Making Good Policy Out of a Bad AMT

Report Taxes

Making Good Policy Out of a Bad AMT

October 31, 2007 19 min read Download Report
J.D. Foster
Former Norman B. Ture Senior Fellow in the Economics of Fiscal Policy
J.D. served as the Norman B. Ture Senior Fellow in Economics of Fiscal Policy

The majority party in Congress threatens to use the annual ordeal of extending the alternative minimum tax (AMT) patch as a ruse to raise taxes. Policymakers and the public need to expose this ruse, reject the tax hike, and turn to practical, substantive reform to end this pernicious tax.

The AMT patch is an increase in the AMT exemp­tion amount that kept the AMT at bay for about 18 million taxpayers in 2006.[1] The patch expired at the end of 2006, and unless Congress acts, a married AMT filer may face a tax increase of up to $5,026.[2] This tax hike would fall almost exclusively on middle-class and upper-middle-class families and would be a big and unhappy surprise for most.

Fortunately, broad bipartisan agreement deems an AMT tax hike unacceptable. However, Congress may try to raise taxes through an AMT patch ruse that may not be immediately apparent from the official scoring because of an oddity in the Congressional Budget Office (CBO) revenue baseline. Specifically, the CBO includes the revenues from a lapsed but long-standing tax provision in the baseline, which is absurd--so absurd that it does not apply the same methodology to spending programs.

Another problem with the AMT patch ruse is that it diverts attention from the important matter of resolv­ing the AMT issue, preferably by repealing the AMT altogether. The AMT today is a tax policy without a purpose, a complication without a virtue.

One device often used to avoid positive action is to argue that the AMT issue is complicated and can be solved only by a massive overhaul of the tax code. This device has been used to delay repeal of the AMT, and it is incorrect. The AMT can be repealed on a revenue-neutral basis (properly defined) by reforming one or two other aspects of the individ­ual income tax that distort economic decision mak­ing. A number of options are available, and the last section of this paper describes the best option for AMT reform.

Three Policy Choices

Congress faces three policy choices on the AMT patch. First, it could enact a huge tax hike to "pay for" extending the AMT patch through 2007. This approach, which relies heavily on confusion about the CBO revenue baseline, is advocated by Representative Charles Rangel (D-NY), chairman of the House Ways and Means Committee, and Senator Kent Conrad (D-ND), chairman of the Senate Budget Committee. Mr. Rangel looks to extend the patch through 2008 and then repeal the AMT altogether thereafter by raising taxes on other taxpayers such as savers, investors, and entrepreneurs. Senator Conrad leans toward retaining the AMT while extending the AMT patch and using the extension as an excuse to raise taxes. Both approaches would increase taxes by nearly $1 trillion over the next 10 years.

The second choice is to extend the patch with­out raising taxes. Extending the AMT patch is a minimalist but acceptable policy in that it pre­vents a massive tax hike. If Congress can do noth­ing else, then it should at least extend the patch for another year.

The third and best choice is to avoid the trap of the CBO baseline by adjusting it downward to reflect a permanent extension of the AMT patch. Using the corrected baseline as a point of reference, Congress should then repeal the remainder of the AMT on a revenue-neutral basis through a sensible reform of the income tax. This reform should recog­nize the importance of preserving economic incen­tives while leaving the distribution of the tax burden as steady as possible. Such a reform is discussed in the final section of this paper.

Setting the Stage for the AMT Debate

The budget deficit for fiscal year (FY) 2007 was $163 billion, according to the U.S. Department of the Treasury,[3] down from $248 billion in 2006. This is a remarkable achievement considering the enor­mous sums still being spent on Katrina recovery efforts and the far greater sums spent on the wars in Afghanistan and Iraq. If not for this extraordinary spending, the budget would have been essentially balanced in 2007.

Recent spending restraint has contributed some­what to the improved budget picture. For example, total federal outlays grew only about 2.8 percent between 2006 and 2007.[4] Yet the real story is how the tax cuts of 2001 and 2003 restored the economy to robust growth, which has resulted in an enor­mous growth in tax receipts. Federal receipts in 2005 were up 14.5 percent from the year before. Federal receipts increased by 11.8 percent in 2006 and by 6.6 percent in 2007. Although the economy will face extraordinary hurdles from time to time, such as the current housing slump and its associ­ated credit crunch, the economy will soon resume a pace of steady growth, generating in turn steady growth in government receipts.

