alternatively have at least 20 percent of the units occupied by house-
holds with incomes at or below 50 percent of median area income. Rents must not exceed 30
percent of the applicable income limit. In practice, most projects have nearly 100 percent of
occupants with incomes at or below 60 percent of median. Projects receive the annual tax credit
over the ten-year period only if occupancy is maintained. Owners must make a commitment to
rent the agreed number of units to households under the income limit for a period of 30 years.10 If
the owners fail to maintain the qualifying status of the project, the tax credits are reduced accord-
ingly, and if the project fails to maintain its qualifying status in the extended commitment period
beyond the ten-year credit period, the owners are required to make a pro-rata payment to the
Internal Revenue Service. This obviously provides a financial incentive for the owners to keep the
property in compliance during the commitment period.

The annual allocation of tax credits is $1.25 per capita. Projects financed by tax-exempt bonds
are not subject to the per capita limit on tax credits. The credits are allocated to eligible projects
by housing agencies of the states and territories, plus city agencies in New York and Chicago.
Credits not used by a state become available for reallocation to other states. The annual amount
of the tax credit is a percentage of the qualifying basis (essentially total project costs if all the

   In the original 1986 legislation the commitment period was 15 years. The Omnibus Reconciliation Act of 1989
extended the commitment period to 30 years. The project owners are allowed to sell or convert the project to
conventional market housing, however, if they so apply to the state credit allocation agency and the agency is
unable to find a buyer (presumably a nonprofit organization) willing to maintain the project under its low-income
restrictions for the balance of the 30 years. If such a buyer cannot be found, the tenants are protected with
assistance only for up to three years or when they move out.

1996 Tri-Country Conference on Housing and Urban Issues
                                     Current Practices for Financing Affordable Housing in the United States      13

project units are qualifying). Two percentage levels are used, depending on the type of project.
For new construction projects not federally-assisted (meaning having a reduced-interest mortgage
or financed with tax-exempt bonds) the nominal percentage is 9 percent. For acquired property
or federally-assisted projects the nominal percentage is 4 percent. These nominal percentages are
adjusted monthly to maintain a discounted present value of 70 percent of the project basis for the
“9 percent” projects and 30 percent of the project basis for the “4 percent” projects, discounted at
an applicable federal rate related to Treasury borrowing rates. Recent percentages for the annual
amount of tax credit were 8.65 and 3.71 percent (Warren, Gorham & Lamont 1997, 813).
Because the tax credit is available over the ten-year allocation period and not as an up-front lump
sum, and because of the legal and accounting services that must be used to “syndicate” the tax
credits to interested buyer/partners, it has been noted above that the actual amount of money
delivered to a typical project for direct project costs is about half of the value of the credits
discounted to the present at the federal discount rate (roughly 50 cents on the present value of
federal costs). This happens because the investors exposed to the risk of recapture of the tax
credits require a much higher effective rate of return than the federal discount rate, and because
the syndication and transaction costs can amount to roughly one-fourth of the amount raised up
front from the investor/owners.

A survey conducted by the National Council of State Housing Agencies (NCSHA) (1995) of its
membership indicates 1994 allocations of tax credits sufficient to generate 2,136 projects with
117,099 tax credit-qualifying units. Seventy percent of the allocations were for new construction
units, with the remainder divided roughly equally equally between substantial rehabilitation and
acquisition and more modest rehabilitation. These planned numbers were increased by the pro-
jects with tax-exempt bond financing, which must fit under some overall limits on tax-exempt
financing but are not specifically limited by the per capita allocation of tax credits. Bond-financed
projects planned in 1994 were 37, representing 4,565 tax-credit qualified units (NCSHA 1995,
48). A separate survey conducted for HUD indicates that the total number of projects placed in
service in 1994 was 1,291 with 57,625 units11 (Abt Associates Inc. 1996).

The tax credit statute provides for a 15 percent set-aside for nonprofit sponsors, unless not
claimed by nonprofit applicants. Based on the NCSHA survey, nonprofit organizations were
allocated 27.3 percent of the tax credit units in 1994 (NCSHA 1995, 53), and the Abt Associates
survey indicates 26.5 percent of the units placed in service in 1994 were by nonprofit entities (Abt
Associates Inc. 1996, 4-11). The Abt Associates survey also obtained location data for as many
projects as possible. Based on the projects that could be geocoded (representing 55 percent of
the units), it appears that the projects sponsored by nonprofits were more likely than for-profit
projects to be in neighborhoods with a high incidence of poverty households, to have substantial
minority concentrations, and to be located in central cities (Abt Associates Inc. 1996), as shown

  These totals cover all tax credit allocating agencies except the City of Chicago and exclude 20 properties on
which no information was available on the number of units.

                                                        1996 Tri-Country Conference on Housing and Urban Issues
14   Michael J. Lea and James E. Wallace

in table 5. Most of the tax-credit projects are targeted to the narrow range of households with
incomes in the range of 50 to 60 percent of area median income (NCSHA 1995), as indicated in
table 6. Those projects intended to reach to the lower end of the income distribution are likely to
be those making use of several forms of assistance, both to reduce capital and carrying costs of
development, as well as rental subsidy (such as project-based Section 8 rental assistance).

                        Table 5. Neighborhood Characteristics by Nonprofit
                                      vs. For-Profit Sponsor

                                                    Nonprofit-Sponsored Proj-        For-Profit-Sponsored
                                                    ects (Percent)                   Projects (Percent)
  Neighborhood Characteristic                       (n = 23,774)                     (n = 68,682)

  Census Tracts with Over 50 Percent Low-                       70.2%                          61.9%
  Income Households (below 80% median)
  Census Tracts with Over 50 Percent Poor                       22.1%                           8.6%
  Households (below the poverty line)
  Census Tracts Where Percent Minority                          68.3%                          54.3%
  Exceeds MSA or County Percentage
  Census Tracts with Median Contract Rent                       72.2%                          55.4%
  at or Below the MSA or County Median
  Central City (n = 123,495)                                    64.5%                          48.0%

Source: Abt Associates Inc. 1996, Exhibit 4–5, p. 4–8 for Central City percentages and special calculations added
to Exhibit 4–11, p. 4–17 for tract characteristics.

             Table 6. Distribution of Credit Units by Household Income Categories

  Income Category (as percent of adjusted median)          Percentage of Units

  Below 30%                                                1.0%
  30 to 50%                                                18.7%
  51 to 60%                                                79.9%

Source: NCSHA 1995, Table 7, p. 51, state percentages weighted by state number of 1994 tax credit qualified
units allocated, excluding Georgia (not reporting).

