CRS Report Explains AMT Interaction With Marriage Penalty Relief

                       CRS REPORT FOR CONGRESS

[1] Several general tax cuts have been addressed in the Congress. Recently, the House passed H.R. 6, legislation to address the marriage penalty by providing across-the-board tax cuts for joint returns and the Senate is scheduled to vote on a similar bill, S. 2346. General tax cut legislation enacted by the Congress in 1999 would have provided even larger tax cuts; this legislation was vetoed by the President.

[2] The Alternative Minimum Tax (AMT) provides for an alternative tax calculation, on a broader base but with a large exemption and a two-tier rate that is below the top tax rates in the regular tax structure. It is paid when the tax liability figured using AMT base and rates is higher than regular tax liability. The AMT is expected to grow rapidly and extend further into the middle class because the exemptions in the AMT are not indexed for inflation. In addition, the tax credits (such as the child credit) enacted in 1997 caused many middle class taxpayers to be affected by the AMT. A temporary provision allowing these credits to be taken against the AMT was adopted last year, but is only effective through 2001.

[3] The marriage penalty legislation, and other proposals for cutting taxes will be limited in their effects for some individuals unless changes are also made in the alternative minimum tax (AMT). Individuals who pay the AMT are not affected by cuts in the regular tax and individuals who switch to the AMT will not receive the full tax cut. This constraint will grow over time. For example, about 28 percent of the tax cuts in H.R. 6 over the next ten years are not received by taxpayers because of the AMT. This effect grows over time; by 2008, 44 percent of the tax cut will not be received.

[4] Cuts in regular tax, without also addressing the AMT, will cause more and more taxpayers to the [SIC] subject to the complexities of the AMT, and also increase the revenue costs of future measures to restrain the growth of the AMT.

[5] There are a number of different policy options that might be considered in evaluating the AMT and its interaction with the regular tax. For some, a priority is making the exclusion for credits permanent, while for others indexing may be the most important priority. Both of these approaches will be costly in the future (about $26 billion for the credit ten years from now and about $14 billion for expensing). Others might wish to eventually phase out the AMT, which will raise about $37 billion by 2010. One can also make a case for expanding the coverage of the AMT as an eventual flat tax, although in some ways the AMT does not conform to certain design principles (such as adjusting exemptions for family size). Another issue is how to adjust the AMT as changes in the regular tax system are made, to keep the relative position and original purpose of the AMT intact. In the latter case, the AMT might be adjusted when fundamental changes are made in the regular tax (rates, bracket widths, standard deductions) but not for proposals that provide special subsidies.

[6] This report will be updated to reflect legislative developments.

CONTENTS

[8] The marriage penalty legislation, and other proposals for cutting taxes will be limited in their effects for some individuals unless changes are also made in the alternative minimum tax (AMT). The AMT is paid when the tax liability figured using AMT base and rates is higher than regular tax liability. Growing numbers of taxpayers will, in the future, be affected by the AMT. Moreover, cuts in the regular tax will shift more individuals into the AMT. Thus, it is important in considering any tax cut proposal to evaluate the role played by the AMT in our tax system.

[9] The first section of this study summarizes the fundamental cause of the AMT interaction with other tax provisions and provides some background data on the AMT. The second section examines the particular interaction between the marriage penalty proposals in H.R. 6 and the AMT, a case for which some estimates of effects are available. The final section discusses potential policy options.

