Fundamental Tax Reform: Options for the Mortgage Interest Deduction
Prepared for Members and Committees of Congress
The mortgage interest deduction, which is one of the largest sources of federal tax revenue loss
with an estimated annual cost of $72 billion, is intended to encourage homeownership. Empirical
studies suggest that the mortgage interest deduction subsidizes mortgage lending, which has more
impact on housing consumption than homeownership rates. Other homeownership subsidies, like
down-payment assistance programs, are proven to be more effective at increasing homeownership
among lower-income families and are less expensive than the mortgage interest deduction.
A recent enhancement to the mortgage interest deduction was enacted in the Tax Relief and
Health Care Act of 2006 (P.L. 109-432),which temporarily allowed, for tax year 2007, mortgage
insurance premiums paid for a personal residence to be tax deductible as mortgage interest. The
Mortgage Forgiveness Debt Relief Act of 2007 (H.R. 3648; P.L. 110-142) extended that provision
through the end of 2010.
Early in the 109th Congress, tax reform was a major legislative issue. President Bush appointed a
bipartisan panel to study the federal tax code and to propose options to reform the code. The
panel produced a report in the fall of 2005 that included a proposal to change the mortgage
interest deduction. Legislation for fundamental tax reform was introduced, and some bills
proposed to change the tax base such that income tax credits and deductions, like the mortgage th
interest deduction, would have been eliminated. Late in the first session of the 110 Congress,
House Ways and Means Committee Chairman Rangel indicated that fundamental tax reform th
might still be considered in the 110 Congress.
There are many possible options to alter the mortgage interest deduction. If the deduction were
eliminated as a result of fundamental tax reform, the resulting effects would depend on a variety
of variables, including the nature of tax reform, the resulting changes in the tax base and tax rates,
changes in interest rates, and other economic variables. If the deduction were eliminated without
other tax policy changes, federal income tax revenues would increase, the tax base could be
broadened, and the federal budget deficit could be reduced.
Modifications to the mortgage interest deduction could take any one of several approaches.
Congress could choose to allow the deduction for only one residence and/or to reduce the
allowable principal debt, which is currently $1 million. These changes would reduce the amount
of tax revenue loss associated with the provision. Congress could choose to improve equity by
allowing more low-income households to claim mortgage interest, either as an above-the-line
deduction or as a tax credit. Finally, the mortgage interest deduction could remain unaltered.
This report, which will be updated to reflect legislative developments, describes the deduction for
home mortgage interest, provides economic analysis, and concludes with possible changes to the
mortgage interest deduction and the potential impacts of those changes.
Legislative Developments in the 110th Congress.............................................................................2
Fundamental Tax Reform and Legislative Developments in the 109th Congress............................2
An Overview of the Deduction for Mortgage Interest....................................................................3
History of the Deduction...........................................................................................................4
Economic Analysis of the Deduction..............................................................................................5
Efficiency: Treatment of Mortgage Interest Under an Ideal Income Tax..................................5
Imputed Rental Income for Homeowners...........................................................................6
Debt and Equity Finance Neutrality....................................................................................7
Empirical Estimates of Efficiency Loss Compared to an Ideal Tax System.......................7
Efficiency: Externalities Associated With Homeownership......................................................8
Efficiency: Empirical Estimates of the Deduction’s Effect on Home Ownership...................10
Low Rate of Return............................................................................................................11
Equity ...................................................................................................................................... 12
Itemizers vs. Non-Itemizers..............................................................................................13
Data on Deductions Claimed............................................................................................13
Simplicit y ..................................................................................................................... ........... 15
Possible Policy Changes................................................................................................................16
Elimination of the Mortgage Interest Deduction.....................................................................16
Elimination as a Result of Tax Reform.............................................................................16
Elimination Not as a Result Tax Reform..........................................................................19
Modification of the Mortgage Interest Deduction...................................................................21
Reduce the Amount of Allowable Indebtedness...............................................................21
Disallow Home Equity Indebtedness................................................................................22
Allow the Deduction for Principal Residences Only........................................................22
Conclusion ..................................................................................................................................... 25
Table 1. Deductions of Mortgage Interest, Claimed in 2001.........................................................14
Table 2. Distribution of Tax Expenditure by Income Class...........................................................15
Author Contact Information..........................................................................................................26
n the 109th Congress, tax reform was a major legislative issue. In January 2005, President
Bush appointed a nine-member bipartisan panel to study the federal tax code and to propose
options to reform the code. Proposals for fundamental tax reform would have changed the tax I
base from income to consumption or changed the way income is taxed. In a pure consumption tax
system, income tax credits and deductions, like the mortgage interest deduction, would be
eliminated. Other proposals for tax reform could have simplified the existing income tax system.
In these cases, modifications to the mortgage interest deduction could have simplified the tax
It was unclear whether changes to the mortgage interest deduction would occur given the strength
of support that exists for the provision. In 2005, President Bush indicated that homeownership tax
incentives are important and should be preserved as tax reform is being considered. Specifically,
in the executive order that established the President’s Advisory Panel on Federal Tax Reform, one
of the very few restrictions was a request that the panel “recognize the importance of 1
homeownership and charity in American society.” Some analysts concluded that the
Administration’s statement indicated its support for preserving the mortgage interest deduction
along with all of the other homeownership tax incentives. In one tax journal it was written “in
plain English, the mother of all tax subsidies, the mortgage interest deduction, shall remain 2
There are three principal tax provisions for owner-occupied housing and one implicit tax benefit.
The deduction for mortgage interest is the most costly provision, with an estimate of $73.7 billion 3
in revenue loss for FY2007. The exclusion of capital gains on the sales of homes is the second
largest tax provision for homeowners, with an estimate of $28.5 billion in tax revenue loss for 4
FY2007. The deduction of state and local real estate taxes is the third provision, with an estimate 5
of $16.8 billion in tax revenue loss for FY2007. The exclusion of net imputed rental income
from taxation, which is not a provision in the tax code but exists implicitly, is estimated to cost 6
$33.2 billion in uncollected tax revenue for FY2007. This would bring the total projected
revenue cost for owner-occupied housing in FY2007 to $152 billion.
The amount of revenue foregone by the federal government to promote homeownership greatly
exceeds government spending on housing. Appropriations made in the Department of Housing
and Urban Development (HUD) budget totaled $31.5 billion in FY2005. Most of the
appropriations for HUD address rental housing problems, most of which are faced by households
with very low incomes or other special housing needs.
1 Executive Order 13369, “President’s Advisory Panel on Federal Tax Reform,” website
http://www.taxreformpanel.gov/executive-order.shtml, visited July 22, 2005.
2 Martin A. Sullivan, “The Economics of the American Dream,” Tax Notes, Jan. 24, 2005, p. 407.
3 U.S. Congress, Joint Committee on Taxation, Estimates of Federal Tax Expenditures for Fiscal Years 2007 to 2011,
JCS-1-05 (Washington: GPO, 2007), p. 27.
6 U.S. Office of Management and Budget, Analytical Perspectives, Budget of the United States Government, Fiscal
Year 2007, Washington, DC: 2007), p. 288.
Although fundamental tax reform has not been on the legislative agenda in the 110th Congress,
changes to the mortgage interest deduction have been enacted. The Tax Relief and Health Care
Act of 2006 (P.L. 109-432) temporarily allowed, for tax year 2007, mortgage insurance premiums
paid for a personal residence to be tax deductible as mortgage interest. The Mortgage Forgiveness
Debt Relief Act of 2007 (H.R. 3648; P.L. 110-142) extended that temporary provision through the
end of 2010.
Mortgage insurance, which guarantees loan repayment in case of death or disability of the
borrower, is often required by lenders for individuals who do not have sufficient funds for a full
down payment on a residence. Premiums paid or accrued for qualified insurance on mortgage
loans can be treated as qualified residence interest and deducted from income tax. “Qualified
mortgage insurance” is mortgage insurance provided by the Veterans Administration (VA), the
Federal Housing Administration (FHA), the Rural Housing Administration (RHA), and private
mortgage insurers as defined under Section 2 of the Homeowners Protection Act of 1998 (12
U.S.C. Sec. 4901).
The deduction of qualified mortgage insurance premiums generally applies to amounts paid or
accrued only during tax years 2007 through 2010, with respect to contracts issued during those
years; the provision terminates for premiums paid or accrued after December 31, 2010. This
deduction is also subject to a phaseout. For every $1,000, or fraction thereof, by which the
taxpayer’s adjusted gross income exceeds $100,000, the amount of deductible mortgage
insurance premiums is reduced (but not below zero) by 10%. In the case of a married taxpayer
filing separately, the amounts are lowered to $500 and $50,000. The purpose of this phaseout is to
prevent taxpayers with adjusted gross incomes greater than $109,000 ($54,500 for a married
taxpayer filing separately) from claiming this tax benefit.