1996 Tri-Country Conference on Housing and Urban Issues
                                     Current Practices for Financing Affordable Housing in the United States       15

Example. For a typical “9 percent” project, the LIHTC captures an effective equity contribution
to the project of about one-third 30 percent of total costs. The remainder of the costs are
financed in a wide variety of ways, including grants and subordinate loans in addition to the first
mortgage loan. A simplified example may show why the tax credit is not enough to provide the
financing required for affordability. In this example shown in table 7, we return to the typical
multifamily project introduced earlier and assume now that 100 percent of the project is rented to
tenants with income equal to 60 percent of median, and that, because of the subsidized nature of
the project, vacancies will only be 3 percent. The maximum market rate loan the project will
support is approximately $1.76 million, or 20.1 percent of the total development cost. Thus, the
developer must raise additional funds for 80 percent of project costs from equity contributions
and gap financing. First, look at the approximate size of the equity raised from the tax credit.
The tax credit is based on total development cost excluding land.12 If we assume that the $8.5
million in total development costs comprise $1.3 million of land costs and $7.2 million in
construction costs, and that the proportion of units occupied by low- and very low income tenants
is 100 percent, the undiscounted sum of tax credits available would be $6,480,000 ($7.2 million ×
0.09 × 10 years).

                           Table 7. Example of Financing a LIHTC Project

Assumptions                                   Inputs
Loan Term                                              240 months
Loan Rate                                          8.00% annual
Total Units                                            100
Project Development Cost                      $8,500,000 Total
Vacancy %                                          3.00%
Per Unit Annual Op Exp                           ($2,600) per unit
Median Income                                    $39,600

% of Total Units (Input)                               0%               0%               0%       100%         0%
Number of Units                                          0               0                 0         100           0
% of Median Income (Input)                          120%             100%              80%          60%        25%
Monthly Income                                    $3,960             $3,300          $2,640      $1,980      $825
25% of Monthly                                      $990              $825             $660        $495      $206
Utility Allowance                                   $120              $120             $120        $120      $120

   The tax credit basis is quite a bit more complicated. The owners are allowed to consider deferred developer’s
fees and other elements, making the basis effectively larger than simply the construction costs.

                                                        1996 Tri-Country Conference on Housing and Urban Issues
16   Michael J. Lea and James E. Wallace

Rent Income Available                           $870           $705           $540       $375      $86
Total Monthly Rent                                 $0            $0             $0     $37,500      $0
Total Annual Rent                                  $0            $0             $0 $450,000         $0

Gross Annual Rent                           $450,000
Vacancy                                     ($13,500)
Total Annual Operating Exp                 ($260,000)
Net Annual Rent Available                   $176,500

Max Loan Given Net Rent                    $1,758,444
Mortgage Payment                           ($176,500)
Loan as % of Development Cost                 20.69%
Financing Gap                              $6,741,556
Net Tax Credit Equity                      $3,500,000     41.18% of development cost
Remaining Financing Gap                    $3,241,556     38.13% of development cost

In order to raise cash for the development, the developer must syndicate or sell the tax credit.
Because the 1986 tax act restricts the ability of individuals to use tax losses (passive loss restric-
tions), the primary market for the credits is large tax-paying corporations. Tax credit investors
typically do not purchase the credits for their full face value, because they are future payments,
and there are also substantial transaction costs in marketing and selling the credits (Stegman
1991). The demand for tax credits has been rising recently, in part due to the fact that the pro-
gram was made permanent. Thus, developers are realizing a larger proportion of the credit in
cash today than in earlier years of the program. A GAO study found that average prices have
increased to over $0.60 per ten years’ worth of tax credits (GAO 1997, 90). If we assume a sale
at $0.55 per dollar (net of transactions costs), the developer will realize approximately $3.5
million, leaving a gap of $3.3 million. This simple example has omitted a number of devices by
which the amount of equity raised is increased, such as spreading out the required investor pay-
ments over a period of years or taking advantage of the depreciation allowances usable by a
corporate investor. The basic fact is illustrated, however. A substantial portion of tax credit
project development costs must be paid out of gap financing of some kind.

Concessionary Financing. A 1997 GAO study provided data from a national probability sample
423 tax credit developments from which owner responses were received on 380 projects on the
costs of tax credit projects and the role of tax credits in financing these low-income housing
developments. On the basis of the sample, the GAO estimated that tax credit projects placed in
service in the period 1992 through 1994 cost a total of about $10.7 billion to develop: about $5.8
billion in construction expenses; about $2.7 billion in construction-related fees, such as those paid

1996 Tri-Country Conference on Housing and Urban Issues
                                      Current Practices for Financing Affordable Housing in the United States   17

to developers and builders; and about $2.2 billion in other costs, including the costs of acquiring
the property (GAO, 1997, 75). The equity investment raised through the award of tax credits
($3.1 billion) amounts to about one-third of the total development costs of the projects sampled,
commercial mortgage loans about one-third, and concessionary financing in some form about
another third (GAO, 1997, 76).

Concessionary financing is necessary because, in a typical project, the restricted rents generate
only enough revenue over and above operating expenses to pay for the debt service on about one-
third of the costs of the project. It helps that the tax credit equity covers another third of project
costs, but that still leaves a large gap. The upshot is that developers must seek grants, donated
land or services, or loans at concessionary terms (low interest rates or deferred repayment) in
order to complete the financing of the project. The GAO estimated that 69 percent of the tax
credit projects placed in service between 1992 and 1994 required subsidies in addition to tax
credits, amounting to about $3 billion in concessionary loans or grants. These concessionary
loans or grants provided 37 percent of the financing for these projects.

Much of the concessionary financing (loans or grants) noted in the GAO study comes from
locally-administered federal block grant programs (Community Development Block Grant or the
HOME Investment Partnership Program) or from the interest-subsidized financing of the Section
515 program administered by the Rural Housing Service. Besides concessionary loans and grants,
tax credit projects may receive rental subsidies for the operating side that effectively lower the
capital financing required for the projects to meet the rental restrictions for low-income tenants.
The GAO estimated from their sample that tax projects placed in service between 1992 and 1994
received about $229 million an year in combined project-based and tenant-based rental assistance
payments. With this type of assistance considered, the percentage of tax credit projects in the
GAO study with assistance beyond tax credits increases to 86 percent (GAO, 1997, 87).

Limitations of the Low-Income Housing Tax Credit.13 As noted in the tax credit example, a basic
inefficiency lies at the heart of the tax credit. The present value to the government of the reve-
nues lost to the tax credits are only partially delivered to a project as funds to cover project costs.
This is partially because the investors discount the value of the tax credits at a higher rate than the
government because of the risk of the projects and the need for a competitive return on invest-
ment. Another component of loss is that the mechanism itself requires accountants, lawyers, and
partnership marketing experts to link the potential investors and a particular project, the so-called
syndication costs. The U.S. Congress has required state agencies that allocate the tax credits to
give priority to projects that use the highest percentage of housing credit dollars for project costs
other than the cost of intermediaries.14

     For a broader discussion of issues on the LIHTC, see Wallace 1998 (forthcoming).
     Revenue Reconciliation Act of 1989.

