Originally published in the NTA Encyclopedia of Taxation and Tax Policy, Second Edition, edited by Joseph J. Cordes, Robert D. Ebel, and Jane G. Gravelle. The encyclopedia is available in paperback from the Urban Institute Press. Order online at www.uipress.org or call toll-free 1-877-847-7377
William G. Gale
The Brookings Institution
A proposal for fundamental tax reform that would replace the income tax system with a consumption tax, to be collected by levying a flat-rate tax on businesses and individuals.
The Hall-Rabushka proposal has served as the blueprint for several proposals to reform the federal tax system, including a proposal introduced by Representative Richard Armey (R-Texas) and Senator Richard Shelby (R-Alabama) and one offered by presidential candidate Steve Forbes (R) in the 1996 presidential primaries.
In comparison to the Hall-Rabushka proposal, the personal income tax in 1994 provided exemptions of $9,800 for a family of four, an earned income tax credit (EITC), and the choice of a $6,350 standard deduction or itemized deductions for mortgage interest, state and local income and property taxes, charity, and large health expenditures. Estimates indicate that in 1996, a family of four taking the standard deduction and the EITC, with all income from wages, would pay no federal income taxes on the first $23,700 of income, 15 percent on the next $31,000 or so, 28 percent on the next $53,000, and higher rates on additional income, reaching 39.6 percent on taxable income above $250,000.
Without the personal exemptions, the flat tax is equivalent to a value-added tax (VAT), but with taxes on wages remitted by households rather than businesses. That is, the flat tax is a consumption tax, even though it looks like a wage tax to households and a variant of a VAT to most businesses. Therefore, other than the exemptions, the economic effects of the flat tax are essentially the same as those of a VAT or a sales tax.
The family exemptions make the flat tax progressive for low-income households. But at the high end of the income distribution, the tax is regressive, just like sales taxes and VATs.
The Hall-Rabushka proposal could be amended in several ways. Princeton economist David Bradford has proposed an X-tax similar to Hall-Rabushka but with graduated tax rates on household wage income to raise progressivity. (The business tax would be set equal to the highest tax rate on wage income.) The flat tax could also be modified to retain the EITC, allow a deduction for charitable contributions, and provide a tax credit (a one-to-one reduction in taxes paid under the flat tax) for payroll taxes paid. The credit would be a huge boon to lower- and middle-income households because most now pay more in payroll taxes than in income taxes. These changes would, of course, require higher rates. But a tax system with these features might be able to retain the progressivity of the current tax system while also reaping most of the gains of the Hall-Rabushka proposal’s broader base, generally lower rates, and simplified compliance. The question remains, though, of how large these gains would be.
Evaluating the effects of adopting a flat tax
Analysts find it hard to predict with precision the effects of minor tax changes, and heated debate continues about the effects of the major 1980s tax reforms. Hence, efforts to evaluate the effects of uprooting the entire tax system must be appropriately qualified. (The economic effects of the flat tax are addressed by a number of contributions in Aaron and Gale 1996 and Zodrow and Mieszkowski 2002.)
A central issue in tax reform is always who wins and who loses. Under the flat tax, low-income households would lose because they now pay no income tax and are eligible for a refundable EITC of up to $3,370. Although the flat tax is more progressive than a VAT, it is more regressive than the current system. A flat tax would provide huge gains for high-income households because their marginal tax rate would fall and because they consume less of their income than do low-income households. As a result, if a flat tax were to raise as much revenue as the current system, the tax burden for the middle class would have to rise. Consumption taxes are generally less regressive when viewed over longer periods of time because income changes from year to year, but they would raise tax burdens on lower- and middle-income households over any time frame. (For further discussion, see Gale, Houser, and Scholz 1996 and Gentry and Hubbard 1997).
Perceptions of fairness may also be difficult to retain when, under the flat tax, some wealthy individuals and large corporations remit no taxes to the government while middle-class workers pay a combined marginal tax rate above 30 percent on the flat tax, state income tax, and payroll taxes.
As for simplicity, the flat tax would likely slash compliance costs for many businesses and households. But for many, the tax system is not that complicated. And fundamental tax reform would not end the demands for special treatment that have so tangled the income tax. Ten years hence, we may find that what started as simplicity has once again become a confused jumble. Thus, some simplification is likely with tax reform, but it is by no means a certain or lasting outcome. Moreover, many simplification gains could be made through changes in the income tax.