A OVERVIEW OF THE AMT

[10] The AMT provides for an alternative tax calculation on a broader tax base than the regular tax. Individuals add back a variety of provisions including not only business provisions but certain itemized deductions (mainly state and local taxes, some medical expenses and miscellaneous deductions), the standard deduction, and personal exemptions. After an exemption of $45,000 for joint returns and $33,750 for single returns, the first $175,000 is taxed at 26 percent and the remainder is taxed at 28 percent. The individual compares AMT liability and regular tax liability and pays the higher one. While the AMT base is broader than the regular tax base, there are many provisions that are not included in its base, notably the benefits of lower capital gains tax rates and the exclusion for tax exempt bonds, and the itemized deduction for home mortgage interest. Indeed, the elimination of the capital gains exclusion in 1986 caused a significant contraction in the number of taxpayers subject to the AMT; the current rate preferences are not part of the AMT. Credits are automatically included in the base (i.e., effectively disallowed), although there is a suspension of this treatment through 2001. ¹

[11] However, unlike the rest of the income tax, the AMT exemptions are not indexed to inflation. The result has been an increase in the number of taxpayers who are covered by the AMT. ² In 1987, about 140,000 returns paid the AMT, constituting 1/10 of one percent of all returns filed. This number was below the 4/10 of a percent of returns that paid the tax in 1984 and fell largely because the deduction for capital gains was eliminated by the 1986 tax act and therefore automatically affected coverage under the AMT. Inflation, however, took its toll. By 1999, the AMT covered 823,000 returns, constituting 6/10 of a percent of all returns filed, an increase in percentage share of 600 percent between 1987 and 1999. This effect occurred even though the exemptions were increased in 1990.

[12] More growth is ahead, however. In 2009, the Joint Tax Committee projects that over 9 million taxpayers will pay the AMT, constituting 6.3 percent of all tax returns filed, an increase in percentage share of over a thousand percent from 1990 to 2009. The AMT would probably constitute a larger share of joint returns filed, since incomes are higher for these returns than for single returns. The AMT would also constitute a larger portion of returns with tax liability; since typically about a quarter of returns filed pay no tax, taxpayers on the AMT would constitute over 8 percent of returns with tax payments. Moreover, another 6 million returns will be subject to limits on tax credits, such as the child credit and education credits enacted in 1997; this surge in the AMT coverage will occur in large part in 2002, as a temporary moratorium expires. (While credits, absent the subsequent moratorium, would increase the number of taxpayers on the AMT, there was a specific extension of the lower capital gains tax rates to AMT taxpayers in the 1997 legislation).

[13] While the AMT will still be concentrated among higher income individuals, it will gradually reach further down into the income distribution. This shift in the distribution is shown in Table 1. Note that this table understates the coverage of the AMT and its reach into the middle income classes, because it does not include those taxpayers whose credits are limited by the AMT. But it does illustrate how the failure to index exemptions will substantially expand the AMT.

[14] As this table illustrates, the shift of exposure to the AMT from the very highest income classes to the middle and upper middle income classes is dramatic over time. In 1998, almost half of AMT taxpayers fell into the 1.6 percent of tax-filers with adjusted income over $200,000; and over three-quarters fell into the top 8 percent of taxpayers who had income over $100,000. In 2008, less than 15 percent of AMT taxpayers have incomes over $200,000 and about half have incomes of $100,000 or more. Moreover, these effects occur despite the fact that taxpayers in these income classes are accounting for a larger share of total taxpayers as incomes rise over time.

   TABLE 1: PERCENTAGE DISTRIBUTION OF AMT TAXPAYERS BY INCOME CLASS

[16] These effects understate the shift of the influence of the AMT toward the middle class that is expected in the future because they do not include the interaction with the significant tax credits, including the child credit and the education tax credit, that were adopted in 1997. Under the AMT provisions, credits are limited to the excess of regular tax over AMT liability. For 2008, there are another 6 million returns that are constrained by the tax credit. Combining this data with the data for 2009 for taxpayers on the AMT, the implication is that 10.5% of taxpayers would be affected by either of these conditions, and about 14% of taxpayers with tax liability.

[17] The effect of the AMT in restricting use of the credits is greater than suggested by these numbers. If one excludes returns that receive no credits or partial credits (although some of these families are phased out of the credit because of high incomes, most of them are probably lower income families without tax liability or without sufficient tax liability to claim the full credits), families that are restricted by the credit through the AMT constitute 24% of all families eligible to claim the child credit (families with one or more children under 17). Adding taxpayers affected by the AMT in general suggests that perhaps a third or so of families with children are affected by the AMT, a very significant share.