                                                        1996 Tri-Country Conference on Housing and Urban Issues
18   Michael J. Lea and James E. Wallace

To increase the efficiency of capture of capital by the LIHTC, that is to reduce the “syndication”
and transaction costs involved in collecting and disbursing the tax credit project investments, a
number of nonprofit tax credit equity funds have been created. These include the National Equity
Fund and other activities in support of nonprofit developers provided by the Local Initiatives
Support Corporation (LISC) and the equity fund of the Enterprise Foundation, the Enterprise
Social Investment Corporation. These equity funds undertake management of the syndication
process and provision of bridge loans to advance funds in anticipation of a series of spread-out
investor payments. These bridge loans serve to increase the amount of money actually delivered
to the project because the bridge loan interest rates are much lower than the effective rate of
return demanded by the investors, so that financing a spread-out series of payments increases the
total amount of money an investor is willing to put into a project. In addition, a number of equity
funds have been set up at the state level by state housing finance agencies or other nonprofit state
organizations (such as the Massachusetts Housing Investment Corporation), as well as several
city funds (such as the Chicago Equity Fund). The Equity Program of the Massachusetts Housing
Investment Corporation (1996) now reports that it is able to deliver 69 cents of equity funds to
project costs, net of fees and syndication costs, for every dollar of total tax-credit amount (ten-
year sum).

Another question about supply-side subsidies for direct production of affordable housing is
whether the housing units produced constitute a net addition to the stock. A study in progress at
the University of Wisconsin on LIHTC production suggests that over the period 1987 through
1994 tax credit units were substituting one-for-one for unsubsidized multifamily construction,
based on an econometric model of factors affecting housing supply at the state level (Malpezzi
and Vandell 1996). It can be argued, however, that these units would not have been supplied to
tenants in the target income range and that, unlike conventional development, these projects may
be located where they can contribute to neighborhood stability.

Housing Block Grants

Federal block grant programs (the HOME Investment Partnership program and the Community
Development Block Grant) provide funds that can be used for affordable housing. The HOME
Investment Partnership program was created in 1990 as part of the National Affordable Housing
Act. It provides federal housing block grants to “participating jurisdictions,” primarily states and
local governments, to undertake a wide range of housing activities, from construction and
rehabilitation of multifamily and single-family housing (including housing for special needs
populations) to use of the funds for tenant-based housing assistance. Federal funds must be
matched by state/local funds at one dollar of matching funds for four dollars of federal funds.
Activities are targeted to those with incomes below 80 percent of area median income.

Most states use the HOME funds in conjunction with the LIHTC program, and many use HOME
funds in conjunction with homeless programs (under the McKinney Act) and the state’s tax-

1996 Tri-Country Conference on Housing and Urban Issues
                                Current Practices for Financing Affordable Housing in the United States   19

exempt bond financing of single-family and multifamily building. In fiscal year 1994, $350 million
in HOME funds were distributed (NCSHA 1995, 86-89).

Tax-Exempt Bond Financing

The main interest rate subsidy program for both owner-occupied and rental housing is the use of
tax-exempt debt. State agencies are allowed to issue bonds for a variety of purposes on which the
interest is exempt from federal income taxes. These are tax-exempt issues sold to investors by
housing finance agencies, which then use the proceeds to fund below-market mortgages for first-
time home buyers or multifamily rental housing. The Mortgage Subsidy Bond Tax Act of 1980
(the Ullman Act) set annual caps on total bond volume for each state, imposed income targeting
requirements on the beneficiaries, and restricted single-family tax-exempt financing to assist first-
time home buyers. Of particular interest with regard to affordable housing are tax-exempt bonds
used to support home purchase and development of multifamily properties. For multifamily
housing, state housing finance agencies issue both taxable and tax-exempt bonds. Only the tax-
exempt bonds are subject to the annual state cap on tax-exempt issues. “New Money” issues
finance new programs or production, and therefore serve to increase the available housing stock.
“Refunding” issues are used to pay off and replace older issues at better interest rates (or under
more favorable terms) or are issued prior to the older bonds’ call date, with proceeds reinvested
until needed, but do not increase the housing stock.

Tax-Exempt Financing of Single-Family Housing. The primary source of funding for state
housing finance agency homeownership programs is the tax-exempt bonds for single-family
finance termed mortgage revenue bonds (MRBs). Housing finance agencies can also convert
MRB authority into mortgage credit certificates (MCCs) that, in lieu of tax-exempt financing,
provide first-time home buyers with a nonrefundable federal income tax credit for a specified
percentage of the annual interest paid on the conventional (not tax-exempt) mortgage of a
principal residence. Table 8 summarizes the 1994 activity for MRBs and MCCs.

Mortgage revenue bonds and mortgage credit certificates are required to be directed to first-time
home purchasers with incomes under 115 percent of area median income. Some state housing
finance agencies are able to direct these loans (credits) to lower income households. In 1994 the
national average of MRBs directed to households with incomes under 50 percent of area median
was 14 percent. For MCCs the percentage to households under 50 percent of median income was
7 percent. The distributions are shown in table 9. Note that interest rate subsidies are a costly
way to support affordable housing as they last for the entire life of the loan, and thus do not take
into account the potential increase in purchasing power of the household.

                                                  1996 Tri-Country Conference on Housing and Urban Issues
20   Michael J. Lea and James E. Wallace

                 Table 8. Mortgage Revenue Bond (Tax-Exempt Single-Family)
                           and Mortgage Credit Certificate Activity

 1994 MRB Issuance
    New Money Issues                                                                            $3.587 billion
    Refunding                                                                                      581 billion
    Issues                                                                                      $9.169 billion

 Home Loans Financed with MRB Issues
    Purchase, New                                                                                15,251 loans
    Purchase, Existing                                                                           74,037 loans
    Home Improvement                                                                              3,404 loans
 1994 MCC Authority Used                                                                        $0.615 billion
 MCCs for New Construction Purchase                                                                      1,987
 MCCs for Existing Home Purchase                                                                         9,333

Source: NCSHA 1995, 26, 29, 34.

               Table 9. Mortgage Revenue Bond and Mortgage Credit Certificate
                              Distribution by Income Category

  Mortgage Revenue Bonds
  Percentage of Applicable Median Income                                       Percentage Distribution

  Under 50%                                                                              14.1
  50+ to 60 %                                                                            15.1
  60+ to 80 %                                                                            34.1
  80+ to 100%                                                                            31.5
  Mortgage Credit Certificates
  Percentage of Median Income                                                  Percentage Distribution
  Under 50%                                                                              17.4
  50+ to 60 %                                                                            11.1
  60+ to 80%                                                                             44.0
  80+ to 100%                                                                            31.7

Source: NCSHA 1995, 29,36
Notes: Averages are computed for reporting states. Of 52 states and territories, five did not report income
distribution data for MRBs. Of 17 states reporting use of MCCs, two did not report income distributions. Among
the remaining 15, five reported income distributions that sum to less than 100 percent below median income. This
can easily occur because the income limit for MRBs is 115 percent of median income. The applicable median
income was defined as the greater of the statewide or area median income.

