A. Introduction

1. The U.S. Administration has recently announced plans to reform the housing finance system and reconsider the overall role of the federal government in housing policy. The U.S. housing finance system is very complex, expensive and regressive. In addition, despite its large cost and complexities, it is unclear whether the current housing finance system has achieved its goals of raising homeownership (particularly for underrepresented groups). Indeed, homeownership rates in neighboring Canada exceed the U.S. rate despite having a cheaper and simpler housing finance system (Table 1). Instead, housing consumption is much higher in the United States, as the current system encourages higher leveraging and thus the purchasing of larger, more expensive houses (Glaeser and Shapiro, 2003; van der Noord, 2005). Given the evidence, reconsidering the broader objectives of housing policy is a step in the right direction.

^{Table 1.}

United States and Canada: Housing Finance

Tax Expenditures

Sources: Finance Canada, personal contacts ; U.S. Office of Management and Budget (2010), Analytical Perspectives, Budget of the U.S. Government Fiscal Year 2011; and author's calculations.

^{Table 1.}

United States and Canada: Housing Finance

Tax Expenditures

		United States	Canada
	Homeowners
	Exclusion of net imputed rental income	√	√
	Deductibility of mortgage interest on owner-occupied homes	√	X
	Deductibility of State and local property tax on owner-ocupied home	√	X
	Deferral of income from instalment sales	√	X
	Capital gains exclusion on home sales	√	√
	Credit for homebuyer (temporary)	√	√
	Exception of sales tax for purchase of resale homes	X	√
	Investors
	Exclusion of interest on rental housing bonds	√	X
	Exclusion of interest on owner occupied mortgage subsidy bonds	√	X
	Exception from passibve loss rules for $25,000 of rental loss	V	X
	Credit for low-income housing investments	√	X
	Accelerated depreciation on rental housing (normal tax method)	√	X
	Disccharge of mortgage indebtedness	√	X
Mortgage Insurance
Public (Federal)		FHA and VA	CMHC
Public (state)		Massachussettes	X
Private loan coverage		20-30%	90%
Wholesale Funding
Mortgage securitization with explicit government guarantee		Ginnie Mae	NHA MBS
Mortgage securitization with implicit government guarantee		GSEs	X
Mortgage securitization with no government guarantee		√	√
Corporate bonds issued by special facilities		GSEs	Canada Mortgage Bonds
Corporate bonds issued by secondary market conduits		√	√
Corporate bonds issued by primary market lenders		√	√

Sources: Finance Canada, personal contacts ; U.S. Office of Management and Budget (2010), Analytical Perspectives, Budget of the U.S. Government Fiscal Year 2011; and author's calculations.

View Table

^{Table 1.}

United States and Canada: Housing Finance

Tax Expenditures

		United States	Canada
	Homeowners
	Exclusion of net imputed rental income	√	√
	Deductibility of mortgage interest on owner-occupied homes	√	X
	Deductibility of State and local property tax on owner-ocupied home	√	X
	Deferral of income from instalment sales	√	X
	Capital gains exclusion on home sales	√	√
	Credit for homebuyer (temporary)	√	√
	Exception of sales tax for purchase of resale homes	X	√
	Investors
	Exclusion of interest on rental housing bonds	√	X
	Exclusion of interest on owner occupied mortgage subsidy bonds	√	X
	Exception from passibve loss rules for $25,000 of rental loss	V	X
	Credit for low-income housing investments	√	X
	Accelerated depreciation on rental housing (normal tax method)	√	X
	Disccharge of mortgage indebtedness	√	X
Mortgage Insurance
Public (Federal)		FHA and VA	CMHC
Public (state)		Massachussettes	X
Private loan coverage		20-30%	90%
Wholesale Funding
Mortgage securitization with explicit government guarantee		Ginnie Mae	NHA MBS
Mortgage securitization with implicit government guarantee		GSEs	X
Mortgage securitization with no government guarantee		√	√
Corporate bonds issued by special facilities		GSEs	Canada Mortgage Bonds
Corporate bonds issued by secondary market conduits		√	√
Corporate bonds issued by primary market lenders		√	√

Sources: Finance Canada, personal contacts ; U.S. Office of Management and Budget (2010), Analytical Perspectives, Budget of the U.S. Government Fiscal Year 2011; and author's calculations.

View Full Size

Select Advanced Economies: Homeownership Rate and Tax Expenditures on Housing

Citation: IMF Staff Country Reports 2010, 248; 10.5089/9781455206759.002.A005

Sources: OECD, Economic Survey of the United States 2010 (forthcoming); Statistics Canada; U.S. Census Bureau; and author's calculations.