[18] The credit limit due to the AMT has been lifted temporarily (through 2001) for nonrefundable credits. Absent further legislation, therefore, the effects of the credit limit will resume in 2002.

[19] Treasury data provided during the debate on the recent marriage penalty bill also indicated growing numbers and even greater numbers of taxpayers subject to the AMT. Their data indicated that currently about 1.5 million taxpayers are on the AMT, but that 17 million will pay the AMT by 2010. Just between 2007 and 2010, AMT taxpayers are projected to increase by 7 million, from 10 million to 17 million. AMT tax liability is projected to rise from $6.4 billion in 2001 to $38.2 billion in 2010, with a total amount of $182 billion over the ten year period.

[20] The cost of correcting the AMT is significant. According to data from the Joint Committee on Taxation, indexing AMT exemptions would cost $13.9 billion by 2008. Eliminating the credit limit provision would cost about $1 billion currently, but would cost many billions of dollars by 2008 or 2009. Eliminating the credit limit provision and adding standard deductions would cost the U.S. Treasury $26 billion by 2009 and $96 billion for the period 1999-2009. ³

[21] Clearly, the AMT will become increasingly important in the years to come, in the number of taxpayers covered and revenue cost of altering the AMT. And any tax cut that reduces regular tax liabilities and does not also alter the AMT will interact with the AMT in two ways: it will increase the number of taxpayers on the AMT and the number affected by the credit limit, and it will cause some or all of the tax cut not to be received by certain families.

[22] Tax provisions that are aimed at reducing taxes for joint returns may particularly interact with the AMT because married couples are most likely to be affected by the credits enacted in 1997 (primarily the child credit, but also education credits) that are being restricted by the AMT. Married couples also tend to have higher incomes and, while their AMT exemptions are also higher, may be more likely to be affected by the AMT. The interaction is also affected by how the tax change is distributed across the income classes, since only joint returns with more than $45,000 of taxable income are potentially subject to the AMT.

INTERACTION BETWEEN MARRIAGE PENALTY RELIEF AND THE AMT

[23] A marriage penalty arises for some families because family income is combined and subject to progressive tax rates. Since the standard deductions and rate brackets, while larger than those of singles, are not twice as large, marriage can cause the loss of standard deductions and cause some income to be taxed at higher rates. Other couples, however, experience bonuses; this outcome tends to arise when earnings are relatively unequal or when there is only one earner, because the exemption amounts and rate brackets are larger for the joint returns filed by married couples than for singles' returns.

PROVISIONS OF THE RECENT MARRIAGE PENALTY LEGISLATION

[24] H.R. 6, which was passed by the House on February 10, proposes to address the marriage penalty for most taxpayers, granting bonuses to many taxpayers who formerly had penalties and expanding the bonuses of those with bonuses. ⁴ Slightly over 61% of joint returns are in the 15% bracket and would have any penalties that did exist eliminated (and bonuses increased) merely through increasing the standard deduction to twice that of single returns. This provision is estimated to cost $7.6 billion by 2009. Another 31% are in the 28% bracket and would have the remainder of any penalties eliminated (and bonuses increased) through both the standard deduction and the widening of the first bracket to twice that of single returns. This provision would be phased in over six years and cost $19.6 billion by 2009. Thus, 92% percent of joint returns, ignoring the earned income tax credit and the AMT, would be covered by these provisions. There are also some provisions for partially reducing the marriage penalty for the earned income tax credit, costing $1.3 billion by 2009. Those provisions plus a correction involving the AMT and refundable credits will cost $30.3 billion by 2009. ⁵ Some individuals whose income is taxed above the 28 percent bracket currently will also have their penalties eliminated, and since the next rate bracket is only slightly higher (31% percent) this approach will eliminate virtually all marriage penalties for the vast majority of married couples (96% are in the 31% bracket or below). ⁶

[25] The Senate marriage penalty proposal also expands the 28 percent bracket, which will increase the coverage of high income taxpayers.