A third concern is how reform would affect different sectors of the economy. Removing the mortgage interest deduction and the deductibility of state and local property taxes may have profound effects on housing prices and home ownership, but the results would depend on how interest rates adjust, the sorts of grandfathering rules introduced, and other factors. How and when health insurance benefits and coverage rates would adjust after the tax-favored treatment of employerprovided health benefits is eliminated is an open question.
Removing the deduction for charitable contributions would reduce overall giving and could affect its composition as well: wealthy donors, for whom the write-off is now worth the most, tend to favor hospitals and universities; low-income donors, religious institutions.
Effects on businesses and investment would be complicated. The flat tax would eliminate corporate income taxes, put all businesses on an equal tax footing, reduce the statutory tax rate applied to business income, and make investment write-offs more generous. But it would also remove the deductibility of interest payments and of state and local taxes, and this could induce dramatic changes. For example, Hall and Rabushka estimate that General Motors’ annual tax liability could rise to $2.7 billion from $110 million, while Intel’s would fall by 75 percent. The effects of a consumption tax on international economic transactions and on the financial sector are potentially far-reaching and need to be examined carefully.
Economic efficiency and growth
Ultimately, increased economic efficiency and growth must be one of the key selling points of a consumption tax. Without a significant gain in living standards, uprooting the entire tax system is probably not worth the risks, redistributions, and adjustment costs it would impose. Efficiency gains might arise from five sources: the change of the tax base from income to consumption; a more comprehensive tax base, which eliminates the differential tax treatment of various assets and forms of income; lower tax rates, which raise the rate of return to working, saving, and investing and reduce incentives to avoid or evade taxes; reduced compliance costs; and the taxation of previously existing assets during the transition to a consumption tax. All but the first and last are attainable under income tax reform.
Although estimates vary, a recent study suggests that a pure flat tax proposal with limited personal exemptions would raise economic output by between 2 and 4 percent over the first nine years and between 4 and 6 percent in the long run. But these results need to be interpreted carefully. First, many of the gains are also available through judicious reform of the income tax, in particular by making the taxation of capital income more uniform. Second, the estimates provided do not allow for child exemptions, as the Hall-Rabushka proposal and all of the recent flat tax proposals do. Allowing exemptions for children reduces the effects by about 2 percentage points (e.g., a gain of 0 to 2 percent over 10 years). Third, the estimates apply to a pure, well-designed consumption tax. Compromises in the design, such as including mortgage interest deductions or allowing a transition, reduce the gains or turn them into losses.
Allowing for transition relief alone is enough to eliminate the impact on growth in the long run. The estimates also show that, even for well-designed consumption taxes, efficiency losses are possible. (The estimates cited in this paragraph are taken from Altig et al. 2001 and private communications with Kent Smetters. For additional analysis of the growth effects of tax reform, see Engen, Gravelle, and Smetters 1997.)
A key element in raising growth and a major motivation for tax reform is increasing saving. Proponents of the flat tax typically point to two reasons why consumption taxes should spur saving. First, a revenue-neutral shift to a consumption tax should raise the after-tax rate of return on saving, while keeping total tax payments constant. Second, consumption taxes reallocate after-tax income toward high-saving households. Such reasoning is straightforward but incomplete. Saving is likely to rise only a little, if at all, for several reasons.
First, the current U.S. tax system is not a pure income tax; it is a hybrid between a consumption tax and an income tax. About half of private savings already receive consumption tax treatment. Funds placed in pensions, 401(k) plans, Keogh plans, and most individual retirement accounts (IRAs) are not taxed until they are withdrawn. The return on these investments, then, is the pretax rate of return. But introducing a consumption tax would reduce the pretax interest rate, so that the rate of return on these forms of saving would fall, which could reduce saving.
Second, pension coverage could fall. Under an income tax, pensions are a tax-preferred form of saving. But a consumption tax treats all saving equally, making it less likely that workers and employers would continue to accept the high regulatory and administrative costs of pensions. Therefore, if workers do not resave all of their reduced pension contributions, saving would fall.