Prepaid mortgage insurance amounts that are allocable to periods beyond the year in which they
are paid are attributed to a capital account and treated as paid in the allocable year (contracts
issued by the VA or RHA are excluded from this provision). If the mortgage is paid off before the
end of its term, a deduction is not allowed for the unamortized balance of the capital account.
Several tax reform proposals were introduced, most of which proposed to enact fundamental tax 7
reform by changing the tax base from income to consumption. H.R. 25, the Fair Tax Act of 2005,
and its companion bill, S. 25, proposed a national sales tax that would have eliminated the current
federal income tax system including the mortgage interest deduction. S. 812, the Flat Tax Act of
2005, proposed to replace the current federal income tax system with a flat rate consumption tax
(also known as a modified value-added tax). The flat tax rate would have been 20% of taxable
7 For more detailed information about fundamental tax reform, see CRS Report RL34343, Tax Reform: An Overview of
Proposals in the 110th Congress; CRS Report RL33619, Value-Added Tax: A New U.S. Revenue Source?; and CRS
Report RL33438, A Value-Added Tax Contrasted With a National Sales Tax, by James M. Bickley.
earned income. The bill defined taxable earned income as the excess of earned income (wages,
salaries, professional fees) over a standard deduction, a deduction for cash charitable
contributions, and a deduction for home mortgage interest. The deductions the bill proposed to
allow were consistent with the previously mentioned mandate given by President Bush to the tax
reform panel. H.R. 1040, the Freedom Flat Tax Act, proposed to allow taxpayers to choose an
election to be subject to a flat tax, in lieu of the existing income tax. The rate would have been
19% of wages for the first two years after an election was made, and 17% thereafter. The bill
proposed to allow a basic standard deduction and an additional standard deduction for each
dependent in the household but made no mention of allowing the mortgage interest deduction. S.
1099, the Tax Simplification Act of 2005, proposed a similar flat tax structure, with a 19% tax
rate on wages for the first two years that would then have been reduced to 17% thereafter.
The nine-member panel appointed by President Bush produced a report in the fall of 2005 that
included a proposal to change the mortgage interest deduction. Specifically, the panel
recommended converting the deduction to a tax credit equal to 15% of interest paid on mortgages.
The principal amount of debt associated with the interest that could be claimed would have been
limited to regional house prices in the range of $227,000 to $412,000. This range would have
been based on the house price ceilings that the Federal Housing Administration (FHA) sets for the
amount of a home mortgage loan that it will insure. These amounts vary by county across the
country. Thus, if the principal amount of a mortgage loan was $500,000, interest on the first
$412,00 of loan principal would have qualified, depending on where the house was located.
The remainder of this report describes the deduction for home mortgage interest and explores
rationales for subsidizing homeowners. The report then provides an economic analysis of the
value and performance of the deduction and concludes with a discussion and analysis of the
possible changes to the mortgage interest deduction that could occur under different tax reform
Taxpayers may claim an itemized deduction for payments of “qualified” interest related to a
primary or secondary residence. The payments of interest that qualify include home mortgage
interest, which is the interest paid on any loans secured by either the taxpayer’s main home or a
second home. The loans may be a mortgage (first or second), a line of credit, or a home equity
product. Qualified residences can include cooperative apartments, condominiums, and mobile
homes. Mortgage interest associated with recreational vehicles and boats can be included in the
deduction as long as the vehicles have sleeping, cooking, and toilet amenities. The debt must also
be collateralized by the vehicles to be eligible for the mortgage interest deduction.
Two types of indebtedness are associated with the interest that can be deducted: acquisition and
home equity. Acquisition indebtedness is debt incurred in acquiring, constructing, or substantially
improving a qualified residence and secured by such residence. Home equity indebtedness is all
non-acquisition debt that is secured by a qualified residence. The interest on such debt can be
deducted even if the proceeds are used for personal expenditures unrelated to the home. The
underlying mortgage loans can represent acquisition indebtedness of up to $1 million, plus home 8
equity indebtedness of up to $100,000.
Any secured debt used to refinance home acquisition debt is treated as home acquisition debt, up
to a certain point. The new debt qualifies as home acquisition debt only up to the amount of the
balance of the old mortgage principal just before the refinancing. Any additional debt is not home 9
acquisition debt, but may qualify as home equity debt. Thus, when refinancing, acquisition debt
may not rise and home equity loans may not exceed $100,000. For example, if a homeowner with
a $150,000 mortgage refinances the debt for $300,000, the original $150,000 would be
considered as acquisition debt, an additional $100,000 would be considered as home equity debt,
and the interest on these amounts would be deductible. The interest on the remaining $50,000 of
debt would not be deductible.
The mortgage interest deduction was not introduced as a provision to subsidize homeownership.
It evolved over time as a result of the tax treatment of interest. Initially, the federal income tax,
instituted in 1913, contained a deduction for all interest paid with no distinction between interest
payments made for business, personal, living, or family expenses. At that time, most interest
payments represented business expenses. For example, farmers borrowed money to buy land that
would be used for both their residence and their business. Mortgages and other consumer
borrowing were much less prevalent than in later years. Over time, and with the increasing
sophistication of lending markets, mortgage interest became distinguishable from other kinds of
For several decades, up until the Tax Reform Act of 1986 (TRA; P.L. 99-514), there were no
restrictions on either the dollar amount of residence interest that could be deducted or the number
of homes for which the interest deduction could be claimed. As part of an effort to limit
deductions for personal interest, the TRA imposed limits on the amount of underlying mortgage
loans for which interest could be deducted. These limits restricted loan amounts up to the
purchase price of the home, plus any improvements, and debt secured by the home but used for
qualified medical and educational expense. The interest deduction was also restricted to mortgage
debt on a first and/or second home.
In addition to changes to the deduction of mortgage interest, the Tax Reform Act also eliminated
the deduction of personal interest, creating tax treatment distortions between homeowners and
renters. Homeowners with sufficient equity and credit worthiness could use home equity loans to
replace personal credit card debt, thus preserving the income tax deduction for interest on their
debt. Renters and homeowners without sufficient home equity or credit worthiness became
unable to transfer personal debt in order to continue the tax savings, and thus were made worse
off relative to some homeowners by the 1986 tax law.
8 The alternative minimum tax (AMT) rules for deducting mortgage interest are different from the rules for regular
taxes. Mortgage interest paid for home equity indebtedness (taken out after June 30, 1982) does not qualify as
deductible for AMT purposes unless the loan proceeds are explicitly used to buy, build, or improve the home.
9 U.S. Department of Treasury, Internal Revenue Service, Publication 936: Home Mortgage Interest Deduction
(Washington: GPO, 2004), p. 8.
As part of a group of revenue raising provisions, the deduction of home mortgage interest was
further restricted by the Omnibus Budget Reconciliation Act of 1987 (P.L. 100-203). That act 10
created an upper limit of $1 million ($500,000 for married filing separately) on the combined
acquisition indebtedness for a principal and second residence. Acquisition indebtedness included
any debt incurred to buy, build, or substantially improve the residence(s). Additionally, the
exception for qualified medical and educational expenses was eliminated and an explicit 11
provision for home equity indebtedness was enacted. In addition to interest on acquisition
indebtedness, interest can be deducted on loan amounts up to $100,000 ($50,000 for married
filing separately) for other debt secured by a principal or second residence, such as a home equity 12
loan, line of credit, or second mortgage.
The Omnibus Budget Reconciliation Act of 1990 (P.L. 101-508) enacted another group of
revenue raising provisions that further influenced the mortgage interest deduction. That act
limited the amount of itemized deductions that upper-income households could claim. Some
high-income taxpayers face reduced marginal benefits from the deduction of mortgage interest
due to the limitation on itemized deductions. In particular, taxpayers with income above the
threshold amount of $145,950 for 2005 are required to reduce the amount of itemized deductions,
including the mortgage interest deduction, by 3% of the excess of their income above the
The mortgage interest deduction is a federal subsidy that treats homeownership favorably
compared to other economic activity. The major justification for the deduction for mortgage
interest is that it encourages homeownership, which proponents deem desirable. Homeownership
is believed to promote continued growth and stability of neighborhoods and communities. This
effect results, in part, because homeowners are more likely to participate in community and
political activities, to promote neighborhood cohesion, and to be more socially stable and less
Like all tax benefits, economic theory suggests the deduction can be evaluated by applying the
criteria of economic efficiency, equity, and simplicity. These criteria are discussed in the next
The tax treatment of owner-occupied housing is complex. In a pure income tax system, the
income from financial assets is taxed and the expenses associated with financial assets are
deducted from taxable income. While this tax treatment applies to financial assets and
10 This amount is not indexed for inflation.
11 It is possible for home equity loans to be used for medical, educational, or other purposes.
12 U.S. Congress, Senate Committee on the Budget, Tax Expenditures: Compendium of Background Material on
Individual Provisions, committee print prepared by the Congressional Research Service, Library of Congress, 103rd nd
Cong., 2 Sess., S. Prt. 103-101 (Washington: GPO, 1994), p. 185.
commercial real estate, it does not apply to owner-occupied housing, which, from an economic
perspective, causes distortions among investment assets.