INTERACTION BETWEEN THE AMT AND MARRIAGE PENALTY LEGISLATION

[26] The amount by which marriage penalties are reduced by the proposed legislation will decline over time, however, because of the AMT. If a taxpayer is on the AMT, marriage penalty relief provisions will not benefit these taxpayers. Moreover, for taxpayers subject to credit limits, a change in the regular tax will be offset by a loss in the credit, so the taxpayer will not benefit from the tax revision. And, the tax cuts in the marriage penalty legislation are likely to substantially increase the number of taxpayers on the AMT, a number that, as noted earlier, is already growing rapidly.

[27] Recently, the Treasury Department has estimated that these marriage penalty provisions would increase the number of taxpayers on the AMT by 49% by 2010, raising the total number from 17 million to 25 million. Since there were 91 million taxable returns in 1996, which would probably not grow much over 1 or 2 percent per year, 22 to 24 percent of taxpayers would then be on the AMT -- a provision that currently affects less than one percent of taxpayers. There would also be additional returns constrained by the tax credits. Thus, it is clear that the growth in the AMT coverage would be sharply increased by this legislation. The revenue collected by the AMT would also increase, by about 48 percent, from $38.2 billion per year to $46.5 billion

[28] Treasury estimates indicate that 28 percent of the marriage penalty tax cuts in H.R. 6 over the next ten years are taken back by the AMT, making the net budget effect $67 billion smaller. ⁷ This take-back rate rises rapidly and reaches 44 percent by 2008. Thus, absent revisions to the AMT, about half the tax cuts in the marriage penalty legislation will disappear after ten years. Eight years after the legislation is enacted, more than 47 percent of couples with two children will be on the AMT.

LESSONS FROM THE MARRIAGE PENALTY EXAMPLE

[29] This analysis of H.R. 6 shows how a tax proposal that affects many ordinary income taxpayers has powerful interactions with the AMT. Depending on the nature of the legislation, the interactions can be larger or smaller. Proposals that lower tax rates or narrow brackets across the board would also be expected to have significant interactions with the AMT because they tend to affect higher income taxpayers proportionally more. Taxpayers already on the AMT would get no tax cut, and some taxpayers would be shifted to the AMT. For example, even in the year 2000, 44 percent of taxpayers with incomes over $50,000 would have received less than the full ten percent tax cut in H.R. 3. ⁸ Tax cuts that add to credits or other provisions disallowed by the AMT would also interact with the AMT. Tax cuts that are directed primarily at lower or middle income individuals would be less affected by AMT interaction, at least in the near future. Left unchecked over a very long period of time, of course, virtually all taxpayers will eventually fall under the AMT provisions as the exemptions erode in value.

[30] There are offsetting effects in the Senate Finance Committee proposal for the marriage penalty (S. 2346). This proposal also makes the ability to offset credits, such as the child credit, against the AMT permanent. This change will reduce the number of middle and upper middle income taxpayers who will have their credits limited as a result of the marriage penalty or who will be switched to the AMT. The expansion of the 28 percent rate bracket, however, will expand the interaction between the marriage penalty legislation and the AMT. For tax year 2000, the top of the 28 percent rate was $105,950, while twice the top of the single bracket is $124,900. Since adjusted gross income is higher than taxable income, this change will affect many higher income individuals.

POLICY OPTIONS AND ISSUES

[31] While the previous analysis describes the importance of AMT interaction with proposed tax cuts, there are a variety of approaches that could be taken to dealing with the AMT. However, one important point to note is that cutting taxes without altering the AMT, by increasing the coverage of the AMT, makes proposals to slow or reverse its growth in importance more costly in terms of revenue loss.