1996 Tri-Country Conference on Housing and Urban Issues
                                   Current Practices for Financing Affordable Housing in the United States   21

Multifamily Financing by State Housing Finance Agencies. In addition to tax-exempt and
taxable bonds issued for financing multifamily housing, state agencies also issue tax-exempt bonds
on behalf of charitable tax-exempt organizations that are referred to as 501(c)(3) bonds, named
for the section of the Internal Revenue Service code section on charitable organizations. Table 10
provides a summary on the 1994 dollar amount of multifamily bond issues and expected units.
“New Money” issues are those used for first-instance financing of multifamily projects, whereas,
as for the homeownership programs, “refunding” issues are used to pay off and replace older
issues at better interest rates (or under more favorable terms) or are issued prior to the older
bonds’ call date, with proceeds reinvested until needed, but do not increase the housing stock. In
1994 projects and units financed with multifamily bonds totaled 85 new construction projects with
6,025 units and 88 rehabilitation projects with 7,072 units.

                       Table 10. Dollar Amount of Multifamily Bond Issues
                                  for 1994 and Expected Units

  Type of Bond Issue              Dollar Amount ($ billion)             Expected Number of Units

  Tax-Exempt New Money                        .432                                     8,185
  Tax-Exempt Refunding                       1.387                                    22,260
  501(c)(3)                                   .133                                     3,642
  Taxable                                     .235                                     6,004
  Total                                      2.185                                    40,091

Source: NCSHA 1995, 68, 69.
Notes: Units under tax-exempt new money overlap with tax-credit units. Numbers may not add up due to

Except for LIHTC projects financed through state agency bonds, mechanisms for income-
targeting, the proceeds of multifamily bonds vary by state. For all 1994 financings combined,
table 11 shows the distribution of units financed by multifamily bonds across income categories.
From the data available it is not possible to determine the exact overlap of tax credit units
allocated in 1994 and the extent to which these are bond financed. It appears, however, that the
targeting requirements of the LIHTC affect a majority of the units funded by multifamily bonds.
For the reporting state agencies, 19 percent of the units are targeted to households with incomes
of 50 percent of area median income or less, whereas 66 percent are targeted to households with
incomes of 60 percent of area median income or less.

                                                     1996 Tri-Country Conference on Housing and Urban Issues
22      Michael J. Lea and James E. Wallace

                Table 11. Income Targeting of Units Financed by Multifamily Bonds

     Income Category                                                Units              Percentage of Total

     Up to 30 Percent of Median Income                               243                           2.2
     30+ to 50 Percent of Median Income                             1,794                         16.5
     50+ to 60 Percent of Median Income                             5,166                         47.4
     60+ to 80 Percent of Median Income                             1,165                         10.7
     80+ to 100 Percent of Median Income                             372                           3.4
     Over 100 Percent of Median Income                              2,165                         20.0
     Total                                                      10,095                           100.0

Source: NCSHA 1995, 72.
Notes: Nine agencies provided no information on income targeting.

In an effort to take advantage of the underwriting capabilities of some state agencies, an
experimental program was started in late 1993 to provide FHA insurance on mortgages issued by
the state agencies, which would share some of the risk of loss. As of 1994, 25 state agencies had
signed an agreement with HUD to participate, but only four states had actually financed projects
(285 units in five projects) (NCSHA 1995, 76).

Affordable Housing Program of the Federal Home Loan Bank Board

The Federal Home Loan Bank Act authorizes the Federal Home Loan Banks (FHLBs) to
contribute the greater of $100 million or 10 percent of the system’s previous year net income
annually to an Affordable Housing Program (AHP).15 This program provides subsidized loans
and other assistance to FHLB members (banks and thrifts) engaged in lending for the acquisition,
construction, and rehabilitation of long-term housing for very low, low-, and moderate-income
households. At the end of 1994, the 12 FHLBs had subsidy commitments exceeding $311 million
to 2,167 active or completed AHP projects, helping to finance the purchase, construction, or
rehabilitation of more than 82,000 housing units (both rental and owner-occupied) (Federal
Housing Finance Board 1995). In addition, the FHLBs also make loans at their cost of funds to
member institutions to finance housing for households with incomes at or below 115 percent of
area median income through their Community Investment Program (CIP). From 1990 through

   The FHLB are liquidity facilities that provide long-term loans to their members. The FHLBs had over $120
billion in loans (advances) to members at the end of 1995.

1996 Tri-Country Conference on Housing and Urban Issues
                                    Current Practices for Financing Affordable Housing in the United States   23

1994 the FHLBs made $7.1 billion in CIP advances for financing over 178,000 housing units and
246 community development projects.

The AHP subsidizes the interest rate on loans (advances) to financial institutions that are members
of the FHLB system. The program also provides direct subsidies to members for qualifying
projects. In both ways, AHP subsidies can fill part of the gap in the financing of affordable
housing. In 1994, 36 percent of AHP subsidies went for owner-occupied housing and 64 percent
were for rental units. Seventy-five percent of the units were for very-low-income households and
25 percent were for low- and moderate-income households.

The CIP is a discounted loan. Under CIP, the FHLBs reduce lending costs by providing advances
priced at the cost of funds plus administrative costs (typically 25 basis points). The discount,
itself a form of subsidy, is modest as normal advance spreads range between 25 and 40 basis
points over the FHLB cost of funds. CIP loans are used to fund mortgages that are typically held
in portfolio and not sold into the secondary market, and thus encourage more flexible
underwriting of loans made for purposes of complying with CRA. In 1994, 72 percent of the
loans supported through CIP were for owner-occupied units and 28 percent for rental units. Both
the AHP and CIP are frequently used in conjunction with the LIHTC program.

The AHP program was established in 1990. Like CRA and the housing goals, AHP can be
considered as a tax placed on the FHLBs—and thus their members—to fund affordable housing
programs in return for their federal backing. Unlike the CRA and housing goals, AHP specifically
mandates the funds to be used as subsidies. The combination of the AHP and the obligations of
the FHLBs to pay the interest on bonds issued to fund thrift resolutions in 1989 is roughly
equivalent to 34 percent of income, thus offsetting the federal income tax exclusion the banks

Insurance and Guarantees

One response to the higher credit risk of affordable lending is to find third parties to underwrite
such risk. Both private and government mortgage insurance is available for loans that meet the
guidelines of the provider. Mortgage insurance provides investors with the benefits of
specialization (all mortgage insurers are mono-line specialists) and nationwide diversification. In
the United States, private mortgage insurers provide “top slice” coverage, typically covering 25 to
30 percent of the mortgage balance. In addition, there are two major government mortgage
insurance programs: the FHA program for low- to moderate-income borrowers, and the
Department of Veterans’ Affairs (VA) program.16 In recent years, the market shares of insured
mortgages by FHA/VA and private companies have been similar (each accounting for between 10

   FHA loans are those insured by theFHA, a part of HUD. VA loans are those insured by the VA for military
veterans. FHA is required by law to operate on an actuarially-sound basis. There is no such requirement for VA.

                                                      1996 Tri-Country Conference on Housing and Urban Issues
24   Michael J. Lea and James E. Wallace

and 15 percent of single-family mortgage originations). Table 12 provides a summary of 1996
FHA and VA activity.