It is important to understand why tax analysts believe that the mortgage interest deduction creates
a subsidy for housing. The underlying issue regarding the taxation of owner-occupied housing is
the exclusion of imputed rent. Net imputed rental income is the difference between the income
homeowners could receive from renting their homes and the associated cost of the home, which
includes mortgage interest, taxes, insurance, maintenance, and depreciation. This net imputed
rental income is not taxed, unlike the rental housing market where owners of rental property
receive income from their tenants and are taxed on that income, net of the same expenses as
homeowners. Economists contend that if net imputed rental income were taxed similarly to
income from rental housing, then the deduction of mortgage interest, real estate taxes,
maintenance and depreciation costs, and other expenses would be considered appropriate for
Congress considered the taxation of net imputed rental income for homeowners during tax reform
efforts in 1986. Congress believed that it would not be advisable to tax imputed net rental income,
though it continued to allow the deduction of mortgage interest. According to the congressional
Joint Committee on Taxation (JCT):
While Congress recognized that the imputed rental value of owner-occupied housing may be
a significant source of untaxed income, the Congress nevertheless determined that
encouraging home ownership is an important policy goal, achieved in part by providing a 13
deduction for residential mortgage interest.
The non-taxation of imputed rental income contributes to asset choice distortions. Individuals
who own a home have an imputed rental income from that home in much the same way as
holders of financial assets receive dividends and interest from the assets they hold. Yet
homeowners are not taxed on their (imputed) income stream while stockholders are taxed on their
income stream. This tax treatment causes households to hold more of their assets in owner-
occupied housing than they may have in the absence of the tax-preferred treatment. The deduction
of mortgage interest further expands the preferential treatment of owner-occupied housing
relative to other investment assets.
Homeownership is usually an attractive investment, even before tax benefits are included. In
recent years, from 1999 through 2004, real estate prices increased more than 56% as compared to 14
the S&P 500 index that declined 6% over the same time period. When the mortgage interest
deduction is combined with other housing tax provisions, like the deduction for state and local
real estate taxes, the value of housing as an investment good rises even further. Some economists
feel that this preferential tax treatment encourages households to overinvest in housing and less in
13 U.S. Congress, Joint Committee on Taxation, General Explanation of the Tax Reform Act of 1986, committee print,
100th Cong., 1st sess., May 4, 1987 (Washington: GPO, 1987), pp. 263-264.
14 Sara Clemence, “Real Estate vs. Stocks,” Forbes website http://www.forbes.com/realestate/2005/05/27/
cx_sc_0527home.html, visited June 6, 2005.
business investments that might contribute more to increasing the nation’s productivity and 15
A justification for the deduction is that it allows neutrality with respect to debt versus equity 16
financing. The cost of equity financing is the after-tax yield that would be earned if funds were
invested elsewhere. The cost of debt financing is the after-tax debt interest rate. Because interest
earned is taxable income and mortgage interest is deductible, the choice of financing the cost of
buying a home is tax neutral. As an example, if a taxpayer chooses to withdraw $250,000 from an
investment account to purchase a home, the cost of that equity financing option is the interest
earnings forgone after taxes. Assuming an interest rate of 6% and a marginal tax rate of 28%, the
annual earnings would have been $15,000 and the after-tax yield would have been $10,800.
Alternatively, if the taxpayer borrows $250,000 to purchase a home, assuming the same interest
rate of 6% and marginal tax rate of 28%, the after-tax cost would be $15,000 in interest paid,
reduced by $4,200 in mortgage interest deducted, for a total cost of $10,800. Thus, the two 17
financing options are neutral with respect to taxation.
For most tax provisions, including the mortgage interest deduction, it is useful to examine
efficiency loss. Efficiency loss, which is also referred to as deadweight loss, is the loss in 18
economic welfare associated with distortions caused by taxes or tax preferences. These
distortions arise when taxpayers are induced to make substitutions in response to the impacts of a
tax or a tax preference. Tax preferences for housing encourage some households to substitute
increased spending on housing for spending on other goods and services.
Most estimates of inefficiency include a broad tax treatment of owner-occupied housing,
incorporating the mortgage interest deduction, the deduction of real estate taxes, and the
exclusion of imputed rental income. Poterba estimated that the deadweight loss for homeowners
in 1990 was $53 for households with incomes of $30,000, $326 for households with incomes of 19
$50,000, and $1,631 for households with incomes of $250,000. Deadweight loss was found to 20
increase as the marginal income tax rate rises and the nominal interest rate declines. Poterba
15 N. Edward Coulson, “Housing Policy and the Social Benefits of Homeownership,” Business Review - Federal
Reserve Bank of Philadelphia, Second Quarter 2002, p. 8.
16 Households with sufficient resources could pay for a home using cash, which is presumably withdrawn from interest-
bearing accounts, thus the reference to “equity” financing.
17 This neutrality exists only if investment earnings are taxed. For many taxpayers, a large portion of their investment
earnings may go untaxed because it is derived from tax-deferred programs, such as 401(k) retirement programs. Other
investment earnings may be taxed at preferential rates. In these cases, neutrality may not exist.
18 James M. Poterba, “Taxation and Housing: Old Questions, New Answers,” The American Economic Review, vol 82,
no. 2, May 1992, p. 238.
20 Poterba used the concept of “economic” income, which equals adjusted gross income plus tax-deductible retirement
and IRA contributions, the excluded component of capital-gains income and business losses, and the deduction for a
dual-earner married couple.
found that the deadweight loss resulting from housing tax preferences declined substantially 21
between 1980 and 1990 because marginal tax rates and nominal interest rates declined.
Gravelle, in a discussion of inefficiency in the tax treatment of capital, found that while “the
estimated welfare cost of the favorable tax treatment of owner-occupied housing has varied in
economic literature,” a welfare loss of 0.1% of consumption has been associated with the 22
favorable treatment of owner-occupied housing.
According to economic theory, in most cases, an economy best satisfies the wants and needs of its
participants if markets operate free from distortions by taxes and other factors. If there is a market
failure and a subsidy remedies that failure, then there is economic justification for the subsidy
based on efficiency. But if there is no market failure, a tax benefit lowers efficiency.
Market failures occur when a market, left on its own, fails to allocate resources efficiently. In
particular, market transactions are inefficient when the marginal benefits are less than the
marginal costs. Market failures may be due to a variety of factors, including the presence of
externalities and common resources; public goods; imperfect competition; and/or asymmetric or 23
There are socially undesirable phenomena that many consider valid targets of public policy, like
inequality, poverty, and inflation, but they are not problems that meet the definition of economic 24
inefficiency. In this context, some of the issues related to homeownership, like affordable
housing and low rates of homeownership among minority groups, are not considered as market
failures. The deduction for mortgage interest is not economically justified on efficiency grounds
in these cases.
The market failure often attributed to homeownership is that of positive externalities. An
externality exists when the activity of an individual directly affects, positively or negatively, the
welfare of another and that effect is not incorporated in market prices. Proponents claim that
homeownership causes positive externalities since it generates not only private benefits for
individuals but also social benefits for the public at large. The individual does not capture all of
these social benefits and, thus, under invests in real estate. From this perspective, the private
demand for homes is less than social demand and too little investment in homeownership occurs.
Government subsidies to homeownership can stimulate private demand for real estate in order to 25
attain more optimal levels of homeownership that achieve market equilibrium outcomes.
21 Ibid., p. 239.
22 Jane G. Gravelle, The Economic Effects of Taxing Capital Income (Cambridge: The MIT Press, 1994), p. 205. In this
text, Gravelle refers to estimates of efficiency gains from the Tax Reform Act of 1986 that were published in the
author’s article, “Differential Taxation of Capital Income: Another Look at the 1986 Tax Reform Act,” National Tax
Journal, vol. 42, no. 4, Dec. 1989, pp. 446-448.
23 For more detailed information about market failures, see CRS Report RL32162, The Size and Role of Government:
Economic Issues, by Marc Labonte, pp. 14-18.
24 Ibid. p. 14.
25 For a more detailed discussion of externalities, see Harvey Rosen, Public Finance (Boston: McGraw-Hill, 1999), pp.
The mortgage interest deduction, as a subsidy to owner-occupied housing, is often justified
because it is claimed that homeownership fosters social stability, social involvement, and socially 26
desirable behaviors among youths and adults. These social benefits from homeownership may
take on the form of better citizenship, increased investment in beautifying property, and civic
involvement that promotes the best interests of the homeowner and the neighborhood. There are
some specific reasons cited as to why homeownership creates positive externalities. They include
homeowners are civically involved because the asset value of the home is tied to the strength of
the community; homeowners are less mobile, which adds stability to communities; and 27
homeowners take better care of their homes.