THE 1997 CREDIT INTERACTION

[32] The Congress has considered the most urgent issue that of dealing with the lack of offset of the credits adopted in 1997, which immediately catapulted many middle class taxpayers into an interaction with the AMT that reduced their credits. Legislation temporarily correcting that problem has already been enacted, and it is reasonable to expect that some further extension of this correction or a permanent change is likely. This change alone could be costly. Eliminating the credit limit provision would cost about $1 billion currently, but would cost many billions of dollars by 2008 or 2009. Estimates for the Senate version of last year's tax cut bill, H.R. 2488, indicated a $1 billion cost currently for both eliminating the restriction and allowing some small additional exemption, a cost that grew to become $26 billion in the tenth year.

INDEXING THE EXEMPTIONS

[33] According to Joint Committee data, indexing AMT exemptions would cost $13.9 billion by 2008. Indexing the exemptions is the step that would be necessary to begin to keep the AMT more or less fixed in relative importance in the tax system, assuming that no other changes in the regular tax structure occurred.

THE AMT AS THE TAX OF THE FUTURE

[34] Some might see the AMT as a desirable, relatively-flat alternative tax with a wider base, and consider the expansion of the AMT desirable. If that is the case, of course, then the AMT structure itself might be examined in light of general tax principles. ⁹ The exemption levels in the AMT are not adjusted for family size or for head of household status; there are marriage penalties within the AMT structure, and tax preferences are not uniformly included or excluded. For example, while the AMT base disallows certain itemized deductions (mainly taxes) and business preferences, it leaves other important preferences intact (capital gains differentials, home mortgage interest deductions, exclusions for tax exempt interest on general obligation state and local bonds, and exclusions for employer-paid fringe benefits). And, if there is concern about the marriage penalty in the regular tax, there is also an issue about the marriage penalty in the AMT.

PHASING OUT THE AMT

[35] Others might see the AMT as an unnecessary and complicating feature of the current tax system. Under this view, adjustments to limit tax preferences should be directed at the preferences themselves and not some overall restrictions on their use as embodied in the AMT approach. These individuals might like to see not only corrections to allow the 1997 credits to be used against the AMT and indexation of the AMT exemption levels, but also steps to eventually eliminate the AMT. Some steps in this direction were already taken for the corporate AMT in 1997, where depreciation rules were brought more in line with regular depreciation.

REVISING THE AMT AND COORDINATING WITH OTHER TAX REVISIONS

[36] Others want to see the AMT continue as a general back-up mechanism to keep tax preferences from being overused, but limited to a small fraction of the population. For them, several issues arise. While the indexation for price inflation of the exemption levels is clearly appropriate to maintain the relative importance of the AMT, other questions are not as easily answered. They include questions as to whether the current base of the AMT is appropriate to its purpose, and how to adjust the AMT in tandem with regular tax changes to ensure that it fulfills its role.

[37] The preferences taken away by the AMT are selective. They include, for example, itemized deductions for state and local taxes, but not for mortgage interest, even though a case might be made that the former is not a preference, while almost everyone agrees that the latter is a preference. They do not include the major investment subsidies (capital gains preferences and tax exempt bond interest), although they include a variety of business related preferences. They include personal exemptions and standard deductions, although the rationale for this inclusion is that the AMT flat exemption is much larger than the sum of the standard deductions and personal exemptions.

[38] How the AMT should be altered as regular tax changes are made is also unclear. For example, if the principal purpose of the AMT is to limit the use of preferences, there is no apparent reason why changes in the basic structure of the tax system (wider brackets, lower rates, larger standard deductions) should trigger additional coverage under the AMT. It would be appropriate to simultaneously adjust the AMT deductions, brackets and rates to conform to the rate changes. In that case, if the standard deduction increases by $500, the AMT exemption should increase by that same amount.

