No. 1, November 2006
This is the first in a new series of analytical essays devoted to key issues in tax policy. Resident scholar Alan D. Viard will be the primary author of Tax Policy Outlook.
Some people have hailed the alternative minimum tax (AMT) as a flat, low-rate, broad-based tax that would be a good replacement for the regular income tax. A closer look at the AMT dispels this rosy view.
As the individual alternative minimum tax (AMT) has spread, there have been more and more complaints about the complexity and inefficiency of dealing with both the regular income tax and the AMT. The usual proposed solution has been to repeal or scale back the AMT.
Lately, though, some people have suggested moving in the opposite direction, repealing the regular tax and keeping the AMT as our sole tax system. They argue that the AMT is a flat, low-rate, broad-based tax--just what tax reformers have always wished for. They claim that the AMT’s lower tax rates would increase work and savings and that its base-broadening provisions would reduce economic distortions by eliminating special tax breaks.
A closer look at the AMT, however, dispels this sanguine view. The AMT’s effective marginal tax rates are, on the whole, no lower than those of the regular tax system. Its base-broadening provisions are limited and some of them may hamper rather than promote economic efficiency.
It would be a mistake to scrap the regular tax for the AMT. Instead, the hard work of moving to a better tax system is best pursued by improving the regular income tax.
Understanding the AMT
Each year, taxpayers must pay either their regular income tax or their AMT liability--whichever is larger. The AMT denies some of the regular tax’s exclusions, deductions, and credits, but it generally imposes a lower tax at any given level of taxable income.
Under the regular income tax, taxpayers subtract deductions and exemptions from gross income to obtain taxable income. A tax schedule then translates taxable income into tax liability before credits. Tax credits are then subtracted. As described below, the AMT has the same basic structure as the regular tax, but with different rules at each stage. My description and comparison of the two systems are based on the provisions currently legislated for 2007.
There are a few differences between the regular tax and the AMT regarding how gross income is computed. For example:
- interest on private-activity municipal bonds, such as those used to finance industrial projects, is included under the AMT, but not under the regular tax;
- in computing income for owners of pass-through businesses (sole proprietorships, partnerships, limited liability companies, and S corporations), the AMT uses slower depreciation schedules and puts more restrictions on the use of one year’s losses to offset other years’ income; and
- the AMT taxes holders of incentive stock options when they exercise the option, while the regular tax waits until the stock is sold.
Under the regular tax, taxpayers can claim the larger of their itemized deductions or the standard deduction ($10,700 for married couples). The AMT has no standard deduction and disallows some itemized deductions. For example:
- state and local income as well as property and sales taxes cannot be deducted, with the exception of property and sales taxes paid on business items;
- medical expenses can be deducted only if they exceed 10 percent of income, compared to 7.5 percent under the regular tax;
- interest on home-equity loans can be deducted only if the loan proceeds are used for home improvements; and
- certain miscellaneous itemized deductions, including workers’ unreimbursed job expenses, investment-related expenses, and attorney’s fees paid by winners of taxable damage awards, cannot be claimed. (The regular tax allows these deductions if they exceed 2 percent of income.)
The AMT generally allows other itemized deductions, such as those for charitable contributions and investment interest, although its rules are more restrictive in some respects. While the regular tax phases out some itemized deductions when income exceeds $156,400, the AMT does not.
The regular tax allows $3,400 per-person exemptions for the taxpayer, spouse, and each child, which are phased out at high income levels. The AMT does not allow the exemptions.
Although taxable income is usually larger under AMT rules, the AMT has a more favorable tax schedule. For married couples, the first $45,000 of income faces a zero rate, and a 26 percent rate then applies up to $150,000. Two higher brackets then follow: income between $150,000 and $206,000 is taxed at 32.5 percent and income between $206,000 and $330,000 is taxed at 35 percent. Above $330,000, though, the rate falls back to 28 percent. The 32.5 and 35 percent brackets effectively phase out the benefits of the $45,000 tax-free amount.
The AMT allows the $1,000 child credit, the earned income tax credit, the adoption credit, and the foreign-tax credit. It does not, however, allow other personal credits, including the child-care credit, the Hope Scholarship credit, and the Lifetime Learning credit. Also, owners of pass-through businesses cannot claim most business-related credits.
The AMT has spread to more tax-payers because its tax brackets, unlike those of the regular income tax, are not indexed for inflation. To slow this spread, Congress has temporarily provided higher AMT tax-free amounts and temporarily allowed certain personal tax credits to be claimed against the AMT. Even so, 3.4 million taxpayers are on the AMT this year, up from a mere 200,000 in 1990. If the temporary measures are allowed to expire at the end of 2006, 21.9 million tax-payers will be on the AMT in 2007.