While subsidies for homeowners are justified by economists because of the presence of positive
externalities from homeownership, there are some ambiguous aspects of these attributes of
homeownership which cast doubt as to whether homeownership deserves extensive subsidies.
A few studies have distinguished difficulties in the research on the positive externalities of
homeownership. In these studies the authors have noted their inability to distinguish whether
homeownership caused the desirable behaviors that led to positive externalities, or, rather, people
who possessed the desirable behaviors tended to become homeowners. Thus, the individual, not 28
homeownership, would be the source of the positive externality. For instance, Cox found that
homeowners as compared to renters were more likely to be civically involved, as did Rohe and 29
Stegman. Yet these studies showed that the type of involvement varied widely across levels of
activity, and was not correlated with any of the expected homeownership and neighborhood
attributes (like house price, level of crime, and voter participation). If the source of the positive
externality lies with the individual and not with their ownership status, then the arguments in
favor of subsidizing homeownership would be diminished.
Even if the existence of positive externalities from homeownership goes unquestioned, the
appropriate magnitude of subsidies for homeownership is difficult to quantify. As previously
mentioned, the tax provisions and benefits for owner-occupied housing add up to $139 billion in
tax revenue loss for FY2005. One could argue that the social benefits of homeownership may not
be worth that amount of cost.
There is a negative externality associated with homeownership: labor immobility.
Homeownership, unlike renting, tends to hinder labor mobility and the ability of individuals to
respond to changes in the demand for labor. Renter-occupied housing, because of its low
transaction costs, is ideal for households that expect to move within a short period of time.
26 George McCarthy, Shannon Van Zandt, and William Rohe, “The Economics Benefits and Costs of Homeownership:
A Critical Assessment of the Research,” Working Paper No. 01-02, Research Institute for Housing America, May
2001, pp. 18-29.
27 Edward Glaeser and Jesse Shapiro, The Benefits of the Home Mortgage Interest Deduction, National Bureau of
Economic Research, Working Paper no. 9284, Oct. 2002, pp. 22-24.
28 K. Cox, “Housing Tenure and Neighborhood Activism,” Urban Affairs Quarterly, vol. 18, no. 1, 1982, pp. 209-207.
29 William Rohe and M. A. Stegman, “The Impact of Home Ownership on the Self-Esteem, Perceived Control and Life
Satisfaction of Low-Income People,” Journal of the American Planning Association, vol. 60, no. 1, 1994, pp. 173-184.
Moving from owner-occupied housing involves higher transaction costs, such as real estate
broker fees and time costs, if the household is selling the property from which it is moving.
Labor immobility interferes with the ability of the economy to adjust to changing market
conditions. This is more true during regional downturns in the economy. For example, a
community experiences the loss of an industry and joblessness rises. Homeowners are reluctant to
move because their house may not sell at a price that exceeds the existing mortgage debt on the
house and the transaction costs of selling the home. Thus, labor becomes more immobile and the
area may remain depressed longer than it might otherwise.
All of the externalities discussed above depend on whether an individual rents or owns. The
externalities do not depend on how much housing is consumed; yet the efficiency of the mortgage
interest deduction depends upon how many marginal or additional households are induced to buy
homes. The deduction is not considered effective if it either fails to increase home buying or
subsidizes home buying that would have been undertaken in the absence of the tax incentive.
The presence of a subsidy for owner-occupied housing creates an increased demand for housing
among renters, some of whom respond by becoming homeowners. Yet, that same subsidy also
creates increased demand for housing among homeowners, some of whom consume more
housing. Most notably, homeowners who increase their demand for housing increase the amount
and quality of housing demanded. In this case, as the mortgage interest deduction subsidizes
existing homeowners, it fails to increase homeownership rates.
It is, however, difficult to prove the mortgage interest deduction’s efficiency, or lack thereof,
because of its continued presence as a subsidy and availability to all home buyers for so many
years. Some economists theorize the value of the mortgage interest deduction is capitalized into 30
housing prices. Buyers of housing, it is thought, bid up the price of owner-occupied housing to
the point where the price of owner-occupied housing, including tax benefits, is equal to the cost
of renting. Thus, the subsidy is perceived to have given homeowners a capital gain in the past, but
new buyers entering the market face higher prices that offset the subsidy. If the deduction is
capitalized this way, it would not induce non-homeowners to become homeowners, though, a loss
of the mortgage interest deduction might adversely affect housing prices.
Some analysts have, in an effort to determine the effectiveness of the mortgage interest deduction,
made comparisons to other countries. In particular, Israel and Australia have homeownership
rates well above those of the United States and yet they do not allow for the deduction of 31
mortgage interest. Canada and Japan have homeownership rates at about the same level as the 32
United States and those countries do not allow mortgage interest deductibility. These rates
suggest that homeownership may not be sensitive to the deductibility of mortgage interest.
30 Michelle J. White and Lawrence J. White, “The Tax Subsidy to Owner-Occupied Housing: Who Benefits?” Journal
of Public Economics, vol. 3, Aug. 1977, pp. 111-126.
31 Bruce Bartlett, “Tax Reform’s Third Rail: Mortgage Interest,” National Center for Policy Analysis, 2001, website
http://www.ncpa.org/bg/bg139.html, visited June 1, 2005.
However, since none of these nations had a mortgage interest deduction and then repealed the
provision, their ability to provide comparative value to the United States is limited.
Few empirical studies have addressed this issue. One study found that the deduction for mortgage 33
interest subsidizes housing consumption, but its impact on homeownership rates was minimal.
Specifically, the authors found that there was “essentially no relationship” between the home 34
mortgage interest deduction and the rate of homeownership. The authors examined more than
30 years of data on the mortgage subsidy rate, which is the marginal subsidy to mortgage interest
from the deduction for the average taxpayer, and the level of homeownership and found that a 1% 35
increase in the mortgage subsidy caused homeownership to rise by .0009%. In this context,
housing consumption reflected the dollar amount spent on housing and homeownership rates
reflected the number of new homeowners. The choice of homeownership, according to the
authors, was influenced by variables other than the mortgage interest deduction, like house
structure, house type, and family size.
To the extent that the mortgage interest deduction promotes homeownership, it can be criticized
on the grounds that the provision yields too little return relative to its cost. In other words, the
amount of increase in homeownership generated by a $1 in tax revenue loss due to the mortgage
interest deduction can be far outweighed by allocating that same $1 to other subsidy programs.
Results of a recent study of mortgage rates and changes in homeownership provided evidence that
direct spending programs and other policy options are more successful at increasing rates of 36
homeownership. The study reviewed underwriting simulations and found that down payment
reductions had larger effects than mortgage rate reductions on increasing rates of homeownership.
In particular, a reduction in the required amount of down payment for a home from 5% to 0% of
the purchase price caused an increase in the number of renters who became owners. This increase 37
equaled a 2 to 2.5 percentage point increase, representing roughly 720,000 renters.
A different simulation, which examined tenure choices, indicated that reducing the amount that
was required for down payment on the purchase of a home increased the probability of ownership
by 4.5 percentage points for all households and 5 percentage points for African-American 38
Further simulations reported in that study showed that cash payments starting at around $5,000
had larger effects than other policies on the ability of renters to purchase homes. A $5,000 down
payment assistance grant was reported to increase the percentage of renters who could buy homes
33 Edward Glaeser and Jesse Shapiro, The Benefits of the Home Mortgage Interest Deduction, National Bureau of
Economic Research, Working Paper no. 9284, Oct. 2002.
34 Ibid., p. 40.
35 Ibid., pp. 39-41.
36 Ron Feldman, “The Quest to Raise Homeownership Rates,” Mortgage Banking, vol. 63, iss. 1, Oct. 2002, pp. 66-73.
37 Ibid., p. 70.
38 Ibid., p. 72.
by 11 percentage points in general, and 13 percentage points for African-American households. A 39
payment of $10,000 per household was reported to cause an effect almost twice as large.
Home equity indebtedness supports homeownership by allowing homeowners to finance
structural improvements, but home equity indebtedness is not limited to this kind of spending.
For example, home equity indebtedness can be used to pay for vacations, reduce credit card debt,
and other personal consumption expenses. Home equity financing used for purposes unrelated to
homeownership does not serve the congressional intent of the provision to encourage
Tax benefits such as those available for homeowners can result in individuals with similar
incomes paying different amounts of tax. In particular, homeowners and renters who are equal in
all other respects, are treated differently as a result of the tax incentives for homeowners. This
differential treatment is a deviation from the standard of horizontal equity, which requires that
people in equal positions should be treated equally.