To be sure, $163 billion is still a significant sum, but the deficit should be interpreted in context. For example, the deficit was about 1.2 percent of gross domestic product (GDP), compared to the average postwar deficit-to-GDP ratio of about 2.2 percent. The decline in the deficit along with continued eco­nomic growth means that the ratio of publicly held debt to GDP is falling, a standard indicator for good near-term fiscal health. As long as the economy continues to expand and the federal government maintains a modicum of spending restraint, the federal budget should reach balance by 2012, if not well before.

Consequently, on the current spending trajec­tory, there is no budgetary case for hiking taxes. On the contrary, Congress should consider additional pro-growth tax cuts, allowing Americans to retain more of the fruits of their labor and giving the American economy more opportunity to grow and raise family incomes.

However, Congress apparently does not intend to keep to the current trajectory on spending. The spring 2007 supplemental spending bill to fund the troops in Iraq and Afghanistan included $20 billion in unnecessary domestic spending. The congres­sional budget resolution laying out the budget blue­print for 2008 calls for increasing non-defense discretionary spending by more than $40 billion (about 9.0 percent).[5] This is an extraordinary increase in spending at a time when inflation is fore­cast at 2.7 percent.[6] In addition, Congress is antici­pating up to $190 billion in increased spending on farm programs, the State Children's Health Insur­ance Program, and other programs.[7]

To enact all of this spending without reversing progress on the budget deficit, Congress would need to raise taxes--a lot. The expiration of the AMT patch provides a convenient and usefully complicated field in which to harvest a lot of tax revenue. However, the correct course of action is to forgo these spending increases, thereby eliminating any need for a huge tax hike.

The AMT and the Patch

The AMT has always been bad tax policy, but today it stands utterly stripped of its original policy justification as an extra tax on a small number of tar­geted high-income taxpayers. It remains relevant only because the federal government is addicted to the enormous stream of AMT revenues. Even with a patch, the AMT is believed to have brought in almost $20 billion in 2006.[8]

Structurally, the AMT is just another poorly designed income tax. Taxpayers calculate their tax liability under the regular income tax and again under the AMT, and pay whichever tax is higher. The AMT differs from the regular income tax in that its top rate is 28 percent compared to 35 percent for the regular income tax. Although it has a lower mar­ginal rate, the AMT also has a larger standard exemption amount ($45,000 for married filers) to keep lower-middle-income taxpayers out of its reach; but there is no personal exemption for chil­dren, so larger families are more likely to be AMT payers. It also denies certain deductions, the most noteworthy of which is the deduction for state and local taxes.

The AMT patch is a significant bump up in the AMT exemption amount. The original patch, set at $4,000 for a married filer, was devised as part of the 2001 tax relief bill.[9] The 2001 tax bill reduced income tax rates, increased the child tax credit, and made other tax changes that had the net effect of significantly lowering regular income taxes. The reduction in regular income tax liability cre­ated potential AMT taxpayers out of some who previously had paid no AMT. The original AMT patch was intended to reduce AMT liabilities somewhat in tandem with the reduction in the regular income tax, ensuring that few taxpayers would lose this regular income tax relief by falling into the AMT.

Since 2001, the "temporary" AMT patch has been extended time and again, preventing a signif­icant tax hike on millions of taxpayers. Congress has also increased the patch over the years, first to index it for inflation and subsequently to hold the number of AMT filers constant. For 2006, the patch was $17,550, raising the total exemption to $62,550.[10] However, from the outset, the patch was at best a crude policy fix, a placeholder waiting for the Administration and Congress to offer a more intelligent means of stopping AMT creep.

Even with the AMT patch extended and indexed for inflation, the number of AMT taxpayers will increase steadily in future years. One cause of this increase is that the base AMT exemption of $45,000 is not indexed for inflation. Furthermore, some income tax deductions that are allowed under the regular income tax but disallowed under the AMT--especially the deduction for state and local taxes--tend to grow faster than inflation.

Thus, even if Congress extends the AMT patch and indexes it for inflation, the AMT will remain a nagging annoyance for taxpayers and contribute to the steady rise in the overall tax burden. Perma­nence of the patch is a good first step, but Congress needs to finish the job by repealing the AMT.

The Administration has consistently acknowl­edged the need for a smarter solution to the AMT, repeatedly suggesting in its budget submissions and testimony that the problem demanded prompt attention and that the large revenues involved would require a more fundamental reform of the federal income tax. In November 2005, the Presi­dent's Advisory Panel on Fundamental Tax Reform made a number of good suggestions, included abolishing the AMT. The Treasury continues to "study" the panel's report, pending a green light from the White House to propound substantive proposals.