                  Table 12. Single Family Mortgage Originations by Loan Type

 Loan and Property Type                                           Mortgage Amount ($ billions)
 FHA 1-4 Unit(s)                                                               72.768
 VA 1-4 Unit(s)                                                                35.211
 Private Insurance 1-4 Unit(s)                                                126.972
 Not Insured 1-4 Unit(s) (includes rural Section 515)                         550.373

Source: 1996 data from Mortgage Insurance Companies of America and HUD Survey of Mortgage Lending
Activity, as reported in HUD 1997, 70.

Single-family Loans. Traditionally, the government insurance programs have been the major
bearers of credit risk for affordable mortgage loans. Households with little cash for down
payments typically obtain mortgages backed by the FHA or the VA instead of mortgages backed
by private insurers, because the agencies insured mortgages with lower down payments and used
more liberal underwriting guidelines when evaluating the creditworthiness of the applicant
(Canner and Passmore 1994). The FHA provides 100 percent insurance up to a congressionally-
set maximum loan limit (currently $152,362, adjusted to the Freddie Mac home price index,
varying by region). The 1990 National Affordable Housing Act allows FHA insurance on loans
(that can also cover various closing costs) for up to 97.75 percent of appraised value (98.75 for
appraised values under $50,000). It also provides for an up-front, partially-repayable premium
and a monthly premium for a period tied to the riskiness of the original loan. The VA programs
guarantee a portion of the loan amount up to a congressionally-established ceiling and are
available only to veterans. They will insure up to 100 percent of the value and have no up-front
fee. In 1993, FHA had a larger share of low- and moderate-income borrowers than conventional
lenders. Fifteen percent of the FHA’s loans were to low-income families, compared with just 6
percent for conventional lenders. The figures stand at 21 and 10 percent, respectively, for
moderate-income loans.

Multifamily Loans. In 1996 there were $47.1 billion of unassisted multifamily loan originations,
representing nearly 6 percent of total long-term mortgage originations (HUD 1997, 71). FHA-
insured loans represented only $2.57 billion of this total, even including refinancings. Most of the
new originations have been coming from commercial banks. Constraints on institutional
investment in multifamily housing and difficulties and creating a secondary market in multifamily
loans have limited the capital market funding of multifamily loans.

1996 Tri-Country Conference on Housing and Urban Issues
                                    Current Practices for Financing Affordable Housing in the United States   25

Regulatory Influence on the Supply of Mortgage Finance

Efforts to increase the flow of mortgage finance to affordable housing have focused on creating
more flexible underwriting guidelines. The impetus for change has come from two sources: the
Community Reinvestment Act (CRA)17 and the affordable housing goals established for Freddie
Mac and Fannie Mae, the government-sponsored secondary mortgage market organizations, or
“government-sponsored enterprises” (GSEs).

There are three possible rationales for CRA (Canner and Passmore 1995) and the GSE affordable
housing goals. The first is that there are inefficiencies in the primary mortgage market due to the
high information costs of lending (particularly affordable housing loans) that lead lenders to under
invest. CRA in this view encourages the pooling of information about borrowers and
neighborhoods that overcome the information externalities. Purchases of loans made under CRA
by the GSEs allow for an efficient geographic diversification of risk. The second rationale is that
there is pervasive discrimination against minorities, and lenders do not extend credit in low-
income neighborhoods because of the high proportions of minority residents in these
neighborhoods. CRA and the housing goals are at best an indirect mechanism for dealing with
discrimination issues. The third rationale is that insured depositories and the GSEs benefit from
government support and that loans made under CRA are a condition for such support. Under this
view, CRA and the goals are a tax, in that the return on such loans is expected to be less than that
under non-CRA loans. (See the recent study by the Congressional Budget Office [1996] on
various assessments of the costs and benefits of the GSEs.)

Effects of GSE Goals on the Secondary Market

The housing goals of the GSEs (Fannie Mae and Freddie Mac) are another impetus for change.
The Federal Housing Enterprises Financial Safety and Soundness Act of 1992 established three
housing goals to target a portion of the GSEs’ mortgage purchases to housing for low- and
moderate-income households and housing located in underserved areas. The specific goals are:

•    Special Affordable Goals: Targets purchase of loans secured by units affordable to low- and
     very-low-income families. The Special Affordable goal is divided into two subgoals: (1)
     Purchases of mortgages for owner-occupied units located in low-income areas and owned by
     low-income families OR units located anywhere and owned by very-low-income families
     (defined for this purpose as under 60 percent of area median income). 1996 Goal: 6 percent

   The Community Reinvestment Act was enacted out of concern that some lenders were engaging in “redlining,”
having a policy of issuing no loans in certain neighborhoods, or simply that some lending institutions were not
making loans in areas from which they derived their business (checking accounts, savings deposits, etc.).
Compliance with CRA requirements is now part of the regular exams of lending institutions conducted by the
federal regulatory agencies. Many of the affordable housing loan consortia carry the term “community
reinvestment” or “community investment” in their corporate names.

                                                      1996 Tri-Country Conference on Housing and Urban Issues
26    Michael J. Lea and James E. Wallace

     of mortgage purchases by each GSE. (2) Purchases of mortgages for rental units located
     anywhere and affordable to very-low-income families, meaning rents requiring no more than
     30 percent of the specified income limit. 1996 Goal: 6 percent of mortgage purchases by
     each GSE.

•    Low and Moderate Income Goal: Targets loans affordable to low- and moderate-income
     families. 1996 Goal: 40 percent of mortgage purchases by each GSE.

•    Underserved Areas Goal: Targets housing for families living in areas not well served by the
     market. 1996 Goal: 21 percent of mortgage purchases by each GSE.

Since these regulations were passed, the GSEs have improved their performance relative to the
housing goals. As noted by HUD (1996b), however, their purchases of low-income loans
continue to lag those of other mortgage market participants, including portfolio lenders. This
result is not surprising, as the GSEs are secondary market institutions and do not deal directly
with borrowers. In order to develop a liquid secondary market, the GSEs have developed
standardized guidelines for documentation and underwriting. One of the criticisms of the
secondary mortgage market is that the pursuit of increased efficiency has overemphasized
standardization at the cost of excluding more flexible underwriting guidelines that are used for
affordable housing loans (Roche 1994).

CRA and the housing goals legislation have led primary market lenders and the GSEs to review
and liberalize their mortgage underwriting guidelines. As reflected in Fannie Mae’s Community
Home Buyer Program and Freddie Mac’s Affordable Gold Program, changes have been made in
four major underwriting areas:

•    Down payments. Allowing 3 percent down payment. Both Fannie Mae and Freddie Mac
     offer to purchase 95 percent LTV mortgages in which only 3 percent of the purchase price is
     paid out of the borrower’s own funds, with an additional 2 percent coming from a gift, grant,
     or other source (“3/2” mortgages). Fannie Mae also offers a 97 percent LTV loan purchase

•    Credit histories. Rather than relying on traditional credit reports, the industry now allows for
     alternative forms of establishing creditworthiness, such as the use of rent, telephone, and
     utility payment receipts.