As mentioned previously, some people have recently hailed the AMT as a flat, low-rate, broad-based tax. Although that view may initially seem plausible, it does not hold up under close scrutiny.
Marginal Tax Rates and Disincentives
A key claim made by AMT supporters is that it penalizes work and saving less than the regular tax does because it has lower marginal tax rates. But is that really the case?
The effective marginal tax rate is the additional tax triggered by earning an additional dollar of income. This effective rate depends on the breadth of the tax base as well as the official marginal tax rate. For example, suppose that people always spend half of their income on apples and half on oranges. Also, suppose that the income tax has a 40 percent rate, but that spending on oranges is deductible. Although this tax has a 40 percent official marginal rate, it has an effective marginal rate of about 20 percent. In other words, its work disincentives are roughly the same as those of a 20 percent tax on both fruits, reflecting the combined impact of a 40 percent tax on working to buy apples and a zero tax on working to buy oranges. (As discussed below, despite the similarity in work disincentives, the 40 percent apple tax is markedly inferior to the 20 percent broad-based tax in other respects, as it distorts consumers’ choices between the two fruits.)
To see how the AMT’s broader base affects its work disincentives, consider state income taxes, a major item disallowed by the AMT. When additional work increases state income tax payments, the deductibility of those payments under the regular tax clearly lowers work disincentives at any given official marginal rate.
Figure 1 displays simplified estimates of the differences in the two systems’ effective marginal tax rates, the rates at which tax increases with an extra dollar of adjusted gross income. Results are shown for married couples with two children, single taxpayers with no children, and unmarried heads of household with one child, at adjusted gross incomes ranging from zero to $500,000. I assume that taxpayers always spend 8 percent of income on items that are deductible under both the regular tax and the AMT and 7 percent on items that are deductible only under the regular tax.
The results differ sharply across three income intervals. At low incomes, the AMT has lower effective marginal tax rates because it offers a larger tax-free amount than the regular tax. Above the tax-free amount, however, the AMT marginal rate jumps above the regular marginal rate. The AMT marginal rate moves back below the regular marginal rate at $359,000* for married couples and at $269,000 for singles and heads of household, settling 5 percentage points below the regular tax marginal rate at the highest income levels.
Rough calculations suggest that the amount of national income earned in ranges in which the AMT has lower marginal rates is comparable to the amount earned in the range in which it has higher marginal rates. For the economy as a whole, the AMT offers little or no reduction in work disincentives.
Turning to saving, the AMT, like the regular tax, is structured as an income tax that taxes the return to saving, not as a consumption tax. The AMT, like the regular tax, ameliorates the tax burden on saving by offering a bewildering variety of tax-preferred savings accounts and preferential rates for dividends and long-term capital gains. As shown above, the AMT’s marginal rates are not systematically lower than the regular tax’s rates. For the economy as a whole, then, the AMT offers little or no reduction in saving disincentives.
AMT supporters also see economic gains from its broader base. Even if replacing a high-rate, narrow-based tax with a low-rate, broad-based tax does not reduce work disincentives, it often promotes economic efficiency in other ways. Consumers are generally better off if the tax system is neutral between different economic activities because the allocation of resources can then be determined by market forces rather than by tax considerations. Removing special tax breaks that favor some activities over others can make the tax system more neutral.
Unfortunately, the AMT’s base broadening largely fails to deliver such benefits. In many areas, the AMT base is the same as the regular tax base. In areas in which the AMT does have a broader base, its base broadening is incomplete and may be harmful rather than helpful.
In many areas, the AMT has the same base as the regular tax. Both tax systems exempt gifts, inheritances, imputed rent on owner-occupied homes, interest on most municipal bonds, personal-injury damage awards, many government transfer payments, and most fringe benefits. Like the regular tax, the AMT allows deductions for moving expenses, charitable contributions, some theft and some casualty losses, and gambling losses. Both systems allow the adoption credit, the earned income credit, and the $1,000 child credit. The AMT falls far short of reformers’ goal of comprehensive income taxation.
In other areas, the AMT has a broader base than the regular tax, but this may not always be beneficial. The problem is that a little base broadening can be worse than none at all. Using the fruit example, repealing the deduction for all oranges would promote efficiency because the tax system would then treat both fruits neutrally. But repealing the deduction only for mandarin oranges would probably reduce efficiency. Although the distortion between mandarin oranges and apples would be eliminated, a new distortion between mandarins and other oranges would be created. Because mandarins are close substitutes for other oranges, consumers’ choice between them is likely to be particularly sensitive to differences in tax rates. The harm from distorting that choice would probably outweigh the gain from apple-mandarin neutrality.