Third, under a pure consumption tax, all capital at the time of the transition is (implicitly) taxed again when it is consumed. But transition rules likely to be added to a consumption tax to avoid this double taxation would reduce or eliminate the long-term effect on saving and growth, as noted above.
The transition: Can we get there from here?
Even if a consumption tax is the right system for an economy starting from scratch, it may not be the right way to reform an existing system. The main transition issue is the taxation of "old capital"-capital assets accumulated earlier out of aftertax income whose principal would not have been taxed again under the income tax. Although some transitional treatment of old capital is typically thought to be likely, not having a transition-that is, implicitly taxing old capital again under a consumption tax-is arguably consistent with the three main goals of tax reform: efficiency, equity, and simplicity.
Certainly, not having a transition is simplest. The transition rules could be very complex, and the transition period could stretch out for years.
Not having a transition is also more efficient. Because only current assets or future wages can finance future consumption, a consumption tax is a tax on future wages and current assets. A consumption tax that exempts old assets is just a tax on future wages. And the same studies that show that a consumption tax (which taxes all old capital assets) is more efficient than an income tax also show that a wage tax is less efficient than an income tax-because not taxing capital requires higher tax rates on wages to raise the same revenue and hence distorts people’s work decisions more. So exempting old capital removes any presumption that tax reform would result in a more efficient system.
Surely, the strongest argument for exempting old capital from taxes is fairness. The assets have already been taxed once; is it fair to tax them again? The answer may not be as obvious as it seems. First, a onetime implicit tax on existing capital is very progressive. The distribution of such capital is more skewed toward wealthy households than is the distribution of overall wealth, which in turn is more skewed than the distribution of income. Second, within any age group, wealthy households do most of the saving. Because these households would benefit most from eliminating the double taxation on future saving under a consumption tax, it is reasonable that they pay for some of the costs. Third, older households tend to have more assets than younger ones, and taxing capital places heavier burdens on older generations. But those older households have received transfers through Social Security and Medicare that far outreach what they have put in. And the vast majority of income and wealth for most elderly households is in the form of future earnings (which have not yet been taxed), housing (which receives extraordinarily preferential treatment under the current tax system), pension income (which already receives consumption tax treatment), Social Security benefits (which are not taxed under the flat tax), and Medicare benefits (which are not and would not be taxed). Relatively few elderly households finance much of their living expenses by other assets, and those that do tend to be very well off.
Pros and cons of the flat tax
In principle, replacing the income tax with a consumption tax, such as the flat tax, offers the possibility of improving the efficiency, equity, and simplicity of the tax system. But these gains are uncertain and depend critically on the details of the reform. At least some of the gains could be made simply by modifying the existing system.
Idealized consumption taxes may always look better than actual income tax systems. Once in place, though, they would be subject to the same compromises and pressures as the income tax is. They could even lead to a system that is less efficient and less fair than the one we have.
- Aaron, Henry J., and William G. Gale, eds. Economic Effects of Fundamental Tax Reform.Washington, DC: Brookings Institution Press, 1996.
- Altig, David, Alan J. Auerbach, Laurence J. Kotlikoff, Ken A. Smetters, and Jan Walliser. "Simulating Fundamental Tax Reform in the United States." American Economic Review 91, no. 3 (June 2001): 574-95.
- Engen, Eric, Jane Gravelle, and Kent Smetters. "Dynamic Tax Models: Why They Do the Things They Do." National Tax Journal 50 (September 1997): 657-82.
- Gale, William G., Scott Houser, and John Karl Scholz. "Distributional Effects of Fundamental Tax Reform." In Economic Effects of Fundamental Tax Reform, edited by Henry J. Aaron and William G. Gale (281–320). Washington, DC: Brookings Institution Press, 1996.
- Gentry, William, and R. Glenn Hubbard. "Distributional Effects of Adopting a National Retail Sales Tax." In Tax Policy and the Economy, vol. 11, edited by James M. Poterba (1–47). Cambridge, MA: MIT Press, 1997.
- Hall, Robert, and Alvin E. Rabushka. The Flat Tax. Stanford: Hoover Institution Press, 1995.
- Zodrow, George R., and Peter Mieszkowski. United States Tax Reform in the 21st Century. Cambridge, U.K.: Cambridge University Press, 2002.