•    Debt ratios. The GSEs and many lenders have expanded traditional debt ratio limits from
     28/36 to 33/38. The first number reflects the maximum housing debt-to-total income ratio
     and the second number sets the maximum total debt-to-total income ratio.

1996 Tri-Country Conference on Housing and Urban Issues
                                   Current Practices for Financing Affordable Housing in the United States   27

•    Reserve requirements. Many lenders have been waiving the requirement that borrowers have
     two months’ worth of principal, interest, taxes, and insurance payments on hand after closing.

As reported by Steinbach (1995), the early results for these programs are not encouraging. Loans
with only 3 percent of a borrower’s own funds invested were experiencing a default rate twice as
high as loans with 5 percent of a borrower’s own funds invested. Loans to borrowers with no
credit history were experiencing a delinquency rate eight times higher than loans made to
borrowers with an established credit and repayment history. Loans to borrowers with housing
and total debt rates higher than 33 and 38 percent were experiencing a 60 percent higher
delinquency rate when compared with loans to borrowers whose debt ratios were below 33 and
38 percent. Loans to borrowers with less than two months of payments in reserve were
experiencing a 40 percent higher delinquency rate than loans to borrowers with reserves in excess
of two months. In light of these results, Steinbach suggests that more proactive borrower
counseling and education programs are needed.

Community Reinvestment Act. The Community Reinvestment Act was enacted in 1977 out of
concern that some lenders were engaging in “redlining,” having a policy of issuing no loans in
certain neighborhoods, or simply that some lending institutions were not making loans in areas
from which they derived their business (checking accounts, savings deposits, etc.). Compliance
with CRA requirements is now part of the regular exams of lending institutions conducted by the
federal regulatory agencies. The CRA is intended to encourage insured commercial banks and
savings associations to help meet the credit needs of the local communities in which they are
chartered.18 CRA is directed at federal supervisory agencies and calls upon them to (1) use their
supervisory authority to encourage financial institutions to help meet local credit needs in a
manner consistent with safe and sound operation, (2) assess an institution’s record of meeting the
credit needs of its entire community, including low- and moderate-income neighborhoods, and (3)
consider the institution’s CRA performance when assessing an application for a charter, deposit
insurance, branch or other deposit facility, office relocation, merger, or acquisition. To enforce
CRA, the regulatory agencies conduct examinations of institutions and evaluate their
performance. When the Financial Institutions Reform, Recovery and Enforcement Act of 1989
(FIRREA) strengthened CRA, regulatory pressure on the lending industry increased. CRA grades
became public, giving community groups greater ability to discuss lending commitments from
banks and thrifts (Steinbach 1995).

One of the principal ways the agencies monitor CRA performance is through publication of loan
volume and rejection rates required by the Home Mortgage Disclosure Act (HMDA). Table 13

   For an in-depth discussion of CRA see Canner and Passmore (1995). CRA only covers banks and thrifts. There
have been proposals in recent years to extend CRA to mortgage banks. Curiously, CRA does not apply to
federally-insured credit unions, which have been the fastest growing intermediaries in recent years.

                                                     1996 Tri-Country Conference on Housing and Urban Issues
28     Michael J. Lea and James E. Wallace

shows the distribution of home purchase loans reported under HMDA, grouped by census-tract
income and distributed by borrower income group in 1993.

Lenders covered by HMDA reported that they originated 2,424,570 home purchase loans in 1993
in an aggregate amount of $269.2 billion.19 Lenders reported the borrowers’ incomes for
2,383,205 of these loans, the distributions of which are reported in table 13. Low income is
defined as household or median family income less than 50 percent of MSA median family
income, moderate income as 50 to 80 percent of MSA median, middle income 80 to 120 percent
of MSA median, and upper income over 120 percent of MSA median income. Low-income
households received 6.6 percent of the loans and moderate-income households received 20.4
percent of the loans extended in 1993. The characteristics of these loans are shown in table 14.
The high LTV for low-income census tracts reflects the widespread use of government-insured
mortgage programs (discussed below) in these neighborhoods.

                   Table 13. Home Purchase Loans Reported Under HMDA, 1993

 Borrower                                  Census Tract Income Group (% of total)
 Income Group              Low          Moderate     Middle       Upper        Number            Percent

 Low                        3.8%             24.7%        55.0%           16.5%      156,299       6.6%
 Moderate                   1.6              15.8         61.0            21.6       487,045      20.4
 Middle                     0.9               9.2         56.7            33.2       720,958      30.3
 Upper                      0.5               4.8         39.1            55.5     1,018,903      42.8
 All                        1.1               9.7         50.0            39.3     2,383,205     100.0

Source: Federal Reserve Bulletin.

              Table 14. Median Values of Selected Characteristics of Home Purchase
               Loans Reported Under HMDA, by Census Tract Income Group, 1993

 Characteristic                           Low         Moderate          Middle    Upper         All
 Income relative to MSA (%)               75.9          79.8             98.7     136.6        109.2
 Loan-income ratio (%)                   188.5         191.4            196.8     200.0        197.6
 LTV ratio (%)                            98.2          89.5             86.1      78.1        832.0
 Income ($ thousand)                       33           35               43         58           47
 Loan amount ($ thousand)                  59           63               81        113           90

Source: Federal Reserve Bulletin.

     Home refinancing accounted for 6,439,021 mortgage loans in 1993.

1996 Tri-Country Conference on Housing and Urban Issues
                                     Current Practices for Financing Affordable Housing in the United States   29

Loan Consortia.20 The Community Reinvestment Act has also motivated the formation of loan
consortia. Loan consortia are a vehicle for lending institutions to pool resources and share the
risk of loans they might not make individually. Participants in a consortium can range from the
largest banks in a state to small community lenders. Most consortia are structured as tax-exempt
organizations under 501(c) of the Internal Revenue Code. Participating lenders typically create a
loan pool, structured as a commitment to lend as loans are approved. Among the more formal
consortia, loans are structured as blind pools, with all members participating in all loans. Formal
consortia also lend almost exclusively to multifamily projects, although they may occasionally
provide the construction loan for an affordable single-family development. Consortia may provide
permanent financing as well as construction loans.

Among the initial reasons for lenders joining consortia are the community lending requirements of
the CRA, although lenders are not interested in joining a consortium solely to meet CRA
requirements. Lenders need to see the loan pool as a way of entering and learning about market
niches that they might not have the resources or experience to enter on their own.

Growth in the numbers of these consortia has been rapid in recent years. Perhaps the earliest of
these affordable housing loan consortia was the Savings Association Mortgage Corporation, Inc.
(SAMCO), started in California in 1970. In 1991 there were about 15 affordable housing loan
consortia. By mid-1995 that number had more than doubled, to 33 affordable housing loan
consortia, including 22 state-level consortia.

