Led by a steep 31.7 percent decline in multifamily production, nationwide housing starts fell 14.4 percent to a seasonally adjusted annual rate of 956,000 units in August, according to newly released figures from the U.S. Department of Housing and Urban Development and the U.S. Census Bureau. Single-family housing starts dropped 2.4 percent to a seasonally adjusted annual rate of 643,000 units. More.
In-Depth Analysis, May 5, 2006 By Tiffany Chaney and Paul Emrath
Estimating the effects of a change in US housing policy can be both complicated and difficult. As a shortcut, it is occasionally tempting to compare the US to some other country. For example, when a change in tax policy is proposed, housing in the US may be compared to housing in a country where the proposed change in tax law has already been adopted. Cross-county comparison of isolated housing statistics is not likely to be very informative, however, as housing markets and government policies differ along many dimensions. Factors such as the physical characteristics of the existing housing stock, social customs, geography, and demographics interact with policy variables in a complex way, and government policies themselves are frequently complex instruments that reflect a variety of political compromises, so looking at one factor while ignoring the others can be misleading.
A further difficulty in cross-country comparisons is a lack of consistency in the way data are collected and reported. It’s not always easy to obtain numbers that are truly comparable.
Recently, an attempt has been made to compile data in a relatively consistent format for countries in the European Union (EU). Eurostat (the umbrella statistical organization for the EU located in Luxembourg) has assembled housing data as part of its more general mission to provide statistics that enable comparisons between countries and regions.  And a recent working paper published by the European Central Bank has brought together information on mortgages, financial deregulation, and housing-related tax policies for countries in the EU.  It is impossible to eliminate all discrepancies from the data, and the data still are often not available for the same time period across all countries. Nevertheless, these new sources permit a somewhat better comparison of the EU and the United States than was previously possible. The purpose of this article is to summarize the data showing different ways housing markets in industrialized countries vary, and how difficult it is to make a valid inter-country comparison from a limited set of statistics. To illustrate this, we look specifically at the relationship between homeownership and tax policy.
Compared to most European countries, the supply of land is abundant in the United States. Population density in the US is only a fraction of the density in the EU overall. The only EU countries with a lower population density than the US are the ones that extend into the arctic circle—Sweden and Finland (Table 1).
Table 1. Economic and Housing Characteristics
Population per Square Kilometer
Rooms Per Housing Unit
Persons Per Room
Share of Occupied Units Built Last 10 Years
Home Ownership Rate
Growth in Real House Prices
GDP Per Capita
Sources: US Bureau of the Census, Eurostat, CIA World Factbook (1) Data from 2005 (2) Data from 1991 except US data from 2004 (3) Data from 2001 except US data from 2004 (4) Data from 1981-1991 except US data from 1989- 1999 (5) Data from 2001 except US data from 2005 (6) Data from 2003 except Austria (2002), Luxembourg (2000) and US (2004-2005) (7) Data from 2005 in US dollars (2-5) EU data from Eurostat (6) EU data from European Central Bank, Working Paper Series; No. 526, September 2005 (1& 7) data from CIA World Factbook (2) US data from US Bureau of the Census, 2004 American Community Survey, Table B25019, Table B25020 (3) US data from US Bureau of the Census, 2004 American Community Survey, Table B25019, Table B09016 (4) US data from US Bureau of the Census, 2004 American Community Survey, Table B25036 (5) US data from US Bureau of the Census, Housing Vacancies and Homeownership, Fourth Quarter 2005
A reasonable hypothesis is that homes would tend to be larger in countries where land is more readily available. Although data on the size of homes in square feet (or square meters) is not available for the EU, there is information about the number of rooms per housing unit. If added to the list of EU countries, US ranks first with 5.6 rooms per unit. Ireland, the EU country with the lowest population density after Finland and Sweden, is second with 5.3. Overall, however, there is no measurable correlation between population density and the size of homes measured by the number of rooms. The countries with fewer than 4 rooms per house (Austria, Finland, and Greece) are not very densely populated by European standards. Number of rooms is not a perfect proxy for house size, of course, as the average size of the rooms may also vary from country to country.
The number of rooms per home is sometimes combined with the number of people living in the homes to develop a measure of overcrowding, which in turn is assumed to be an indicator of inadequate housing. In the United Kingdom there is legal definition of overcrowding that takes into account factors such as marital status, the age of the occupants, and the size of the rooms, in addition to persons per room. In the US, the simple measure of 1.00 or more persons per room has been used as a less formal definition of overcrowding for several decades. By that standard the occupied housing stock is not, on average, overcrowded in any EU country. Greece comes closest with 0.71 persons per room, despite the fact that it has one of the lower population densities of the countries in Table 1.
Like house size, overcrowding is not consistently associated with population density the way we might expect. The Netherlands has the highest population density of any EU country, but the least crowded living conditions (0.38 persons per room).
