Special Studies, July 1, 2008
By Robert Dietz, Ph.D.
Report available to the public as a courtesy of HousingEconomics.com
The major tax preferences benefiting homeowners are the mortgage interest and real estate tax deductions, which allow home-owning taxpayers who itemize their deductions to reduce their income tax liability. In the tax policy community, a common, although misleading, criticism of these housing tax preferences is that they are claimed by a relatively small number of taxpayers and the benefits accrue mostly to higher-income taxpayers. When viewed relative to the reporting of taxable income, the distribution of tax liability, and the use of other tax preferences, these claims lack merit. These inaccurate observations also lead to flawed conclusions regarding the distribution of impacts associated with these housing deductions. Building on analysis provided in a previous HousingEconomics.com publication, this article demonstrates that the homeowner tax preferences make the income tax system more progressive and are justified by economic and tax policy principles.
Number of Benefiting Taxpayers
An often-reported estimate finds that only 25% of taxpayers benefit from the mortgage interest deduction. This is a misleading statistic for several reasons. First, according to the Congressional Budget Office, 94% of all mortgage interest payments are in fact deducted for federal income tax purposes. Why the disparity between these two estimates? In addition to the obvious exclusion of renting taxpayers, many homeowners have no outstanding mortgage. Indeed, according to the Census Bureau’s 2006 American Community Survey, 32% of all owner-occupied homes have no mortgage. With no mortgage, these homeowners clearly do not deduct (non-existent) mortgage interest payments. Of course, many taxpayers do not itemize (approximately 64% of taxpayers in 2005, according to IRS Statistics of Income [SOI] data).
This leaves the inability to itemize as the primary reason for the 6% of mortgage interest payments that are not deducted. Nevertheless, the vast majority of mortgage interest is deducted because home-owning taxpayers are much more likely to itemize their deductions than renting taxpayers (63% for home-owning taxpayers vs. 36% for all taxpayers). Combining these results, it is easy to reproduce the 25% estimate cited earlier. Assume that 60% of taxpayers are homeowners (the homeownership rate is 67.8%, but some households file multiple returns), and multiply that estimate by 68% (the percentage of homeowners with a mortgage) and 63% (the percentage of homeowners who itemize). This calculation yields 25.7%.
However, the calculation itself demonstrates the limits of this statistic as a measure of taxpayer benefit because it requires accounting for the situations in which taxpayers have no reason to expect benefits from the mortgage interest deduction. The relevant number (i.e. 94% of mortgage interest payments are deducted) proves that the deduction, in general, accomplishes its purpose of allowing home-owning taxpayers to deduct interest expense, and as a matter of tax policy, interest expense is a permitted deduction because it is a cost of investment.
Moreover, it is important to note that decisions regarding renting, homeownership, and mortgage status/size are linked to an individual’s age and lifecycle (e.g. married, children, etc). In the stylized case, an individual begins his/her working years as a renter, accumulating savings for a down payment on a home. In time, the individual purchases a home and his/her annual mortgage interest payments are at a lifetime maximum because in the early years of a self-amortizing mortgage most of the housing payment constitutes interest and not principal. As the homeowner ages, he/she pays less mortgage interest in each payment and may in time cease to itemize their deductions. This lifecycle-model gives rise to the following stylized presentation of lifetime annual mortgage interest deductions in Figure 1.
This lifecycle is the reason for the failure of the 6% of mortgage interest payments to be itemized later in life, but it also illustrates another weakness of the 25% estimate. Taxpayers benefit from the homeownership tax deductions at specific times during their lives. There is clearly a time when an individual rents a home and a time when an individual considers homeownership. The 25% estimate reflects a combination of taxpayers who are not currently benefiting because they are renting or have significantly or completely paid down their mortgage. To claim that the 25% estimate suggests that very few homeowners benefit from the housing tax preferences is thus inaccurate.
As an analogy, consider the following non-housing example. The 2005 IRS SOI data reveal that only 8 million taxpayers benefited from the tax code’s interest deduction for student loans (another form of investment expense – this one associated with human capital). This represents approximately 6% of all taxpayers. Nonetheless, the student loan interest deduction is, like the mortgage interest deduction, a tax preference claimed at a particular time in an individual’s life, and does not represent a tax preference that benefits only a narrow set of taxpayers, despite its low number of claimants in a single year.