Based on the compilation of consortia by the National Affordable Housing Lenders Association
(Schon 1995), one can estimate the national size of their activities. For 30 reporting entities, the
1994 total loan volume was $444 million, distributed over approximately 770 single-family loans
and 405 multifamily projects (number of units not known). No comprehensive data are available
on the extent to which these loans are used in conjunction with other aids to affordable housing.
The Schon sourcebook indicates that most consortia have a policy of working in conjunction with
practically all of the other aids (LIHTC, Community Development Block Grants, the HOME
program, affordable housing programs of the Federal Deposit Insurance Corporation and FHLB,
state housing trust funds, and federal Section 8 rental assistance). The president of the
Community Investment Corporation in California estimated that 40 to 45 percent of the dollar
volume of their loans goes to projects using the LIHTC.

In a telephone interview, the president and CEO of the Massachusetts Housing Investment Fund
(MHIC) estimated that their loan volume is roughly equally divided between tax credit and non-
tax credit projects, that about 60 percent of the tax credit projects are with nonprofit sponsors,
and that MHIC is involved in about half of the state’s tax credit developments. He also indicated

  Material in this section is from the sourcebook on loan consortia compiled by the National Association of
Affordable Housing Lenders (Schon 1995).

                                                       1996 Tri-Country Conference on Housing and Urban Issues
30    Michael J. Lea and James E. Wallace

that affordable housing is benefitting, at least in the short run, from the flurry of bank acquisitions
and mergers. Massachusetts passed an interstate banking law in the late 1980s that requires that a
bank acquiring an out-of-state bank set aside 90 basis points (nearly 1 percent) of the assets of the
acquired bank for affordable housing projects administered by the Massachusetts Housing
Partnership (MHP). This has provided MHP with a substantial source of funds with which to do
permanent financing of affordable projects.

Housing Trust Funds21

As of 1993, 37 states had established state housing trust funds to assist in the development and
finance of affordable housing. Most of these have grown up in the wake of the federal withdrawal
from finance of affordable housing production starting in the early 1980s. Only one state housing
trust fund existed before 1982 (the Delaware Housing Development Fund, established in 1968).
Even the equity capital captured by the LIHTC program often does not provide enough equity
that the full remaining capital costs can be serviced with conventional debt. Housing trust funds
are one of a variety of mechanisms that have been established, in part, to fill this gap. Only three
state housing trust funds have no ties to a state housing finance agency or state government
agency (Michigan, Rhode Island, and Vermont). Research conducted for the NCSHA (Petherick
1993, 6) indicates that state housing trust funds collectively have provided more than $780 million
in financial assistance and have assisted more than 80,000 housing units. Not counting new trust
funds authorized since July 1991, state housing trust funds provide annual assistance of more than
$100 million (Petherick 1993, 28).

Sources of funds for the state housing trusts include:

•    Dedicated real estate revenues (such as or interest earnings on escrowed funds provided by
     buyers or sellers of real estate until the transaction closes, or from real estate transfer fees and

•    Other dedicated revenues (unclaimed property, voluntary check-off on state income tax
     returns, excess balances in tax-exempt bond reserve debt service accounts)22

•    State appropriations

•    Loan repayments

   The description in this section is summarized from Petherick 1993. This volume includes appendix material
providing detailed profiles of each of the state housing trust funds.
   The use of excess balances in state housing finance agency accounts to fund state housing trust funds is probably
reflected in the numbers cited above on agency funds other than bonds or tax credit allocations used to finance
multifamily housing.

1996 Tri-Country Conference on Housing and Urban Issues
                                Current Practices for Financing Affordable Housing in the United States   31

Real estate sources provided the substantial source of funds for at least 16 of the state housing
trust funds. State appropriations can vary from year to year, but of the 37 state housing trust
funds, nearly half have received state appropriations at some time. Annual funding levels vary
widely, from zero to as much as $32.5 million (Florida’s Housing Trust Fund for FY 92-93).
Most state housing trust funds receive $4 million or less per year.

State housing trust funds vary widely in how they are targeted, types of housing financed, and
eligible housing activities. As characterized by Petherick (1993, 20), most state housing trust

•   Address the most pressing housing needs in their state

•   Finance a wide range of physical and legal types (homeownership, rental, cooperative) of
    affordable housing

•   Target assistance to very-low-income (under 50 percent of local median) and low-income
    households (under 80 percent of median income)

•   Often combine their assistance with other federal support from mortgage revenue bonds, the
    LIHTC, the HOME program, or the McKinney homeless program

•   Attempt to expand the capacity of nonprofit organizations to develop, preserve, and operate
    affordable housing

•   Provide financial assistance mainly in the form of deferred payment (not amortized), low-
    interest loans

Trust funds with more limited scope include those limiting activity to assistance for rental housing
(Hawaii and Missouri), or focusing primarily on housing to serve the homeless (Georgia,
Nebraska, and Utah).


Unaided private provision of affordable housing is very limited because of its basic economics.
Efforts to encourage private sources of financing through marginal incentives such as reduced
interest rates on mortgages are, perforce, limited to marginal changes that primarily benefit
moderate-income households. In areas where public intervention and subsidy are needed to
create affordable housing, the tools presently available typically reach only to the range of
households at 50 to 60 percent of median income. Direct federal support for affordable housing is
essentially confined to the low-income housing tax credit. Even the tax credit program requires

                                                  1996 Tri-Country Conference on Housing and Urban Issues
32   Michael J. Lea and James E. Wallace

substantial gap financing (roughly one-third of development costs) even to reach its 50 to 60
percent of median income target. To reach lower-income levels the tax credit program requires
additional subsidy, such as the Section 8 rental assistance program. The HOME program is used
for housing, although these funds and those of the Community Development Block Grant
program often are used to help bridge the affordability gap for tax credit projects to reach their 50
to 60 percent of median income requirement.

Other federal programs play a role in affordability. Multifamily FHA insurance plays a minor role,
but does provide some implicit assistance toward affordability. Tax-exempt bond financing
provides some additional assistance toward affordability. Other federal efforts at encouraging
financing of affordable housing take the shape of requirements that can be viewed as implicit
taxes. These include the community lending requirements of the CRA and loan consortia formed
largely in response to it, the Affordable Housing Program of on-lending to member banks by the
Federal Home Loan Bank Board, and the housing goal requirements of the government sponsored
secondary mortgage market enterprises (Fannie Mae and Freddie Mac).

At the state and local levels, the primary independent sources of finance for affordable housing are
the various housing trust funds. These often have sources of funding independent of the federal

Finally, we observe that the diffusion of the incentives and the use of implicit taxes to encourage
development of affordable housing make it difficult to assess the incidence of provision or the
efficiency with which it is provided.

1996 Tri-Country Conference on Housing and Urban Issues
                                   Current Practices for Financing Affordable Housing in the United States   33


Michael J. Lea is President of Cardiff Consulting Services, Cardiff, California. James E. Wallace
is Vice President of Abt Associates Inc., Cambridge, Massachusetts.


Abt Associates Inc. 1996. Development and Analysis of the National Low-Income Housing Tax Credit Database
(Research in Progress for the U.S. Department of Housing and Urban Development). Cambridge, MA.

