February 12, 1996
Last year, the National Association of Realtors hired Data Resources Inc./McGraw-Hill, a leading economic firm, to analyze the economic effects of a flat tax. The resulting study concluded that the enactment of this tax, which would treat all taxpayers and all income equally, would result in a 15 percent drop in housing values, 1due largely to the elimination of the home mortgage interest deduction. This study is now being used by opponents of the flat tax. Yet numerous experts have pointed out that the DRI/McGraw-Hill study has serious methodological flaws that make its conclusions highly suspect. A more reasonable method of analysis-one that looks at the historical relationship between personal income growth and the value of owner-occupied housing, combined with the range of estimates showing faster economic growth under a flat tax-shows that home values actually would rise by as much as 7 to 14 percent by the fifth year after enactment of a flat tax.
These results should come as no surprise. As Chart 2 illustrates, home values consistently have mirrored changes in personal income. When the economy is strong and incomes rise, consumers spend more on almost all goods and services. As a major component of family budgets, housing has followed this pattern. Although projecting the value of housing into the future involves a level of uncertainty, a wide range of studies indicate that adoption of a flat tax would boost the annual rate of growth in gross domestic product (GDP) by at least 1 percentage point for a period of five or more years.3 Because personal income would rise with increases in GDP, it is possible to use the strong historical relationship between income and house values to generate estimates of how home values will change under the flat tax.4
Critics of the flat tax argue that it is improper to use historical data to predict the future value of homes under a flat tax because, they say, the relationship between home values and personal income exists only in a world where mortgage interest is deductable. Yet if the deduction were indeed the critical variable, there should be a strong relationship between housing values in the past and the tax value of the deduction. In 1980, before the Reagan marginal tax cuts, for instance, a dollar of home mortgage interest cost an upper-income taxpayer only 30 cents. In 1988, that same taxpayer paid 72 cents out of pocket for every dollar in interest payments. As Chart 2 indicates, however, the tax value of the deduction had no discernible impact on the value of housing over the period.
Why the Realtors’ Study Is Flawed
If the historical evidence and economic projections suggest that a flat tax will increase home values, why did the realtors’ study predict a sharp decline? The reason is that it used several very questionable assumptions. For example:
DRI/McGraw-Hill analyzed a flat tax with an initial rate of 25 percent, gradually phasing down to 21 percent. The DRI/McGraw-Hill study used a much higher tax rate than is used in existing flat tax plans. Crucially, the lower the rate, the larger the economic gains under the a flat tax, and the higher the positive impact on home values. Even the Clinton Administration’s Treasury Department, no friend of the flat tax, refused to use a rate higher than 21 percent when estimating the effects of a flat tax on the economy.
DRI/McGraw-Hill did not properly estimate the drop in interest rates under a flat tax. According to the Kansas City Federal Reserve Bank, interest rates would drop by 25 percent to 35 percent under a flat tax.5Other experts have similar estimates.6 DRI/McGraw-Hill stands alone in basing its estimate of likely interest rate declines-and thus the positive effect on home values-on the comparison between municipal bond rates and 10-year federal bonds, rather than interest-bearing assets of equal risk such as high-grade corporate bonds.
DRI/McGraw-Hill overestimated by 30 percent the average marginal tax rate which applies to homeowners. Doing so significantly overstates the tax benefits of the deduction and exaggerates the change in a taxpayer’s cost of mortgage under the flat tax.
The study assumed that financial assets would lose value under a flat tax. This assumption is at odds with academic analyses of the flat tax and defies common sense. Lower taxes on savings and investment actually would boost the value of financial assets by increasing the after-tax stream of income they produce.
DRI/McGraw-Hill failed to include the impact of higher disposable incomes on home values. The study does conclude that the flat tax would increase economic growth but fails to factor back in the effects on house prices. This is inexplicable.
The flat tax represents a direct threat to powerful vested interests that feel they benefit from today’s deductions. Unfortunately, many of these special interests are willing to disseminate erroneous information to Americans in an effort to preserve the tax breaks they now enjoy. The DRI/McGraw-Hill study is based on an unsound methodology which leaves it alone in its dire predictions that housing values will fall. Historical evidence on the relationship between performance of the economy and house prices, however, clearly shows us that a flat tax would be a bonanza for homeowners.
1. Roger Brinner, Mark Lasky, and David Wyss, “Market Impacts of Flat Tax Legislation,” DRI?McGraw-Hill U.S. Review, June 1995.
2. For example, see Bruce Bartlett, “Will the Flat Tax KO Housing?,” The Wall Street Journal, August 2, 1995
3. See, for example, Alan Auerbach, “Tax Reform, Capital Allocation, Efficiency, and Growth,” unpublished paper, December 21, 1995, forthcoming, Brookings Institution, February 15, 1996; Michael J. Boskin, ed., Frontiers of Tax Reform (Stanford, Cal.: Hoover Institution Press, 1996); Dale Jorgenson, “The Economic Impact of Fundamental Tax Reform,” Frontiers of Tax Reform Conference, Hoover Institution, May 11, 1995; and Laurence J. Kotlikoff, “The Economic Impact of Replacing Federal Income Tax with a Sales Tax,” Cato Institute Policy Analysis No. 212, April 1993
4. Heritage economists projected changes in housing values as a function of changes in personal income and gross domestic product. Regressions of personal income against GDP and housing values against personal income produced r2s of .93 and .99, respectively. The F and t statistics were significant at .05 and .025, respectively.
5. John E. Golub, “How Would Tax Reform Affect Financial Markets?,” Federal Reserve Bank of Kansas City Economic Review, Fourth Quarter, 1995
6. Robert Hall and Alvin Rabushka, The Flat Tax (Stanford, Cal.: Hoover Institution Press, 1995).