Another component of equity in taxation is vertical equity, which requires that tax burdens be
distributed fairly among people with different abilities to pay. Housing tax deductions, like all
deductions, benefit those who have sufficient income to owe federal taxes, and the higher the 40
income, the greater the benefit. Those individuals without sufficient income do not have the
opportunity to benefit from the provision. The disproportionate benefit to individuals with higher
incomes reduces the progressivity of the tax system, which is often viewed as a reduction in
An example of the effect an income tax deduction has on vertical equity can be seen by
identifying two individual homeowners, both of whom incur $10,000 in mortgage interest. The
tax benefit to the two differs if they are in different tax brackets. A homeowner with lower
income, who may be in the 15% income tax bracket, receives an exclusion with a value of
$1,500, while the other homeowner, with higher income in the 28% bracket, receives an
exclusion worth $2,800. Thus, the higher income taxpayer, with presumably greater ability to pay
taxes, receives a greater tax benefit than the lower income taxpayer.
Besides renters, there are a few reasons why the mortgage interest deduction may not be available
to certain homeowners.
Some homeowners may not have mortgage debt on which they pay interest. According to the
Census Bureau, 30% of all homeowners own their homes free and clear with no mortgage interest 41
to deduct; in particular, seniors comprise a large portion of homeowners with no mortgage debt.
39 Ibid., p. 73.
40 Some high-income taxpayers face reduced marginal benefits from the deduction of mortgage interest due to the
limitation on itemized deductions. Taxpayers with income above the threshold amount of $145,950 for 2005 are
required to reduce the amount of itemized deductions, including the mortgage interest deduction, by 3% of the excess
of their income above the threshold amount.
41 U.S. Department of Commerce, Census Bureau, American FactFinder, http://factfinder.census.gov/home/saff/
Thus, these individuals do not benefit from the tax deduction for mortgage interest, though they
still benefit from the non-taxation of imputed rental income. In the case of homeowners without
mortgage debt, their net imputed rental income is relatively higher than those with mortgage debt.
So as mortgage debt falls, along with mortgage deduction claims, the value of the exclusion of net
imputed rental income rises.
Other homeowners may not claim the mortgage interest deduction because claiming the standard
deduction is more advantageous.
The mortgage interest deduction can be claimed only by those taxpayers who itemize. Itemizers
tend to be primarily middle-and upper-income households. Nationally, the percentage of 42
taxpayers who itemized was 34% in 2003, 35% in 2002, 34% in 2001, and 33% in 2000. Lower
income taxpayers generally do not itemize and claim the standard deduction amount instead. In
2005, the standard deduction amounts are $10,000 for a married couple filing jointly, $7,300 for
heads of household, and $5,000 for single individuals. When the value of the standard deduction
is increased, the difference between it and the value of the mortgage deduction diminishes for
lower income households. This effect reduces the incentive to purchase owner-occupied housing,
but does not make the taxpayer worse off.
For those households on the homeownership margin, the tax incentive’s influence on their
decision to own as opposed to rent depends on the amount by which the household’s total
itemized deductions exceeds the standard deduction. Taxpayers would need to have itemized
deductions, including deductible taxes such as real estate taxes, state and local income or sales
taxes, or other deductions, that exceed the standard deduction amount to make itemizing
Income tax and tax expenditure data indicate that high income households claim the majority of
the deductions for mortgage interest and receive the majority of tax savings from claiming those
deductions. The most recent tax year data available from the Internal Revenue Service, 2001,
shown in Table 1, indicated that nearly 67% of taxpayers claiming a deduction for mortgage
interest had incomes of $50,000 or more. These households represented only 29% of the total 43
number of taxpayers filing returns. Additionally, households with income of $50,000 or more 44
claimed a majority (76%) of the value of the deduction.
main.html?_lang=en, website visited on Nov. 15, 2004.
42 As reported in the Internal Revenue Service, Statistics of Income Bulletin (Washington: Winter 2003-2004), pp. 105-
107 and at http://www.irs.gov/pub/irs-soi/03in01pd.xls.
43 Internal Revenue Service, Statistics of Income, website http://www.irs.gov/pub/irs-soi/01in21id.xls, visited on
October 25, 2004.
Table 1. Deductions of Mortgage Interest, Claimed in 2001
Share of Tax
Income Group Number of Tax Returns Returns by Income Total Amount Share of
(Adjusted Gross Claiming Group of Deduction Amount of
Income) Deduction Claiming ($ in thousands) Deduction
Under $5,000 159,867 0.4% $1,291,310 0.4%
$5,000 under $10,000 320,826 0.9% $2,490,082 0.8%
$10,000 under $15,000 544,410 1.5% $3,471,107 1.0%
$15,000 under $20,000 818,701 2.3% $5,248,461 1.6%
$20,000 under $25,000 1,167,740 3.2% $7,459,455 2.3%
$25,000 under $30,000 1,400,622 3.9% $9,169,083 2.8%
$30,000 under $40,000 3,551,427 9.8% $23,575,153 7.1%
$40,000 under $50,000 3,977,911 10.9% $27,316,887 8.3%
$50,000 under $75,000 9,545,780 26.3% $74,679,450 22.6%
$75,000 under $100,000 6,331,711 17.4% $57,410,870 17.4%
$100,000 under $200,000 6,574,062 18.1% $77,691,655 23.5%
$200,000 or more 1,938,132 5.3% $40,888,865 12.4%
Total 36,331,190 100.0% $330,692,376 100.0%
Source: Internal Revenue Service, Statistics of Income, website http://www.irs.gov/pub/irs-soi/01in21id.xls, visited
on October 25, 2004.
While the IRS lists the amount of deductions claimed by households, the amount of tax savings
received by households, or alternatively, the amount of federal tax revenue loss is calculated by
the congressional Joint Committee on Taxation (JCT). The actual cost of the mortgage interest
deduction is determined by the amount of federal tax revenue loss that results from taxpayers
claiming the deduction. Tax revenue loss depends on many factors, which include the
composition of the household (i.e., number of adults and dependents), the household’s marginal
income tax rate, and the household’s income. As an example, two households could both claim
$20,000 in mortgage interest deduction that results in different tax revenue loss. If one household
is in the 33% income tax bracket, they would reduce their taxable income by $6,600. If the other
household is in the 25% bracket, their taxable income is reduced by $5,000. The tax revenue
savings to individuals, thus tax revenue loss to the government, differs across households.
As shown below in Table 2, the JCT estimated that 94.2% of the 2001 tax savings associated with
the mortgage deduction were from households with adjusted gross income of $50,000 or higher.
Table 2. Distribution of Tax Expenditure by Income Class
(The Mortgage Interest Deduction at 2001 Income Tax Rates and Income Levels)
Amount of Tax Share of Tax
Adjusted Gross Income Class Expenditure Expenditure
(in millions of dollars)
Below $10,000 11 0.0%
$10,000 to $20,000 117 0.2%
$20,000 to $30,000 433 0.7%
$30,000 to $40,000 981 1.5%
$40,000 to $50,000 2,235 3.5%
$50,000 to $75,000 7,927 12.3%
$75,000 to $100,000 12,204 18.9%
$100,000 to $200,000 23,978 37.2%
$200,000 and over 16,664 25.8%
TOTAL 64,530 100.0%
Source: U.S. Congress, Joint Committee on Taxation, Estimates of Federal Tax Expenditures for Fiscal Years 2002
to 2006, JCS-1-02 (Washington: GPO, 2002), p. 34.
The deduction of mortgage interest for homeowners contributes to the complexity of the tax code
and raises the cost of administering the tax system. Those costs, which can be difficult to isolate
and measure, are often excluded in the cost-benefit analysis of the provision. The complexity of
the tax code adds to the time cost of taxpayers in either learning how to claim various incentives
and doing so, or an increased direct cost of paying tax professionals to perform the service for the
taxpayer. Also, the more complex the tax code is, the more opportunity for tax evasion and
Federal tax policy changes cause unintended consequences for housing subsidies enacted through
the tax code. Several significant tax cuts have been enacted recently, all of which have affected
the pool of potential beneficiaries of housing tax policy. Most notably, the Economic Growth and
Tax Relief Reconciliation Act of 2001 (EGTRRA, P.L. 107-16) increased the value of standard
deductions and reduced marginal income tax rates for individuals, causing lower tax burdens. In
providing tax relief that was designed to stimulate the economy, a likely unintended consequence
was to reduce the number of taxpayers claiming the housing tax deduction. As an example, the
standard deduction for married taxpayers filing jointly rose from $7,600 in 2001 to $9,700 in
period, mortgage interest was likely to be lower such that some households who may have been
itemizing on their tax returns began to claim the standard deduction.
45 This statement assumes the homeowner has a fixed-rate mortgage and is making the same amount of payments
towards interest and principal each month.
Additionally, the Jobs and Growth Tax Relief Reconciliation Act of 2003 (JGTRRA; P.L. 108-
27), which was passed in May 2003, accelerated many of the tax provisions of EGTRRA, thus
further expanding unintended consequences of reducing the number of taxpayers benefitting from
the mortgage interest deduction.