Advisory Commission on Regulatory Barriers to Affordable Housing. 1991. “Not in My Back Yard.” U.S.
Department of Housing and Urban Development, July.

American Council of Life Insurance. 1996. Investment Bulletin, March.

Bradley, Don, and Peter Zorn. 1996. Fear of Homebuying. Secondary Mortgage Markets 13(2): 1, 24–29.

Canner, Glenn, and Wayne Passmore. 1994. Private Mortgage Insurance. Federal Reserve Bulletin October:

Canner, Glenn, and Wayne Passmore. 1995. Home Purchase Lending in Low-Income Neighborhoods and to Low-
Income Borrowers. Federal Reserve Bulletin, February: 71–103.

Case, Karl E. 1991. Investors, Developers, and Supply-Side Subsidies: How Much Is Enough? Housing Policy
Debate 2(2): 341–56.

Congressional Budget Office. 1996. Assessing the Public Costs and Benefits of Fannie Mae and Freddie Mac.
Washington, DC: Congressional Budget Office, May.

Cotterman, Robert, and James Pearce. 1996. The Effects of FNMA and FHLMC on Conventioanl Fixed Rate
Mortgage Yields. HUD Studies May: 97–168.

Danter, Kenneth. 1995. Inside the Apartment Market. Mortgage Banker July: 54-62.

DiPasquale, Denise, and Jean Cummings. 1992. Financing Multifamily Rental Housing: The Changing Role of
Lenders and Investors. Housing Policy Debate 3(1): 77–116.

Engelhardt, Gary. 1994. Do Targeted Savings Incentives for Homeownership Work: The Canadian Experience.
Paper presented at the Fannie Mae Roundtable on Understanding Household Savings for Homeownership,

Fannie Mae. 1995. Annual Report.

Federal Housing Finance Board. 1995. Report on the Low Income Housing and Community Development
Activities of the Federal Home Loan Bank System.

                                                     1996 Tri-Country Conference on Housing and Urban Issues
34   Michael J. Lea and James E. Wallace

Ferguson, B., J. Rubinstein, and V. Dominguez Vial. 1996. The Design of Direct Demand Subsidy programs for
Housing in Latin America. Review of Urban and Rural Development Studies 8.

Freddie Mac. 1995. Annual Report.

Harvard Joint Center for Housing Studies. 1995. The State of the Nation’s Housing.

Hebert, Scott, Kathleen Heintz, Chris Baron, Nancy Kay, and James E. Wallace. 1993. Nonprofit Housing: Costs
and Funding. HUD-1435-PDR. Washington, DC: U.S. Department of Housing and Urban Development.

Hendershott, Patric. 1990. The Tax Reform Act of 1986 and Real Estate. In D. DiPasquale and L. Keyes, eds.
Building Foundations 241–62. University of Pennsylvania Press.

Lea, Michael, and Bertrand Renaud. 1994. Contract Savings for Housing: Relevancy for U.S. Housing Finance.
Paper presented at the Fannie Mae Roundtable on Understanding Household Savings for Homeownership,

Malpezzi, Stephen and Kerry Vandell. “Evaluation of the Low Income Housing Tax Credit: Interim Report.”
University of Wisconsin-Madison, School of Business. Working Draft, January 24, 1996.

Massachusetts Housing Investment Corporation. 1996. 1996 Annual Report. Boston, MA.

Miles, Mike E., Richard L. Haney, Jr., and Gayle L. Berens. 1996. Real Estate Development: Principles and
Process, 2nd edition. Urban Land Institute.

National Council of State Housing Agencies. 1995. State HFA Factbook: 1994 NCSHA Annual Survey Results.
Washington, DC.

Nelson, Kathryn P. 1994. Whose Shortage of Affordable Housing? Housing Policy Debate 5(4): 401–42.

Nothaft, Frank. 1994. Higher Costs of Commercial Financing. Secondary Mortgage Markets, Mortgage Market
Review, 20.

Petherick, Glenn D. 1993. State Housing Trust Funds: Innovative Sources for Financing Affordable Housing.
Washington, DC: National Council of State Housing Agencies.

President’s Commission on Affordable Housing. 1993. Report of the President’s Commission on Affordable
Housing. Washington, DC.

Roche, Ellen. 1994. Challenge and Change in the U.S. Mortgage Market in the 1990s. Housing Finance
International 8(3): 28–34.

Schon, David, ed. 1995. Affordable Housing Loan Consortia Sourcebook. National Association of Affordable
Housing Lenders.

Steinbach, Gordon. 1995. Ready to Make the Grade. Mortgage Banker 55(9): 36–43.

Stegman, Michael. 1991. The Excessive Cost of Creative Finance: Growing Inefficiencies in the Production of
Low Income Housing. Housing Policy Debate 2(2): 357–73.

1996 Tri-Country Conference on Housing and Urban Issues
                                   Current Practices for Financing Affordable Housing in the United States   35

Theobald, H. Scott, and Michele Walczak. 1995. Multifamily Moves Ahead. Secondary Mortgage Markets
Mortgage Market Review, 11–15.

U.S. Department of Housing and Urban Development. 1996a. Rental Housing Assistance at a Cross-roads: A
Report to Congress on Worst Case Housing Needs. Office of Policy Development and Research.

U.S. Department of Housing and Urban Development. 1996b. Statement by Nicolas Retsinas, Assistant Secretary
for Housing and Federal Housing Commissioner before the Senate Subcommittee on HUD Oversight and Structure,
Committee on Banking, Housing and Urban Affairs. Washington DC, March 5.

U.S. Department of Housing and Urban Development. 1996c. U.S. Housing Market Conditions, First Quarter
1996, May.

U.S. Department of Housing and Urban Development. 1997. U.S. Housing Market Conditions, 2nd Quarter 1997,

U.S. General Accounting Office. 1997. “Tax Credits: Opportunities to Improve Oversight of the Low-Income
Housing Program” (GAO/CCD/RCED-97-55). Report to the Chairman, Committee on Ways and Means; and the
Chairman, Subcommittee on Oversight, Committee on Ways and Means, House of Representatives, March.

Vandell, Kerry, Walter Barnes, David Hartzell, Dennis Kraft, and William Wendt. 1993. Commercial Mortgage
Defaults: Proportional Hazards Estimation Using Individual Loan Histories. AREUEA Journal 21(4): 451–80.

Van Order, Robert. 1990. The Hazards of Default. Secondary Mortgage Markets, Fall: 29-31.

Van Order, Robert, and Ann Schnare. 1994. Finding Common Ground. Secondary Mortgage Markets 11(1):

Wallace, James E. 1995. Financing Affordable Housing in the United States. Housing Policy Debate 6(4):

Wallace, James E. 1998 (forthcoming). “Chapter 3: Evaluating the Low-Income Housing Tax Credit” in
Evaluating Tax Expenditures. San Francisco: Jossey-Bass Inc.

Warren, Gorham, and Lamont (publ.). 1997. Tax Credit Briefs. Housing and Development Reporter 24: 813.

                                                     1996 Tri-Country Conference on Housing and Urban Issues

To top