The Threatened AMT Tax Hike

If the AMT patch is allowed to expire and no other changes are made in tax policy, then taxes on about 18 million taxpayers will increase by up to $5,026 in 2007.[11]

Congress is responsible for tax policy. Under the Constitution, tax legislation must originate in the House of Representatives, and historically, the House has appropriately and zealously defended this prerogative. The President can sign or veto tax legislation, but only Congress can enact tax legisla­tion. If the patch or its tax relief equivalent is not extended, then Congress bears the full responsibil­ity for the massive tax increase.

Some advocates for higher taxes, both in and out of Congress, argue that allowing the patch to expire or allowing the 2001 and 2003 tax cuts to expire is not really a tax increase because the expirations are already the law.

For its part, the Administration has maintained a muddled position on the AMT patch, supporting an extension of the patch in the short term but cloak­ing its expiration thereafter in vague references to fundamental reform. By contributing to the confu­sion rather than stating a clear position on taxes, the Administration has been at least complicit in cur­rent efforts to raise taxes.

Regrettably, certain CBO scoring rules, as estab­lished by Congress, contribute to this confusion. Under the scoring rules, allowing the patch to expire is not classified as a tax hike. The CBO scores the tax law as it stands. If the AMT patch were extended for 20 years and allowed to expire in year 21, CBO scoring would show higher receipts in its baseline in the 21st year, and any legislation that extended the patch for the 21st year would there­fore be shown as a tax cut.

The purpose of constructing a revenue or spending baseline is to indicate the future conse­quences of current policy, assuming that current policy is maintained. This is why the spending baseline is sometimes called the current services baseline. The federal government's day-to-day spending is authorized and appro­priated annually, yet the spending baseline assumes that current spend­ing levels will continue from one year to the next, adjusted for infla­tion and population growth.

Similarly, some spending pro­grams, such as the highway bill and the State Children's Health Insurance Program, are authorized for multiple years. Such programs are likewise assumed to continue in the spending baseline. The reasonable presump­tion is that these programs reflect current policy and that, by assump­tion, current policy continues in the baseline.

The revenue baseline is constructed based on current law. The spending baseline is constructed based on current services. This lack of symmetry is utterly without justification. This is not a criticism of the Congressional Budget Office, which is carry­ing out the dictates of Congress, but it is important for correcting the record when advocates of tax hikes attempt to explain away a tax hike.

Confusing revenue baselines may carry the day in congressional caucus rooms, but they will not fool American taxpayers. This is the tax policy equivalent of the old carnival game of "watch the pea." If the patch is allowed to expire or if other taxes are raised in its stead, then there will be no question about the reality of a tax hike to the tax­payers facing higher tax bills. The proof will be in black and white in their reduced paychecks and on their Form 1040 tax filings. If taxpayers and poli­cymakers do not watch the pea carefully, this is exactly what will happen.

Table 1 provides a guide for determining whether the net result of the AMT legislation is a tax hike. The first line shows the CBO baseline revenue forecast, which reflects all tax provisions as slated under current law, whether elements of the tax law expire, phase up, phase down, or are indexed for inflation.

The second line, "Tax Cuts Extended," shows the CBO revenue forecast year by year assuming that Congress and the Administration extend the 2001 and 2003 tax provisions. This line is con­structed using the Joint Tax Committee (JTC) esti­mates included in the CBO's analysis of the President's FY 2008 budget. The third line in the table, "Corrected Baseline," makes the further adjustment of extending the AMT patch indexed for inflation through 2012. The Corrected Baseline shows total federal tax collections if Congress nei­ther increases nor cuts taxes.[12]

In today's debate, another way to gauge whether or not taxes are rising through legislation is to look at the resulting five-year growth rates in receipts. If the CBO's receipts forecast shows receipts growing at an average of less than 4 percent per year over the next five years, then Congress is adhering to the Corrected Baseline.[13] If receipts are forecast to rise by closer to 5.5 percent per year, it is because Con­gress has raised taxes by the equivalent of allowing the AMT patch to expire, not because the CBO has assumed faster economic growth and thus faster revenue growth. If receipts are forecast to increase by around 6 percent or more per year over the next five years, then Congress has enacted a massive tax hike equivalent to allowing the AMT patch and all of the 2001 and 2003 tax cuts to expire.