This analysis likely applies to some of the AMT’s base-broadening provisions. For example, the AMT limits the deduction for out-of-pocket medical expenses, but retains the tax preference for employer-provided health insurance. Some research indicates that so long as health insurance is tax-preferred, limiting the deductibility of out-of-pocket medical spending may be inefficient, because it further promotes insurance. Similarly, the AMT taxes interest on private-activity municipal bonds, but retains the exemption for interest on other municipal bonds, which may simply cause state and local financing to be directed away from private-activity projects.
Base broadening can be harmful for other reasons. Efficiency calls for a deduction for expenditures that are costs of earning income. It also generally calls for a partial deduction for goods toward which people tend to shift their purchases as they work more. As such goods--sometimes called mixed business-personal expenses--become cheaper, people work more. Unfortunately, some of the AMT’s base broadening targets these items.
The AMT denies a deduction for employees’ unreimbursed job expenses. This treatment is clearly inappropriate for costs of earning income such as workers’ purchases of tools, and may be too harsh for mixed expenses such as meals and lodging costs incurred at temporary jobs. The AMT’s denial of deductions for investment fees, subscriptions to investment publications, and other investment-related expenses also seems questionable.
The AMT’s disallowance of the credit for child-care expenses may also be harmful, since extensive statistical evidence indicates that increases in child-care costs reduce maternal employment. A similar point may apply to the AMT’s denial of the Hope Scholarship and Lifetime Learning credits, since increases in higher education costs may result in a less-educated labor force and thereby reduce the productive value of work.
In some cases, the AMT imposes heavy, even confiscatory, taxation. For example, a plaintiff who wins a taxable damage award in a nonbusiness lawsuit may not deduct attorney’s fees. As a result, the AMT liability can be larger than the net award. Indeed, the taxpayer owes AMT even if the legal fee absorbs the entire damage award. Also, workers who buy stock by exercising an incentive stock option and later suffer a loss may owe AMT on gains they did not retain. Consider a worker who exercises an incentive stock option, paying $100,000 to buy shares worth $500,000, and later sells the stock after its value falls to $200,000. Under the regular tax, the worker is taxed on a $100,000 capital gain at the time of sale. Under the AMT, the worker is taxed on $400,000 at the time of purchase and has a $300,000 capital loss at the time of sale. However, only $3,000 of that loss can be deducted against income other than capital gains per year. Some workers at high-tech firms, whose stock values plummeted after 2000, have AMT bills that they may never be able to pay.
The AMT is on firmer ground in denying owners of pass-through businesses tax credits that favor some industries and activities over others. The disallowed credits include those for enhanced oil recovery, renewable electricity production, nonconventional fuels, disabled access, electric vehicles, Indian employment, new markets, work opportunity, research, and low-income housing. Even here, though, the AMT may paint with too broad a brush. For example, the research credit may offer an appropriate subsidy for basic research that provides social benefits greater than what firms can capture in the market.
The AMT disallows the regular tax’s $3,400 per-child exemption and offers relatively little rate reduction for married couples and heads of household. As a result, the AMT is significantly less favorable to marriage and childbirth (or significantly more favorable to singleness and childlessness) than is the regular tax. Since the proper tax treatment of marriage and children lies beyond the scope of this article, no evaluation of this difference is offered here.
The denial of the itemized deduction for state and local taxes, which is the AMT’s largest single base-broadening provision, merits separate discussion.
Treatment of State and Local Taxes
In general, the federal income tax probably should not distort states’ and localities’ division of functions between the public and private sectors. Under this principle, many state and local taxes should be nondeductible. For example, suppose that residents of one community pay fees to use a privately owned golf course while another community taxes its residents to provide a no-fee municipal golf course. Since there is no federal tax deduction for the fees paid in the first community, there should probably be no deduction for the taxes paid in the second community. Similarly, if a community taxes some residents to finance transfer payments to other residents, the taxes should probably be nondeductible unless the transfer payments are taxable.
To be sure, some state and local taxes should be deductible under this principle. Suppose that businesses in one community hire private security guards to protect their assets while another community imposes taxes to hire police officers to perform this function. Since there is a federal tax deduction for the business expense of hiring security guards in the first community, there should probably be a deduction for the taxes paid in the second community. This point may apply to part of the 8 percent of state and local government spending that goes toward public safety. Also, since most private medical expenditures escape federal income tax, it may be appropriate to allow a deduction for expenditures financed through Medicaid (roughly 10 percent of state and local spending). A similar analysis may apply to the roughly 8 percent of state and local government spending that goes to higher education.
Still, this analysis suggests that most state and local taxes should be nondeductible. One might then think that the AMT’s disallowance of the itemized deduction for income, sales, and property taxes is an improvement. It is important to realize, though, that neither the AMT nor the regular tax follows a consistent approach to state and local taxes.