Tax reform in the 109th Congress may have caused moderate or fundamental changes in the
federal tax system. These changes could have been as far reaching as eliminating the current
income tax system and replacing it with an alternative tax system, like a national sales tax. This
change could have eliminated deductions and credits, including the mortgage interest deduction.
Alternatively, Congress could have chosen to simplify the existing tax code by modifying or
eliminating provisions, like the mortgage interest deduction. Finally, tax reform could have
occurred and the mortgage interest deduction could have been unchanged.
In the absence of tax reform, Congress may or may not make changes to the mortgage interest
deduction. Congress may choose to modify or eliminate the deduction for homeowners for a
variety of reasons, including revenue raising strategies to reduce the deficit or addressing the
equity issues surrounding the provision. Alternatively, Congress may choose not to alter the
mortgage interest deduction.
Two possibilities can be examined in analyzing the elimination of the mortgage interest
deduction. The first case is that the provision is eliminated. The second case, which may be more
likely to occur than the first case, is in the event of tax reform that fundamentally changes the
federal tax system and eliminates the mortgage interest deduction.
As mentioned previously, several tax reform proposals, if enacted, would have eliminated the
mortgage interest deduction. There were, however, differences among proposals concerning the
elimination of the mortgage interest deduction.
In the 109th Congress, several bills were proposed to enact a flat tax. A flat tax is levied on wages
and earnings, but is not an income tax. The flat tax is considered a consumption tax because it
does not tax savings and investment. The difference between a flat tax and a sales tax, both of
which are types of consumption taxes, is where the tax is collected. The flat tax is levied on
income as it is earned and the sales tax is levied on income as it is spent.
Two of the proposed flat tax bills (H.R. 1040 and S. 1099) would have eliminated the mortgage
interest deduction. S. 812 proposed to preserve the mortgage interest deduction. H.R. 1040 and S.
retirement distributions, and unemployment compensation after subtracting a basic standard
deduction and an additional standard deduction for each dependent. S. 812 proposed a similar
definition of taxable income but would also allow deductions for cash charitable contributions
and home mortgage interest.
In general, flat taxes would eliminate the three primary tax provisions for owner-occupied
housing: the mortgage interest deduction, the deduction of state and local property taxes, and the 46
exclusion of capital gains on sales of primary residences. As mentioned previously, the cost of
these tax provisions combined with the exclusion of net imputed rental income is estimated to be 47
$139 billion in tax revenue loss for FY2005. While that amount is sizeable, it is unclear as to
what the net change in federal revenue may be as a result of moving to a flat tax. It is notable that
the fundamental tax advantage of owner-occupied housing, the non-taxation of imputed rental
income, would remain unless specifically addressed by legislation. However, the flat tax would
eliminate the favorable treatment of owner-occupied housing relative to other assets because the
returns to all new investment would be untaxed, not just the returns to homeownership.
Estimates of the effect of the eliminations of these tax provisions were made in the mid-to late
1990s, when fundamental tax reform was last being considered. A 1996 report estimated that the
elimination of both the mortgage interest and property tax deductions would reduce house prices
by an average of 17% but could range from as low as 11% to as high as 34% depending on the 48
region of the country.
In 1996 CRS examined the effect of a flat tax on housing.49 The principal finding was that if a flat
tax were instituted, eliminating the mortgage interest deduction, demand for owner-occupied
housing and homeownership would likely contract. CRS found that the overall price of housing
could fall by 22%, but that only one-third of that amount of price decline would be due to the loss 50
of housing tax deductions. The remaining portion of the price contraction would occur as a
result of other factors, one of which would be the elimination of the preferential tax treatment of
housing relative to other investment assets. Assets would shift from housing to business
investment, which would improve economic efficiency. The report stated,
Economists would typically view this shift as a benefit of tax reform, because the current tax
system encourages an inefficient overinvestment of funds in housing and consequent 51
underinvestment in business assets.
CRS found much smaller effects than private sector reporting, which had suggested that the
elimination of the mortgage interest deduction would cause larger, more adverse effects on 52
housing prices and on the economy. For example, a private sector report concluded that enacting
46 Under a flat tax there is no capital gains tax.
47 U.S. Congress, Joint Committee on Taxation, Estimates of Federal Tax Expenditures for Fiscal Years 2002 to 2006,
JCS-1-02 (Washington: GPO, 2002), p. 33.
48 Dennis R. Capozza, Richard K. Green, and Patric Hendershott, “Taxes, Mortgage Borrowing, and Residential Land
Prices,” in Henry Aaron and William Gale, eds., The Economic Effects of Fundamental Tax Reform, (Washington,
D.C.: The Brookings Institution 1996), p. 192. Due to differences in interest rates and other economic factors, these
numerical estimates would likely be smaller if the elimination of the mortgage interest deduction occurred in the
49 CRS Report 96-379, The Flat Tax and Other Proposals: Effects on Housing, by Jane Gravelle, Apr. 29, 1996 (Out of
print. Available from the author).
50 Ibid., p. 6.
51 Ibid., p. 4.
52 Ibid., p. 1.
a flat-rate income tax would cause home prices to fall by 15% specifically because of the
elimination of the deduction for mortgage interest. The price decline would in turn eliminate $1.7 53
trillion of homeowners’ equity.
Capozza, Green, and Hendershott examined the effects of a flat tax on housing and found house 54
price declines ranging from 10 to 30%. The authors noted that house price changes that could
occur under a flat tax would vary based on what changes occurred with interest rates and with
property taxes. The higher these two variables were found to be, the greater the percentage 55
decline in house prices. After performing regional analysis, the authors further concluded that
the degree of house price decline would vary by region and would be dependent on the rate of
property taxation and the existing price level of houses in the region. High-priced metropolitan
areas in California, as an example, would be hardest hit, while low-priced metropolitan areas in
the south would experience the smallest degree of house price decline.
The issue of fundamental tax reform and housing was further examined in a working paper from 56
the National Bureau of Economic Research (NBER). The authors simulated tests of housing
market response to fundamental tax reform and found that tax reform proposals would cause
contraction in the quantity of housing stock, though the degree of contraction varied depending 57
on the type of tax reform. The authors reported that the removal of the deductions for mortgage
interest and property taxes would cause a 1.9% decline in the housing stock, while a flat tax 58
enactment would cause an 8% decline. The authors note that their results are constructed under
a partial equilibrium assumption that the long-run supply of housing is perfectly elastic and that if
housing demand were to grow more elastic, the impact of eliminating tax benefits for owner-59
occupied housing causes a greater decline in the stock of housing.
Movement to a sales tax system from the current income tax system would eliminate existing tax
subsidies to owner-occupied housing, including the mortgage interest deduction and would also
eliminate the significance of the net exclusion of imputed rental income. A national retail sales tax
would repeal the personal and corporate income tax code and replace it with a tax on all final
sales of goods and services to consumers.
A national retail sales tax would impose a tax on newly built homes, which would cause existing
homes to enjoy a tax advantage. Under most concepts of a national retail sales tax, the sale of an
existing home would be categorized as the sale of used property and would not be taxed. In this
context, the value of existing homes, relative to newly built homes, would rise. If rental payments
for housing were taxed, owners of existing homes would also enjoy a tax advantage over renters.
53 Roger E. Brinner, Mark Lasky and David Wyss, Residential Real Estate Impacts of Flat Tax Legislation (Lexington,
MA: DRI/McGraw-Hill, 1995).
54 Capozza, Green, and Hendershott, “Taxes, Mortgage Borrowing, and Residential Land Prices,” p. 190.
55 Ibid. It is important to note that these estimates were made when interest rates were higher than in 2005 so the effects
would be much smaller given today’s interest rates.
56 Donald Bruce and Douglas Holtz-Eakin, “Apocalypse Now? Fundamental Tax Reform and Residential Housing
Values,” National Bureau of Economic Research Working Paper 6282, Nov. 1997.
57 Ibid., p. 9.
58 Ibid., p. 10.
Generally, eliminating the mortgage interest deduction separate from tax reform would harm
itemizing homeowners with the highest amounts of mortgage debt and, in particular, those
homeowners who could not easily replace that debt with equity. Homeowners with mortgages
would be expected to reduce their debt and new homebuyers would incur less debt than they 60
would have previously. For some, homeownership would no longer be an optimal choice. If this
were the case, a reduction in the demand for homes would occur and housing prices would
decline, at least in the short run, and lead to a decrease in the quantity of owner-occupied housing
in the long run.
A 1996 study examined the possibility of eliminating the mortgage interest deduction, noting that
the benefits of the mortgage interest deduction accrued overwhelmingly to young households and 61
to high income households. The authors found that the younger households had limited wealth
and were forced to borrow to finance homeownership, while higher income households’
participation with the mortgage interest deduction was to engage in tax arbitrage. For the higher
income households, the loss of the deduction would lead these households to use financial assets
to reduce or eliminate their mortgage debt. This alternative, however, is not available to
households without financial assets.