An even simpler way to test congressional actions is to look at the basic pattern of receipts over time. To the extent that Congress resorts to using expiring tax provisions to cover raising taxes, the growth in receipts in coming years will look much like the top line in the graph. If Congress holds the line on tax cuts, both with respect to the AMT patch and with respect to the 2001 and 2003 tax cuts, then the growth in tax receipts will look much like the Corrected Baseline.

The Corrected Baseline depicts the receipts forecast assuming that tax policy is treated symmetrically to spending policy. That is, it assumes extension of the 2001 and 2003 tax cuts and the 2006 AMT patch in­dexed for inflation. As such, it depicts the true policy-neutral forecast for tax receipts. However, the Corrected Baseline is not the target outcome for tax policy. At 18.8 percent of GDP, federal tax receipts are again well above their postwar average of 18.3 percent. This historic average should be regarded as a ceiling above which Congress should not permit the tax share to rise.

A Smart Path to AMT Repeal

The expiration of the AMT patch should not be used as a cover to raise taxes. However, it does offer a goodopportunity to examine smarter alternatives to ex­tending the AMT patch.

The starting point on the smart path to reform is to measure changes in the level of tax collections against the Corrected Baseline that assumes exten­sion of the AMT patch. The next step in reform is to repeal the AMT. This can be done without raising or lowering taxes in the aggregate through offsetting income tax changes. These income tax changes would shift tax liability among some taxpayers but would not create a net tax increase as long as the level of aggregate receipts is held at or below the Corrected Baseline.

The best revenue offset for an AMT repeal would be to phase out the income tax deduction for state and local taxes. There are many ways to phase out the deduction. A simple way would be to allow married filers to deduct the full amount of state and local taxes on the first $14,300 of income, 50 per­cent of taxes on the next $14,300, and nothing for taxes paid on income above $28,600.[14] The combi­nation of repealing the AMT and phasing out the deduction for state and local taxes respects eco­nomic incentives, minimizes the shifting of tax bur­dens among taxpayers, and minimizes the shifting of tax burdens among states.

There are good tax policy arguments both for and against preserving the state and local tax de­duction, but a review of these arguments is beyond the scope of this paper. However, respected voices on both the left and the right have suggested re­pealing the state and local deduction in the context of AMT reform.[15]

These theoretical and political arguments aside, one crucial fact makes the phaseout of the state and local tax deduction the logical "payfor" to eliminate the AMT: The AMT itself is repealing the deduction already. Taxpayers lose some income tax deductions in calculating their AMT liability. One such deduc­tion is the state and local tax deduction. The pro­jected steady rise in AMT receipts is due in part to the AMT's steady de facto repeal of the state and local tax deduction.

Eliminating the AMT and offsetting the revenue effects by phasing out the state and local tax deduc­tion would have other important advantages as well. For example, it would offer significant, real simplification of the tax system by eliminating the AMT altogether without significantly changing effective marginal tax rates.

The combination of AMT repeal and the phase­out of the state and local tax deduction also has the advantage of minimizing the tax shifting involved. Those taxpayers whose income tax liabilities would rise without the state and local tax deduction would often be the same taxpayers who would receive relief from repealing the AMT. There is no reason to allow revenue-neutral AMT reform to produce significant shifts of federal tax liability among taxpayers.

Likewise, there would be relatively little shifting of the federal tax burden among taxpayers in their respective states with this formulation. States that impose lower tax burdens on their citizens also have fewer citizens subject to the AMT. These low-tax states also tend to have fewer citizens choosing to itemize in lieu of taking the standard deduction and thus have fewer citizens materially affected by the phaseout of the state and local tax deduction for upper-income taxpayers.


A strong consensus exists to reform the AMT and, preferably, to eliminate it altogether. While the consensus is not new, there is a renewed drive for results. This newfound enthusiasm for serious AMT reform is rooted in part in the desire by the majority in Congress to kill not just two, but possibly three birds with one stone: resolving the AMT issue, rais­ing taxes to fund a splurge in new federal spending, and possibly shifting even more of the tax burden onto upper-income taxpayers.

The necessity of dealing with the AMT should not be used to disguise a tax hike. The deficit is fall­ing as a result of economic growth and modest spending restraint. The only justification for raising taxes is to jack up spending without worsening the deficit picture, but there is no justification for rais­ing spending.

At 20.3 percent of GDP, the federal government already consumes over one-fifth of total output in the United States. If new spending priorities arise, as they inevitably will, then Congress, at a mini­mum, should reprioritize existing levels of spending while holding the aggregate total unchanged. Better yet, Congress should take this occasion to repriori­tize spending and give relief to taxpayers by slowing the growth in total spending.