In addition to land availability and cultural factors that impact household and house size, overcrowding may be related to income. It seems reasonable to suppose that, all else equal, people with larger incomes would buy larger homes, and in general the countries in Table 1 tend to fit that pattern. The US has the second highest GDP per capita of any country in the table and the third lowest number of persons per room. Greece combined the largest number of persons per room with the second lowest per capita GDP.
Another measure often regarded as an indicator of how well off households are is the homeownership rate. By this measure, the US ranks roughly in the middle of the pack of EU countries with a homeownership rate of 69 percent (Figure 1). However, homeownership is not the only possible indicator of well-being. In Greece, for example, the homeownership rate is very high (over 84 percent) in comparison with the other countries. However, Greece also has a low GDP per capita, some of the smallest homes in terms of rooms per unit, and the highest level of overcrowding. Italy displays similar economic and housing characteristics, although not quite to the extreme degree Greece does. The US falls well below Greece and Italy in terms of homeownership, but the US is at or close to the top of the list in GDP per capita, rooms per house, and rooms per person in the house. Few analysts would be willing to validate a claim that citizens of Greece and Italy are better off than citizens of the US, despite the relative homeownership rates.
A particular housing variable that seems to be positively correlated with homeownership rates is the percentage growth in real (adjusted for inflation) home prices. In particular, the two countries where real house prices fell by more than two percent (Germany and Austria) are at the bottom of the homeownership chart in Figure 1. There are many economic and policy-related reasons this could occur. On the policy front, proposed changes that tend to depress real estate values may face less political opposition in countries where relatively few people own their own homes. On the other hand, low homeownership and declining prices could be evidence of policy changes that made home owning less attractive. Without some other kind of evidence, it’s difficult to say which way the causal relationship runs. Moreover, low homeownership and declining house prices can be the result of economic forces that have little to do with policy decisions.
Taxes and Homeownership
Economic factors that can have an effect on the homeownership rate include, income and wealth, the price of housing, interest rates, and tax treatments. All else equal, households with greater income will more easily be able to afford a home of a given price. The price of a home will be related to factors such as, size and amenities of the homes and the scarcity of land. The after tax cost of a home will depend not only on the house price, but also prevailing interest rates and how owner-occupied housing is treated in the tax code.
The US income tax code gives owner-occupied housing advantages such as a mortgage interest deduction, a limited exemption of capital gains taxes on homes, and a property tax deduction. Mortgage interest is deductible on a homeowner’s first and second homes with a million dollar cap on the loan. Home equity interest is also deductible on a loan of up to $100,000. Capital gains taxes are not applicable to the sale of a home in the US on principal residences up to a $500,000 gain (for a married couple). The federal property tax deduction is a partial offset to a tax disadvantage for housing, as most local jurisdictions in the US levy a tax on residential property. Most types of state and local taxes are deductible on the federal income tax forms.
In 2006, Congress estimates that the tax expenditure (that is, the extra revenue the U.S. Treasury would realize if a certain aspect of the tax code were changed and tax payers didn’t change their behavior as a result) associated with the mortgage interest deduction is $402.7 billion over five years. The exemption of a capital gains tax on a home is second at $128.4 billion, and the property tax deduction comes next at $73.8 billion.
How does this compare to the treatment of housing in the tax systems in Europe? In the European Union, the tax code varies considerably. All but three of the 15 countries have a mortgage interest deduction, but the limitations and marginal tax rates vary greatly. Each country has some form of capital gains tax exemption on housing. Rather than a property tax, some of the EU countries have a tax on imputed rent (Table 2). Where this tax is used, the justification for it is an attempt to treat the housing services that flow from owning a home like the rate of return that flows from other types of investment.
Tax deductible as special expense up to a limit that goes to 0 as annual income increases
Tax deductible up to income from immovable property, but some additional deduction possible. Deduction decreases over time, for at most 12 years
Deductible from captial income that is subject to income tax, unlimited
29% (flat rate)
Normally deductible from capital income that is subject to income tax
Gradually abolished over the period 1991-2000. Only tax credit in very special cases for loans before 1998
Abolished since 1986 with the introduction of a subsidy scheme (Eigenheimzulage)
Since 2003, tax credit of 15% for annual mortgage interest, subject to limits
"Tax credit" at 20% standard income tax rate, at source, with (low) limits
Tax credit of 19% for annual mortgage interest, with limit
Tax deductible, with (low) limits
Tax deductible without limit, for at most 30 years
Tax credit of 30% of mortgage interest and amortisation, with a limit
Tax credit of 25% of amounts paid (principal and interest) in first two years, 20% thereafter up to a threshold, and 15% for payments above the threshold
Interest not attributable to any source of income is deductible from capital income
Abolished in a number of steps (1983, 1988, 1991, 2000)
Qualified residence (1) interest expenses incurred on up to 1 mil. (500,000 for married filing seperately) deductible.