Some economists theorize that the loss of the mortgage interest deduction separate from tax
reform would adversely affect housing prices because the value of the subsidy is capitalized into
housing prices. Buyers of housing, it is thought, bid up the price of owner-occupied housing to
the point where the price of owner-occupied housing, including tax benefits, is equal to the cost 62
of renting. Thus, the subsidy is perceived to have given homeowners a capital gain in the past, 63
the loss of which could be factored in to housing prices.
The full economic implications of eliminating the mortgage interest deduction depend upon
whether or not the subsidy is capitalized into real estate prices. If it is fully capitalized eliminating
the subsidy would not affect the cost of owning but many owners would experience significant 64
changes in wealth. Gyourko and Sinai found that in 2003, the value of the subsidy amounted to
about a fifth of property value, on average, in the United States. In particular, they found there
were more than 40 metropolitan areas, including many densely populated ones in and around
Boston, New York City, Washington, D.C., Los Angeles, and San Francisco for which the present
value of the subsidy flow is greater than 25% of house values. The increased revenue associated
with eliminating the subsidy could be redistributed back to taxpayers in the form of lower taxes.
According to the authors, the net wealth effect of lower taxes would likely be significant enough
60 The presence of high transaction costs make selling seem impractical for homeowners. Renters, on the other hand,
could choose to forgo home purchases.
61 Dennis R. Capozza, Richard K. Green, and Patric Hendershott, “Taxes, Mortgage Borrowing, and Residential Land
Prices,” in Henry Aaron and William Gale, eds., The Economic Effects of Fundamental Tax Reform (Washington, D.C.:
The Brookings Institution 1996), pp. 174-175.
62 Michelle J. White and Lawrence J. White, “The Tax Subsidy to Owner-Occupied Housing: Who Benefits?” Journal
of Public Economics, vol. 3, Aug. 1977, pp. 111-126.
63 Ibid., p. 112.
64 Joseph Gyourko and Todd Sinai, “The Spatial Distribution of Housing-Related Ordinary Income Tax Benefits,” Real
Estate Economics, Winter 2003, vol. 31, iss. 4, pp. 529-531.
to offset, at least in part, the loss of the mortgage interest deduction.65 The report did not address
how these decreased property values would effect state and local government revenue.
If the tax subsidy is not capitalized into real estate prices, the cost of ownership rises when the
deduction is eliminated. In 2003, Gyourko and Sinai determined that the rise in cost of ownership
could equal between 4% and 6% of annual household income in over half of the nation’s
metropolitan areas. However, in 20 metropolitan areas, including many in California, Hawaii, and
Massachusetts, costs of homeownership amounted to 10% of income or more. Their analysis
indicated that if the tax savings from eliminating the subsidy were rebated through the
implementation of a tax credit, this rebate would more than offset the increase in the cost of
ownership in most areas. However, the authors noted that returning the tax savings through the 66
existing tax system would not provide a similar offset.
Repealing the mortgage interest deduction would cause responses outside the housing market.
Demand for mortgage debt financing would decline causing mortgage lending to decline.
Capozza, Green, and Hendershott estimated that loan-to-value ratios, the amount of debt relative
to equity value in the home, would fall by more than 30% and that house price declines of 10% 67
could accompany mortgage lending declines of about 40%. Mortgage interest rates could fall in
response to lower demand for mortgage debt. Lower mortgage rates could attract new
homebuyers which could increase demand for homes and cause house price increases, thereby
reversing declines in mortgage lending and house prices. The net results, however, would likely
yield lower house prices, lower before-tax mortgage rates, higher after-tax mortgage rates, and
lower mortgage lending.
Bourassa and Grigsby examined the possibility of the elimination of the mortgage interest
deduction and offered a plan for phasing out the provision that would minimally affect housing 68
demand. Citing that proposals to limit the mortgage deduction are opposed on the “grounds that
existing homeowners would suffer huge capital losses,” the authors proposed a 15-to 20-year
period of phasing in the elimination of the tax provision. This type of elimination, the authors
argued, would not adversely affect demand for housing or the capital values of the existing 69
Finally, the repeal of the deduction could be justified as a second-best policy to address the
absence of taxation on net imputed rental income. Net imputed rental income from
homeownership has never been included as taxable income in the federal income tax. This is due
in part to limited acceptance by non-economists of the idea that owning a home (or other
consumer durables) provides owners with implicit income that should be taxed, and in part
65 Ibid., p. 530.
66 Ibid., p. 557.
67 Capozza, Green, and Hendershott, “Taxes, Mortgage Borrowing, and Residential Land Prices,” p. 196.
68 Steven C. Bourassa and William G. Grigsby, “Income Tax Concessions for Owner-Occupied Housing,” Housing
Policy Debate, vol. 11, iss. 3, (Washington, DC: Fannie Mae Foundation, 2000), p. 541.
69 Ibid., p. 541.
because of the administrative difficulty of taxing such income. Further, many economists doubt
that imputed rent could be accurately taxed in the United States as a practical matter. Little is
known, for example, about the probable rental value of owner-occupied homes; only rough
estimates could be made from available data on assessed house values. Available data on rental
rates is not useful because of the differences in size and quality of rental units as compared to
Congress could choose to modify the provision to either simplify the tax code, increase tax
revenue, or both. In the event of tax reform, these modifications could be considered and included
in various reform proposals. For instance, Congress could choose to implement a flat tax and
allow some modified version of the mortgage interest deduction. The possibilities in this regard
could be endless. Thus, the modifications listed below describe possible policy options that could
be taken without regard to a particular form of systematic tax change.
One possible way of modifying the mortgage interest deduction would be to reduce the maximum
allowable mortgage debt limit from $1 million to a lesser amount. This choice would reduce the
tax benefit for those homeowners with debt in excess of the allowable amount, presumably high
income households with very large houses. The reduction in tax savings could cause the prices of
homes worth more than the allowable debt amount to decline, which would reduce demand for
There are a few possible consequences of reducing the debt limit. Some homeowners would use
equity to pay down some or all of their mortgage debt. Other households may reduce their level
of housing consumption by moving to less expensive homes. Marginal homeowners, who are
typically not high income households or purchasers of expensive homes, are unlikely to be
affected by a reduction in allowable debt.
In February 2005, the Congressional Budget Office (CBO) examined this policy option. It
estimated that reducing the amount of mortgage principal that could be allowed from $1 million
to $500,000 would affect 700,000 taxpayers with large mortgages and raise $2.7 billion in 70
FY2006. The number of taxpayers affected by the year 2010 would rise to 1.3 million with an
increase in revenue of $4.3 billion that year. Over the period of 2006 through 2010, the total
revenue increase from limiting the principal debt that could be claimed for the mortgage interest
deduction would be $47.9 billion. CBO estimated that less than 1% of all homeowners and about 71
This policy option would not be any more complex than the existing tax code, though there would
be increased administrative costs associated with the change. There would be increased costs to
make taxpayers aware of the change and also increased costs to change tax forms. If the reduction
70 U.S. Congress, Congressional Budget Office, Budget Options, Congressional Study, (Washington: GPO, 2005), p.
in allowable mortgage debt caused increases in attempts to avoid or evade taxation, IRS costs
Disallowing home equity indebtedness is another possible modification to the mortgage interest
As mentioned previously, home equity debt often finances personal consumption rather than
financing homeownership, which causes disparate treatment among similarly situated taxpayers.
As an example, two taxpayers can incur $20,000 in consumption debt. The taxpayer who finances
this debt with a credit card, as opposed to a home equity loan, is unable to receive a tax benefit by
deducting the interest charges. The taxpayer who uses a home equity loan to pay off the credit
card debt can deduct the interest and thus is favorably treated by the tax code.
The interest deduction’s disparate treatment of homeowners also reflects differences in housing
appreciation and values. Homeowners with housing price appreciation can take advantage of
home equity debt while other homeowners, with flat or declining home values, may not be able to 72
benefit from home equity debt. Allowing home equity indebtedness deductibility also favors
those homeowners who have sufficient credit worthiness to incur home equity debt. This effect
further contributes to disparate treatment of taxpayers to the extent that some homeowners are
unable to obtain home equity loans. On the other hand, eliminating home equity deductibility
negatively affects households who have no significant pool of funds from which to finance
needed repairs or improvements. These households, who are more likely to be lower-income
and/or elderly households, would lose the benefit of the tax savings if home equity indebtedness
were disallowed. Yet this loss only applies to those households who were benefitting from the tax
savings, which may have been a small portion of the affected population.
This policy choice has been presented as an option to improve tax compliance and reform tax
expenditures by the Joint Committee on Taxation (JCT). The option was estimated to raise $0.3
billion in tax revenue for FY2006 and $22.6 billion in tax revenue over the 10-year period 73
beginning in FY2006.