The AMT should not be reformed; it should be eliminated. To the extent that elimination of the AMT must be accompanied by offsetting revenues without raising receipts above the corrected base­line, the best "payfor" is to phase out the income tax deduction for state and local taxes. This reform would avoid increasing the disincentives to eco­nomic growth in the tax system and significantly simplify the tax code. It would also minimize the shifting of the tax burden among taxpayers across income levels and across states.

J. D. Foster, Ph.D., is Norman B. Ture Senior Fel­low in the Economics of Fiscal Policy in the Thomas A. Roe Institute for Economic Policy Studies at The Heritage Foundation.

[1]See Joint Tax Committee, Present Law and Background Relating to the Individual Alternative Minimum Tax, staff report, JCX- 38-07, June 27, 2007, at (October 29, 2007).

[2]This is the estimated amount of tax relief provided in 2007 when an AMT payer at the 28 percent rate uses the full amount of the AMT patch. The AMT patch for 2006 was $17,550. Through the third quarter of 2007, the Consumer Price Index was up 2.4 percent over the third quarter of 2006. Assuming that this rate of inflation carries through the year, the 2007 patch indexed for inflation would be $17,950 using the usual rounding convention, which would be worth $5,026 to an AMT taxpayer in the 28 percent bracket.

[3]See Henry M. Paulson, Jr., and Jim Nussle, "Budget Results for Fiscal Year 2007," October 11, 2007, at (October 29, 2007).

[4]Total federal spending rose by just 2.8 percent from 2006 to 2007. See Ibid.

[5]Committee on Appropriations, U.S. House of Representatives, "FY 2008 Appropriations Subcommittee Allocations-- 302(b)s," at /static/reportimages/01D69C847F99C5858410F3880AB3B143.pdf (October 29, 2007).

[6]The Congressional Budget Office forecast for 2007 GDP deflator is 1.7 percent. Congressional Budget Office, "The Budget and Economic Outlook: Fiscal Years 2008 to 2017," January 2007, at (October 29, 2007).

[7]Republican Caucus, Committee on the Budget, U.S. House of Representatives, "The Democrats' Fiscal Year 2008 Budget Conference Report: Higher Spending," at /static/reportimages/24464C56DD78C9BB648365A6CE01B858.pdf (October 29, 2007).

[8]Joint Tax Committee, Present Law and Background Relating to the Alternative Minimum Tax, p. 9, Figure 1.

[9]Ibid., p. 6.


[11]Ibid., p. 9, Figure 1.

[12]There are three noteworthy caveats associated with this table, each of which essentially implies that the figures in this table will change from time to time as the CBO and the JTC update their estimates. First, baseline revenue forecasts change periodically, sometimes significantly. For example, in August, the CBO updated its revenue forecast for 2007 and beyond, raising the forecasts somewhat. Thus, the figures six months or a year from now could look somewhat different. Second, the JTC has not updated its estimates of the effects of extending the 2001 and 2003 tax cuts in light of the new CBO economic forecast and new data on tax receipts. Based on the new CBO forecast, the JTC would likely modestly alter its forecast for the amount of tax increase prevented by extending the 2001 and 2003 tax provisions. Thus, in the details there is a modest degree of incomparability between the first line and the second. The third caveat is that the CBO has provided a scoring for only a 2007 AMT patch. The revenue estimates for a 2008 AMT patch were developed for this paper and reflect the expected rapid growth in the AMT in the coming years. The implication of these caveats is that the exact figures will change in the coming months, but the basic pattern will hold true.

[13]These benchmarks are all relative to the CBO's receipts forecasts and do not necessarily reflect the author's views on how receipts are likely to grow.

[14]These figures are estimated for 2008 and were provided by the Center for Data Analysis at The Heritage Foundation. The corresponding figures for single filers are half of the amounts for joint filers.

[15]Leonard E. Burman, William G. Gale, and Jeffrey Rohaly, "The Expanding Reach of the Individual Alternative Minimum Tax," testimony submitted to the Subcommittee on Taxation and IRS Oversight, Committee on Finance, U.S. Senate, May 23, 2005, at (October 29, 2007), and Daniel J. Mitchell, "Sales Tax Deduction Would Subsidize Bigger Government, Undermine Tax Reform," Heritage Foundation WebMemo No. 520, June 9, 2004, at


J.D. Foster

Former Norman B. Ture Senior Fellow in the Economics of Fiscal Policy