[39] However, the adjustments in the AMT are limited and imperfect. For example, there is no adjustment for family size and no adjustment for head of household status. There are only two rate brackets and the width of the first bracket does not bear a close relationship to the width of the regular income brackets. The exemption is phased out at high income levels. Thus, in the marriage penalty proposal, while it might make sense to increase the AMT exemption by the increase in the standard deduction, it is not clear what, if any, conforming change the expansion of the 15 percent rate bracket should induce. Thus, it is not clear that changes of these nature should trigger conforming changes in the AMT.

[40] Also, under this view of the AMT there appears no clear reason to allow credits under the AMT although there is a reason to adjust the AMT for the expansion of the standard deduction in the marriage penalty legislation. A case might be made for an adjustment in the child credits, on the grounds that these credits are the equivalent of increasing personal exemptions and that such credits should be allowed against the AMT. There is less of a justification for other types of credits, and the case for the child credits is complex because only some taxpayers receive those credits, but the AMT exemption is uniform (distinguishing only between single and joint returns).

CONCLUSION

[41] The growth in the AMT has been considered a potential problem for some time, and its importance increases with tax cuts, such as those passed in 1997, and those, such as the marriage penalty provisions in H.R.6, that are now under consideration. Because of the AMT, not all taxpayers receive the full amount, or even any, tax cut. Moreover, every reduction in the regular tax that is not accompanied by adjustments in the AMT increases the number of taxpayers who pay the AMT and the complications for those taxpayers in filing their tax return. The marriage penalty legislation combined with the 1997 tax cuts would, in particular, cause a very large fraction of families with children to be subject to the AMT. And each time this issue is not addressed, the higher the cost grows for doing so at some future time

² The data in this and the following paragraph, as well as in Table 1, are taken from the Joint Committee on Taxation's pamphlet, JCX-39-99 Present Law And Background Relating To The Marriage Tax Penalty, Education Tax Incentives, The Alternative Minimum Tax, And Expiring Tax Provisions June 22, 1999

³ See Joint Tax Committee document JCX-55-99 for revenue estimates for the Senate version of last year's tax cut bill, H.R. 2488.

⁴ Penalties will still exist because of the AMT, the earned income tax credit, because of taxpayers in very high rate brackets and because of some other minor tax provisions. The marriage penalty discussed in this paper and addressed by legislation is the penalty relative to single returns. There is a special head-of-household return and some penalty relative to these returns will continue for families with children.

⁵ These data are from two reports of the Joint Committee on Taxation. JCX-11-00, Estimated Revenue Effects Of The Chairman's Modified Amendment In The Nature Of A Substitute To The "Marriage Tax Penalty Relief Act Of 2000," Scheduled For Markup By The Committee On Ways And Means On February 2, 2000 (10-year numbers), February 2, 2000 and JCX 8-00, "Estimated Revenue Effects Of The Chairman's Mark Of The "Marriage Tax Penalty Relief Act Of 2000," Scheduled For Markup By The Committee On Ways And Means On February 2, 2000 (10- year numbers) February 1, 2000

⁶ Data in this paragraph are based on Internal Revenue Service Statistics of Income, Individual Income Tax Returns 1996.

⁷ Al Davis. The Marriage Tax Penalty Relief Act: "Cheap" Tax Relief or Not. Tax Notes. February 28, 2000, pp. 1300-1302.

⁸ Al Davis. Candidate Bush's Tax Cut Plan. Tax Notes. January 10, 2000, pp. 271-277.

⁹ For a discussion of alternative views of the AMT discussed here and in the following sections, see Michael J. Graetz and Emil M. Sunley, Minimum Taxes and Comprehensive Tax Reform, in Uneasy Compromise: Problems of a Hybrid Income-Consumption Tax, Washington, D.C., Brookings Institute, 1988, pp.385-418. See also comments of discussants Jane G. Gravelle and Donald C. Lubick, pp. 419-429. Also, see David Weiner, Alternative Minimum Tax, In The Encyclopedia of Taxation and Tax Policy, ed. Joseph J. Cordes, Robert D. Ebel, and Jane G. Gravelle, Washington, D.C.: Urban Institute, 1999.