• Download Figure
• Download figure as PowerPoint slide

Tax Expenditures on Housing in Select OECD Countries, 2006 ^1/

Sources: Eurostat; Haver Analytics; OECD, Economic Survey of the United States 2010 (forthcoming) ; author's calculations.

^1/Data on tax expenditures refer to 2004 for Canada, and 2008 for Spain and the United States.

^2/Latest available year.

Tax Expenditures on Housing in Select OECD Countries, 2006 ^1/

	CAN	DEU	NLD	ESP	GRB	USA
% of GDP	0.2	0.2	0.1	0.4	1.2	1.1
% of Central Gov. Revenues	1.3	2.0	0.1	1.1	3.3	5.9
Number of Tax Expenditures	1	10	2	3	7	11
Homeownership rate 2/	68.4	55.6	67.8	89.1	73.2	67.4

Sources: Eurostat; Haver Analytics; OECD, Economic Survey of the United States 2010 (forthcoming) ; author's calculations.

^1/Data on tax expenditures refer to 2004 for Canada, and 2008 for Spain and the United States.

^2/Latest available year.

View Table

Tax Expenditures on Housing in Select OECD Countries, 2006 ^1/

	CAN	DEU	NLD	ESP	GRB	USA
% of GDP	0.2	0.2	0.1	0.4	1.2	1.1
% of Central Gov. Revenues	1.3	2.0	0.1	1.1	3.3	5.9
Number of Tax Expenditures	1	10	2	3	7	11
Homeownership rate 2/	68.4	55.6	67.8	89.1	73.2	67.4

Sources: Eurostat; Haver Analytics; OECD, Economic Survey of the United States 2010 (forthcoming) ; author's calculations.

^1/Data on tax expenditures refer to 2004 for Canada, and 2008 for Spain and the United States.

^2/Latest available year.

B. Impediments of the Current System

2. The U.S. housing finance system is extremely complex and multi-faceted, with most policies dating back almost a century ago. The system includes numerous tax incentives for homeownership, such as mortgage interest and state and local property tax deductions, in place since 1913, and exclusions on capital gains from the sale or exchange of primary residence. Similarly, numerous institutions have been in operation, some since the end of the Great Depression, to facilitate housing finance, many of which with competing functions. For example, Ginnie Mae, Fannie Mae, and Freddie Mac offer mortgage securitization with either explicit or implicit government guarantees, competing with private sector security issuers and the regional Federal Home Loan Bank system (FHLBs).² Similarly, government insurance offered by the Federal Housing Administration (FHA) competes with private mortgage insurers.

View Full Size

A Schematic View of the Mortgage Intermediation Process

Citation: IMF Staff Country Reports 2010, 248; 10.5089/9781455206759.002.A005

Sources: U.S. Department of Housing and Urban Development (2006), Evolution of the U.S. Housing Finance System; author's estimates.

• Download Figure
• Download figure as PowerPoint slide

3. Housing finance in the United States is also very expensive for taxpayers (Figure 1). The U.S. Treasury estimates that tax expenditures on housing amount to $200 billion in FY 2010 (1.4 percent of GDP), with the mortgage interest deduction (the second most expensive tax expenditure) accounting for nearly half of the cost. For the medium term, housing-related tax expenditures are expected to rise to close to $330 billion by FY2015. These numbers do not include the implicit and explicit subsidies provided to the Government Sponsored Enterprises, with the Congressional Budget Office (CBO; 2001, 2004) estimating that in 2003, subsidies to Fannie Mae and Freddie Mac were around $25 billion (in 2006 dollar terms), three times their 1995 level;³ while FHLB system subsidies were estimated at around $3 billion in 2000.

View Full Size

Housing Finance in the United States and Other OECD Countries

Citation: IMF Staff Country Reports 2010, 248; 10.5089/9781455206759.002.A005

Sources: Australian Bureau of Statistics; Committee on the Budget, United States Senate (2008), Tax Expenditures: Compendium of Background Material on Individuals' Provisions; Haver Analytics; International Monetary Fund, World Economic Outlook; U.S. Census Bureau (1989), Historical Statistics: Colonial Times to 1970; U.S. Office of Management and Budget (2010), Analytical Perspectives, Budget of the U.S. Government Fiscal Year 2011; Yelten (2006); and author's calculations.

• Download Figure
• Download figure as PowerPoint slide

4. These large fiscal costs end up benefiting middle-to high-income households (Figure 1). The majority of the tax measures are not well targeted, and they are largely enjoyed by upper income households that itemize deductions. The richer households benefit disproportionately since they face higher marginal income tax rates and usually have larger mortgages. However, their decision to buy a house is little influenced by government financial incentives, since on the margin they are not fiscally constrained on their decision to buy a house (Glaeser and Shapiro, 2003).