The lifecycle aspect of homeownership also produces another interaction with the housing tax preferences. It is often claimed that the mortgage interest deduction encourages homeowners to purchase a larger home. While there may be some validity to this claim, it is also the case that this effect works in both directions due to lifecycle considerations. Homeowners with a larger family (e.g. children) need a larger home and thus will have a larger mortgage interest deduction. The need for a larger home created the larger mortgage interest deduction in this case, not the other way around. Moreover, some households file multiple tax returns, with mortgage interest payments only claimed on one tax form. This causes the deduction, which benefits the entire household, to be concentrated on one tax form.
Together, these considerations recommend caution with respect to using aggregate tax return statistics to evaluate the economic merits and distributions of provisions that are ultimately targeted to households.
The IRS SOI data also illustrate that it is incorrect to claim that the mortgage interest and real estate tax deduction benefit higher-income taxpayers at the expense of lower-income taxpayers (i.e. make the income tax system less progressive). A progressive tax system is one for which low-income taxpayers pay a smaller percentage of their income in taxes than high-income taxpayers pay. A policy that reduces tax liability for low-income taxpayers lowers their average tax rate and thus makes the income tax system more progressive. The housing tax preferences have this effect on the federal income tax system.
Using the 2005 IRS SOI data, Figure 2 demonstrates the distribution by Adjusted Gross Income (AGI) classes of the mortgage interest deduction relative to all other itemized deductions. First note that the mortgage interest deduction is claimed more than proportionally relative to AGI for taxpayers with AGI of $200,000 or less and less than proportionally for taxpayers with AGI $200,000 or more. This means that mortgage interest is more likely to be claimed by lower-income taxpayers than higher-income taxpayers than their reporting of income would suggest. Figure 2 also shows that the mortgage interest deduction is claimed in greater proportion (relative to AGI) for taxpayers with AGI of $20,000 to $200,000 than all other itemized tax deductions, thereby demonstrating that the mortgage interest deduction is more progressive in income distribution terms than the set of all other itemized deductions.
With respect to the set of all other deductions, why is the mortgage interest deduction more progressive? This distributional consequence arises because of two factors. First, the payment of mortgage interest as a percentage of income (AGI) declines as income increases. In other words, higher-income households are more likely to purchase larger homes but not one-for-one relative to the increase in income. That is, someone whose income increases 100% is unlikely to purchase a 100% more expensive home. Furthermore, higher-income individuals are more likely to obtain a proportionally smaller mortgage, thereby generating a lower loan-to-value (LTV) ratio. These effects produce the pattern reported in Figure 3, which illustrates that the aggregate amount of mortgage interest deducted (by AGI class) tends to fall (as a percentage of aggregate income) as income increases. In fact, for the top income class ($1 million AGI or higher), this percentage approaches zero.
Second, the effect described above tends not to be observed for other large deductions. Consider the deduction for state and local taxes. Except for the lowest income class, the deduction tends to be relatively constant as a percentage of AGI. That is, it does not exhibit the downward slope of the mortgage interest deduction, as seen in Figure 3. As a more dramatic example, examine the deduction for investment interest expense, which is essentially the deduction for all interest expense reported by individual taxpayers, except for interest allocable to mortgages and student loans. This deduction is mostly associated with higher-income taxpayers with business investments. Not surprisingly, as seen in Figure 3, the deduction exhibits a positive slope with respect to the amount of the deduction relative to AGI, with higher-income taxpayers reporting proportionally more of the deduction relative to income.
The consequence of the interplay of these two factors is that the mortgage interest deduction is reported in more progressive terms relative to other deductions because it is used less than proportionally relative to AGI.
Figure 4 illustrates similar income distributions properties for the deduction for state and local real estate taxes assessed against owner-occupied homes. Like the mortgage interest deduction, the real estate tax deduction is claimed more than proportionally, relative to AGI, for taxpayers with AGI up to $200,000. The real estate tax deduction is also claimed more than proportionally to AGI than the set of all other itemized deductions for taxpayers with AGI between $40,000 and $500,000. The real estate tax deduction is thus somewhat less progressive than the mortgage interest deduction relative to all other itemized deductions, but this is in part due to the fact that the mortgage interest deduction itself is included in the “all other deductions” measure.