Both systems have built-in deductions for some state and local taxes, simply because such taxes reduce the income on which the federal tax is paid. For example, if state and local corporate income taxes result in lower dividends and capital gains, then stockholders pay less federal income tax; if these taxes are shifted to workers in the form of lower wages, then workers pay less federal income tax.
Even with its explicit deduction for income, sales, and property taxes, the regular tax treats them less favorably than taxes with built-in deductions, since taxpayers must choose between deducting income and sales taxes, and non-itemizers cannot deduct any of these taxes. By removing the deduction for only these taxes, the AMT engages in the same type of partial base broadening seen in many of its other provisions. Once again, it is hard to know whether the result is a net improvement.
The AMT Falls Short
Recent claims that the AMT should replace the regular income tax do not hold up under close scrutiny. For the economy as a whole, the AMT’s work and savings disincentives are about as large as those posed by the regular tax. While comprehensive base broadening would make the tax system more efficient, the AMT’s haphazard and incomplete base broadening offers little or no economic benefit.
The shortcomings of the regular income tax have received much attention, but those of the AMT are at least as great. On balance, a switch to the AMT would be a detour from, rather than a first step on, the road to tax reform.
*Correction, January 4, 2007: The AMT effective marginal tax rate for married couples falls below the regular effective marginal tax rate at $359,000, not $242,000, as previously stated. Figure 1 has been edited to correct an erroneous application of 2007 tax law.
Alan D. Viard (firstname.lastname@example.org) is a resident scholar at AEI.
AEI research assistant Cindy Soo and AEI editorial associate Nicole Passan worked with Mr. Viard to edit and produce this Tax Policy Outlook.
1. For example, see Jonathan Rauch, “A Bad Tax with Good Timing,” National Journal, March 18, 2006; Editorial, “Fear of Filing,” Wall Street Journal, April 15-16, 2006; and Robert H. Nelson, “An Accidental Tax Boon,” Washington Post, June 1, 2006.
2. I assume, however, that the tax “extenders,” including the research tax credit and the sales tax deduction, are reinstated and extended through 2007.
3. Urban-Brookings Tax Policy Center, Table T06-0246 (October 2006), available at www.taxpolicycenter.org/TaxModel/tmdb/Content/PDF/T06-0246.pdf.
4. Alan J. Auerbach and Joel Slemrod make this point in “The Economic Effects of the Tax Reform Act of 1986,” Journal of Economic Literature 35, no. 2 (June 1997): 602.
5. Calculations are available from the author upon request. The 8 percent and 7 percent values roughly match the ratio of the relevant itemized deductions to income for a significant range of taxpayers. The regular-tax computations include the standard deduction, personal exemptions, and phaseouts of exemptions and itemized deductions. Tax credits and differences in adjusted gross income between the two systems are not included. Heads of household are unmarried taxpayers caring for children.
6. See William Jack and Louise Sheiner, “Welfare-Improving Health Expenditure Subsidies,” American Economic Review 87, no. 1 (March 1997): 206-21; and John F. Cogan, R. Glenn Hubbard, and Daniel P. Kessler, Healthy, Wealthy, and Wise (Washington, DC: AEI Press/Hoover Institution, 2005), available at www.aei.org/book831/.
7. See Patricia M. Anderson and Phillip B. Levine, “Child Care and Mothers’ Employment Decisions” (National Bureau of Economic Research working paper 7058, March 1999), and the studies surveyed therein. For a study looking specifically at the child-care credit, see Susan L. Averett, H. Elizabeth Peters, and Donald M. Waldman, “Tax Credits, Labor Supply, and Child Care,” Review of Economics and Statistics 79, no. 1 (February 1997): 50-67.
8. Warren Rojas, “Bankruptcy Deadline Looms for Those with AMT/ISO Problems,” Tax Notes, October 3, 2005, 32-33; and Floyd Norris, “Windfall Never Came: Big Tax Bill Did,” New York Times, March 31, 2005.
9. Corporations claim far more of these credits than do pass-through businesses. Corporations are not subject to the individual AMT discussed in this article but may be subject to the corporate AMT, which also disallows the credits. The corporate AMT lies outside the scope of this article.
10. Leonard E. Burman and David Weiner, “Suppose They Took the AM out of the AMT?” Proceedings of the 97th Annual Conference on Taxation (Washington, DC, National Tax Association, 2005), 447-58; and George K. Yin, “Enacting ‘Tax Reform’ in the Face of Fiscal Calamity,” Tax Notes, July 3, 2006, 61-63, also conclude that the AMT would not be a good replacement for the regular tax.