5. It is thus unclear whether subsidies affect homeownership rates, although they raise housing consumption. Data suggest that there is no strong correlation between homeownership rates and subsidies to housing, both across time and in cross country comparisons (Figure 1). In fact, U.S. tax expenditures on housing and homeownership rates have been negatively correlated in the last two decades, while homeownership rates are much higher in some countries (e.g., Australia, Canada, Spain) than in the United States, despite less generous subsidies and simpler housing finance systems; for instance, neither Canada nor Australia have a mortgage interest deduction. It appears on the other hand, that the mortgage interest deduction affects housing consumption (i.e., the size of the house) with Australian mortgages around 40 percent lower than the ones in the United States. Van den Noord (2005) shows that a tax system that favors homeownership through subsidies and tax deductions raises the equilibrium price of housing and price volatility, using a sample of euro area countries. In addition, Lehnert et al. (2008) find little evidence that GSEs lower interest rate spreads or raise homeownership.

View Full Size

Homeownership Rates, latest available data

(in percent)

Citation: IMF Staff Country Reports 2010, 248; 10.5089/9781455206759.002.A005

Sources: OECD (2010), Economic Survey of the United States, forthcoming; U.S. Census Bureau; and author’s calculations.

• Download Figure
• Download figure as PowerPoint slide

C. Lessons from Other Countries

6. Experience from other countries indicates that considering a less complex and cheaper system for the United States is warranted. For example, other Anglo-Saxon countries such as Australia, Canada and the United Kingdom are enjoying higher homeownership rates than the United States, without having a mortgage interest deduction. Australia’s experience also indicates that public mortgage insurance is not necessary for a well-functioning housing finance system.⁴ Similarly, Canada’s explicit government guarantees on mortgage funding have shielded the system from the ambiguities related to the pre-crisis hybrid public/private status of the GSEs in the United States. In that respect, reforms to the GSE’s ambiguous public/private status, which proved unsustainable, could be considered, as well as streamlining their mandates and making explicit the guarantees on their business activities that provide public goods. Privatizing their retained portfolio, which bore little relation to the core bundling and guarantee businesses could also be considered.

7. Reforming tax incentives in housing finance has been successfully undertaken in many countries. The importance of subsidies to homeownership, notably of tax deductibility of interest payments, has decreased over time in many countries; for instance, the effective marginal tax rate at which mortgage interest tax relief can be claimed was reduced over time in Ireland from 47 percent to 20 percent and in Denmark from about 33.5 percent to 25.5 percent above a certain threshold, and has been progressively reduced in the United Kingdom over 12 years and completely eliminated in 2000, with minimal implications for the housing market and homeownership rates (Figure 1).⁵

8. In contrast, attempts thus far to reduce the generosity of the U.S. mortgage interest deduction have not borne fruit. In 2005, a bipartisan tax reform commission, proposed ending the mortgage interest deduction, but the plan stalled in Congress. In 2009, the Administration proposed cutting the deduction rate for itemized expenses for those making more than $250,000 to the rate paid by the middle class, but again it stalled in Congress; a similar proposal is pending again this year. Such reform would result in significant savings: the CBO (2009) recommended that by reducing the $1 million cap on the size of mortgage for which interest is deductible by $100,000 a year beginning in 2013 and ending at $500,000 in 2018, the Administration could generate $41.4 billion in additional revenues over 10 years. Alternatively, by changing the mortgage interest deduction to a 15 percent tax credit on mortgage interest for everyone with mortgage amount below the declining limits in the aforementioned option, revenues would increase by $387.6 billion over 10 years.

9. The pace and timing of any adjustment need to bear in mind the state of the housing market. The disruptive experience in Sweden in the early 1990s and the favorable experience in the United Kingdom with gradual phasing out of interest relief for homeowners suggests that spreading out reforms of housing finance over time could avoid large disruptions to housing markets and minimize any adverse macro-financial implications. It is also important to choose a period when the housing market is in a relatively healthy state. For example, when Sweden first reduced the maximum deductible tax rate that could be applied to mortgage interest payments to 50 percent (from 80 percent) in 1985, the housing market was not hurt; indeed, real house prices experienced spectacular growth since the country was in the midst of strong economic growth. However, when the rate was subsequently reduced further to 30 percent in 1991—in the midst of one of Sweden’s worst recession—the housing market and economic activity tumbled (with nominal house prices declining by an average annual rate of 10¼ percent in 1992–93).

D. Conclusions and Policy Implications

10. The U.S. Administration is appropriately considering fundamental reforms to its housing finance system. The road to reform would be complicated given that most measures and institutions have been in place for almost a century. However, lessons could be learned from other countries that have successfully phased down government intervention in housing finance. In general, gradually moving away from the current complex, regressive and costly financing system would be desirable.