The mortgage interest deduction could be limited to acquisition indebtedness associated with
principal residences only. Presumably it is higher income households that are able to afford more
than one residence, so disallowing the mortgage interest from debt on a second residence (or
mobile home, yacht, or other qualifying property) would contribute to making the provision more
equitable by reducing the amount of benefit that higher income households can receive.
This modification may contribute to a decrease in demand for housing, vacation homes in
particular, which could lead to a short-term decrease in the price of vacation homes. The supply
72 Ibid., p. 55.
73 U.S. Congress, Joint Committee on Taxation, Options to Improve Tax Compliance and Reform Expenditures, JCS-
02-05 (Washington: GPO, 2005), pp. 52-56.
response to that decrease in demand may cause a reduction in the quantity of vacation homes
which would result in higher prices and a lower quantity of vacation homes.
This change would increase tax revenue by reducing the amount of mortgage interest taxpayers
deduct and could be simpler to administer. It is, however, possible that this change could increase
attempts at tax avoidance or evasion if individuals attempted to claim residences that are not their
principal residence. In this case, IRS administration costs would rise.
There are many possibilities that Congress could choose to modify the mortgage interest
deduction. Some policy options would not necessarily simplify the tax code or increase tax
revenue but might, instead, improve the equity of the provision. This section discusses those
The deduction as it currently exists does not limit participation by taxpayers to households with
particular income levels. The imposition of an income threshold, as in the case of certain other tax
provisions, would increase tax revenue by reducing the number of taxpayers eligible to claim the
deduction. As an example, the tax credit for new homebuyers in Washington, D.C., is a
nonrefundable credit against federal taxes of up to $5,000 for the first-time purchase of a
principal residence in the District of Columbia. The credit is phased out for individuals with
adjusted gross income (AGI) of $70,000 to $90,000 and for joint filers with AGI of $110,000 to
$130,000. In particular, for every $1,000 of AGI above $110,000 for joint tax filers, the credit is
reduced by $250. Once AGI exceeds $130,000, no credit may be taken.
Imposing an income cap could increase equity because the cap would disallow high-income
households from claiming the deduction. In doing so, the provision would subsidize
homeownership for households with less income which may add progressivity to the tax code. It
would still be the case that low-income households with no tax liability, or those who claim the
standard deduction, would not benefit from the provision.
This policy choice would add complexity to the tax code by further complicating the tax
provision. This option would require taxpayers to do more work in determining their eligibility
for the provision. The Internal Revenue Service would also have increased administrative effort
to enforce income restrictions for those claiming the provision.
An above-the-line deduction, which would be available to taxpayers regardless of whether they
claim the standard deduction or itemize deductions, could increase equity. This possibility would
allow the deduction of mortgage interest to be available to more modest income taxpayers than is
currently allowed by tax law. This change would contribute to progressivity in the tax code, but
would also increase tax revenue loss by allowing more individuals to claim the deduction.
At its initial introduction, an above-the-line deduction would add complexity to the tax code and
tax filing forms. Because of its newness, the provision would increase the administrative work
taxpayers experience in filing their income tax returns and the work the IRS performs in
processing those returns. Yet, over time, an above-the-line provision would not be any more
complex than the existing deduction. The enactment of an above-the-line deduction could,
however, be perceived as undermining the concept of the standard deduction, which exists in part
to simplify the tax code.
Another possible change that would increase the equity of the deduction would be to transform
the mortgage interest deduction into a partial tax credit. This change would benefit lower income
individuals by allowing them to benefit from claiming mortgage interest. Those taxpayers who
claim the standard deduction could also claim a tax credit for mortgage interest.
This policy option would provide taxpayers at different income levels with the same rate of
subsidy for mortgage interest payments, unlike the current deduction which varies with the
marginal tax rate. The tax credit would also allow lower income households that currently do not
claim mortgage interest, perhaps because the interest they pay is less than the value of the
standard deduction, to benefit by claiming a tax credit for mortgage interest.
This option could either increase or decrease federal revenues, depending on the amount of credit
chosen. Estimates for this policy option were made by the Congressional Budget Office (CBO) 74
25 years ago. At that time, CBO determined that moving to a 25% tax credit would increase
revenues by about $2.4 billion in the year the study was conducted, and by $4.3 billion in the next
year. At that rate of 25%, only a small group of wealthy taxpayers would be worse off. A credit of
30% or more would have decreased revenues. CBO also found that a tax credit of this amount
would raise house prices for less expensive homes and lower them for higher-priced units. This
effect could lead to some upper-income homeowners experiencing a decrease in the value of their
homes as well as an increase in their tax payments. Allowing current owners the choice of a
deduction or credit could have limited those capital losses, but the revenue tax revenue losses 75
from doing so would have been substantial.
Higher income taxpayers may be less well off by the transformation of the current mortgage
interest deduction to a tax credit if the tax credit were only a partial credit. For higher income
taxpayers, the value of the current mortgage interest deduction is likely to be greater than the
value of a proposed tax credit which could cause an after-tax reduction in price for high income
households. Yet, higher income households are less sensitive than lower income households to
changes in price, so the loss in value of the mortgage interest deduction if it were to become a tax
credit may not be significant enough to cause high income households to abandon
homeownership. If high income households responded at all, they may reduce their housing
consumption, perhaps by purchasing less expensive housing.
As a result, this proposal could increase homeownership because it is most likely to alter
decisions about homeownership at the margin. More families in the lower-and middle-income
groups may choose to own, rather than rent. Others in those same groups may choose to buy more
housing than they currently own.
74 The more recent policy proposal by the tax reform panel established by President Bush in 2005 did not include cost
75 U.S. Congress, Congressional Budget Office, Tax Treatment of Homeownership: Issues and Options, Congressional
Study (Washington: GPO, 1981), p. 19.
It is notable, however, that in terms of the measurement of income for taxation purposes, tax
analysts believe a tax credit is not appropriate under a pure income tax system. Tax credits are
dollar-for-dollar reductions in a taxpayer’s federal and/or state income tax liability, while tax
deductions are offsets to a taxpayer’s pretax income. Since income taxes are meant to raise
revenue from taxpayers’ according to their ability to pay, tax deductions alter income to best
reflect taxpayers ability to pay. Deductions, unlike credits, reduce the base amount of income to
be taxed and in doing so, affect the variable benefits correlated to a taxpayer’s particular income 76
tax rate or bracket.
Tax reform proposals made in the 109th Congress promoted consumption-based tax systems, most
of which would have eliminated the mortgage interest deduction. In the case of either a flat tax or
a sales tax, all credits and deductions as they exist in the current income tax system would have
been eliminated. The deduction of mortgage interest would no longer be allowed, unless
specifically addressed by legislation. As an example, S. 812 proposed a flat tax that included the
preservation of the mortgage interest deduction along with the charitable contributions deduction.
Although the mortgage interest deduction is generally defended on the basis of supporting access
to homeownership, most studies conclude that its primary effect is on housing consumption, not
homeownership. Most of the benefits of the deduction accrue to households with well-above-
average incomes, while households on the cusp of homeownership tend to have lower income.
Depending on how they are structured, most studies suggest that lower income households would
be aided more by down-payment assistance than by the deduction for mortgage interest.
Eliminating the mortgage interest deduction separate from tax reform would have several effects.
It would tend to reduce debt relative to equity by ending the neutrality between equity-and debt-
financing. It would lead some households who may have been considering homeownership to
choose to continue to rent. Some owner-households may choose to consume less housing, either
by choosing smaller homes or fewer amenities. The elimination could lower home prices relative
to rents in the market and could cause households to convert housing assets to business
If the deduction were eliminated as a result of fundamental tax reform, the resulting effects would
depend on a variety of variables. These variables include the nature of tax reform, the resulting
changes in the tax base and tax rates, changes in interest rates, and other economic variables. If
the deduction were eliminated without other tax policy changes, federal income tax revenues
could increase, the tax base could be broadened, and the federal budget deficit could be reduced.
This elimination could lead to greater economic efficiency, increased equity, and simplification of
the current income tax system.
Modifications to the mortgage interest deduction could also have several effects. Congress could
choose to allow the deduction for only one residence, rather than two, or reduce the allowable
principal debt from $1 million to some lower amount, perhaps $500,000. These changes would
reduce the amount of tax revenue loss associated with the provision without adding complexity to
76 Babette B. Barton, et al., eds., Study of Federal Tax Law: The Taxation of Income, 19th Ed., (Chicago, IL: CCH
Incorporated, 1994), p. 341.
the tax code. Congress could also choose to improve the equitable nature of the provision by
allowing more low income households to claim mortgage interest, either as an above-the-line
deduction or as a tax credit.
Finally, the mortgage interest deduction could remain unaltered. It would continue to be a benefit
to more than 37 million taxpayers and provide more than $70 billion in annual tax savings to
Pamela J. Jackson
Section Research Manager