(1) Qualified residences are up to 2 residences occupied for 2 out of the last 5 years. Guido Wolswijk, "On Some Fiscal Effects on Mortgage Debt Growth in the EU," Working Paper Series no. 526, European Central Bank, September 2005
Imputed rent is the rent that would have to be paid, if the home was rented out rather than owner occupied. In owner-occupied housing, the tenant and the landlord are the same person, so imputed rent is an estimate of the rent that would have been paid to the landlord, had the tenant and the landlord been different people.
Imputed rent is an estimate of what property of a certain value would rent for, so the tax on imputed is ultimately a fraction the property’s estimated value. In that sense a tax on imputed rent is somewhat similar to a property tax. In the US, as most readers know, property taxes are collected by local governments, and these taxes are then deductible on federal income tax forms. Of the fifteen EU countries, only Belgium, Italy, Luxembourg, and the Netherlands collect a tax on imputed rent.  The European Central Bank reports that the total tax levy in these in these countries is fairly small, because imputed in rent is based on out-of-date property values.
The EU countries that tax imputed rent have substantially different homeownership rates. None of the three EU countries with homeownership rates above 80 percent tax imputed rents, but that’s not enough evidence to prove that taxing residential property or imputed rents lowers homeownership.
Some European countries also offer partial or total relief from consumption taxes (which include General Sales Taxes and Value Added Taxes) on home sales.  In the US there is no value added tax or national general sales tax, although sales taxes are imposed in many states. These sales taxes generally do not apply to housing in the US, although some of them do apply to residential construction materials.
In general, the EU countries do not collect capital gains taxes on the sale of owner-occupied housing. Each country has either no capital gains tax, or housing is largely exempt from capital gains taxation. As a result, the European countries provide no evidence of a link between capital gains taxes and homeownership.
A mortgage interest deduction, however, is a feature of many of the countries’ tax codes. The United Kingdom, France, and Germany do not have a mortgage interest deduction in place, as seen in Table 3. Figure 1 shows that homeownership rate is not directly correlated with whether or not a country has a mortgage interest deduction.
It is possible that tax incentives could not only cause more people to buy homes, but cause them to buy homes that are larger with more amenities. The mortgage interest deduction may also impact the price of homes or the amount of new construction, as well as the homeownership rate. In the early 1990’s during the housing market crisis in Sweden, a dual income tax system was introduced that lowered the maximum tax rate for mortgage interest deductions. In addition, a tax on imputed rent was replaced with a property tax, a Value Added Tax was imposed on new dwellings, and subsidies were reduced to the construction sector. Home prices and construction rates were drastically affected, and suffered significant negative effects. One paper estimated that the reduced mortgage interest deduction was responsible for 10-15% out of a total of 25% of the decline in housing prices. Moreover, housing construction declined by 25%. 
In Figure 1 among the countries that do not have a mortgage interest deduction, the UK is the only one that is above the US in homeownership rate at 71.7%. Germany and France are both significantly below the US, at 43.6% and 63.1% respectively. All three countries previously had a mortgage interest deduction, as noted in Table 3. Germany’s was abruptly abolished, while the UK and France gradually abolished their mortgage interest deductions over a period of time very recently.
Figure 1 shows that countries with high homeownership rates have or recently had mortgage interest deductions, but this is still insufficient evidence to establish that a mortgage interest deduction causes high homeownership. The mortgage interest deduction interacts with many other aspects of countries’ tax codes, including how rental property is treated relative to owner-occupied units (something not addressed in the EU data sources on which this article is based).
In addition, tax policies are not static, but frequently change over time. Countries phase out certain policies, as the UK and France have done with their mortgage interest deductions. Moreover, marginal income tax rates may change significantly over time. Establishing an accurate connection between tax policy and the state of housing in a country’s economy can be thwarted by such changes in tax policies.
Also, many economic and social factors influence homeownership, and mortgage interest deductions provide only one of the reasons to own a home.
Consider a country such as Greece with a very high homeownership rate (84.6 percent) and a mortgage interest deduction. Greece also has small homes, a low GDP per capita, and the highest instance of overcrowding. It’s unclear that other countries would want to emulate Greece simply because of its homeownership rate.
Given not only the complex way many components of national tax codes impact housing, but also the differences in economic and housing characteristics among various countries, it’s difficult to say much about how one aspect of the US tax code (e.g., the mortgage interest deduction) impacts one aspect of US housing markets (e.g., the homeownership rate) by comparing the US to European countries.
 Guido Wolswijk, “On Some Fiscal Effects on Mortgage Debt Growth in the EU,” Working Paper Series no. 526, European Central Bank, September 2005. from another source. Return to the article
 Some EU countries have started adopting a property tax, according to a 1996 reference cited in Wolswijk. However, that reference is not readily available from any US library, and it has not been possible to track down more precise information from another source. Return to the article
 David Crowe and Karen Danielsen, Comparing Worldwide Tax Treatment of Housing, NAHB draft white paper, 2002. Return to the article
 P. Englund, “The Swedish Banking Crisis: Roots and Consequences”, Oxford Review of Economic Policy, 15, No. 3, 1999, pp. 80-97. Return to the article