Given this interaction, as in the previous HousingEconomics.com article on this topic, it is necessary to report a combined or simultaneous estimate to examine the distributional consequences of the combined set of the homeowner tax deductions. Figure 5 presents this result and demonstrates that the housing deductions, in tandem, primarily benefit taxpayers with AGI between $40,000 and $200,000 relative to the set of all other itemized deductions. Like the separate, individual estimates detailed above, the combined set of housing tax preferences is claimed more than proportionally relative to AGI for taxpayers with AGI less than $200,000.
Ultimately, however, it is the reduction in tax liabilities generated by these deductions that matters most when examining the distributional consequences of the itemized deductions. Figure 6 records the tax expenditure, or tax reduction benefit, produced by the housing tax preferences as aggregated by AGI classes. Indeed, Figure 6 illustrates the critical point when considering the income distribution impacts of the housing tax preferences. Relative to actual taxes paid, the housing tax preferences increase the progressive nature of the federal income tax. This can be seen by noting that the housing deductions’ tax expenditures are accrued in greater proportion than final tax liabilities for AGI classes up to $200,000. In other words, if these tax expenditures were eliminated, the income tax system would become less progressive. Moreover, these deductions are more progressive than the set of other itemized deductions, being claimed more than proportionally for taxpayers with up to $500,000 of AGI.
Despite the existence of multiple, cautionary issues concerning the measurement of the tax expenditures for owner-occupied housing and their distributional impact, it is nevertheless true that homeowners receive favorable tax treatment due to the various housing tax preferences. To conclude this paper and the previous paper on housing tax expenditure analysis, the following question is posed. What is distinct about housing and homeownership that justifies such treatment on economic grounds?
In general, homeownership, when attained at the appropriate moments of a household's lifecycle, offers documented benefits for individuals and communities. To the extent that the benefits of homeownership are realized not just by the homeowner but also by the homeowner’s community, homeownership yields what economists call a positive externality, and economic theory permits tax preferences that encourage such beneficial economic outcomes.
The benefit from homeownership occurs because ownership confers upon an individual or household a set of property rights for a dwelling. A homeowner thus possesses a financial stake, and a corresponding responsibility, with respect to the status of the dwelling and its surroundings. However, distinguishing homeownership from other forms of investment, this financial claim is defined by its fixed geographic location. Hence, a component of the financial responsibility associated with homeownership is a concern for the neighborhood of the home.
The social science research literature detailing the impacts of homeownership includes documented studies from the social sciences, medicine, psychology, and other fields. Overall, economists, sociologists and other social scientists have found significant, positive homeownership-related impacts on a large set of outcomes associated with households and communities.
This research finds a homeowner’s stakeholder status encourages the owner to take an active role in bettering their neighborhood. As an example, research indicates that homeowners are more likely to maintain and improve their homes and be involved in local social and political organizations and activities. As a result, communities with large numbers of homeowners possess increased public amenities, reduced crime and greater environmental awareness.
Moreover, homeowners are more likely to improve and maintain their residences due to their ownership stake. As a result, neighborhoods of homeowners tend to be better maintained, leading to more attractive and safer communities. To the benefit of their neighborhoods, home-owning households are more socially involved in community affairs. This is due to both the fact that homeowners expect to remain in the community for a longer period of time and that homeowners have an ownership stake in the neighborhood. Dipasquale and Glaeser (1999) find that homeownership has a positive impact on the provision of a variety of types of social capital including membership in non-professional organizations, knowledge of local officials, voting in local elections, involvement with community problems, church attendance and home security.
Homeowners are often involved in community activism and local affairs. Furthermore, homeowners are also more likely to voice dissatisfaction to local government and political agencies. Perkins et al. (1996) demonstrate that homeowners are more likely to be involved in grassroots community organizations. Homeowners are likely to be sensitive to local activities that generate negative externalities or spillover costs. Indeed, these households are more likely to participate in political activity to prevent the location of such nuisance activities in their communities. For these and other positive impacts, homeownership has a favorable place in the tax code.
While the tax exclusion for net imputed rent (discussed in an earlier article) or the deductions allowed for mortgage interest and real estate taxes can be justified on such positive externality grounds, the current treatment of owner-occupied housing can also be justified as a tax simplification measure. There are many examples in the tax code where theoretically-ideal tax treatment is not obtained because the compliance cost and taxpayer burden would rival or exceed the revenue gain for the Treasury. For example, tax theory requires any increase in wealth to be treated as income and to be a taxable event. However, as a matter of tax law, capital gain is not taxed upon accrual but upon realization. That is, taxpayers do not pay tax when stock or other asset values increase. Instead, taxes are only assessed upon sale of the asset. This tax treatment is present policy because the administrative cost of taxing capital gains upon accrual would be excessive.
Similarly, taxing imputed rent would be a complicated task for both the government and individual homeowners. Comparable issues arise for taxing capital gains attributable to the sale of a principal residence. Because of these administrative and compliance burdens, the tax exclusion of net imputed rent for owner-occupied housing and the present-law exclusion of capital gain income from the sale of a principal residence can also be justified on tax simplification grounds. Combined with the general tax policy principle permitting deductions for investment expense, including taxes and interest, these administrative issues and the benefits conferred by homeownership justify the present-law tax treatment of owner-occupied homes.
This article has estimated the distributional consequences of the major housing tax preferences. The analysis reveals that these deductions make the income tax system more progressive, not less as is commonly claimed. Furthermore, this article concludes that the tax expenditures for owner-occupied housing are justified in terms of public policy due to the positive externalities generated by homeownership, as well as tax administration and simplification grounds.
 The local use of these deductions was previously examined in The Local Use of the Mortgage Interest and Real Estate Tax Deductions. HousingEconomics.com, June 9, 2006. http://www.nahb.org/generic.aspx?genericContentID=58984
 An example: Who Benefits from the Home Mortgage Interest Deduction? Tax Foundation, February 6, 2006.
 Housing Tax Expenditure Analysis. HousingEconomics.com, May 27th, 2008. http://www.nahb.org/generic.aspx?sectionID=734&genericContentID=96447&channelID=311
 This estimate can be calculated from the annual reporting of taxpayers claiming the mortgage interest deduction relative to the total number of taxpayers from the Joint Committee on Taxation Tax Expenditure report - http://www.house.gov/jct/s-3-07.pdf. The most recent: Estimates of Federal Tax Expenditures for Fiscal Years 2007-2011. September 4, 2007. Joint Committee on Taxation. JCS-3-07.
 Table A-4, footnote 3, page 31 of Taxing Capital Income: Effective Rates and Approaches to Reform.
 Tax Expenditures for Owner-Occupied Housing: Deductions for Property Taxes and Mortgage Interest and the Exclusion of Imputed Rental Income. Jim Poterba and Todd Sinai. Working Paper. http://real.wharton.upenn.edu/%7Esinai/papers/Poterba-Sinai-2008-ASSA-final.pdf
 See footnote #3.
 See footnote #3.
 This kind of economic policy is called Pigovian tax policy, named after the English economist Arthur Pigou, who recommended taxing negative externalities (that impose costs on 3rd parties) and subsidizing positive externalities (that impose benefits on 3rd parties). Unfortunately, many tax economists view Pigovian policies with skepticism, seeing the primarily role of the tax system as raising money for the government and not pursuing other legitimate policy objectives.
 Two recent, comprehensive literature reviews detailing the impacts of homeownership are:
W. M. Rohe, G. McCarthy, S. Van Zandt, The social benefits and costs of homeownership: A critical assessment of the research, Research Institute for Housing America, Working Paper No. 00-01 (2000).
R. Dietz and D. Haurin, The social and private micro-level consequences of homeownership, Journal of Urban Economics 54 (2003) 401-50.
 N. S. Mayer, Rehabilitation decisions in rental housing: An empirical analysis, Journal of Urban Economics 10 (1981) 76-94.
 D. DiPasquale, E. L. Glaeser, Incentives and social capital: are homeowners better citizens? Journal of Urban Economics 45 (1999) 354-384.
 K. R. Cox, Housing tenure and neighborhood activism, Urban Affairs Quarterly 8 (1982) 107-29.
 D. D. Perkins, B. B. Brown, R. B. Taylor, The ecology of empowerment: Predicting participation in community organizations, Journal of Social Issues 52 (1996) 85-110.
 See footnote #3.